Technical Report Documentation Page

1. Report No. 2. Government Accession No. 3. Recipient's Catalog No.

SWUTC/99/472840-00003-4
4. Title and Subtitle 5. Report Date
COMPENDIUM: GRADUATE STUDENT PAPERS ON ADVANCED SURFACE August 1999
TRANSPORTATION SYSTEMS, 1999 6. Performing Organization Code

7. Author(s) 8. Performing Organization Report No.

Luke P. Albert, Scott P. Booth, Marcus A. Brewer, Debra A. Burdette, Randy W.
Carroll, R. Matthew Estes, Scott Harry, Kevin P. Hoppers, Renee Hulett, Jacqueline
Jenkins and Jonathan K. Lammert
9. Performing Organization Name and Address 10. Work Unit No. (TRAIS)
Texas Transportation Institute
Texas A&M University System 11. Contract or Grant No.
College Station, Texas 77843-3135 DTRS95-G0006
12. Sponsoring Agency Name and Address 13. Type of Report and Period Covered
Southwest Region University Transportation Center
Texas Transportation Institute
14. Sponsoring Agency Code
The Texas A&M University System
College Station, Texas 77843-3135
15. Supplementary Notes
Faculty & Editor: Conrad L. Dudek
Mentors: Thomas Hicks, Patrick Irwin, Wayne Kittelson, H. Douglas Robertson, William Spreitzer and Carol Zimmerman
Supported by a grant from the U.S. Department of Transportation, University Transportation Centers Program
16. Abstract

This document is the culmination of the ninth offering of an innovative transportation engineering graduate course at Texas A&M
University entitled, “Advanced Surface Transportation Systems”. The ninth offering of the course was presented during the summer
1999 term. As part of the course, a Mentors Program provides students with unique learning experiences. Six top-level transportation
professionals from private enterprise and departments of transportation, who are leaders in their field and who have extensive
experience with Intelligent Transportation Systems, were invited to Texas A&M University to present a 1½-day Symposium on
Advanced Surface Transportation Systems at the beginning of the summer term. Immediately following the Symposium, the students
enrolled in the course participated in a Forum and a Workshop with the transportation professionals and course instructor. Each student
had discussions with the transportation professionals and the course instructor to identify a topic area for a term paper. Based on
mutual interests, each student was assigned to one of the professionals who served as a mentor (along with the course instructor) for
the remainder of the summer term. Each student worked with his/her mentor and course instructor to identify a topic area and
objectives for a term paper. In addition to discussions with the course instructor, the students (communicating via telephone, fax, e-
mail, and mail) worked directly with the mentors throughout the term while preparing their term papers. The mentors returned to the
Texas A&M University campus near the end of the summer term to hear and critique the students’ presentations.

17. Key Words 18. Distribution Statement

Intelligent Transportation Systems, Advanced Traffic No Restrictions. This document is available to the public
Management Systems, Advanced Traveler Information through NTIS:
Systems, Advanced Parking Systems, Automated Enforcement, National Technical Information Service
HOV Facilities, Vehicle Siren Systems, Advanced Traffic 5285 Port Royal Road
Management Centers, Traffic Data Bases, Travel Time, Special Springfield, Virginia 22161
Events, Adverse Weather, Adverse Pavement Conditions,
Hurricane Evacuation, At-Grade Railroad Crossings, Optical
Radio, Radar, Safety, Public/Private Partnerships
19. Security Classif.(of this report) 20. Security Classif.(of this page) 21. No. of Pages 22. Price
Unclassified Unclassified 344
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized
 COMPENDIUM:
GRADUATE STUDENT PAPERS ON
ADVANCED SURFACE TRANSPORTATION SYSTEMS
AUGUST 1999

Class Instructor and mentors (front row, from left) Conrad Dudek, Carol Zimmerman, Wayne
Kittelson, Partrick Irwin; (back row) Thomas Hicks, H. Douglas Robertson, William Spreitzer
Preface Page ii

PREFACE

This document is the culmination of the ninth offering of an innovative transportation engineering graduate
course at Texas A&M University entitled, "Advanced Surface Transportation Systems," which was
presented during the 1999 summer term. As part of the course, a Mentors Program provided the students
with unique learning experiences. Six top-level transportation professionals from private enterprise and
departments of transportation, were invited to Texas A&M University to present a 1½-day Symposium on
Advanced Surface Transportation Systems at the beginning of the summer term. Immediately following the
Symposium, the students enrolled in the course participated in a Forum and a Workshop with the
transportation professionals and course instructor. Each student held numerous discussions with the
transportation professionals and course instructor to identify a topic area for a term paper. Based on mutual
interests, each student was assigned to one of the professionals who served as a mentor (along with the
course instructor) for the remainder of the summer term. Each student worked with his/her mentor and
course instructor to identify a topic area and objectives for a term paper. In addition to discussions with the
course instructor, the students (communicating via telephone, e-mail, fax and mail) worked directly with the
mentors throughout the term while preparing their term papers. The mentors returned to the Texas A&M
University campus near the end of the summer term to hear and critique the students' presentations.

One important objective of the program was to develop rapport between the students and the transportation
professionals. The opportunity for the students to communicate and interact with top transportation officials,
who are recognized for their knowledge and significant contributions both nationally and internationally, was
a key element to the students gaining the type of learning experiences intended by the instructor. Therefore,
extra care was taken to encourage interaction through the Symposium, Forum, Workshop and social events.

Thomas Hicks, Patrick Irwin, Wayne Kittelson, H. Douglas Robertson, William Spreitzer and Carol
Zimmerman devoted considerable time and energy to this program. We are extremely grateful for their
valuable contributions to the educational program at Texas A&M University.

The opportunity to bring top-level transportation professionals to the campus was made possible through
financial support provided by the "Advanced Institute" at Texas A&M University which is sponsored by the
University Transportation Centers Program of the U.S. Department of Transportation, and from funds
received from the Zachry Teaching Program from the College of Engineering at Texas A&M University.

Sherry Burr and Elizabeth Post, Senior Secretaries with the Texas Transportation Institute, coordinated the
Symposium and Workshop in a very efficient and professional manner.

Congratulations are extended to the transportation engineering graduate

students who participated in this course. Their papers are presented in
this Compendium. Appendix A and B show the transportation
professionals who graciously served as mentors in previous years and the
students who participated in the Advanced Institute Program since 1991.
A listing of all the papers that were prepared since the first offering of
the course in 1991 is shown in Appendix C of this Compendium.

Conrad L. Dudek
Professor of Civil Engineering & Associate Director, SWUTC
1999 Mentors Page iii

1999 MENTORS

THOMAS HICKS

Thomas Hicks is presently the Director of the Office of Traffic & Safety
for the Maryland State Highway Administration, one of six engineering
offices reporting to the Chief Engineer. He is responsible for
coordinating the work of six Divisions - Traffic Safety Analysis, Traffic
Engineering/Design, Traffic Operations (maintenance and operations),
Traffic Development Support (studies and research), Motor Carrier, and
the Maryland Highway Safety Office. In addition he is responsible for
the Highway Sign and Signal Shops and serves as one of 8 members
comprising the Board that guides and directs the State’s ITS program -
CHART.

Mr. Hicks has served as the State Traffic Engineer for the Maryland
State Highway Administration since 1968. Under the old State Roads
Commission, he served as Assistant Chief Engineer for Traffic Safety,
and later Assistant Chief Engineer for Traffic Engineering. In the early
1980s, as some states were cutting back on traffic engineering activities,
the Maryland SHA began a significant effort to expand theirs, and Mr. Hicks position was elevated to Deputy
Chief Engineer. In 1991, the Office of Traffic Engineering assumed control of the State’s safety grant and
Motor Carrier programs. The office then became known as the Office of Traffic & Safety. Maryland’s
traffic engineering program today is one of the most comprehensive in the country.

Prior to his Maryland assignments, Mr. Hicks was the State Traffic Engineer for the Oklahoma Department
of Highways, becoming their first in this position. He is a graduate Civil Engineer with a BSCE from the
University of Maryland and he completed the graduate school program of the Bureau of Highway Traffic
at Yale University. Mr. Hicks taught traffic engineering for three years at the University of Oklahoma.

Mr. Hicks has served in leadership rolls with several professional organizations including ITE, AASHTO,
NCUTCD, TRB, ATSSA, and the Safety Council of Maryland. He is currently a member of the AASHTO
Standing Committee on Highway Traffic Safety, Vice-Chairman of the AASHTO Committee on Traffic
Engineering, and a Member of NCUTCD where he served as Chairman, Construction & Maintenance
Technical Committee. He is a member of the Washington DC ITE; served as a member of various TRB
committees; and was a Co-Founder and first President of the Oklahoma Traffic Engineers Association. Mr.
Hicks is the recipient of the National Safety Award from ATSSA; the 1991 Highway Safety Award from
AASHTO; and the Community Transportation Award from the Washington DC Section of ITE. Mr. Hicks
is the recipient of the 1999 Theodore M. Matson Award. He is a registered Professional Engineer in
Maryland.
1999 Mentors Page iv

PATRICK L. IRWIN

Patrick Irwin was born and raised in San Antonio , Texas. He received
a Bachelor of Science degree in Civil Engineering from Texas A&M
University.

He began his career with the Texas Department of Transportation in

1974 with an assignment to the San Antonio District Traffic Section in
freeway operations. His operations study of I-410 in San Antonio
justified the first implementation in the state of turnarounds as well as
x-configuration of ramps. An assignment in Atascosa County followed
which involved the rural design of I-37. Next came an appointment to
the San Antonio Special Design Group where he was the design team
leader for several projects including the Fratt Interchange Project which
in 1979 and at $67 million was the largest Department contract. A six
and a half year assignment as Supervising Resident Engineer for Frio
and LaSalle Counties followed. In 1986, Irwin returned to the San
Antonio District as District Traffic Engineer, responsible for traffic
operations for some 11,800 lane-miles of highways in sixteen South Texas counties. In 1992, he was
appointed Director of Transportation Operations for the District. Mr. Irwin is currently responsible for the
on going development of the San Antonio Advanced Traffic Management System (TransGuide) which is
envisioned to be the most advanced system in the nation. TransGuide was the 1995 recipient of the ITS
America Board Chair Award and was selected as a United States DOT Model Deployment Initiative Site.

Mr. Irwin has addressed and participated in numerous national, state and local Transportation groups. He was
a primary developer and instructor of the Advanced Traffic Engineering course of TxDOT’s Professional
Development training program. He also has published papers in the fields of both transportation operations
and transportation management.

He is a past Director of ITS Texas. In addition, he has been a guest lecturer at universities, servers as a
resource for the USDOT’s Peer-To-Peer Program and has provided consultation to 16 foreign counties in
transportation management. Mr. Irwin also supports the Texas Transportation Federal Credit Union as a
Director, Rotary Club past resident and a Superintendent of the Walter Gerlach Junior Livestock show &
sale.

He was honored by TxDOT in 1995 with the prestigious High Flyer Award. This award was given for
innovative risks in supporting TxDOT’s mission of providing safe, effective and efficient movement of
people and goods.
1999 Mentors Page v

WAYNE K. KITTELSON

Wayne Kittelson is a Principal in the transportation consulting firm

Kittelson & Associates, Inc. Throughout his career, Mr. Kittelson has
been involved in a wide variety of projects related to transportation
planning, traffic engineering, highway design, public involvement, and
transportation research. He has also taught transportation related
courses, and has developed and applied computer programs to facilitate
various analytic procedures.

For over twenty-five years, Mr. Kittelson has been directly involved in
a wide variety of applied research activities. Whether as a project
participant, Project Manager, Panel Member, or Principal Investigator,
he has assisted in the development and application of state-of-the-art
techniques for accurate and realistic analyses of traffic operational and
safety analyses on all types of facilities. Examples of recent research-
related activities in which Mr. Kittelson has had significant professional involvement include: NCHRP 3-46
(development of an unsignalized intersection analysis method); NCHRP 3-49 (Evaluation of the Operational
and Safety Effects of Mid-Block Left-Turn Treatments); NCHRP 3-54 (Uniform Display for Protected-
Permitted Left-Turn Phasing); NCHRP 3-55(2) (Development of Planning Applications for Estimating Speed
and Level-of-Service); NCHRP 3-55(3) (development of a two-lane roadway analysis method); NCHRP 3-
55(6) (Development and Delivery of a Year 2000 Highway Capacity Manual); TCRP A-15 (development
of a Transit Capacity Manual); and an FHWA-sponsored project to develop a Roundabout Design Guide.

Mr. Kittelson also regularly serves as an instructor in numerous highway capacity analysis and intersection
design courses given throughout the United States. He has been associated with the Transportation Research
Board’s Committee on Highway Capacity and Quality of Service since 1978, when he was a major
contributor to an FHWA-sponsored research project entitled Quality of Flow on Urban Arterials. He has
continued to stay involved in Committee activities since then, serving as both its Secretary and as Chairman
of the Subcommittee on Interpretations and User Liaison. He participated directly in the development of the
1985 Highway Capacity Manual, the 1994 update, and the 1997 update.

In the area of design, Mr. Kittelson has previously served as an instructor for an FHWA-sponsored course
entitled, “Design of Urban Streets.” This course was given throughout the United States and in Puerto Rico,
and continues to serve as a basic reference on urban street design standards and guidelines. Mr. Kittelson
has presented short-courses on urban intersection design and operations issues.

Mr. Kittelson received his Bachelor of Science degree in Civil Engineering from Northwestern University
and his Master of Science degree in Civil Engineering with specialization in Traffic and Transportation from
the University of California at Berkeley. He has been a member of several professional organizations and
served as the President of the Oregon Section of ITE. He is a registered professional engineer in Oregon and
Florida.
1999 Mentors Page vi

H. DOUGLAS ROBERTSON

As a Vice President and Regional Operations Manager for SAIC’s

Transportation Consulting Group, Dr. Robertson directs the operations
of four divisions located in six area offices currently serving clients in
Virginia, Maryland, Connecticut, Pennsylvania, the District of
Columbia, Florida, Alabama, Georgia, Arizona, New Mexico, and
California. He leads business development efforts and contributes to
application projects in the areas of intelligent transportation systems,
transportation planning, traffic engineering, and highway design.
Having worked in the public and private sectors and academia, he has a
broad perspective of past, present, and future transportation needs and
services.

Dr. Robertson also currently serves as SAIC’s Senior Advisor for the
Northeast Consultants (a joint venture with PB Farradyne Systems) who
provide management support to the I-95 Corridor Coalition. The
Coalition is made up of twenty-eight transportation agencies in twelve
eastern states. He is also the Program Manager for the Development of Rural Intelligent Transportation
Systems (ITS) Project, a $10 million effort to provide technical support to U.S. DOT’s rural ITS program.

While Director of Plans and Programs at ITS America, he co-chaired a joint effort with U.S. DOT to develop
the National ITS Program Plan. In other previous positions, Dr. Robertson was a professor of civil
engineering at the University of North Carolina at Charlotte and prior to that served in both the National
Highway Traffic Safety Administration and the Federal Highway Administration of the U.S. Department of
Transportation.

He is a Fellow of the Institute of Transportation Engineers and a former Group 3 Council Chair of the
Transportation Research Board. He was the senior editor and a principal author of the Manual of
Transportation Engineering Studies for the Institute of Transportation Engineers and has published over 60
journal articles and reports.

Brigadier General Robertson serves in the U.S. Army Reserves and is the Commanding General for the 108th
Division (Institutional Training), a unit of over 3,400 soldiers with subordinate units located in the two
Carolinas, Georgia, Florida, and Puerto Rico. He is a graduate of the U.S. Army War College and the
recipient of numerous decorations and awards.

He received his Bachelor’s degree in Civil Engineering from Clemson University, his Master’s in
Transportation Engineering from the University of South Carolina and his Doctorate in Civil Engineering
from the University of Maryland. He is a registered professional engineer.
1999 Mentors Page vii

WILLIAM M. SPREITZER

Mr. Spreitzer received his B.Ae.E. in Aeronautical Engineering, from the

University of Detroit in 1952 and P.Ae.E., Professional Aeronautical
Engineering degree (Honorary) from the University of Detroit, 1957. He
is a recognized world leader in Intelligent Transportation Systems
(including AHS) and transportation research.

Mr. Spreitzer brings forty-six years of relevant experience to Intelligent

Transportation Studies including advanced automotive gas turbine
engine development; full-scale and on-the-road concept vehicle
development in applications of gas turbines, advanced transmissions and
automatic vehicle controls to automobiles (Firebird I, Firebird II and
Firebird III), buses (TurboCruiser I and TurboCruiser II), heavy trucks
(TurboTitan I and TurboTitan II) and a variety of wheeled and tracked
military vehicles; direction of research development programs in
advanced transportation systems (U.S. Department of Housing and
Urban development Study of New Systems for Urban Transportation)
and interdisciplinary studies of transportation systems of the future--public and private/personal and transit
and commercial (freight). Mr. Spreitzer retired from General Motors in January 1998 as Technical Director,
General Motors ITS Program where he was responsible for planning and coordination of General Motors ITS
Programs, corporate-wide and world-wide.

Mr. Spreitzer is active in national and international ITS efforts. He is past Chair of the ITS America
Coordinating Council (1994-97); past Chair of the ITS America Futures Group; Chair of the Society of
Automotive Engineers Technical Standards Board ITS Division and a member of the SAE ITS Program
Office and past Chair of the United States Delegation to the International Standards Organization (ISO)
Technical Committee 204, Transport Information and Control Systems (TICS).
1999 Mentors Page viii

CAROL A. ZIMMERMAN

Dr. Zimmerman has 20 years experience in marketing, education and

consulting. Her areas of expertise have included new product analysis
and definition for consumer and business markets, market management
of intelligent transportation systems, social analysis of natural resource
utilization, and quantitative research to support regional economic
development. She has been a consultant to federal and local agencies,
has served in a number of marketing roles for a global
telecommunications firm, and has taught graduate and undergraduate
students at U.S. colleges.

Dr. Zimmerman is Vice President and Director of ITS in Battelle

Memorial Institute's Transportation Sector and is Relationship Manager
for the ITS Joint Program Office at USDOT. She currently serves as
project manager for Partners In Motion, a public/private partnership for
delivering traveler information in the Washington, D. C. region. Battelle
is the systems integrator for this $12.2 million project, which began
operation on July 1, 1997. Dr. Zimmerman is responsible for all aspects of the project, which consists of
twenty-six public agencies and thirteen private sector firms working together to implement a system for
providing real-time information to travelers over a wide variety of delivery platforms.

Dr. Zimmerman also serves as program manager for Battelle's work on ITS Program Evaluation Support
(IPAS), a broad-ranging effort by USDOT's ITS Joint Program Office to assess the benefits and impacts of
ITS projects around the country, such as the Metropolitan Model Deployment Initiatives, CVISN, Federal
Operational Tests, and other projects.

Previously, Dr. Zimmerman managed end-user testing and public relations for the Atlanta Traveler
Information Showcase during the summer of 1996. She had direct responsibility for planning,
implementation, and assessment of the users of the information delivery technologies being showcased, in
addition to supervising the entire public relations program for the Showcase. Dr. Zimmerman designed and
managed a multi-million dollar market research effort for the USDOT ITS Joint Program Office to gauge
acceptance of intelligent transportation technologies among consumer, commercial, and government markets
and has been a consultant on ITS telecommunications issues at USDOT.

Prior to Battelle, Dr. Zimmerman initiated AT&T's program in intelligent transportation systems and later
managed the cross-business unit team marketing communications infrastructure products and services to state
and local transportation agencies. Dr. Zimmerman served as Assistant Professor at Rutgers University, where
she taught courses and conducted research in environmental management. At Southern Connecticut State
College and Central Connecticut State College, Dr. Zimmerman served as Lecturer in Geography, teaching
courses in physical and cultural geography, urban geography, and natural resource conservation. At Harconn
Associates, a consulting firm in Connecticut, Dr. Zimmerman led the firm's activities in computer modeling
and analysis of regional industry and labor markets.

Dr. Zimmerman received her Ph.D. from Yale University, where she focused on the sociology of resource
utilization within the School of Forestry & Environmental Studies. She is an active member of ITS America,
where she has served on the Coordinating Council and many committees. She serves on the Advisory Board
of the Center for Advanced Transportation Technology at the University of Maryland.
Contents Page ix

CONTENTS

An Evaluation of the Potential of Public/Private Partnerships for the Management

Of Archived ITS Data by Luke P. Albert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Simplifying Data Entry for Roadway and Pavement Condition Information Systems
by Scott P. Booth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Development of Plans to Use ITS Technologies to Facilitate Traffic After Special Events
by Marcus A. Brewer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

An Investigation of Advanced Parking Information Systems at Airports

by Debra A. Burdette . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Methods for the Development and Use of a Traffic Violation Data Base to Enhance
Safety Enforcement Efforts by Randy W. Carroll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

Determining Local Travel Time Thresholds for Use in Transportation Decision-Making

by R. Matthew Estes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144

ITS Applications for Severe Weather Conditions

by Scott Harry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

Opening HOV Lanes to General Traffic During Major Incidents and Severe Weather
Conditions by Kevin P. Hoppers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201

Application of ITS Technology to Hurricane Evacuation Routes

by Renee Hulett . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231

Suitability of Optical Radio and Radar Technologies as Potential Emergency Vehicle

Siren System Enhancements by Jacqueline Jenkins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264

Application of Automated Enforcement for At-Grade Railroad Crossings

by Jonathan K. Lammert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283

APPENDIX A - Transportation Professionals Who Served As Mentors . . . . . . . . . . . . . . . . . . . . 316

APPENDIX B - Advanced Institute Former Students . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320

APPENDIX C - Advanced Surface Transportation Systems Papers (1991-1999) . . . . . . . . . . . . 330

 AN EVALUATION OF THE POTENTIAL OF PUBLIC/PRIVATE
PARTNERSHIPS FOR THE MANAGEMENT OF ARCHIVED ITS DATA

by

Luke P. Albert

Professional Mentors
Patrick L. Irwin, P.E.
Texas Department of Transportation

and

Carol A. Zimmerman, Ph.D.

Battelle

Course Instructor
Conrad Dudek, Ph.D., P.E.

Prepared for
CVEN 677
Advanced Surface Transportation Systems

Department of Civil Engineering

Texas A&M University
College Station, Texas

August 1999
Luke P. Albert Page 2

EXECUTIVE SUMMARY

Transportation professionals must have reliable data for their planning and operations purposes. Every day,
vast amounts of valuable data are generated by ITS throughout the country, but much of these data are not
effectively archived for future use. It is expensive to archive ITS data into a manageable form, and most
public sector agencies that produce these data cannot afford to archive the data efficiently. Public/private
partnerships have been developed for the dissemination of real-time traveler information, and the same types
of partnerships can be developed for the management of archived ITS data.

In this research the author identified transportation systems that archive ITS data; identified potential users
of archived ITS data; determined public and private sector benefits from investing in the archival of ITS data;
evaluated existing public/private partnerships developed to access real-time traffic information; developed
potential strategies of public private partnerships for managing archived ITS data; and performed a case study
by implementing the results in San Antonio, Texas.

A review of literature was performed that focused on the following: primary ITS data-generating
technologies, agencies that archive ITS data, public/private partnership business models, and successful
public private partnerships. To supplement the literature review, the author interviewed five public sector
agencies that archive or manage ITS data and three private sector individuals that have an interest in archived
ITS data.

Using the results of the literature review and the interviews, guidelines were developed to help the public
and private sectors form partnerships in the management of archived ITS data. These guidelines included the
following:

• Create framework for the archived ITS data system.

• Select private sector partners.
• Negotiate the contract.
• Amend differences among partners.
• Create the archived ITS data management system.

These guidelines were applied to a case study in San Antonio, Texas, where a hypothetical data management
system was developed.
Luke P. Albert Page 3

TABLE OF CONTENTS

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Research Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

STUDY DESIGN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Public Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Private Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Data Analysis and Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

ARCHIVED ITS DATA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

ITS Data Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Loop Detectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Automatic Vehicle Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
ITS Data Management Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Data Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Data Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Data to Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

STATE-OF-THE-ART LITERATURE REVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Agencies That Archive ITS Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
INFORM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Houston TranStar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Illinois Traffic Systems Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Los Angeles (Caltrans District 7) Traffic Management Center . . . . . . . . . . . . . . . . . . . . . . . . . 10
Michigan Intelligent Transportation Systems Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Minneapolis, Minnesota Traffic Management Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
North Seattle Advanced Traffic Management System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Phoenix Traffic Operations Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
San Antonio Transguide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Toronto’s COMPASS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
TRANSCOM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

PUBLIC/PRIVATE PARTNERSHIPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Advanced Traveler Information Systems (ATIS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Public Sector and Private Sector Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Performance Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Public/Private Partnership Business Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Public Sector Emphasis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Contracting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Franchising . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Competitive Licensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Asset Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Joint Ownership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Cost Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Luke P. Albert Page 4

Private Sector Emphasis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Public/Private Partnership Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Partners In Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Smart Trek . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

INTERVIEW RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Public Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
ITS Data Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Data Collection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Data Archival and Aggregation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Data Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Data Access. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Public/Private Partnerships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Benefits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Private Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Market for Archived ITS Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Uses of Archived ITS Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Monitoring of System Performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Traffic Impact Assessments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Improved Travel Demand Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Prediction of Future Traffic Conditions in the Short-Term. . . . . . . . . . . . . . . . . . . . . . . . . 18
Improvement of Wireless Telephone Networks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Incident Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Public/Private Partnerships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Benefits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Example Application of Archived ITS Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

GUIDELINES FOR DEVELOPING PUBLIC/PRIVATE PARTNERSHIPS . . . . . . . . . . . . . . . . 20

CASE STUDY: SAN ANTONIO, TEXAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Luke P. Albert Page 5

INTRODUCTION

Large quantities of valuable traffic data are generated by Intelligent Transportation Systems (ITS). These
data are collected by Traffic Management Centers (TMCs) and used primarily for real-time applications,
including transit, freeway, and incident management, and traffic signal control. If these data are stored for
future use in a manageable and useful format, the ITS-generated data could also be utilized in other
applications, such as transportation planning and operations purposes. The use of these data for other
purposes results in a value-added situation. In this report, the term “ITS data” will refer to data generated
using various ITS technologies.

Loop detectors are an important method of collecting traffic volume, lane occupancy, and average speed data.
Loop detector stations are permanent, thereby allowing greater quantities of information to be collected than
with manual collection methods. Local controller units (LCUs) store and aggregate the loop detector
information in the field. Computers at the traffic management center (TMC) retrieve the information from
the LCUs. For example, the TransGuide loop detector system in San Antonio, Texas, collects continuous
volume, occupancy, and speed data, which is aggregated into twenty-second intervals, totaling approximately
150 megabytes of data each day. It is important to save these data in a way that they can be aggregated into
any desired time period and accessed with an easy-to-use computer interface (1). If information from loop
detectors is properly saved, it can be easily used by transportation professionals who need accurate
information about traffic performance characteristics (1). The characteristics include vehicle volumes, travel
times, and speed data.

In recent years, the private sector has become financially involved in the transportation system through
public-private partnerships. The public sector has an infrastructure of roadways and the private sector invests
money with the expectations of making profit from their investments. For example, the private sector invests
in toll roads and receives a percentage of the revenues. Other private sector business ventures include the
installation and sale of excess telecommunications lines and the creation and sale of traveler information
services.

Problem Statement

Transportation professionals must have reliable data for their planning and operations purposes. Vast
amounts of valuable data are generated by ITS equipment throughout the country, but much of these data are
not effectively archived for future use. It is expensive to archive ITS data into a manageable form, and most
public sector agencies that collect these data cannot afford to archive the data efficiently. Public-private
partnerships have been successful in increasing the number of travelers that access real-time traffic
information. Some similar business strategies must be developed to involve the private sector in the
archiving of ITS data. To justify private sector participation, benefits from investing in the archiving of ITS
data must be determined.

Research Objectives

The primary objective of this study was to investigate the potential for private sector investment in the
archiving of ITS-generated data. The objectives of this study were as follows:

• Identify transportation systems that archive ITS data;

• Identify potential users of archived ITS data;
• Determine private sector benefits from investing in the archival of ITS data;
• Evaluate existing public-private relationships for accessing real-time traffic information;
• Develop potential strategies of public-private partnerships for managing archived ITS data; and
• Perform a case study by implementing the results in San Antonio, Texas.
Luke P. Albert Page 6

STUDY DESIGN

In examining the potential of developing public/private partnerships for managing archived ITS data, the
author’s research consisted of three primary tasks: a literature review, data collection, and data analysis and
application. This section of the report discusses these tasks.

Literature Review

A review of current literature was conducted, focusing on the technologies used to collect ITS data, agencies
that archive ITS data, uses and benefits of ITS data, public/private partnership strategies, and successful
public/private partnerships.

Data Collection

To supplement the information from the literature review, a survey of six state agencies and three private
individuals that are involved with archived ITS data was conducted. The survey was performed through
telephone interviews and written correspondence. Separate sets of questions were developed for the public
and private sectors. The questions asked were the following:

Public Sector

1. Does your organization generate information through the use of loop detectors?
2. Does your organization generate information through other types of ITS?
3. Is the ITS information saved by your organization?
4. Who uses your organization’s archived ITS data, and what type do they use?
5. For what applications do they use the data?
6. How do the users of your organization’s data acquire it?
7. What method is there for accessing the information that your organization saves?
8. Do you see the potential for the development of public/private partnerships in managing archived ITS
data?
9. What are the benefits and problems associated with public/private partnerships for managing archived
ITS data?

Private Sector

1. Do you perceive a market for archived data generated by ITS?

2. What types of data are needed and in what form are they needed?
3. How is the data acquired?
4. For what purposes do you need the archived data?
5. Is your company currently involved in public/private partnerships?
6. Do you see the potential for the development of public/private partnerships for managing archived ITS
data?
7. What are the benefits and problems associated with public/private partnerships for managing archived
ITS data?

Data Analysis and Application

The information gathered in the literature review and interviews were used to develop guidelines for
public/private partnerships for managing archived ITS data. The guidelines are comprised of existing
strategies used in partnerships that disseminate real-time traveler information and new guidelines that apply
to archived ITS data. Finally, a case study was performed that used the guidelines in the development of a
hypothetical data management system in San Antonio, Texas.
Luke P. Albert Page 7

ARCHIVED ITS DATA

Intelligent transportation systems include many different areas of transportation, one of which is the
Advanced Traffic Management Systems (ATMS). The purpose of ATMS is to monitor the current traffic
situation and to identify any unusual occurrences in the traffic flow, such as incidents (2). These systems can
also be adapted to provide information to Advanced Traveler Information Systems (ATIS). ITS data are
primarily used in real-time for the purposes of monitoring traffic conditions. The increasing deployment of
ITS across the nation generates data that can be used for purposes other than real-time traffic monitoring.

ITS Data Technologies

The two primary technologies that produce data that are archived are loop detectors and Automatic Vehicle
Identification (AVI.).

Loop Detectors

Loop detectors consist of a wire buried in or placed just below the surface of the pavement. An electric
current is sent through the loop by attaching it to a power source, which creates an electromagnetic field.
When a vehicle passes over the loop, the energy of the system is reduced causing the inductance to decrease.
The decreased inductance causes the detector's electronic circuits to generate a pulse, which is sent to traffic
counters or signal controllers to monitor traffic (3). Loop detectors are positioned in either single-loop or
double-loop configurations. Single loop detectors acquire vehicle volume and lane occupancy information.
Double-loop detectors can also be used to acquire volume and occupancy information using the first loop,
while spot speeds are determined using both loops by calculating the difference in arrival times between
consecutive loops (1). Loop detector information is sent to local controller units (LCUs), which aggregate
the information in the field. Computers at the traffic management center (TMC) retrieve the information
from the LCUs.

An example of twenty-second loop detector data from the TransGuide loop detectors is shown in Figure 1.
These data were obtained from the website, “ftp://www.transguide.dot.state.tx.us/lanedata/.” Each line is
a twenty-second report for a specific detector, which includes the date, time, location, speed, volume, and
lane occupancy for that detector.

DATE TIME LOCATION SPEED VOLUME OCCUPANCY

06/18/99 18:24:28 L1-0090E-571.712 Speed=77 Vol=005 Occ=005

06/18/99 18:24:28 L1-0090W-571.712 Speed=78 Vol=002 Occ=002
06/18/99 18:24:28 L2-0090E-571.712 Speed=60 Vol=006 Occ=006
06/18/99 18:24:28 L2-0090W-571.712 Speed=73 Vol=004 Occ=014
06/18/99 18:24:28 L3-0090E-571.712 Speed=64 Vol=007 Occ=010
06/18/99 18:24:28 L3-0090W-571.712 Speed=67 Vol=002 Occ=002
06/18/99 18:24:28 L4-0090E-571.712 Speed=63 Vol=011 Occ=009
06/18/99 18:24:28 L4-0090W-571.712 Speed=67 Vol=003 Occ=003
06/18/99 18:24:28 L5-0090E-571.712 Speed=65 Vol=009 Occ=009
06/18/99 18:24:28 L5-0090W-571.712 Speed=64 Vol=004 Occ=004
06/18/99 18:24:28 L1-0281N-146.354 Speed=65 Vol=003 Occ=003
06/18/99 18:24:28 L2-0281N-146.354 Speed=63 Vol=007 Occ=007
06/18/99 18:24:28 L3-0281N-146.354 Speed=56 Vol=006 Occ=006
06/18/99 18:24:28 L1-0281S-146.354 Speed=45 Vol=012 Occ=013

Figure 1. Sample of Loop Detector Data

Luke P. Albert Page 8

Automatic Vehicle Identification

AVI allows traffic center operators to track probe vehicles uniquely through wireless communication (5).
An AVI system consists of four primary components:

• Probe vehicles carrying electronic transponders;

• Roadside antennas that detect transponders;
• Roadside readers that bundle data; and
• A central computer facility that collects and interprets data and calculates travel time.

In an AVI system, roadside antennas broadcast radio frequency energy over the freeway. The area that is
covered by this is called the read zone. When a vehicle is in the read zone, the transponder deflects a small
part of the energy back to the antenna. These radio waves recognize the tag’s unique ID and other data. The
captured information is sent to the roadside reader via coaxial cable, and the vehicle is assigned a unique time
and date stamp. These data are typically sent to the central computer facility via telephone line where they
are processed and stored (4,5).

AVI is most useful for direct calculation of travel times and can be as accurate as 99 percent for this use.
Since there are no personnel required for field data collection, travel times can also be calculated for long
periods of time. The major disadvantage of AVI is that the data collection is limited to the probe density,
so AVI cannot be used for volume counts. Another disadvantage of AVI is that the capital, installation, and
maintenance costs are very high. Motorists are also concerned about their personal privacy, since each AVI
transponder has a unique identification number (4).

An example of AVI data from the TranStar AVI system is shown in Figure 2. Each line indicates a vehicle
with a transponder reaching an AVI station. Each record includes the date, the station, the cumulative time
in seconds, the vehicle identification number, and the travel time in seconds between that station and the next
station.

Day Station Time ID TT

1 1 7 44082 468
1 1 8 21321 403
1 1 46 134902 397
1 1 64 6537 406
1 1 259 151320 219
1 1 291 95711 377
1 1 438 23445 545
1 1 442 23446 542
1 1 464 20760 217
1 1 494 122907 467
1 1 506 73088 388
1 1 516 1812 440
1 1 553 35864 522
1 1 582 132015 471
1 1 604 124372 513
1 1 615 147352 451
1 1 646 155844 442

Figure 2. Sample of AVI Data

Luke P. Albert Page 9

ITS Data Management Systems

Due to the large quantities of data generated by ITS equipment, data management systems must be designed,
so that they can store these data and allow users to quickly access the data that they desire. This section of
the report will examine some important issues that need to be considered when developing a data
management system.

Data Storage

Enormous quantities of data are generated by ITS. For example, the TransGuide loop detectors produce over
50 gigabytes (GB) of data per year. There are many options for storing these data, which include magnetic
disk, compact disk (CD), digital video disk (DVD), optical disk, and hard drive (6,7).

Data Access

Most databases require users to know a programming language, but it is preferable to allow users to access
data on their desktop computer without knowing any special languages. A simple, easy-to-use interface
would allow a wide variety of users to access the data. A web browser interface would be ideal because of
the familiarity of most users with the internet (7).

Data to Information

It is ideal to convert the large quantities of ITS data into easy to understand information. Transportation
planners need to analyze trends in freeway corridors or systems for long periods of time, which could mean
converting gigabytes of data into one page of useful information. This process is known as “data mining”(7).

STATE-OF-THE-ART LITERATURE REVIEW

Agencies That Archive ITS Data

There are many agencies that archive ITS data. This section contains information about specific TMC’s and
what types and quantities of data they are archiving.

INFORM

INFORM (INformation FOR Motorists) monitors 35 miles of the central corridor of the Long Island
Expressway. The Center polls the loop detectors 60 times per second. The data are aggregated into quarter-
second, 1- minute, and 15-minute intervals at the Center. The single loop detectors are located in each lane
at half-mile intervals. The 15-minute data are saved for three months to tape; then the tapes are overwritten
(6).

Houston TranStar

The Houston TranStar Traffic Management Center provides traffic management for 160 miles of the Houston
freeway system. The loop detectors cover about 30 miles of the freeway system and are spaced at half-mile
intervals. Loop data are sent to the center approximately every twenty seconds. Houston also has an
extensive AVI that has over 300,000 tags in circulation. The AVI system reads are sent to the center in real-
time as they occur. The AVI data are aggregated into 15-minute intervals and saved for future use (6,7,9,
10).
Luke P. Albert Page 10

Illinois Traffic Systems Center

The Traffic Systems Center monitors 130 miles of freeway with approximately 2,250 single loop detectors.
The detectors are located every three miles in all freeway lanes and collect volumes and lane occupancies.
Approximately every half-mile a loop detector is located in the center lane, which is used to estimate travel
time. The data are collected every 20 seconds and are aggregated to the five-minute level. Only the travel
time and lane occupancy during peak hours are saved at this level. Hourly volumes across all lanes are also
saved. The data are saved on magnetic tape (6,7).

Los Angeles (Caltrans District 7) Traffic Management Center

The Los Angeles (Caltrans District 7) TMC monitors 748 directional miles of roadway. The system consists
of over 1,000 loop stations at half-mile intervals. The stations consist of single loops that are polled every
30 seconds. The 30 second data have been permanently archived to magnetic tape since the center opened.
The TMC is in the process of developing a new system for archiving and accessing ITS. The new system
will make 13 months of ITS data available online through a relational database. Beyond the 13 month period,
the data will be saved to magnetic tape (6,7).

Michigan Intelligent Transportation Systems Center

The Michigan Intelligent Transportation Systems Center consists of two systems that combine to cover over
180 miles of freeway in the Detroit area. The old system consists of single loop detectors that cover 32 miles
and are spaced at approximately one-third-mile intervals. The new system consists of double loop detectors
that cover 150 miles spaced at two-mile intervals. The data are aggregated in the field at 20-second intervals
and sent to the center. The data are aggregated at the center into one-minute volume, occupancy, and speed
and are saved. The center keeps data on-line for one week; then the data are transferred and saved on tape
(6,7).

Minneapolis, Minnesota Traffic Management Center

The Minnesota Traffic Management Center monitors 175 miles of freeways in the twin cities metropolitan
area. The system consists of approximately 3,000 loop detectors that are in place at half-mile intervals. The
loops collect average volumes per lane across lanes at each location, lane occupancy, and vehicle length.
The average speeds are estimated from the loop detector data. Loops are polled by the Center every 30
seconds as an average across lanes. Every five minutes the data are aggregated and saved on hard disk in
a compressed format. Daily information is compressed and stored on a CD. (6,7).

North Seattle Advanced Traffic Management System

The North Seattle Advanced Traffoc Management System monitors approximately 100 miles of freeway in
the Seattle Area. Loop detectors that collect volume, speed, and lane occupancy information are placed at
half-mile intervals. These stations send data to the center every 20 seconds. The loop detector data are
stored at the five-minute aggregation level. Six months of these data are saved onto a CD and are available
to anyone who requests it (6).

Phoenix Traffic Operations Center

The Phoenix Traffic Operations Center monitors approximately 41.5 miles of freeway with single loop
detectors that are spaced at approximately one-third mile intervals. The loops collect volume, vehicle length,
speed, and lane occupancy. The TOC has saved all 20-second data since the center opened in 1995. Five-
minute summaries of the data are also created and saved (6,7).
Luke P. Albert Page 11

San Antonio Transguide

The city of San Antonio was selected as one of four cities for national deployment of the Intelligent
Transportation Infrastructure (IFI). The TransGuide system covers approximately 55 miles with loop
detectors placed at approximately half-mile intervals. The loop detectors collect volume, lane occupancy,
and speed data. The loop detector data are sent to the TransGuide computers approximately every 20
seconds. The loop detector data are saved for future use and also made available on the internet. The AVI
system in San Antonio covers 97.5 miles. The AVI system collects travel time and average speed data. The
ID numbers are scrambled so that the individual drivers cannot be identified (6-8).

Toronto’s COMPASS

COMPASS is a traffic management system that monitors parts of highway 401 in Toronto, Canada. There
are approximately 400 loop detector stations with 2,800 loops that cover 22 miles and are spaced at intervals
of half-mile or less. The stations are polled every twenty seconds and the data are aggregated into five-
minute, 15-minute, one hour, daily, and monthly time periods. Traditionally, the data were archived on 8
mm tapes, but since 1997 the system has been saving the data on CD (6,7).

TRANSCOM

TRANSCOM (the Transportation Operations Coordinating Committee) coordinates traffic management in

New York, New Jersey, and Connecticut. TRANSCOM has a project called TRANSMIT that uses vehicles
with AVI transponders (6). The purpose of these AVI stations is the collection of speed and travel time data
to be used for incident detection. The organization saves information into 15-minute time periods.

Summary

The loop detector systems that were discussed in this section are summarized in Table 1. The amount of
centerline miles that the system covers, the approximate spacing between consecutive detector stations, the
level of aggregation, and the media to which the data are aggregated are summarized in the table. The same
information is summarized for the AVI systems in Table 2.

Table 1. Summary of Traffic Management Center Loop Detector Data Archiving Practices

TMC Coverage Spacing Archival/Availability

San Antonio TransGuide 55 mi 0.5 mi 20-second data available on internet

Houston TranStar 30 mi 0.5 mi None

Minneapolis TMC 175 mi 0.5 mi 5-min data available on CD

Phoenix TOC 41.5 mi 0.33 mi 20-sec data saved. 5-min data easily accessed

Michigan ITS 32 mi (1) 0.33 mi (1) 1-min data saved to tape. One week available
150 mi (2) 2 mi (2) online.

Illinois TSC 130 mi 3 mi 1-hr data saved to tape

Los Angeles TMC 748 mi 0.5 mi 30-sec data saved to tape

North Seattle ATMS 100 mi 0.5 mi 5-min data saved to CD

Toronto Compass 22 mi 0.5 mi 5-min data saved to CD since 1997

INFORM 35 mi 0.5 mi 15-min data saved for 3 months

Luke P. Albert Page 12

Table 2. Summary of Traffic Management Center AVI Data Archiving Practices

TMC Coverage Spacing Archival/Availability

San Antonio 97.5 mi 1.0 mi - 2.0 24-hour data saved

Transguide mi
Houston TranStar 227 mi 0.9 mi - 6.7 15-min data saved
mi
TRANSCOM 1.5 mi 15-min data saved

PUBLIC/PRIVATE PARTNERSHIPS

A gap exists between the ITS technology that is available and the funds that the public sector has to spend
on ITS. The public sector has become involved in public/private partnerships to remedy this problem. A
public/private partnership is a combining of resources among entities within the public and private sectors
that achieves the goals of the individual partners by providing a service to others. In the partnership it is
important that both parties benefit: public agencies better utilize their resources, private companies increase
the market for their goods and services, and users get a more complete service than either group could
provide alone. Most existing public/private transportation partnerships involve Advanced Traveler
Information Systems (ATIS), which will be discussed in this section of the report (11-13).

Advanced Traveler Information Systems (ATIS)

ATIS allows motorists to make more informed travel decisions, such as route planning and selection of travel
mode, by providing travelers information prior to a trip or during a trip. ATIS also provides travelers real-
time information such as traffic congestion, roadway incidents, and other travel conditions. The information
provided can be expanded to an intermodal level, providing information about transit and airline departures
and arrivals (14).

There are three major functions of traveler information services: data gathering, data fusion and processing,
and data dissemination. The data gathering phase describes the receiving of information. This stage includes
data collection and transforming the collected data into relevant information. The data fusion and processing
phase includes consolidating data from many sources in the same area. Any additional value added to the
data is also performed in this step. Data dissemination is the delivery of the information to consumers (11).

Public Sector and Private Sector Differences

In forming partnerships, the public and private sectors view operations differently, so for a partnership to be
successful, it is important to try to bridge the gap between the two views. According to Hallenbeck et al.,
the public and private sectors view ATIS very differently.

Goals

In forming public/private partnerships, the primary goal of the public sector is to improve the quality of
transportation for the citizens. This includes improved safety, decreased congestion, improved economic
situation of the region, decreased fuel consumption, and improved air quality. The primary goal of the
private sector is to make a profit (15).
Luke P. Albert Page 13

Costs

The public sector is concerned with political costs, costs of using scarce resources, and the environmental
and social costs. The private sector is concerned with the capital and operating cost of the project (15).

Performance Indicators

The level of service indicators that the public sector is concerned with are measures such as level of service,
ridership, air quality, and public support. The measures that the private sector is concerned with are sales,
market share, and net profit (15).

Public/Private Partnership Business Models

There are many forms that public/private partnerships can take. This section of the report will examine some
of the public/private partnership “business models.”

Public Sector Emphasis

In the public sector emphasis model, the local or state transportation agency has full authority and
responsibility to operate the traveler information system. The data are primarily collected, processed, and
disseminated by the public sector. In this model, large quantities of information are given to the public by
the public sector, which reduces the business opportunities of the private sector. The lack of financial
interest by the private sector makes the public sector provide most of the funding, but it also allows the public
agency to have a high level of control over the operation of the system. If more than one agency is involved,
it is referred to as a public-public partnership (13-15).

Contracting

This model is a variation of the model described above. In contracting, the public agency hires a private firm
to perform a specific portion of the project. This allows the private sector to utilize personnel that are more
likely to be experts in the processing and dissemination of data and reduce the manpower needs of the public
sector (14,15).

Franchising

Franchising relies on the public sector only for data collection, which is a function that is already performed
by the Advanced Traffic Management Systems. The data are collected by the public sector, and the
government awards an exclusive distribution franchise to a single information provider after a competitive
bidding process takes place. In return, the provider is allowed to sell the data to other Independent Service
Providers (ISPs) and consumers as a way to generate revenue. The advantage of this model is that it pushes
the majority of the operation and maintenance expenses to the private sector. This model also allows the
private sector to tailor the product to the needs of the end-users. Franchising raises the question of social
equity. Since the product is purchased by consumers, people who are poor are less likely to have access to
travel information. Another disadvantage is that if the public sector uses software that remains owned by
the private sector, it is difficult to change contractors at a later date (13-15).

Competitive Licensing

Competitive licensing is an agreement in which the public sector allows two or more franchises in the same
area to obtain traffic data collected by the public agency and market it to consumers. This situation is not
as profitable for the private sector, but the competition creates better, lower cost access to traveler
information (14,15).
Luke P. Albert Page 14

Asset Management

In asset management, the public sector produces the data and disseminates the data through the services of
an outside broker. The broker sells the data to wholesalers and retailers in exchange for a commission. The
public sector maintains ownership of the data, but new markets are opened up through the expertise of the
broker (14).

Joint Ownership

In joint ownership, the public agency agrees to jointly operate a traveler information system with one or more
private sector firms. Typically, both partners are responsible for the collection of data, the private sector
merges the data, and the both groups disseminate the information (14).

Cost Sharing

Cost sharing involves the public and private sectors sharing the costs in the design, construction, operation,
and maintenance of the traveler information system. The revenue generated is split between the two groups.
This strategy is used when both groups benefit from the system (14).

Private Sector Emphasis

In the private sector emphasis model, the private company performs all the data collection, processing, and
dissemination functions. This information is collected by state-of-the-art technologies and is processed into
useful traffic reports. These reports are provided to subscribing TV and radio stations in a format that they
desire. In this model, the private sector owns and operates the entire system. This gives the private company
more incentive to produce a product that has a great potential to generate revenue. One drawback of this
model is that it may not generate a sufficient amount of revenue for future expansion of the transportation
data collection system (14).

Summary

In this section, various types of public/private partnerships were described. Table 3 summarized each
sector’s role in each type of partnership regarding the data gathering, data fusion, and data dissemination
aspects of ATIS.

Public/Private Partnership Examples

This section of the paper will examine examples of existing successful public/private partnerships.

Partners In Motion

Partners in Motion is a public/private partnership that consists of 26 public and 13 private sector entities in
the Washington D.C. area. It was officially announced in 1997 and is an information system that was
developed to help travelers make better decisions and help agencies better coordinate their efforts in
managing congestion. The partnership has a $12.2 million budget for the first three years, which is divided
into $8.3 million in public funds and $3.9 million in private funds. The purpose of the public funding is to
assist in the formation of a market for the traveler information. After three years, the traveler information
service is designed to become self-supportive and will operate totally privatized (16).
Luke P. Albert Page 15

Table 3. Public and Private Sector Roles and Various Partnership Models

Business Model Data Gathering Data Fusion Data Dissemination

Public Sector Public Public Public/Private
Emphasis
Contracting Public Private Public
Franchising Public Private Private
Competitive Licensing Public Private Private
Asset Management Public Public Private
Joint Ownership Public/Private Private Public/Private
Cost Sharing Public/Private Public/Private Public/Private
Private Sector Private Private Private
Emphasis

The public and private sectors have different risks and rewards from this project. The public sector’s risk
is that they may start a service that travelers rely on with the possibility of the loss of funding for the project.
The public sector is rewarded and by having better informed travelers and a new source of ITS funds. The
private sector risk is that they may not make a profit on their investment. The private sector is rewarded by
a profitable business venture (16).

There were some technical considerations when the partnership was in its developmental stages. Two of the
public agencies involved were in the process of developing new traffic management systems, so Partners in
Motion would have to wait for the completion of those projects before data in those areas would be available.
The second consideration was that the management of data flows is typically labor intensive, so they wanted
a design that would reduce the workload of the operations staff. The third consideration was that the system
needed to conform to the ITS National Architecture as close as possible (16).

Smart Trek

As a part of the United States Department of Transportation Model Deployment Initiative (MDI), a group
of twenty-five public agencies and private companies are working together to improve Seattle’s traffic
management system. The purpose of this program is to better utilize the Seattle area ITS infrastructure.
Existing data sources will be integrated with new data sources to establish an information network that is
integrated, regional, and multi-modal. The initial project budget was $17.8 million, with the United States
Department of Transportation providing 77 percent, the Washington Department of Transportation providing
10.5 percent, and the private sector providing 12.5 percent (17).

The primary project goal is to improve transportation for the regional traveler by providing the following
(17):

• reduce travel time by 15 percent;

• increase system efficiency;
• increase acceptance and awareness;
• increase safety by 10 percent;
Luke P. Albert Page 16

• improve traveler information distribution by 25 percent;

• decrease emissions from energy consumption; and
• demonstrate the cost-effectiveness of an overall integrated ITI.

INTERVIEW RESULTS

Five public sector and three private sector representatives were interviewed about the uses of ITS data and
the potential of forming public/private partnerships in the management of archived ITS data. The public
sector was interviewed about the types of data that they collect or manage and the current practices involving
the archival and management of the data. The private sector was interviewed regarding the potential uses
of archived ITS data. Both sectors were asked their opinions about their archived data needs and the
formation of public/private partnerships for managing the archived data.

Public Sector

All of the public sector representatives interviewed either represented an agency that collects ITS data or an
agency that manages ITS data that another agency collects. Each interviewee’s name, the agency that they
represent, and the agency that collects the data that they are reporting about are listed in Table 4. The results
of the public sector interview questions are summarized in this section of the report.

Table 4. Public Sector Interviews

Name Agency (Agency that collects data)

Steve Connell Texas Department of Transporation (TransVISION)
Mark Hallenbeck University of Washington (Washington DOT)
Pat Irwin Texas Department of Transportation (TransGuide)
Brian Smith Virginia Transportation Research Council (Virginia DOT)
Sally Wegman Texas Department of Transportation (TranStar)

ITS Data Practices

Data Collection. The data that are presently collected are primarily collected through the use of inductive
loop detectors, automatic vehicle identification, and closed caption television (CCTV). TransGuide also
collected data through the use of acoustic detectors. The inductive loop detectors are used to collect traffic
conditions data and incident data. Traffic conditions data consist of the following: lane occupancy, average
speed, and traffic volume. Incident information includes location, lanes affected, duration, and queue length.
The incidents are predicted by detecting variations from expected traffic conditions. Automatic vehicle
identification is primarily used to collect travel time information. CCTV uses video cameras to monitor
traffic and collect incident information. Acoustic detectors and loop detectors collect similar information
(18-20).

Data Archival and Aggregation. There were no consistencies in the data archival and aggregation practices
of the agencies that were represented by the interviewees. TransGuide archives all of the loop detector data
that they generate and aggregates it into 20-second increments. The Washington Department of
Transportation aggregates their loop detector data to 5-minute intervals and saves it on CD. TranStar saves
Luke P. Albert Page 17

all of their AVI data, but none of their loop detector data. The Virginia DOT does not archive any data for
long-term purposes, but they are experimenting with the possibility of archiving their loop detector and AVI
data for future use. TransVISION is still under development, and they plan to retain large quantites of
information generated by their existing and proposed detection technologies (18-22).

Data Management. There are two ways that public agencies are currently managing the data that they
archive: hire an internal data manager or have a research organization manage the data for them. The
TranStar representative cited that an internal data manager is difficult to retain at a public agency because
they cannot afford to pay a database expert a competitive salary. Since researchers are the primary users of
archived ITS data, they acquire this information from the traffic management centers in a raw form and
archive the data based on their needs (18,21).

Data Access. The agencies that archive or plan to archive ITS data for the purposes of letting others access
it use two primary methods: internet or CD. Two other ways that data are made available are electronic
spreadsheets and paper. The Washington DOT archives 5-minute data and makes it available to anybody
that wants it. TransGuide makes all of their 20-second data available on the internet to be downloaded
(18,19).

Public/Private Partnerships

Benefits. There were mixed opinions from the public sector representatives regarding their opinions about
public/private partnerships. Several benefits resulting from the formation of public/private partnerships were
cited. In the management and access of the data, the private sector would have better resources to determine
how to market the data and tailor it to the needs of the end user. Also, by having greater quantities of data
reach the public, there would be increased support of future ITS activities (18).

Problems. There were also many problems resulting from the formation of public/private partnerships that
were cited. For security reasons, the public sector is concerned about allowing information on the other side
of the firewall. There are also questions about the accuracy of the ITS data, and in order for the private
sector to market the data, some accuracy must be established. There is no evidence that adding value to ITS
data and selling it is profitable, so until a market is established, the private sector will be hesitant. Because
AVI tracks individual vehicles, it has the potential to violate the privacy of individual motorists. These
privacy concerns may delay the acceptance of ITS data by the general public. A final concern about the
formation of public/private partnerships is the varying needs of the end users. Some end users prefer to
receive raw data and aggregate it to their needs, while other end users only want hourly or daily volumes.
It may be too expensive to create a archived data system that will meet the needs of all potential end users
(19 -21).

Private Sector

The individuals from the private sector that were interviewed and the organizations that they represent are
summarized in Table 5. The results of the interview questions are summarized in this section of the report.

Market for Archived ITS Data

Since the private sector is more familiar with the marketing of products, they were asked who they thought
would be the potential users of ITS data. The following potential users were mentioned (23-25):
Luke P. Albert Page 18

Table 5. Private Sector Interviews

Name Organization
Uday Nagendran U.S. Wireless
Raymond Starsman ITS America
Robert Winick Motion Maps

• Academic;
• Transportation agencies;
• Information service providers;
• Insurance companies;
• Wireless telephone providers; and
• Private consulting companies.

Uses of Archived ITS Data

The primary users of archived ITS data have been planners and researchers from the public sector. The
applications of ITS data extend beyond the current users of the data. The most important of the other
potential users are the private sector companies that see benefits because this could spark investment in the
archival of ITS data. This section of the report examines public and private sector uses of archived ITS data.

Monitoring of System Performance. Transportation professionals from state departments of transportation

must justify the benefits that resulted from prior funding to receive funding for future transportation projects.
Loop detector data that is collected daily can be used to track trends in a transportation system. These daily
comparisons will allow public sector transportation professionals to show system improvements from prior
funding and the potential for more improvements from additional funding. (23).

Traffic Impact Assessments. It is important to be able to assess the effects that land development has had
on traffic and parking patterns. Traditionally data that are used for these types of studies are collected for
short periods of time and adjusted. With archived ITS data, the traffic conditions before and after land
development for cases all over a region can be easily acquired (23,24).

Improved Travel Demand Models. Travel demand models are periodically updated to reflect people’s
behavior changes as their surrounding conditions change. The data input into these models typically come
from traditional sources, such as interviews and responses to travel questions for the United States Census.
If AVI data were used to supplement the origin-destination inputs, large quantities of ITS data could
supplement the existing data (26).

Prediction of Future Traffic Conditions in the Short-Term. The purpose of ATIS is to inform travelers of
real-time travel conditions. An added benefit of ATIS could be short-term predictions of future traffic
conditions based on historical traffic information. With access to an archived ITS database, the Independent
Service Providers (ISPs) could provide this information (20).

Improvement of Wireless Telephone Networks. A new ITS technology tracks users of wireless telephones.
The wireless telephone companies can access the archived data to follow the travel patterns of their users.
From this information, the wireless phone companies can expand their networks to reflect their users’ travel
patterns (25).
Luke P. Albert Page 19

Incident Information. The automobile insurance industry acquires accident information. Since TMCs note
incidents that happen within their boundaries, insurance companies would probably find value in acquiring
this information from the TMCs (24).

Public/Private Partnerships

Benefits. All of the public sector representatives that were interviewed felt that the formation of
public/private partnerships for the management of archived ITS data would be beneficial. Several benefits
resulting from the formation of these partnerships were cited. It will be an expensive process to deploy a
standard archived ITS data format across the nation, so the formation of partnerships would probably make
this goal more attainable. Some ITS data resides in the private sector and some resides in the public sector,
so by forming partnerships, these data could be combined. The private sector has better knowledge of what
the market desires than the public sector, so the private sector feels that it can add beneficial qualitative value
to the quantitative public sector data (23-25).

Problems. There are also some potential problems that could slow the formation of partnerships that the
private sector representatives mentioned. It is believed that the public sector tends to delay projects more
than the private sector, so by forming partnerships, the private sector would experience longer project delays.
There appears to be a market for archived ITS data, but until somebody tries to sell it, nobody knows if it will
be profitable. Therefore, there is a risk associated with the investment in the archived ITS data (23-25).

Example Application of Archived ITS Data

The results of the interviews were used to demonstrate how the public sector and private sector can form a
partnership that benefits the public sector, private sector, and general public. This section uses TransGuide
as an example, and demonstrates how those groups of people can benefit from the formation of a
public/private partnership.

The TransGuide loop detector system collects over 150 megabytes of loop detector data daily. They are
archiving this data and making it available on the internet, but they do not have the resources to effectively
manage this data. This is valuable traffic data that the private sector could add value to and market to
potential users of the data.

Information service providers can acquire this data from TransGuide, and create a database of TransGuide
loop detector data. From this archived data, they would be able to predict short-term future traffic conditions
and have a San Antonio traffic volume database. They could make these predictions of short-term future
traffic conditions available to motorists through the internet or telephone for a fee. TransGuide could also
acquire access to the database in return for supplying the data and use it to predict incidents. In the case of
incidents, the data could be used to route the motorists to the most appropriate detour. The traffic volumes
in the database could be sold to local contractors, local transportation consulting firms, and public agencies
that have use for traffic volumes

There are several ways that the general public benefits from this partnership. First, motorists would benefit
from a savings in travel time, resulting from the improved traveler information that is provided by the
information service provider and TransGuide. The archival and management of data by the private sector
would allow TransGuide to direct more resources towards serving the public, improving the quality of
service that they provide. Finally other state agencies that would acquire traffic information from the private
sector could use these resources for other purposes.
Luke P. Albert Page 20

GUIDELINES FOR DEVELOPING PUBLIC/PRIVATE PARTNERSHIPS

This section of the report outlines the guidelines that public agencies must use when they are entering into
a partnership with the private sector with a goal of developing a system to manage archived ITS data. These
guidelines should allow the public and private sectors to create a partnership that minimizes potential
financial, organizational, and legal problems that could occur down the road. The guidelines consist of the
following five steps that are described in detail later.

Table 6. Guidelines for Developing Public/Private Partnerships

Step 1: Create framework for the archived ITS data system.

Step 2: Select private sector partners.
Step 3: Negotiate the contract.
Step 4: Amend differences between partners.
Step 5: Create the archived ITS data management system.

Step 1: Create framework for the archived ITS data system

A successful archived ITS data system must be able to store data, manage data, and allow users to easily
access specific data. A framework must be developed to determine how these features can be achieved.
Multiple public sector agencies reside over an area, so the initial partnerships will be public-public
partnerships that will utilize the resources of that area. One public agency needs to take the lead and
coordinate the efforts of all partners. From this agency, one project leader must oversee the entire project.
This leader will need to be able to communicate with the upper management of other agencies and private
sector companies. The final product of this partnership must be something that will satisfy the data needs
of the potential consumers, so these consumers must be interviewed about their needs. Knowing what data
and technologies are available, the lead agency must determine what product can be delivered to the
consumers (14,26).

Step 2: Select private sector partners

The project leaders must evaluate potential private sector partners and determine which ones are the most
compatible with the framework that was created in Step 1. The private sector partners are critical because
they are familiar with the market and are able to cater to it. The private sector participants will primarily be
interested in maximizing their profits, therefore, it is important to select partners that will make decisions
that benefit the partnership (14,26).

Step 3: Negotiate the contract

In this step, the details of the contract should be laid out. The components of the archived data management
system should be assigned to the partners, based on their strengths and interests. It is not necessary to have
formal legal contracts, but at a minimum, all of the duties need to be determined in informal written
agreements. The agreement should include details about which partners will be responsible for data storage,
management, and access. It is also important to determine how the revenue generated from the data
management system are distributed among the partners (14,26).
Luke P. Albert Page 21

Step 4: Bridge differences among partners

Throughout the contract negotiation stage there will be significant differences of opinion among all of the
partners. It is important to lessen the problems of partner-to-partner interactions as much as possible, so
each group needs to try to understand the needs and goals of the other groups. A couple of strategies that
help the communication among groups are the empowerment of key individuals and personnel exchanges.
A mediator that is trusted by all parties can be appointed to translate and negotiate the needs and objectives
of the participating organizations. This individual would work to resolve the differences among the groups.
In addition, exchanging personnel among involved agencies allows the exchanged members to gain insight
by seeing the day-to-day operations of other agencies (26).

Step 5: Create the Archived ITS Data Management System

Throughout the process of developing the data management system, the partnership should receive feedback
from the users of the product. This will allow the team to develop a product that will satisfy the needs of the
users. Finally, like any other product developed, a key to the success of the product will be how well the
product is marketed. The partnership must determine the appropriate outlets to market the product and then
market it accordingly.

CASE STUDY: SAN ANTONIO, TEXAS

San Antonio, one of four cities chosen for the Model Deployment Initiative (MDI) is the third largest
metropolitan area in Texas and is the 8th largest city in the nation. TransGuide, the ATMS program in San
Antonio, produces over 150 megabytes of ITS data daily and expressed interest in private sector investment
in the archival of their data.

Application of Guidelines for Developing Public/Private Partnerships for Managing Archived ITS
Data

This section of the report uses guidelines described earlier to implement a case study in San Antonio, Texas.

Step 1: Create framework for the archived ITS data system

San Antonio Transguide is currently involved in public/public partnerships with transportation research
agencies, public transportation agencies, the police department, and the fire department. The TransGuide
building is shared by the following public agencies; the Texas Department of Transportation (TxDOT), the
San Antonio Police Department, the San Antonio Fire Department, VIA Metropolitan Transit, Texas
Transportation Institute, and the City of San Antonio. Each of these agencies benefits from the real-time
traffic information that TransGuide provides.

Transguide has installed 590 in-vehicle navigation units (IVN) in law enforcement, fire, and public
transportation vehicles. They use Global Positioning Satellite signals to identify exactly where the vehicle
is located, which can help guide airlift helicopters or other emergency vehicles to accident sites. TransGuide
can also use these IVN units to offer alternative routes to the destination, show the route with highlighted
map segments, and notify the driver when to turn or if the car is off the route. Transguide also has two-way
video, audio, and data teleconferencing links between hospitals and EMS units. These links allow doctors
to monitor vital signs of patients by actually seeing the patients and talking to paramedics. TransGuide also
disseminates real-time traffic information on their internet website.

With access to an archived ITS database, TransGuide could use the historical data to track trends in their
transportation system and predict future traffic conditions in the short-term. By tracking trends in their
Luke P. Albert Page 22

operations, TransGuide would be able to determine how well their system is performing on a day-to-day
basis. This would make it easier for them to determine which parts of their traffic operations needs
improvement and help them make adjustments. By predicting future traffic conditions in the short-term
TransGuide would be able to predict more efficient routes on their changeable message signs, IVN units, and
webpage.

TxDOT coordinates the other agencies in the TransGuide real-time operations, so they should also serve as
the lead agency in the archived data partnership, also. This data partnership will be comprised of all the
agencies that occupy the TransGuide building because each of these agencies will receive great benefit from
using the archived data.

TTI's Translink® research program has an "ITS Data Management" focus area, in which Translink®
personnel have developed Datalink, a prototype on-line ITS data management system. The Datalink system
warehouses data from TransGuide loop detectors that are aggregated into 5-minute intervals. The 5-minute
intervals were chosen based on user data requirements and database performance. This database is a 40+
gigabyte relational database, where users can access the data using a web browser interface. In the
public/private partnership, the loop detector data will be archived in a similar manner that it is currently
being archived. A similar system will have to be set up to archive the AVI data (6).

Step 2: Select private sector partners

TxDOT will search for private sector companies that could benefit from joining the archived ITS data
management system public/private partnership. Local information service providers that could potentially
increase their business with access to archived data would have to be educated on the potential benefits.
TxDOT could market the idea to local transportation agencies.

Step 3: Negotiate the contract

In this step, the specifics of the archived ITS data management system are partitioned among the partners.
The partners located in the TransGuide building will be responsible for providing and maintaining the
computer equipment. The private sector partners will assess the market and determine the technical details
of the service. These details for the loop detector system would include modifying the Datalink system to
meet the needs of the potential users. A similar system would have to be designed by the private sector for
the AVI data. TransGuide would retain the ownership rights to the data, with the private sector having
unlimited access to it. The ownership of intellectual property rights of the algorithms and software
developed would be an issue that would depend on exactly what the private sector develops.

Step 4: Amend differences among partners

There should not be any major problems among the public sector partners, since these groups have been
working together at TransGuide for years. There may be problems with the public/private relationships
because of the general differences between the two sectors, especially the differing goals and motivations
between the two groups. Somebody in the TxDOT organization will have to step forward and coordinate
the efforts of the two sectors.

Step 5: Create the archived ITS data management system

Throughout the developmental process the private sector should be determining the needs of the users and
designing a product that caters to their needs. The data should be available in the least aggregated form the
market desires, and users should also be able to aggregate the data to the level that satisfies their needs. This
product should be marketed to transportation professionals and other potential users throughout the San
Antonio metropolitan area.
Luke P. Albert Page 23

CONCLUSIONS

Every day, there are large quantities of valuable ITS data produced throughout the United States for real-time
traffic information purposes. These data have many applications other than real-time purposes, but if the data
are not properly archived, valuable traffic information is wasted. It can be costly to archive this data, so the
permanent archival of this data may only be possible through the formation of public/private partnerships.
The public sector has the infrastructure, and the private sector has the knowledge of the market. This
combination allows the private sector to add value to the public sector data, such that it is more attractive to
consumers.

There are also many potential problems that could result from the development of archived ITS data
public/private partnerships. Since archived ITS data has never been marketed to the public, the potential for
making profit from the sale of the data is unknown. Also, the public sector is concerned about allowing too
much information to cross their firewall. The various end users of the data have varying needs for the data,
so it may not be feasible to accommodate all of these needs.

The guidelines developed in this report provide recommendations for the development of a public private
partnership to create an archived ITS data management system. These guidelines serve as guidelines for
overcoming some of the barriers that exist due to the differences in the public and private sectors. These
guidelines are flexible so that they can be applied to different geographic regions and different types of
public/private partnerships.

ACKNOWLEDGMENTS

The author would like to thank Dr. Conrad L. Dudek for affording him the opportunity to participate in this
unique program that allows students to interact with transportation professionals from across the nation. The
author would also like to think Pat Irwin and Carol Zimmerman for the guidance that they provided in
helping the author conduct this research and finish the report. A special thanks also goes out to the other
mentors that helped with this program: Wayne Kittleson, Tom Hicks, Bill Spreitzer, and Doug Robertson.
The author would also like to thank Chris Poe and Shawn Turner for helping in the development of the list
of contacts. Finally, the author would like to thank the following professionals that were contacted for taking
time to contribute to this research:

• Steve Connell, Texas Department of Transportation

• Mark Hallenbeck, University of Washington
• Brian Smith, University of Virginia
• Sally Wegman, Texas Department of Transportation
• Uday Nagendran, U. S. Wireless
• Raymond Starsman, ITS America
• Robert Winick, Motion Maps

REFERENCES

1. Turner, S. M., and C. Poe. Using ITS Data for Planning: Corridor an d System Congestion Measures.
Proceedings of the 1998 Annual Meeting of the Institute of Transportation Engineers. Toronto, 1998.

2. O’Neill W. and W. Klancher. The Use of ITS Data for Transportation Planning: Future Research
Needs. Proceedings of the 1998 Annual Meeting of the Transportation Research Board. Washington,
D.C., January, 1998
Luke P. Albert Page 24

3. Woods, D. L., B.P. Cronin, and R. A. Hamm. Speed Measurement With Inductance Loop Speed Traps,
Texas Transportation Institute Research Report 1392-8, College Station, Texas, August 1994.

4. Ford, G. L., Travel Time Estimation on Freeways Using Loop Detectors and AVI Technologies.
Compendium: Graduate Student Papers on Advanced Surface Transportation Systems. Southwest
Region University Transportation Center Report SWUTC/98/78240-00003-3, Texas Transportation
Institute, College Station, Texas, August, 1998.

5. Arendonk. J. A Comparison of Real-Time Freeway Speed Estimation Using Loop Detectors and AVI
Technologies. Compendium: Graduate Student Papers on Advanced Surface Transportation Systems.
Southwest Region University Transportation Center Report SWUTC/96/78240-00003-1, Texas
Transportation Institute, College Station, Texas, August, 1996.

6. Turner, S. M., W. L. Eisele, L. P. Albert, B. J. Gajewski, and R. J. Benz. Archived ITS Data: State-of
the-Practice and Lessons Learned From Analyses of TransGuide Data. Texas Transportation Institute
Research Report 407470-01-1F, College Station, Texas, May 1999.

7. Turner, S. M., R. E. Brydia, J. C. Liu, and W. L. Eisele. ITS Data Management System: Year One
Activities. Texas Transportation Institute Research Report 1752-2, College Station, Texas, August 1997.

8. TransGuide WWW page. www.transguide.dot.state.tx.us

9. TranStar WWW page. http://traffic.tamu.edu/central2.html

10. Turner, S. M., G. A. Carlin, and R. H. Henk. Quantifying the Benefits of High-Occupancy Vehicles
Using Automatic Vehicle Identification. Southwest Region University Transportation Center Report
SWUTC/95/465020-1, Texas Transportation Institute, College Station, Texas, November 1995.

11. Nelson, J. R., and T. Tran. Public/Private Partnerships in Traveler Information Services. Intelligent
Transportation: Realizing the Benefits. Proceedings of the 1996 Annual Meeting of ITS America.
Houston, Texas: April 15-18, 1996. Volume 2, pp. 916-927.

12. Hedley, J. Using the Private Sector in the Implementation of ITS: A Look at the Public/Private
Partnerships. Intelligent Transportation: Realizing the Benefits. Proceedings of the 1996 Annual
Meeting of ITS America. Houston, Texas: April 15-18, 1996. Volume 2, pp. 884-887.

13. Williams, E. Leaders of the Pack . ITS World, Vol. 2, No. 1, January/February 1997, pp. 21-25.

14. Jackson, D. D. An Evaluation of Public-Private Partnership Strategies Within the Traveler Information
Services Market. Compendium: Graduate Student Papers on Advanced Surface Transportation Systems.
Southwest Region University Transportation Center Report SWUTC/98/78240-00003-3, Texas
Transportation Institute, College Station, Texas, August, 1998.

15. Hallenbeck, M., S. MacLachlan, and K. Godat. Towards a Framework for Creating ATIS Business
Plans. Transportation Technology for Tomorrow. Proceedings of the 1998 Annual Meeting of ITS
America. Detroit, Michigan: May 4-7, 1998.

16. Marston, P., C. A. Zimmerman, and K. C. Miller. Partners in Motion: The Evolution of an ATIS
Public/Private Partnership in the Washington, D. C. Region. Transportation Technology for Tomorrow.
Proceedings of the 1998 Annual Meeting of ITS America. Detroit, Michigan: May 4-7, 1998.

17. Smart Trek WWW page. www.smartrek.org

Luke P. Albert Page 25

18. Written Correspondence from Pat Irwin from Texas Department of Transportation, TransGuide, June
1999.

19. Written Correspondence from Mark Hallenbeck from Washington State Transportation Center (TRAC),
June 1999.

20. Written Correspondence from Steve Connell from Texas Department of Transportation, July 1999.

21. Telephone Interview with Sally Wegman from Texas Department of Transportation, June 1999.

22. Written Correspondence from Brian Smith from University of Virginia, June 1999.

23. Telephone Interview with Robert Winick from Motion Maps, July 1999.

24. Telephone Interview with Raymond Starsman from ITS America in Washington, D. C., June 1999.

25. Written Correspondence from Uday Nagendran from U.S. Wireless, July 1999.

26. Winick, R. M. Using ITS-derived Data for Transportation Planning, Programming, and Operations.
Proceddings of the 1998 North American Travel Monitoring Exhibition and Conference. Charlotte,
North Carolina, 1998.

27. Peterson, C. N. and D. S. Olson. Overcoming Barriers to Innovative Public/Private Partnerships. ITS
Quarterly, Vol. III, No. 1, Summer 1995, pp. 22-32.

LUKE P. ALBERT

Luke P. Albert received his B.S. in Civil Engineering in May

1998 from Texas A&M University. Luke is currently pursuing
his M.S. in Civil Engineering at Texas A&M University. He is
employed as a Graduate Research Assistant in the Mobility
Analysis Program at the Texas Transportation Institute (TTI). He
also worked for TTI as an undergraduate student worker in the
Mobility Analysis Program. University activities include:
Institute of Transportation Engineers, American Society of Civil
Engineers, and Intelligent Transportation Society of America,
where he is Vice President for the Texas A&M chapter. His areas
of interest include transportation planning, traffic operations, and
intelligent transportation systems.
 SIMPLIFYING DATA ENTRY FOR ROADWAY
AND PAVEMENT CONDITION INFORMATION SYSTEMS

by

Scott P. Booth

Professional Mentor
Patrick L. Irwin, P.E.
Texas Department of Transportation

Course Instructor
Conrad L. Dudek, Ph.D., P.E.

Prepared for
CVEN 677
Advanced Surface Transportation Systems

Department of Civil Engineering

Texas A&M University
College Station, TX

August 1999
Scott P. Booth Page 27

SUMMARY

With the increasing popularity of home computers and the ease with which one can access the Internet,
motorists can look up travel information from the comfort of their own home. Many states have some
form of road condition advisory available on the Internet now. It seems like the perfect way of providing
important information for the motorists, who can access it when they want, at relatively low cost.

For a motorist to take advantage of this information, it must be current and relevant to the individual.
Anytime these criteria are not met the user loses confidence in the system. For the Texas Department of
Transportation (TxDOT), it has been determined that a lack of time coupled with a complex data entry
and retrieval system is a primary cause of this type of problem, non-current information.

A literature review was conducted to determine the current practice in developing a data entry application
of this type. The guidelines presented in this paper are summarized from the literature reviewed. Also, a
survey was conducted of other state DOTs to determine how their road condition data entry systems
operated and how the information is presented to the motorist. A survey of transportation individuals in
the San Antonio district of TxDOT was also conducted to aid in the evaluation of the current system and
to determine any specific needs they might have.

This report contains guidelines for the development of a graphically based data entry system that is
designed to save time and make road condition data entry easier. It is believed that with the aid of this
new system, those responsible for entering the information will be able to do so more frequently, reducing
the problem of non-current data. These guidelines are then used to develop a preliminary prototype that is
compared with the current system.
Scott P. Booth Page 28

TABLE OF CONTENTS

INTRODUCTION.....................................................................................................................................29
Problem Statement................................................................................................................................29
Objectives...............................................................................................................................................29
Report Organization .............................................................................................................................29

STUDY METHOD ....................................................................................................................................29

Evaluate Current System......................................................................................................................29
Literature Review..................................................................................................................................30
Data Collection ......................................................................................................................................30
Analysis and Application ......................................................................................................................30

BACKGROUND........................................................................................................................................30
Human-Computer Interface Guidelines .............................................................................................30
Consistency..........................................................................................................................................31
Simplicity.............................................................................................................................................31
User Issues ..........................................................................................................................................31
Prompts and Feedback........................................................................................................................32
Error Management ..............................................................................................................................32
Prototype Development.........................................................................................................................32
Prototyping Steps ................................................................................................................................32
Tips and Techniques............................................................................................................................32
Evaluation ...........................................................................................................................................32

SURVEY RESULTS .................................................................................................................................33

State Departments of Transportation..................................................................................................33
Virginia ...............................................................................................................................................34
Texas ...................................................................................................................................................35
Area Supervisor Survey........................................................................................................................35
Data Collection and Entry ..................................................................................................................36
Data Entry and Retrieval ....................................................................................................................37
Improvements ......................................................................................................................................38

PROTOTYPE DEVELOPMENT GUIDELINES..................................................................................38

Determine User Needs...........................................................................................................................38
Initial Prototype Design ........................................................................................................................39
Prototype Evaluation ............................................................................................................................40

SAMPLE PROTOTYPE DEVELOPMENT ..........................................................................................40

CONCLUSIONS........................................................................................................................................42

ACKNOWLEDGEMENTS......................................................................................................................42

REFERENCES ..........................................................................................................................................43

APPENDIX ................................................................................................................................................43
Sample Questionnaire ...........................................................................................................................43
Scott P. Booth Page 29

INTRODUCTION

Currently in the State of Texas, motorists can use the Internet or the telephone to contact the Texas
Department of Transportation (TxDOT) for information regarding weather conditions on roads
throughout the state. The parties that are primarily responsible for providing this information are not
always able to report it in a timely manner and as consequence the information is sometimes not accurate.
These inaccuracies can cause the public to lose confidence and no longer seek the information.

Problem Statement

It is believed that the problem lies in the current system that is used to enter the information. The system
is text based and requires a fair amount of time to enter the data (1). Oftentimes, those who are
responsible for entering the data (usually the maintenance supervisor for that area) simply do not have the
time to enter new data on a regular basis, especially when they are coping with a weather or roadway
emergency. It is believed that if the system could be simplified, it would allow the maintenance
supervisor to enter the data quickly and more frequently. This in turn would help keep the information
current and ensure motorist confidence in the system.

Objectives

There are four primary objectives to this study. These are to:

• Examine and evaluate the current system from data entry to public access;
• Determine what technology or information is available that can be used to improve the current
system;
• Using the information obtained in the previous objective, develop a set of guidelines for the
development of a prototype system; and
• Develop sample prototype based on guidelines.

Report Organization

There are six main sections to this report. This section contains an overview of the problem and details
the objectives of this study. The next section contains the methodology behind the study. Section 3
contains background information and details the results of the literature and technological review, while
the fourth section contains the survey results.

The fifth section uses the information from the previous four sections to develop a specific set of
guidelines for the development of a new data entry system. Finally, section 6 contains an initial prototype
design that is based upon the information presented throughout this report.

STUDY METHOD

Evaluate Current System

This is the first and an important step in the development of a new system. The evaluation of the current
system will provide insight to its weaknesses and strengths. The evaluation will be accomplished by
interviewing those who work with the system, the maintenance supervisor survey, and a hands-on
evaluation.
Scott P. Booth Page 30

Literature Review

The purpose of the literature review was to determine the current practice in human-computer interface
systems. The search for related information was divided into two categories, transportation and human
factors. The results of this review are presented as background information in the next section.

Data Collection

Data collection took two forms. First, a review of other state’s road condition information systems on the
Internet was performed. Individuals who participate in upkeep of these systems were contacted in an
effort to determine how data is entered into the system. The selected states consisted of Utah, Oregon,
Virginia, Maryland and Pennsylvania.

The first step in this process was to examine the road condition information Internet site maintained by
each state. Individuals involved were contacted and asked questions focused on how information was
entered into the system and who was responsible for entering it. Using this information, each system was
compared to the system used by TxDOT.

The second part of the data collection process consisted of surveying area maintenance supervisors of
TxDOT to gain greater understanding of the current system and the limitations of it. A secondary purpose
of this information was to validate the original theory that the current system is cumbersome (1). Items of
interest from the survey were:

• Who collects the data and how is it collected?

• Who enters the data into the system?
• How often is the data updated?
• How hard is it to enter the data?
• What would make the process simpler?
• Would a simpler process affect the frequency of data entry?

A copy of the survey that was administered appears in the appendix.

Analysis and Application

The analysis portion of the paper consisted of taking the information acquired from the literature review
and the surveys and developing guidelines for the prototyping of a new data entry system. Using these
guidelines, a trial prototype was developed. The trial prototype was then compared to the existing system
detailing what improvements would be made.

BACKGROUND

Human-Computer Interface Guidelines

If a new system or application is going to be used for the entry of road condition data, it should follow
certain guidelines. These guidelines have been developed over a period of many years by human-factors
and human-computer interface (HCI) researchers. Since the development of the first cathode-ray tube
(CRT), human-computer interaction techniques have been studied (2). The following section contains an
overview of some general guidelines.
Scott P. Booth Page 31

Consistency

This may be the most important guideline given. Interactive control actions should be consistent in form
and consequences from one task to another (3). Consistency in the data entry format allows users to
develop a mental model of how the system works, thereby reducing human error rates and the amount of
training required (3, 4). Consistency includes the wording of messages, placement of action buttons, even
the color scheme used.

The best way to obtain consistency is to develop a set of design standards from the start. There are
industry standards already available from IBM (1993) and Microsoft (1995) which help define many of
the needs (95 to 99 percent) of the designer. Which industry standard the designer uses should be based
on what operating system, i.e. Windows 95, is being used. Any remaining holes in the design standard
can be filled at the designer’s discretion.

If consistency and design standards are used, users of the data entry system will most likely have prior
knowledge of how to work the data entry system. Consistency between applications and tasks means the
rules of operation do not need to be repeated, nor is there a need for great detail in explaining the use of
each and every feature (4). Use of the system quickly becomes second nature.

Simplicity

Simplicity is a characteristic of good design. As a general rule, the data entry program should not be
more complex than is needed. Excessively busy displays (overall density greater than 40 percent) are
difficult to understand and consequently difficult to use (4). Colors should be used sparingly and
consistently throughout the application and follow the contrast rule, light on dark or dark on light.

The designer is encouraged to use fonts that are easy to read. Use of more than two to three different fonts
on a particular screen should be avoided, keeping in mind that a change in size, style (bold, italics,
underlining, etc.), and colors constitutes a different font. Following these rules will yield a simpler and
easier data entry system.

Associated items, such as road condition classifications, should be grouped together while keeping
unassociated items separate. Actions should occur in the areas where the user is working and should
follow a logical order. Data entry tasks should be able to be completed with a minimum number of
actions and should not require repetitive entry of information (3).

User Issues

Given the variability in users, the data entry system should support both novice and expert users. This
entails allowing users to modify some options of the system according to their needs. As novices learn to
use the data entry program, they can explore the more advanced options. The use of program functions
that are available only by clicking a certain item or area should be avoided. This is called hidden
functionality and can prevent users from gaining experience in the operation of the data entry system.

It is common practice to define a default button on every screen. The default button is invoked if the
Enter/Return key is pressed. Oftentimes, users will unintentionally hit the Enter/Return key actuating the
default function. For this reason, the default button should not be associated with a potentially destructive
function, such as the deletion of information.

Other user related issues include keeping short-term memory requirements of the user to a minimum.
This can be accomplished by using appropriate displays and making interactive sequences self-evident.
Also, the names of control functions should be semantically congruent with normal usage, i.e. Up goes
with Down, not Lower (3).
Scott P. Booth Page 32

Prompts and Feedback

The purpose of prompts is to obtain required information from the user and provide orientation of the
computer’s processes. Prompts should be clearly displayed in a consistent location, indicate what
information is expected, and remain visible until appropriate action is taken. It is desirable that users
have the ability to control the level of prompting detail.

Feedback is an important part of the data entry system. Feedback includes any message that is generated
by the computer for the user and can take the form of sounds, actions or the actual display of a message.
This may include routine messages, error messages, or user-requested messages. Also, if a requested
operation takes more than two seconds to process, the system should provide feedback in the form of a
“working” indicator or a message. For extra long processing times, feedback should be given upon the
completion of the task (3).

Error Management

Errors will occur from time to time and mechanisms for dealing with them need to be available. User
errors should be displayed as quickly as possible allowing the user to correct the problem immediately.
Where possible, options for correct input should be displayed. Also, if an error occurs during the
execution of a multi-step function, users should be able to correct the error without having to repeat the
whole process.

Prototype Development

Prototyping is an iterative analysis technique in which users are actively involved in the design of the user
interface. Through the solicitation of user input, the designer can avoid costly errors in the application.

Prototyping Steps

The first step is to determine the needs of the user. These requirements can be gathered in a number of
ways including interviews or modeling sessions. Once the uses are determined the prototype is built and
evaluated. The main goal of the evaluation step is to verify that the prototype meets the needs of the
users. Once the evaluation is complete, the designer needs to determine if the prototype is complete.
Figure 1 is a simple flow chart diagramming these four steps.

Tips and Techniques

For prototyping to be successful, it is essential to use the actual users, in this case the area supervisors and
office personnel, for the evaluation and to get them to work with it. Set a schedule to get together with
the users/testers. Use interface flow diagrams, which offer a high-level view of the relationships and
interactions between the interface objects.

Use a prototyping tool to help minimize the time needed to make adjustments to the prototype. Begin by
using a hand-drawn prototype. Move to one that shows the screens without data and then one with data.
By using this process, less time will be wasted in the early stages where changes are frequent (4).

Evaluation

The interface between humans and computers is highly complex. Given this complexity, and the fact the
user interface developers are also human, there are bound to be errors. Therefore, the evaluation of the
these interfaces is a critical part of the design process. The sooner a problem is discovered, the less it will
cost to fix it.
Scott P. Booth Page 33

Figure 1. Prototype Design Flow Chart

Evaluation of user interface systems requires the acquisition of data on the interface in question. Because
it is not possible to cover all contingencies, data are collected by taking samples (5). Once collected, the
data must be analyzed so the results can be summarized. The conclusions are then fed back into the
design loop so improvements can be made.

The methods for collecting data on user interfaces are numerous. Specific types include personal
interviews and questionnaires, the collection of protocols (such as think-aloud or video), or the program
may be designed to collect data on its own. The method used will depend on the details of the interface,
such as, how often it is used or for how long (5).

SURVEY RESULTS

State Departments of Transportation

Part of the process of this survey was to examine the road condition information that each of the six states
made available on the Internet. The primary purpose of this survey was to a point of contact in that
state’s DOT. Once the point of contact was determined, e-mail messages were sent, requesting
information on how data were entered into the system and who was responsible for its entry. Another
purpose was to compare the other states Internet information with that of Texas.

From the study of the Internet, it was determined that each of the six states offered some form of road and
pavement condition information, including information about construction areas, road closures due to
severe weather, traffic reports, and detour information. All six states offered this information through the
Internet. In all cases, information about weather related road conditions was provided on a seasonal basis.
Scott P. Booth Page 34

Virginia, Pennsylvania, Maryland and Oregon primarily used graphics, in the form of color-coded maps,
for informing motorists of road conditions. Figure 2 is a sample of the system used by Pennsylvania. The
major routes are shown on the map and are color-coded to indicate their condition. On this sample the
colors have been replaced by different shades of gray. The other three states had similar setups to
Pennsylvania.

Figure 2. Road Conditions Information Page for Pennsylvania

Virginia

From the five states surveyed, the only response came from the State of Virginia. To better understand
how the road condition information system works, the hierarchy of the Virginia Department of
Transportation (VDOT) needs to be explained. Basically, Area Headquarters report to the Residency, the
Residency reports to the District, which then reports to the Central Office. The Transportation
Emergency Operations Center (TEOC) is located at the central office and is responsible for receiving,
analyzing, and disseminating all incident data it receives (6).

All road condition information is transmitted through the Virginia Operational Information System
(VIOS) which allows many state agencies to share the information. Road condition information can be
input at any level, but usually begins with the Residency. The Residency supervisor enters condition
information that is then translated into one of the 5 color codes used by VDOT. Once the information is
entered, it immediately becomes available over the whole system (6).

From the information provided, it appears that there are three simple steps involved in entering the
information. They are, selection of the road in question, specify a section of that road, entry of a condition
for selected section. It was unclear if these steps were accomplished using keyboard commands (text) or
by pointing and clicking with a mouse (graphical).

In contrast, both the Texas and Utah systems where primarily text based. In the case of Utah’s system,
the state is divided into five regions. When a motorist enter the road condition information section, they
Scott P. Booth Page 35

are instructed to select a region. Upon selecting a region, the motor is presented with a list of roads and
their current condition for the entire region. It is then the responsibility of the motorist to find the road of
interest.

Texas

TxDOT’s system is setup very similarly to that of Utah’s. When an individual first enters the road
conditions page, a map of the state is presented with several little blue (dark gray) boxes covering it (see
Figure 3). Each blue box represents an area for which road condition information is available. To access
this information, motorists select the square nearest to the area of interest.

Figure 3. Texas Road Conditions Sample Map

After selecting an area, road conditions for the entire area are then presented in text format as is shown in
Figure 4. Note that Figure 4 is only a short listing and that in most cases there were more than a dozen
listings for a selected area. Given these multiple listings, the motorist was required to scroll through the
available information to find the roadway in question.

For Texas, the entry of road condition data begins at the area level. Each area office acquires road
condition information through a variety of sources. It is the responsibility of the area supervisors to see
that these conditions are reported to the central office in Austin, where the data are entered into the road
conditions web page and made available to the public. This process is detailed later in this report.

Area Supervisor Survey

The purpose of the survey was to gain a better understanding of the data entry system used by TxDOT
and its associated limitations. Coupled with the results of this survey, is a review from a first-hand look
by the author and an accompanying interview of a TxDOT employee. It was believed that the responses
could have bias because the maintenance supervisor’s fear of possible reprisal for criticizing the system,
so the survey data were collected in such a way to promote anonymity.
Scott P. Booth Page 36

Interstate 410 OTHER DELAY in Bexar County

N-S-bound from US 90 INTERCHANGE to US 90 INTERCHANGE
JULY 17-18, SAT.-SUN., 7:30AM-4PM, FOR THERMOPLASTIC WORK. NO LANES CLOSED, BUT
TRAFFIC WILL MOVE VERY SLOWLY.
(Texas Travel Map 15-R, San Antonio District, Jul 15, 1999 7:21AM CST)

Interstate 410 CONSTRUCTION DELAY in Bexar County

S-bound from CULEBRA RD. to US 90
ALT. LANE & RAMP CLOSURES 7-26 THRU 8-6, 10PM-5:30AM, SUN. THRU THURS. ALL LANES
CLOSED AT MIDNIGHT. USE ACCESS RD.
(Texas Travel Map 15-R, San Antonio District, Jul 31, 1999 6:57AM CST)

Interstate 410 CONSTRUCTION DETOUR DELAY in Bexar County

N-S-bound from N. OF US 87 to S. OF SOUTHCROSS
SB DETOUR TO NB. NB 1 LANE OPEN. NO LOADS OVER 9FT. WIDE. SB WIDE LOADS MUST EXIT
RIGSBY. NB WIDE LOADS EXIT SOUTHCROSS.
(Texas Travel Map 15-R, San Antonio District, Jul 31, 1999 6:57AM CST)

Figure 4. Sample Text of Road Condition Information (Texas)

There were a total of 15 responses to the survey. Most of the responses came from area supervisors, but a
couple of the responses came from individuals who work in the area offices. The first five questions were
designed to provide insight into the operation of the system. The sixth question asked individuals to rate
the system’s ease of use. The last two questions were posed in an effort to determine possible changes
that would be beneficial to the system.

Table 1 contains a summary of the results. The first column lists the question asked. The second column
indicates possible responses, with the third column listing the percentage associated with each response.
Answers to the fourth and fifth questions were not limited to one response, so percentages do not total to
100 percent.

Data Collection and Entry

Road condition data is collected by a combination of crew observation and the law enforcement, which
includes both local and Highway Patrol. The primary collector of this data is the maintenance supervisor
of the area with the crew providing additional help from time to time. The majority of those questioned
indicated that the supervisor was the responsible for overseeing data entry, but 27 percent indicated that
another party would sometimes enter data.

Comparison of the survey results and information acquired during the interview revealed what appeared
to be an inconsistency. In practice, it is very rare that the area supervisor will enter road condition data.
This responsibility typically falls to an office assistant who is at the area office throughout the day (7).
Thus, it appears that the area supervisors feel they a directly responsible for the entry of the data, but
delegate the responsibility as necessary.

Answers to the forth question were not limited to a single answer. Nearly half of those questioned
entered road condition data on a daily basis, while only 13 percent updated it every 2-4 hours. But, three-
quarters stated that they enter data based on need, which indicates that many will engage in more frequent
updating when necessary.
Scott P. Booth Page 37

Table 1. Survey Results Summary (15 responses)

Survey Question Responses Percent

Law Enforcement -
How is data relating to road conditions collected? Crew Observation -
Combination of Both 100
Maintenance Supervisor 60
Who collects the data? Maintenance Crew 33
Other 7
Maintenance Supervisor 73
Who is responsible for entering the data?
Other 27
Hourly -
Every 2-4 Hours 13
How often is the data entered/updated?
Daily 47
Need Based 73
Hourly -
Every 2-4 Hours 27
How often should the data be updated?
Daily 73
Need Based 60
Easy -
How do you rate the current system used for Fairly Easy 13
entering and transmitting the data? Fairly Hard 47
Hard 40
Graphical Interface (no text) 20
What do think would simply your data entry? Graphical Interface (with text) 13
Do Not Know 67
Yes 40
Do you believe that a simpler entry format would
No 13
promote more frequent updating?
Do Not Know 47

The fifth question asked how often the road condition data should be updated. 27 percent stated that it
should updated every 2-4 hours, while 73 percent stated daily is minimum requirement. In fact, the
system requires the information to updated daily. If it is not, it is deleted from the system (7). This helps
to ensure non-current information does not remain in the system due to oversight.

Again multiple responses to the question were allowed. The result is that 60 percent of those questioned
indicated that data should be updated whenever need dictates. It is interesting to note that this 60 percent
came from the 73 percent stating a daily updating is the minimum. Given these results, it appears that the
majority (87 percent) of those responsible for entering the data understand the importance of keeping the
data current.

Data Entry and Retrieval

Frequent entry of road condition data appears to be a desire among those responsible for it, but given the
results of the sixth question, the data entry system does not appear facilitate this. Referring again to Table
1, it can be seen that an overwhelming majority rated the system as hard or very hard. The results of the
survey appeared to be in conflict with information from the interview and first-hand evaluation.

Data entry begins by logging into the system. The operator then designates a roadway for which
information will be entered by typing its name. Next, the affected section is specified and finally, a
Scott P. Booth Page 38

condition is specified by entering a numeric code from a predetermined list. The operator is also given
the option of enter more detailed information for the affected section. Once these steps are complete, the
information is immediately made available to the entire network.

The three steps do not appear to be very complex to someone who is familiar with the system. But, for
those unfamiliar with it, it becomes a complex system in which one can get lost (8). From the survey, it
was noted that the two responses indicating the system was fairly easy to use did not come from area
supervisors, but office workers. Apparently, familiarity with the system directly affects one’s assessment
of it.

Even though the system is not complicated, data entry still takes time. The text-based entry of data into
the system can be a disadvantage at times when frequent updating is necessary. An example of this is
severe weather conditions that change frequently throughout the day (7). A graphically based system that
requires three clicks with a mouse to accomplish the same steps would greatly reduce the time required
for data entry.

The true weakness of the current system is not in the data entry portion, but the in the retrieval of the data.
As was indicated motorists can call either the TxDOT Hotline or a local maintenance office to acquire
information about road conditions throughout the state. This requires the operator of the system to scan
through multiple areas that contain lists similar to those of Figure 4. This is very time consuming and
inefficient compared to a graphical system that would have statewide conditions at a glance.

Improvements

The final two questions were designed to explore improvements to the system. Question 7 was designed
to gain feedback in what would make the system simpler. One-third of those surveyed indicated that a
graphically based data entry system (with or without text) would be simpler, but nearly two-thirds
indicated they did not know what would make the system simpler. The final question was split with
almost half saying a simpler format would increase entry frequency, while the other half was unsure of
the benefits.

Although the survey does not overwhelmingly indicate the need for improvements, individual interviews
did. Many of the problems have already been presented and possible solutions given, but more discussion
is warranted. The next section contains the development of preliminary prototype and compares the
existing system with the proposed improvements.

PROTOTYPE DEVELOPMENT GUIDELINES

Determine User Needs

The most basic needs of the user that have already been identified are:

• Data entry and retrieval system is simple to use.

• Move from the text based system to one that is graphically based.
• Program should be easy to learn.

The needs specified are generalizations for the group as a whole. To make the system even more
effective the needs of each individual should be addressed. This is accomplished by seeking input from
the end user, in this case the maintenance supervisor. Methods for obtaining these data include surveys,
interviews or hands-on observation.
Scott P. Booth Page 39

It is impractical to develop an application for each individual supervisor. But, with specific input from
them, it is possible to develop user-defined options and functions that allow for individualization.
Supervisors and office personnel should be involved from the earliest stages of the design process to
maximize the benefits of user input.

Initial Prototype Design

The first step in the development of a prototype is to create an interface-flow diagram. The purpose of
these diagrams is to provide the designer with a greater understanding of how the system is to work.
Also, by sketching out the initial design in a crude manner for review by the user, specific needs can be
assessed with a minimum of work. Figure 5 is a representation of what the initial interface flow diagram
might look for this project.

Figure 5. Sample Interface-Flow Diagram

Scott P. Booth Page 40

From the main menu, the supervisor would enter the data entry application. Here they would be
presented with a map of the state from which a specific area could be selected. Upon selection of an area,
a new map would appear displaying the roads of that area. Supervisors then select a road or section of
road to enter condition information.

Upon selection of a section of roadway, a pop-up window would open with the options to input a general
condition class for the section or to enter text information about the section. If the supervisor desires to
input text another pop-up box would appear for text input. From here, they can return to the selected
road’s information screen or go back to the area map to select another road.

Once the supervisor is finished entering the data they can transmit the data to the central area for posting
on the web. Upon transmitting, there would be a confirmation of the transmission. At this point the
supervisor would be taken back to the main menu. This is a simple representation, but it illustrates the
basic process.

Prototype Evaluation

Once the initial design is complete, testing of the prototype can begin. There are a number of ways to
evaluate the design. One such way is to have a supervisor actually enter information in the prototype and
evaluate it performance. Possible questions include:

• Was navigation through the program difficult?

• Was the information presented in a logical form and easy to understand?
• Do you believe the information entered was sufficient for the motorist to understand? Why or why
not?
• How long did it take to enter the information? Did it take less time than the current system?

There are many more questions that could be asked, but the most important thing is to make sure the users
needs are addressed. Knowing what the user thinks about the product goes a long to way to seeing if it
meets their needs. If the needs of the supervisor are not met, then the whole process begins again
incorporating the new information obtained during the evaluation.

SAMPLE PROTOTYPE DEVELOPMENT

This section contains the development of a new data entry system based on the interface-flow diagram
shown in Figure 5 and guidelines discussed in the both the background and previous section. Differences
between the proposed and current system will be highlighted and discussed. The discussion of users
needs and prototype evaluation have been presented in previous sections and will not be covered here.

Referring to Figure 5, the likely place to begin would be just after the user elects to run the data entry
program. This would bring the user to a map of the State of Texas that would be broken into sections or
districts such as those shown in Figure 6 below. The next step is for a user to select a district in which
data are to be entered.

A new screen would be presented providing greater detail of the area in question. All major roadways
through the area would be shown. An example of how the section would look can been seen in Figure 7.
In this screen the user would select a section of roadway for which information is going to be entered.

The issue of determining roadway sections needs to be addressed. In the current system, sections are
designated by entering text to describe the section affected. In the new system, major highway
intersections or towns could be used to define sections. Further study is needed to resolve the issue.
Scott P. Booth Page 41

Figure 6. Sample Map of TxDOT Districts (9)

Figure 7. Sample Area Map

Once the section of roadway is selected a new window would appear, overlaying the area map requesting
the input of information. The new window would contain the name of the road and several condition
options. The user would then select a condition by clicking on it. A sample layout of this screen can be
seen in below in Figure 8.
Scott P. Booth Page 42

Interstate 10 Additional
Normal Information

Closed
Construction

Figure 8. Sample Road Condition Input Screen

Included in this screen would be an option to enter any additional information the user may feel is
necessary to clarify conditions. By selecting the additional information button, a text-input box would
appear. Also, at any time, the user could return to any of the previous screens through commands
available on the menu bar. Once finished, the user can close out of the data entry system, which will
activate an update feature making the new information available across the system.

CONCLUSIONS

From the survey of the area maintenance supervisors it is evident that there are failings in the current
system. Time consuming data entry and complex data retrieval are preventing the system operators from
providing current information to the motorist. It appears that by improving the way road condition data is
entered into and accessed from the system many of these problems could be limited or possibly
eliminated.

The development of a new data entry system would be based upon the steps of determining the user’s
needs, development of a prototype, and evaluation of this prototype. These steps are performed in a loop
fashion incorporating the results of the evaluation into the design. Prototype development is based upon
several guidelines, such as consistency in the design and use of industry standards, simplicity of use and
design, and error management.

Clearly, there are other issues that need to be examined. One issue that could be developed further is the
use of mobile computers, both laptop and handheld. No matter what the focus is, the weak link of the
system is the human factor. If the need for human interaction can be minimized or removed, then the
system should operate more efficiently and with less error.

ACKNOWLEDGEMENTS

The author would like to thank Dr. Conrad L. Dudek for putting forth the effort and time to organize such
a great program. The author would also like to extend his personal gratitude to Patrick L. Irwin for taking
the time to help with this project. Special thanks go to each of the mentors for taking the time to share
Scott P. Booth Page 43

their experience and knowledge with the students: Thomas Hicks, Wayne K. Kittelson, H. Douglas
Robertson, William M. Spreitzer, and Carol A. Zimmerman. Lastly the author expresses his gratitude to
all the professionals that were contacted for taking the time to share their knowledge.

REFERENCES

1. Patrick L. Irwin. Texas Department of Transportation. Interview in June 1999.

2. Curricula for Human-Computer Interaction. ACM Special Interest Group on Computer-Human

Interaction. 1996. http://www.acm.org/sigchi/cdg/.

3. Human Factors Design Guide. FAA Technical Center. January 1996.

http://www.tc.faa.gov/hfbranch/htdg/index.htm.

4. Ambler, Scott W. User Interface Design Tips and Techniques. 1998.

http://www.ambysoft.com/userInterfaceDesign.html.

5. Perlman, Gary. User Interface Development. November 1989.

6. Perry C. Cogburn. Virginia Department of Transportation. E-mail response July 1999.

7. Linda Steel. Texas Department of Transportation. Interview in July 1999.

8. Neal Munn. Texas Department of Transportation. Interview in July 1999.

9. http://www.dot.state.tx.us/insdtdot/geodist/.

APPENDIX

Sample Questionnaire

1) How is data relating to weather and road conditions collected?

2) Who collects the data?

3) Who is responsible for entering the data?

4) How often is this data entered or updated?

5) How often should the data be updated?

6) How would you rate the system used for entering and transmitting the data?
(easy, fairly easily, fairly hard, hard)

7) What do you feel would simplify you data entry?

8) Do you feel that simpler entry format would allow to update the information more frequently?
Scott P. Booth Page 44

SCOTT P. BOOTH

Scott P. Booth received his B.S. in Civil and Environmental

Engineering from the University of Utah in June 1998. He is
currently working towards his M.S. in Civil Engineering at Texas
A&M University and is employed by the Texas Transportation
Institute as Graduate Research Assistant. As an undergraduate,
Scott worked for the Utah Traffic Lab as a Research Assistant on
the Roadway Weather Information Systems project. University
activities have included membership in the Institute of
Transportation Engineers, American Society of Civil Engineers,
ITS America, and Chi Epsilon. His areas of interest include
traffic operations, transportation system management, and
intelligent transportation systems.
DEVELOPMENT OF PLANS TO USE ITS TECHNOLOGIES TO FACILITATE
TRAFFIC AFTER SPECIAL EVENTS

by

Marcus A. Brewer

Professional Mentor
Thomas Hicks, P.E.
Maryland Department of Transportation

Course Instructor
Conrad L. Dudek, PhD., P.E.

Prepared for
CVEN 677
Advanced Surface Transportation Systems

Department of Civil Engineering

Texas A&M University
College Station, TX

August 1999
Marcus A. Brewer Page 46

SUMMARY

This paper contains the review of plans to facilitate exiting traffic at special events such as concerts, fairs,
and sporting events. These special events traffic plans are either in existence or in development in various
locations across the United States. This review was accomplished by the use of a survey questionnaire
completed by mail, e-mail, fax, or telephone interview. Where applicable, reports or other documentation
were used to supplement these surveys. After the survey results were compiled, those results were used to
identify priorities, tactics, and strategies used by the various government agencies and private groups in the
development of their special events traffic plans. The similarities between those priorities, tactics, and
strategies were compared and discussed. Also, where applicable, the relative successes of tactics and
strategies were compared. In situations where changes or modifications have been made to the original plan,
the success of those changes was also evaluated. Based on the success of the various strategies indicated by
survey participants, a list of recommendations was created as suggestions to aid in the development of future
plans to facilitate special events traffic. To illustrate the effectiveness of these recommendations, they were
used in a case study to devise a strategy to expedite special events traffic at the Astrodome in Houston,
Texas.
Marcus A. Brewer Page 47

TABLE OF CONTENTS

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Study Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Definition of “Dump Time” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

LITERATURE REVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Strategies and Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

SURVEY RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Anaheim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Kansas City . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Milwaukee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Orlando . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
San Antonio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Seattle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Washington/Baltimore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

SUMMARY AND DISCUSSION OF SURVEY RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

Nature of Events and Venues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Traffic Plan Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Application of ITS Technologies and Other Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Law Enforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Measures of Effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Traffic Management/Operations Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Exit Routes and Multi-Modal Concerns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

RECOMMENDATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

APPLICATION OF RECOMMENDATIONS: A CASE STUDY . . . . . . . . . . . . . . . . . . . . . . . . . . 69

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Traffic Management Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Priorities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Public Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Interagency Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Technologies and Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Measures of Effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Adaptability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

APPENDIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Marcus A. Brewer Page 48

INTRODUCTION

Background

Special events such as concerts, fairs, and sporting events draw large numbers of people every year. Because
of the large attendance at many of these events, they often have a tendency to create traffic problems for
travelers during times when these events take place. These problems affect not only those who attend such
events, but also others who may be simply traveling through the area at the time. Traffic engineers and law
enforcement officials in many cities and urban areas across the country are examining new solutions to these
traffic problems, and an increasing number are using ITS program applications with existing infrastructure
to manage special events traffic. ITS program applications can be used in several ways: on a small scale to
provide a fine-tuning improvement to current conditions in specific locations; on a larger scale as part of a
plan to prepare for projected growth and congestion in a community; or on a massive scale as a separate
project to manage a freeway corridor and the additional traffic generated by the surrounding area events.

Problem Statement

While there is often a great attempt to provide the means for traffic to arrive before events, there is not
always as much effort taken to facilitate the movement of traffic after the events are over. As a result, post-
event traffic can generate significant congestion on the transportation system surrounding the event venue.
This increases the delay and frustration experienced by motorists in the area. If weather or an incident affects
the venue area, the problem is compounded. However, the vast majority of special events are scheduled well
in advance and expected traffic related to those events is somewhat predictable. If this is taken into
consideration and used in a positive way, proper planning can significantly reduce congestion and increase
the efficiency of the flow of exiting traffic.

The purpose of this paper is to provide recommendations to aid cities, counties, and other entities in the
development of strategic plans to facilitate the movement of traffic after special events. All of the special
events under consideration are based at fixed locations, such as stadiums, multipurpose arenas, or
fairgrounds.

Objectives

There are six objectives in this paper. Each objective builds on the efforts of the previous objectives, with
the first four objectives consisting primarily of collection of information. The fifth and sixth objectives use
that information to produce a final product.

1. Examine existing plans used to facilitate exiting special events traffic.

2. Determine priorities used in the creation of the existing plans in Objective 1.
3. Identify ITS solutions currently used to satisfy the priorities in Objective 2 in order to gain an indication
of possible trends and successful applications.
4. Where possible, examine results of existing plans, as well as modifications and improvements made to
those plans, to identify effective solutions and changes.
5. Based on the priorities, solutions, and results from Objectives 2, 3, and 4, provide recommendations for
the development of future special events traffic plans.
6. Apply the recommendations from Objective 5 in a case study of a location with significant and varying
special events traffic.
Marcus A. Brewer Page 49

Scope

The scope of this paper consists of plans and strategies used by various government agencies and private
groups to expedite the flow of traffic from special events. Those plans and strategies include the examination
of the nature of special events, priorities and corresponding solutions in facilitating traffic, enforcement
issues, use of Traffic Management Centers, primary exit routes and modes, effectiveness of existing plans
and strategies, and length of time that the plans have been implemented.

Study Methodology

1. Develop Survey Questionnaire

The primary source of information for this paper is a variety of government agencies and private
organizations that have influence on the development of plans in their city or area to expedite the flow of
traffic after special events at a specific venue. In order to obtain the same types of information from each
agency, a standard list of questions was necessary. This list of questions is provided in the appendix.

2. Contact Agencies to Complete Survey and Provide Additional Information/Prepared Reports

Personnel from a variety of agencies and private organizations involved in developing special events traffic
plans were contacted by phone and asked if they would be willing to complete the survey and, if possible,
provide supplemental documentation. After agreeing to participate, these contacts were provided with a copy
of the survey questions by fax or e-mail, and a deadline was established to complete the survey, either by
phone, by fax, by mail, or by e-mail.

3. Compile and Analyze Results

After receiving the completed surveys and any additional information, the results were compiled for
comparison and analysis. Comparison involved identifying similarities among the various agencies and their
plans, particularly among the problems and corresponding solutions specified by the agencies. Analysis
involved evaluating the effectiveness of the various plans and solutions, and any changes made to increase
effectiveness. This comparison and analysis allowed identification of trends and patterns both in problems
and in solutions.

4. Develop Recommendations

After identifying trends and patterns in the analysis of results, a list of recommendations was made for use
in the development of a new or revised special events exit traffic plan. These recommendations used the
information gained from the efforts of the agencies surveyed, and they helped to identify possible priorities
in facilitating traffic as well as potential problems and corresponding solutions that will determine whether
those priorities are satisfied.

5. Perform a Case Study

The recommendations developed from the survey results were applied to the development of a plan to
facilitate exiting traffic at special events at the Astrodome in Houston, Texas. A set of priorities was
established based on the trends identified in analysis, some potential problems were identified pertaining to
the specific characteristics of the events and the city, and solutions to those problems were proposed.
Marcus A. Brewer Page 50

Definition of “Dump Time”

When describing the efficiency or effectiveness of an exiting traffic plan, the term “dump time” is often used.
“Dump time” is also used frequently in this paper as a measure of effectiveness; however, the specific
definition of this term varies depending on the city, agency, or event under consideration. A general
definition of “dump time” could be considered to be: “the amount of time needed to empty a venue’s parking
area to the point that the remaining vehicles are still in the parking area by the choice of the drivers.” Implied
in this definition is that traffic on arterials and freeways used as exit routes is free-flowing, or operating
above level of service F.

LITERATURE REVIEW

Contained within this section is a brief discussion of literature related to special events traffic and traffic
management. This discussion is intended to provide a foundation for the information and recommendations
provided in the following sections of the paper.

Introduction

Transportation, in general, involves the movement of persons, goods, and services from place to place. An
ever-growing proportion of transportation takes place in urban areas, specifically on urban freeways.
However, the flow of traffic on urban freeways can be adversely affected by incidents such as traffic
accidents, maintenance operations, and temporary blockages due to large traffic volumes. Many studies
document that incidents can have serious impacts on traffic, manifested in terms of congestion or delay. It
is recognized that each minute of blockage during commuter periods on the freeway results in five minutes
of motorist delay. An objective of each urban transportation agency is to respond to disruptive incidents
within its respective jurisdiction and relieve congestion as quickly and safely as possible (1).

Incidents related to special events are somewhat unique in nature, in that they are not high-frequency
recurrent incidents similar to peak-period congestion, which occurs daily. However, event-related incidents
are still somewhat predictable, in contrast to random incidents such as traffic accidents or effects of adverse
weather. In general, a significant portion of event-related congestion occurs on certain freeway segments
at or near the event venue. In many cases, there are alternate routes for both event and non-event traffic, but
often they are not fully utilized because drivers are either unaware of them or they have no knowledge of the
severity of congestion in the venue area. Although the effects of many special events can be predicted from
historical data by traffic planners and are expected by motorists who regularly attend these events, the
congestion that develops is often unexpected by non-event motorists in the area (2).

Strategies and Solutions

Operational studies have shown that managing traffic after special events will result in extremely high
payoffs in terms of reduced congestion and delay. Strategic plans to facilitate the movement of post-event
traffic can be very beneficial both to event-related traffic and non-event traffic. The determination of specific
event venue exit routes and the supply of relevant information to all motorists in the area can produce a
variety of benefits. Some of these benefits are: advance decisions on route selection and time of travel;
diversion of non-event traffic to alternate routes; improved merging of event traffic on area freeways; and,
ultimately, a reduction in congestion.

Incident management requires coordinated and preplanned procedures to restore traffic to normal operation
through the use of all available human and electronic/mechanical resources. In general, these procedures
include: 1) detecting the incident; 2) identifying the incident; 3) identifying the response requirements; and
4) providing the appropriate aid to the motorists involved to minimize the effects on the transportation system
Marcus A. Brewer Page 51

by clearing the incident area as quickly as possible. The benefit to event-related incident management is that,
to a certain degree, the first three procedures can be completed in advance. Using this fact in developing an
exit traffic plan allows the implementation of the plan to be focused on actually providing aid to clear the
event area quickly (3).

Of fundamental importance in any traffic management plan is multi-agency communication. Response needs
for most major freeway incidents extend beyond the capability of a single agency, and special events are no
exception. Different agencies have different responsibilities and objectives when responding to event-related
traffic incidents. These responsibilities have different priorities in the overall traffic plan; however, decisions
and actions by any one agency must be compatible with those of the other agencies involved. Part of proper
planning for post-event traffic is a means of establishing the chain of command, flow of relevant information
to aid motorists, specific responsibilities of each agency, and resources that will be provided by each agency.
As a result, it is necessary to establish multi-agency consensus for needs and priorities, establish interagency
cooperation agreements, and develop interagency communication protocols. After the basic structure for
interagency cooperation and coordination is established, the involved agencies can improve their
preparedness for post-event traffic (1).

There are a variety of aspects that need to be considered when implementing technologies and solutions to
facilitate post-event traffic. One aspect is traveler information, which is vital for the reasons stated earlier;
that is, providing drivers with accurate information that their intended route may be blocked or severely
congested can help eliminate a proportion of excess travel. There are several methods commonly in use for
this purpose; some of them are: traveler advisory radio, specialized maps, temporary signing, and kiosks.
Traveler advisory radio can be used to warn drivers within a localized area of short-term delays caused by
preplanned activities such as special events. Drivers who receive these messages can make routing decisions
before entering the affected area, thus avoiding post-event congestion altogether. Specialized maps can be
used to inform event attendees of preferred exit routes and their designated directions of travel; these maps
can be distributed before or during events to allow drivers to decide on their exit route before entering the
post-event traffic stream. Variable message signs and temporary static signs can be placed to provide route-
and location-specific messages regarding changes in traffic conditions. These signs can be used to indicate
downstream congestion, lane closures, channelization, or suggested detour routes. Informational kiosks can
be placed inside event venues to display updated area traffic conditions and information on preferred exit
routes; this will further enable event attendees to make informed decisions about their travel away from the
event venue.

Signal and traffic control systems are also important in post-event traffic. Because of the large volumes of
traffic generated by special events, signal systems on adjacent intersections often cannot function adequately.
As a result, changes must be made to expedite the flow of traffic. One solution is real-time signal system
adjustment, which allows an operator to make changes in timings and phases based on the current level of
demand. This ability can be used to provide more green time to the heaviest traffic volumes, while reducing
green time for lesser movements, thus improving the efficiency of the signal. Another method of traffic
control is physical channelization, which involves the use of devices such as barrels, cones, and barricades
to physically block access to a particular street or lane. This method is often supplemented by the use of
traffic police or other law enforcement to help direct the flow of traffic and prevent violation of control
devices.

A third major aspect of post-event traffic is freeway and corridor control; it is necessary to make sure that
all of the other strategies and solutions are working properly in order to maintain traffic flow. A popular
method to accomplish this is the use of a traffic management center. Traffic management centers are vital
requirements for integrating advanced technologies to improve conditions on urban freeways. These centers
provide capabilities of monitoring traffic flows, providing information to motorists through variable message
signs and advisory radio, and communicating with personnel from various agencies participating in the traffic
Marcus A. Brewer Page 52

plan. These capabilities are all used to optimize the use of corridor capacity by altering the flow of non-event
traffic to better accommodate the large volumes of event-related traffic (4).

Implementation

A number of the strategies and solutions discussed above either have been or are being implemented in
various locations around the country. The remainder of this section contains examples of these strategies
and solutions in selected cities.

Perhaps no annual event in the United States attracts as much attention, and as many people, as the Super
Bowl. As part of preparations for Super Bowl XXXII at Qualcomm Stadium in January 1998, officials in
San Diego, California, worked to implement the Mission Valley ATIS-ATMS (Advanced Traveler
Information System-Advanced Traffic Management System) program. Specifically, the first phase of this
program dealt with improvements in and around Qualcomm Stadium, and included elements for corridor
control, traveler information, and a transit alternative. Closed-circuit television (CCTV) cameras were
installed to be aimed at the streets outside the stadium, thus enabling surveillance of area traffic conditions.
Variable message signs were installed to guide drivers into various lane configurations and parking lots. The
new Mission Valley Trolley Line was built to encourage attendees to park elsewhere and ride to Qualcomm
Stadium, where the largest station on the trolley line is located. Phase two of the plan is designed to integrate
these special-events features with the California Department of Transportation area traffic management
system. Although the official evaluation of the program’s performance was unavailable, project officials
were optimistic about the results (5).

Another Super Bowl-related project involved the use of a video detection system in Tempe, Arizona, outside
Sun Devil Stadium, site of Super Bowl XXX in 1996. Officials there decided to use this system for managing
the surge of event traffic in part because it would avoid adding to traffic problems by tearing up streets to
install loop detectors. Traffic managers used the surveillance data to adjust signal timings at the primary
intersection just outside the stadium, through which 40,000 people crossed daily during Super Bowl week (5).

The Rose Garden arena in Portland, Oregon is a relatively new facility that attracts thousands of basketball
fans to Trail Blazers games and hosts other events such as musical concerts. However, the arena was built
on a site in an already congested area. To help manage the congestion, city officials are utilizing a real-time
traveler information system with significant use of variable message signs. There are at least two such signs
directing drivers to parking lots or shuttle sites. It is the hope of city transportation officials that the new
information system will be successful enough to encourage its use in other areas of the city.

As the use of these solutions becomes more common, more definitive results will become available concerning
their effectiveness. However, officials in each of the cities mentioned above were encouraged by the
possibilities of increased efficiency and optimized use of the transportation system in their areas (5).

SURVEY RESULTS

Introduction

This section contains the findings from the survey questionnaire, a copy of which is provided in the appendix.
A total of 20 persons, representing 17 agencies in 11 cities, were contacted to participate in the survey. Of
those contacted, 11 persons from eight cities agreed to participate, and 10 persons from seven cities actually
completed and returned the survey. Table 1 provides a summary of the cities and regions represented in the
survey, as well as the agencies that provided the information for the survey.
Marcus A. Brewer Page 53

The remainder of this section contains summaries of the information obtained about each city, based on
answers to the survey questions and any supplemental documentation provided by the agencies surveyed.

Table 1. Listing of Cities and Agencies Represented in the Survey

City/Region Agency Nature of Agency

Anaheim, CA Anaheim Dept. of Traffic Engineering Public
Kansas City, KS/MO TranSystems Corporation Private Consultant
Milwaukee, WI WI DOT – MONITOR Public
TransCore Milwaukee Private Consultant
Orlando, FL City of Orlando Public Works Public
San Antonio, TX VIA Metro Transit Public
City of San Antonio Public Works Public
Seattle, WA WA DOT Public
Washington, DC / MD State Highway Administration Public
Baltimore, MD MD State Police Public

Anaheim

The city of Anaheim has a permanent population of about 300,000; however, the city also experiences an
estimated 20-30 million visitors each year. Special events are frequent in Anaheim, and there are four main
venues that will be discussed here:

• Edison Field – Home of the Major League Baseball (MLB) Anaheim Angels, with 81 regular-season
home games per year and a capacity of 45,050. Edison Field is also the site of various Monster Truck
events, soccer games, and large concerts.
• Arrowhead Pond of Anaheim – Home of the National Hockey League (NHL) Mighty Ducks of Anaheim,
hosting about 40 games each season with a capacity of 17,174. The Pond also hosts other events
including NBA exhibition games and boxing matches.
• Anaheim Convention Center – Hosts a wide variety of conferences and conventions each year; exact
attendance figures were unspecified.
• Disneyland – World-famous theme park with millions of visitors annually. Disneyland also contains a
resort area with hotels, dining, and entertainment.

The city of Anaheim has had some form of a traffic plan in place for well over a decade. In 1989, the city’s
first Traffic Management Center was established; this center operates on a fixed schedule under normal
conditions, but it also has an event-based schedule that allows officials to concentrate on conditions related
to event traffic. Barring technical difficulties or construction problems, the Traffic Management Center is
the central point of operation for all event traffic plans.

Anaheim’s priorities in establishing the current version of their event traffic plan were based on a Traffic
Management Center-operated system. The city is divided into sections, each with its own individual traffic
system; this sectional strategy allows the city to deal with each event separately. Each plan then becomes a
modular plan; if a section is affected, a plan can be introduced to deal with problems in that localized area.
Marcus A. Brewer Page 54

The City of Anaheim uses a wide variety of technologies and other solutions to address issues related to
exiting special events traffic; some of them are listed below:

• Changeable message signs on arterials and (in conjunction with CalTrans) on freeways
• Dynamic traffic signal adjustments
• Highway advisory radio for arena and stadium events and the 91 Freeway corridor
• A fiber optic backbone for the transfer of information
• Video detection
• Traffic Police, employed primarily for directing and calming traffic and coordinating movement of
pedestrians; law enforcement is generally confined to illegal parking
• Kiosk system, with real-time updates on freeway conditions and other information

City officials hold weekly meetings with event managers to evaluate current performance and investigate
opportunities for improvement. Some specific measures of success or effectiveness include “dump time”, cost
of resources used, and complaints about construction or traffic flow. Of those three measures, cost is
considered to be the least important; Anaheim is willing to spend some extra money to improve the flow of
traffic. Officials are more concerned with the amount of time it takes to restore normal traffic and with the
public’s perception of their job performance. To this point, Anaheim has had great success with their traffic
strategies, achieving better efficiency with more traffic.

Each of the venues listed above was designed to be in close proximity to Interstate highways. Thus they are
all accessible to the large number of passenger cars and other vehicles that travel from the events. Estimated
proportions of traffic leaving stadium and arena events are as follows: 90 percent passenger cars, 8-10 percent
shuttle buses, and the remainder on tour buses. Traffic from the Convention Center is estimated to be a 70/30
split between passenger cars and shuttle buses, and Disneyland traffic is approximately a 50/50 split.

Transit options are available for all events at these venues, and they are highly encouraged. In addition to
public transit, many hotels run shuttles to and from Disneyland, sporting events, and/or the Convention
Center. There are also private shuttles and tour buses commonly available.

There are very large numbers of pedestrians at many of these venues, particularly at Disneyland. In fact, there
are far more pedestrians at Disneyland than any other of these venues, but interaction between pedestrians and
vehicles is much higher at sporting events due to the configuration of parking areas. Despite these high
pedestrian volumes, Anaheim officials have not indicated significant problems with pedestrian accidents or
injuries following special events (6).

Kansas City

The Kansas City metropolitan area straddles the Missouri River, which acts as the border between the states
of Kansas and Missouri in the northern part of the metropolitan area. This unique geographical characteristic
creates some interesting situations involving traffic and the solutions available to deal with it. There are a
number of facilities on both sides of the river that generate significant special event traffic:

• Arrowhead Stadium – Home of the National Football League (NFL) Kansas City Chiefs with eight
regular-season and two exhibition home games each year, Arrowhead is also the host of the Major League
Soccer (MLS) Kansas City Wizards with 16 home games per year. The facility seats about 79,100 for
football and about 20,500 for soccer, and it has a variety of seating configurations for large concerts.
• Kauffman Stadium – Home of the Major League Baseball (MLB) Kansas City Royals, Kauffman Stadium
hosts 81 home games annually and seats approximately 42,000. Originally named Royals Stadium,
Kauffman joins Arrowhead as the major components of the Harry S. Truman Sports Complex, which was
opened in 1972. The Sports Complex is located in eastern Kansas City, Missouri.
Marcus A. Brewer Page 55

• Kemper Arena – The Kansas City area’s major multipurpose arena, Kemper Arena is home to the Kansas
City Blades hockey team, the World Team Tennis Kansas City Explorers, and the Kansas City Attack
indoor soccer team. Kemper also hosts the Big XII Conference men’s basketball post-season tournament,
other sporting exhibitions, and a number of concerts each year. The facility has a capacity of about
17,000, and is located just east of the Missouri River in Kansas City, Missouri.
• Sandstone Amphitheatre – Located in Bonner Springs, Kansas, Sandstone Amphitheatre hosts numerous
concerts and musical events each year, including nearly every weekend during the summer months.
Sandstone has an estimated capacity of 3,000.
• Kansas International Speedway – Currently in early stages of construction, the Speedway is located on
the western edge of Kansas City, Kansas. The Speedway design consists of a 1.5 mile oval racetrack on
1,200 acres with an initial seating capacity of 75,000, which could be expanded in the future as demand
necessitates. This facility is designed for NASCAR Winston Cup and Busch Grand Nationals auto races.
The Speedway is expected to be open for the 2001 racing season and is anticipated to host four three-day
events per year.

Each of the venues currently in operation has a plan in place to facilitate traffic leaving events, but much of
the discussion in this paper will deal with the plan in development for the Kansas International Speedway
(KIS).

It was important for speedway designers and transportation officials to thoroughly plan and prepare for the
large volumes of traffic that would be generated for events at the facility, and a high level of communication
among various city and county representatives, law enforcement personnel, and transportation officials was
required to develop an adequate traffic plan. Five priorities were established in the formulation of this plan:

1. Maintain non-event traffic capacity on adjacent Interstates and local non-event traffic access,
2. Minimize delay time entering and exiting the event site,
3. Mitigate event traffic in neighborhood areas,
4. Maximize use of existing infrastructure, and
5. Provide positive traffic control, ITS applications, and information systems.

In order to properly address these priorities, officials are using three main strategies:

1. Computer analysis of service levels and traffic routing,

2. Proposed central traffic control center to monitor traffic queues, arrival rates, and direction and phasing
of traffic flow alternatives, and
3. ITS applications to emphasize travel information and routing.

Each of these strategies has elements that are discussed in more detail in the following paragraphs.

The site of the Speedway is positioned to have access to Interstate 70 on the south and Interstate 435 on the
east; these Interstate highways will be the primary routes for Speedway traffic. Direct access to and from the
Speedway will be provided by five-lane arterial routes with reversible lanes. These arterial routes are served
by three interchanges with I-70 and I-435. Arterial routes also serve traffic exiting the Speedway to the north
and west. Ultimately, these routes also lead to I-70 and/or I-435; this strategy will allow traffic using the
Interstate highways to be spread out over a larger area and lessen its effect on through and local traffic.

Law enforcement priorities involve meeting the demand of on-site parking availability and limiting the amount
of exiting traffic that cuts through surrounding neighborhoods. For these priorities to be met, coordination
between city police, the Kansas Highway Patrol, and KIS officials will be necessary to maintain the flow of
traffic. In the proposed plan, city police will be responsible for managing traffic on local streets and Highway
Patrol officers will deal with traffic on Interstate routes. This leaves KIS personnel with the responsibility
for Speedway access spokes and their intersections with local arterials. To aid these personnel, a variety of
Marcus A. Brewer Page 56

solutions have been proposed, but the most commonly utilized will be variable message signs used in
coordination with traffic police and Highway Patrol cruisers. The signs will be used to provide advance
information and direction to motorists, while enforcement officers will be on hand to manually direct traffic
and enforce temporary lane closures and channelization on arterials and freeway ramps.

The Kansas City metropolitan area, in conjunction with the Kansas and Missouri Departments of
Transportation, is also in the process of implementing an area-wide ITS plan, which includes the construction
of a Traffic Management Operations Center. The Speedway management plan being developed will be
included within this metropolitan initiative, and the center will be used in coordination with efforts by law
enforcement personnel to encourage the most efficient use of the existing roadway infrastructure. Transit
service will be available through three separate transit agencies in the Kansas City metropolitan area; this
service will be highly promoted and served by a separate access facility.

Based on computer simulations, supplemented by observations at similar facilities, traffic proportions are
expected to be approximately 85 percent passenger cars, 10 percent transit service, 3 percent tour bus, and
2 percent unspecified. This combination is expected to keep the external street system relatively free of
pedestrians and allow officials to concentrate on vehicle traffic.

Officials have established three criteria to measure the success or effectiveness of their plan: 1) “dump time”,
with a target maximum of three hours; 2) travel speed for through vehicles on Interstate routes, with a target
minimum of 40 mph; and 3) number of enforcement personnel necessary to carry out the proposed plan, a
target number for which was not specified. Based on computer simulations, the agencies believe their plan
will be able to achieve these goals, but actual results will not be measured until 2001 (7, 8).

Milwaukee

The Milwaukee metropolitan area is currently in a state of significant change in regard to special events
facilities and venues; a key component of these changes is the construction of a new stadium, Miller Park, to
house the MLB Milwaukee Brewers. At the same time, the Wisconsin Department of Transportation
(WisDOT) is developing two major transportation improvements in the area: an ITS program, called
MONITOR, for traffic management and traveler information (TM/TI) along the Interstate 94 Corridor through
Milwaukee; and the resurfacing of US Highway 41, which intersects with the Corridor. Miller Park’s
proposed location is adjacent to this corridor, as is County Stadium, where the Brewers currently play. As part
of the Miller Park Infrastructure Development Project, WisDOT conducted a study to identify preliminary
event management needs and develop conceptual TM/TI alternatives within the corridor and into the US-41
design. Their plan was to create a large, multi-stage project, which is integrated with the new stadium, to
improve TM/TI along the entire corridor.

The I-94 Corridor serves as the main transportation link between Waukesha County and the Milwaukee
Central Business District. The Corridor experiences several hours of congestion each day, and projections
indicate that this congestion will increase. In addition to daily traffic, and excluding event locations that are
open all year, there are more than 450 annual special event occurrences within the Corridor. In particular,
there are a number of major special event sites located within or near the Corridor that account for the
majority of the extra event congestion. Some of those sites are:

• Milwaukee County Stadium – Home of the MLB Milwaukee Brewers, it hosts approximately 80 home
games per year from April to October.
• Miller Park – New home of the Milwaukee Brewers currently under construction, it will replace County
Stadium and will hold about 43,000 fans. To be complete by Opening Day (April) 2000.
• Bradley Center – Home of the National Basketball Association (NBA) Milwaukee Bucks, International
Hockey League Milwaukee Admirals, Indoor Soccer League Milwaukee Wave, and the Marquette
University Golden Eagles men’s basketball team, it also hosts several major concert and entertainment
Marcus A. Brewer Page 57

events. The Center averages between three and five major events per week, and has an estimated annual
attendance above two million.
• Midwest Express Center – Milwaukee’s new convention center currently in Phase II of Construction. The
National Governor’s Convention was hosted at the Midwest Express Center shortly after completion of
Phase I in 1998. It is anticipated that large, high profile events will quadruple due to the new facilities,
with as many as 40-50 major conventions each year.
• Wisconsin State Fairgrounds – Home of the Miller 200 and the 11-day Wisconsin State Fair. The
fairgrounds staff has recently looked to generate additional revenue by increasing the number of auto
races. Also, a plan has been approved to remodel the grounds and increase promotion as a year-round
facility. The Pettit National Ice Arena is also located on the fairgrounds; the ice arena is the home of the
national speed skating time trials. Estimated annual attendance at the Fairgrounds is more than 900,000.
• Lakefront Festival Grounds – Home of Summerfest, the world’s largest music festival. Over 11 days in
late June and early July, the event typically draws more than one million people, not including several
hundred thousand who flock to outlying areas to watch fireworks displays on the opening night of the
event and on July 3rd. The Lakefront Festival Grounds also hosts several ethnic festivals such as
Germanfest, Festa Italiano, Polishfest, Irishfest, Asian Moon Festival, Indian Summer, and African World
Festival during summer months; these festivals each attract between 35,000 and 170,000 visitors over
three to four days.

Some events and venues have had exit traffic plans in place for decades, while others have only begun to plan
in the last few years. As a result, in order to accommodate long-term increases in traffic volumes, anticipated
special event growth, and major freeway reconstruction in the area, a need was identified by both public and
private sectors to improve and/or enhance special event-related traffic management and traveler information
activities throughout the corridor. WisDOT officials began their efforts by determining what event
management activities already existed. This was done through a series of phone interviews and interagency
workshops with officials from special event organizations, city officials, and police departments in the area;
based on the information received from these efforts, WisDOT began their strategies to deal with traffic at
each venue.

Virtually all of the venues and events discussed here have some form of access to the I-94 freeway and involve
significant numbers of passenger cars:

• Milwaukee County Stadium – Immediately adjacent to freeway with direct access. Predominantly
passenger car traffic (90%) with a small portion of tour buses (10%).
• Miller Park – Immediately adjacent to freeway. Traffic anticipated to be the same as County Stadium,
with perhaps a small increase in the use of tour buses.
• Bradley Center – Located downtown, two blocks away from existing freeway spur. Predominantly
passenger car traffic.
• Midwest Express Center – Located downtown, three blocks from freeway. Local events generate
passenger car traffic, while large conventions produce traffic from taxis and airport shuttles.
• Wisconsin State Fairgrounds – Adjacent to freeway. Large number of passenger cars, but also significant
transit and tour bus activity.
• Lakefront Festival Grounds – A portion of the grounds is located under an elevated freeway; however,
city streets are the primary routes used to access parking. Most visitors arrive by passenger car (about
72%), transit or shuttle bus (25%), or on foot.

Transit options are available for Summerfest and Wisconsin State Fair. Strategies that have been highly
effective include providing shuttle service to remote park-and-ride lots and parking structures along major
arterials at reduced rates. Very heavy pedestrian activity is not widespread, but an intersection in the vicinity
of Summerfest routinely becomes saturated with pedestrians and vehicles. Also, a large number of pedestrians
cross major arterials during State Fair because residents adjacent to the facility sell parking on their property
at a reduced rate.
Marcus A. Brewer Page 58

Since each venue has different specific needs, a separate traffic plan is developed for each one. The Miller
Park plan is currently under development, and is more extensive than the plan for County Stadium, largely
because all of the elements necessary for the plan can be designed and built into the new stadium. Specifically
for Miller Park, several technologies and other solutions for facilitating exiting traffic have been or are in the
process of being instituted. These are:

• Additional Freeway Variable Message Signs,

• Initial Deployment of Arterial Variable Message Signs in the Village of West Milwaukee,
• CCTV Video Sharing with Village of West Milwaukee Police and Fire Dispatch,
• Direct Video / Communications Link into Miller Park to the Milwaukee County Sheriff’s office
(Milwaukee County Sheriff is responsible for patrolling adjacent freeways and security on-site),
• On-ramp Gates to assist Milwaukee County Sheriff in restricting access at locations that are frequently
closed,
• Traveler Advisory Radio,
• Dynamic Message Signs specifically for special event information, and
• Informational Kiosks inside Miller Park.

Despite the unique characteristics of each venue, there are some common issues in each venue’s plan that can
be addressed, which is why WisDOT is developing the MONITOR program. A key component of this
program is the WisDOT MONITOR Freeway Traffic Operations Center, which is located in downtown
Milwaukee. Typically engineers and/or operators from the TOC participate in traffic task force meetings to
determine how MONITOR elements may be best utilized during special events. Representative actions
typically exercised to assist in the management of exiting special event traffic include:

• Operate freeway and portable variable message signs,

• Operate traveler advisory radio,
• Freeze important CCTV images to the media and Milwaukee County Sheriff’s Department, and
• Modify ramp meter timings.

Officials in the Milwaukee metropolitan area have made multi-agency communication and cooperation the
highest of priorities in implementing their traffic plans. The initial I-94 Corridor Special Event Traffic
Management & Traveler Information Study involved city officials and police department representatives from
four cities, as well as officials from Milwaukee County, the Milwaukee County Transit System, the
Milwaukee County Sheriff’s Department, the State Fair, the Brewers, Summerfest, and WisDOT. This same
high level of communication also involves detailed documentation of plans and strategies, and it has remained
a priority in ongoing traffic operations and planned improvements (9-10.)

Orlando

The city of Orlando has two primary fixed venues that cause significant traffic congestion:

• Orlando Arena – Home of the NBA Orlando Magic, it also hosts other basketball games, Arena Football,
hockey, and various shows and concerts. Events occur on an average of three per week.
• Florida Citrus Bowl (FCB) – Hosts the annual CompUSA Florida Citrus Bowl game every January 1st,
and is the football home of the University of Central Florida. This stadium also hosts local high school
football games, occasional NFL pre-season games, concerts, and Monster Truck and Motocross events.
Approximately 20 events are held annually at the Florida Citrus Bowl.

Both of these facilities have had exit traffic plans in place since they opened, but the City of Orlando has
modified these plans over time to continue to meet traffic demands. Combined, these two facilities host a
wide variety of events; consequently, there are a number of priorities that city traffic officials have established
in moving traffic after events. Some of these priorities include: one-way and street closure schemes; shuttle
Marcus A. Brewer Page 59

bus operations; traffic assignment locations for police; parking flow schemes; disabled parking; charter bus
parking; and limousine parking.

The Orlando Arena is located in the middle of downtown and is bordered by an Interstate freeway and two
major state highways. The estimated traffic is 80 percent passenger cars, 10 percent transit, and 10 percent
walk-up. The transit service for Arena events is a bus circulator that utilizes exclusive lanes. The FCB is
located west of downtown, in a residential/commercial area, but is also bordered by two major state roads and
an expressway. For the annual CompUSA Florida Citrus Bowl, traffic proportions follow a pattern of 50
percent passenger cars, 20 percent charter buses, 15 percent shuttle buses, and 15 percent walk-up traffic.
Other events at the FCB are less predictable and modal splits can vary significantly. Both venues have
numerous local, arterial, and collector streets to facilitate traffic flow.

Based on this knowledge of their facilities and events, the City of Orlando implemented a variety of
technologies and other solutions to address post-event traffic problems. A key component of the plan for the
Arena involves the use of fiber optic parking directional signs, which are used for both pre- and post-event
traffic. For events at the FCB, variable message signs are used to alert motorists of congestion related to event
traffic and to direct fans to shuttle park-and-ride locations.

Since the amount and modal split of traffic can vary widely, particularly for FCB events, the City of Orlando
keeps a high priority on the ability to change their traffic plans “on the fly.” For example, local college
football games often generate major traffic volumes on the west side of the stadium, while the fan base for
a Monster Truck or Motocross event often originates south of the stadium. In these situations, it is of utmost
importance to be able to adapt and not be over-committed to the plan originally designed for the event.

Orlando has a Traffic Management Center that works with event staff to adjust pre- and post-event signal
timing for Arena events. However, there has been much less success in implementing a signal timing plan
for FCB events due to heavy pedestrian/vehicle volumes and conflicts. Both venues experience a high degree
of pedestrian/vehicle interaction, largely due to the location of parking facilities and the large number of
vehicles that park in adjacent neighborhoods and on private property. Traffic officials have, on occasion,
attempted to establish post-event “pedestrian corridors”, but they achieved little to no success. Policing of
these corridors proved expensive and difficult, and it became a public relations problem as people were
directed along specified routes and then often had to double-back to get to their destination. Vehicle traffic
moves very slowly immediately after events, which usually allows ample time for pedestrians to make their
way through parking lots and adjoining streets. Thus, the pedestrian corridor plan was discontinued.

As mentioned above, transit options are available for events at Orlando Arena by the use of a downtown bus
circulator. It is a free shuttle service that encourages fans to park at remote sites and ride to and from the
Arena. There are 11 shuttle stops in the downtown area, with the primary station located in a parking garage
at the Arena. During selected events at the FCB, the City of Orlando and their local public transit service
partner an agreement to provide round-trip shuttle service for $3 per person. This service is not provided for
all events, however, as ridership for certain events has historically been extremely low.

Due to the pedestrian interaction noted earlier, law enforcement after special events is generally confined to
encouraging the continuous flow of traffic. Jaywalking is rarely enforced, and illegally parked vehicles are
ticketed or towed only if they obstruct line of sight or create a hazard to pedestrians. The city must also
address the “business as usual” aspects of local businesses in the area. Based on this principle, they do not
implement one-way traffic plans on streets surrounding the Arena, primarily due to a large hotel that is
adjacent to the arena. Attempting to keep local traffic flowing through the area is a challenge that the city
must face after each event.

Orlando generally measures their success in “dump time”. The Orlando Arena has a capacity of
approximately 16,000 and has a target of about 35-40 minutes. The FCB has a capacity of approximately
Marcus A. Brewer Page 60

72,000 and has a target of about 75 minutes. The plans and strategies used by the City of Orlando generally
meet these target times.

In the summer of 1994, Orlando was a host site for the 1994 Men’s World Cup. Soccer games were held at
the FCB on five different days in late June and early July. Many of the strategies used in the World Cup
exiting traffic plan were adapted from normal exiting traffic plans used for FCB events. A 29-member team
from the Orlando Department of Public Works was stationed at strategic sites throughout the area surrounding
the event. Each was equipped with a radio and assigned to a specific aspect of the traffic plan, such as street
closure or traffic signal operations. Key components of the World Cup post-event traffic plan were: extensive
use of lane closures and directional lane schemes; dedicated lanes for shuttle buses and law enforcement;
widespread use of signing for temporary detours and directions for local traffic; adjustment of signal timings;
and an extensive park-and-ride shuttle bus system (12, 13).

San Antonio

San Antonio hosts a number of events each year, ranging from sporting events to seasonal festivals. Primary
venues for these events are:

• Alamodome – A multipurpose stadium, it is the home of the NBA San Antonio Spurs with four pre-season
and 41 regular-season games, plus playoffs. Most Spurs games draw approximately 20,000 per game,
but attendance can reach as much as 32,000 for selected games. The Alamodome is also the site of the
annual Alamo Bowl and hosts the Big XII Conference football championship every other year; with
attendance at both events estimated at 65,000. Other annual events include about six major concerts and
shows (attendance above 40,000) and 10-12 smaller shows (attendance above 20,000).
• Downtown San Antonio – A multi-block area in the heart of the city that includes the Riverwalk and is
the site of several major festivals each year. Fiesta is a 10-day festival in April highlighted by: three
major parades that each draw 250,000; “Night in Old San Antonio”, which is a four-night street party with
100,000 in nightly attendance; and Fiesta Oyster Bake, a two-day event that draws 50,000 each day. The
Texas Folklife Festival is a four-day event in August at the Institute of Texan Culture; it brings in between
40,000 and 50,000 each day. The Holiday River Lighting immediately after Thanksgiving draws 50,000-
100,000 attendees, and the New Years Celebration attracts 100,000-250,000 to a street party and
fireworks.

San Antonio has had a traffic/transit operations plan for the Alamodome since it was opened in 1993;
operating plans for other events have been developed and modified as needed. The vast majority of the
discussion below will focus on the Alamodome plan.

City officials gathered the input of the community, business interests, political entities, and transportation
professionals to develop a set of priorities for their traffic plans. Those priorities include:

• Provide for safe ingress and egress of event attendees,

• Control on-street parking in adjacent neighborhoods,
• Maintain non-event traffic and transit flow, and
• Accommodate downtown pedestrian movements.

In order to address these priorities, city officials implemented several solutions. They built the Robert
Thompson Transit station adjacent to the Alamodome, where they can transport more than 20,000 attendees
per hour from the Alamodome to parking shuttles and Park & Ride lots. They also implemented the
TransGuide traffic management system. TransGuide is a regional traffic management/operations center. It
is utilized on a daily basis, but is particularly helpful for events. This center coordinates activities such as
ramp closures, adjustments for accidents and congestion, and the display of messages on variable message
signs to describe parking availability and suggested exit routes. There is a TransGuide work station in the
Marcus A. Brewer Page 61

Robert Thompson Transit Station, which allows traffic and transit operations professionals to access all of
the TransGuide cameras and communications from the middle of an event.

Related to the community’s priorities, law enforcement is generally limited to on-street parking in adjacent
neighborhoods. City officials monitor complaints from residents in those neighborhoods to gauge their
performance. Other performance measures include “dump time”, with a target under one hour, and the
percentage of attendees using transit. Based on results from previous events, the dump time target is almost
always met. In addition, officials use on-site observations and assessments to evaluate performance in specific
situations; records of post-event traffic are compared to conditions recorded in previous years to determine
the effectiveness of changes made from year to year.

Modal split for Dome events largely depends on the size, type, and attendee demographics for the event. The
Alamodome has only 3,000 parking spaces on site; a 48-foot wide pedestrian walkway connects the Dome
with downtown parking. Events less than 20,000 generally do not require transit service; at 20,000 in
attendance, transit usage is about 10 percent. As the size of the event grows, so does the transit percentage,
up to about 50 percent for a 65,000-person event. Passenger car access to the area is plentiful, and is the usual
mode of choice. There are dedicated direct entrance ramps from Thompson Station to Interstate 37 for Dome
events. I-10, I-35, and I-37 form a loop around downtown, providing easy access to the entire area.

As mentioned above, available transit service varies with the event. San Antonio has five basic Special Event
Park & Ride sites around the city. Only one may be utilized for small events, while all five may be necessary
for very large events. Downtown parking shuttles are only used for very large events and the length and
number of shuttle routes may be adjusted to better meet event needs.

The pedestrian walkway to downtown parking is fully grade-separated as it crosses over and under parts of
I-37 in the downtown area; there are no traffic conflicts until the walkway empties at a downtown intersection,
which is closely monitored by San Antonio police. This walkway focuses most pedestrian activity on one
route, but there are still occasional problems with pedestrians jaywalking across expressway entrance ramps,
which police address when necessary (14, 15).

Seattle

Seattle hosts many sports-related special events, and they take place at a variety of facilities. This is a
selection of some of those facilities:

• Kingdome – Home of the NFL Seattle Seahawks for 10 games each season, Kingdome is a multi-purpose
stadium that seats 66,400 for football. It was also the home of the NBA Seattle Supersonics until 1995
and home of the MLB Seattle Mariners until July 1999.
• Key Arena – Current home of the Supersonics for approximately 40 games per season, Key Arena opened
in 1995 and seats 17,072.
• Safeco Field – Opened on July 15, 1999, it is the current home of the Mariners for 81 games per season.
A retractable-roof baseball-only facility, it seats about 47,000.

Less-frequent events in the Seattle area include the Goodwill Games, an annual PGA golf tournament, and
two NHRA drag-racing events per year.

The Seattle area has had a comprehensive special events traffic plan in place for about five years. In that plan,
officials gave high priority to moving large volumes of people in the most efficient manner. This was done
by emphasizing high-occupancy vehicles, buses, and preferred parking. Some other strategies used by Seattle
officials to facilitate exiting traffic include: police control at intersections; variable message signs; fixed signs;
highway advisory radio; and helicopter observation.
Marcus A. Brewer Page 62

The information provided to variable message signs and highway advisory radio is coordinated by Seattle’s
Traffic Operations Center, which is also used to inform the public of traffic conditions in non-event situations.

All of these events in the Seattle area are located in relatively close proximity to area freeways, which are the
primary exit routes for special events. The estimated modal split for these events is 40 percent transit, 30
percent passenger car, 10 percent on foot, and 20 percent other. Transit options are available for all events,
and use of these options is encouraged to the public, as evidenced by the modal split. There is a high degree
of pedestrian/vehicle interaction, but there was no indication of any significant problem with accidents.

Seattle officials use complaints from the public as a way to measure effectiveness of their post-event traffic
plan; other measures are cost of resources used, and press reaction to implemented strategies (16).

Washington, DC / Baltimore, MD

The large metropolitan area surrounding our nation’s capital is constantly busy with a wide variety of special
events. The Maryland State Police (MSP) at the Forestville Barrack and the Maryland State Highway
Administration (MSHA) focus on three primary venues in Washington at which they jointly, along with local
police and public works departments, provide assistance in mitigating traffic congestion:

• Jack Kent Cooke Stadium – Home of the NFL Washington Redskins for 10 games per year, it is also the
site of concerts and other sporting events such as World Cup soccer. Capacity is approximately 80,000.
Predominant modes of travel are: passenger cars 92%, tour bus 5%, transit bus 3%.
• Six Flags of America – An amusement park in the Six Flags chain that draws about 30,000 attendees per
day. Predominant modes of travel are: passenger cars 95%, tour bus 5%.
• Andrews Air Force Base – Hosts an Open House each year, which draws about 800,000 visitors over a
three-day period. Passenger cars make up virtually all of the event traffic.

The involved police and highway agencies, working closely with facilities staff, have prepared Special Orders
outlining their purpose and procedures for each of the three venues; these orders were prepared in 1997 for
the stadium, in 1999 for Six Flags, and revised annually for Andrews AFB.

Their main priority is keeping traffic flowing into and out of the event venues and eliminating queues on
adjacent Interstate highways. In addition, they work to reduce the traffic on major arterials and throughout
local neighborhoods. Pursuant to these priorities, the flow of traffic is emphasized over enforcement. Illegal
parking is targeted, particularly in residential neighborhoods, when it impedes efficient operation. Otherwise,
their main concern is the direction of traffic.

Both the stadium and Six Flags are in close proximity to both an Interstate highway and major expressways.
A primary interchange for the stadium was built on the Interstate for stadium use only. There are also two
major arterial highways with lane control signal systems that lead to the Interstate from the stadium. Six Flags
and Andrews AFB are located between the Interstate and a major highway and are both on a major commuting
thoroughfare.

Transit options vary widely for these venues. For stadium traffic, free parking is allowed at surrounding
Metro Center (subway) stations. Shuttle buses are also available to transport people to and from the stadium
for a minimal fee; this service is widely publicized in both print and media coverage. Six Flags has no transit
option other than Metro, which carries an insignificant number of riders to the park. Andrews AFB has
sufficient parking for all event traffic and there is no transit service available.

Pedestrian conflicts can be very significant at these facilities. The stadium complex is not equipped to
accommodate parking for all patrons; therefore, approximately 1,000-1,500 vehicles are parked about a mile
away at an adjacent sports arena. This results in a large number of pedestrians (about 3,000) crossing the
ramps of the Interstate interchange at the stadium. Six Flags experiences a high number of vehicles that drop
Marcus A. Brewer Page 63

off children at the park and return later to pick them up. There is no designated place for parking or stopping
for pickup, and this creates a high level of interaction between vehicles and children near the entrance to the
park. MSP is working with the park to establish an area for this purpose. The ample parking at Andrews
virtually eliminates any pedestrian-vehicle interaction outside of the parking area.

The various agencies utilize the MSHA Traffic Operations Center and a local Command Post located inside
Jack Kent Cooke Stadium. Both sites use cameras to give real-time reports of post-event traffic conditions,
which assists them in adjusting their strategies to maintain the flow of traffic. As traffic operations at each
venue progressed in a positive manner, the agencies were able to reduce the level of manpower needed to
adequately staff these venues, which caused a corresponding reduction in expenses. They still observe “dump
time” at these venues, and they continue to monitor any complaints they receive.

Some specific technologies and solutions used by the agencies include: electronic traffic detection; cones and
arrow boards; portable and stationary variable message signs; traveler advisory radio; broadcast audio
messages to AM/FM radio stations; police officers at selected intersections; and aerial surveillance.

The MSHA is also involved in other events across the state, most notably in Baltimore. MSHA’s traffic
management plans are created as new stadiums are built or events are planned. In addition to Washington-area
events, MSHA is involved with traffic generated by the MLB Baltimore Orioles (Oriole Park at Camden
Yards) and the NFL Baltimore Ravens (Ravens Stadium at Camden Yards). Less frequent events include
major golf tournaments and concerts.

MSHA’s primary concerns are public safety and the orderly flow of traffic, which is directly related to efforts
by their major partner, the MSP. Orderly and safe traffic flow requires timely and accurate traffic information
supplied to motorists, stadium officials, and parking personnel. Multi-agency coordination is vital to the
success of their strategy, and planning is key to anticipating potential problems.

The technologies used in the Baltimore area are similar to those used in Washington: CCTV; real-time signal
timing plans; variable message signs; traveler advisory radio; and cooperation with commercial radio stations,
either through traffic reporters or through the flagship station of a particular event in coordination with both
the statewide Traffic Operations Center and local Traffic Operations Centers. Again, use of these centers to
monitor and observe conditions has proven very successful in alleviating post-event traffic congestion.
Credibility is of primary importance when relating information to the public, and the controlling agencies have
had great success in establishing credibility and persuading motorists to use the information provided to them
(17-19).

SUMMARY AND DISCUSSION OF SURVEY RESULTS

This section of the paper contains a summary of the information obtained from the survey questionnaire.
Contained in the summary will be a discussion of the trends and patterns observed from the survey
information. This discussion will be the basis for conclusions and recommendations outlined later in the
paper.

Nature of Events and Venues

Information in this section is based on event and venue details revealed in the survey results. This is provided
primarily for background information and does not lend itself to significant trends or patterns to be used in
recommendations. Table 2 contains a list of the events described in the previous section and the number of
cities in the survey that consider those events significant. The maximum possible number of cities for any
event is seven, equal to the number of cities in the survey.
Marcus A. Brewer Page 64

Table 2. Listing of Special Events and Their Frequencies in the Survey

Number Number
Event Event/Venue
of Cities of Cities
Sporting Events 7 Concerts 6
Basketball 6 Conventions 2
Baseball 5 Amusement Parks 2
Football 5
Soccer 5 Other 3
Hockey 4 Festival/Fair 2
Auto Racing 2 Air Force Base 1
Other Sports 5

Table 2 indicates a heavy emphasis on sporting events traffic as being important to transportation officials.
In fact, each city in the survey had more than one kind of sporting event that was considered significant. The
wide variety of sporting events also reveals the large number of facilities that are in use or are being built to
accommodate these events. Table 3 contains a listing of sports venues described in the survey, and their
respective frequencies. As in Table 2, the maximum possible number of cities for any venue is seven.

Table 3. Listing of Sports Venues in the Survey

Number
Venue
of Cities
Multi-Purpose Arena 5
Multi-Purpose Dome 2
Baseball Stadium 5
Football Stadium 3
Racetrack/Speedway 2

Based on Table 3, multi-purpose facilities are quite common; however, current trends toward single-occupant
facilities are also evident in the survey results. In relation to exiting traffic, multipurpose facilities can
simplify the preparations needed for an event traffic plan, because many events will occur in the same place
and utilize the same infrastructure; personnel would be accustomed to the characteristics of that venue, and
implementation of the plan could be more efficient. Conversely, single-event facilities have the advantage
that post-event traffic plans for those facilities can be adapted to specific events and specially designed
infrastructure to maximize efficiency.
Marcus A. Brewer Page 65

Traffic Plan Characteristics

All of the cities surveyed have a plan in place to deal with traffic exiting facilities currently in use and are
developing or adapting plans for post-event traffic from future facilities. The history of these plans largely
depends on the facilities being considered, as most cities are implementing plans when new venues open.
Indeed, for the most recent of these venues, a primary concern is the accommodation of traffic in the most
efficient means possible. This factor influences the location of new venues, and it usually results in major
infrastructure improvement projects included as part of the construction of the facility.

Most of the agencies participating in the survey indicated that their main priorities in establishing a post-event
traffic plan were protecting the safety of the public and maintaining traffic flow in the vicinity of the event
site. These two main priorities were related to other responses given: encouraging transit use, monitoring on-
street parking in the vicinity, minimizing delay, and maintaining capacity and speed on area streets and
freeways.

A number of respondents indicated that they employed significant input from the community, area
neighborhoods, and business interests in establishing their priorities. This process also provided an added
benefit in the form of increased credibility and positive public relations, which often made implementation
of plans easier and more successful.

Application of ITS Technologies and Other Solutions

A wide variety of strategies were employed by the cities and agencies surveyed. Table 4 lists some of the
technologies and solutions identified by survey participants as useful in facilitating exiting special event
traffic. Similar to Tables 2 and 3, the highest possible number of cities for any solution is seven.

Table 4 indicates some significant patterns in the widespread use of certain strategies. All cities employ
TMCs in their post-event traffic plans. Each city also uses VMSs for traveler information, on-site law
enforcement officers, and some form of transit service. All but one city indicated they utilize on-site
communications among the various agencies involved in traffic plans, and all but two noted the use of
physical changes to the normal flow of traffic by channelization, reversible lanes, and/or ramp or lane
closures. These results would seem to indicate a general acceptance of these strategies and their effectiveness.
Indeed, on-site law enforcement and transit services are basically universal at special events and have been
for quite some time. While TMCs and VMSs are newer strategies, they have also gained acceptance and have
proven to be effective in many different situations and in many locations.

The remaining strategies listed in Table 4 are not necessarily ineffective; in fact, the cities that use them have
indicated great success with them. However, most of these solutions are still relatively new for use in event
traffic, and they are somewhat more difficult and/or costly to implement effectively. This has kept these
strategies from becoming as widespread as the others. Still, three of these solutions – signal adjustment,
advisory radio, and CCTV – have been implemented by more than half of the respondents; this would seem
to indicate a growing acceptance for these strategies as well.

Law Enforcement

The majority of the cities surveyed reported that law enforcement after special events was largely limited to
illegal parking and, in some instances, jaywalking. Enforcement personnel were more concerned with
maintaining the flow of traffic and keeping it on designated exit routes in the proper directions. Parking
enforcement was designed to remove any impediments to traffic along designated exit routes; jaywalking and
pedestrian enforcement was designed to maintain traffic flow and increase safety.
Marcus A. Brewer Page 66

Table 4. Listing of Technologies and Solutions Implemented by Survey Participants

Number
Technology or Solution
of Cities
Monitoring by Traffic Management Center (TMC) 7
Variable Message Signs (VMS) 7
On-site Law Enforcement (Traffic Police/Sheriff) 7
Shuttle Bus/Park & Ride 7
On-site Multi-agency Communications 6
Channelization/Reversible Lanes/Closures 5
Dynamic/Real-time Traffic Signal Adjustments 4
Traveler Advisory Radio 4
Video Detection/CCTV 4
Kiosks 2
Helicopter/Aerial Surveillance 2

Measures of Effectiveness

Five of the seven cities in the survey indicated “dump time” as a means of evaluating the performance of their
exiting traffic plans; five cities also reported that citizen complaints were a key factor. Four cities also
considered cost of resources or manpower used in carrying out their plan. Measures used by single cities
included traffic flow on surrounding routes, press reaction to the traffic plan, and percentage of attendees
using transit. These results indicate a trend toward public involvement. As the time to restore normal traffic
decreases, the public’s opinion of the traffic plan generally improves. Lower “dump time” also decreases the
cost of resources and manpower, but cities that use cost as a measure of success generally indicated it was not
the most important factor in gauging success. City officials were more than willing to spend a few extra
dollars to improve the performance of the traffic plan and gain the confidence of the public.

Respondents also indicated that, based on their selected measures of effectiveness, their respective traffic
plans enjoyed great success in moving large amounts of traffic efficiently. Some cities reported that they also
compared results with records from similar previous events to determine relative successes and identify
opportunities for improvement.

Traffic Management/Operations Center

As mentioned earlier, each city in the survey either has or is planning a Traffic Management/Operations
Center (TMOC) as a key component of its special events traffic plan. Some cities in the survey use the
TMOC primarily for events, and then monitor regular traffic patterns in non-event scenarios. Other cities
utilize the event management capabilities of the TMOC as an added benefit of a facility that is required for
everyday use. In either case, all TMOCs mentioned in this survey are primary elements in the successful
implementation of their respective special events traffic plan. In fact, most of the cities in the survey use the
TMOC as a central command post, where representatives from all involved entities and agencies can observe
post-event traffic conditions and communicate with one another. If, for example, a particular intersection
Marcus A. Brewer Page 67

becomes oversaturated with exiting traffic, officials can make real-time adjustments to signal timings and/or
deploy traffic police to manually direct traffic through the intersection. The existence of a TMOC is an
invaluable benefit to the efficient and successful implementation of event traffic strategies.

Exit Routes and Multi-Modal Concerns

Virtually all of the event venues considered in this survey are, or are designed to be, in close proximity to
Interstate highways or other freeways. Some venues have dedicated interchanges, used only for venue events;
other venues require short-distance travel on arterials or collectors before accessing the freeway. In either
case, these configurations provide easy access for the large number of passenger cars that are anticipated for
all events.

Passenger cars have been the traditional choice of event attendees, and probably will continue to be for the
foreseeable future, based on modal split estimates in this survey. Outside of Seattle, the lowest reported
percentage of attendees who use passenger cars for any event or venue in this survey is 50 percent. Seattle
has placed great emphasis on the use of transit in their area, and has been able to reduce passenger car usage
to an estimated 30 percent for some events. However, even if 30 percent of attendees use passenger cars, the
result is a high number of vehicles that must all exit the same venue at approximately the same time. This is
why the majority of respondents indicated that one of their priorities was moving event traffic to freeways as
efficiently as possible. This involved minimizing event traffic in surrounding neighborhoods and significantly
altering the normal flow of traffic on streets and roads in the adjacent area.

Despite the large volumes of passenger cars, transit use appears to be increasing in popularity where viable
transit options exist. In addition to the high transit usage in Seattle, the cities of Anaheim, Orlando, and San
Antonio also have events or venues with large numbers of attendees utilizing transit. Anaheim reported that
use of shuttle buses approached 30 percent for Convention Center events and 50 percent for Disneyland
traffic. Orlando indicated that as many as 35 percent of attendees of the annual Florida Citrus Bowl game
utilized charter buses or shuttle buses. San Antonio officials said that up to half of attendees at the largest
Alamodome events utilize transit options. Cities are encouraging a variety of transit options such as subways,
light rail systems, and several variations of buses: regularly-scheduled public transit; event-based Park & Ride
shuttles; downtown circulators; and privately-operated transit and tour buses. As transit use increases, cities
and agencies can increase the efficiency of their traffic plans; more people leaving in fewer vehicles reduces
the demand on the roadway system and allows event traffic to clear more quickly.

Pedestrian volumes at all of the events under consideration are very large, as one would expect. However,
most respondents reported no serious problems with pedestrian accidents or other issues. In a few cases, a
large number of pedestrians have to cross freeway ramps and/or arterials in order to access their parking area
or transit station. These situations are generally dealt with by on-site enforcement personnel, but a prolonged
problem may require a more permanent solution such as a grade-separated walkway. On the whole, event
traffic moves slowly enough immediately after events that pedestrian/vehicle interaction is not a significant
problem.

RECOMMENDATIONS

This section of the paper contains a list of recommendations for use in new or developing plans to facilitate
traffic exiting from special events. These recommendations are based on the results of the survey
questionnaire and discussion of those results contained in the previous section.
Marcus A. Brewer Page 68

• Utilize a Traffic Management/Operations Center as the central point of coordination

A well-run traffic plan requires a great deal of coordination among event personnel and among all
participating agencies. The most efficient way to facilitate this coordination is to have a single location where
all involved agencies can have representatives who can communicate with each other. A TMOC can serve
as a base of operations and can be furnished with the personnel and equipment necessary to carry out the
traffic plan. TMOCs are designed to monitor traffic conditions throughout an area and be able to make
immediate decisions concerning the re-routing of traffic and the assignment of emergency personnel. These
are functions that are utilized extensively in post-event traffic plans, and the use of a TMOC to carry out these
functions makes good use of resources.

• Determine priorities that should be met in the exiting traffic plan

The obvious priorities in any post-event traffic plan should be the most efficient movement of traffic away
from the venue and the maintenance of public safety. However, it is necessary to go beyond that basic
foundation and establish what elements are necessary to make those initial priorities possible. These
additional elements could include the direction of traffic to freeways, the rapid removal of event traffic from
adjacent neighborhoods, or the maximization of transit use. Definition of these detailed priorities will provide
reachable goals and objectives that can be pursued.

• Obtain and utilize feedback from the community and affected parties
In order to properly establish the necessary priorities in the plan, it is necessary to obtain the opinions and
concerns of the community. Ultimately, the success of any plan will depend on how well the traveling public
follows that plan. Any plan that has public support will be more likely to succeed, because they will be
inclined to understand and agree with the methods and goals of that plan. In particular, the concerns of
affected parties should be taken into account. These parties can include officials from the event venue and
event organizers, area businesses, residents of adjacent neighborhoods, and local media. The support of these
groups will greatly increase the effectiveness of any plan that is implemented, and the use of their feedback
will make the plan more relevant to the needs of the community.

• Maintain a high level of communication and cooperation between agencies

The success of any plan depends on the contributions of many different people from a number of different
agencies. Elected officials, law enforcement personnel, transportation officials, and emergency services all
have key roles to play in the implementation of a special event traffic plan. Each group has certain interests
and aspects of the plan on which they must focus, and they require certain tools to do their jobs effectively.
Therefore, it is necessary that all of these agencies understand each of their roles, understand what is
necessary to fulfill those roles, and work together to ensure the best possible performance of each participant.

• Determine what technologies and solutions are necessary to implement the plan and meet objectives
After objectives have been established and each agency knows their responsibilities, it becomes necessary to
define what solutions can be applied to meet the objectives effectively. A means of providing information
to travelers is necessary, whether it be through the use of on-site traffic police manually directing traffic, or
through the use of variable message signs, informational kiosks, or advisory radio to supply advance notice
of changes to normal traffic patterns. If transit use is to be encouraged, there need to be viable and practical
transit options in existence, with good access to parking areas and the event venue. The proper use of the
correct applications is vital to accomplishing the objectives of the traffic plan.

• Establish measures of success or effectiveness and use them to improve the plan
In order to determine how well the plan works, there needs to be a means of measuring success or
effectiveness. This can be accomplished by monitoring “dump time”, responding to public complaints,
reducing the cost of resources used, or a number of other criteria. These criteria can be used as targets to
strive for, or they can simply be used as a means of recording performance over time. After analyzing the
results of implementing the plan, decisions can be made as to whether the plan was effective and whether
improvements can be made.
Marcus A. Brewer Page 69

• Be willing and able to adapt the plan during implementation

It is extremely unlikely that a plan is going to be perfect, especially on the first attempt. Therefore, it is
important that the plan be flexible enough to make changes during implementation. If estimates incorrectly
predict certain aspects of post-event traffic conditions, unexpected problems will arise. Contingencies should
be available to compensate for those unexpected problems and deal with them as efficiently as possible. The
ability to adapt during implementation will not only improve the plan’s effectiveness, but it will also improve
the public’s image of the performance of agencies involved.

APPLICATION OF RECOMMENDATIONS: A CASE STUDY

The information contained in this section of the paper is intended to be an illustration of the development of
a plan to facilitate traffic due to special events at a specific venue. It was formulated by the author to
demonstrate the use of the procedures and recommendations contained in this paper. Specifically, it will rely
heavily on the recommendations listed in the previous section. Each segment in the scenario described in this
section corresponds to one of those recommendations.

Introduction

Houston, Texas, has the second-largest metropolitan area in the state, with a population of approximately three
million occupying most of Harris County and parts of neighboring counties. Houston is the site of many
special events of various natures. One venue for special events in Houston is the Astrodome, which is the
home of the MLB Houston Astros for 81 regular-season games with a capacity of 54,313. The Astrodome
also was once the home of the NFL Houston Oilers, and it is still the site of many other sporting events,
concerts, conventions, and the annual Livestock Show and Rodeo (20). The Astrodome is located in south-
central Houston, bordered by Interstate Loop 610 on the south, Old Spanish Trail (US Highway 90A) on the
north, Kirby Drive on the west, and Fannin Street on the east. Six Flags AstroWorld amusement park is
located across I-610. Figure 1 is a detail map illustrating the Astrodome area (21).

The information in the remainder of this section of the paper is based on the scenario created and developed
by the author. With the exception of the advisory commission created specifically for this scenario, all of the
agencies mentioned in this section actually exist; however, any actions by those entities described in this
scenario are created by the author and do not necessarily reflect conditions in reality.

The scenario created by the author is based on the assumption that city, county, and event officials have
decided that the plan to facilitate exiting traffic at Astrodome events needed to be updated to reflect
infrastructure changes and current attendance and travel patterns. In order to update their plan, officials
decided to completely reexamine their existing plan for ideas about possible improvements. They formed an
advisory commission to study potential options and perform a wholesale reevaluation of strategies and
priorities. Upon completion of their study, the commission would make recommendations to the city and
county for their approval.
Marcus A. Brewer Page 70

Figure 1. Astrodome Area Map

Traffic Management Center

Houston has a very active traffic management program, called TranStar. This program includes a traffic
management center that is used to monitor traffic on area freeways, staff personnel from law enforcement and
emergency services, and maintain lines of communications with agencies throughout Harris County.

The commission decided to utilize the TranStar center as the command center for their updated special events
traffic plan. This would allow them to use existing facilities with trained personnel and appropriate
equipment. Use of the TMC also centralized the operations of several agencies, streamlined inter-agency
communications, and simplified preparations for implementation of the new traffic plan.

Priorities

In order to move event traffic away from the Astrodome quickly and safely, the commission decided to utilize
Interstate 610 as the primary exit route. This strategy would involve channeling the majority of traffic to
interchanges and frontage roads at Kirby and Fannin. Specifically, traffic would be divided such that all
traffic exiting the Astrodome grounds by the southern two exits to Fannin Street would be directed to
eastbound I-610. Similarly, traffic on the southern two exits onto Kirby Drive would be directed to westbound
I-610. This is shown graphically in Figure 2.

To expedite the flow of traffic onto I-610, the eastbound ramps at Kirby would be temporarily closed, as
would the westbound ramps at Fannin.

Much of the remaining traffic could be directed to the northern exits on Kirby or Fannin or to North Stadium
Drive leading to US-90A where they would have two options; they could either turn southbound on US-90A
and access I-610 away from the Astrodome or continue northbound and travel through central Houston to their
destination.

A related priority that would be pursued would be the minimization of event traffic in area neighborhoods.
If the majority of traffic can access the freeway and frontage roads, this would remove large amounts of traffic
that would otherwise be creating congestion on adjacent streets.

A third priority was the use of public transit. Houston Metro transit has a number of regular bus routes that
run on streets adjacent to the Astrodome, and a Park & Ride system could be easily established.
Marcus A. Brewer Page 71

Figure 2. Exit Routes for I-610 Traffic

Public Feedback

After establishing some basic priorities, the commission decided to gather public input and feedback on their
concerns and needs. Some groups that were specifically addressed include Astros season ticket holders,
residents in adjacent neighborhoods, and owners of businesses along the I-610 corridor. In addition, officials
from Six Flags AstroWorld theme park, located across I-610 from the Astrodome, were consulted to obtain
their opinions and concerns. Comments from these groups were supplemented by feedback from
advertisements made in the Houston Chronicle.

Analysis of public responses led the commission to include a few more considerations to their list of priorities.
Nearby residents echoed the need to minimize neighborhood traffic. Season ticket holders wanted to be able
to have preferred parking with easy access to exit routes, while local business owners were worried that the
changes in traffic flows would adversely affect access to their establishments. Six Flags officials wanted to
be able to coordinate activities with Astrodome officials to guard against both facilities releasing large traffic
volumes at the same time.

General feedback indicated that many attendees found existing traffic control somewhat confusing, and they
felt the time required to leave the event and make their way onto I-610 was too long. Furthermore, non-event
traffic in the area often was caught in event-related congestion that could have been avoided had there been
advance notification and opportunities to detour around the affected area. The commission decided to take
all of these factors into account in their revised plan.

Interagency Communication

The commission was made up of members from the following entities: City of Houston, Harris County,
Houston Police Department, Houston Fire Department, Harris County Sheriff’s Department, TranStar, Harris
County-Houston Sports Authority, Houston Astros, Houston Livestock Show and Rodeo, Houston Metro
Transit, Texas Department of Public Safety, and Texas Department of Transportation. Each of these agencies
and organizations had a vested interest in the implementation and the success of the new traffic plan.

Commission members worked to establish responsibilities and obligations for each entity. These
responsibilities were designed to be related to each entity’s specific interests and areas of expertise. As
mentioned above, TranStar facilities would be used to have a central point of coordination. Police and
sheriff’s officers would share enforcement and traffic duties on local streets, while DPS officers would be
responsible for activity on I-610. The fire department would have EMT personnel on-site with a full
firefighting crew stationed nearby. Metro officials would be responsible for transit service, including normal
bus service and park-and-ride shuttles. TxDOT officials would work with the city and the county to design
specific plans for re-routing and channelization of traffic. Event officials and the Sports Authority would be
Marcus A. Brewer Page 72

responsible for helping to publicize the new plan and helping to pay for the cost of resources needed to
implement the plan.

Commission members maintained an open dialogue among themselves, and worked to eliminate any
jurisdictional or geographical obstacles. All of these responsibilities and obligations were included in a
formal report outlining the entire Astrodome traffic plan, copies of which would be made available to each
agency for their use. After creation of the plan was completed, commission members held regular meetings
and special post-event meetings to discuss performance and problems.

Technologies and Solutions

In order to effectively route traffic to I-610, transportation officials felt it necessary to utilize a high degree
of positive guidance and channelization. Traffic flow on both Kirby and Fannin adjacent to Astrodome
grounds would be redirected to maximize capacity for exit traffic; this would be accomplished through the
use of arrow boards, cones, barrels, and traffic police. Barrels and cones would be placed at each exit from
the Astrodome grounds to guide traffic in the directions outlined in the exit traffic plan, so that traffic on
southern exits would be directed to the south and traffic on northern exits would be directed northward.
Arrow boards would be placed on Kirby and Fannin to guide traffic to freeway entrances on the frontage
roads. An example of this guidance is shown in Figure 3, based on the section of Kirby Drive from Murworth
Drive to I-610.

The high volume of exit traffic would result in a need for signal re-timings at the I-610 access road
interchanges; personnel at TranStar facilities would be given the ability to monitor traffic levels and make
adjustments to timings as necessary. Buses would be assigned to restricted-use lanes on the Astrodome
grounds and given the use of temporary bus-only lanes on Kirby and Fannin.

Non-event traffic on area streets would be notified of pending increases in volume and changes in flow by
portable variable message signs and temporary static signs, which would display informational messages and
suggested detour routes. These signs would be installed on affected streets in advance so that messages could
be displayed through the duration of the event as well as after the completion of the event. Traffic police
would also be placed on patrols throughout the area to reinforce the messages, provide assistance, and enforce
the law when needed. Suggested signing locations are shown in Figure 4.

Non-event traffic on I-610 would be reduced by one lane to allow for the large volumes of merging vehicles.
This would be accomplished by channelization through the use of barrels and arrow boards, preceded by
variable message signs. DPS officers would be strategically stationed throughout the corridor to observe
traffic and enforce the law if necessary. An example of the positioning strategy for non-event traffic on
eastbound I-610 is shown in Figure 5.

Law enforcement would primarily be limited to the removal of obstacles to traffic flow. Parking tickets would
be issued for vehicles blocking exit routes, and citations would be issued to drivers who used bus lanes.
Vehicle traffic on area streets would be periodically stopped by traffic police to allow pedestrians to cross and
to create breaks in traffic entering I-610. Attendees who parked in areas south of the stadium would be
encouraged to use an existing walkway over I-610 that connects to the grounds of Six Flags; this strategy
would significantly reduce the number of pedestrians attempting to walk across freeway ramps and access
road intersections.
Marcus A. Brewer Page 73

Figure 3. Guidance Strategy for Traffic Exiting on Kirby Drive

Figure 4. Suggested Deployment of Informational Signs for Local Traffic

Marcus A. Brewer Page 74

Figure 5. Positioning Strategy for Eastbound I-610 Non-Event Traffic

In order for the commission’s strategies to be effective, it would be necessary to provide adequate information
to event attendees. This would be accomplished through several means. First, small detail maps outlining
all exit routes and their direction of travel would be distributed to each Astros season ticket holder and
included with group ticket sales. Additional copies would be made available for all other event attendees;
these copies would be distributed two ways: the driver of each vehicle entering the Astrodome parking area
would receive
a copy after paying the parking fee at the entrance gates; and supplemental copies would be available at the
information office located inside the Astrodome. In addition, the maps would be reproduced in all media
guides
and programs. Second, at the end of each game or event, images of the detail maps would be placed on the
video board with accompanying announcements by the public address announcer. Third, kiosks would be
installed at strategic locations throughout the Astrodome with information on exit routes and area traffic
conditions. Use of these kiosks would be encouraged through public address announcements during and after
each event. Finally, fixed and portable signs would be placed throughout the Astrodome parking area to guide
travelers to the proper exits.

Measures of Effectiveness

Commission members agreed upon three primary measures of effectiveness for the new plan: “dump time”,
public complaints, and cost of resources and manpower. The order in which these measures are listed is
identical to their order of importance. The commission decided that “dump time” would be the single most
important factor that attendees would consider. As the amount of time required to restore normal traffic flow
decreased, the time that attendees spent waiting in the parking lot would also decrease. This would indicate
that
implementation of the new traffic plan was successful in accomplishing a primary objective.A decrease in
“dump time” would also lead to a decrease in the number of complaints received, according to the
commission. People who spend less time waiting in the traffic queue will be more satisfied with the results
of implementation. The other primary cause of complaints the commission anticipated was due to accidents
or other factors that infringed on safety. However, it was the commission’s belief that a decrease in accidents
Marcus A. Brewer Page 75

was related to more efficient traffic flow and a reduction in “dump time”. By attempting to eliminate the two
main sources of complaints, the commission felt it could then focus on solutions to smaller problems that
would help them fine-tune the plan.

Finally, cost of resources was considered as an important measure. This was partially due to the concerns of
event personnel who wanted to keep their costs down. However, it was also related to the safety and morale
of personnel involved. When traffic officers and management personnel can reduce the amount of time
needed to complete their job, officers are exposed to fewer potential accidents and all personnel can go home
earlier. Cost is also related to “dump time”; a decrease in “dump time” should translate to a decrease in cost,
all other things remaining constant. Thus, the commission was primarily concerned with moving event traffic
out of the area as quickly and safely as possible; the other measures would be by-products of that primary
measure.

Adaptability

In order to know exactly how well the new plan was going to perform, it was necessary to implement it in the
field under actual event conditions. Before doing that, however, commission members wanted to have some
contingencies in place to deal with unforseen problems. Key among those contingencies was ensuring that
each officer and event personnel could be in constant communication with each other and with TranStar
headquarters. The field officers could then make first-hand observations and relate them to other personnel
for response. Incidents such as traffic accidents, stalled vehicles, medical emergencies, or fires could be
reported instantly and the proper authorities could respond with necessary equipment and personnel. This
would greatly reduce the time needed to deal with incidents and help restore traffic to normal.

Provisions were also made to allow traffic officers to manually direct traffic at intersections if necessary. If
signal re-timings were inadequate or if an incident developed, signal operations could be coordinated between
TranStar and police.

Traffic police also communicated conditions to each other, in the event that traffic patterns needed to be
changed. For example, if too many vehicles were going south on Kirby to I-610, police at stadium exits could
then direct more people north to US-90A to route them back to I-610 at a different interchange.

Implementation

Having identified priorities, participants, and solutions, the commission approved the use of the new
Astrodome post-event traffic plan and recommended it to city and county officials. The recommendations
were supported with thorough documentation of the goals and methodology the commission used in reaching
their decision. The commission’s recommendations were then reviewed and approved by the City of Houston
and by Harris County.
Marcus A. Brewer Page 76

ACKNOWLEDGMENTS

This paper was prepared for Advanced Surface Transportation Systems, a graduate course in Transportation
Engineering at Texas A&M University. The instructor for this course was Dr. Conrad L. Dudek. Professional
mentors were Thomas Hicks, Patrick Irwin, Wayne Kittelson, H. Douglas Robertson, William Spreitzer, and
Carol Zimmerman. The author gratefully acknowledges these professionals for their time and effort in
supporting this course. The author would like to extend special appreciation to Thomas Hicks and Dr. Dudek
for their support and comments in preparing this paper. Finally, special thanks are offered to the people listed
below and referenced throughout the paper for taking the time to share their experiences and opinions with
the author.

• Mr. Nathaniel Behura, Anaheim Department of Traffic Engineering

• Lt. Suzanne Jordan, Maryland State Police
• Mr. Frank Juares, City of San Antonio Engineering & Traffic Division
• Mr. Mark Kenneally, TranSystems Corporation, Kansas City, MO
• Mr. Hebbani Lokesh, Wisconsin Department of Transportation
• Mr. Alvin Marquess, Maryland State Highway Administration
• Mr. Ed Mulcahy, TranSystems Corporation, Kansas City, MO
• Mr. Steve Scott, City of Orlando Public Works Department
• Mr. Roger Steinert, Washington Department of Transportation
• Mr. Todd Szymkowski, TransCore, Milwaukee, WI
• 1st Sgt. Ken Thrasher, Maryland State Police
• Mr. Cris Young, VIA Metro Transit, San Antonio, TX

REFERENCES

1. McCasland, W.R., M.A. Ogden, J.M. Mounce, G.L. Ullman, and D.R. Middleton; Guidelines for
Response to Major Freeway Incidents – Response Manual; Research Report 1345-1, Texas Transportation
Institute, Texas A&M University System, College Station, Texas; April 1994.

2. Dudek, C.L.; “Freeway Incidents and Special Events: Scope of the Problem,” Transportation Research
Circular 326; Transportation Research Board, National Research Council, Washington, DC; December
1987.

3. Ogden, M.A., J.M. Mounce, D.R. Middleton, and G.L. Ullman; Traffic Management in Response to Major
Freeway Incidents; Research Report 1345-2F, Texas Transportation Institute, Texas A&M University
System, College Station, Texas; August 1994.

4. Davies, P., N. Ayland, C. Hill, S. Rutherford, M. Hallenbeck, and C. Ulberg; Assessment of Advanced
Technologies for Relieving Urban Traffic Congestion; NCHRP Report 340, Transportation Research
Board, National Research Council, Washington, DC; December 1991.

5. Taylor, S.T.; “Crowd Pleaser,” ITS World; Volume 3, Number 1; January/February 1998.

6. Survey questionnaire completed by Nathaniel Behura, Anaheim Department of Traffic Engineering.

Answers received by telephone interview conducted July 2, 1999.

7. Survey questionnaire completed by Ed Mulcahy and Mark Kenneally, TranSystems Corporation, Kansas
City, Missouri. Answers received by mail July 12, 1999.

8. Kansas International Speedway Traffic Report: Summary of Traffic Projections and Associated Roadway
Improvements; TranSystems Corporation, Kansas City, Missouri; September, 1998.
Marcus A. Brewer Page 77

9. Survey questionnaire completed by Todd Szymkowski, TransCore, Milwaukee, Wisconsin. Answers

received by e-mail July 2, 1999.

10. Survey questionnaire completed by Hebbani Lokesh, Wisconsin Department of Transportation. Answers
received by fax July 13, 1999.

11. I-94 Corridor Special Event Traffic Management & Traveler Information Study; Wisconsin Department
of Transportation and HNTB Corporation, Milwaukee, Wisconsin; January, 1997.

12. Survey questionnaire completed by Steve Scott, City of Orlando Public Works Department. Answers
received by e-mail June 28, 1999.

13. Orlando World Cup Soccer Traffic Control Plan; City of Orlando Transportation Engineering Bureau,
Orlando, Florida; 1994.

14. Survey questionnaire completed by Cris Young, VIA Metro Transit. Answers received by e-mail July
14, 1999.

15. Survey questionnaire completed by Frank Juares, City of San Antonio Engineering & Traffic Division.
Answers received by fax July 14, 1999.

16. Survey questionnaire completed by Roger Steinert, Washington Department of Transportation. Answers
received by e-mail July 7, 1999.

17. Survey questionnaire completed by Lt. Suzanne Jordan and 1st Sgt. Ken Thrasher, Maryland State Police.
Answers received by fax June 30, 1999.

18. Barrack Special Order No. 61-9904, No. 61-9906, and No. 61-9907; Barrack “L”, Central Region,
Maryland State Police, Forestville, Maryland; Spring 1999.

19. Survey questionnaire completed by Alvin Marquess, Maryland State Highway Administration. Answers
received by e-mail July 12, 1999.

20. Houston Astros Internet home page. http://www.astros.com.

21. Microsoft Network Houston Sidewalk home page. http://houston.sidewalk.com.

APPENDIX
Survey Questions

1. What is the nature of events at fixed venues that cause significant traffic congestion in your city or area?
Specifically, what types of events (sports, concerts, fairs) and what frequency (annually, weekly, etc.)?
2. How long have you had a plan in place to expedite the flow of traffic leaving these events?
3. What elements were given priority in your plan to facilitate event traffic, and how did you determine those
priorities?
4. What technologies or other solutions did you implement in order to address those priorities?
5. What issues (i.e., jaywalking, illegal parking, etc.) were emphasized in enforcement?
6. What did you use to measure success or effectiveness? (i.e., “dump time”, cost of resources used, number
of complaints received, etc.)
7. Based on those measures in Question 6, what were the results of implementing the plan in your area?
8. Do you have a Traffic Management/Operations Center in your city or area, or at the site? If so, how is
it used in coordination with your special events traffic plan?
Marcus A. Brewer Page 78

9. What are your primary exit routes for special events? Specifically, is the event venue in close proximity
to Interstate Highways or other freeways?
10. What are the predominant modes of travel to the events and their approximate percentages? (i.e.,
passenger car, tour bus, transit bus, light rail, bicycle, on foot)
11. Are transit options available and/or adapted to special events? If so, is the use of these options
emphasized and encouraged to the public?
12. Is there a high degree of pedestrian/vehicle interaction? (i.e., Do large numbers of pedestrians need to
cross a major exit route in order to access a parking area or transit station?)
13. Do you have further documentation or a formal report outlining your event traffic plan? If so, are copies
available to those who request them?

MARCUS A. BREWER

Marcus A. Brewer received his B.S. degree in Civil Engineering

with honors from the University of Kansas in December, 1997. He
is currently pursuing an M.S. degree in Civil Engineering at Texas
A&M University with an emphasis on transportation science.
Marcus is a recipient of the TTI Robert C. Lanier Distinguished
Research Assistantship and a Graduate Merit Fellowship from the
Texas A&M University Department of Civil Engineering. Marcus
has served as president of the University of Kansas Student
Chapter of American Society of Civil Engineers, and he currently
serves as treasurer of the Texas A&M University Institute of
Transportation Engineers Student Chapter. He is also a member
of ITS America, ASCE, and NSPE. While at the University of
Kansas, Marcus spent five summers as an intern with the Kansas
Department of Transportation. Currently, Marcus is employed by the Texas Transportation Institute as a
Graduate Research Assistant and is investigating the effects of design elements on operating speeds on
suburban arterials. His areas of interest are: design speed and operating speed issues, signalized intersection
operations, traffic accident analysis, and Intelligent Transportation issues.
AN INVESTIGATION OF ADVANCED PARKING INFORMATION SYSTEMS
AT AIRPORTS

by

Debra A. Burdette

Professional Mentor
H. Douglas Robertson, Ph.D., P.E.
Vice President/Operations Manager
Science Applications International Corporation

Course Instructor
Conrad Dudek, Ph.D., P.E.

Prepared for
CVEN 677
Advanced Surface Transportation Systems

Department of Civil Engineering

Texas A&M University
College Station, Texas

August 1999
Debra A. Burdette Page 80

SUMMARY

As a result of the recent increase in air traffic volume and the high degree of dependence on the automobile
as the dominant mode of ground transportation to and from the airport, the access roads and parking lots at
many of the nation’s airports are experiencing additional demands. Additionally, when parking areas near
capacity, circulation problems are compounded especially when travelers are unaware of the location of
available spaces. Although all airports have parking guidance in the form of static signing as navigational
aids, advanced parking information (API) systems attempt to alleviate some of the problems and hassles
associated with searching for available parking spaces by using ITS technologies to first direct travelers to
open lots or garages and then to specific bays.

The overall goal of this research was to develop an advanced parking information system using ITS
technologies for communicating real-time parking information to airport users. In order to familiarize the
author with issues related to airport parking and advanced parking information systems, a review of current
literature and previous research was performed. To fill gaps in the literature reviewed regarding applications
of API systems at airports, officials from airports with existing and planned systems were contacted and
interviewed on the telephone. Additionally, data regarding user preference in parking information (i.e.,
information, location, and sources) was collected using a self-administered survey questionnaire.

Using the information gathered from the literature review, airport interviews, and survey results, guidelines
for implementing an advanced parking information system to provide users with real-time parking availability
or alternate parking suggestions at airports were developed. The guidelines include the following steps.

• Determine which parking facilities are to be included in the API system;

• Identify the parking guidance approach;
• Divide the global system into zones;
• Select the information dissemination strategy;
• Specify system characteristics;
• Determine the maximum capacity of the parking facilities;
• Develop and test system software;
• Deploy API System; and
• Evaluate API System.

To illustrate the use of the guidelines, they were implemented on the hypothetical expansion of the
advanced parking information currently installed in Terminal C at Houston’s George Bush
Intercontinental Airport.
Debra A. Burdette Page 81

TABLE OF CONTENTS

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Research Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Organization of Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

STUDY DESIGN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Data Analysis and Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Airport Parking Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Types of Airport Parking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Parking’s Contribution to Landside Access Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Opportunities for Integrating ITS to Manage Parking Facilities . . . . . . . . . . . . . . . . . . . . . . . 86
Overview of Advanced Parking Information (API) Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Approaches to Parking Information Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Total Space Guidance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Zone-Specific Guidance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Space-Specific Guidance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Strategies for Information Dissemination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Descriptive Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Prescriptive Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
How the API System Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Vehicle Detection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Parking Control Center. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Information Displays. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Telecommunications Network. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Benefits of API Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
User Response to API Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Driver Awareness and Use of API. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
API’s Impact on Search Behavior. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Potential ITS Technologies for Disseminating Advanced Parking Information . . . . . . . . . . . 92
Variable Message Signs (VMS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Static Signs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Highway Advisory Radio (HAR) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Internet Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Telephone Information Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
In-vehicle Navigation Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Commercial Television Broadcasts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

CURRENT APPLICATIONS OF API SYSTEMS AT AIRPORTS . . . . . . . . . . . . . . . . . . . . . . . . 94

Raleigh-Durham International Airport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
George Bush Intercontinental Airport (Houston, TX) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Metropolitan Washington Airports (Ronald Reagan National and Washington Dulles) . . . . 95
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Debra A. Burdette Page 82

SUMMARY OF SURVEY RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

Parking Information Desired . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Preferred Location to Receive Parking Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Desired Method of Information Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

ASSESSMENT OF POTENTIAL ITS TECHNOLOGIES USED IN API SYSTEMS . . . . . . . . . 98

GUIDELINES FOR IMPLEMENTING API SYSTEMS AT AIRPORTS . . . . . . . . . . . . . . . . . . 100

CASE STUDY: GEORGE BUSH INTERCONTINENTAL AIRPORT . . . . . . . . . . . . . . . . . . . . 102

CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

APPENDIX A: SURVEY QUESTIONNAIRE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

APPENDIX B: SURVEY DATA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

Debra A. Burdette Page 83

INTRODUCTION

The implementation of the Intermodal Surface Transportation Efficiency Act in 1991 by the United States
encouraged the development and use of technology to help manage urban transportation systems. This
legislation has resulted in the advancement of intelligent transportation systems (ITS) technologies to provide
enhanced mobility, safety, and traveler convenience. Advanced traveler information systems (ATIS) are a
sub-component of ITS which offer static and dynamic pre-trip and en-route information (1). Examples of
static information include transit schedules and fares, construction schedules, driving directions, and flight
schedules and fares. Dynamic information includes, but is not limited to real-time traffic conditions,
accidents, delays, parking availability, and route guidance.

In recent years the popularity of air travel throughout the United States has continued to rise. With this
increase in air traffic volume and the high degree of dependence on the automobile as the dominant mode
of ground transportation to and from airports, the access roads and parking lots at many of the nation’s
airports are experiencing additional demands (2). Parking area performance is a function of the time required
to find a parking space. As parking areas near capacity, circulation problems are compounded when travelers
are unaware of the location of available spaces. Repeated circling of the airport property while searching
for an open space increases lost time and puts unnecessary strain on airport roads (3). Advanced parking
information (API) systems (sometimes called parking guidance information (PGI) systems) use ITS
technologies to help alleviate some of the problems and hassles associated with searching for available
parking spaces by first directing travelers to open lots or garages and then to specific bays. These systems
have the potential to benefit the overall operations of the parking and access facilities, improve the level of
service for airport users, and possibly increase the airport’s revenue (4).

Problem Statement

Air travelers are predominately concerned with getting to the airport on time by the quickest, most
convenient, and most reliable means possible since they can usually ill-afford to miss their flights and are
frequently on tight schedules. Time spent traveling by air can be divided into three categories including time
in the air, time at the airport, and time on the ground going to and from an airport. Results from a 1994 study
conducted by the Airport Council International-North America revealed that nearly 75 percent of airport
operators surveyed indicated that passengers encounter more delay on access and circulation roads than they
do on the airfield. These findings support the feeling, held by passengers and airport operators, that time
spent on the ground is rapidly increasing (2). Since time spent searching for available parking is a large
portion of the ground access time, utilization of an advanced parking information system to provide
passengers with accessible and accurate parking availability and routing information at critical decision
points could result in a reduction in travel time (5). However, in order to recommend guidelines for the
application of advanced parking information systems at airports, research is needed to ascertain the specific
type of information required by air travelers, especially considering real-time parking information.
Moreover, the optimum location and preferred method of dissemination also need to be identified in order
to assist travelers in making informed decisions prior to and during their trips.

Research Objectives

The overall goal of this research was to develop guidelines for the implementation of an advanced parking
information system using ITS technologies to communicate real-time parking information to airport users.
The specific objectives of this research included:

• Establish the information needs of motorists intending to park at an airport;

• Identify and assess potential ITS technologies to be used in disseminating real-time parking information;
+ Investigate current advanced parking information systems;
Debra A. Burdette Page 84

+ Develop guidelines for an improved advanced parking information system for airports; and
+ Apply the guidelines to a case study at Houston's Gorge Bush Intercontinental Airport.

Scope

This research was limited to developing an improved method of disseminating parking information to airport
users. It included a review of the process of collecting parking availability data to be used as traveler
information, but did not address these issues in the guidelines formulated. In order to meet the objectives
of the study, a survey of Texas Transportation Institute personnel who use airport parking facilities (i.e.,
short-term, long-term, and remote lots), an investigation of potential ITS technologies to be used in providing
parking information, and a review of existing API systems were conducted.

Organization of Report

This report is organized into nine sections. The second section includes the methodology used to conduct
this research. Background information including an overview of airport parking facilities, a discussion on
advanced parking information systems, and a description of potential ITS technologies to be used in
dissemination parking information are presented in the third section. A review of airports currently using
advanced parking information systems is provided in Section 4. The results from the user information needs
survey are presented in the fifth section. The sixth section includes an assessment of ITS technologies based
on findings from the literature review and the results of the parking information needs survey. Guidelines
for implementing advanced parking information systems at airports are found in Section 7. The application
of the guidelines to Houston’s George Bush Intercontinental Airport is in Section 8. Finally, conclusions
formulated based on the results of this research are presented in the last section.

STUDY DESIGN

The procedures followed in carrying out this research to develop guidelines for the implementation of an
improved advance parking information system to communicate real-time parking information to airport users
consisted of four main tasks: literature review, data collection, data analysis, and application. These tasks
are discussed in greater detail in the following sections.

Literature Review

A review of current literature was performed in order to examine previous research and present issues related
to airport parking and advanced parking information systems. Specifically, previous studies of the impact
of parking on airport landside operations, as well as the current methods by which air travelers receive
parking information, and case studies of applications of advanced parking information systems were
examined.

Data Collection

To fill gaps in the literature reviewed regarding applications of API systems at airports, officials from
airports with existing and planned systems were contacted and interviewed on the telephone. The following
questions were proposed to airport officials regarding the airport’s API system.

• What is the capacity of your API system (how many spaces are covered by the API system)?
• What type of parking information do you give to passengers (no. spaces available, lot status, guidance
to alternative lots)?
• How is parking availability data collected?
• How do you disseminate parking information to passengers?
Debra A. Burdette Page 85

Additionally, data regarding user preference in parking information (i.e., information, location, and sources)
were collected using a survey questionnaire. The survey was distributed to random employees at the Texas
Transportation Institute in College Station, Texas. A copy of the survey is provided as Appendix A to this
report.

Data Analysis and Application

Using the information gathered from the literature review, airport interviews, and survey results, guidelines
for implementing an advanced parking information system to provide users with real-time parking availability
information or alternate parking suggestions at airports were developed. The guidelines were formulated
using two methods. Existing strategies used by city-wide systems were adopted, and new strategies to
address air traveler information needs and airport parking characteristics were developed. An enhanced
parking guidance system for Houston’s George Bush Intercontinental Airport was designed using the
guidelines.

BACKGROUND

Airport Parking Facilities

When traveling by air, passengers are often on strict schedules and therefore generally rely on the automobile
to get them to and from the airport as quickly, conveniently, and reliably as possible. This dependence on
the private auto dictates that the airport provide a sufficient supply and proper allocation of parking spaces
in order to ensure convenient access to the airport (2). This section of the paper includes a discussion of the
types of parking available, how parking affects landside access performance, and applications of intelligent
transportation systems technologies at airports as background to illustrate the potential for integrating an API
system.

Types of Airport Parking

Airport parking consists of separate facilities for employees, rental cars, and passengers and visitors, with
the latter broken up into short-term (i.e., hourly parking), long-term (i.e., daily parking), and remote lots (i.e.,
economy or park and ride). Short-term spaces are those located most conveniently to the terminal building
and are available for a premium rate. Long-term parking is also typically located near the main terminal
building, but is less convenient than short-term lots. As a trade-off, long-term parkers are usually given a
discounted rate compared to short-term parkers. Remote parking areas are usually far removed from the
airport terminal complex and are generally served by courtesy vehicles to transport passengers to and from
their vehicles. Again, passengers are compensated for the remote location by receiving the most economical
rate (6). Figure 1 illustrates the three parking area locations relative to Terminals A and C at Raleigh-
Durham International Airport (7).

Parking’s Contribution to Landside Access Performance

The overall goal when designing the access portion of an airport is the safe and efficient movement of
passengers. The design and operation of parking areas significantly affect the performance of all airport
access components. These elements, known as landside facilities include the parking areas, access roadways,
and terminal curbside. Parking area performance is predominately determined as a function of the time
required to find a space. Additionally, the time spent traveling (walking or riding) between a vehicle and the
terminal and the conditions in which the trip takes place influence the level of service provided to passengers
who park at an airport (2, 3).
Debra A. Burdette Page 86

Figure 1. Parking Area Location at Raleigh-Durham International Airport

High turnover rates in short-term lots caused by high hourly rates imposed to discourage long-term parking,
and parking areas operating at near capacity affect the performance of the roadways leading to and
circulating through the airport. As a result, to ease the strain on airport roads caused by circulating vehicles,
the number of required spaces calculated in the design phase of the parking lot is often greater than the
expected demand. Overall, to avoid congestion on access roads and terminal curbsides, public parking that
is conveniently located and provides sufficient capacity or immediate directions to available spaces is
essential (2, 3).

Opportunities for Integrating ITS to Manage Parking Facilities

There are several ways ITS technologies can be utilized by airport authorities as a means of effectively
managing their parking lots. These methods include the use of automatic vehicle identification (AVI) and
automatic vehicle location (AVL) for traveler information systems and traffic management. By installing
detectors in the entrances and exits or in each individual space of a parking lot, and connecting them to a
central detector unit, the use of AVI and AVL concepts to gather information on space availability and
vehicle parking duration is feasible. In the case of parking availability, the information is then distributed
to incoming passengers via variable message signs, highway advisory radio, or telephone information systems
so they can locate an open space as quickly as possible. Parking duration information can be used to
calculate appropriate fees, which can be automatically debited if the vehicle operator has an existing account
with the airport. In addition to parking guidance and automatic fee calculation and collection, ITS
technologies can be effective in detecting illegally parked vehicles, increasing airport parking lot security,
and reducing traffic flow inside the parking lot during congested periods (4). Table 1 includes a partial list
of airports throughout the United States that currently or plan to use ITS to manage parking and ground
transportation operations.

Overview of Advanced Parking Information (API) Systems

The first advanced parking information system was installed in Achen, Germany in the early 1970s making
parking guidance one of the oldest forms of traveler information. Advanced parking information systems
use parking availability data to provide motorists with directions to open lots and vacant spaces in order to
more efficiently operate a parking lot and reduce search times. There are several approaches and strategies
to parking guidance and information dissemination that incorporate both static and dynamic message signs
Debra A. Burdette Page 87

Table 1. Examples of ITS Projects at U. S. Airports in 1993 (4)

Airport ITS Initiative Project Objectives Status

Denver Revenue collection, congestion control & ground

AVI system Operational
International transportation management.

John F. Kennedy AVI system Ground transportation management. Operational

Tracks commercial vehicles in and out of the Fully

Los Angeles
AVI system airport, billing purposes, and to bill certain groups operational
International
(courtsey vehicles) using accurate circuit counts. since 1990

Commercial billing
Orlando Fee collection. Operational
system

Collects commercial roadway fees and provides

Pittsburgh AVI system shuttle buses access to both the public and Operational
employee lots.

The system has 3 transmitters on the access road to Operational

Washington-Dulles HAR system
the airport to provide parking information. since 1983

Washington Traffic information Provides information on parking availability and Operational

National system road construction in the airport. since 1982

to display lot status or the number of free spaces for individual lots or a city-wide system such that drivers
can make effective decisions when choosing lot location, in addition to being guided efficiently within the
facility. With studies showing that up to 33 percent of all drivers circulating within a city center are looking
for parking, API systems have the potential to be an integral part of traffic management in and around the
parking facility, especially under congested conditions (8-11). A discussion on the approaches used in
parking information systems, the strategies used in disseminating information, how an API system functions,
the benefits to implementing a system, and motorists’ response to receiving parking guidance is included in
the following sections.

Approaches to Parking Information Systems

The operation of parking information systems follow the same principles regardless of the overall guidance
objective of the system. Whether guidance is provided on an individual lot or area-wide basis, the system
will take on one of three approaches. These approaches are total space, zone-specific, and space-specific
guidance.

Total Space Guidance. Total space parking guidance systems are currently used by airports, universities, and
medical facilities among others to provide parking information to drivers on a single lot. In a total space
guidance system, the number of vehicles entering and exiting a parking facility are counted and the current
number of spaces available or lot status for that facility only are continuously displayed via a variable
message sign just outside the lot entrance. This system does not provide alternative parking suggestions, as
its purpose is to let patrons know the lot’s status prior to them searching for an available space. Because of
the limited guidance afforded by the system, it is the least expensive to implement (12).

Zone-Specific Guidance. Zone-specific guidance systems function similarly to the total space guidance
system as they also detect all vehicles entering and exiting the parking area. However, in this system, the
parking lot or garage is sub-divided into zones (e.g., levels of a garage or quadrants of large surface lots) and
Debra A. Burdette Page 88

vehicle movements are also detected for each section. As a result, real-time information is available on a
zonal basis in addition to the total facility. This allows the system to be more specific in its guidance as
drivers can be directed to bypass fully-occupied zones, thereby reducing search times. Zone-specific
guidance is especially beneficial in multiple story parking lots with helicon ramps such as downtown and
airport parking garages. In this design, drivers can drive upwards through the facility until they reach the
first section with available spaces as indicated by the API system without traversing parking aisles.
Additionally, large surface lots can also use this system when divided spatially (12).

Space-Specific Guidance. For a space-specific guidance system, detection equipment is installed in each
individual parking bay within the parking area. Therefore, upon entering the parking area, drivers can be
directed to an open bay based on the space occupancy information. Generally, the patron is directed via signs
and/or traffic lights. Because detectors have to be placed in each individual space, this system can be
extremely costly to implement and maintain (12).

Strategies for Information Dissemination

Advanced parking information systems employ a variety of different approaches when disseminating parking
information to drivers. Two main strategies are descriptive and prescriptive approaches.

Descriptive Approach. A descriptive approach to disseminating parking information emphasizes increasing

the parking choices available to the driver. Specifically, these systems attempt to communicate as much
information as possible about the status of the parking system. This information would include a list of all
the parking lots in the system along with directions to each lot. Additionally, the status of each lot would
be given. As a result, the decision-making responsibilities are placed with the individual driver (13).

Prescriptive Approach. The prescriptive approach removes all decision-making responsibility from the driver
and places it solely on the guidance system. Throughout the system, drivers are only relayed the information
deemed necessary to proceed on to the next stage, even if information is withheld (13). For instance, if a
parking area is full, only parking information for lots with similar or cheaper fares would be given as opposed
to information on all lots in the system.

How the API System Works

Advanced parking information systems generally consist of four basic components including:

• the vehicle detectors;

• a parking control center;
• information displays; and
• a telecommunications network (10, 12).

Figure 2 shows a diagram illustrating the elements of a parking guidance system.

Vehicle Detection. The advanced parking information system must have a means of collecting parking space
availability data to serve as input for the guidance system. The method of counting entering and exiting
vehicles has a significant influence on the success of the system. If the data are inaccurate, calculation of
available spaces by subtracting the number of vehicles entering and adding the number of vehicles exiting
to the total spaces available will be erroneous. As a result, drivers will loose confidence in the information
and the whole system will suffer (10, 12, 14).
Debra A. Burdette Page 89

Information Displays

Vehicle
Detectors

Telecommunications
Network

Figure 2. Schematic of an Advanced Parking Information System

(10)

Methods of collecting traffic in-flow and out-flow counts include a wide range of technologies from barrier
contacts, which count vehicles as entry/exit gates are opened, to inductive loops, ultrasonic, infra-red,
microwave, laser, piezo, and machine vision sensor technologies that detect vehicles as they pass over or
through sensors (10, 12). A summary and comparison of some of the main technologies according to Kevin
Gavin of Appian Technologies is in Table 2.

Parking Control Center. The main function of the control center is to receive and edit parking availability
messages generated by data collected from parking facilities, and then transmit them to individual variable
message signs or other means of disseminating information for use by travelers. The hardware and software
required to accomplish these tasks include a central controller with system software, a computer monitor,
and a printer. These elements are available through various companies and are typically tailored to meet the
specific needs of a particular system. Generally, the software automatically generates messages for
information distribution based on the current lot status. However, using a color computer monitor and a
graphical user interface, the system can be manually overridden for special events or unusual circumstances.
Additionally, potentially problematic and non-transmitted parking availability information messages are also
displayed on the monitor for manual intervention (10).
Debra A. Burdette Page 90

Table 2. Comparison of Vehicle Detection Technologies (12)

Method of Vehicle Detection

Performance
Inductive Loop Ultrasonic Infra-red Machine Vision
Criteria

Accuracy 2 6 6 6

Cost 5 2 4 2

Ease of Installation 2 1 5 3

Reliability 2 4 5 4

Lifespan 3 5 5 4

Effectiveness of
2 6 5 6
Guidance Solution

INDEX 16 24 30 25
1 - Very Poor, 2 - Poor, 3 - Adequate, 4 - Good, 5 - Very Good, and 6 - Excellent

Information Displays. Use of effective information displays is critical to the success of the advanced parking
information system as drivers need to feel comfortable accepting the information. Most importantly,
motorists need information regarding the availability of parking. A combination of static and dynamic signs
which provide drivers with parking availability information and directions to parking areas has proven
effective. However, several other technologies have also been identified as means of providing parking
information. These include highway advisory radio, the Internet, telephones, commercial television, and in-
vehicle navigation. Additionally, distributing information on the number of spaces that the driver can expect
to find when they reach the parking lot rather than the current number of spaces available is a recommended
practice so that patrons are never faced with finding no available spaces when the signs indicated that at least
one was free (10, 14).

Telecommunications Network. The telecommunications network is used to link the system together allowing
the transmission of information between the parking facility, the control center, and the information displays
by coding data into bit patterns. Typically, exchange of information takes place via one of two types of
communications networks. The first option, dedicated networks consisting of specially-laid cables and
infrastructure, is the more expensive alternative. Existing public networks like city telephone or signaling
equipment control cables can also be used; however, limits on the number data channels available or on the
quality of the transmitted signal sometimes dictate the use of a dedicated system (13, 15).

Benefits of API Systems

While parking information systems cannot create parking spaces, they have the potential to positively impact
individual and city-wide parking areas, providing benefits to operators, users, and the surface transportation
system. First, use of parking guidance can increase the efficiency or utilization of the parking facility as the
distribution of accurate availability information results in empty spaces being filled faster, and thus higher
occupancy rates. This increased utilization and higher occupancy rates, as well as quicker turn around times,
directly translate into increased revenue for the parking operator. Additionally, zone and space-specific
parking information systems provide parking operators with detailed management information such as
occupancy, usage patterns, and throughput, which can be used in designing effective payment structures and
Debra A. Burdette Page 91

in checking the fraud detection process. Finally, operators can cut costs since manpower requirements are
reduced in parking areas with an API system (8, 12).

Parking patrons benefit from a parking information system, since guiding them directly to an open facility
reduces search time, queues, delay at the entrance, and ultimately driver frustration. When drivers are not
forced to circle or wait for a parking space, they are more satisfied with the service and are likely to be repeat
customers (8, 12, 16).

Installation of an advanced parking information system benefits the surrounding area by improving traffic
flow and reducing congestion in and near the parking area. These enhanced conditions along with shorter
search times contribute to a reduction in fuel emissions as car engines are running for less time (8, 12).

User Response to API Systems

The effectiveness and performance of the advanced parking information system as a traffic management tool
is dependent on voluntary compliance by motorists. If a significant number of drivers do not follow the
suggested guidance, the system will fail to serve its intended purpose. This premise is different from other
forms of traffic control where drivers are required to conform (11, 13). Recently, several surveys have been
carried out to discern the nature of drivers’ awareness and use of advanced parking information systems as
well as the system’s impact on parking search behavior. A discussion of the results from three of these
investigations is in the following sections.

Driver Awareness and Use of API. The first study, by Polak, et al., was conducted in Nottingham, England.
After two years of using local radio stations to broadcast real-time parking information during Christmas
time, drivers parking in public lots were surveyed to determine their awareness and use of the parking
information service. Through the investigation, it was found that of the 1548 respondents, approximately
two-thirds of drivers were aware of the system, while 25 percent of the respondents who knew of the system,
had not used it (5). The second investigation was performed in Kingston-upon-Thames, an important
suburban shopping center near London. Results from a survey of 180 shoppers who had parked at public off-
street parking lots showed that there was a high degree of awareness (only 4 percent were unaware) of the
system, which monitors occupancy levels at eight parking lots. However, of those that were aware of the
system, almost half had never used it. Of the remaining respondents who used the system, 30 percent had
used the signs recently, and 19 percent of them had used the system on the interview day (11). The final
study was conducted in Frankfurt, Germany. Before and after surveys were completed by downtown
shoppers who used on and off-street parking on Tuesdays and Saturdays in order to determine drivers’
exposure to and use of the system. After three months of operation, 80 percent of the 3,506 participants
(3,025 off-street, 481 on-street) were aware of the parking information system. The results also indicated
that about 50 percent of off-street and 25-percent of on-street parkers have used the parking information at
least once (17).

API’s Impact on Search Behavior. Advanced parking information systems have the potential to establish or
modify the method in which drivers search for desired parking locations. A survey of drivers parking on
Saturdays in Frankfurt’s shopping district revealed that only a couple of drivers completely depend on the
parking guidance information at first. Rather, they use it primarily as a back up plan to find an open parking
space if their first choice is not available. The results also showed that motorist are skeptical of giving
complete control to the system since the number of drivers indicating they use the system as a fallback plan
is about the same as those who simply parked illegally. Overall, very few respondents actually rely on the
system for guidance. Most users only check or update previous decisions using the system (17).
Debra A. Burdette Page 92

Potential ITS Technologies for Disseminating Advanced Parking Information

Intelligent transportation systems have the potential to greatly enhance the operation of the surface
transportation system through dissemination of accurate, real-time information. When informed mode, route,
and time of travel decisions are made, the entire system benefits through improved safety, mobility, and
traveler convenience. Several technologies, which are currently used in distributing traveler information to
commuters, have been identified as possible mediums for disseminating parking availability and guidance
information to air travelers. These technologies are included in the following list:

• Variable message signs;

• Static signs;
• Highway advisory radio;
• Internet service;
• Telephone information system;
• In-vehicle navigation systems; and
• Commercial television broadcasts.

Variable Message Signs (VMS)

Variable message signs are increasingly used as traffic control devices on freeways and urban roads to
provide information such as alerting motorists of road closures, changes in traffic conditions, and lane
closures, along with other traveler information. Although they are used to display information on daily
occurrences and scheduled events, their primary function is to alert motorists to unpredicted, unscheduled
events in an effort to help reduce speed, avoid congestion and accidents, choose alternate roues, and lessen
frustration. Messages displayed on VMS are activated based on the existing road conditions and can be
controlled on-site or remotely. There are numerous preprogrammed messages to choose from or operators
can tailor them to meet specific needs. VMS are among the most effective ways to disseminate traveler
information because they provide motorists traffic information about the very roads they are traveling on,
right on the highway (8, 18, 19, 30). Figure 3 illustrates the contents of a variable message sign providing
accident information in Atlanta, Georgia.

Figure 3. Example of Variable Message Signing

Static Signs

Static road signs are usually implemented to provide navigational information to travelers. When utilized
correctly, their legends follow specifications outlined in the Manual on Uniform Traffic Control Devices
(MUTCD) and are placed well in advance of key decision points to allow drivers to detect, respond, and
decide on appropriate actions. Additionally, color coding for specific facilities or directions is recommended
(2). The following figure depicts a static directional sign showing air passengers the appropriate route to
parking facilities at Ronald Regan National Airport near Washington, D.C.
Debra A. Burdette Page 93

Figure 4. Static Signing Directing Travelers to Parking Areas

Highway Advisory Radio (HAR)

Highway advisory radio is a means of disseminating traveler information and advisories to motorists using
low-power AM radio signals. Recorded messages, dispatched and monitored by radio personnel, are
transmitted via fixed or portable field antennas and picked up on dedicated AM radio stations. The
information can be accessed free of charge by anyone who is equipped with a radio and is within the
frequency range. Static or variable message signs displaying the HAR broadcast band and mounted flashing
lights are used to alert motorists of message broadcasting. Currently, highway advisory radio is used to
distribute information on traffic conditions, roadway construction, incidents, and airport parking. However,
it can also be used for disseminating travel restrictions, event information, directions, and safety information
(21-23).

Internet Service

With the number of computers with access to the Internet having grown exponentially since 1993, the use
of the Internet as an information source has quickly become wide-spread. As a result, providing real-time
traveler information to a wide variety of people, at any time, for a minimal cost (users pay for their service
provider subscription) can ideally be accomplished using the Internet. Interested travelers only need a
computer, a modem, and Internet service to gain access. Currently, travelers have 24 hour access to weather
forecasts, airport and flight status information, public transit schedules, fares, and routes through a user-
friendly graphical user interface. Additionally, a number of state and local transportation agencies have
developed web sites in order to distribute real-time traffic information allowing commuters to view graphics
and pictures of traffic conditions, which can in turn be used for trip planning. Along with public agencies,
independent information service providers such as Microsoft and SmartRoute are also getting involved with
traveler information dissemination by setting up traveler information web sites (24-27).

Telephone Information Systems

Telephone information systems have been used to provide travelers with traveler information for some time.
This generally inexpensive alternative is low-tech, requiring only a telephone and an easy to remember access
number (e.g., 817-1717 for the San Francisco Bay Area). These systems offer several advantages over the
traditional radio and television broadcasts as callers can obtain information via a recorded voice message for
a particular region, route or location on demand rather than waiting for the next broadcast. Although the
Debra A. Burdette Page 94

system is easy to use and only requires that the user have a touch-tone telephone, travelers generally consider
picking up the phone and placing a call burdensome compared to alternatives that require little to no effort.
Additionally, travelers are sometimes worried that the information they receive will be out of date by the time
they begin the trip; however, this is less of a concern with the popularity of cellular phones. Currently,
traveler information accessed through telephone information systems is free. Most advertising and money
making schemes have not worked because of the notion that comparable information is distributed for free
over the radio or television (28-30).

In-vehicle Navigation Systems

In-vehicle navigation systems include onboard computers that are used by drivers to obtain driving directions
and point-of-interest information visually or audibly while en route. The system consists of a small
computer, route selection software, and map-matching software or global positioning satellites to determine
the location of the vehicle. Additionally, some systems are able to receive FM subcarrier broadcast signals
allowing travelers to access real-time traffic and congestion information. Because of the amount of
equipment involved in operation of the system, installation typically costs between $2,000 and $3,000.
Travelers in the United States have been slow to embrace these navigation systems, but the number installed
in vehicles continues to rise, thereby increasing the potential for a reduction in the system’s price. Overall,
in-vehicle navigation systems are particularly helpful in unfamiliar areas as they increase driver safety and
decrease travel time and anxiety caused by the driving task (27, 31).

Commercial Television Broadcasts

Commercial television is used prior to and during regular commute hours to broadcast limited amounts of
current traffic conditions on an area-wide basis. The information distributed during these broadcasts come
from private traffic reporting companies who use state-of-the-art technologies, aerial surveillance, incident
and construction reports, and reports from cellular callers among other means to gather traveler information.
This information is then processed into abbreviated traffic reports for dissemination. Because TV air time
is expensive, the actual minutes dedicated to traffic information is limited, therefore; very little information
can be given and follow-up reports are rarely made, thus affecting the effectiveness of the information.
However, commercial television is still a popular means of receiving traveler information because it requires
no additional effort, and like HAR and telephone information systems, access is very economical (you just
have to pay for the power for the television) (28, 32).

CURRENT APPLICATIONS OF API SYSTEMS AT AIRPORTS

Several airports throughout the United States and Europe have implemented advanced parking information
systems in an effort to manage their parking lots, provide relief to crowded access and recirculating
roadways, and expedite passenger trips to and from the airport. A discussion of four U.S. airports with
existing or planned systems is included in this section.

Raleigh-Durham International Airport

The Raleigh-Durham International Airport (RDU) serves the Research Triangle region of North Carolina.
In 1998, the 24 major and regional carriers offering more than 500 daily flights transported approximately
7.2 million passengers. To accommodate originating passengers, which makes up most of the passenger
traffic mix, the airport has more than 10,000 parking spaces in one short-term, one long-term, and four
remote park and ride lots. However, due to the construction of a new parking garage due to open in 1999,
parking near the terminals is limited (7).
Debra A. Burdette Page 95

The planned advanced parking information system at RDU will be used to direct passengers to available
parking in the first of a four-phase terminal garage plan, which includes 2,700 spaces on 5 levels. The system
will consist of two components, a message board placed at the entrance to the garage and variable message
signs which are currently located on the main inbound access roadways. The message board will indicate
parking space availability (within 5 to 10 percent) for each level based on data gathered using counters and
loops embedded in the pavement. As passengers travel up helical ramps, the same information is distributed
for the specific level and a gate arm will be lowered, indicating the level is closed if no spaces are available.
The variable message signs, which are operated manually, display lot status (e.g., “FULL” or “OPEN”) and
suggest lower cost alternatives if a lot is full. The parking manager monitors the parking facilities and
changes messages using pre-programmed alternatives.==Eventually the airport operators want to integrate the
access roadway signs with the parking garage signing so that all data collection and parking guidance are
done automatically (33).

George Bush Intercontinental Airport (Houston, TX)

George Bush Intercontinental Airport (IAH), located in north Houston, is classified by the Federal Aviation
Administration as the 12th busiest commercial airport in the United States. In 1998, the 20 airlines flying to
approximately 150 destinations throughout the world served nearly 31 million passengers. Currently, the
airport has over 19,000 public parking spaces and ample room to expand depending on customer demand
(34).

Travelers approaching IAH can obtain airport-wide parking information by tuning their radio to 530 AM or
from an advanced parking information system currently installed in the East and West garages at Terminal
C to assist drivers in traversing the 6,500 spaces, six levels, and seven and a half miles of parking aisles.
Drivers are alerted to parking availability upon entering the garage in an attempt to assist drivers in finding
an available space without repeated circling of the lot and access roadways. A level-by-level sign displays
“CLOSED” (level is within 20 vehicles of capacity) and “OPEN” backlit in red and green respectively,
depending on the level status. The same system is available for each level, and a gate arm comes down
prohibiting entrance into the parking area if no spaces are available. Additionally, the east or west sections
of a level can be closed off to help passengers navigate the garage. However, since travel is not permitted
between floors, passenger have to exit the garage via payment booths and begin the process over again if they
leave a open floor in an attempt to park in another area. Data for the system are gathered using wire sensors
embedded in the garage floor to count cars entering and exiting each parking level. This information is used
by the central computer to update the information signs. A 5 percent buffer is used in calculating the number
of spaces available to account for handicapped and double-parked vehicles. A similar system will be
implemented in the upcoming AB parking garage, which is expected to be complete in 2001 (34-36).

Metropolitan Washington Airports (Ronald Reagan National and Washington Dulles)

The Metropolitan Washington Airport Authority maintains and operates Ronald Reagan National (DCA) and
Dulles (IAD) airports, which are located just south and west, respectively, of Washington, D.C. Last year,
Reagan National Airport served nearly 16 million passengers. To accommodate the 50,000 vehicles that
travel on the airport roadway, DCA has parking for nearly 7,200 vehicles. On the other hand, Dulles
transported almost 15 million passengers in 1998 and provides parking for 19,000 vehicles (37).

Both airports have existing parking information systems that rely on telephone information systems, highway
advisory radio and message signs to disseminate parking guidance. In total, the parking guidance system
covers approximately 7,500 spaces including hourly, daily, and remote facilities. At the entrance to each
facility and on each level of one of the two parking garages, signs are used to distribute lot status based on
availability information collected by counters in lot entrances and exits, and inductive loops embedded in
helical ramps. Figure 5 illustrates the use of signing to distribute parking availability information. In
addition` to availability information, alternative lot suggestions are displayed by way of neon/fiber optic
signs on the exterior of the parking areas. Finally, traveler’s may soon be able to access all of this
Debra A. Burdette Page 96

information from the Internet, as the operators are working towards integrating their Computerized Parking
Control System (CPCS) screen with the Authority’s web site (38).

Figure 5. Parking Lot Status Information Dissemination via Message

Signs

Summary

This section has provided several examples of airports with an advanced parking information system planned
or in place. The details for each system including airport parking and airline passenger statistics, and API
system characteristics are in Table 3.

Table 3. Summary of API Systems at U.S. Airports

1998 Enplaned Total Method of Type of

Parking Spaces
Airport Passengers Parking Vehicle Information
Covered by API
(millions) Spaces Detection Displayed

RDU 7.2 10,000 2,700 Counters/Loops Level-by-level

spaces available

IAH 31 19,000 6,500 Loops Level-by-level

garage status

DCA 16 7,200 Level-by-level

7,200 Loops status at 1 garage
IAD 15 19,000
Debra A. Burdette Page 97

SUMMARY OF SURVEY RESULTS

The success of advanced parking information systems is dependent on the motorists’ awareness and use of
the information. However, usage of the system is directly related to whether or not drivers receive
information they feel is relevant to their decision and how conveniently the information can be obtained.
In order to determine the type of parking information drivers need to make parking decisions, their preferred
location to receive the information, and the method of receiving the information a survey of 30 Texas
Transportation Institute employees who have used parking facilities at various airports throughout the United
States was conducted. The results from the user needs survey along with any additional comments provided
by respondents are presented in this section. A tabulation of survey responses are included as Appendix B.

Parking Information Desired

Results from the investigation showed that passengers are especially concerned with parking availability and
directions to lots when making parking decisions at airports as both of these categories received a rank of
one (highly valuable) or two (valuable) from all of the respondents. On the other hand, parking cost and
alternative lot suggestions were perceived to have little value by 40 percent and 30 percent, respectively of
participants. Other parking information needs receiving high value rankings include method of payment,
shuttle times between parking areas and terminal buildings, and any vehicle clearance requirements. Figure
6 illustrates the results of the desired parking information survey question.

These findings appear to be reasonable as they reflect the types of information currently distributed by
parking guidance systems that are in place.

Preferred Location to Receive Parking Information

The results from the investigation of preferred location to receive parking information are found in Figure
7. It can be seen from the figure that even though participants could select more than one option, the
majority of passengers want to obtain information upon entering the airport property. In fact, 90 percent of
the participants responded positively to this location.

80.0

1 - Very Valu able

70.0 2 - Valuable
3 - Som ew h at Valua ble
60.0 4 - N ot Valua ble
N /A

50.0

40.0

30.0

20.0

10.0

0.0
P arking Co st P arking Directio ns to Distance/Travel A lternative Other
A vailability Lo ts Time to Suggestio ns
Terminal

Figure 6. Desired Parking Information Results (n = 30)

Debra A. Burdette Page 98

100.0

90.0 Upon entering airport grounds

Just before the parking area
80.0
Before leaving for the airport
70.0
Proportion of Responses

60.0

50.0

40.0

30.0

20.0

10.0

0.0

Figure 7. Preferred Information Dissemination Location Results (n = 30)

These results are surprising since previous studies have indicated that travelers tend to feel that information
given too early will be stale by the time they need it to make decisions. However, the findings may reflect
passengers’ desire to obtain guidance information at a location where there is still sufficient time to make
an informed decision.

Desired Method of Information Transmission

Figure 8 contains the results from the portion of the study to determine passengers’ desired methods for
receiving parking information. These findings show that a combination of variable message and static
signing would result in the most preferred system. However, radio broadcasts via highway advisory radio
and in-vehicle navigation systems may be feasible technologies.

Similar to the parking information desired, the findings regarding the preferred method to receive information
seem reasonable since these are the two technologies currently used in advanced parking information
systems.

ASSESSMENT OF POTENTIAL ITS TECHNOLOGIES USED IN API SYSTEMS

Seven technologies currently used in providing traffic information were identified as possible methods to
disseminate parking information to air travelers. These technologies were researched and evaluated based
on the following criteria:

1. Perceived cost of technology to user;

2. What is the timeliness of the information received by the technology; and
3. Does the technology meet the needs of airport users as identified in the survey?
Debra A. Burdette Page 99

100.0
Variable message sign
90.0
Static signing
Radio/In-vehicle
80.0
Internet
70.0 800-telephone number
Television
60.0

50.0

40.0

30.0

20.0

10.0

0.0

Figure 8. Desired Method for Information Dissemination Results (n = 30)

Table 4 shows a summary of the assessment results. By examining the table, it can be concluded that use of
television broadcasts, an 800-telephone number and the Internet to distribute parking guidance information
would not promote use of the system. First, each technology reported low acceptance rates in the user
survey. Additionally, they all provide pre-trip information, which has the potential to be out dated by the
time the passenger gets to the parking lot entrance. Finally, in the case of the Internet, there is some level
of cost associated with obtaining the information. A better system design would include technologies that
minimize user cost and the possibility of outdated information, yet have a high degree of user acceptance
such as VMS, static signs, and highway advisory radio.

Table 4. Parking Information System Technology Assessment Results

Technology Cost to User Timeliness of % Survey Participants

Information Willing to Use

VMS None En-route 100

Static Signs None En-route 56.7

Highway Advisory Radio None En-route 30

In-vehicle Navigation High En-route 30

Internet Moderate Pre-trip 23.3

800-Telephone Number None Pre-trip/En-route 10

Television None Pre-trip 0

The use of a cellular phone to access an 800-telephone number logically appears to be an acceptable way to
receive parking information; however, the 800-telephone number received a low percent willingness to use
Debra A. Burdette Page 100

by survey participants. This could be attributable to the participant misunderstanding of the survey question,
as they thought the telephone information system could only be accessed at home before beginning the trip.
Also, a toll-free number could be too long to easily remember and therefore accessing the system could
appear burdensome.

GUIDELINES FOR IMPLEMENTING API SYSTEMS AT AIRPORTS

Guidelines to be used by airport parking operators when implementing an advanced parking information
system at the airport are developed in this section. These guidelines are designed to ensure that the airport
parking system will adequately meet the information needs of air travelers, while maintaining efficient
management of the parking facilities. The formulation of these guidelines was accomplished by adopting
current strategies used in city-wide parking guidance systems, and adding new strategies to address air
traveler information needs and airport parking characteristics identified in this investigation. Table 5
includes a list of the nine guidelines, followed by a description of each guideline.

Table 5. Guidelines for Implementing an API System at Airports

Guidelines

Step 1: Determine which parking facilities are to be included in the advanced parking information system.

Step 2: Identify the parking guidance approach.

Step 3: Divide the global system into zones.

Step 4: Select the information dissemination strategy.

Step 5: Specify system characteristics.

Step 6: Determine the maximum capacity of the parking facilities.

Step 7: Develop and test system software.

Step 8: Deploy the API system

Step 9: Evaluate the API system

Step 1: Determine which parking facilities are to be included in the API system.

A first step in introducing an advanced parking information system to an airport is to decide which parking
areas should be covered by the dynamic system. It was shown in Section 3 that airports with multiple
parking areas typically have short-term, long-term, and remote lots. An effective API system must provide
guidance to as many passengers possible; however, depending on the lot usage and the availability of funds,
certain areas may be eliminated from the system altogether or implemented in later phases of a project.

Step 2: Identify the parking guidance approach.

Three potential parking guidance approaches including total-space, zone-specific, and space-specific
guidance were introduced in Section 4. Total-space guidance provides the broadest information and is the
most economical, while space-specific guidance is the most detailed and can be expensive. Airport operators
Debra A. Burdette Page 101

can choose the appropriate guidance approach depending on the airport’s budget and desired level of
information to be given to passengers.

Step 3: Divide the global system into zones.

The use of parking zones in an overall parking information system facilitates the movement of vehicles by
allowing for more specific guidance without overloading travelers with information. Additionally, search
times can be reduced as passengers are directed to by-pass entire zones that are fully occupied. Once the
boundaries of the system are specified and the guidance approach is selected, the areas should be divided into
zones depending on the layout of the airport and parking lot. Guidance systems utilizing the total space
strategy will require that zones are specified based on the location and the function of the parking area, most
likely following the hourly, daily, and remote designations as well as the division by terminal area (e.g.,
Terminal A and hourly parking). A system built on zone-specific or space-specific guidance will require that
zones be established based on lot location and function, and then further divided by level for multi-level
garages or area for large surface lots.

Step 4: Select the information dissemination strategy.

Descriptive and prescriptive methods are two broad strategies used in information dissemination. Descriptive
parking information systems provide drivers with as much information concerning the parking areas as
possible. As a result, these systems increase the choices for drivers and allow them to make decisions. On
the other hand, prescriptive parking guidance attempts to limit drivers’ opportunity to exercise choice,
thereby leaving decisions up to the system itself. This is accomplished by limiting the information
distributed to motorists to include only what is necessary to proceed to the next decision point. A descriptive
or prescriptive approach can be selected by airport operators based on their desired level of responsibility
in guiding passengers to available parking and the information requested by drivers.

Step 5: Specify system characteristics.

Based on the results from Steps 1 through 4, technologies to be used to distribute parking information and
the location of information dissemination should be specified. Section 5 included a detailed summary of
seven potential technologies to be used in disseminating parking information. Possible technologies include,
but are not limited to highway advisory radio, the Internet, telephone information systems, television
broadcasts, variable message signs, static signs, and in-vehicle navigation systems. When choosing system
technologies, airport parking managers should select alternatives that address the needs of the airport, as well
as reflect preferences of airport users.

Location of information dissemination depends on the display method used. For instance, information
gathered using the Internet or the television will only be available before the trip (i.e., at home or in the
office); whereas, highway advisory radio, variable message signs, and static signing can be used to deliver
parking information en-route. Airport operators should also be careful to ensure the information is
distributed when air travelers feel it is helpful and in advance of critical decision points. Examples of critical
decision points include locations where roadways diverge toward different terminals and entrances to parking
facilities.

Step 6: Determine the maximum capacity of the parking facilities.

Because the successful operation of a parking guidance system depends on its use by air passengers wanting
to park, airports need to gain passenger acceptance of the system. Two of the quickest ways to loose
passenger confidence are to direct them to parking areas that are fully occupied and to make them by-pass
areas that clearly have available spaces. Therefore, a cut-off capacity needs to be established so that the
system knows when to close and re-open lots. To account for the difficulty associated in finding the one or
two last spaces in a lot, vehicles that are double parked, and handicapped spaces, airport parking facilities
Debra A. Burdette Page 102

typically operate at practical capacity, between 85 to 95 percent occupancy, during peak periods. For parking
guidance purposes, it is recommended that the maximum capacity of a parking lot be set to the total number
of spaces in the lot or garage less the average number of vehicles that are expected to arrive in the time it
would take a driver to reach the parking lot from outermost airport access point. Therefore, if 5 new vehicles
enter a specific lot every 15 minutes, and if the drive time from the airport entrance to that lot is 15 minutes,
the maximum capacity of a 1000 space lot is 995 spaces. Although this seems to be a relatively small margin
of error for finding an available space, the system uses arrival rates rather than turnover rates, which account
for vehicles exiting the parking lot. Any vehicle exiting the parking lot will only increase the number of
spaces available.

Step 7: Develop and test system software.

The requirements specified in Steps 1 through 6 will dictate the development of the software used to operate
the system. The software will use data acquired from vehicle detection equipment to formulate parking
availability information that is then transmitted to dissemination technologies. Since each system is unique,
operators of airport parking information systems will have to work with software designers to ensure it is
tailored to reflect the requirements of their specific system.

Step 8: Deploy the API System

Once the system approach and characteristics have been specified, the advanced parking information system
should be deployed for use by drivers intending to park at the airport. Ideally, the entire system would be
implemented at one time; however, staged deployment may be necessary due to financial constraints.

Step 9: Evaluate the API System

To test the effectiveness of the API system, it should be evaluated to determine user awareness of and
compliance with the parking information. The change in parking lot occupancy rates and stated preference
behavior surveys are two ways to measure the effect of an advanced parking information system. The first
method simply monitors an increase or decrease in parking lot utilization to infer the effect of API. Stated
preference survey data are collected using survey questionnaires. For the state preference survey,
respondents are asked what effect the API system had on their parking choice behavior. Additionally, the
survey could be used to check driver awareness of the system or inquire about potential modifications to the
system in order to increase its use.

CASE STUDY: GEORGE BUSH INTERCONTINENTAL AIRPORT

This section presents the application of the guidelines developed to illustrate the implementation of an
advanced parking formation system that effectively provides guidance to air passengers in making parking
decisions. The guidelines are applied to Houston’s George Bush Intercontinental Airport as a hypothetical
extension to the current parking guidance system available at Terminal C. Additionally, for the purpose of
this demonstration, it is assumed that the results from the user needs survey described in Section 6 and the
findings from the evaluation of the potential ITS technologies found in Section 7 are representative of the
Houston, Texas, population.

The following figure contains a diagram illustrating the parking lot layout in relationship to the passenger
terminals at IAH. There are a total of seven lots available to air passengers. Six of them, Terminal A hourly
and daily lots, Terminal B hourly and daily lots, Terminal C hourly and daily garage, and Terminal AB daily
lot are located near their respective terminal building, while the city lot (not shown if the figure) located at
the intersection of JFK Boulevard and Greens Road serves as the remote economy lot.
Debra A. Burdette Page 103

Terminal A Terminal B Terminal C Terminal IAB

Hotel

H/DP DP H/DP DP H/DP H/DP DP

Airline Terminal Assignments

JFK
Blvd
Remote Lot

Greens Road

HP: Hourly Parking DP: Daily Parking

Figure 9. Parking Layout at George Bush Intercontinental Airport

Step 1: Determine which parking facilities are to be included in the API system.

Intercontinental Airport currently has a parking guidance system in place for the Terminal C garage, but has
the potential to increase the system to include all seven of the airport parking lots. Although the airport
currently provides parking for 19,000 vehicles, the installation of a parking guidance system to cover all the
lots is not as intimidating at it first seems since 6,500 of the spaces fall under the control of the existing
system. Additionally, the remote lot is not operated by the airport; therefore, the consent and cooperation
of the city of Houston is required if this lot is to be incorporated into the system. Ideally, the parking
information system would include the remote lot, as it is commonly used as an alternative parking area
suggestion. Therefore, assuming the funds are available and full cooperation is given by the city, the
expanded advanced parking information system will include all seven lots.

Step 2: Identify the parking guidance approach.

The parking guidance system currently in place in the Terminal C parking garages utilizes a zone-specific
approach to provide parking information. The zone-specific strategy is an economical alternative that will
allow drivers to reduce parking space search times by bypassing full parking areas. In keeping with the
strategies currently in use so as to maintain a homogeneous system and because of the large number of
parking facilities and spaces covered by the API system, a zone-specific approach is selected for the airport-
wide parking information system.

Step 3: Divide the global system into zones.

Now that it has been established that all seven parking facilities will be included in the parking guidance
system and that a zonal approach will be used, the facilities must be broken up into zones in order to provide
navigational and availability information. First, the parking facilities are already divided by location and
directions are provided through static signing. These zones include Terminals A and B, Terminals C and
IAB, and the city lot. The first two zones will be further partitioned into daily and hourly facilities. Finally,
if the parking area is a garage structure with helical ramps, the facility will be zoned by level.
Debra A. Burdette Page 104

Step 4: Select the information dissemination strategy.

Similar to choosing a parking guidance approach, the information dissemination strategy will be selected so
that a homogeneous system is maintained. Therefore, a descriptive strategy was selected since the dynamic
information signs located at the entrance to the Terminal C garage report parking availability for each level,
allowing the driver to decide on which level to search for a vacant parking space. This descriptive approach
will be used system-wide in disseminating overall lot status in addition to level by level availability for
parking garages.

Step 5: Specify system characteristics.

Based on the assessment of the seven technologies identified as potential methods of disseminating parking
guidance information, a combination of variable message signs to provide dynamic parking information and
static signs to help drivers navigate the airport access roads was determined to be the most effective
information distribution system. This decision is supported by the fact that the selected alternatives are
already in use in the existing Terminal C garage system.

The type of information displayed by the signing system depends on the utilization of the facility. This
should be kept in mind when designing the information dissemination system. Since people generally want
to park as close to their destination as possible, the hourly and daily lots will be used more often than the
remote facility. Therefore, signs on the access roadways directing passengers to parking at Terminals A and
B, Terminals C and IAB, and remote lots should include static navigational guidance, as well as dynamic
information (e.g., lot status or parking availability) for all zones except the remote lot since the remote lot
is rarely filled to capacity. As the driver approaches the individual facilities within the zone, static signing
should provide directional guidance, while variable message signs continue to alert passenger to the status
of the parking area. Finally, when entering the facility, variable message signs at garage structures should
display level by level parking information. At surface lots, the VMSs will simply indicate availability
information for the entire lot.

In an effort to improve the existing system, the variable message signs on access roadways will only display
the lot status as open or closed using backlit green and red letters, respectively, while as the driver
approaches the parking area, information on the approximate number of spaces will be distributed .
Additionally, a suggestion for alternative facilities will be given as part of the dynamic message when the
parking area is closed. Figure 10 illustrates a possible layout and design of variable message signs to provide
drivers with parking information and guidance to Terminal C at George Bush Intercontinental Airport.

Step 6: Determine the maximum capacity of the parking facilities.

As suggested in the guidelines, the maximum capacity was set as a function of the number of vehicles
arriving during the time it takes a vehicle driving the speed limit to travel from the furthest variable message
sign to the parking lot entrance. To determine the arrival rate of vehicles at the various parking lots, vehicle
counts were performed at each facility entrance and the 95 percentile arrival rate was used to account for
double parking and handicapped spaces. In this system each parking area will have a different maximum
capacity, which will have to be factored into the API system software.

Step 7: Develop and test system software.

Software incorporating the requirements specified in the first 6 steps will be developed and tested to ensure
accurate availability information will be displayed. The software will collect input data from the lots
providing dynamic information, process the availability information, and then transmit it to variable message
signs in order to provide drivers with parking guidance on a zonal basis in a descriptive manner.
Debra A. Burdette Page 105

Variable and Static Message Signs

Parking Lot Spaces
C - HOURLY CLOSED
HP: Hourly Parking DP: Daily Parking
C - DAILY 690
USE REMOTE LOT FOR
ADDITIONAL PARKING

Terminal A Terminal B Terminal C

Hotel Terminal IAB

H/DP DP H/DP DP H/DP H/DP DP

Parking Lot Spaces

Parking Lot Spaces LEVEL 1 CLOSED
C - HOURLY CLOSED Airline LEVEL 2 CLOSED
C - DAILY OPEN Terminal LEVEL 3 54
IAB - HOURLY OPEN Assignments LEVEL 4 141
IAB - DAILY OPEN JFK LEVEL 5 285
USE REMOTE LOT FOR Blvd LEVEL 6 215
ADDITIONAL PARKING USE REMOTE LOT FOR
ADDITIONAL PARKING

Figure 10. Example of API System at IAH

Step 8: Deploy the API System

An initial survey to determine the parking information needs of motorists intending to park at Intercontinental
Airport served as the basis for the development of the API system; however, the extent that the system will
actually be used has yet to be determined. Therefore, implementation of the system will be completed and
evaluated in phases in order to reduce the initial capital cost associated with deploying the system. In the
first stage, an API system serving Terminals C and IAB will be constructed. The second stage will
incorporate Terminals A and B into the overall system.

Step 9: Evaluate the API System

Evaluation of the first phase of the system will be completed prior to the commencement of the second phase
in order to evaluate its effectiveness in reducing parking search time and congestion on access roadways and
at the terminal curbfront. The effectiveness will be measured using stated preference surveys that are
distributed in the parking areas (most likely on the windshields of parked vehicles) and mailed back to the
airport authority for analysis. Based on the results of the survey, modifications to the system will be made
in order improve it or the second phase will be deployed such that all drivers will have the benefit of
advanced parking information.

CONCLUSIONS

Airport landside access facilities are beginning to experience more congestion as the popularity of air travel
has and continues to rise. Passengers using their personal automobile as a means of conveniently accessing
the airport are faced with longer delays in finding available spaces, thereby contributing to the number of
vehicles on airport roadways as they circle trying to find an open parking bay. Utilization of advanced
parking information systems to provide passengers with accessible and accurate parking availability and
routing information at critical decision points results in a decrease in parking search time, reduces roadway
and parking facility congestion, and encourages efficient use of parking areas.
Debra A. Burdette Page 106

When designing parking guidance information systems at airports, it is important to consider the needs of
the patrons using the information as well as requirements specified by the operators of the system since
voluntary compliance directly affects the success of the system. Results from a survey conducted as p
reduces art of this research showed that drivers intending to use airport parking facilities feel that parking
availability and directions to open lots are the most valuable pieces of information when choosing where to
park. Additionally, the survey concluded that a combination of variable message signs, static signing, and
possibly highway advisory radio are the preferred methods to receive the desired information. Finally, it was
determined that air travelers would like to receive parking information when they enter the airport grounds.

The guidelines developed for implementing an advanced parking information system provide a means for
airport operators to install a system that effectively manages parking facilities in an efficient manner while
getting passengers into and out of the airport as quickly as possible. The guidelines were applied to
Houston’s George Bush Intercontinental Airport in order to illustrate the implementation process, which can
be used by any airport with parking congestion.

ACKNOWLEDGMENTS

This paper was prepared for Advanced Surface Transportation Systems, a graduate course in Civil
Engineering at Texas A&M University. The author would like to express her gratitude to her professional
mentor, Doug Robertson, for his assistance and guidance in conducting this research and preparing the report.
A sincere thanks also goes out to all the other mentors Tom Hicks, Pat Irwin, Wayne Kittelson, Bill Spreitzer,
and Carol Zimmerman, who took the time out of their busy schedules to participate in this program and share
some of their professional expertise with the students. Special thanks goes to Dr. Conrad Dudek for the
opportunity to participate in the unique program and for providing the author opportunity to work with a
diverse group of transportation professionals. Finally, the author would like to acknowledge the following
individuals and previous Advanced Institute participants who were instrumental in conducting this research:

Lee Anne Dixon, Kimley-Horn and Associates, Inc.

David Hathaway, JKH Mobility Services
Peter Mandel, Leigh Fisher Associates
Kelly D. Parma, Lee Engineering
Greg Parzych, TCS Incorporated
Tom Quisenbury, Raleigh-Durham International Airport
Bridgette D. Shamburger, Kimley-Horn and Associates, Inc.
Ron Stehman, HNTB Corporation
Mark Yedinak, Metropolitan Washington Airport Authority

REFERENCES

1. Polydoropoulou, A. and M. Ben-Akiva. The Effect of Advanced Traveler Information Systems (ATIS)
on Travelers Behavior. Prepared for 77th Annual Transportation Research Board Meeting, Washington,
D.C., January 1998.

2. Bellomo-McGee, Inc. Intermodal Ground Access to Airports: A Planning Guide. Project No. DOT-T-97-
15, prepared for the Federal Highway Administration and the Federal Aviation Administration, December
1996.

3. Hoel, L. A., and H. W. Shriner. Evaluating Improvements in Landside Access for Airports. Virginia
Transportation Research Council, Virginia Department of Transportation and the Federal Highway
Administration, Richmond, Virginia, 1998.
Debra A. Burdette Page 107

4. Al-Ayat, R. A., P. Hung, and A. Kanafani. Integrating the Airport System into the Intelligent
Vehicle/Highway System Programs. Institute of Transportation Studies Working Paper UCB-ITS-WP-93-
16. University of California at Berkeley, Berkeley, California, November 1993.

5. Polak, J. W., P. Vythoulkas, and I. Chatfield. Broadcast Parking Information: Behavioral Impacts and
Design Requirements. Transportation Research Record 1403, TRB, National Research Council,
Washington, D.C., 1993, pp. 36-44.

6. Horonjeff, R. and F. X. McKelvey. Planning and Design of Airports - Fourth Edition. McGraw-Hill,
New York, New York, 1994, pp.451.

7. Raleigh-Durham International Airport web site. www.rdu.com.

8. Wei, Wen-Bin. Variable Message Signs.

http://www.path.berkeley.edu/~leap/TTM/Driver_Info/message.html.

9. Amsign Corporation. http://www.amsign.com/system/Default.htm.

10. DAMBACH-WERKE GmbH Enterprise Brochure. Gaggunau, Germany.

11. Allen, P. A. Driver Response to Parking Guidance and Information Systems. Traffic Engineering and
Control, Vol.34, No.6, June 1993, pp. 302-307.

12. Gavin, K. Parking Guidance Comes of Age. http://www.parking-

net.com/authors/publications/K.Gavin/0001/0001.htm.

13. Polak, J. W., I. C. Hilton, K. W. Axhausen, and W. Young. Parking Guidance and Information Systems:
Perfomance and Capability. Traffic Engineering and Control, Vol.31,No. 10, October 1990, pp. 519-524.

14. Urban & Environmental Graphics web site. http://www.ozemail.com.au.

15. A Guided Parking System. http://library.urbanmobility.net/ViewHtml.asp?Item=34&Action=Html.

16. Keller, B. D. Integrating Advanced Parking Information Systems with Traffic Management Systems.
Compendium: Graduate Student Papers on Advanced Surface Transportation Systems. Southwest Region
University Transportation Center Report SWUTC/95/72194-4, Texas Transportation Institute, College
Station, Texas, 1995.

17. Axhausen, K. W., J. W. Polak, M. Boltze, and J. Puzicha. Effectiveness of the Parking Guidance
Information System in Frankfurt am Main.
http://library.urbanmobility.net/ViewHtml.asp?Item=29&Action=Html.

18. Variable Message Signs (VMSs) Washington Department of Transportation.

http://www.wsdot.wa.gov/regions/northwest/NWFLOW/brochure/vms.htm.

19. Siemens. Signs and Signing Systems. http://www.siemens.co.uk/traffic/signs.htm.

20. Pang, G. Y. K. Electronic Road Signs Turn Off Some Windward Residents. Honolulu Star-Bulletin Local
News. Honolulu, Hawaii, January 1998.

21. Smith, B. et al. An Investigation of Operational Procedures for Highway Advisory Radio System.
Virginia Transportation Research Council. Charlottesville, Virginia, September 1995.
Debra A. Burdette Page 108

22. Highway Advisory Radio (HAR) System. Washington Department of Transportation.

http://www.wsdot.was.gov/regions/northwest/NWFLOW/brochure/har.htm.

23. Highway Advisory Radio of New Jersey Turnpike Authority. http://www.state.nj.us/turnpike/har.htm

24. Lyons, G. D. and M. McDonald. Traveller Information and the Internet. Traffic Engineering and Control,
Vol. 39, No. 1, January 1998, pp. 24-31.

25. Fogle, K. L. The Development of an Internet Home Page to Provide Information for Regional Travelers.
Compendium: Graduate Student Papers on Advanced Surface Transportation Systems. Southwest Region
University Transportation Center Report SWUTC/97/72840-00003-2, Texas Transportation Institute,
College Station, Texas, 1997.

26. Shull, L. Use of the Internet as an Effective Tool for Disseminating Traveler Information. Compendium:
Graduate Student Papers on Advanced Surface Transportation Systems. Southwest Region University
Transportation Center Report SWUTC/96/72840-00003-1, Texas Transportation Institute, College
Station, Texas, 1996.

27. Williams, D. F. Atlanta to Showcase ITS Traveler Information.

http://www.tfhrc.gov/pubrds/summer96/p96su35.htm.

28. Dudek, C. L. and J. D. Carvell, Jr. Feasibility Investigation of Audio Modes for Teal-Time Motorist
Information in Urban Freeway Corridors. Texas Transportation Institute Research Report RF 953-8,
College Station, Texas, April 1973.

29. Kitamura, R., L. Laidet, and F. Mar. An Evaluation Study of a Low-Cost, Telephone-Based Traveler
Information System. California PATH Program, Institute of Transportation Studies, University of
California, Davis, 1995.

30. Charles River Associates. User Acceptance of ATIS Products and Services: A Report of Qualitative
Research. Project No. 852-03, prepared for the U.S. Department of Transportation, January 1997.

31. Tisdell, R. K. and E. J. Karkiwski. Navigation and Tracking Systems, Part II. ITS World, Volume 2, No.
2, March/April 1997, pp. 35-38.

32. Orski, C. K. Is There a Consumer Market for Traveler Information Services? Traffic Technology
International, August/September 1996, pp. 52-55.

33. Quisenbury, T. Telephone Interview, Raleigh-Durham International Airport, RDU Airport, North
Carolina, June 1999.

34. Houston Airport System web site. http://www.ci.houston.x.us/has.

35. Stehman, R. Telephone Interview, HNTB Corporation, Falls Church, Virginia, June 1999.

36. IAH garage uses “smart technology” to aid drivers’ search for vacant parking. Houston Airport System
Newsletter. Fourth Quarter 1997.

37. Metropolitan Washington Airport Authority web site. http://www.mwaa.com.

38. Yedinak, M. E-mail Interview, Metropolitan Washington Airport Authority, Washington, D.C., July,
1999.
Debra A. Burdette Page 109

APPENDIX A: SURVEY QUESTIONNAIRE

Dear Airport Traveler:

This survey is being conducted as part of a research project to determine your airport parking information
needs. So that you may be better served, please take a few minutes to answer the following questions. Your
cooperation in this effort is greatly appreciated. All information you provide will remain strictly confidential.

1. Please indicate the number of times you parked at the fllowing types of parking facilities in the last
12 months.

Short-term __________ Long-term __________ Satellite/reduced rate __________

2. Please rate the following types of parking information according to your perceived value -(1 =
highly value, 4 = not valuable)

___ Parking cost ___ Distance/time from parking to terminal

___ Parking availability ___ Alternative parking lot suggestions
___ Directions to parking lots ___ Other _________________________

3. When would you like to receive parking information? Check all that apply

Before leaving for the airport When entering the airport grounds
Just before entering the parking area

4. How would you like to receive airport parking information? Check all that apply

Radio/In-vehicle
Variable message signs
Static signing
Internet
800-telephone number
Television

5. Please list three airports where you have experience in parking, and indicate the level of difficulty
you experienced in finding a parking space.

Airport Location Not Difficult Somewhat Difficult Very Difficult

____________________
____________________
____________________

In the space provided below, please feel free to give us any general comments concerning your parking
information needs as an air traveler.
____________________________________________________________________________________
____________________________________________________________________________________
____________________________________________________________________________________
That completes this survey. Your time and effort are greatly appreciated. Thank you again.
Debra A. Burdette Page 110

APPENDIX B: SURVEY DATA

Question1:

Number of times parked in the last 12 months

0-2 3-5 6+ Total
Short-term 19 5 6 30
Long-term 28 2 0 30
Satellite/reduced rate 27 3 0 30

Question 2:

Parking information ratings (number of responses)

Very Valuable Somewhat Not N/A Total
Valuable Valuable Valuable
Parking Cost 9 7 11 1 2 30
Parking Availability 22 6 0 0 2 30
Directions to lots 19 9 0 0 2 30
Distance/travel time to terminal 7 14 7 1 1 30
Alternative suggestions 4 15 7 2 2 30
Other 2 2 1 0 16 21

Question 3:

When would you like to receive parking information?

(number of responses)
Before leaving for the airport 8
Just before the parking area 18
Upon entering airport grounds 27

Question 4:

How would you like to receive airport parking information? (number

of responses)
Radio/In-vehicle 9
Information message sign 30
Static signing 17
Internet 7
800-telephone number 3
Television 0
Debra A. Burdette Page 111

Question 5:

Ease of parking at various airports (number of responses)

Not Somewhat Very Not Somewhat Very
Airport Airport
Difficult Difficult Difficult Difficult Difficult Difficult
DFW 3 2 SEL 1
IAH 4 8 1 MCO 1
CLL 6 1 BZN 3
HOU 4 1 SMF 1
ATL 3 MCI 2
GSP 1 MSY 1
DIA 3 ANC 1
BWI 1 CMH 4
IAD 1 1 BUF 1
AUS 4 CVG 2
PHL 1 PHX 1
SAT 1 TPA 1
KUL 1 DAY 1
CLT 2 CAE 2
AGS 2 PIT 1
JNB 1

Additional Comments:

• Give shuttle times between parking and terminal buildings (mentioned twice).
• Parking should be cheaper.
• More curbside pick-up and drop-off spaces.
• Give proximity of parking to each airline.
• Just point me in the right direction and tell me where the empty spots are. Price doesn’t really matter
since I just spent $300 for a ticket, what’s another $20?
• I like the idea of variable message signs at the entrance to the airport grounds. In car radio, the Internet,
etc. is too difficult to follow, and the Internet will be outdated. I think the parking areas need better
pavement markings.
• Air travelers need to be informed of available spaces before they get to decision points.
• I like to use satellite parking for trips longer than a couple days, and I like knowing how much that
parking costs and where it is located soon after entering the airport grounds.
Debra A. Burdette Page 112

DEBBIE BURDETTE

Debbie Burdette received her B.S. in Civil Engineering from

Clemson University in May 1998. While at Clemson, she was an
active member of the American Society of Civil Engineers,
serving as the 1996 Social Coordinator and the 1998 Steel Bridge
Project Manager. Additionally, she was an undergraduate
research assistant at Clemson’s Wind Load Test Facility and was
enrolled in the cooperative education program, working four
semesters for CSX Transportation’s Atlanta District Design and
Construction office. Debbie is currently pursuing her M.S. in
Civil Engineering at Texas A&M University and is currently
employed as a graduate research assistant at the Texas
Transportation Institute. She is presently a member of Texas
A&M’s Institute for Transportation Engineers and Intelligent
Transportation Society of America student chapters, serving as the ITE Vice-president. Her other
professional experience includes a summer engineer position at Wilbur Smith Associates in Atlanta, GA,
where she participated in highway design and airport planning projects. Her career interests include:
transportation planning, airport design, public transportation, public involvement, and intermodal ITS
applications.
 METHODS FOR THE DEVELOPMENT AND USE OF A TRAFFIC
VIOLATION DATABASE TO ENHANCE SAFETY ENFORCEMENT EFFORTS

by

Randy W. Carroll

Professional Mentors
William M. Spreitzer
General Motors Corporation

and

Professional Mentor
Wayne K. Kittelson, P.E.
Kittelson & Associates, Inc.

Course Instructor
Conrad L. Dudek, Ph.D., P.E.

Prepared for
CVEN 677
Advance Surface Transportation Systems

Department of Civil Engineering

Texas A&M University
College Station, TX

Summer 1999
Randy Carroll Page 114

SUMMARY

The objective of this research was to document the process and equipment needed to build and operate a
traffic violation database that would allow law enforcement to scientifically document problem locations to
more efficiently and effectively deploy their resources. The database would be capable of archiving violation
data related to red-light running, highway-railroad grade crossing gate violations and speeding.

Designing, developing and implementing the traffic violation database system involves assembling a
collection of resources including; software, hardware, capital, staff and training. These resources must then
be integrated in an effective and efficient manner. Implementation of a traffic violation database system can
be accomplished in several ways, from developing and operating the entire system in-house to outsourcing
key components to independent contractors. Outsourcing could be in the form of installation and
maintenance of hardware to custom database creation. Twelve essential steps were identified in the
development and successful operation of a traffic violation database.

Law enforcement agencies from 29 states were contacted and asked to participate in a survey. Responses
were received from ten law enforcement agencies in eight states. Eighty-three percent of respondents
indicated that a traffic violation database would address an existing need at their agency. All of those 83
percent indicated that this system would assist in identifying problem areas and allow deployment of
manpower based on objective data. Eighty-three percent of respondents indicated that this type of system
would become a regular/integral part of their enforcement activities, while only 17 percent indicated that it
would only be used in special situations.

The basic structure of the traffic violation database was based upon the results of the law enforcement
survey. The survey indicated that the traffic violation database should: operate on a Windows platform;
allow data to be published to the Internet; allow data to be input both automatically and manually, and utilize
a relational type database with indexed searches. For the traffic violation database to operate as a true
relational type database, a large amount of information must first be entered. This information includes
geometry, traffic control, signal timings, current ADT, and police jurisdiction. Historical data indicating the
violation rate, accident rate, 85th percentile speed, and ADT must also be included. Both custom and
prepacked database programs are available for use as the traffic violation database.

It is recommended that two rectangular inductive loops should be used in all lanes of interest at an
intersection to detect violations. Two inductive loops allow a minimum speed threshold to be used to avoid
erroneous detections. These inductive loops should be placed between one half and five meters apart from
front to back and should be located near the stopbar. At highway-railroad grade crossings one inductive loop
should be placed downstream on the crossing. This detector would only detect the passage of a vehicle. The
inductive loop should be placed in an area where there are no driveways or turning traffic, only vehicles
driving around the gates should be registered by the inductive loop. Both installations require a processing
unit to accompany the detection equipment. This processing unit must be custom made and allow a remote
access connection data downloading. This processing unit would be wired into the signal control at the
intersection or highway-railroad grade crossing. This would allow the system to become active only when
violations could occur.

Speed trailers should be used to record speed violations throughout the city street system. The speed trailers
should use laser radar to detect vehicle speeds, along with a reader board to display the vehicle’s speed. The
laser radar also needs to be accompanied by a processing unit that would determine if the vehicle is in
violation of the speed limit for that particular street. This processing unit only needs to be capable of manual
data downloading.
Randy Carroll Page 115

TABLE OF CONTENTS

INTRODUCTION............................................................................................................................. 117
Problem Statement....................................................................................................................... 117
Research Objectives..................................................................................................................... 117
Scope.............................................................................................................................................. 118
Work Plan...................................................................................................................................... 118
Task 1: Review Literature...................................................................................................... 118
Task 2: Develop Law Enforcement Survey............................................................................ 118
Task 3: Administer Law Enforcement Survey........................................................................ 118
Task 4: Develop Traffic Violation Database Concept........................................................... 118
Task 5: Apply the Traffic Violation Database Concept......................................................... 118

BACKGROUND................................................................................................................................ 119
Automated Enforcement.............................................................................................................. 119
Red light Running......................................................................................................................... 119
Victoria, Australia.................................................................................................................. 119
New York, New York............................................................................................................... 120
Polk County, Florida.............................................................................................................. 120
Howard County, Maryland..................................................................................................... 120
Montgomery County, Maryland............................................................................................. 121
Oxnard California.................................................................................................................. 121
Charlotte, North Carolina...................................................................................................... 121
Speed Limit Enforcement............................................................................................................ 121
Edmonton, Alberta.................................................................................................................. 122
Victoria, Australia.................................................................................................................. 122
National City, California........................................................................................................ 122
Paradise Valley, Arizona........................................................................................................ 122
Grade Crossings............................................................................................................................ 123
Jonesboro, Arkansas............................................................................................................... 123
Los Angeles, California.......................................................................................................... 123
Ames, Iowa.............................................................................................................................. 123
Texas....................................................................................................................................... 124
Law Enforcement.......................................................................................................................... 124
Legal Issues.................................................................................................................................... 124
Automated Enforcement Opponents.......................................................................................... 125

LAW ENFORCEMENT SURVEY................................................................................................... 126

Survey Design................................................................................................................................. 126
Survey Administration.................................................................................................................. 127
Survey Results................................................................................................................................ 127

TRAFFIC VIOLATION DATABASE.............................................................................................. 128

Conceptual Design.......................................................................................................................... 129
Database Structure......................................................................................................................... 129
Software........................................................................................................................................... 130
Violation Detection - Intersections and Highway-Railroad Grade Crossings.......................... 131
Air Tubes................................................................................................................................... 131
Piezoelectric Sensors................................................................................................................ 131
Laser Detectors......................................................................................................................... 132
Video Loops.............................................................................................................................. 132
Randy Carroll Page 116

Inductive Loops......................................................................................................................... 132

Earth Magnetic Loops.............................................................................................................. 133
Violation Detection - Speed............................................................................................................ 133
On-Site Communication and Storage........................................................................................... 133
Intersections and Highway-Railroad Grade Crossings........................................................... 133
Speed Monitoring...................................................................................................................... 133

RECOMMENDATIONS..................................................................................................................... 134
Location Identification................................................................................................................... 134
Violation Detection Equipment..................................................................................................... 134
Intersections............................................................................................................................. 134
Highway-Railroad Grade Crossings........................................................................................ 136
Speed Monitoring..................................................................................................................... 137
Traffic Violation Database Software............................................................................................ 137

TRAFFIC VIOLATION DATABASE APPLICATION.................................................................. 138

REFERENCES..................................................................................................................................... 140

APPENDIX A - Law Enforcement Survey.......................................................................................... 142

Randy Carroll Page 117

INTRODUCTION

Automated enforcement is defined as the use of technology to capture images for the purpose of monitoring
and enforcing traffic control laws, regulations, or restrictions (1). The image capture technology negates the
need for law enforcement to directly witness the event when enabling legislation has authorized its use.
Numerous legal and social issues surround the use of automated enforcement and limit its attractiveness as
a cost-effective solution. Opponents argue that automated enforcement violates personal privacy and
constitutional rights for the sole purpose of generating revenue. These programs are seen as a dishonest form
of enforcement that is less effective than traditional enforcement efforts.

In a highly litigious country, such as the United States, where citizens’ rights are held to the highest regard,
innovative solutions that do not violate the perceived constitutional rights of motorists must be found to help
mitigate traffic violations. Testimony given in 1997 by Ricardo Martinez, the Administrator of the National
Highway Traffic Safety Administration, estimated that in recent years one-third of all crashes and two-thirds
of the resulting fatalities could be attributed to aggressive driving behavior (2). A reduced presence of law
enforcement on our nation’s roadways is thought to contribute to, if not encourage, aggressive driving.
However, the limited financial resources available to law enforcement agencies prohibit an increased
presence.

Problem Statement

Historically, automated enforcement has been based on an all-or-nothing approach with little to no regard
for the officer responsible for conventional enforcement. The social climate of today severely limits the
acceptance of technology as an alternative to law enforcement. A public opinion poll conducted in 1989 by
the Insurance Institute for Highway Safety indicated that 58 percent of respondents approved of automated
enforcement and 35 percent disapproved (3). Although opponents are outnumbered, they are still capable
of prohibiting or discontinuing automated enforcement programs by lobbying elected officials or encouraging
motorists to contest mailed citations in court. Law enforcement estimates that if two percent of traffic
citations issued were contested in court, the extra burden on the judicial system would be overwhelming (4).
Automated enforcement may be inevitable in the future. However, cost-effective, non-invasive solutions
exist today that would enhance the current capabilities of law enforcement. A simple, yet highly effective
alternative to automated enforcement would incorporate technology into law enforcement activities to
enhance their capabilities without sacrificing the objectiveness of the officer in the field.

Research Objectives

The objective of this research was to determine the process and equipment needed to build and operate a
traffic violation database that would allow law enforcement to scientifically document problem locations to
more efficiently and effectively deploy their resources. The database would be capable or archiving violation
data related to red-light running, highway-railroad grade crossing gate violations, speeding and high-
occupancy vehicle lane violations. Data would be collected using key components of existing automated
enforcement systems. Once a problem has been clearly documented, it can be used to describe traffic
violation problems and demonstrate the need for more enforcement to legislators, city government and the
general public. States that currently have enabling legislation for automated enforcement, as well as those
that do not, would benefit from the ability to report on trends in violations over time. In order to prove the
validity of the traffic violation database, the findings and recommendations of this study are applied in a
general sense to the city of College Station, Texas.
Randy Carroll Page 118

Scope

The scope of this research was limited to the conceptual development of a traffic violation database and the
supporting technology needed to gather the data. As indicated in the problem statement and research
objective, this system should be structured around law enforcement capabilities and constraints. The study
focused on alternate means of existing technology to enhance the enforcement capability of officers in the
field. The scope of this paper was limited to red-light running, highway-railroad grade crossings gate
violations and speeding.

Work Plan

In order to accomplished the research objective outlined in this proposal and to provide a thorough
understanding of the issues related to enhanced enforcement, the following tasks were completed.

Task 1: Review Literature

A literature review was performed in order to examine previous research and issues related to automated
enforcement. Specifically, case studies documenting the use of automated enforcement technology in the
United States and its application to typical law enforcement. Legal and social issues that prohibit the use
of automated enforcement technology were also identified.

Task 2: Develop Law Enforcement Survey

A telephone and fax-back survey was constructed to elicit specific information regarding the use and
acceptance of technology in enforcement and the structure of the proposed traffic violation database.

Task 3: Administer Law Enforcement Survey

The survey was administered in fax-back form to law enforcement personnel throughout the United States.
Law enforcement contacts were gathered from Federal Highway Administration (FHWA) Demonstration
Project No. 86 - Incident Management Workshop sign-in sheets. Additional contacts were also obtained from
mentors and various Texas A&M University staff and Texas Transportation Institute employees.

Task 4: Develop Traffic Violation Database Concept

Using the results of the law enforcement survey and literature review, the structure of the traffic violation
database was determined. This included determining what technologies can be used to gather, transmit, store
and display the traffic violation data to law enforcement personnel in accordance to law enforcement desires.
The environments in which the database will function was also determined.

Task 5: Apply the Traffic Violation Database Concept

In order to prove the validity of the traffic violation database, the findings and recommendations of this study
were to be applied conceptually to the city of College Station, Texas. Problem locations were to be identified
through the help of the College Station Police Department and the appropriate technologies, as determined
by the study, were to be described for each location in terms of installation, data collected and potential
benefits. The hypothetical application was also to describe what information police officers in the College
Station Police Department would want from each problem location and how it would be utilized to more
effectively and efficiently deploy their resources. Since contacts with the city of College Station were not
obtained, this task was pursued in a general sense.
Randy Carroll Page 119

BACKGROUND

Enforcement of traffic laws is the primary means through which hazardous driving behavior can be
controlled. Maintaining a strong police presence on city streets and roadways has traditionally been used
to enforce traffic laws. However, during the past two decades, the number of licensed drivers, vehicles and
vehicle miles of travel (VMT) have increased steadily. According to the Federal Highway Administration,
the total number of licensed drivers in the United States in 1997 was179,539,340. There are roughly 880
licensed drivers to every 1000 persons of legal driving age. The total number of all licensed publicly owned,
private and commercial vehicles, including motorcycles is 210,236,393 (5). The ability of law enforcement
to respond to these rising demands has clearly diminished over the years.

A 1992 survey conducted by the Insurance Institute for Highway Safety showed police enforcement resources
in relation to need from 1978 to 1989. The findings from 33 states showed only a 6 percent increase in the
number of officers whose duties included traffic law enforcement, while the number of licensed drivers
increased almost 19 percent. The number of traffic officers per million licensed drivers fell approximately
11 percent while the annual number of miles traveled increased 38 percent (6). This suggests that a more
efficient and effective allocation of resources is needed to maintain the integrity of traffic law enforcement.
The National Highway Traffic Safety Administration (NHTSA) reports that approximately 66 percent of all
annual traffic fatalities are caused by aggressive driving behavior such as red light running and speeding (7).

Automated Enforcement

Automated enforcement was previously defined as the use of technology to capture images of license plates
and drivers for the purpose of monitoring and enforcing traffic control laws, regulations or restrictions.
Automated enforcement systems typically consist of three key components, the detection equipment, a
processing unit and a camera to record the violation. In states where enabling legislation has been enacted,
this image capture technology negates the need for law enforcement to directly witness a violation.

Red Light Running

Enforcing red-light violations at intersections has become one of the most common uses for automated
enforcement. Traditional enforcement methods for red light runners are limited in efficiently addressing the
increasing frequency of aggressive drivers. Red light violations are easily discernable and perhaps the least
ambiguous traffic violation. Yet, the United States has been relatively slow to embrace the idea of automated
enforcement. However, it appears that automated enforcement technologies have been used abroad for a
number of years and have proven to be effective and reliable.

Detection equipment typically consists of buried loops or piezo sensors. These detectors are typically placed
at or near the stop bar to detect approaching vehicles and vehicle speeds. Once the traffic signal has switched
to the red phase, the camera used to record the violation becomes active. A small period of time and a preset
speed are incorporated into the system to differentiate turning vehicles and vehicles coming to a stop. If both
of these criteria are exceeded the image capture device records an image. Image capture devices can either
be standard 35 mm cameras or digital image processors. The following are brief descriptions on several red
light running automated enforcement programs throughout the world.

Victoria, Australia

Australia has used automated enforcement for red light violations since 1979 (8). Red light cameras were
introduced in Victoria to reduce the number of collisions occurring as a result of motorists disobeying traffic
control signals. After an initial six month trial period in 1981, ten red light cameras were purchased by the
Road Traffic Authority and first used on August 16, 1983. The number of cameras has progressively
increased to 35 which are rotated through 132 sites within the Melbourne metropolitan region.
Randy Carroll Page 120

New York, New York

New York, New York currently has the largest and oldest automated enforcement program for red light
runners in the United States. The New York City Department of Transportation had been researching
automated enforcement technology since 1983. Program success in countries like Europe and Australia
prompted the City of New York to conduct a demonstration project to prove the red light running technology.
Through the demonstration of the automated enforcement technology the Department of Transportation was
able to gain the support of the City Government and begin the Request For Proposal (RFP) process.
However, no public education campaign was undertaken and no advance warning was given as to the
placement of the cameras (9).

A contract was awarded to Electronic Data Systems by the City of New York on the premise that the cost
of the system would be offset by ticket revenue. The total price for the three and a half year contract was
$13,900,000 (10). This contract was awarded after the State of New York passed a law allowing the use of
automated enforcement technology for red light violations. Privacy issues were raised during the programs
inception and subsequently only the registered owner of the vehicle can be cited because photographs are
taken of the rear license plate (9). Initially, 20 locations were selected for camera installation. These twenty
locations were determined by using accident information and contacting local police precincts to identify
locations that were receiving special attention (9). Since the inception of the program, violations at
intersections with automated enforcement technology have decreased by 20 percent.

Polk County, Florida

In September of 1994 Polk County, Florida began using automated enforcement technology for red light
runners as part of a Federal Highway Administration demonstration project. The demonstration was
conducted at four intersections in several different environments in Fort Meade, Haines City and Lakeland
(11). A public education campaign was initiated with funding by the Federal Highway Administration to
reinforce the problems associated with red light running. This program followed the recommendations given
in the Red Light Running Program Strategic Planning Guide by Plosky (12). Since Florida does not currently
have enabling legislation to allow citations to be issued, warning letters are mailed to drivers who are caught
running red lights. Statistics are not currently available on the effectiveness of the system in terms of
violation reductions (1).

Howard County, Maryland

Howard County has completed phase I of a Federally funded demonstration project to test automated
enforcement systems for red light violations. This program demonstrated the feasibility and reliability of
photographic based surveillance systems to detect and record red light violations at problem intersections.
The use of the cameras was in response to the high cost of traditional enforcement methods and a rising need
for enforcement at intersections. A significant public awareness campaign was undertaken through numerous
media outlets. Signs were posted on roadways that contained cameras but the exact intersection where the
camera was placed was not publicized. During the test period, warning notices were sent to the registered
owner of the vehicle. As a result of that study, legislation was passed by the 1997 Maryland General
Assembly, House Bill 391, that authorizes law enforcement to use such automated enforcement technology
to enforce traffic laws at signalized intersections. The program was implemented on February 3, 1998, with
two cameras. The 20th camera became operational in November 1998. In 1998, violation rates at the
monitored intersections were reduced by an average rate of 48 percent (13).

Phase II of the demonstration project began in 1998 with plans to demonstrate the feasibility of digital red
light running automated enforcement systems. Vendors in Australia, Great Britain and Israel are participating
in the demonstration and installation of digital camera systems at four sites has begun. Digital cameras can
Randy Carroll Page 121

electronically transmit pictures to a central computer for identifying owners and issuing notices. The most
important benefit will come in the form of time and manpower savings as the cameras no longer need
servicing on a daily basis (13).

Montgomery County, Maryland

Montgomery County is participating with Howard County in testing the feasibility of digital camera
technology. The first camera was installed in the Bethesda area and the Police Department and the
Department of Public Works and Transportation are currently working to identify locations for future
installations. Initially, warning notices will be sent to violators during a test period. Following the test
period tickets will be sent to violators. The fine is $75.00 and no points will be issued against the vehicle’s
owner. Failure to pay a ticket can result in the inability of the vehicle’s owner to renew his/her driver's
license or vehicle registration (14).

Oxnard, California

The automated enforcement of red light violations began in Oxnard, California on April 4, 1997. The
program began to combat the growing number of red light violations and the need for 24-hr enforcement.
The program was initiated with 15 intersection approaches. A public education and awareness campaign was
initiated to inform the public about the purpose of the automated enforcement program. California Vehicle
Code SB833 authorizes the city of Oxnard to use automated enforcement technology with frontal
photography to identify both the driver and the vehicle. Results of this enforcement effort could not be
obtained (1).

Charlotte, North Carolina

In 1998, Charlotte became the first city in the state of North Carolina to implement an automated
enforcement program, called SafeLight, for red light violations. The North Carolina State Legislature ratified
enabling Legislation in 1997 and the Charlotte City Council approved the program in 1998, paving the way
for automated enforcement technology to be used at signalized intersections. Charlotte contracted with
Lockheed Martin IMS and in July of 1998 the first two cameras were installed. Eighteen more cameras were
added in September. All the cameras are located at high accident locations and crashes were reduced by 32
percent on the monitored approaches (15).

Speed Limit Enforcement

Automated speed limit enforcement is one of the earliest automated enforcement applications in the United
States and abroad. Traditional enforcement methods for enforcing speed limits are limited in their ability
to efficiently address the increasing frequency of aggressive drivers. Automated speed limit enforcement
is considered by many to be the most controversial type of enforcement due to the variability associated with
setting speed limits (1). However, it appears that the use of automated speed limit enforcement is on the
rise.

Speed enforcement systems usually consist of a radar device, a processing unit to identify violators and a
camera to identify the vehicle. When the processing unit determines that a vehicle is exceeding a preset
speed, an image is recorded with either a standard 35 mm camera or a digital image processor. Pertinent
information regarding the violation is also superimposed upon the image. Automated speed limit
enforcement equipment may be installed in a stationary vehicle, portable trailer, tri-pod, or it may be
mounted in a self-contained cabinet to a pole.

Speed limit awareness programs are also gaining popularity at home and abroad. These programs use radar
device to measure speeds and an electronic display panel to provide the speed to the passing motorist. Speed
Randy Carroll Page 122

trailers are mobile roadside devices that measure and display the speed of a passing vehicle. Speed trailers
include a radar device to detect and measure the speed of a passing vehicle, as well as a display board which
displays the speed of the detected vehicle and the legal speed limit of the roadway. Usage of speed trailers
had grown steadily since their development in the late 1980's. In late 1998 500 devices were in operation
across the United States.

Speed enforcement trailers have been used to enhance typical officer enforcement. However, they have
typically been used as a public education device. Speed trailers are also used for automated enforcement
purposes in a few states. When the speed trailer is used for automated enforcement purposes an image
capture device is installed which takes of picture of a vehicle if it is above a predetermined threshold. This
image is typically stored internally and downloaded through a modem connection or manually in the field
(18). The following are brief descriptions of automated speed limit enforcement programs throughout the
world.

Edmonton, Alberta

In 1987, the Edmonton Police Service began exploring using photo radar. Between 1987 and 1993, the
Edmonton Police Service evaluated several photo radar systems for future use. It was decided that photo
radar would be an effective way to conduct enforcement and improve officer safety. A secondary benefit
would be the reduced number of officers needed for traffic enforcement. The system became operational and
enforcement officially began in March 1993. A second unit was introduced in 1997. The photo radar units
are housed in two unmarked police vehicles. The Edmonton Police Service estimates that they would need
over 30 police officers to maintain the current level of enforcement without the use of the two photo radar
units. The photo radar units are used at high speed locations, high collision locations, areas of heavy traffic
volume and school zones. Also, the photo radar units can be used to deal with citizen complaints about
speeding vehicles in their communities, as well as to monitor traffic in those areas (16).

Victoria, Australia

Photo radar enforcement began in Victoria in 1989 with the introduction of 54 speed cameras. The photo
radar units are mobile and are operated by the Victoria Police. The Traffic Camera Office (TCO) was
established in 1990 to administer the program. Other automated enforcement efforts in Victoria were also
consolidated into the TCO. This consolidation effort was essential to the continued success of the programs.
The high level of enforcement provided by the Victoria Police and the TCO is intended to modify driving
behavior and reinforce the perception of strict enforcement. Since the program was initiated in 1989 the total
number of accidents decreased by 22 percent (1).

National City, California

Photo radar enforcement began in National City in 1991 with the introduction of one camera. This camera
was mounted in a mobile vehicle and rotated among 90 enforcement locations. This program issues the ticket
to the driver of the vehicle, not the registered owner. This requires the owner of the vehicle to identify the
driver in the event he/she was not driving. If the owner does not identify the driver the case is thrown out
of court due to a loophole in California law. In a six year period from 1991-1997 the number of accidents
decreased by 51 percent (1).

Paradise Valley, Arizona

In 1987, a legislative change enabled the City of Paradise Valley to pass a city ordinance allowing the use
of photo enforcement for speed limit violators. The program began using manned automated enforcement
equipment to enforce speed limits at approximately 60 locations throughout Paradise Valley and has grown
Randy Carroll Page 123

from there. The threshold for violations occurs at 125 percent of the posted speed. This program is the
longest running operation in the United States and is credited with substantial accident reductions (17).

Grade Crossings

Automated enforcement of highway-railroad grade crossing violations is one of the newest forms of
automated enforcement. A typical system provides surveillance of one or more traffic lanes on one or both
approaches. Inductive loop detectors are used to detect vehicles weaving through the gates. A connection
between the control circuitry of the highway-railroad grade crossing allows the system to activate itself and
await violations. Video, digital image processing or a standard 35 mm camera are used to provide images
of the violations. Pertinent information relating to the time of the violation, the location of the violation, and
the gate, signal and train operation are superimposed onto the image. The following are brief descriptions
of grade crossing automated enforcement programs throughout the United States.

Jonesboro, Arkansas

The first United States installation for automated enforcement at highway-railroad grade crossings was
demonstrated in the City of Jonesboro, Arkansas in 1991. This system consisted of field equipment along
with a direct telephone link to playback equipment in the Jonesboro Police Department. Police viewed the
transmitted images and determined if a citation was warranted. Approximately a year and a half after the
project began, the transmitted images were no longer of suitable quality for use as evidence due to equipment
and maintenance problems. As of 1998 the system is still not operational but the City of Jonesboro remains
optimistic that the system will be operational in the near future. No studies were conducted to determine the
effect of the system on accident reduction. However, it is believed that the crossing was safer when the
system was operational (1).

Los Angeles, California

In 1992, the Metropolitan Transit Authority began five demonstration projects to determine the feasibility
of using automated enforcement equipment at highway-railroad grade crossings. Citations were issued in
three of the five demonstrations projects. Large reductions in the number of grade crossing violations were
seen at each test site after the installation of the equipment. Based on the positive results of the five
demonstration projects, the Metropolitan Transit Authority decided to proceed with the installation of
automated enforcement equipment at 17 crossings. The Metro Blue Line Grade Crossing Safety Program
was initiated in 1993 by the Los Angeles Metropolitan Transit Authority. The program was designed to
discourage and prevent illegal movements being made by vehicles at both gated and non-gated grade
crossings (19).

Ames, Iowa

An automated enforcement system became operational in Ames, Iowa in 1996. The need for automated
enforcement was sparked by concern for motorist safety and a lack of cooperation on the part of the railroad
to install a four-quadrant gate system. A digital image capture system was used to photograph the violation.
However, Iowa state law requires that the driver of the vehicle be identified and that the citation cannot be
delivered through the mail. Thus, it was required that an officer hand deliver the citation to the home of the
driver. During July of 1997 through September 1998, the system was taken off-line to upgrade the detection
and image capture equipment. This was hoped to resolve problems associated with identifying drivers at
night, as well as license plate legibility (1, 20).
Randy Carroll Page 124

Texas

In 1995 the Texas Legislature passed Senate Bill 1512 which required the Texas Department of
Transportation to conduct an automated enforcement demonstration project at a highway-railroad grade
crossing. Automated enforcement equipment was installed and demonstrated at three crossings with
automatic gates. The equipment photographed, either on 35 mm film or as a data file, vehicles violating the
gate arms. This program successfully demonstrated how automated enforcement equipment can be used to
record violations and identify the license plate and owner of the vehicle. It did not demonstrate the process
of issuing citations or examine implementation efforts (20).

Law Enforcement

Interest in automated enforcement equipment has been building for a number of years. Experience with
automated enforcement equipment in other countries has provided a basis for a number of programs
operating in the United States today; all of which would not exist without the cooperation of law enforcement
agencies. A study conducted by the National Highway Traffic Safety Administration (NHTSA) in the mid
1980's dealt with field testing automated enforcement equipment by three state police agencies (21). The
law enforcement personnel involved in the study generally thought that automated enforcement was a worthy
concept and were in favor of implementing a speed enforcement program using the equipment. However,
interest in the use of automated enforcement equipment diminished after the study was completed and other
enforcement needs arose. Again, interest was expressed in the use of automated enforcement technology in
late 1989 and another report by the NHTSA documented the experiences of eight police departments across
the United States. Most indicated that automated enforcement was a viable alternative to traditional law
enforcement (17).

Increased efficiency in terms of processing citations and resource deployment was a theme common to all
the automated enforcement programs outlined in the literature review. The law enforcement agencies in
these areas were no longer able to provide the necessary enforcement to ensure a safe city street system.
Thus, automated enforcement technology was seen as a tool to broaden their capabilities without increasing
their presence. Programs that initiated public education/awareness campaigns appeared to be more
successful and have a more favorable public opinion.

However, not all law enforcement agencies support the use of automated enforcement equipment. Tom
Wagoner, Chief of Police for the city of Loveland, Colorado decided against using photo radar because he
felt it ultimately undermines respect for the law and police officers. “It's just not our style, we want our
officers to have a personal relationship with our citizens, and I think that if people started receiving tickets
in the mail, it would be perceived as some sort of fundraiser. You can give out ten times as many tickets with
photo radar, but I think that it angers as many people as it pleases (22).”

Legal Issues

Currently, only 20 states have installed some form of automated enforcement equipment. However, all 20
states do not have enabling legislation that allows automated enforcement. Enabling legislation is necessary
to uphold constitutional, state and local standards. A growing interest in the use of automated enforcement
technology prompted the Institute of Transportation Engineers’ (ITE) Traffic Engineering Council to form
a committee to gather and disseminate information on the use of automated enforcement technology (23).
As of November 1998, ITE has adopted Policy G-9 regarding automated enforcement (1):

Enforcement: “It is the policy of ITE to encourage and support enforcement of traffic laws and
regulations, and obedience of traffic control devices where such devices are implemented under
Randy Carroll Page 125

proper guidelines, to promote safety, to maintain operational efficiency, and generate respect for
traffic control devices.”

Facilitating Enforcement: “It is the policy of ITE to support enforcement by using, among other
methods, automated enforcement techniques that use image-capturing technologies to identify
illegal driver behavior. Further, the support for automated enforcement is subject to effective
measures being taken to protect the privacy of vehicle occupants, prompt notification of alleged
offenses, due process being available to alleged offenders, and the use of the collected data being
restricted to law enforcement purposes.”

Numerous case studies both in the United States and abroad have proven the effectiveness of these
technologies. However, strong opposition surrounding the use of automated enforcement has centered on
issues such as ticket revenue distribution, ticketing procedures and their effectiveness, associated costs and
the right to privacy. ITS America adopted the Interim ITS Fair Information and Privacy Principles in 1996.
This document provides guidance on how to implement ITS technologies while respecting individual privacy.
However, these principles recommend against the use of advanced technologies for enforcement purposes
(24).

The legal issues surrounding the use of automated enforcement are rooted in the rights guaranteed to the
citizens of the United States by pertinent amendments to the Constitution. The First and Fourth Amendments
deal with the right to privacy. Currently, there is no court case that has defined a person’s right to privacy
with respect to operating a vehicle under the First Amendment. However, several Supreme Court decisions
have been used to assert that automated enforcement technology does not violate an individual’s right to
privacy because the vehicle they are operating is in public view and on a public roadway (1).

The Fourth Amendment states that a person has the right to freedom from unreasonable searches and
seizures. Again, the Supreme Court has ruled that the Fourth Amendment is based on a reasonable
expectation of privacy. Since the individual is traveling in view of the public on a public road they should
have no expectation of privacy. Thus, several legal experts have concluded that automated enforcement
technology does not violate and individual’s rights under the Fourth Amendment (1).

Concerns regarding due process and equal protection deal with the Fifth and Fourteenth Amendments. In
order for an automated enforcement program to issue tickets under the Fifth and Fourteenth Amendments
the citations must be sent in a timely manner. Reasonable notification should also be given to the public to
warn that automated enforcement technology is in use and citations are being issued. Citizens must also be
provided the right to declare their innocence and present a defense and view the evidence prior to appearing
in court. The right to present a defense is provided to an individual by the Sixth Amendment (1). Most legal
experts believe automated enforcement of traffic laws does not violate the constitutional rights of a citizen.
However, the perceived violation of an individuals rights is just as damaging as an actual violation.

Automated Enforcement Opponents

Privacy and loss of individual freedoms are major concerns to a somewhat small, but extremely vocal group
of citizens. Public education campaigns on the documented success of automated enforcement efforts does
little to persuade these individuals of their merits. These groups of individuals can hinder the ability of a city
or state to pass enabling legislation allowing the use of automated enforcement. They can also hamper the
effectiveness of programs that are already operational. Organizations such as the National Motorists
Association (NMA), founded in 1982, are dedicated to protecting the rights of motorists and have lobbied
successfully on local, state and federal levels (25). NMA opposes all camera-based law enforcement and is
working toward a ban on its use in all of North America. NMA is funded entirely through membership dues
by individuals, families and small businesses. NMA provides a number of services to it members including
legal services aimed at successfully contesting a camera-based citation. They stress that all citations received
Randy Carroll Page 126

through an automated enforcement program should be contested, if for no other reason than to overload the
court system (25).

Another organization that is based in Canada but fundamentally opposes the use of automated enforcement
is the Society for Safety by Education Not Speed Enforcement (SENSE) (26). SENSE provides many of the
same resources as the NMA. They are also a strong and successful lobbying group that disseminates
information to the public and elected officials on the dangers of automated enforcement. They also stress
the importance of contesting all citations issued under an automated enforcement system. As stated
previously, law enforcement estimates that if two percent of traffic citations issued were contested in court,
the extra burden on the judicial system would be overwhelming (4).

The City of Glendale, Arizona is carefully examining alternatives to photo radar as an enforcement tool for
speed and accident reduction. For more than three years, the city’s traffic engineers have been studying the
traffic volumes and accident rates at major intersections. They have scientifically documented these statistics
at major arterial streets and residential streets in Glendale. It is the goal of the city, through increased
education and enforcement, to gain voluntary compliance from drivers to obey existing speed limits and
traffic regulations by providing them proof of their actions, not implementing a system that may not be
warranted (27).

The American Automobile Association (AAA), a powerful lobbying organization with millions of member
across the United States also opposes the use of some forms of automated enforcement. Art Kinsman, an
AAA government-relations staffer in Massachusetts, indicated AAA has always been opposed to roadside
surveillance and citation without contact with an actual police officer. “The lack of a face-to-face meeting
with a police officer results in an inability to face one’s accuser. It also makes it more difficult to come up
with an adequate defense should an individual choose to fight it in court. In the long run, we feel that it
undermines respect for traffic laws as well as police officers (22).”

The traffic violation database concept could be used to counter claims by these types of organizations that
automated enforcement is solely a revenue generating venture. Scientifically documenting problem areas,
along with public education and awareness efforts could form the basis by which further automated
enforcement efforts could build. If it could be shown that these enhanced enforcement methods and public
education campaigns were not adequate to deter violators then the process of passing legislation and gaining
public support wouldn’t be as difficult.

LAW ENFORCEMENT SURVEY

Survey Design

The first step in the creation of a traffic violation database was to design the law enforcement survey. Two
objectives were identified in the design process. First, each respondent will be able to complete the survey
within 10 minutes and second, easily understand all questions presented therein. Given the limited amount
of time available to complete this research, these requirements were essential in ensuring adequate response
by law enforcement personnel. The law enforcement survey was constructed to be administered by telephone
or in a fax-back format. The survey was designed to elicit specific information regarding the use and
acceptance of technology in traffic enforcement, as well as to help determine the process and equipment
needed to create a traffic violation database. The six areas of inquiry identified for the survey were as
follows:

• Potential application/acceptance of the traffic violation database;

• Perceived benefit of the traffic violation database;
Randy Carroll Page 127

• Computer capabilities of the agency and officer;

• Current methods for deploying resources in the field;
• Ability to use historical violation data as a public relation/education tool, and
• Specific information of use to officers in enforcing traffic law.

Survey Administration

The survey was to be administered to law enforcement personnel who worked in a traffic enforcement
division. It was hoped that one response could be obtained from each state in the continental United States.
However, this goal was not achieved due to the limited amount of time available to complete the survey.
Law enforcement contacts were gathered from sign-in sheets for FHWA Demonstration Project No. 86 -
Incident Management Workshop. Additional contacts were also obtained from CVEN 677 Mentors, Texas
A&M University staff, Texas Transportation Institute (TTI) employees and the Internet. However, everyone
contacted requested that the survey be faxed to their office so they could review its content before
proceeding. Law enforcement agencies from 29 states were contacted and asked to participate in the survey.
Responses were received from 10 law enforcement agencies in 8 states. Responses by state can be seen
below in Table 1.

Survey Results

Thirty-seven surveys were sent to law enforcement agencies in 29 states. Of those 37 surveys 12 responses
were returned from 10 different law enforcement agencies. The following is a list of the 10 responding
agencies:

• Beaverton Police Department – Beaverton, Oregon;

• Bryan Police Department – Bryan, Texas;
• Carlsbad Police Department – Carlsbad, California;
• Casper Police Department – Casper, Wyoming;
• Cheyenne Police Department – Cheyenne, Wyoming;
• Lincoln Police Department – Lincoln, Nebraska;
• Montrose Police Department – Montrose, Colorado;
• Norfolk Police Department – Norfolk, Virginia;
• Prescott Police Department – Prescott, Arizona, and
• San Antonio Police Department – San Antonio, Texas.

A copy of the survey that was used to elicit the 12 responses is in Appendix A. The survey began with a
brief introduction to the database concept. Eighty-three percent of the 12 respondents indicated that this type
of system would address an existing need at their agency. The majority of those 83 percent indicated that
this type of system would assist in identifying problem areas and allow deployment of manpower based on
objective data. Eighty-three percent of respondents indicated that this type of system would become a
regular/integral part of their enforcement activities, while 17 percent indicated that it would only be used in
special situations. Again, the majority of those 83 percent indicated that this type of system would be an
integral part of traffic enforcement. Ninety-two percent of respondents indicated that this type of system
would influence decisions on where to conduct enforcement activities. The majority of those 92 percent
indicated that this type of system would allow them to identify problem areas and deploy their manpower
accordingly. Respondents were then asked where their agency would experience benefits in enforcement
efficiency and/or effectiveness from this type of system. One hundred percent of respondents indicated a
benefit in manpower utilization, 67 percent indicated a time savings, 17 percent indicated a benefit in
equipment utilization and 8 percent indicated possible benefits may come in the form of additional, or
alternate, traffic control devices.
Randy Carroll Page 128

Respondents were asked about their computer skills, as well as the computer capabilities of their agency.
Forty-two percent of respondents indicated their agency currently used DOS based programs. Forty percent
of those 42 percent indicated they were comfortable in using DOS based programs. One hundred percent
of respondents indicated their agency currently used Windows based programs. Seventy-five percent of
those 100 percent indicated they were comfortable using Windows based programs. Ninety-two percent of
all respondents indicated their agency currently had Internet capabilities with a browser. Eighty-two percent
of those 92 percent indicated they were comfortable using the Internet.

Respondents were asked how their agency typically establishes an intersection, grade crossing or segment
of roadway as a problem location. Fifty percent of respondents indicated that their agency uses current
violation rates. Ninety-two percent of respondents indicated their agency uses accident rates. Seventy-five
percent of respondents indicated their agency uses the opinion of the public to establish problem locations.
Sixty-seven percent of respondents indicated problem locations are determined by personal observation.
Thirty-three percent indicated other methods of identifying problem locations, which included information
from the traffic engineering office as well as special situations like construction activities.

Respondents were asked to indicate how their agency currently makes decisions on where and when to
deploy their resources. Seventy-five percent indicated they utilize citizen input or complaints in deploying
resources. Fifty percent of respondents indicated their agency uses some form of accident statistics in
deployment. Twenty-five percent of respondents indicated their agency uses their local traffic engineering
office, personal observation, and/or violation rates to help in deployment decisions. Eight percent of
respondents indicated that past enforcement efforts and experiences are used in deployment decisions.

Respondents were asked how useful they felt a system that could historically document traffic violations to
be in educating the public or city officials on law enforcement concerns. One hundred percent of respondents
indicated that the system would be beneficial in some form. It was indicated that such a system would be
useful in educating the public on problems within the local city street system from a traffic enforcement point
of view. One respondent felt it would be beneficial to display current violation rates on a web page. The
ability to show historical, objective data was also cited as an important feature of the system.

Respondents were asked how much time per week they would be willing/able to spend manually entering
information into the system. Twenty-five percent indicated they were unwilling or had no time available to
manually enter information. Eight percent indicated they could spend one hour per week. Seventeen percent
indicated they could spend two hours per week. Eight percent indicated they could spend three hours per
week. Twenty-five percent indicated they could spend four hours per week. Eight percent indicated they
could spend five hours per week. Eight percent indicated they could spend more than six hours per week
entering data into the system. A common comment regarding data entry was that clerical staff should be
required to perform data entry in most circumstances.

Respondents were asked what information they would like to be able to easily access regarding a specific
location or street in terms of enforcing traffic laws and regulations. Sixty-seven percent indicated the number
of violations, 42 percent indicated time and date of citation, 25 percent indicated number of accidents, 25
percent indicated number of speeders, 25 percent indicated the location of the violation or accident, 17
percent indicated the traffic volume, 17 percent indicated the type of citation, 17 percent indicated the speed
of the roadway, and 8 percent indicated contributing factors.

TRAFFIC VIOLATION DATABASE

Instead of relying on automated enforcement to alleviate engineering or educational problems, agencies

looking to automated enforcement should first consider a more contemporary approach that balances
engineering, education and enforcement. Instead of pushing for a fully automated enforcement system, law
Randy Carroll Page 129

enforcement agencies and local government should consider enhancing traditional enforcement methods
through the use of existing technology.

Any agency interested in correcting a problem must first identify and scientifically document the problem
to assure the problem is real and not perceived. The traffic violation database concept is a tool for
identifying and documenting problems related to red light running, speeding and highway-railroad grade
crossing gate arm violations. It is the first step in what may evolve into a fully automated enforcement
system. However, this information should first be used to address educational problems or alleviate
engineering deficiencies thereby negating the need for automated enforcement. Law enforcement agencies
may be capable of continuing conventional enforcement if their efforts can be targeted at those areas where
problems are real and can be documented. In today’s ever changing technological climate, it is easy to
overlook some of the more simple solutions that can be applied nationwide without the benefit of legislative
action.

Conceptual Design

Designing, developing and implementing a traffic violation database system involves assembling a collection
of resources including; software, hardware, capital, staff and training. These resources must then be
integrated in an effective and efficient manner. Implementation of a traffic violation database system can
be accomplished in several ways, from developing and operating the entire system in-house to outsourcing
key components to independent contractors. Outsourcing could be in the form of installation and
maintenance of hardware to creating a customized database. If the entire system is to be built and operated
in-house, appropriate staffing and training should be accomplished. For the equipment to be effective, a
thorough understanding of the detection equipment must be obtained.

Just as it is important for the public to be informed about the implementation of any new enforcement
program, law enforcement agencies and related personnel must be provided the necessary support, training
and supervision to ensure successful operation of a traffic violation database. Training for all individuals
who have access to the traffic violation database should be completed before the system becomes operational.
This will ensure the success of the system through proper operational methods.

Database Structure

The basic structure of the traffic violation database was determined from the results of the law enforcement
survey. The survey results indicated that the traffic violation database should:

• Operate on a Windows based platform;

• Allow data to be published to the Internet;
• Allow data to be input automatically, as well as manually, and
• Utilize a relational type database with indexed searches.

The basic structure of the traffic violation database is illustrated below in Figure 1. For the traffic violation
database to operate as a true relational type database, a large amount of information must first be entered.
This information can be entered all at one time or on a location by location basis as more detection
equipment is put in the field. The information that is critical to the operation of the traffic violation database
includes:

A list of all streets that are to be monitored;

• Roadway data relating to the highway-railroad grade crossing, intersection or section of roadway. This
includes the geometry, traffic control, signal timings, current Average Daily Traffic (ADT), and police
jurisdiction (if the database will be linked to multiple jurisdictions), and
Randy Carroll Page 130

• Historical roadway data indicating the violation rate, accident rate and historical ADT of the highway-
railroad grade crossing, intersection or segment of roadway. This database should also give the 85th
percentile speed for all roadways.

Speed Violation Red Light Gate Arm

Data Violation Data Violation Data

Roadway Data Historical Roadway

•Geometry Data
•Traffic Control •No. of Accidents
•Channelization •ADT By Lane
•ADT By Lane •No. of Violations
•Police Jurisdictions
•85th Percentile
Speed

User Interface - Search Engine

Figure 1. Relational Database Concept

In the law enforcement survey, respondents were asked what information they would like to be able to easily
access regarding a specific location or street in terms of enforcing traffic laws and regulations. Sixty-seven
percent indicated the number of violations, 42 percent indicated time and date of citation, 25 percent
indicated number of accidents, 25 percent indicated number of speeders, 25 percent indicated the location
of the violation or accident, 17 percent indicated the traffic volume, 17 percent indicated the type of citation,
17 percent indicated the speed of the roadway, and 8 percent indicated contributing factors. Since this
system will be used by law enforcement agencies, it is important to provide the information they have
requested.

It may be possible to import this information from an existing database maintained by local police, the
highway patrol or a state department of transportation. However, this would depend upon the sophistication
of the original database and its compatibility with the traffic violation database. Importing this information
may not be appropriate if the traffic violation database will only be used to monitor a small number of
locations, as the volume of information may be prohibitive. Once this information has been entered into the
database, the detection equipment may be brought on-line and violation data can be accumulated.

Software

The concept of the traffic violation database was to create an effective and efficient means of deploying law
enforcement resources at minimal cost. Thus, it was important to identify software packages that offered
Randy Carroll Page 131

the desired features at a reasonable price. A database is a collection of tables and descriptions. The database
describes the properties, rules, views, connections and stored procedures of the tables. In a database, design
features allow you to create, add, modify and remove tables, relations, views, connections, rules, triggers and
procedures. The descriptive information that is entered about each intersection, grade crossing or segment
of roadway would be contained in different tables. The violation data for each location would also be
contained in these tables. It is the function of the database to provide a relational link between certain fields
or sets of data. Relations can be one-to-one or one-to-many. This simply means when one item is indicated
it is related to one other item or several other items. Combinations of these relations allow many-to-many
relations to be obtained and provide a truly relational database.

Several prepacked software programs are commercially available for creating large relational databases. One
program of note is Microsoft’s Visual FoxPro. Visual FoxPro can be used for data retrieval, manipulation,
querying and reporting. It is also capable of storing vast amounts of data and batch processing. Visual
FoxPro components can be used to build high-powered Internet database applications which allows
publication of data to the Internet. The Traffic Law Enforcement division of the Texas Department of Public
Safety uses Visual FoxPro to manage accident data from across the State of Texas (28).

Software can also be specially designed to meet the specific needs of the traffic violations database. Several
software companies that will take your specifications and create a relational type database application that
provides indexed searches on a Windows platform. They can also integrate web publishing capabilities into
the software. These programs all allow some or all of the data to be published to a website on the Internet.

Violation Detection - Intersections and Highway-Railroad Grade Crossings

In order to be able to detect a traffic violation, some mechanism must be in place to accurately determine if
and when a violation occurred. Detection equipment can be simple or extremely complex. The type of
violation being detected determines the equipment to be used. The detection equipment is the means through
which the recording equipment in an automated enforcement system is activated when a violation occurs.
In this application, the detection equipment is used only to detect and record that a violation has occurred.
The traffic violation database is not used to issue citations or warnings. Thus, image capture technology is
not needed to record the violation. This greatly reduces the cost per intersection of this type of system. This
not only reduces the amount of hardware needed in the field it also reduces the overall size of the traffic
violation database. For the traffic violation database concept the detection equipment must be relatively
inexpensive, dependable and durable. It should be noted that pedestrian and bicycle violations are not
included in the traffic violation database.

Air Tubes

Air tubes are one of the simplest and most inexpensive triggers available. Air tubes can be used to detect
the passage of vehicles and can also be used to determine vehicle speeds. However, air tubes should not be
used for detecting violations due to their limited life.

Piezoelectric Sensors

Piezoelectric sensors can be secured to the roadway in order to detect vehicles that violate the traffic control
device. When a vehicle crosses the piezoelectric sensors it is compresses and generates a voltage that then
is used to trigger a timer. Two piezoelectric sensors can be used to measure speed. While piezoelectric
sensors can be very accurate, they are also sensitive to excessive vibration. As with the air tubes,
piezoelectric sensors are inexpensive and are secured to the road and are exposed to moving traffic. Thus,
piezoelectric sensors should not be used for this application.
Randy Carroll Page 132

Laser Detectors

Laser technology is typically used in conjunction with digital image capture technology as an alternative to
loop detectors. Laser technology is expensive and provides a level of sophistication that is beyond the
capabilities of the traffic violation database concept. A number of high resolution traffic cameras incorporate
laser technology into the camera itself.

Video Loops

Video loops are a recent development in triggering technology. Video loops make use of a computer to
create a virtual loop on the image of an intersection. Changes in each frame are compared to predefined
screen areas and used to trigger devices such as counters and cameras. These systems are capable of
detecting traffic in multiple locations within the camera’s field-of-view. These systems are marketed for use
in highway management, ramp control, vehicle counts as well as enforcement. The major benefit of using
video loops is that they do not require any cuts to be made in the roadway. The cost of installing video loops
is equivalent to other detection equipment when eight or more lanes or traffic are to be monitored at one
location (29). Thus, these systems can be considered for use with the traffic violation database when the
location being monitored has more than eight lanes.

Inductive Loops

Vehicle detection systems frequently combine a digital loop detector and an inductive loop embedded in the
roadway. The combination of these two components provides a highly sensitive field for detecting vehicles
within the detection zones. A rectangular loop detector zone is typically used to assure the magnetic field
is equal throughout the zone. This configuration has proven successful in detecting and photographing
vehicles that are running a red light or going under or around a rail crossing arm during the red signal
sequence. Thus it holds that this system would be an effective tool in detecting violators. A fully-functional
inductive loop is basically no better than more sophisticated and expensive detection equipment such as
video loops and lasers (30). Thus, this system would be an effective detection device to be used with the
traffic violation database.

This detection system allows the direction and speed of vehicles traversing the loops to be monitored. When
used at highway-railroad grade crossings or intersections, directional sensitivity can be useful in assuring a
vehicle actually committed a violation and not a legal maneuver. Speed sensitivity also allows for a
minimum speed threshold to be defined. This is important to distinguish vehicles that are resting on an
inductive loop or making a right hand turn during a red phase at a signalized intersection from a vehicle
violating the red light. Inductive loops are fairly inexpensive and dependable. However, correct installation
is critical to insure reliable operation. Placement, choice of materials and thorough testing are also very
important to reliable operation.

Inductive loops are usually placed one half to five meters apart from front to back. The distance between
the loops depends upon the anticipated speed of the roadway. Inductive loops should also never be placed
closer than one meter apart from side to side. Inductive loops for use with higher speed traffic are usually
placed further apart. Standard six foot diamond or round loops are not suitable for automated enforcement
detection. These are the standard shapes for traffic monitoring when only simply presence detection is
required. However, smaller loops are more effective for passage detection and thus used for automated
enforcement (31). Thus it is assumed that this would hold true in the traffic violation database concept.
Accurate detection is required to maintain the integrity of the traffic violation database.
Randy Carroll Page 133

Earth Magnetic Loops

Another type of loop available for use with traffic violation database is the Earth Magnetic Loop (EML)
made by 3M. Unlike surface or buried loops a small hole is drilled into the pavement into which the EML
is placed. These may be useful in situations where other loops already exist or extensive cuts are not advised
or feasible. The EML transforms changes in magnetic field intensity into inductance changes which can be
sensed by loop detector units. EML are intended for point detection applications. One, two or three earth
magnetic loops installed across a lane will replace a typical three- turn, six foot wire loop. Earth magnetic
loops will usually not work in the vicinity of large metal objects and thus would not be suitable for
applications near rail lines (29).

Violation Detection - Speed

There are several systems on the market for electronically enforcing speed laws. The most common is radio
detection and ranging (radar). While, radar is the most used detection device in the enforcement of speed
laws, there are other effective methods. One product exists which uses reflected light as a trigger. This
electro-optic technology cannot be detected or jammed. It is also not subject to interference, and can detect
two vehicle speeds when they are next to each other. Another alternative to traditional radar is laser radar.
This system measures speed, distance and direction by using an extremely narrow beam width. Laser radar
provides a higher degree of accuracy than traditional radar systems. The beam from the laser radar can only
target one vehicle at a time and minimizes the possibility of false readings (29).

This is the only application of the traffic violation database that is designed to be mobile. Again, the system
that will be used to detect speed violations is the same as that used in photo enforcement minus the image
capture technology. Typically, photo radar units can be mounted for fixed or mobile deployments. When
the equipment is mounted for mobile use it is typically mounted inside a specially equipped vehicle, on a
tripod or in a specially designed speed trailer.

On-Site Communication

Intersections and Highway-Railroad Grade Crossings

Once the detection equipment is in place and the database is capable of receiving data, a communication link
must be established. When the detection equipment detects a violation a signal must be sent to a processing
unit. This processing unit would essentially be a small computer capable of remote access via a telephone
modem or manual access through a data port. This device would store data related to the violation, including
the time and date of the violation as well as the speed of the vehicle if applicable. This device must also be
wired into the signal controller at an intersection or the gate controller at a highway-railroad grade crossing.
This would allow the computer to determine when the red phase has begun or if the gates had been lowered.
Once this indication is received the computer would start recording violations from the detectors (30). A
simple program would need to be written in order to extract the data from the storage device. This program
would allow the data to be extracted and automatically imported in the correct fields in the database. Most
of the database software listed above has this capability or it could incorporated into a database that was
custom designed.

Speed Monitoring

The detection equipment used to identify a violation must be incorporated into a processing unit. Threshold
speed values must be determined for each location where the device is placed. If the detection equipment
registers a speed greater than the threshold value the computer will record that a violation has occurred, along
with the time, date and speed of the vehicle. Since this detection technology is designed to be mobile the
computer only needs to be manually accessible through a data port. The same program that is written to
Randy Carroll Page 134

extract data from the intersections and highway-railroad grade crossings can be used to extract and import
the violation data into the traffic violation database.

RECOMMENDATIONS

The recommendations for the conceptual creation and use of the traffic violation database are based upon
the information presented throughout this paper. It should be noted that a number of individuals did not
respond to the law enforcement survey. In addition no responses were received to a follow-up questionnaire
regarding implementation of a traffic violation database were also not received. A number of the individuals
contacted in regard to the use and operation of existing automated enforcement systems also did not respond.
Thus, the objectives of this research could not be fully completed in the time available to prepare this report.

In order to develop an operational traffic violation database a law enforcement agency must complete twelve
essential steps. The twelve steps are listed below:

1. Obtain the cooperation and support of city officials,

2. Identify potential monitoring locations,
3. Determine the type of detection equipment needed for each application,
4. Determine the peripheral hardware that must accompany the detection equipment in the field,
5. Determine the software needed to create, operate and manage the violation data,
6. Determine the equipment and process needed to transfer the data from the field to the database,
7. Determine if the hardware installation, operation and maintenance can be done in-house or will the work
need to be contracted out,
8. Install and test all necessary equipment at one site for each type of violation,
9. Assess and complete all necessary changes and complete the full installation,
10. Complete training on the capabilities and proper use of the traffic violation database,
11. Expand the system when feasible to do so,
12. Maintain the equipment.

Each of the twelve steps is important in the successful operation of a traffic violation database. However,
only steps two thru six are within the scope of this research and will be discussed below

Location Identification

The law enforcement survey indicated that the majority of agencies typically establish an intersection, grade
crossing or segment of roadway as a problem location based on accident rate statistics. These agencies also
use current violation rates, public opinion, personal observation and information from the traffic engineering
office. This system could be used to monitor locations that have already been established as problem areas
or to identify locations that are in need of an increased law enforcement presence. The system can also be
used to alleviate or address citizen concerns. A prioritized list of locations for each type of violation
detection system should be created and continually updated.

Violation Detection Equipment

Intersections

Two rectangular inductive loops should be used in all lanes of interest to detect violations of the red phase.
The placement of the loops depends upon the geometry of the intersection being monitored, the speed of the
vehicles traversing the intersection and the presence of other detection devices. The loops should be placed
between one half meter and five meters apart from front to back on center. Two adjacent loops should not
be closer than one meter from side to side on center. Figure 2 illustrates the placement of the inductive loops
Randy Carroll Page 135

on all approaches to an intersection. The downstream inductive loop should be positioned at or near the
stopbar. The exact location will vary from site to site. If there is already a rectangular inductive loop present
on the approach for actuated signal control, only one additional inductive loop may be needed. This decision
depends upon the type of existing loop, its location and function. Laser detection systems and video loops
may be necessary if the existing detection equipment prohibits additional inductive loop installations.

Inductive loops provide accurate detection at a reasonable cost. They are readily available and most city
traffic engineers are familiar with their operation. It is also assumed that most city traffic engineering
divisions and maintenance personnel have installed inductive loops before. However, the installation and
maintenance of the inductive loops may also be contracted out if it is technically or financially warranted.
The degree of sophistication in operation and the cost to purchase detection devices such as a laser or video
loop make them inappropriate for this application, unless the existing equipment prohibits additional
inductive loop installations. The placement of the loops depends upon the geometry of the intersection being
monitored and the presence of other detection devices. Figure 2 illustrates the placement when two inductive
loops must be installed.

Inductance
Loops Stopbar

Crosswalk

Traffic Light
Pole

Figure 2. Red Light Violation Detection System

The processing unit that needs to accompany the detection equipment must be custom made (30). This
device would allow a dial-up modem to download the stored data to the database. This system would also
determine when a detection was a violation. The computer would be wired into the signal control and
become active once the red phase had initiated. A buffer must be programed into the system to allow
vehicles to pass through the intersection under an all red condition without detecting the violation. A one
second buffer is sufficient for use with the traffic violation database. Also, the speed of the vehicle must be
Randy Carroll Page 136

determined and compared with a preset threshold. A speed of 10 mph will be used as the threshold above
which the detection is considered a violation. By allowing a speed comparison vehicles which come to rest
on the inductive loops or make right hand turns will not be counted as a violation.

Highway-Railroad Grade Crossings

One digital loop detector and inductive loop should be used on the downstream side of the crossing for the
highway-railroad grade crossing application. Figure 3 illustrates the placement of the inductive loop in each
travel lane. Depending on the geometry of the crossing, the inductive loop should be placed so that vehicles
driving around the gate do not register as a violation. The loops will detect the violators after they have
already driven around the gates and are proceeding down the roadway. The inductive loops should also be
placed in an area where there are no driveways or turning traffic. The inductive loops are connected to a
processing unit that is wired into the gate controls. It will determine when the gates are down and begin to
record violations. This processing unit must be capable of connecting to a dial-up modem for downloading
the data. It should also have manual download capabilities. As with the intersection application, inductive
loops provide accurate detection at a reasonable cost. They are readily available and most city traffic
engineers are familiar with their operation. It is also assumed that most city traffic engineering divisions and
maintenance personnel have installed inductive loops before. However, the installation and maintenance of
the inductive loops may also be contracted out if it is technically or financially warranted. Video loops and
laser technology are extremely expensive and for this application may not be able to reliably detect
movement through the gates without detecting the train itself.

Gate Arm

Vehicle path as it
passes through the
gate arms

Gate Arm

Buried Loop Detector

Figure 3. Highway-Railroad Grade Crossing Violation Detection System

Randy Carroll Page 137

Speed Monitoring

Monitoring sections of roadways or spot locations should be done using speed trailers. These devices are
truly portable and serve as an educational tool as well. A typical speed trailer is illustrated in Figure 4 below.
The technology used to detect vehicle speeds can be either laser or conventional radar. This depends upon
the preference and the accepted method for measuring speeds in the city of interest. The speed trailer will
have an on-board processing unit that will be used to store the violation data and determine if a vehicle is
in violation of the posted speed limit. A threshold value will need to be entered for each location being
monitored. This value depends upon the law enforcement agency. The processing unit only needs to be
capable of transferring data manually since this device will be moved periodically and an officer can
download the violation data after each placement. Most speed trailers usually have a reader board that
displays the speed of the vehicle as well as the speed limit of the roadway. This serves to educate the driver
as they pass the speed trailer. This is also a key objective of the traffic violation database concept.

Figure 4. Speed Trailer from Montgomery County, Maryland

Traffic Violation Database Software

In order to run the traffic violation database a computer must be available at the law enforcement agency of
interest. This computer may be used for other purposes. However, it is recommended that the computer be
used only to manage the traffic violation data. If an employee of the law enforcement agency is proficient
in the use and creation of databases, Visual FoxPro may be used and training must be provided to everyone
who will be using the software. The task of creating the database should be contracted out if no one is
Randy Carroll Page 138

proficient in the use and creation of databases. The memory requirements for each system will depend on
the number of locations being monitored and the amount of explanatory information that is required. The
computer should be at least a Pentium II, 400 MHz with 64 megabytes of RAM and 10 gigabytes of hard disk
space (30).

TRAFFIC VIOLATION DATABASE APPLICATION

The traffic violation database concept was to be hypothetically applied to the city of College Station, Texas.
Several attempts were made to contact the College Station Police Department to discuss this hypothetical
application. Unfortunately, due to the limited amount of time and resources available to conduct this
investigation, it was not possible to identify a list of problem locations in the city of College Station. Thus,
the hypothetical application can not be tailored to meet the specific needs of the College Station Police
Department. Therefore, the application described in the remainder of this section has been designed to be
more generic in nature, and will be applied to the intersection of Texas Avenue and Southwest Parkway in
the city of College Station. It should be noted that this intersection has not been identified as a problem area,
it is only used as an illustration in the application of the traffic violation database. It should also be noted
that no highway-railroad grade crossing withing the city of College Station had suitable geometry for the
application of the traffic violation database. The following paragraphs describe the process necessary to
implement a traffic violation database in the city of College Station and illustrates the equipment needed to
detect violations at the intersection of Texas Avenue and Southwest Parkway.

The first step in developing the system is to involve local government in the city of College Station. This
will ensure cooperation between law enforcement and city officials. It is then important to determine a
source of funding for the creation and operation of the traffic violation database. One aspect of the funding
issue is to place a monetary figure on the expected increase in deployment efficiency. It is also important
to determine if the detection systems that will be placed in the field can serve a dual purpose.

A list of problem locations must then be compiled. These locations may be identified a number of different
way; accident statistics, personal observation and citizen complaints can all be used. Again, this information
was not obtained. However, once this information is gathered, the type and amount of detection equipment
and peripheral hardware can be determined. The decision must now be made as to whether the system can
be built and installed in-house or if it must be contracted out. This is where the cooperation of the city
becomes important. The city traffic engineering office may have the expertise available to complete the
design and installation. However, the creation of the database software may still need to be contracted out
to a qualified firm. This will ensure the integrity of the data over time.

Once the equipment is purchased installation of the equipment may begin. Again, this can be done in-house
or contracted out. While the installation is being completed, all law enforcement personnel from the College
Station Police Department that will have access to the traffic violation database should complete training on
the capabilities and proper use of the system. A test system should be installed at one intersection and one
highway-railroad grade crossing. One speed trailers will also need to be assembled and deployed in the
field. This should be done to test the design and determine if the processing unit on-site will transfer the data
into the database. The overall system should be checked rechecked and all necessary changes should be
made. Full installation may then begin at all the specified locations. Thorough testing should then be
completed on the entire system to ensure the detection equipment is functioning properly. The data transfer
procedure should also be reevaluated.

The traffic violation database system should now be fully functional and gathering violation data to create
a means by which the College Station Police Department may better monitor the city street system. It is
important that the system be expanded when the resources are available. Expanding the system will assure
the largest possible coverage area and increase deployment efficiency. The equipment should also be
continually maintained to ensure the longevity of the program.
Randy Carroll Page 139

For illustrative purposes detection equipment will be applied to the intersection of Texas Avenue and
Southwest Parkway. All approaches have two thru lanes and a left-turn bay. Two inductive loop detectors
are placed in each lane on all four approaches. The downstream detector is placed with its upstream edge
adjacent to the stopbar. The upstream detector is placed two meters from the downstream detector front to
back on center. A speed of 10 mph will be used as the threshold above which the detection is considered a
violation. By allowing a speed comparison vehicles that come to rest on the inductive loops or are making
a turning movement will not be counted as a violation. The processing unit will be housed in the signal
cabinet and allow for dial-up data downloading through a modem connection. The processing unit will also
be wired into the signal controller to active the appropriate loops during their associated red phase. The
violation detection configuration for Texas Avenue and Southwest Parkway is illustrated in Figure 5.

2 Inductive Loops

Signal Mast Arm

Stopbar

Signal Cabinet

Southwest
Texas Ave

Pkwy

Figure 5. Vehicle Detection Equipment (Texas Avenue & Southwest Parkway)

Randy Carroll Page 140

REFERENCES

1. Automated Enforcement in Transportation - Draft. Institute of Transportation Engineers Committee

96-5, Traffic Engineering Council, December 1998.

2. Martinez, R. Statement of the Honorable Ricardo Martinez, M.D., Administrator, National Highway
Traffic Safety Administration, Before the Subcommittee on Surface Transportation, Committee on
Transportation and Infrastructure, U.S. House of Representatives.
http:\\www.nhtsa.dot.gov/nhtsa/announce/testimony/aggres2.html.

3. Freedman, M., Williams, A.F., and Lund, A.K. Public Opinion Regarding Photo Radar.
Transportation Research Record No. 1270, Transportation Research Board, Washington, D.C., 1990
pp. 59-65.

4. Photo Radar Steering Committee Meeting Minutes. Minnesota Department of Transportation, Roseville,
MN, 1995.

5. U.S. Department of Transportation, Federal Highway Administration , Office of Highway Policy

Information. http://www.fhwa.dot.gov/ohim/1996/

6. Freedman, M. and Paek, N.N. Police Enforcement Resources in Relation to Need: Changes During
1978-89. Insurance Institute for Highway Safety, Arlington, VA, January 1992.

7. Aggressive Driving Facts. State Farm Insurance. http://www.statefarm.com/consumer/crash3.htm.

8. Coleman, J.A., Paniati, J.F., Cotton, Maj. R.D., Parker Jr. M.R., Covey, Lt. Col., Pena Jr. H.E., Graham,
D., Robinson, M.L., McCauley, J., Taylor, W.C. and Morford, G. FHWA Study Tour for Speed
Management and Enforcement Technology. Federal Highway Administration, Washington, D.C.,
December 1995.

9. Popolizio, R.E. New York City’s Red Light Camera Demonstration. Compendium of Technical Papers,
Institute of Transportation Engineers, 1995.

10. Popolizio, R.E. Cameras at Stoplights. Better Roads, Vol. 66, No. 6, June 1996 pp32-36.

11.

12. Plosky, M. Red Light Running Campaign Strategic Planning GuideK. Federal Highway
Administration, Washington, D.C., 1994.

13. Howard County Department of Public Works, Bureau of Highways, Traffic Engineering Division.
http://www.co.ho.md.us/co.ho.md/redlight.htm.

14. Montgomery County Department of Public Works and Transportation, http://www.dpwt.com/redlight/.

15. City of Charlotte, Department of Transportation, The SafeLight Program,

http://www.ci.charlotte.nc.us/citransportation/programs/safelight.html.

16. Photo Radar Detail, Edmonton Police Service. http://www.police.edmonton.ab.ca/photordr.htm.

Randy Carroll Page 141

17. Blackburn, R.R. and Gilbert, D.T. Photographic Enforcement of Traffic Laws. National Cooperative
Highway Research

18. Parham, A. H. and Fitzpatrick, K. Handbook of Speed Management Techniques. Report 1770-2, Texas
Transportation Institute, Texas A&M University, College Station, Texas, September 1998.

19. PhotoCop - Rail. http://www.photocop.com/rail.htm.

20. Fitzpatrick, K., Bartoskewitz, R.T. and Carlson, P. J. Demonstration of Automated Enforcement
Equipment Systems At Selected Highway-Railroad Grade Crossings In Texas. Research Report 2987-2F,
Texas Transportation Institute, Texas A&M University, College Station, Texas, October 1997.

21. Blackburn, R.R. and Glauz, W.D. Pilot Tests of Automated Speed Enforcement Devices and Procedures.
National Highway Traffic Safety Administration, Report No. DOT-HS-806573, Washington, D.C.,
February 1984.

22. Perez-Giese, T. Rush To Judgement. http://www.westword.com/1997/040197/news2.html.

23. Turner, S. and Polk, A.E. Overview of Automated Enforcement in Transportation. ITE Journal, June
1998 pp. 20-29.

24. Interim ITS Fair Information and Privacy Principles. ITS America, Washington, D.C., April 1996.

25. National Motorist Association. http://www.motorists.org/.

26. Society for Safety by Education Not Speed Enforcement. http://sense.bc.ca/.

27. City of Glendale, Traffic Safety Issues.

http://www.ci.glendale.az.us/localgov/transportation/traffic/safety.html.

28. Visual FoxPro, Microsoft Corporation. http://www.msdn.microsoft.com/vfoxpro/.

29. Video Loops. http://photocop.com/red-light.htm.

30. Conversation with Leonard Ruback, Texas Transportation Institute, June 1999.

31. Farradyne Systems, Inc. Technical Memorandum Report: MBL Grade Crossing Safety Program.
May, 1993.
Randy Carroll Page 142

Appendix A - Law Enforcement Survey

As part of the Advanced Institute program at Texas A&M University, I am surveying law enforcement personnel to help determine the process and
equipment needed to create a traffic violation database where automated enforcement equipment and other technologies are used to assist law
enforcement in determining where their limited resources are most needed. This system would not be used to issue citations, record license plate
numbers or identify drivers, but instead would be a way of monitoring driver behavior in critical parts of the city street system. Law enforcement
personnel could access the traffic violation database and search by intersection or street for data relating to red-light running, highway-railroad grade
crossing gate violations or speeding. Copies of this survey may be given to any law enforcement personnel interested in participating. Your time
and comments are greatly appreciated and will be beneficial in determining the usefulness of such a system to law enforcement personnel. Please
fax all completed surveys to Randy Carroll at (409) 845-6481 by July 2, 1999.

Please complete the following questions.

City: State:
Agency: Position:

1. Would this type of system address an existing need at your agency? Yes No If yes, what:

2. Would this type of system become a regular / integral part of your enforcement activities or just used in special situations?

Regular / Integral Special Situations Why:

3. Would this type of system influence decisions on where to conduct enforcement activities? Yes No If yes, how:

4. Where would your agency experience benefits in enforcement efficiency and/or effectiveness from this type of system?

Manpower In what ways: .

Time In what ways: .
Equipment In what ways: .
Other In what ways: .

5. What computer programs and capabilities does your agency currently possess?

DOS Based Programs…. Are you comfortable using this computer format? Yes No
Windows Based Programs…. Are you comfortable using this computer format? Yes No
Internet / Web Browser…. Are you comfortable using the Internet? Yes No

6. Typically, how do you establish an intersection, grade crossing, etc. as a problem location?
Current violation rates Accident Rates Public Opinion Personal Observation

Other:

7. How do you currently make decisions on where and when to deploy your resources?

8. How useful would you expect a system that could historically document traffic violations to be in educating the public or city officials on
law enforcement concerns?

9. How much time per week would you be willing / able to spend manually entering information, that could not be automatically gathered, into
the system?

0 hrs 1 hrs 2 hrs 3 hrs 4 hrs 5 hrs 6 hrs More than 6 hrs
Randy W. Carroll Page143

10. In terms of enforcing traffic laws and regulations, what information would you like to be able to easily access regarding a specific location
or street?

Thank you, your responses are greatly appreciated. If you would be willing to discuss this issue further, please include your name and telephone
number below.

Name:

Phone Number: ________________________________________________________

RANDY W. CARROLL

Randy W. Carroll received his B.S. in Civil Engineering from

Montana State University in May 1998. Randy is currently
pursuing his M.S. in Civil Engineering from Texas A&M
University. He has been employed by the Texas Transportation
Institute as a Graduate Research Assistant since August 1998.
Prior to graduation from Montana State University he was
employed for 2½ years by the Western Transportation Institute,
a research institution dedicated to advancing Intelligent
Transportation Systems in the rural environment. He also
completed a summer internship with the Wyoming Department of
Transportation in 1994. University activities include involvement
in the A&M student chapters of ITS America and the Institute of
Transportation Engineers. His areas of interest include:
Intelligent Transportation Systems, public involvement, and traffic operations.
DETERMINING LOCAL TRAVEL TIME THRESHOLDS FOR USE IN
TRANSPORTATION DECISION-MAKING

by

R. Matthew Estes

Professional Mentors
Wayne K. Kittelson, P.E.
Principal
Kittelson & Associates

and

Thomas Hicks, P.E.

Director
Office of Traffic & Safety
Maryland State Highway Administration

Course Instructor
Conrad L. Dudek, Ph.D., P.E.

Prepared for
CVEN 677
Advanced Surface Transportation Systems

Department of Civil Engineering

Texas A&M University
College Station, TX

August 1999
Matthew Estes Page 145

SUMMARY

Mobility is a complex component of level of service in transportation. There are many different measures
which may be used to describe it, yet few of them are both understood by the public and capable of being
used alone to describe a facility. Travel time has the potential to be a mobility measure that could satisfy
both goals, yet it has not been used in the past because of data collection concerns.

In this paper, a methodology is presented for using travel time data to describe mobility. The method uses
public perception data to help grade mobility based on the responses of surveyed travelers. In this way, the
method links the level of service concept to a specific rating system.

An extensive literature review was conducted for this research, focusing on mobility, travel time, perception
data, and preference models. The results of the literature review were supplemented by interviews with
transportation professionals.

The result of this research is a method for using travel time data and public perception data together to
determine level of service criteria for local decision-making. However, it is recommended that further
research be conducted to determine more suitable ways of obtaining perception data.
Matthew Estes Page 146

TABLE OF CONTENTS

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Identify Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Develop Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

QUALITY (OR LEVEL) OF SERVICE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Accessibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Measuring Quality of Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Volume to Capacity (v/c) Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Load Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Flow Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Reserve Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Percent Time Delay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Delay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Travel Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

TRAVEL TIME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

Related to Goals and Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Clearly Communicate Results to Audiences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Include Urban Travel Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Consistency and Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
Illustrate Effect of Improvements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Applicable to Existing and Future Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Applicable at Several Geographic Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Use Person and Goods Movement Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Use Cost Effective Methods to Collect and/or Estimate Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

OBTAINING TRAVEL TIME DATA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

Test Vehicle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
License Plate Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
ITS Probe Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Non-Traditional Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Extrapolation Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Vehicle Signature Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
Platoon Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Aerial Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

OBTAINING PUBLIC PERCEPTION DATA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

Acceptable/Non-Acceptable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Rating Schemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Matthew Estes Page 147

PREFERENCE MODELS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

Random Utility Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Homogeneous Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Limited Alternatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Utility of an Option . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Maximum Utility Choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
Maximum Likelihood Estimate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
Logit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Nested Logit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Probit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Perception vs. Preference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Use of Perception Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168

METHOD DEVELOPMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

Step 1: Evaluate Need . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Step 2: Collect Travel Time Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Step 3: Collect Public Perception Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Step 4: Employ Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
Step 5: Evaluate Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

CONCLUSIONS AND RECOMMENDATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Matthew Estes Page 148

INTRODUCTION

Mobility is a complex, but very important, component of overall quality of service in traffic applications.
It may be defined as “the ability to reach a destination in a time and cost that is satisfactory” (1). With the
advent of Intelligent Transportation Systems (ITS), the potential to record and use data that measure mobility,
such as travel time, is increasing. If this potential is realized, decision-makers will be able to use up-to-the-
minute mobility data to make decisions about transportation projects. Moreover, the public is increasingly
aware of and interested in transportation issues. Such interest is displayed in the demand for public meetings
to discuss transportation improvements as well as the increased use of traffic information services. People
are interested in not only the effects of large projects, but also the effects of daily congestion on the routes
they use.

With this public interest, it becomes necessary for transportation professionals to be able to communicate
effectively with the public and non-technical decision-makers. The terms used to describe the benefits of
a project need to be terms that the public can understand, yet they also need to be complete in their
description. Travel time is one mobility measure that may be suitable to describe a large number of
transportation projects.

There is also a need to define thresholds for use of any measure in a decision-making process. Thresholds
define the limits of acceptable service and help to prioritize projects. Ideally, thresholds should be based,
to some degree, on the perception of local users, not on the consensus of government or even transportation
officials. A transportation system, just like any other product, must serve the needs of its users, or it will not
be used as intended. The focus of this research is to develop a method for determining travel time thresholds
for use in transportation decision-making.

Objectives

There are two primary objectives of this research. They are to:

• Identify sources of public perception and travel time data, and

• Develop and recommend a method for determining travel time thresholds for use as a decision-making
tool in the evaluation of local transportation system options.

The outcome of this research is a recommended source of these data and a recommended method for
determining local travel time thresholds. This method allows local transportation agencies to calibrate and
use a travel time-based model in decision-making that can be understood by the public and non-technical
decision-makers. The method may also find future use as a component of overall quality of service in the
Highway Capacity Manual.

Scope

The scope of this research is limited because of the short time allowed to complete the tasks. While
investigations were made of the sources of travel time and user perception data, no attempt was made to
actually obtain these data. Instead, the focus is on discovering the types of data that are available for use and
creating a method that can use the available data.

Tasks

Four primary tasks were identified for this research. These tasks focused first on obtaining relevant
information and then on using that information to develop the proposed method.
Matthew Estes Page 149

Literature Review

A state-of-the-art literature review was conducted, focusing on mobility measures, public comprehension of
mobility, and factors that influence local acceptance of travel time. In this literature review, current research
and methods in mobility measurement and public perception of traffic conditions were also identified.
Finally, potential models for use in the development of travel time thresholds were found.

Interviews

In order to further the knowledge gained in the literature review, interviews were conducted of four
transportation professionals interested in mobility measures and travel time data. In addition, these
interviews were conducted in an attempt to discover the sources of travel time and public perception data for
the models that were developed.

Identify Data Sources

As a result of the literature review and interviews, a summary was prepared of the sources of travel time and
user perception data. In addition, recommendations were made regarding the most appropriate sources of
this information.

Develop Models

Finally, models were developed for use in determining local travel time thresholds. Three different models
were explored, and their relative advantages and disadvantages were discussed. Recommendations were
made regarding the most appropriate model for use by local agencies.

Organization

This paper is organized in eight sections. In the first section, an introduction to the research is made. The
second section contains a discussion of quality of service and mobility. In the third and fourth sections, a
discussion of travel time and methods of collecting travel time data is presented. The content of the fifth and
sixth sections is a discussion of perception data and the preference models that were investigated for this
research. In the seventh section, the efforts in model development are documented, and the eighth section
presents the recommendations and conclusions of this research.

QUALITY (OR LEVEL) OF SERVICE

Since the second version Highway Capacity Manual was written in 1965, quality (or level) of service has
been the criterion for evaluating transportation decisions. Level of service was then defined as “a qualitative
measure of the effect of a number of factors, which include speed and travel time, traffic interruptions,
freedom to maneuver, safety, driving comfort and convenience, and operating costs.” (2) As such, level of
service is a rather broad term, encompassing the whole of travel conditions available on a particular route.

Quality of service has two main components. Mobility and accessibility, which are themselves very related
terms, can be thought of as making up the two parts of quality of service. Mobility describes how easy the
transportation system is to use, while accessibility describes how effectively the transportation system joins
types of origins and destinations. Together, these two factors describe overall quality of service.
Traditionally, however, only mobility measures have been used to estimate level of service.

Over the years, a number of different measures have been used to define quality of service. Practically, it
is necessary to have one measure define level of service on a single facility. If more than one measure were
Matthew Estes Page 150

used to define a particular level of service, it would be possible for traffic conditions to be rated differently
depending upon the measure. It would be desirable to use the same measure for all types of facilities, but
this ideal measure has not yet been discovered and implemented. The measures that have been proposed are
either not useful for all facilities or not easily estimated or observed.

Mobility

Mobility is a term that refers to the ease with which people and goods can move. Lomax and Schrank
defined mobility as “the ability to reach a destination in a time and cost that is satisfactory.” (1) Mobility
tends to describe the traffic conditions in the system at a given point in time and space. When travelers
consider their transportation options, mobility affects such decisions as when to leave and which route to use.
Mobility measures answer some of the immediate questions of a traveler. How fast is traffic moving? How
long will the trip take? Is it better to leave early or late? Is there congestion? What is the best route to use?
These questions all focus on the ease with which a traveler can move from one point to another in the
transportation system.

The “satisfactory” part of mobility may vary depending on the time of day and location. Commuters going
to work in a large city may find mild levels of congestion satisfactory. It would not, however, be satisfactory
to find congestion on the same trip late at night. Likewise, travelers might find that speeds of 30-40 mph are
acceptable on an urban collector but not on the freeway. Thus it is important that mobility measures be
specific with regard to time and location of travel.

The opposite of mobility may be thought of as congestion (3). It is possible for a long trip to have high
mobility if it is smooth and vehicles move at or near free speeds. However, a short trip may have low
mobility if congestion and low speeds characterize it.

Accessibility

Accessibility is a term that refers to the ease with which desired activities can be reached from any and all
locations. A more formal definition of accessibility is “the achievement of travel objectives within time
limits regarded as acceptable.” (1) The concept of an acceptable time limit appears in the definitions for both
mobility and accessibility, but accessibility refers more to the availability of activities. Accessibility affects
transportation decisions such as where to go, what mode to use, and whether to travel at all.

Travelers are typically concerned with accessibility before they are concerned with mobility. Accessibility
answers questions about where and how far away activities are. Where is the nearest convenience store?
How long will it take to get to work from this neighborhood? Is there a school nearby? These questions
focus on the relationship between the place of origin and the location of the desired type of activity. An
activity (shopping, working, dining out) is said to be accessible to a location if it can be typically reached
within an acceptable time limit. These time limits vary depending on the nature of the activity.

Typically, accessibility is more difficult to measure than mobility. Land use and travel time measures seem
to capture accessibility concerns most completely. Accessibility is more easily addressed on a large
geographic scale, such as a corridor or a region. Therefore, it does not capture the effects of points in the
system except as they in turn affect the larger geographic scope.

Measuring Quality of Service

Many different measures could be (and have been) used to define quality of service. Since 1965, the
Highway Capacity Manual has used eight different measures to delineate level of service thresholds (2, 4,
5, 6). Others have been proposed that have not been included in the Highway Capacity Manual. Most (if
Matthew Estes Page 151

not all) of these measures are mobility measures. The following subsections describe some of these
measures.

Volume to Capacity (v/c) Ratio

The volume to capacity ratio was one of the first level of service criteria adopted in the Highway Capacity
Manual (2). Estimates of capacity are made by either observing a facility or comparing the facility with
others like it. Similarly, the demand volume on the facility is estimated based on field observations or
planning models. Low v/c ratios are indicative of freely flowing traffic, while high v/c ratios reflect
congested conditions.

The v/c ratio has fallen out of favor with analysts for two reasons. First, it does not directly relate to the
driving experience. As a result, the general public has a hard time understanding the terms that are used as
inputs to the measure. Second, the misuse of terms has caused confusion among analysts. The difference
between demand and actual volume is a subtle concern that is often misinterpreted, with critical results.

Load Factor

In the 1965 Highway Capacity Manual, the load factor was used to evaluate signalized intersections. It is
defined as “the ratio of the total number of green signal intervals that are fully utilized by traffic during the
peak hour to the total number of green intervals for that approach during the same period” (2). Besides being
difficult to calculate and understand, data collection for the load factor was difficult. Estimating the load
factor for proposed intersections required the use of a number of nomographs provided in the Manual. The
next version replaced load factor with delay (4).

The term load factor changed meanings in the 1985 Highway Capacity Manual. It was now used for analysis
of public transit, and referred to the ratio of the number of people on the vehicle to the number of seats
provided. Other factors are also used to describe transit level of service, such as the flow rate of transit
vehicles in dedicated lanes. Transit level of service is especially difficult to estimate because it is comprised
of two components: passenger comfort and traffic conditions. The two measures employed by the Highway
Capacity Manual attempt to capture both sides of transit operations (4).

Flow Rate

Flow rate has been used since the 1965 Highway Capacity Manual to assign level of service to ramp
junctions (2). Ramp junctions are the only facility to use this measure, primarily because so many factors
contribute to their operations. At a ramp junction, a level of service value is assigned to both the ramp and
the freeway based on the flow rates on each. The lowest (worst) of those levels of service is assumed to
govern and is assigned to the junction.

Ramp junctions are difficult to understand because they are complex. The wide range of geometric
considerations requires a large number of equations and/or nomographs for their analysis. In addition, the
behavior of drivers at junctions is not fully understood, preventing more complex or detailed analyses.

Speed

Travel speed is the last level of service criteria used in the 1965 Highway Capacity Manual (2). It has
survived four decades of use and is still used to determine level of service on arterial segments and
uninterrupted facilities (6).
Matthew Estes Page 152

Speed is an advantageous measure because it is applicable at a number of points. Time mean speeds can be
used to assess the quality of service on an uninterrupted facility while space mean speeds can be used to
evaluate interrupted facilities. Speed is also a concept that is readily understood by motorists.

However, speed can also be a difficult measure with which to work. The difference between space mean and
time mean speed is a very important and subtle one that the general public (and traffic analysts, in some
cases!) does not understand. In addition, extensive tables must be constructed to define level of service
thresholds for various facility types. Finally, the use of speed as a level of service criteria might tempt some
planners to increase speed limits to improve service. In so doing, safety may be jeopardized, and the source
of the service problem is not properly addressed.

Density

In 1985, the Highway Capacity Manual began to use density to assign levels of service to freeways and multi-
lane highways (4). In 1997, weaving sections were also evaluated using density (6). Density was chosen
for these analyses of uninterrupted facilities because it is a measure that relates to the driver’s freedom to
maneuver and proximity to other vehicles (4). Density has an advantage over speed (which used to be the
primary measure for these facilities) in that it does not need to be recalibrated for every speed limit.

Unfortunately, density is another term that is not understood by the general public. While drivers may have
some sense of density’s inverse (space headway), they have little feel for density itself. Density is also not
applicable to interrupted facilities, as the measured value changes based on the operations of the opposing
traffic stream or the indications of a traffic signal. Finally, density is a difficult quantity to measure directly.
Loop detectors, which are most commonly used for these measurements, cannot be used to measure density
directly, but instead are used to measure occupancy, which must be converted to density based on
assumptions about traffic composition.

An alternative to using inductive loops to measure density is aerial photography. A Maryland-based firm
specializes in using photographs captured while flying over a city to estimate level of service on both
freeways and arterials. Density is directly measured in an aerial photograph, although the measure is
computed only for the time of the photograph (7).

Reserve Capacity

Reserve capacity was used in the 1985 Highway Capacity Manual to assign level of service to unsignalized
intersections. Reserve capacity is the unused capacity of a lane approaching a stop sign. It seems to have
been an interim measure used while theories on how to estimate stopped or control delay at unsignalized
intersections were being developed and accepted. Strong language was included in the Manual warning
analysts that the levels of service used for unsignalized intersections were not to be associated with the delay
values used to analyze signalized intersections. In the 1994 revision, average total delay replaced reserve
capacity (5).

Percent Time Delay

Percent time delay is used to analyze two-lane highways. This measure reflects the unique condition of these
highways whereby a faster vehicle can experience delay when it is forced to follow a slower vehicle because
passing sight distance is not adequate. It also incorporates to some degree the effects of extended grades
along the highway (4).

In practice, the percent time delay is never actually calculated. The defined threshold values are used to help
define peak-hour volumes that form the thresholds for the various levels of service. These volumes are
calculated specifically for each highway and compared with the actual volume observed or estimated.
Matthew Estes Page 153

Delay

Delay is most often used for estimating level of service at both signalized and unsignalized intersections.
The public easily understands delay, although many drivers’ perception of delay is much greater than that
actually experienced at an intersection. Delay has been adopted by the Highway Capacity Manual for use
with signalized intersections in 1985 and unsignalized intersections in 1994 (4, 6).

Three forms of delay have been proposed to analyze intersections. Stopped delay refers to delay experienced
while a vehicle is actually stopped at an intersection. Control delay includes stopped delay, but also takes
note of the extra delay caused by the need to slow down and start up again. Total delay includes control
delay and any additional delay caused by geometric conditions (i.e., curb radii, length and width of the
intersection, etc.) at the intersection. All three measures can be measured in the field or estimated based on
projected conditions. Currently, the Highway Capacity Manual uses control delay for all intersections (6).

Travel Time

Travel time, like delay, is a measure that is easily understood by the general public. It can be measured based
on field observations or estimated based on projected conditions. However, the Highway Capacity Manual
has never formally adopted it for use in establishing levels of service. One reason that it has never been
adopted is that, until recently, travel time data was not cost-effective to measure in the field. Measurement
was possible, through means such as probe vehicles or license plate surveys, but such studies were expensive
and had to be customized to facility for which travel time data were requested. Another reason is that if
levels of service were to be determined based on travel time, they would have to be customized for the
facility. No procedures exist (other than those presented in this paper) to develop levels of service on an
individual facility based upon travel time.

However, with the advent of Intelligent Transportation Systems (ITS), travel time is becoming a more easily
measured quantity. Vehicle tracking systems such as Automatic Vehicle Identification (AVI) and Automatic
Vehicle Location (AVL) have created a virtual flood of travel time information. Also, many of the other
sensors being placed in the field for other purposes (such as loop detectors and video cameras) have the
potential to estimate travel time in addition to their intended purpose.

TRAVEL TIME

Travel time is a characteristic of traffic conditions that is readily understood by the general public. It is also
a measure that can be applied to a wide variety of facilities. Travel time data are also beginning to be
collected in large quantities in large urban areas. With these characteristics, it is conceivable that travel time
could be used to establish level of service criteria.

Lomax and Schrank identified nine characteristics of an ideal mobility measure (1). These characteristics,
if satisfied, will produce a measure that can be used with confidence to analyze transportation facilities.
Travel time has the potential to satisfy many of these characteristics. In the discussion that follows, these
characteristics are described. In addition, travel time is compared to the demands of each to evaluate its
performance as a mobility measure.

Related to Goals and Objectives

An ideal mobility measure will help further the goals and objectives of a comprehensive transportation
program. It will measure progress toward those goals and also identify areas that have not yet reached their
goal. It should also be able to be continually monitored so that the fulfillment of these objectives can be
documented (1).
Matthew Estes Page 154

Travel time is a measure that is clearly related to many transportation goals and objectives. Low travel times
are generally indicative of an efficient system and high travel times identify areas where either mobility or
accessibility is impeded. The use of ITS-related monitoring systems will allow municipalities to continually
measure travel time as travel conditions change. In addition, the automated systems are relatively low-cost
as compared to using technicians to gather the same amount of data.

Travel time may not be a good indicator of some transportation goals, however. Safety and comfort are not
easily measured by travel time. In fact, the use of travel time alone may result in changes to higher speed
limits, which would be detrimental to both goals.

Clearly Communicate Results to Audiences

The need for an easily understood mobility measure has already been presented. Analysts and engineers are
not the only people who need to understand the chosen measure or how to apply it. Decision-makers and
the general public, many of whom do not have an extensive technical background also need to understand
these measures. This qualification does not relate to the difficulty of calculating the measure, only the ease
of understanding it when it is presented (1).

Travel time is an easily understood measure, because it relates directly to the driving task. When travelers
make a decision about transportation, travel time is nearly always a factor. Travel time communicates to
travelers information about the distance of a trip and traffic conditions along the chosen route. Depending
on the expected travel time, the traveler may choose to change modes, change routes, leave early, leave late,
or even to cancel the trip.

Include Urban Travel Modes

An ideal mobility measure is one that can be computed for any of the various alternative modes available to
a traveler. It should not favor one mode over another by incorporating a bias. It should also be equally
representative of each alternative mode (1).

Travel time is a measure that can be computed for every transportation mode. However, it may not
completely describe the quality of the trip, particularly transit trips. Travel time cannot capture information
about how crowded a bus or subway is; yet, these factors contribute to the overall quality of service.

Consistency and Accuracy

The need for consistency exists on several levels. Similar levels of mobility should have similar mobility
measures. This thought has been the basis for converting mobility measures into a graded scale. Ideally,
each level of service grade would be perceived to result in similar levels of mobility, regardless of the type
of facility. Consistency also relates to precision; a consistent measure should produce similar value when
it is used to measure similar conditions (1).

A consistent measure is of no use unless it is also accurate. Analysts have expounded the need for accuracy
for a long time. In order to achieve a given level of accuracy, a corresponding level of complexity is needed
in the method. It is important to identify that level of complexity, and it is equally important that this level
remain constant.

Travel time is a measure that can be either directly observed with great accuracy or estimated with little loss.
When it is estimated, bounds can usually be developed for the estimate. Direct point-to-point travel time is
a consistent measure primarily because it is easily interpreted and validated.
Matthew Estes Page 155

Illustrate Effect of Improvements

Regardless of the scope of the improvement, the ideal mobility measure should be able to demonstrate the
effectiveness of the improvement. It is important that the scope of the measure match the scope of the
improvement (1).

For many planning or operations improvements, travel time can be used to demonstrate the effectiveness of
the improvement. Changes in travel time can show the effects of retiming a traffic signal as well as the
effects of adding another lane to an existing roadway. It can also be used to evaluate operations
improvements in transit operations as well as other traffic.

Applicable to Existing and Future Conditions

The ideal mobility measure is one that can be estimated as well as observed. This quality makes the measure
useful for analysis and planning. The existence of empirical or theoretical means of estimating the measure
is also indicative of careful study and understanding (1).

Travel time is already being both observed and estimated in transportation systems. The Travel Time Data
Collection Handbook describes a number of ways in which travel time data can be observed in the field (8).
Theoretical models, such as Webster’s delay equations, also result in travel time estimates. Travel time can
also be estimated by simple means using only speed limits and distances.

Applicable at Several Geographic Levels

An effective mobility measure can be applied with equal precision and credibility at a variety of geographic
levels. This quality makes it useful for analyzing points in the system, such as intersections, or wide areas,
such as corridors. It also allows the measure to be used for both macroscopic and microscopic analyses (1).

Travel time is a quantity that can be measured or estimated almost anywhere in a transportation system. As
long as there is a non-zero distance between two points, the travel time between them can be found. Even
at intersections, which are considered to be points in the system, travel time can be found in terms of average
delay. Travel time is also a measure that can be computed microscopically (for individual vehicles) or
macroscopically (for a platoon).

Use Person and Goods Movement Terms

Ideally, a mobility measure will describe travel conditions in terms of people and goods. System efficiency
is best documented in these terms, describing how well or how often the system is used. They also give some
idea of the scope of problems. When problems happen to large numbers of people or goods, the problem is
considered more important. An ideal mobility measure will use terms that help to quantify the magnitude
of concern (1).

Travel time cannot measure system throughput, such as persons or tons of freight moved. Some travel time-
based measures, such as time per person or time per ton, may show some improvement in efficiency, but they
do not capture this goal as well as volume measures do.

Use Cost Effective Methods to Collect and/or Estimate Data

Finally, an ideal mobility measure should not be difficult or expensive to measure. This concern is always
present when budgets are limited. If the data are already available or inexpensive to collect, more can be
Matthew Estes Page 156

done with it. An efficient allocation of resources requires the use of cost effective data collection methods
(1).

In the past, travel time data were relatively difficult to collect. It required a large amount of human labor,
which was expensive. New ITS technologies, however, have significantly reduced the amount of data
collection effort required. Much of this data collection can be performed by a computer, without ever
needing a human operator’s input until the final data are requested (8).

OBTAINING TRAVEL TIME DATA

The Federal Highway Administration (FHWA), in cooperation with the Texas Transportation Institute,
created a document called the Travel Time Data Collection Handbook. In the publication, many of the most
popular and cost-effective methods used for collecting travel time data are reviewed (8).

Travel time is generally composed of two components. Running time is the time the vehicle spends in
motion. Stopped delay time is the time the vehicle spends stopped, or moving sufficiently slow as to be
stopped (usually less than 5 mph). This concept is illustrated in Figure 1.

Stopped
Delay Running
Running Time Time Time

60

50
Speed (km/h)

40
30
20
10

10 20 30 40 50 60
Time (seconds)

Figure 1. Illustration of Running and Stopped Delay Time (8).

Travel time can also be estimated by assuming that a spot speed observed at a particular point is constant
over a short distance (usually 0.5 miles or less). This assumption is most applicable to uninterrupted
facilities (8).

Several methods for measuring travel time are described in the Handbook. In the following sections, a brief
review of each of these methods is presented.
Matthew Estes Page 157

Test Vehicle

The test vehicle technique is the oldest and most commonly used method for obtaining travel time
information. A test vehicle is driven along a specified route and travel times are recorded at checkpoints
along the path. For this reason, this technique is sometimes called the “active probe” technique. The travel
times may be recorded in one of three different ways: manually, using an electronic distance measuring
instrument (DMI), or using global positioning system (GPS) information (8).

Manual data collection requires a passenger to accompany the driver on the trip. As the driver passes the
checkpoints, the passenger records the travel time information. Electronic DMIs have automated this
process, using a computer that receives information from the transmission of the vehicle. The transmission
sends pulses of information to the computer that allow it to compute the speed and distance elapsed since
the last pulse was sent. The GPS method further automates the process by using signals from Department
of Defense satellites to determine the position of the test vehicle. The test vehicle is equipped with a GPS
transmitter and the laboratory computer is equipped with a GPS receiver. One computer can track an entire
fleet of test vehicles using the GPS method (8).

In the test vehicle technique, the driver is an important part of the data collection process. The
aggressiveness of the driver can have a great impact on the results of the data collection effort. Therefore,
three different driving methods have been developed to control variation among drivers.

• Average car: the driver drives according to his/her judgement of an average car in the vehicle stream.
• Floating car: the test vehicle “floats” in traffic by attempting to safely pass as many vehicles as pass the
test vehicle.
• Maximum car: the driver attempts to drive at the speed limit as much as possible unless impeded by other
vehicles or traffic conditions (8).

In practice, some combination of the average car and floating car methods are used because of the difficulty
of keeping track of the number of cars passed.

The advantages of the test vehicle technique include:

• Provides for the determination of driving styles (e.g., “floating car”), which provides consistent data
collection;
• Advanced test vehicle techniques (e.g., DMI or GPS use) result in detailed data that cover the entire
study corridor; and
• Relatively low initial cost.

The disadvantages of the test vehicle technique include:

• Sources of possible error from either human or electric sources that require adequate quality control;
• Advanced and detailed data collection techniques (e.g., every second) can provide data storage
difficulties; and
• The travel time estimates for the corridor are based on only one vehicle that is in the traffic stream,
traveling at one specific time.

By making provisions for quality control, adequate data storage, and a large sample size, the test vehicle
technique can provide an accurate estimate of travel time in a corridor. However, it is often difficult to
obtain the large sample size that is statistically necessary due to budgetary constraints. Other techniques
must be employed if a large sample is needed.
Matthew Estes Page 158

License Plate Matching

License plate matching is a data collection method that is capable of recording a large number of travel times
between observation points. License plate numbers and observation times are recorded at each of the
observation stations. When vehicles are observed at more than one station, they are assumed to have traveled
between the two posts without stopping and the time between consecutive observations is recorded as their
travel time. The basic principle of the method is illustrated in Figure 2 (8).

There are several different methods used to record license plate matching data. The manual method uses
human observers to record license plates and the times they were observed. The portable computer method
automates the time stamp process by using a portable computer to keep track of when each vehicle was
observed. The video with manual transcription technique uses video cameras to record traffic at each site
and human technicians record the license plate numbers and observation times in the office. The video with
character recognition technique uses advanced computer imaging to automatically record license plate
numbers as they are recorded on video (8).

License plate matching has several advantages:

• Able to obtain travel times from a large sample of motorists, which is useful in understanding variability
of travel times among vehicles within the traffic stream;
• Provides a continuum of travel times during the data collection period and ability to analyze short time
periods (e.g., 15-minute averages for continuous data); and
• Data collection equipment relatively portable between observation sites.

License plate matching also has several disadvantages:

• Travel time data limited to locations where observers or video cameras can be positioned;
• Limited geographic coverage on a single day;
• Manual and portable computer-based methods are less practical for high-speed freeways or long sections
of roadway with a low percentage of through-traffic;
• Accuracy of license plate reading is an issue for manual and portable computer-based methods; and
• Skilled data collection personnel required for collecting license plates and/or operating electronic
equipment.

0.8-3.2 km on
arterial streets
1.6-8.0 km on
freeways

Travel time = difference between arrival times.

Figure 2. License Plate Matching Technique (8).

Matthew Estes Page 159

License plate matching is an effective means of capturing many point-to-point travel times. However, it is
limited by the amount of equipment and personnel available for data collection.

ITS Probe Data

Another method for using probe vehicles is to use ITS means for obtaining the data. This method is
sometimes called the “passive probe” technique, in contrast to the “active probe” technique presented earlier
(8).

There are a number of means of obtaining ITS travel time data from vehicles, depending upon the extent of
ITS implementation in an area. Five such means were presented in the Travel Time Data Collection
Handbook: Signpost-Based Automatic Vehicle Location (AVL), Automatic Vehicle Identification (AVI),
Ground-Based Radio Navigation, Cellular Geo-location, and Global Positioning System (GPS). Each of
these technologies rely on devices placed inside the vehicle that allow external tracking systems to monitor
the locations of these ITS probes. An example of an AVI system is shown in Figure 3 (8).

The advantages of ITS techniques are:

• Once the necessary infrastructure and equipment are in place, data may be collected easily and at low
cost;
• Travel time data may be collected 24 hours per day with ITS probe vehicle systems. If the infrastructure
is permanently installed, data are collected as long as probe vehicles continue to travel through the
system;
• Data can be collected electronically; and
• Since data are collected from probes within the traffic stream, the traffic is not influenced by the
experimenter.

Tag I.D. #, Toll Plaza, Sign Bridge,

Central
Time Overpass, or Gantry
Computer
Stamp,
Date
Stamp, READER
Antenna UNIT
Leased phone lines Antenna Transceiver
I.D.
AVI Tag
Tag I.D. #
READER
UNIT

Coaxial Antenna Spacing

Cable, Tag I.D. # Varies, Typically
Radio Wave, or 2-5 km
Microwave

Figure 3. AVI System (8).

Matthew Estes Page 160

The disadvantages of ITS techniques are:

• Probe vehicle systems typically have a high initial cost to purchase necessary equipment, install the
equipment, and train personnel to operate the system and collect data;
• Once the fixed infrastructure of receiving antennas is implemented, it is generally not financially feasible
to make adjustments in the size and orientation of the system coverage area (GPS is the exception);
• The software that performs the data collection tasks are complex programs and are typically designed
in-house or by a consultant;
• Motorists are concerned that if they have a probe they may be more likely to receive traffic citations or
have their travel habits monitored; and
• Probe vehicle systems generally have large implementation costs, and they are most cost-effective for
collecting data within a large study area.

A large, permanent, continuous data collection effort can most effectively be made using ITS techniques.
The high cost of infrastructure improvements makes them unsuitable for single-application uses. In addition,
the high cost requires a commitment to operate and maintain these devices.

Non-Traditional Methods

There are other non-traditional or experimental methods that are used to collect travel time data. Several of
these methods focus on using point vehicle detection equipment (such as loop detectors and cameras) to
either track vehicles or estimate speeds at a location. These techniques are not commonly used in data
collection efforts.

Extrapolation Method

The extrapolation method uses observed point speed measurements to approximate travel time through a
corridor. It is generally assumed that in the vicinity of the observation, speeds are constant. This assumption
allows the use of simple algorithms for the calculation of travel times between detector stations. The most
commonly used instruments are loop detectors and cameras (8).

The advantage of these methods is that they use equipment that is already in place. Loop detectors and
cameras have already been installed in many cities and are used for volume counts and other data collection
purposes. By using existing equipment, the implementation costs are reduced. The disadvantage is that these
methods only estimate travel time between stations. The assumption of constant speed in the vicinity of the
detector may be invalid, particularly during peak periods. If detectors are not sufficiently close to one
another, this may increase the error.

Vehicle Signature Matching

The vehicle signature matching techniques use cameras, loop detectors, weigh-in-motion detectors, or some
other means to identify a vehicle as it passes a detection station. If the same signature is identified at a
downstream detector, it is assumed that the vehicle traveled between the two stations without stopping (8).

This method is able to either use equipment that is already installed in the field, or equipment that can be
easily installed. It has the advantage over extrapolation techniques that it measures real travel times instead
of estimating them. However, signatures may be confused so that not every vehicle that travels between the
two points is correctly matched. Even worse, individual vehicles may be matched with another similar
vehicle resulting in a false travel time.
Matthew Estes Page 161

Platoon Matching

Platoon matching techniques use point sensors to observe platoons of vehicles as they travel along an arterial.
The identifiable characteristics of the platoon (position and spacing of headways, for example) are recorded
and matched to those of the same platoon as it is recorded downstream (8).

Platoon matching offers the same benefits as vehicle signature matching. However, it is susceptible to the
same types of concerns, particularly that when a platoon reaches the downstream sensor it may not be
recognizable. If there are many driveways between detectors or if the detector spacing is too large, the
platoon may break apart before it reaches the downstream sensor.

Aerial Surveys

Aerial surveys typically use fixed-wing aircraft to photograph a particular study area. Although fixed-wing
aircraft are most often used, other aircraft such as balloons, satellites, and remote-controlled gliders have
been proposed. An alternative to photographing an entire area is tracking particular vehicles as they travel
through the network. Most of these methods are only experimental – they have yet to be used and validated
(7, 8).

Aerial surveys have the advantage that they can cover a wide geographic area in a single flight. However,
photographs can only record vehicle density on a highway, which must be converted to vehicle speeds before
they can be used to estimate travel time. Aerial vehicle tracking has potential as a travel time data collection
method, but it has never been tested.

OBTAINING PUBLIC PERCEPTION DATA

Public perception data is not very common in transportation applications. Transportation planners use
preference data to make choices about future growth, but they rarely ask the public to rate transportation
projects. Interviews with several transportation professionals revealed no sources of data on real-time public
perception of traffic conditions, although most thought it would be useful (9,- 12).

If mobility criteria are to be graded in a fashion similar to that used in the Highway Capacity Manual, it is
beneficial to use perception data to help develop the boundary thresholds between grades. Perception data
helps analysts understand how an individual makes a choice. This type of data can also be used to develop
threshold boundaries for grading level of service on individual transportation facilities. The use of
perception data will help planners maintain an equal perception of level of service across many different
facility types in many locations.

The most likely method to be implemented for obtaining these data is some type of survey. Surveys have
been used for planning purposes for many years. Household surveys are very common for estimating origin-
destination patterns. Travel diaries have also been used for these purposes (13).

Obtaining real-time perception data is a unique and necessary challenge if perception data are to be used to
grade mobility. The literature and interviews revealed no methods of obtaining such data. However, the
author submits that it may be possible to collect these data in conjunction with some of the ITS travel time
data collection methods.

The ITS travel time data collection methods automate the process of collecting travel times of individual
vehicles. In most cases, these methods require probe vehicles to be equipped with transponders of some
fashion. This transponder technology could be used to transmit not only the position of a probe vehicle but
also the driver’s level of satisfaction with traffic conditions. A push-button device could be installed in the
Matthew Estes Page 162

vehicle to allow the driver to communicate his perception of current traffic conditions to analysts.
Alternatively, subject drivers could be asked to use cellular phones to inform analysts when conditions reach
an unacceptable level. Motorist safety is a concern for both of these methods, however. This area of data
collection has yet to be explored to its full potential, and further research is required to appreciate the
contributions of perception data.

When data collection plans are made, there is a need to decide what type of data is requested. Perception
data is generally in one of two forms: acceptable/non-acceptable or rating schemes.

Acceptable/Non-Acceptable

If survey responders are asked to rate the system based on whether they feel that their experiences are
acceptable, the data that are collected are binary in nature. There are only two possible responses: acceptable
or non-acceptable.

Using this type of data tends to simplify the problem of data analysis. Less complex methods may be
employed to find the threshold between acceptable and non-acceptable conditions.

Binary data also allows transportation planners to identify areas of primary concern. If a significant
proportion of users find a particular route unacceptable, the route could then be identified as a problem area.
If a route is identified as being unacceptable during unusual congestion (such as that during an incident or
construction), the threshold values obtained may be used to help monitor that route. When conditions worsen
so that they are below the threshold, the area should be submitted as a candidate for improvement.

Rating Schemes

The second option available for data collection is that of a rating or grading scheme. This option is similar
to the level of service grades that are currently used in the Highway Capacity Manual. Individuals would
be asked to rate their experiences on a scale, reporting in real time when their perception of traffic conditions
changes.

Rating schemes are more difficult to analyze, as they require potentially more complex data analysis
techniques. Depending on the model chosen for use, a closed-form solution may not exist.
However, a rating scheme offers advantages over binary data. First, drivers are able to identify similar levels
of service at different facilities. Use of this type of data would allow the construction of a rating scale that
can truly be compared across facilities. Second, there is more detail in this type of rating. By identifying
boundaries between several levels of service, analysts can monitor the degradation or improvement of
conditions at a particular facility.

When using a rating scheme to evaluate travel time data, it is important to pick the allowable responses
carefully. An intuitive approach would be to allow individuals to rate the system on a scale such as “A” to
“F”, or 1 - 10. However, if travel times become so short that they are perceived as dangerous or
uncomfortable, a fast travel time might receive the same rating as a slow travel time. In order to avoid
confusion, the rating scheme should be chosen so as to allow the individual to distinguish between travel
times that are both so fast and so slow as to be uncomfortable.

A good rating scheme would be presented as follows:

Too Accept- Too

Fast Slow
Fast able Slow
Please rate the travel conditions you experienced 1 2 3 4 5
Matthew Estes Page 163

This scheme would allow the surveyed individual to place upper and lower boundaries on safe and
comfortable travel times, as well as give an indication of the level of service on a facility. In order to
understand the range of travel times encompassed by a particular level of service, it would be necessary to
obtain ratings for a number of different observed travel times.

PREFERENCE MODELS

A number of different choice or preference models are used in transportation planning. Among these are the
maximum likelihood estimate (used for critical gap estimation), the logit, and the probit models. In their
basic form, each of these models is a binary choice model: that is, they model the choice between two
mutually exclusive options. The logit and probit models can be modified to include a number of different
choices.

In this chapter, the three models mentioned above are described. A short background in utility theory, which
is necessary to understand the logit and probit models, is also given. Finally, a discussion of the similarities
and difference between perception and preference data illustrates the applicability of these models to
perception data.

Random Utility Theory

In order to use most of the models presented below, the utility of each choice available to the user needs to
be calculated. Utility is a mathematical construct devised for finding the solution to a preference problem,
and is defined as the quantity the user seeks to maximize. Utility is usually thought of as being the personal
benefit derived from the attributes of a particular choice. Positive attributes contribute positive utility while
negative attributes contribute negative utility (13).

There are four postulates in random utility theory. These postulates describe the inherent properties of both
the decision-maker and the options among which he must choose.

Homogeneous Population

Individuals belong to a homogeneous population, act rationally, and have perfect information about their
choices. In other words, individuals always select the option that maximizes their individual net utility,
subject to legal, social, physical, and/or budgetary constraints (13).

Limited Alternatives

Individuals are able to select from a predetermined, limited set of alternatives. There is a particular set A
= {A1, …, Aj, …, AN} of available options from which the individual must choose. In addition, there exists
a set X of measured attributes of individuals and their alternatives. An individual q with a set of attributes
x  X must choose among a set of alternatives A(q)  A. The set of alternatives available to each individual
is based upon the constraints identified in the first postulate (13).

Utility of an Option

Each option Aj  A available to an individual has a unique utility Ujq for user q. The modeler, because he
is necessarily outside of the individual’s decision-making process, does not possess perfect knowledge of
all the factors that contribute to the individual’s choice. Therefore, the utility of an option is thought of as
having two distinct components:
Matthew Estes Page 164

• A measurable, systematic, or representative part Vj which is a function of the measured attributes x; and
• A random part εjq that explains the variation not captured by the measured attributes. This random part
represents the tastes and biases of the user and explains why two individuals with the same choices and
the same measured attributes may make different decisions. It also explains why some users may not
select the “best” alternative given the limited set of attributes considered by the modeler (13).

Therefore, the specific utility of an option Aj to a user q can be written:

UjqVjqεjq Eq. 1

Maximum Utility Choice

It is assumed that each individual will choose the option that maximizes his/her individual utility. That is,
when presented with choices A, the individual will choose option Aj if and only if:
Ujq#Uiq,œAi0A(q) Eq. 2

When dealing with perception data such as that presented earlier in this paper, the overall meaning of these
assumptions is that a surveyed individual will pick the response that most suits his experience. For example,
if a rating of “B” best describes travel conditions to that person, he will pick “B,” not “A” or “C,” if polled.
The individual will not pick a rating that is not given to him (say “G” if the scale was from “A” to “F”).
Finally, two individuals that have the same travel time experience may give different ratings based on other
criteria. These different ratings show up in models as a random departure from the average (13).

Utility theory has no use outside of choice models that can use utility as an input. Two of the models
described below, the logit and probit models, can use utility inputs. The other model, the maximum
likelihood estimate, uses direct observations from individuals without converting them to utility.

Maximum Likelihood Estimate

The maximum likelihood model is the most respected model for estimating critical gaps. Critical gap theory
is based on the assumption that when a driver seeks to enter or cross a traffic stream, he seeks a gap that is
equal to or greater than some minimum size. If the gap is larger than the minimum, it is accepted. The
choice in this situation is whether or not to accept a gap, based on its size (14).

The model is based on the assumption that there is a distribution of critical gaps consistent with a distribution
of drivers. Within the population of drivers, there is a certain amount of randomness that is accounted for
as the aggressiveness of the driver. Because aggressiveness is not a quantifiable variable, it is measured as
a deviation from the mean critical gap of the population as a whole (14).

Another fundamental concept inherent in the maximum likelihood technique is that a driver’s critical gap
is not directly observed. All that can be observed is whether or not a gap of a particular size was accepted.
The critical gap should represent the boundary between accepted and rejected gaps. In practice, the largest
rejected gap is compared with the gap the driver accepted. If no gaps are rejected (i.e., the driver accepted
the first gap, or lag, presented), then the data for that driver are omitted (14).

The means of obtaining the maximum likelihood estimate of a critical gap begins with two variables
describing one driver d:

rd = largest rejected gap; and

ad = accepted gap.
Matthew Estes Page 165

The objective of the maximum likelihood method is to calculate the probability that the critical gap tc is
between the largest rejected and the accepted gap. The modeler must specify a general form of the
distribution of critical gaps fc(t) among drivers in the population. The likelihood that a driver’s critical gap
will be between rd and ad is given by fa(ad) – fr(rd). The likelihood L* within a sample of n drivers for which
a largest rejected and an accepted gap have been observed is:
n

L  k (fa(a d)fr(r d)) Eq. 3
d1

The logarithm L of the likelihood L* is given by:

n
L j ln(fa(a d)fr(r d)) Eq. 4
d1

When L is maximized, so is L*. The values of the parameters describing the critical gap (the mean and
variance) are found by differentiating L with respect to each parameter and setting the partial derivative equal
to zero. Two equations result, the solution of which is found through iterative numerical methods (14).

As a level of service model, the maximum likelihood estimate could be used to estimate the threshold
between acceptable and unacceptable conditions. A suitable number of travelers could be observed to travel
the same route a number of times during a specified time interval (i.e., the morning peak period). The travel
conditions for each trip would be rated by each individual as acceptable or unacceptable. Then, for each
driver, the largest acceptable travel time, ad, and the smallest unacceptable travel time, rd, can be observed.
These terms replace the ad and rd terms in the discussion above, and the result of Equation 4 is the threshold
between acceptable and unacceptable conditions during that travel period.

Logit

Logit choice models have received extensive use in the estimation of modal split. It is the simplest and most
practical discrete choice model, generated by assuming that the random variable ε described earlier is a set
of independent and identically distributed (IID) residuals following a Gumbel distribution. If this assumption
is true, the probability of an individual making a particular choice is:
exp(Viq)
Piq Eq. 5
jA jA(q) exp(Viq)

A linear utility function is employed because it is easy to use and still approximates the true utility function
well. One of the benefits of the logit model is that a number of different variables can be used to determine
the utility of a particular choice to an individual. In this study, however, travel time is the only variable in
use. In addition, the individual can make a choice among several alternatives, such as level of service ratings.

When a rating scheme is used, choices are assumed to be made sequentially based on the values of the
observed variables. In other words, as the results of the utility function change (either increase or decrease),
the choice of a user changes in sequence along the rating scale (say, from “A” to “B” to “C”, not from “A”
to “C” to “B”). The boundary points where an individual would be just as likely to make one choice as
another can be found by setting the probability of the two choices equal:

P1q = P2q Eq. 6

Matthew Estes Page 166

When using the logit model with travel time and perception data, it is necessary to first develop the utility
relationships described earlier. The utility of each rating choice must be established, and then the standard
form of the logit model can be employed to find the threshold boundaries between ratings.

Nested Logit

Nesting is used to distinguish between alternatives that are deemed similar. The need for nesting arises when
it is observed that choices are made between dissimilar options based on different criteria than that used in
selecting between similar options. In these cases, one logit model is used to model the choice between
dissimilar options, while another is used to model the choice between similar options (13).

Consider the following case where a nested logit model is an appropriate choice. An area is serviced by both
subway and bus transit to the same destination. The travel time for each is the same, and so people choose
between modes based on more personal motives (the inherent utility of each mode). Two-thirds of the people
choose the subway; one-third choose the bus.

A second bus line is introduced in the area. It performs exactly the same as the first bus line. It would be
expected that the proportion of people using the subway would stay the same, and the bus passengers would
split between the two bus lines. However, one of the properties of a logit model is called the axiom of
independence of irrelevant alternatives (IIA). This axiom states that:

Where any two alternatives have a non-zero probability of being chosen, the ratio of one
probability over the other is unaffected by the presence or absence of any additional
alternative in the choice set (13).

The ratio of the probability of using the subway over the first bus line was originally equal to two. It would
be expected that a new bus line with exactly the same characteristics as the first would experience the same
ratio. It is also expected that the ratio of two exactly similar bus lines will be equal to one. However, if this
is the case, subway riders will be reduced to half the population, while one-fourth will use the first bus line
and another one-fourth will use the second. This is not a realistic solution.

It is more likely that two-thirds of the population will continue to use the subway and the remaining one-third
will split evenly between the two bus lines. Two logit models are now needed: one to model the choice
between subway and bus, and another to model the choice between the two bus lines.

The IIA axiom is an important property of the logit model that needs to be understood before applying it.
Rating schemes are typically not affected by the IIA axiom because when an additional choice is introduced,
the nature of each choice is redefined.

Probit

The probit model is another model based on the principle of utility. It was used by Manadat, et al., to
develop a rating scheme for comfort on transit buses (15). This model makes the simplifying assumption that
the random variable ε is normally distributed with a mean zero and an arbitrary covariance matrix. This
assumption allows the terms to have any variance, but restricts them to a normal distribution (13).

Only two choices will be used here so that the closed form of the model can be written. Probit models that
incorporate more than two choices are very difficult to write in closed form and are usually solved using
numerical methods (13).
Matthew Estes Page 167

As defined earlier, the probability of choosing Option 1 is:

P1(Θ,Z)Prob[ε1ε2#V1V2] Eq. 7

As a property of the Normal distribution, (ε1 - ε2) is distributed univariate N(0,σε) where:
σε2σ12σ222ρσ1σ2 Eq. 8

Dividing (ε1 - ε2) by σε gives a standard N(0,1) variate. This allows the binary probit choice model to be
written concisely as:
P1(Θ,Z)Φ[(V1V2)/σε] Eq. 9

where Φ[x] is the accumulated standard Normal distribution function that is presented in tabular form in most
statistics textbooks. Probit models for more choices can be developed; however, they cannot be written so
easily in closed form (13).

The probit model is used in a manner similar to the logit model. Once utility functions have been developed,
the probit model can be used to find the travel time where two consecutive ratings have an equal probability
of being chosen. It is at this point where the threshold is defined. The probit model allows for more
flexibility than the logit model; however, it is more difficult to compute, and, as mentioned before, cannot
be written in closed form for a large number of choices. Table 1 compares the three models discussed in this
section.

Table 1. Preference Model Comparison.

Maximum Likelihood Logit Probit

Binary Data Yes Yes Yes
Rating Data No Yes Yes
Utility Input No Yes Yes
Closed Form No Yes Yes, choices <= 3
Software No Yes, needs Yes, needs

The comparison in Table 1 shows that the logit model is the most generally applicable model for the analysis
of perception data. However, any of the models could be employed, depending on the complexity of the data
gathered. Also, the assumptions of each model should be checked prior to employing that model.
Matthew Estes Page 168

Perception vs. Preference

The usefulness of each of these models is heavily dependent on the type of perception data obtained. One
section of this paper is devoted to a description of the means of obtaining perception data. It is important
to note that these models were originally intended to model user preference, not user perception. Typically,
preference data would be used to understand how users make a choice, whereas perception data would be
used to understand how users evaluate the different choices they have. However, when compared, the two
types of data look very similar.

Preference, or choice, data are generally recorded as either the subject user’s stated or revealed preference
for one choice over another. The choices are usually enumerated (Option 1, Option 2, etc.), and each user’s
choice is recorded.

Perception data are recorded as the rating the user would give an individual option. These ratings are usually
enumerated (on a scale of one to ten, for example), and the user’s perception is recorded.

The difference between preference and perception data is that while preference data reveals a user’s choice
given a range of options, perception data reveals the same user’s opinion of that option. In addition, when
the data are enumerated, preference data are always integers, whereas perception data may be continuous.
The following example demonstrates the difference between perception and preference data.

Say a commuter has three options for travel from home to work. The first option is to drive; the second is
to take the bus; and the third is to take the subway. If he drives, the trip will take 35 minutes. If he takes the
bus, the trip will take 40 minutes. If he takes the subway, the trip will take 25 minutes. Assuming he makes
his decision based solely on the trip time, he will choose the subway. Taking the subway is his preference:
he prefers the subway over driving and taking the bus. Now, if he were asked to rate each mode on a scale
of one to ten (ten being the best), it would be expected that he would rate the subway higher than driving,
and driving higher than the bus. However, consider the two hypothetical sets of ratings shown in Table 2:

Table 2. Illustration of Perception Data.

Mode Rating 1 Rating 2

Driving 3 7.5
Bus 1 6
Subway 6 9

The first set of ratings shows that he does not think highly of any of his choices. Riding the bus received a
rating of one, the lowest possible rating. Riding the subway received a rating of six, which suggests that he
barely felt that the subway had an acceptable travel time. Each mode has much room for improvement in
his opinion. The second set of ratings shows a much more satisfied user. Each rating is higher than five,
which indicates that the user thinks the travel times are acceptable. The subway still receives the highest
rating, but the user is much more pleased with his choice. The rating of 7.5 for driving (halfway between
seven and eight) shows how perception data can be continuous.
Matthew Estes Page 169

Use of Perception Data

At first, perception data may seem to be a rather imprecise form of data. The idea of relying on non-technical
sources to categorize technical options may offend some of the technical elite. There are many areas,
however, where the public’s opinion is an invaluable tool. Politicians and businesses use opinion data to find
out what the needs of the public are. Because transportation serves the public, transportation planners ought
to seek the public’s opinion, also.

Surveys have been used in many settings where important decisions need to be made. In transportation
settings, surveys are most often used to collect data about travel patterns. Origin-destination surveys and
travel diaries have been used for many years to predict understand traveler’s choices and prepare for future
growth. Other surveys present travelers with hypothetical situations and ask them to choose a transportation
option (13).

The level of service grades used in the Highway Capacity Manual are intended to reflect the probable
thresholds that users might choose. They are also supposed to bridge the gap between technical
transportation analysts and the largely non-technical public. Level of service is often used as a decision-
making criteria. Projects where the existing conditions exhibit a low level of service are granted higher
priority than projects with a higher existing level of service. Ideally, similar level of service grades should
reflect similar perceived levels of mobility, regardless of the facility type. However, no test has even been
made to verify this assumption.

The assertion is often made that congestion is a relative term — relative to time, space trip purpose, and other
factors. The use of perception data allows transportation planners to evaluate facilities with each of these
variables in mind. While it is unlikely that perception data will change the design period (rush hour), it is
possible that different standards could be employed for different types of facilities in different locations.
Perception data is also consistent across facility types, so a consistent level of service could be maintained
throughout a region.

Several econometric techniques exist for analysis of preference data. With random utility theory as their
basis, these techniques include logit and probit analysis, among others. Logit and probit analyses have found
uses in transportation planning applications, and have many other uses besides (13).

METHOD DEVELOPMENT

In this section, a potential method for combining travel time and public perception data to create a local
decision-making criteria is outlined. The proposed method consists of five steps that, when completed, will
result in a level of service criteria that is based on travel time and public perception data specific to the study
corridor.

Step 1: Evaluate Need

It has been implied throughout this paper that an undertaking to redefine level of service criteria for an
individual corridor or study area can be a difficult and expensive process. There must exist a need for such
a process. This need will most often be expressed in the level of exposure the potential project receives,
either from political bodies, the news media, or concerned citizens’ groups. The level of exposure must be
balanced with the cost of performing such a study to see if it is a worthwhile venture. It is likely that this
method will not be employed except for the largest projects until methods of gathering perception data have
improved.
Matthew Estes Page 170

Step 2: Collect Travel Time Data

Any of the travel time data collection techniques identified in the Travel Time Data Collection Handbook
are applicable for use. Most agencies can continue to use any means that are already employed to collect
travel time data without concern. The most important consideration in selecting the data collection technique
is the geographic scope of the study.

It is important that these data be correlated, at least in time and space, with the perception data. While the
actual travel time of the individual responding to the survey is not required, an estimate of that travel time
should be made.

For large data collection efforts, ITS technologies provide the most reliable means of obtaining both travel
time and perception data. The use of GPS systems, in particular, does not limit the geographic scope of the
study to areas where large investments in infrastructure improvements have been made.

Step 3: Collect Public Perception Data

While the means of collecting public perception data can vary, it is very important that each response be
correlated with a travel time. The most important concern in selecting a data collection method is the
required sample size. Large samples can require great expenses if inefficient data collection methods are
used.

The geographic scope of the study area can greatly influence the necessary sample size. In order to obtain
a representative sample of travelers using a major freeway, it may be necessary to include a large number
of people in the survey. On the other hand, if the geographic scope is limited to a particular intersection, the
survey may be completed using only a small number of participants.

It would be most efficient to conduct a public perception survey in conjunction with some other type of
survey, if it is possible. For example, if a major planning project has proposed to use travel diaries to
estimate origin-destination patterns, the diaries should include a user perception survey as well. These
surveys should be of the form recommended earlier in this paper, grading the facility from “Too Fast” to
“Too Slow” rather than from “A” to “F” or 1 to 10.

Step 4: Employ Model

Using the travel time and public perception data gathered in Steps 2 and 3, the logit model should be
employed to obtain level of service thresholds. If the data collection processes have been carefully controlled
to obtain accurate information, the model should provide threshold values that can be used to evaluate
options for the proposed project.

Step 5: Evaluate Options

Each of the options proposed for the project should now be evaluated using the level of service thresholds
obtained in Step 4. The option that meets the appropriate level of service criteria should be chosen for the
project. This choice should be confirmed by engineering judgement to ensure that the resulting project will
not be excessive in cost or hazardous to motorists.

The result of this process is a set of level of service thresholds that can be applied to one particular segment
of the transportation system. The thresholds can be used to help select the best option, as in Step 5, or to
monitor a facility to see when conditions worsen to the point that an improvement is necessary. If another
facility is deemed similar to the one under study (say, two similar intersections or entrance ramps), the
thresholds can be used at both facilities to save the cost of another study.
Matthew Estes Page 171

CONCLUSIONS AND RECOMMENDATIONS

In this paper, a methodology for identifying local travel time thresholds has been presented. The benefit of
this method is that it is locally calibrated and therefore reflects local conditions. There is no need to rely on
nationally developed guidelines that may have little bearing on the particular transportation facility being
studied.

The use of travel time data in the method allows it to be easily understood by the general public, while
capturing accurately the conditions in the study area. The method also uses direct public perception data to
create thresholds for decision-making. This dependence on public input makes the method more valuable
to both the public and non-technical decision-makers.

Travel time meets many of the criteria for a mobility measurement suggested by Lomax and Schrank (1).
In the past, mobility measurements have been used to develop level of service criteria for evaluating
transportation facilities. Because travel time can be measured on most facilities, it has much potential for
use as a performance measure for defining level of service. As a result, travel time was chosen as the
performance measure for use in this paper.

While much research has been done in the area of obtaining travel time data, relatively little has been done
in obtaining public perception data. An investigation of the means of obtaining this type of data in a cost-
effective manner is recommended for further research.

In addition, further research is recommended for the development of specific forms of the models
investigated in this paper. Software solutions for these models should be developed that would allow
transportation analysts to use travel time and public perception data together to calibrate these models.

ACKNOWLEDGMENTS

This report was prepared for the graduate level civil engineering course, Advanced Surface Transportation
Systems, at Texas A&M University, under the direction of Dr. Conrad L. Dudek. The professional mentors
for this course were Thomas Hicks, Patrick Irwin, Wayne Kittelson, Douglas Robertson, William Spreitzer,
and Carol Zimmerman. The author would especially like to thank Thomas Hicks and Wayne Kittelson for
their encouragement and guidance in writing this paper. The author would also like to thank those
professionals who provided further information for this paper:

Dr. Tim Lomax — Texas Transportation Institute

Dr. Rod Troutbeck — Queensland University of Technology
Mr. Greg Jordan — Skycomp, Inc.
Ross Hasson — Smart Tag

REFERENCES

1. Lomax, T. and D. Schrank. The Keys to Estimating Mobility in Urban Areas: Applying Definitions and
Measures That Everyone Understands, Draft. Texas Transportation Institute, Texas A&M University,
College Station, Texas, January 1999.

2. Special Report 87: Highway Capacity Manual. HRB, National Research Council, Washington, D.C.,
1965.
Matthew Estes Page 172

3. Lomax, T., S. Turner, G. Shunk, et al. NCHRP Report 398: Quantifying Congestion. TRB, National
Research Council, Washington, D.C., 1997.

4. Special Report 209: Highway Capacity Manual. TRB, National Research Council, Washington, D.C.,
1985.

5. Special Report 209: Highway Capacity Manual. TRB, National Research Council, Washington, D.C.,
1994.

6. Special Report 209: Highway Capacity Manual. TRB, National Research Council, Washington, D.C.,
1997.

7. Traffic Quality on the Atlanta Regional Highway System, Draft Report. Skycomp, Inc., Columbia,
Maryland, Fall 1998.

8. Turner, S.M., W.L. Eisele, R.J. Benz, and D.J. Holdener. Travel Time Data Collection Handbook (CD-
ROM). Research Report 07470-1F. Texas Transportation Institute, Texas A&M University, College
Station, Texas, March 1998.

9. Interview with Tim Lomax, Texas Transportation Institute, College Station, Texas, June 7, 1999.

10. Interview with Rod Troutbeck, Queensland University of Technology, Australia, June 7, 1999.

11. Phone interview with Greg Jordan, SkyComp, Inc., Columbia, Maryland, June 23, 1999.

12. Phone interview with Ross Hasson, Smart Tag, Reston, VA, July 8, 1999

13. Ortúzar, J. de D. and L.G. Willumsen. Modelling Transport, 2nd Edition. John Wiley and Sons,
Chichester, England, 1994.

14. Tian, Z., M. Vendehey, B.W. Robinson, W. Kittelson, M. Kyte, R. Troutbeck, W. Brilon, and N. Wu.
Implementing the Maximum Likelihood Methodology to Measure a Driver’s Critical Gap.
Transportation Research Part A 33, April 1999, pp. 187 – 197.

15. Madanat, S.M., M.J. Cassidy, and W.W. Ibrahim. Methodology for Determining Level of Service
Categories Using Attitudinal Data. In Transportation Research Record 1457, TRB, National Research
Council, Washington, D.C., 1994, pp. 59 – 62.
Matthew Estes Page 173

R. MATTHEW ESTES

Matthew Estes received his B.S. in Civil Engineering from

Virginia Polytechnic Institute and State University in May 1998.
Matthew is currently pursuing his M.S. in Civil Engineering from
Texas A&M University. He has been employed by the Texas
Transportation Institute as a Graduate Research Assistant since
August 1998. During the final two years of his work at Virginia
Tech, Matthew was employed as a site development technician
by Anderson and Associates, Inc., a civil engineering firm in
Blacksburg, Virginia. University activities include active
involvement with the A&M chapter of ITS America where he
serves as chapter president, and with the Institute of
Transportation Engineers. Other memberships include the
American Society of Civil Engineers, Chi Epsilon Civil
Engineering Honor Society, Phi Kappa Phi Honor Society, and Golden Key National Honor Society. His
areas of interest include traffic operations, transportation system management, multimodal ITS applications,
and signal systems.
ITS APPLICATION FOR SEVERE WEATHER CONDITIONS

by

Scott Harry

Professional Mentor
Bill Spreitzer

Course Instructor
Conrad L. Dudek, Ph. D., P.E.

Prepared For
CVEN 677
Advanced Surface Transportation Systems

Department of Civil Engineering

Texas A&M University
College Station, TX

August 1999
Scott Harry Page 175

EXECUTIVE SUMMARY

Adverse weather and other harsh environmental features are common throughout the United States due to
the country’s vast size and diverse range of geographic characteristics. Severe winter weather, involving ice
and snow occurs throughout the northern and central states and in mountainous areas of southern regions.
The impacts of these conditions on highway users are considerable. Statistical analyses have shown that a
disproportionate number of accidents and fatalities occur under adverse environmental conditions. It has
been estimated that between 25 and 35 percent of all interurban accidents occur during adverse weather
conditions.

Due to the high frequency of accidents during severe weather conditions, and the increasing maintenance
costs related to these conditions, several systems have been developed to predict severe weather conditions.
The objectives of this research were to evaluate the current state of practice for dealing with severe weather
conditions, make recommendations as to which sensors or weather prediction systems work best in different
situations, and make recommendations on the implementation of Road Weather Information Systems (RWIS)
in the San Antonio, Texas area.

To accomplish these objectives a literature review was conducted encompassing all ITS applications for
severe weather conditions. Several phone calls and personal interviews were also conducted due to the lack
of published material in the specific area of pavement sensors.

From the findings it was recommended that the active pavement sensors be used in areas where high
concentrations of salt are regularly present (greater than four percent concentration). In this situation the
Climatronics FRENSOR sensor is recommended. In areas where typically low concentrations of salt are
present, less than four percent concentration, it is recommended that passive pavement sensors be used. The
Surface System, Inc. (SSI) passive pavement sensor is recommended under these conditions. The most
commonly used Road Weather Information System (RWIS) in the United States seems to be the SSI system.
In situations where this system is present in neighboring states or countries, it is recommended that this
system be implemented for weather prediction purposes so that information can be shared across boundaries.
It was also determined that San Antonio, Texas is a good candidate for implementation of an RWIS.
Scott Harry Page 176

TABLE OF CONTENTS

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178

PROBLEM STATEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178

OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179

SCOPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179

BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Road Weather Information Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
ATWIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Colorado Weather Prediction System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Road Weather Information System- Nevada . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Test & Evaluation Plan for the Effectiveness of RWIS-Based Snow and Ice Control
Practices - Iowa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Washington RWIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Idaho Storm Warning System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Lake Tahoe Weather Information System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Sierra Project - “Snow Wars” - California . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
State Route 22 (Crescent Mountain) Visibility Project - Pennsylvania . . . . . . . . . . . . . . . . . . 183
Travel Aid - Washington . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Variable Speed Limit Project - Nevada . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
Visibility Warning Project - California . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Infrared Pavement Temperature Sensor on Maintenance
Trucks - Indiana/Vermont . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Thermal Mapping of Road Temperature - New York . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Off the Shelf RWIS for Rural ITS Applications - California . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
The Traffic Management Policy of Variable Message Signs for Weather-Controlled
Road Between Pyhtaa and Hamina, Finland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Effects of Weather Controlled Variable Speed Limits and Warning Signs on Driver
Behavior - Finland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Examining the Institutional Issues Related to Implementing Road and Weather
Information Systems (RWIS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Assessment of Road Surface Freezing Point Sensors for the UK . . . . . . . . . . . . . . . . . . . . . . . 187
Test of Road Weather Monitoring Systems and Sensors - Sweden . . . . . . . . . . . . . . . . . . . . . . 188
Environmental Sensor Systems for Safe Traffic Operations . . . . . . . . . . . . . . . . . . . . . . . . . . 189
Fog Warning Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Fog Warning System - Georgia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Fog Warning System - Utah . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
I-526 Fog Mitigation Project - South Carolina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
I-75 Fog Detection/Warning System - Tennessee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Avalanche Warning Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Technical Applications in Avalanche Hazard Mitigation - Utah/Wyoming . . . . . . . . . . . . . . . 191
Development and Trial Deployment of an ITS Avalanche Hazard Management
System - Utah, Wyoming, Idaho . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

ROAD WEATHER INFORMATION SYSTEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192

RECOMMENDATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
Scott Harry Page 177

RWIS IMPLEMENTATION CRITERIA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

Funding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Staffing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Partnerships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Expandability, Transferability, and Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

SAN ANTONIO, TEXAS CASE STUDY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

Funding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Staffing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
Partnerships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
Expandability, Transferability, and Compatibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Scott Harry Page 178

INTRODUCTION

The safety, mobility, and economic impacts of weather on transportation are considerable. The White House
Office of Science and Technology Policy (OTSP) estimates that weather causes or contributes to 6,000
fatalities on U.S. highways each year. According to the General Estimate System crash database, in 1995
there were approximately 435,000 injury crashes in the United States due to adverse weather. Despite the
fact that North America spends more than $2 billion a year on snow and ice control, road accidents increase
during adverse weather by a factor of from two to five (1).

Through the use of weather prediction systems not only do agencies have the ability to decrease accidents,
they also have the ability decrease the amount of chemicals applied to roadways. Ice prevention, or anti-
icing, requires 30-75 percent less chemical treatment than does deicing, and is much more cost-effective in
snow and ice control (2). Pinet & Associates suggest that benefit-to-cost ratios resulting from the
implementation of Road Weather Information Systems (RWIS) can be as high as 11 to 1 (3).

The most popular weather prediction systems in use throughout the United States and several other countries
are the Road Weather Information Systems (RWIS). RWIS are defined by the federal Weather Summary
and Synthesis Report as “a combination of technologies and decision making that uses weather information
to apply labor, equipment and materials as cost-effectively as possible (2).” In general, the RWIS consists
of on-site temperature and atmospheric sensors, data processing units, telecommunication capabilities, and
workstation displays. The information typically output from these systems includes air temperature, wind
speed and direction, humidity, precipitation, dew point, visibility, and road condition (wet, dry, icy).
Pavement temperatures are obtained from solid-state electronic devices installed in the roadway. Surface
sensors can either be active or passive, but passive sensors have gained more wide-spread acceptance with
over 2,000 installed in the U.S., Canada, the UK, Europe, and Iceland (2).

Heavy snowfall not directly attributable to winter storms can also contribute to roadway accidents and
damage if avalanching occurs above the roadway. Several projects around the world are in the process of
testing and evaluating avalanche detection systems. These systems not only have the ability to make
conditions safer for travelers, but also for the maintenance personnel working near avalanche prone areas
(9).

Fog warning systems are also being tested extensively throughout the country. These systems go above and
beyond the simple visual detection provided by the RWIS systems. For example, some of these systems
incorporate video cameras and other visual detection technology that better predict the presence and intensity
of fog (4).

PROBLEM STATEMENT

Many weather prediction systems throughout the world have been implemented trying to predict severe
weather conditions before they occur. However, many of these projects are still in the developmental stages
and their progress is unknown. Research is needed so that these weather prediction systems can be expanded
to their full potential. Through the use of new technologies the reactive manner of dealing with severe
weather conditions can be changed into a more proactive response which will decrease maintenance costs
and make for a safer driving experience.
Scott Harry Page 179

OBJECTIVES

The major objectives of this research were to:

• Identify field operational test projects related to the use of ITS technologies for severe weather
conditions;
• Identify the reliability and accuracy of these technologies, specifically in-pavement sensors;
• Make recommendations on their use and effectiveness; and
• Make recommendations on the feasibility of these technologies to a traffic management program
(Transguide) in San Antonio, Texas.

SCOPE

The scope of this research included all ITS technologies that have been developed with regards to severe
weather conditions. These weather conditions include: wind, low visibility conditions (fog), avalanches, ice
and snow. Pavement sensors reviewed included both in-pavement and out-of-pavement sensors.

BACKGROUND

Several projects around the world are being conducted with the objective to predict severe weather conditions
before they occur. The projects reviewed are summarized into three categories: Road Weather Information
Systems (RWIS), fog warning systems, and avalanche warning systems.

Road Weather Information Systems (RWIS)

The projects reviewed in this section are briefly summarized in Table 1. The full description of the
projects are also given in this section.

ATWIS

North and South Dakota are developing an Advanced Transportation Weather Information System (ATWIS)
for several of their interstate highways. Weather forecasting and analysis technologies are used to produce
short-range site-specific weather forecasts using road and weather sensors and weather satellite observations.
Nowcasts and forecasts are made subjectively, by people interpreting the data, and by a prediction model
developed for the project that is run on a supercomputer. Forecasts are updated every three hours and cover
a six hour period. The electronic forecast is converted to a speech file so that travelers can access it through
cellular phones. Maintenance departments in both states receive forecasts daily through electronic mail.

Three types of data collection technologies are used to maximize data collection: Surface Systems, Inc. (SSI)
weather stations, Geostation Observation Weather Satelite (GEOS) observations, and an agricultural weather
system. Road and weather data are collected from approximately 30 weather stations located throughout
North and South Dakota. Data collected include: pavement surface and subsurface temperature, air
temperature, wind speed and direction, and pavement status (dry, wet, amount of chemicals on surface).
GEOS provides surface and upper air observations and weather predictions. The agricultural mesonet is used
to obtain the following data: air temperature, moisture, wind speed and direction, radiation, and precipitation.
Data from the weather and pavement sensors and the mesonet are transmitted over telephone lines (4).
Scott Harry Page 180

Table 1. Road Weather Information Systems (RWIS) Summary

PROJECT NO. OF WEATHER LOCATION MANUFACTURER

STATIONS
Advanced Transportation 30 North and South Dakota Surface Systems, Inc. (SSI)
Weather Systems (ATWIS)
Colorado Weather Prediction 100 sensors Colorado Surface Systems, Inc. (SSI)
System
Road Weather Information 22 Nevada SSI, Coastal Environment,
System and Vaisala
Test & Evaluation Plan for the 22 Iowa ?
Effectiveness of RWIS-Based
Snow and Ice Control Practices
Washington RWIS 200 Washington Surface Systems, Inc. (SSI)
Idaho Storm Warning System ? Idaho Santa Fe Systems, SSI, and
Handar
Lake Tahoe Weather ? Nevada ?
Information System
Sierra Project - “Snow Wars” 4 California Surface Systems, Inc. (SSI)
State Route 22 (Crescent 1 Pennsylvania ?
Mountain) Visibility Project
TravelAid 6 Washington Surface Systems, Inc. (SSI)
Variable Speed Limit Project 22 Nevada SSI, Coastal Environment,
and Vaisala
Visibility Warning Project 10 California Surface Systems, Inc. (SSI)
Infrared Pavement Temperature ? Indiana and Vermount Control Products, Inc.
Sensor on Maint enance Trucks
Thermal Mapping of Road ? New York Vaisala, Lockheed Martin
Temperature
Off the Shelf RWIS for Rural ? California ?
ITS Applications
The Traffic Management Policy ? Finland ?
of Variable Message Signs for
Weather-Controlled Road
Between Pyhtaa and Hamina,
Finland
Effects of Weather Controlled ? Finland ?
Speed Limits and Warning
Signs of Driver Behavior
Examining the Institutional Does not apply (–) Nationwide –
Issues Related to Implementing
Road and Weather Information
Systems (RWIS)
Assessment of Road Surface – United Kingdom ?
Freezing Point for the UK
Test of Road Weather – Sweden Many
Monitor ing Systems
Environmental Sens or Systems – Nationwide Many
for Safe Traffic Operations

Colorado Weather Prediction System

Colorado is using Surface Systems, Inc. (SSI) weather stations. Approximately 100 road sensors throughout
the state collect road condition data every 15 minutes that is transmitted to remote processing units (RPU’s).
Scott Harry Page 181

All pavement sensors monitor pavement temperature; newer sensors also monitor chemical concentration
on the roadway and road refreezing temperature. Weather data collected includes air temperature, relative
humidity, wind speed and direction, and whether there is precipitation. The newer weather monitors also
collect data on precipitation accumulation and rate, traffic counts, and visibility. These data are used by SSI
to prepare location-specific road and weather forecasts.

The Hanging Lake Tunnels division in Glenwood Springs, can access the RPU’s in their jurisdiction in real-
time. This allows the maintenance staff to provide current conditions to travelers, including commercial
vehicle operators in the northwest portion of the state (4).

Road Weather Information System - Nevada

Twenty-two weather stations have been installed throughout the state of Nevada. The sensors monitor air
temperature, wind speed and direction, humidity, precipitation, dew point, visibility, road condition (wet,
dry, ice), pavement surface and subsurface temperature, and the amount of chemical on the roadway. The
weather stations are from three different providers: Surface Systems, Inc. (SSI), Coastal Environment, and
Vaisala.

Thermal maps have been prepared for 275 miles of highway in conjunction with pavement temperature
measurements. The thermal maps, which remain the same over time as long as no changes are made to the
topography or to the vegetation or buildings along the highway, provide for more efficient winter road
maintenance.

The Nevada DOT developed a Performance-Based Training Package to educate maintenance employees on
how to use the data provided by the Road Weather Information Systems (RWIS). All personnel who would
use the RWIS to make decisions are trained on its use.

SSI provides localized pavement and weather forecasts based on data obtained from its weather monitors.
Data from the Coastal Environment and the Vaisala weather stations are integrated into one system, allowing
the NV DOT staff to use LEADS software, developed by Lockheed Martin, to prepare weather and pavement
forecasts using data collected from both of these stations. Atmospheric forecasts are provided by Northwest
Weathernet, a local provider.

Maintenance personnel can access the data and forecasts by computer. Based on forecasts, weather
advisories are issued through variable message signs (VMS), highway advisory radio (HAR), and a local
television station (4).

Test & Evaluation Plan for the Effectiveness of RWIS-Based Snow and Ice Control Practices - Iowa

The Iowa DOT developed this plan with two main goals: 1) demonstrate how snow and ice control decisions
are most effectively performed using various components of Road Weather Information Systems (RWIS),
and 2) evaluate the extent to which the benefits predicted by research are actually achieved. Weather data
will be obtained from the state’s 22 RWIS sites, as well as other public and private sources.

This plan was developed to evaluate the existing RWIS in Iowa to determine deficiencies and develop an
approach to integrated RWIS implementation. Existing weather data sources will be identified and
evaluated, including the 22 weather stations installed throughout the state. Three locations in Iowa were
selected to be used as test and evaluation (T & E) sites. For these locations, detailed pre-, during, and post-
storm snow and ice control decisions and actions will be monitored to determine their effectiveness. Also,
an effort will be made to document actions that would have been taken without RWIS. The project also
includes a review of weather forecast services to determine their appropriateness (4).
Scott Harry Page 182

Washington RWIS

Approximately 200 weather stations are located along roadways throughout Washington State. The types
of data collected at the stations include air temperature, wind speed and direction, and visibility. Most of
the weather stations are manufactured by Surface Systems, Inc. (SSI).

The weather stations are used by maintenance personnel throughout the state; however, the stations are not
connected to each other. Each maintenance section uses its own monitors. The Washington Department of
Transportation is working on connecting all of the weather stations so all maintenance sections can access
data from any weather station. The state is also working to put the weather data on an Internet web page (4).

Idaho Storm Warning System

Road, weather, and visibility data are collected using three different sensors which are located at the same
site. Data from the sensors are downloaded to a master computer every five minutes. The objectives of the
project are to evaluate and compare the performance of the three sensors. A video camera has been set up
at the site to assist with sensor evaluation. If the visibility level observed by any of the sensors falls

below a threshold value, an alarm is activated. Changeable message signs are used to warn motorists of the
low visibility. The data are not currently being used for forecasting. The following sensors are being
evaluated (4):

• LIDAR (Light Identification and Ranging) manufactured by Santa Fe Systems-. This sensor uses a laser
to measure visibility. While this sensor is more expensive per unit than the other two sensors, only one
sensor is needed to monitor the visibility of a path up to 10 kilometers.
• SCAN sensor manufactured by Surface Systems, Inc. (SSI)- The SCAN sensor monitors weather and
pavement conditions, including visibility, wind direction and speed, air temperature, precipitation, and
pavement temperature.
• Handar sensor- The Handar Sensor monitors weather conditions only, such as visibility, wind direction
and speed, air temperature, and precipitation.

Lake Tahoe Weather Information System

The goal of this project is to use an advanced weather information system to reduce the use of sand, salt, and
de-icing chemicals in the Lake Tahoe Basin, which includes parts of Nevada and California. The de-icing
materials enter the lake and deteriorate water quality. Weather stations, installed by Nevada DOT (NDOT),
provide early warning of changing conditions. The advanced warnings allow NDOT to warn travelers of
potentially hazardous weather conditions and allow road crews to place salt, sand, and other de-icing
chemicals in a more timely manner. The weather station information has been used on a small scale around
Lake Tahoe. Funding for this project will allow for expansion of the system Basin-wide (4).

Sierra Project - “Snow Wars” - California

Four Surface System, Inc. (SSI) weather stations, which include pavement sensors, collect data on weather
and roadway conditions. Weather data include air temperature, humidity, wind speed and direction, and
visibility. The pavement sensors collect data on surface and subsurface temperature, pavement condition
(wet, dry, icy), and the amount of de-icing chemical on the roadway. The monitors are located at altitudes
above 4,000 feet. Traffic detectors are also used to monitor traffic speeds. The Kingvale Snow Management
Control Center collects the weather and traffic information.

SSI provides localized pavement and weather forecasts to the Kingvale Snow Management Control Center
using the data obtained from the weather stations. In addition, weather forecasts for the region are obtained
Scott Harry Page 183

from a local weather forecasting service, DTN. DTN monitors are located in the highway maintenance
dispatch rooms of the Kingvale Snow Management Control Center.

Based on forecasts and traffic information, the Kingvale Snow Management Control Center issues traffic and
weather advisories, if warranted, through variable message signs (VMS), highway advisory radio (HAR),
and the California Highway Information Network (CHIN) (4).

State Route 22 (Crescent Mountain) Visibility Project - Pennsylvania

In order to improve safety on the section of State Route 22 that crosses Crescent Mountain, a weather station
was placed along the roadway to monitor weather conditions. If road safety is compromised by inclement
weather (snow, ice, or fog), variable message sign (VMS) are manually programmed and activated to alert
drivers to the diminished roadway conditions. Two VMS are used, one located at each approach to the
mountain (4).

TravelAid - Washington

Weather and pavement sensors monitor conditions along the highway through the Snoqualmie Pass. Six
weather stations monitor air temperature and wind speed and direction; only two or three of the weather
stations also measure visibility. The pavement sensors monitor pavement conditions and temperature. Radar
traffic detectors are also positioned along the 40-mile stretch of highway to monitor vehicle speeds. Based
on weather and pavement conditions and travel speeds, highway maintenance personnel can reduce speeds
using variable speed limit signs (VSL), and alert travelers using highway advisory radio (HAR), variable
message signs (VMS), and in-vehicle units (IVU) radio receivers with alphanumeric displays.

Surface Systems, Inc. (SSI), the manufacturer of the weather stations and pavement monitors, will provide
pavement and weather forecasts to WA DOT using the data obtained from the monitors. In addition, the
National Oceanic and Atmospheric Administration (NOAA) has weather sensors and provides the state with
its forecasts, concentrating on avalanche potential. NOAA will incorporate data from the SSI sensors along
with data from its own sensors.

The state will distribute approximately 200 IVU to state police patrols, commuters, and commercial vehicle
operators. The IVU are called Traffic Master and will provide information such as weather, revised speed
limits, incidents, and whether chains are necessary, as well as a map showing areas of reduced conditions.
The idea of using Traffic Master in-vehicle devices is not cutting edge; however, the concept of providing
in-vehicle units with weather information to travelers is innovative and not widely practiced (4).

Variable Speed Limit Project - Nevada

Twenty-two weather stations have been installed throughout the state. The sensors monitor air temperature,
wind speed and direction, humidity, precipitation, dew point, visibility, road condition (wet, dry, icy),
pavement surface and subsurface temperature, and the amount of chemical on the roadway. One of the
weather stations is located along a portion of highway that began experiencing increased incidences of
reduced visibility and reduced road surface conditions (icing) as a result of a power plant expansion. At this
location, variable speed limit signs and warning signs are being installed to warn drivers of diminished road
and visibility conditions.

Vehicle speeds will also be monitored along I-80, 15 miles east of Reno. The 85th percentile speed will be
calculated based on prevailing traffic conditions. Data from the RWIS will then be used to calculate the
appropriate speed from current weather conditions. This speed will be shown on the variable speed limit
signs. If speed limits are reduced, signs will be activated that read “Reduce Speed When Flashing.”
Scott Harry Page 184

Data from the weather station are sent to the central processing unit, allowing the NV DOT staff to use
LEADS software, developed by Lockheed Martin, to prepare weather and pavement forecasts. Maintenance
personnel can access the data and forecasts by computer (4).

Visibility Warning Project - California

Ten Surface Systems, Inc. (SSI) weather stations, which include pavement sensors, collect data on weather
and roadway conditions. Weather data includes air temperature, humidity, wind speed and direction,
precipitation, dew point, and visibility. The pavement sensors collect data on surface and subsurface
temperature and pavement condition (wet, dry, icy).

SSI provides localized pavement and weather forecasts based on data obtained from the monitors.
Maintenance personnel can access the data and forecasts by computer. Based on forecasts, weather
advisories are issued through variable message signs (VMS).

Closed circuit television (CCTV) cameras are located at two of the weather stations along Interstate 5 to
verify visibility readings from the monitors. This portion of the highway is prone to dust storms in the spring
(4).

Infrared Pavement Temperature Sensor on Maintenance Trucks - Indiana/Vermont

The goal of this project is to increase the efficiency of applying anti- and deicing materials to the road
surface. An infrared (IR) sensor is installed on maintenance patrol trucks to monitor pavement temperatures.
As the truck travels at highway speeds, continuous pavement temperature readings are displayed inside the
truck cab. Material application rates are based on pavement temperature and maintenance staff estimates
of the ice or snowpack thickness. This system has been used in Vermont and Indiana. Vermont’s sensor
supplier is Control Products, Inc. (4).

Thermal Mapping of Road Temperature - New York

A thermal profile of roadway sections will be developed based on road surface temperature readings obtained
from sensors in the roadway. This will provide maintenance personnel with information on which areas of
the roadway are prone to freezing before other areas. This information can be used to provide more efficient
winter road maintenance.

Two technologies will be used to create thermal maps of about 130 miles of highways in the state. The
results from the two methods will be compared, and the accuracy of both methods will be evaluated against
actual pavement temperatures.

Conventional thermal mapping will be performed by Vaisala. Their method involves conducting a route
survey to identify characteristics of the roadway which may affect the temperature, such as trees or tall
buildings. A vehicle with an infrared sensor will then be used to collect pavement temperature data.
Approximately five thermal surveys will be made for each weather category (calm and clear, windy and
cloudy, and intermediate conditions). Thermal maps, which will be prepared for each weather category, will
be provided in paper format and in digital form for use on a PC.

Lockheed Martin will perform thermal mapping using innovative technology that is considerably less
expensive per mile than Vaisala’s method. An infrared sensor and a global positioning system (GPS) will
be used to record thermal data. Pavement temperature and outside air temperature will be recorded with a
Sprague Controls ROADWATCH road and air temperature sensor mounted on the truck. A Trimbel SveeSix
GPS receiver will record latitude and longitude. Approximately five surveys will be made for each weather
category (damped, intermediate, and extreme). Temperature differences for each segment and weather
Scott Harry Page 185

category will be prepared and entered into the Modified Central File Server. The data can be viewed using
PC-based software (4).

Off the Shelf RWIS for Rural ITS Applications - California

WELS Research developed a two-part software package that provides detailed forecasts of surfaceweather
conditions customized for specific regions. Part I uses National Weather Service (NWS) raw observational
weather data to produce 24-hour weather forecasts for the U.S. The software also produces forecasts for
subareas (for software users). The subarea forecasts are produced twice daily. Part II of the software is
accessed by the users. It provides a user-friendly, interactive display of the forecasts.

This software is unique in two ways. The first is that it provides an interactive graphics user interface which
allows users to input actual conditions and change forecasts. The second innovation of this software is that
the forecast algorithm, which provides results similar to that of the NWS, runs on a desktop personal
computer (with a pentium processor) whereas the NWS program runs on a supercomputer (4).

The Traffic Management Policy of Variable Message Signs for Weather-Controlled Road Between Pyhtaa
and Hamina, Finland

The Finnish National Road Administration has used road transport telematics on a weather-controlled road
to reduce problems caused by bad weather and road conditions. The weather-controlled road stretches
between Pyhtaa and Hamina for 25 kilometers along a section of European road E18 on the southern coast
of Finland. Along the E18 roadway drivers are informed about varying weather and road conditions in real
time through automatic road weather monitoring stations and variable message signs. The weather-controlled
road contains five automatic road weather monitoring stations, 66 variable message signs and 13 variable
warning signs. The weather-controlled road aims to improve the efficiency and safety of transport as well
as to increase traveling comfort by controlling variable message signs according to existing road conditions.

The sensors connected to the road weather monitoring system measure wind speed and direction, air
temperature, roadway surface temperature, humidity, intensity and state of precipitation, visibility, and road
surface state (wet, dry, icy). The mainframe computer collects road weather information and warnings from
road weather monitoring stations for data storing and analysis in five minute intervals. Road weather
information are then analyzed and assigned to four road condition categories A, B, C, or D. Category A is
the good road condition category which includes conditions where the road is dry or moist and no deviations
are observed from summer like road conditions. Category B is the normal conditions category. The poor
road condition category C includes situations of clearly worsened conditions. The very poor road conditions
category D means that the conditions are very severe.

The effects of the weather-controlled variable message signs were studied. Ninety-five percent of the drivers
studied expressed that varying speed limits according to prevailing road conditions was useful and enhanced
traffic safety. The weather-controlled traffic management system helped decreased both the mean speed and
the standard deviation of speeds along this section of E18 in Finland (5).

Effects of Weather Controlled Variable Speed Limits and Warning Signs on Driver Behavior - Finland

As part of phase two of the previously mentioned paper, this study was designed to investigate the effects
of weather controlled speed limits and slippery road signs on driver behavior on Finnish road E18. Local
weather and road conditions were monitored from two unmanned road weather stations. The speed limits
were lowered automatically in adverse road conditions and in some cases slippery road signs were displayed
as well. Speed and headway data were obtained from loop detectors. The results showed that in winter time
the change of the posted speed limit from 100 km/h to 80 km/h decreased the mean speed of cars traveling
in free-flow traffic by 3.4 km/h, in addition to the average mean speed reduction of 6.3 km/h caused by the
Scott Harry Page 186

adverse weather and road surface conditions. When the poor road conditions were difficult to detect, i.e.,
there was no rain or snowfall or the rain was insignificant, the effect was 1.9 km/h higher (i.e., the reduction
was 5.3 km/h). When the road conditions were such that the slippery road sign was also displayed, the
system reduced the mean speed by 1.8 km/h, while the reduction of the weather was 9.3 km/h. In addition
to the effects on the means speed, the lowering of the speed limit decreased the speed variance, but had no
remarkable effect on headways. The main implication of the study was that the system improved traffic
safety by decreasing the mean speeds and speed variation.

System reliability is a key factor in influencing driver behavior. Therefore, the system reliability for this
study was evaluated. 139 observations of the weather and road conditions were collected, as well as friction
measurements. In 70% of the cases, the speed limit and the use of the slippery road sign was estimated to
by appropriate. Furthermore, the results indicated that there was a tendency to use speed limits that were too
high. In 26% of the cases the speed limit was assessed to be too high or the slippery road sign was not
displayed when it should have been (6).

Examining the Institutional Issues Related to Implementing Road and Weather Information Systems (RWIS)

Aurora, a multi-agency consortium consisting of several state DOT’s and federal agencies, funded this
project aiming to identify and document institutional issues relating to the implementation and development
of RWIS. From the research, the issues that were encountered could be classified into four categories:
funding, staffing, partnerships, and the expandability, transferability, and compatibility of RWIS systems.

Funding. The key to resolving funding issues may be the ability to present the benefits of implementing
RWIS and how it can improve maintenance activities within an agency. At the outset of RWIS
implementation, agencies should draw upon the previous experiences of others and present to their
organization a clear picture of how RWIS can benefit their agency. As with securin g funding for any
endeavor, the competition of funds will be a constant struggle. However, if it can be widely understood
within implementing agencies that RWIS can significantly improve the efficiency and appropriateness of
maintenance and operation procedures, these technologies should eventually evolve into a well-accepted tool
in the snow and ice control process.

Staffing. In some instances, RWIS innovations have been perceived negatively, as another tool to replace
manual labor. Additionally, winter storms have not been perceived by maintenance staff as being an
unfortunate experience. The conviction was that if RWIS can be used effectively in snow and ice control
activities, it would reduce the amount of overtime money (7).

The negative mindset of some personnel is an obstacle that may be overcome by adequate training. It is
apparent that training is required initially to alleviate resistance to using RWIS technologies, and continually
to increase levels of familiarity and comfort. From the surveys that Aurora conducted, there was found to
be a noticeable void and training is not emphasized like it should be. In the cases where training has taken
precedence, personnel are generally openly embracing RWIS technologies.

Partnerships. Although there is tremendous interest in partnerships, there appears to be some hesitation by
public agencies to form partnerships with other private and public agencies. The prevailing reluctance stems
from liability and ownership issues. Additionally, partnerships proved to be largely uncharted territory for
agencies within the United States, particularly public/private ventures.

Before public agencies seek out partnerships, they must determine the needs of their potential partners and
be able to clearly define the roles each will play and how each will benefit from the alliance. For instance,
private sector partners need to be confident that their cooperative venture will prove profitable. However,
it is recognized that government agencies need to maintain the competitive bid process and be continually
cognizant of not intentionally favoring one company over another in the quest for potential partners.
Scott Harry Page 187

Furthermore, clearly delineated roles and responsibilities in the ownership of data can alleviate some of the
potential areas of conflict within partnerships.

Expandability, Transferability, and Compatibility. Incompatibility of various RWIS sensors and systems is
a primary concern since public agencies are looking towards deploying a network of RWIS technologies.
Not only is it an issue with the boundaries of a state, but it appears to be problematic for neighboring
countries who would like access to centralized road and weather information to aid in maintenance activities
along common boundaries. Presently, the technology exists to implement RWIS at a statewide and even
nationwide level, but the incompatibility of systems from different vendors is one factor which impedes full-
scale deployment at an acceptable cost.

Agencies should attempt to remain abreast of the RWIS-related protocols and standards development process
by becoming actively involved with the organizations and committees established to deal with these issues.
Being involved in these activities encourages proactive problem solving and reaching solutions which meet
the needs of a broad range of organizations. Also, by directing energy into standardization and protocols,
duplicate efforts can be minimized by having cooperative participation by numerous agencies.

Conclusions. From the perspectives of the agencies that were interviewed, perhaps the greatest challenges
are training of operations and maintenance staff and the shortage of funds to provide sufficient and proper
equipment. In addition, changing the cultural climate of agencies and mindests of those decision-makers who
have control of funding that could be utilized for RWIS are other impediments to advancing RWIS
technologies. The lack of coordination between agencies with parallel efforts was seen as an obstacle for
widespread implementation of systems.

Although there appear to be many hurdles to clear, RWIS programs prove to be successful and beneficial to
public agencies. RWIS is a valuable tool in the winter maintenance decision-making process. The
information gathered from these systems and distributed to maintenance staff and the general public cannot
reduce the severity of weather and winter storms. However, it can help reduce the effects from such storms
(7).

The following three papers deal with evaluations of different sensor technologies.

Assessment of Road Surface Freezing Point Sensors for the UK

Between 1992 and 1997, the Transportation Research Laboratory was commissioned by the UK Highway
Agency to assess whether the displayed freezing point temperatures (FPT) from different systems were
consistent and accurate, taking into account the residual deicing che mical on the road. A series of
experiments on in-service roads monitored the performance of both existing types of sensors (passive and
active) installed in the highway. Two test methods were adopted: spraying and ponding of the salt solution
directly onto the sensor. Each test was carried out with increasing concentrations of salt solution before final
washing off to determine the ability of a sensor to recover. To increase confidence in the results found on
the road, the sensors were tested under controlled laboratory conditions. The purpose was to investigate the
ability of the sensors to detect the freezing point (or concentration) of 0.5 mm thickness of aqueous rock salt
solutions under idealized conditions.

The performance of passive sensors, on UK roads, shows that the FPT’s of applied salt solutions at ambient
temperatures at and below 2°C are slow in response and salt solutions are not accurately measured when the
concentration is above four percent. Increasing salt concentrations are measured more accurately than
decreasing salt concentrations.

The active sensor performed satisfactorily and measured the FPT accurately at test temperatures of 2°C and
-5°C for applied salt solutions of 0, 3.8, and 10.7 percent concentration. In general, the sensor responded
Scott Harry Page 188

quickly to changes in salt concentration and test temperatures and performed better than the passive sensors
under this test procedure (9).

Test of Road Weather Monitoring Systems and Sensors - Sweden

As part of COST 309, the Swedish National Road Administration (SNRA) conducted an objective study
comparing four different roadway weather systems and sensors. The four different systems compared where
the van den Berg BV & BG Engineering Ice Warning System, the CDS/Campbell Weather Monitoring
System, the FFV Road Weather Information System, and the Thies Weather Monitoring System (11).

Surface Temperature. Although no comparisons were made between the different road temperature sensors,
some general comments were made with regards to these sensors. The most common type of sensor for
measuring road surface temperature is the pt100-sensor. This sensor measures the electronic resistance over
a platinum thread usually placed in a steel tube. The accuracy for this sensor has been shown to be good,
less than ±0.2 °C. Another widely used surface temperature sensor is the Thermocouple sensor. This sensor
is inexpensive and only consists of the wires between the logger and the road pavement. In some
applications an IR-sensor can be used, mounted above the road. However, this sensor is expensive and is
not useful in ordinary road monitoring systems.

The location of the road surface temperature sensor is very important. The high speed lane (with less traffic)
is often some degrees cooler than the low speed lane. Within the lane, the temperature differs between the
wheel tracks, the area right under the car, and the area just beside the car. In the Swedish system, the sensors
are located in the middle of the lane between the two wheel tracks. If the sensors were located in the wheel
track, the studs from the tires would wear out the sensor too fast.

Humidity. There are two purposes for measuring humidity when monitoring road conditions. First, it gives
general information about the weather conditions. Second, by comparing road temperature and dewpoint
temperature, one sees the conditions for dew and frost.

The Rotronic YA 100 and the MP-100F both utilize the same sensor (Hygromer). It is very small,
approximately 0.002 grams. It can measure the relative humidity from 0% to 100% and operate at
temperatures of -50°C to +150°C. The values of the relative humidity from these sensors did not differ
significantly from the values of the sensor installed in 1989/1990 (11).

The Thies hygro-thermo-transmitter was also tested for accuracy. The concern for accuracy occurs when
the humidity is between 85-90%. The paper suggests that this sensor, a hair transmitter sensor, needs to be
recalibrated somewhat more often than electronic sensors and that the accuracy is best when the humidity
is not close to saturated. The calibration should be done in the field because the indoor ambient air
temperature can affect the hygrotransmitter (11).

The FFV Road Monitoring System uses Lambrechts hair hygrometer, 809L-100 with a synthetic hair-element
(PERNIX). This sensor is identical to the one manufactured by Friedrich. The Swedish Road Administration
have had good experience with this sensor and have used it for several years. Usually they recalibrate it once
in the beginning of the season, once in the middle, and occasionally some extra calibration is needed (11).

Freezing Point and Road Condition. BG Engineering uses a circular shaped sensor, 6 cm in diameter that
is mounted flush with the road surface. The sensor measures the conductivity of the road surface emulsion,
using two concentric steel pipes. It is recommended that when these sensors are used, the system should not
classify the road condition automatically but rather present the value of the conductivity. The road officers
themselves then can classify the road condition, using information both from the road sensor and from other
sources (11).
Scott Harry Page 189

The FFV Frensor sensor is an active sensor which directly measures the freezing point temperature of the
road surface film. The size of the device is 40 X 80 X 80 mm, and is made of epoxy and stainless steel. This
sensor showed very good accuracy, and after an adaption of the linear error, the total error is less than tenths
of a degree (11).

Environmental Sensor Systems for Safe Traffic Operations

Castle Rock Consultants went into great detail comparing different weather prediction sensors. For the
purposes of this paper, the discussion is limited to pavement sensors.

Climatronics (U.S.)-FRENSOR. This is an active device that is imbedded into the concrete and directly
measures the freezing point using a peltier element. As earlier stated, the Swedish National Road
Administration COST 309 report indicates that the freezing point is accurately measured by the FRENSOR
sensor.

Surface Systems, Inc. (U.S.) - SCAN. The SCAN sensor is an in-pavement sensor that measures temperature,
determines if the road surface is wet or dry using a capacitive technique, and measures conductivity to
determine salinity. The sensor is thermally passive and fabricated using materials with the same thermal
characteristics and color as the roadway. The system has been well tested and is in widespread use with over
200 installations reported. The cost of each ice detection station, including commissioning and installation,
is approximately $35,000.

Vaisala (Finland) - Road Surface Sensor DRS12. This sensor, normally connected to a Vaisala MILOS
weather station, is passive and buried into the roadway surface. The thermodynamic properties and thermal
emissivity are designed to match the roadway so that the sensor’s thermal measurements match the roadway.

AANDERAA Instruments - Road Surface Temperature Sensor (3304) and Conductivity Sensor (3330). This
conductivity sensor uses four copper electrodes to measure the conductivity of any moisture present on the
sensor. The sensor must be located in the roadway shoulder because only a thin film of moisture can be
present.

The temperature sensor is mounted flush with the road surface in the travel lanes. The sensing element is
a platinum resistor in a Wheatstone half-bridge. Only the surface temperature is measured and not a below-
grade temperature. The road temperature and conductivity sensors together determine the freezing point.

Findlay Irvine (Scotland) - Road Surface Sensor. This sensor is mounted flush with the road surface and
measures surface temperature, surface condition (dry, wet, ice/snow), and freezing point. The freezing
temperature can be measured to four levels: 0°C, -3°C, -6°C, -10°C.

BG Engineering (Holland) - Road Condition Sensor. This device measures conductivity of the road surface.
The sensor is 60 mm in diameter and consists of two concentric steel pipes. It is mounted flush with the
roadway and measures the conductivity of the roadway emulsion using the inner and outer pipe as electrodes.
Using this sensor it is difficult to determine the difference between dry conditions and icy conditions since
dry and ice conditions both have low conductivity. One way around this is to use two sensors, one heated
and one not.

Vibrometer SA (Switzerland). The Vibrometer is a steel pavement sensor which is 50 mm in diameter which
determines the presence of moisture by constantly vibrating a surface plate. Analysis of the vibration
determines if the road surface is dry, moist, or frozen. The surface plate is cooled and warmed using a peltier
element. The phase change from ice to water is used to determine the freezing point rather than from water
to ice to avoid problems with supercooled water. How accurately the sensor measured water and ice films
Scott Harry Page 190

was not determined, but the study indicated there was a good correlation between indicated and actual film
thickness.

Schrack System, Inc. (Austria) - Road Condition Radar (RCR). Road radar has the potential of measuring
the thickness of the water layer and the salinity of the solution, from which the freezing point can be
calculated. A radar transmitter bounces energy off the highway at a 60° angle of incidence into a microwave
receiver. The radar device has an advantage over in-pavement sensors because it looks at a large area of the
road across the width of the highway. Advantages are that the pavement is not disturbed during installation,
and a longer sensor life can be seen since no traffic wear occurs on the sensor.

Fog Warning Systems

Fog Warning System - Georgia

Sensing equipment will be used to monitor visibility. When visibility falls below a preset level, variable
message signs (VMS) will be activated automatically to alert drivers of the low visibility conditions ahead
and variable speed limit (VSL) signs will show reduced speed limits.

Two visibility sensors will be used at one location. Both use forward scatter technology, however, one is
standard size and the other, developed through the federal Small Business Innovative Research (SBIR)
program, is smaller and cheaper than standard meters (4).

Fog Warning System - Utah

Sensing equipment will be used to monitor visibility and traffic speed. When visibility falls below a preset
level, variable message signs (VMS) will be activated automatically to alert drivers of the low visibility
conditions ahead and variable speed limit (VSL) signs will show reduced speed limits based on visibility and
actual traffic speeds. The visibility sensors will use forward scatter technology (4).

According to the principle investigator, this project has experienced some equipment troubles that will need
to be ironed out before full implementation.

I-526 Fog Mitigation Project - South Carolina

One weather station and five fog detectors are used to monitor the visibility of a section of I-526 which
crosses the Cooper River in South Carolina. When sensors indicate that visibility is reduced, the
maintenance dispatcher observes the location from a video camera to verify the monitor readings. If visibility
is reduced, the dispatcher turns on the roadway lights and activates an appropriate message on the VMS. The
roadway lights are airport lights which have been embedded in the roadway to assist drivers during reduced
visibility conditions.

The weather station, manufactured by Met One, measures wind direction and speed, air temperature, and
relative humidity. The five fog detectors use back scanner technology (4).

I-75 Fog Detection/Warning System - Tennessee

This project encompasses a 3-mile fog prone area of Interstate 75 at the Hiwassee River crossing and 8-mile
approaches on each side. The project was initiated as a result of a fog-related accident that occurred in 1990
and involved 99 vehicles and resulted in 12 fatalities and 42 injuries. During the first two years of system
operation (1993 - 1995), no fog-related accidents were reported.
Scott Harry Page 191

Eight visiometers (fog detectors) and two weather stations monitor conditions near the fog prone area. When
conditions reach levels conducive to fog generation, personnel at the Central Control Center (CCC) at the
Highway Patrol Office in Tiftonia, Tennessee, are automatically alerted. Safety personnel can then activate
any of 10 Variable Message Signs (VMS) along with the flashing lights on the fixed message signs. The
VMS are programmed with pre-determined messages, including warnings, speed limit changes, notification
of activation of highway advisory radio (HAR), and activation of six swing gates on on-ramps if the highway
is closed. Variable speed limit signs are used along this portion of I-75 so the posted speed can be reduced
, if necessary.

This stretch of roadway is also equipped with 44 radar vehicle flow detectors which monitor speed and
number of vehicles. When either of these two parameters reaches a pre-determined level, the VMS are
automatically activated to notify travelers.

The system was designed to require minimal staffing requirements. As such, safety personnel were able to
incorporate the operation of this system into their daily responsibilities.

Communication between system components on-site is provided by buried fiber optic cable. Data are
transmitted by microwave through the two repeater sites to the CCC in Tiftonia (4).

Avalanche Warning Systems

Technical Applications in Avalanche Hazard Mitigation - Utah/Wyoming

This project consists of two parts, monitoring traffic in avalanche-prone areas and sensing the onset of
avalanches. A traffic monitoring system will be developed to monitor a two-way, two lane segment of
highway. The system will count the number of vehicles entering and leaving the highway segment. If an
avalanche occurs, highway maintenance personnel will be able to determine if any vehicles are trapped by
the avalanche.

The second part of the project involves using and evaluating avalanche sensors. (Project participants
anticipate using different sensors in this project from those used in the Idaho Automatic Avalanche Sensing
System project described later in this paper). The sensors to be used for this project as of 1997 had not yet
been selected. While the sensors that will be used are available commercially, they have not historically been
used for avalanche detection. Thus, the project could be considered state-of-the-art.

Initially, sensors will be used to detect the onset of an avalanche. If any funding remains, the project may
also include evaluation of sensors that monitor precursor conditions to an avalanche which would further
increase traveler safety (4).

Development and Trial Deployment of an ITS Avalanche Hazard Management System - Utah, Wyoming,
Idaho

Increased travel demand on the winter/alpine roads of the Western United States has resulted in a dramatic
increase in the hazard to motorists and maintenance section personnel from snow avalanches. As part of the
ITS Ideas Deserving Exploratory Analysis (IDEA) project, this study was established to evaluate three sites
in the northwest that currently have avalanche prediction systems in place. The three sites being evaluated
are: Idaho State Highway 21, the Canyon Creek section; Utah State Route 210, Little Cottonwood Canyon;
and Wyoming/U.S.R. 189-191, Hoback Canyon (8).

The IDEA concepts investigated and reported for this study include the testing and integration of technology
to manage the avalanche hazard to motorists and maintenance section personnel using an automated system.
This includes sensing the initial release of the avalanches in conjunction with automated traffic management
Scott Harry Page 192

actions along the hazardous roadway corridor. These data are then used in one of two ways: On many
roadways avalanche descent times from their starting zones to the roadway below are upwards of 10 to 120
seconds. In these situations sensing the initial release of the avalanche will allow the road, which will very
shortly be overrun, to be closed prior to the avalanche’s arrival. Typically, the length of road which will be
overrun is on the order of 100 to 200 feet. This form of knowledge based ITS avalanche hazard management
is known as time-of-descent management. Alternatively, the avalanche starting zones may be relatively close
to the road or, as is often the case, the hazardous corridor is made up of numerous avalanche paths in close
proximity. In this case sensing the onset of avalanching is used to initiate closure of the entire corridor and
to simultaneously notify highway and emergency management agencies. This form of knowledge bases ITS
avalanche hazard management is known as corridor management.

Utah S.R. 210/Alta Test Bed. On Utah S.R. 210 in Little Cottonwood Canyon near the community of Alta,
an operational Utah Department of Transportation avalanche road closure gate was equipped with a wireless
communication interface (cellular phone) and photo-voltaic electrical re-supply capabilities. Using alpha-
numeric menu commands, either manually input or from an avalanche detection sensor, it was possible to
close (or open) this avalanche road closure gate remotely. The structure was destroyed by an avalanche
during the winter of 1997/1998 and was not replaced (8).

Wyoming/U.S.R. 189-191 Time-of-Descent System. Wyoming/U.S.R 189-191 in the Hoback Canyon has
a specific avalanche path that has a history of striking highway maintenance and motorist vehicles on the
roadway. This avalanche path is long and the descent time of the avalanche is on the order of 30 seconds.
The avalanche detection elements of the system are located 800 feet above the roadway. These sensors are
limit switches. When the sensors are excited a radio frequency (RF) communications interface is enabled
and an alarm is sent simultaneously to the lights of motorist information signs on either side of the avalanche
path, and to any of several in-vehicle audible avalanche alarm boxes that can be located in highway
maintenance vehicles that may have need to be working on the roadway immediately below the avalanche.
This ITS avalanche hazard reduction system provides highway maintenance personnel and road users around
12 seconds of warning prior to the arrival of the avalanche on the roadway. During the 1998/1999 season,
WyDOT personnel recorded four avalanche events. Of these four avalanches, at least two were successfully
detected. In addition, there was one false alarm, most likely due to wind generated motion of the avalanche
detection sensor, and one avalanche which occurred that was not successfully detected. The unsuccessful
detection was a result of the avalanche starting below the sensor line (8).

Idaho S.H. 21 Corridor. Highway maintenance and emergency management agencies along this route are
in need of real time notification on the onset of the avalanche process. This route is the only all season route
connecting the Stanley basin area with the Boise area. The alternative route adds around three hours to a one
way trip. There are over 50 separate avalanche paths over an 18 mile stretch of roadway. As part of this
study, two avalanche detection sensors systems were installed along this route. One of the two sensor sites
experienced several problems with the equipment and due to the conditions the data from both of the sensors
had to be down loaded manually after the season. It is suspected that the sensors, while operational,
successfully predicted avalanching. However, the snowfall was so heavy during the 1998/1999 winter season
that the highway was closed and therefore no visual confirmation could be given for sensor predicted
avalanches (8).

ROAD WEATHER INFORMATION SYSTEMS

There are several major manufacturers of weather prediction components; Surface Systems, Inc., Vaisala,
and Climatronics are three of the most popular. This section summarizes the components and the prediction
capabilities of a typical RWIS.
Scott Harry Page 193

The RWIS consists of on-site temperature and atmospheric sensors, data processing units, telecommunication
capabilities, and workstation displays.

Surface (Pavement) Sensors. Pavement temperatures are obtained from electronic devices installed in the
roadway. Surface sensors can either be active or passive, but passive sensors have gained more wide-spread
acceptance.

The thermally passive surface devices are usually fabricated of material with thermal characteristics similar
to those commonly used for pavement contractions. Sensors, Typically 13 centimeters in diameter, are
installed in a 14 centimeter diameter hole flush with the pavement surface and color matched with the
surrounding pavement so that both respond to incident light energy at the same rate. The sensor’s output
signal is designed to reflect the condition on top of the device, which closely approximates conditions on the
surrounding pavement surface;

• Dry
• Wet (above 0 degrees C)
• Wet (not frozen but at or below 0 degrees C)
• Snow/ice alert (at or below 0 degrees C)
• Dew
• Frost
• Absorption

Other measurements include a chemical concentration factor (a relative indicator of the deicing or anti-icing
chemical present in the moisture on the sensor surface). Sensors typically have a stable operating range from
-30 degrees to 50 degrees C.

Subsurface Temperature Sensors. Probes are installed directly below the surface sensor at a depth of around
40 centimeters. These provide heat flux information primarily for computer models designed to predict
pavement temperatures. Frost depth data are also used by highway agencies to regulate truck routing on the
basis of the frost level beneath the road surface.

Atmospheric Sensors. Air temperature is typically measured over a range of -62 degrees C to 70 degrees C,
and relative humidity (RH) measurements over a range of 10-100%. Dew point is calculated from
temperature and RH readings. Precipitation is detected by sensing interruption in an infrared optical beam.
Wind direction and speed detectors have an operating range of 360 degrees and can record speeds of up to
215 km/h.

Data Processing Units. All data collection by the sensors are transmitted to a remote processing unit (RPU),
which in turn converts the analog signal into a digital format and relays the information to the central
processing unit (CPU). The RPU can typically receive data from up to four surface sensors, four subsurface
sensors, and the atmospheric sensors. Atmospheric sensor measurements are sued in the system logic
processed by one or several RPU’s. Agencies configure RPU’s is various ways. Airports, for example,
typically have 2-15 RPU’s; there are 178 in the highways surrounding Columbus, Ohio; and 32 are in service
in the Wisconsin statewide road weather information system (2).

The CPU analyzes, stores, and arranges the data it receives from up to 100 RPU’s for information display
on the workstation. The workstation displays the location of each RPU and surface sensor, the surface status
of each sensor, atmospheric sensor data, and the current date and time. The system can typically update this
information every 15 seconds.

Communications. Several telemetry options are available for system configuration, including dial-up and
leased telephone lines, radio, microwave, fiber optics, satellite, and value-added networks.
Scott Harry Page 194

Workstation Formats and Displays. In addition to displaying the information in tabulated test formats, the
systems can provide customer-specified individualized color graphics. Critical pavement conditions are color
coded for instant recognition of situation that may require proactive decisions (2).

Figure 1 below shows a typical atmospheric and temperature sensor station. Figure 2 shows the installation
of a pavement sensor.

Figure 1. Atmospheric and Temperature Sensor Station

Figure 2. Installation of an SSI Pavement Sensor

RECOMMENDATIONS

From the literature review it can be seen that implementation of an RWIS can not only increase safety, it can
also significantly decrease maintenance costs. One article states that the benefit to cost ratio of implementing
an RWIS can be as high as 11 to 1 (3). With this in mind, it should be noted that some systems and sensors
work better in certain situations than others.
Scott Harry Page 195

One of the purposes of this paper was to make recommendations on which pavement sensors to use in
different situations. It is recommended that active sensors be used in areas where high concentrations of salt
typically are applied to the roadways (greater than four percent concentration). One sensor in particular is
the Climatronics FRENSOR sensor that has been used several places throughout the U.S. The SSI active
pavement sensor would be recommended for this situation because more of these systems have been
implemented and incompatibility would not be an issue. However, no analysis of this sensor was found and
therefore the reliability and accuracy is not fully known.

In areas where typically low concentrations of salt are regularly present, less than four percent concentration,
it is recommended that the SSI passive pavement sensors be used. These sensors have been shown to be
accurate and reliable in this situation. As stated above, the SSI sensors are in wide-spread implementation
across the United States and implementing these sensor technologies avoids problems with compatibility and
allows full information exchange.

With regards to the implementation of full RWIS systems. Each situation is different. For general weather
information and prediction, the SSI system has been shown to be accurate and has had good acceptance
among agencies. However, in situations where fog is a major cause of traffic problems, other technologies
should be implemented to help mitigate these problems. Several different systems have been shown to be
effective using forward scatter techniques.

The different types of severe weather events that can be experienced along the roadways pose unique
problems. The different ITS technologies that have been developed to help predict these events are also
unique. Therefore, it seems ambitious to think that there will ever be one system that works best for all
weather situations. For best results when implementing these weather prediction systems, it is recommended
that a good blend of accuracy and compatibility is used to maximize the benefits of these systems.

RWIS IMPLEMENTATION CRITERIA

From the literature review a set of criteria for selecting RWIS implementation locations was developed. The
issues listed below are issues that need to be addressed before implementing an RWIS system. These criteria
can help agencies decide whether or not they should implement such a system.

Funding

Being able to acquire funding for RWIS projects can be difficult. The key issue is being able to present the
benefits of implementing RWIS and how it can improve maintenance activities within an agency. Drawing
on previous experiences of others and being able to present to your organization a clear picture of how RWIS
can benefit your agency is also advised. The general idea is that if it can be widely understood within
implementing agencies that RWIS can significantly improve the efficiency and appropriateness of
maintenance and operation procedures, these technologies should eventually evolve into a well-accepted tool
in the snow and ice control process.

With these issues in mind, it is recommended that any agency considering implementation of an RWIS
system should have a public relations type of person on staff. Significant time and effort needs to be spent
organizing and presenting an RWIS so that acceptance is widespread. Not only should this person be able
to devote all their time to such an effort, they should also have good communication skills and training so
that the plans and objectives of the implementing agency are clearly and concisely presented.
Scott Harry Page 196

Staffing

In some instances, RWIS innovations have been perceived negatively, as another tool to replace manual
labor. The conviction was that if RWIS can be used effectively in snow and ice control activities, it would
reduce the amount of overtime money paid to maintenance crews.

The negative mindset of some personnel is an obstacle that may be over come by adequate training. It is
apparent that training is required initially to alleviate resistance to using RWIS technologies, and continually
to increase levels of familiarity and comfort. For this reason, any agency thinking about implementing an
RWIS needs to have the ability to develop a training program for personnel. This training program can be
developed using in house personnel or personnel from elsewhere depending on what is best for the
sponsoring agency.

Partnerships

Although there is tremendous interest in partnerships, there appears to be some hesitation by public agencies
to form partnerships with other public and private agencies. The prevailing reluctance stems from liability
and ownership issues. Before public agencies seek out partnerships, they must determine the needs of their
potential partners and be able to clearly define the roles each will play and how each will benefit from the
alliance.

In order for an agency to be successful in implementing an RWIS system it is beneficial that they have good
relationships with other public and private agencies in the area. If good relationships are not present, it
should be determined if anything can be done to make the relationships work. Each area is unique when it
comes to partnerships, therefore, each situation needs to be looked at to determine how past experiences with
partnerships have turned out.

Expandability, Transferability, and Compatibility

When deploying an RWIS incompatibility between components of various RWIS sensors and systems is a
primary concern. Not only is it an issue within the boundaries of a state, but it appears to be a problem for
neighboring countries who would like access to centralized road and weather information to aid in
maintenance activities along common boundaries.

When an agency is determining which RWIS sensors they should implement, the agency needs to determine
which sensor technologies are already in place and which other agencies might consider accessing the
information. In practical terms, this should be taken to the statewide or possibly even the multi-state level.

SAN ANTONIO, TEXAS CASE STUDY

As part of a case study, the city of San Antonio, Texas was chosen to determine if an RWIS should be
implemented, and more specifically should pavement sensors. Transguide, the traffic management center
in the area, has expressed interest in RWIS technologies and asked if an analysis could be conducted.

Pavement sensors aid in the prediction of roadway surface conditions, typically ice formation. However, due
to the fact that San Antonio only experiences icy conditions between 0 and 8 days per year, the focus of this
case study is on the RWIS system as a whole and not just the roadway sensors. The storms that pass through
the San Antonio area typically have no set arrival pattern or approach direction. Therefore, to accurately
predict conditions around the city, a pattern of weather stations would have to be implemented around the
city on several of their major arterials. The criteria that were developed in the previous section were applied
to the San Antonio area and the results are listed below.
Scott Harry Page 197

Funding

With the geographical make up of San Antonio and the network of roads surrounding the city, it is suggested
that a minimum of five stations be installed on incoming routes outside of loop 410: one on I-10 northwest
of town, one on I-35 northeast of town, one on I-10 east of town, one on I-37 southeast of town, and one on
I-35 southwest of town. Concentrations of salt along the roadways in San Antonio are low, for this reason
it is suggested that the SSI RWIS with passive pavement sensors be installed at the five locations previously
mentioned. A conservative estimate of cost of these systems is $50,000 per system or a total implementation
cost of around $300,000. This does not include annual maintenance costs.

The following figure shows a map of San Antonio, Texas and the proposed RWIS locations (indicated by
the stars).

Figure 3. Proposed RWIS Locations; San Antonio, Texas

Transguide’s maintenance division has shown interest in this type of system and has indicated interest in
raising funding for an RWIS system. Specific dollar amounts for maintenance costs of these systems are
unknown. However, with the public relations people and the reputation Transguide has earned, it should not
be hard for Transguide to raise and support this type of system.
Scott Harry Page 198

Funding could possibly be shared between the maintenance division and the operations division. Transguide
has shown interest in these systems for operational purposes as well, and might be willing to share some of
the cost and responsibility between the two divisions.

Staffing

Personnel will be needed for the maintenance and operation of these systems. Transguide has indicated a
willingness to devote the equivalent of 1.5 to 2 positions for this purpose. The duties that these personnel
would take on would be add-on duties to what they are already assigned. Training of these personnel should
be of high priority. A key in making a system like this work is to have good people running the system.
Therefore, very thorough training programs should be implemented so that the RWIS is allowed to develop
to its full potential. Transguide has conducted similar training programs and staffing problems should not
be an issue.

Partnerships

Public/private partnerships are one area that Transguide has little experience. Typically RWIS have seen
efforts to develop partnerships between public and private agencies to aid in the success of these systems.
Transguide does have minimal experience with partnerships and the author believes that implementing this
system would be a good opportunity for Transguide to expand on their partnership experience.

For private agencies to buy into the idea of RWIS systems, they need to be able to see the benefits and know
that profits will be seen. Transguide has a successful public relations program and has seen success in selling
their ideas to public agencies. Therefore, it is not foreseen that Transguide will have problems getting
private sector firms involved in this type of operation.

Expandability, Transferability, and Compatibility

In the review of the literature, very few RWIS systems were found to exsist in the states neighboring Texas.
For this reason, incompatibility should not be a problem. Most of the projects reviewed from other areas
have chosen to install SSI systems for weather prediction purposes. Since this geographical area is new to
these types of weather prediction systems, Transguide should strive for compatibility with regions and states
that have already seen successful implementation of these types of systems. The most common system used
throughout the United States seems to be the Surface System, Inc. weather stations.

Of the four criteria developed, San Antonio shows strong potential in the areas of funding and staffing.
Transguide has minimal experience with partnerships, and therefore this criteria was a little weak. However,
as stated above it is not foreseen that there will be problems getting private sector firms involved due to
Transguide’s strong reputation. The issue of compatibility of components is virtually non existent in this
area due to the fact that San Antonio would be one of the first cities in its region to implement such a
program. With this said, the author believes that San Antonio would be a good candidate for implementation
of an RWIS system.

ACKNOWLEDGMENTS

The author would like to thank Dr. Conrad L. Dudek for planning and organizing the Advanced Institute
program which allows students to interact with a diverse group of transportation professionals. The author
would also like to thank Mr. William Spreitzer for his guidance and time throughout the summer. A special
thanks is extended to the other professional mentors for their participation and insight: Wayne Kittelson,
Thomas Hicks, Patrick Irwin, Douglas Robertson, and Carol Zimmerman.
Scott Harry Page 199

REFERENCES

12. Transportation Technology Plan. National Science and Technology Council. Committee on
Technology, Subcommittee on Transportation Research and Development. November, 1998.

13. Road Weather Information Systems by Surface Systems, Inc. http://www.dp105.net/partners/surface.htm

14. Road Weather Information Systems (RWIS). Mark F. Pinet & Associates Limited.
http://www.comnet.ca/~mfpinet/

15. Weather Summary and Synthesis Report. Federal Highway Administration. Prepared for a Weather
Workshop June 17-18, 1997. McLean, VA.

16. Portaankorva, Petteri. The Traffic Management Policy of Variable Message Signs for Weather-
Controlled Road Between Pyhtaa and Hamina, Finland. Finnish National Road Administration. 1996.

17. Rama, Pirkko. Effects of Weather Controlled Variable Speed Limits and Warning Signs on Driver
Behavior. Technical Research Centre of Finland. 1998.

18. Road and Weather Information Systems (RWIS) Feasibility Workshop Summary and Findings.
Mn/DOT and ENTERPRISE. January, 1996.

19. Examining the Institutional Issues Related to Implementing Road and Weather Information Systems
(RWIS). Presented at the 1998 World Congress on Intelligent Transportation Systems, Seoul, Korea.
October 1998

20. Decker, Rand, Jensen Newel, and Patterson, Ralph. Development and Trial Deployment of an ITS
Avalanche Hazard Management System. Winter/Alpine Engineering Laboratory. University of Utah.
May, 1999.

21. Burtwell, Marilyn. Assessment of Road Surface Freezing Point Sensors for the UK. Transportation
Research Laboratory. 1998.

22. Test of Road Weather Monitoring Systems and Sensors. Swedish National Road Administration. 1990.

23. Environmental Sensor Systems for Safe Traffic Operations. United States Department of Transportation,
Federal Highway Administration. DTFH61-92-C-00012. January 1995.

24. Riley, Betsy. Surface Systems Helps Get a Jump on Weather. St. Louis Business Journal. October 12,
1998.

25. Adams, Edward E. Proof of Concept for Prediction of Pavement Temperature: A Tactical Decision Aid
for Highway Safety. Montana State University, Montana Department of Transportation. 1999.

26. Kwon, Taek Mu. Next Generation RWIS. University of Minnesota at Duluth. 1998.

27. Takle, Gene. Maximizing the Use of RWIS. Center for Transportation Research, Iowa State University.
October 1998.
Scott Harry Page 200

28. Shao, J. Improving Nowcasts of Road Surface Temperature by a Backpropagation Neural Network.
Weather and Forecasting, Vol. 13. October 1997.

29. Croft, Paul J. Fog Forecasting for the Southern Region: A Conceptual Model Approach. Weather and
Forecasting, Vol. 12. April 1997.

SCOTT R. HARRY

Scott R. Harry received his B.S. in Civil Engineering from

Montana State University in May 1998. Currently, Scott is
pursuing his M.S. in Civil Engineering/Transportation at Texas
A&M University, and should graduate in December 1999. Scott
has been employed since August of 1998 by the Texas
Transportation Institute as a graduate research assistant. He also
worked for the Western Transportation Institute located at
Montana State University for two summers. Current TTI projects
with which he is involved include: “Geometric Design
Guidelines for At-Grade Intersections Near Railroad Highway
Grade Crossings.” Scott has also been involved in the American
Society of Civil Engineers, Intelligent Transportation Society of
America, and the Institute of Transportation Engineers. His main
interests are: Intelligent Transportation Systems, geometric design, and traffic operations.
OPENING HOV LANES TO GENERAL TRAFFIC DURING MAJOR
INCIDENTS AND SEVERE WEATHER CONDITIONS

by

Kevin P. Hoppers

Professional Mentor
Thomas Hicks, P.E.
Director of the Office of Traffic and Safety for the Maryland State Highway Administration

Course Instructor
Conrad L. Dudek, Ph.D., P.E.

Prepared for
CVEN 677
Advanced Surface Transportation Systems

Department of Civil Engineering

Texas A&M University
College Station, Texas

August 1999
Kevin P. Hoppers Page 202

ABSTRACT

A restriction in roadway capacity results in long queues, long motorist delays, greater emissions and a higher
potential for secondary accidents. In order to avoid the impacts of major incidents or severe weather
conditions, many areas containing HOV facilities have implemented HOV diversion strategies in their
incident management strategies. Some locations use various criteria for determining if the HOV lane should
be opened for general traffic use. Often the officer on the scene is responsible for the decision of opening
the HOV lane to single occupant vehicles. Once a decision is made to open the HOV lane to general traffic
a diversion plan must be followed and instructions must be clearly expressed to the drivers.

The objectives of this research were to:

• Establish current practices for diverting general traffic into HOV lanes during major incidents or severe
weather conditions.
• Develop a set of general criteria for HOV facilities where the HOV lane should be opened to
accommodate all traffic.
• Develop guidelines for a typical diversion plan for opening HOV lanes to the general traffic.
• Apply guidelines to a specific location.

Through a literature review and a series of interviews with six agencies that operate HOV facilities
recommended criteria was developed to determine when the HOV lane should be opened to general traffic.
A volume-capacity ratio of 0.7, which is a level of service D according to the 1997 Highway Capacity
Manual, is a reasonable threshold value that should exist before the general traffic is diverted to the HOV
lane. Once an incident occurs, the goal of the diversion plan is to maintain a volume-capacity ratio of 0.95
throughout the duration of the incident. A freeway can theoretically sustain a volume-capacity ratio of 1.0.
However, the volume-capacity ratio of 0.95 was chosen because 1.0 is potentially unstable and anything
above this value could cause the freeway to breakdown. The threshold value for reduction in capacity for the
freeway is suggested by the author to be 50 percent for an estimated time to clear the incident of 1 hour.
Fifty percent was chosen because a single lane blockage on a three-lane freeway results in a reduction of fifty
percent of the capacity. A duration of the incident of 1 hour was chosen because the average incident takes
around 45 minutes to clear. If the capacity is reduced by 50 percent on a freeway for more than an hour with
a volume-capacity ratio of 0.7 then the alternative routes must be able to handle 33 percent of the traffic on
the freeway in order to maintain the desired volume-capacity ratio of 0.95. If the alternative routes off of
the freeway and the remaining mainlanes cannot handle this amount of traffic then the HOV lane should be
opened.

These criteria should be used by the response team to assess the situation and determine if lifting the HOV
lane restrictions is warranted. The response team would be in communication with the traffic management
center in order to receive operating conditions on the mainlanes and HOV lanes. The implementation plan
used in the HOV diversion strategy would include the following issues.

• Identification of the corridor

• Formation of agency partnerships
• Determination of alternative routes
• Defining HOV diversion criteria
• Development of signing and other traffic devices
• Coordination with the media
• Final implementation

These issues provide the framework of a typical HOV diversion implementation plan. They would need to
be modified for specific locations and situations. The criteria and diversion plan is applied to I-270 in
Montgomery County, Maryland.
Kevin P. Hoppers Page 203

TABLE OF CONTENTS

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Report Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206

STUDY METHODOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206

Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
Data Analysis and Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Incident Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
HOV Lanes Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
Concurrent HOV Flow Lanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
Reversible HOV Flow Lanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Contraflow HOV Flow Lanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210

SURVEY RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210

Dallas, Texas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Houston, Texas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Maryland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Minnesota . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
Seattle, Washington . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
Virginia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215

ISSUES TO IMPLEMENTING AN HOV DIVERSION PLAN . . . . . . . . . . . . . . . . . . . . . . . . . . . 218

Agency Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Motorist Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Public Acceptance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Development of HOV Lane Diversion Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219

DEVELOPMENT OF GENERAL CRITERIA FOR DIVERSION TO THE HOV LANE . . . . . 219

Severity of Situation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Time of Day . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Impact on Mainlane Traffic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Availability of Alternative Routes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

GUIDELINES FOR TYPICAL HOV DIVERSION PLAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

Identify the Corridor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Form Agency Partnerships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Determine Alternate Routes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Define Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
Development of Signing and Other Traffic Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
Coordination With Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
Final Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222

CASE STUDY: I-270 IN MONTGOMERY COUNTY, MARYLAND . . . . . . . . . . . . . . . . . . . . . 223

CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Kevin P. Hoppers Page 204

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228

APPENDIX -Guideline Questions for Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229

Kevin P. Hoppers Page 205

INTRODUCTION

Many states such as Maryland,, Minnesota and Texas, Virginia and Washington divert general traffic to the
High Occupancy Vehicle (HOV) lane in order to by-pass a major incident during peak and off-peak hours.
The diversion process is dependent on the configuration of the HOV lane, severity of the incident and agency
policy. A restriction in roadway capacity results in long queues, long motorist delays, greater emissions and
a higher potential for secondary accidents. In order to avoid the impacts of major incidents or severe weather
conditions, many areas containing HOV facilities have implemented HOV diversion strategies in their
incident management strategies. Some locations use various criteria for determining if the HOV lane should
be opened for general traffic use. Often the officer on the scene is responsible for the decision of opening
the HOV lane to single occupant vehicles (1). Once a decision is made to open the HOV lane to general
traffic a diversion plan must be followed and instructions must be clearly expressed to the drivers.

Opening the HOV lane to general traffic is a very politically volatile issue. HOV users expect a certain level
of service which would be disrupted by the adding general traffic. General traffic users also expect efficient
movement of vehicles on the mainlanes which sometimes requires adding the HOV lane to increase the
capacity of the freeway. Criteria boundaries of when opening the HOV lanes to general traffic should be
allowed has not been clearly defined in most locations. Specific criteria used to decide if opening the HOV
lane to general traffic is warranted is presented in this paper along with a typical diversion plan used to
divert traffic. The goal of this strategy is to improve the overall operation of the freeway. The investigation
included an examination of the criteria used to open HOV lanes to general traffic and the actions taken by
agencies during severe weather conditions or major incidents. This research provided a typical diversion
plan that can be applied to an HOV facility. The issues discussed in the diversion plan included how drivers
know to divert to the HOV lane; the ability of a driver to reach a desired exit ramp; the length of time or
distance that the driver will be allowed to travel in the HOV lane as a single occupant; and the reason for
diversion. The diversion plan also included a discussion of possible ITS technologies that can be used in the
plan. However, no specific technology is endorsed because the general plan must be altered for the specific
technologies that are available for each facility.

Objectives

The objectives of this research were to:

• Establish current practices for diverting general traffic into HOV lanes during major incidents or severe
weather conditions.
• Develop a set of general criteria for HOV facilities where the HOV lane should be opened to
accommodate all traffic.
• Develop guidelines for a typical diversion plan for opening HOV lanes to the general traffic.
• Apply guidelines to a specific location.

Scope

The paper will include only freeway applications of HOV lane diversion plans during unplanned events such
as major incidents or severe weather conditions. The HOV facilities are located in major metropolitan areas.
Six agencies responsible for incident management on a variety of freeway HOV facilities were selected for
interviews. Through the interview process the analyst received detailed information of each agency’s
diversion plan. The findings were used to determine a set of general criteria of when to open an HOV lane
to general traffic and to develop a typical diversion plan.
Kevin P. Hoppers Page 206

Report Organization

The report is divided into nine sections. The first section includes the objectives and study approach of the
research project. The study methodology which was used during the research is explained in the next
section. The study methodology includes information on the literature review, data collection methods and
how the results were analyzed. The background section of the report includes a discussion of general HOV
terminology and the importance of incident management.

The results of the survey questions asked of each agency are provided in the fourth section. General
information on each HOV facility analyzed such as occupancy level, type of facility and time of operation
is given in this section. The study results will also provide the agency policies on opening HOV lanes to
general traffic during severe incidents and the HOV facility’s diversion strategy. The fifth section includes
the issues involved in implementing an HOV diversion plan. The developed general criteria used to
determine the opening of HOV lanes to general traffic is included in the sixth section. The next section
provides a set of guidelines for a typical HOV diversion plan. The developed diversion strategy for an actual
HOV facility is described in the eighth section of the report. The final conclusions which include a
discussion of the importance of developing criteria for decision makers and the public acceptance of a
decision are included in the ninth section of the report.

STUDY METHODOLOGY

Literature Review

A literature review was performed to identify a minimum of five HOV facilities to be studied and analyzed.
Information on pertinent HOV lane characteristics including time of operation, occupancy level, type of
system, incident management and diversion policies and any safety and operational issues associated with
such plans were obtained. Also, the literature review provided information on HOV terminology and the
importance of incident management. The information gathered in the literature review search is presented
in the Background and Study Results sections of the report.

Data Collection

The data collection task of the research consisted of questioning six transportation agencies to determine
current practices of diverting general traffic to HOV lanes during major incidents or severe weather
conditions. The questioning process of the transportation officials was guided by a series of questions that
are shown in the Appendix. The individuals that were contacted have a direct connection to the operation
of the incident management strategies on the HOV facilities. The agency officials were asked a variety of
questions that were used to determine the policy on opening HOV lanes to general traffic, criteria used to
determine when the HOV lanes should be opened and the strategies used to carry it out. The agency officials
were also asked to describe the coordination between all of the agencies involved in the incident
management. The following table provides a list of individuals contacted and the organization with which
they are associated.
Kevin P. Hoppers Page 207

Table 1. Questionnaire Contacts

Contact Organization

John Gaynor Texas Department of Transportation, Houston

Mahesh Kuimil Dallas Area Rapid Transit

Joel Katz Minnesota Department of Transportation

Jimmy Chu Virginia Department of Transportation

Michelle Hoffman, Saed Rahwanji Maryland State Highway Administration

Les Jacobson Washington Department of Transportation

Data Analysis and Application

The information presented in the Study Results section of the report was examined to determine the incident
management plans used by various agencies. A list of criteria that determines when an HOV lane should be
opened to general traffic was gathered from the agency interviews. A level of importance of each criterion
was determined from interviews and then included in a general list of criteria of when to allow general traffic
to travel on the designated HOV lanes. From the agency interviews the recommended implementation
strategy was developed. A case study for the proposed guidelines and diversion plan is presented to illustrate
the proposed process.

BACKGROUND

The concept of diverting general traffic to HOV lanes involves two aspects of transportation: incident
management and HOV system operations. The following sections describe these two areas.

Incident Management

The public and transportation officials have recognized the need for well-developed incident management
strategies in order to reduce congestion on the roadways. The capacity on a three lane freeway can be
reduced by 50 percent when an incident blocks only one lane. The incident also has a significant effect on
opposing traffic because of rubbernecking. In a one hour period during an incident, the traffic back-up can
reach 8.5 miles when the traffic flow is near capacity. Observations in Los Angeles indicated that during
off-peak travel periods each minute of incident duration results in 4 to 5 minutes of delay (2).

Incident management is a process to reduce the time to detect an incident, initiate an appropriate response,
clear the incident, communicate appropriate information to motorists and to divert and manage traffic until
the roadway returns to full capacity. Effective incident management involves a coordinated and cooperative
approach among all agencies. Agencies that may be involved in an incident management team are state
departments of transportation, transit agencies, police and fire departments, emergency medical services and
towing companies. The benefits of a sound incident management program includes reduced business costs
associated with delays, fewer secondary incidents and improved air quality due to lower vehicle emissions
(3). An incident can be defined as an accident, disabled vehicle, debris on the roadway and other random
events that reduce the capacity of a roadway. The National Highway Traffic Safety Administration defines
a major freeway incident to include motor vehicle crashes that involve serious fatalities and serious injury,
motor vehicles on fire, crashes where cargo is spilled, overturned vehicles, downed power lines or structural
Kevin P. Hoppers Page 208

failures (4). Unpredictable timing and location of incidents are the main reasons they are difficult to manage.
Most travelers plan for peak hour congestion conditions but delays due to incidents cannot be foreseen. Due
to the increase in traffic volumes, the frequency and impacts of incidents have also become higher.
Secondary accidents are also a concern in incident management (2).

Traffic management during an incident is extremely important. Proper traffic management will reduce the
impact of the incident and provide a safer environment for motorists and responders. In a situation where
diversion from the freeway is necessary, inadequate traffic management will result in lost motorists,
additional surface street congestion and more secondary accidents. Traffic accidents require that the
responders, police, service patrols and private wreckers are equipped with proper traffic control devices and
the traffic management strategies are well planned and understood. Roll up signs, cones and other portable
devices should be standard equipment for first response teams, those first arriving on the scene of an incident
(2).

Motorist information is also a vital aspect of an incident management plan. A well informed motorist can
make a travel decision based on the given information. The motorist can make the decision to divert around
or stay away from an incident site, delay or change the time of a trip or even change travel modes. Variable
message signs (VMSs), highway advisory radio, radio and television reports and traffic reporting services
are valuable in advising motorists of incidents and other traffic related situations (2).

Many strategies are used in incident management plans. Some agencies have used HOV lanes in their
incident management plans as a diversion strategy. Agencies in Dallas, Houston, Maryland, Virginia and
Washington, among others use the HOV lane as a diversion route for vehicles to by-pass a major incident
(1,5,6). The diversion process depends on agency policy, HOV lane configuration and the severity of the
incident (1).

HOV Lanes Description

HOV facilities are being used in many areas to lessen the growing problem of traffic congestion. Many areas
are faced with the dilemma that the projected travel demands cannot be reasonably served at current
occupancy levels. Therefore, the goal of encouraging travelers to use HOV facilities is to increase the
average person-carrying capacity of the roadway without being forced to construct new general purpose
lanes. The characteristics of an HOV lane include occupancy level, type of facility and time of operation.
Each of these characteristics is important to the implementation of a diversion strategy. The type of facility
determines the accessibility of the general traffic to the HOV lane while by-passing an accident (7). Three
types of facilities are discussed in the following sections: concurrent flow lanes, reversible flow lanes and
contraflow lanes.

Concurrent HOV Flow Lanes

A concurrent flow lane operates in the direction of adjacent general purpose lanes and is designed for use
by HOV’s during all or a portion of the day. The lane is normally separated from general traffic by a double
stripe or a buffer area. Concurrent lanes are usually located on the inside lane but they may also be
positioned on the outside lane. Entry to the lane may be limited to specific access points or allowed
anywhere on the facility. The concurrent flow lane is the most common HOV lane type because it is less
expensive to implement than the other types. However, the occupancy level on the concurrent flow lane is
much more difficult to enforce (8). Figure 1 shows an example of a concurrent lane facility.
Kevin P. Hoppers Page 209

Figure 1. Concurrent HOV Flow Lane Along I-405 in Seattle, Washington (9).

Figure 2. Reversible HOV Flow Lane in the Center of I-394 in Minneapolis (10).

Reversible HOV Flow Lanes

Reversible HOV lanes allow the direction of traffic flow to be changed at different times of the day. The
inbound traffic is served during the morning peak while the outbound traffic travels in the lane during the
evening peak. Reversible lanes are separated from the general purpose lanes by concrete barriers and access
is limited by gates and pylons. Figure 2 shows an example of a reversible flow lane (8).
Kevin P. Hoppers Page 210

Contraflow HOV Flow Lanes

Contraflow lanes are lanes taken in the off-peak direction of travel and designated for exclusive use by HOVs
traveling in the peak direction. The lane is separated from the off-peak direction general purpose travel lanes
by some type of changeable treatment such as a moveable concrete barrier, plastic posts or pylons.
Contraflow lanes are typically operated only during the peak period. They are only used when a significant
directional distribution exists on a facility and there is available capacity in the non-peak direction. Figure
3 provides an example of a contraflow lane (11).

Figure 3. Contraflow HOV Lane Along I-93 in Boston, Massachusetts (12)

SURVEY RESULTS

In order to determine specific criteria and develop a typical diversion plan, six agencies responsible for
operations or incident management on HOV facilities were interviewed by telephone. The characteristics
of each HOV facility were determined. The characteristics included type of facility, occupancy level, time
of operation and number of general purpose lanes. Diversion policies used by these agencies were
determined as well as any criteria used to determine when the HOV lane restrictions should be lifted. The
agencies interviewed are listed in alphabetical order. Summary tables of facility characteristics and diversion
policies are provided at the end of this section.

Dallas, Texas

Facility Characteristics

The East R.L. Thornton Freeway Contraflow HOV (I-30) lane opened and began operation in Dallas in
October 1991. The lane is open to vehicles with an occupancy level of 2 or more between 6 a.m. to 9 a.m.
and 4 p.m. to 7 p.m. The inside freeway lane in the off-peak direction of travel is dedicated for HOVs in the
peak direction. Traffic is separated by a moveable concrete barrier. The reversal process which involves
changing and constructing a temporary barricade of pylons takes approximately two hours. There are no
emergency access gates along the length of the barrier. The I-30 HOV lane is 8.4 kilometers in the morning
Kevin P. Hoppers Page 211

(westbound) and 5.3 kilometers in the evening (eastbound). The number of general purpose lanes is three
(5).

The two concurrent HOV facilities operating in Dallas are I-35E and I-635. These lanes are open to vehicles
with at least two occupants 24 hours a day and are separated from the general purpose lanes by a double-wide
stripe. The lanes designate two entrances and two exits. The 24 hour operating time was determined based
on the belief that motorists would be less confused and mainlane off-peak volumes are not heavy enough to
require extra capacity. I-635 is one of the busiest freeways in the United States and is the most used HOV
lane in Texas. Both freeway facilities have three general-purpose lanes in each direction. I-35E and I-635
have been very successful projects for Dallas. The violation rates for the two HOV facilities is 6 percent
which is well below the accepted rate of 15 percent (5).

Diversion Policies

The HOV system in Dallas is operated by the Dallas Area Rapid Transit (DART). DART allows the
diversion of general traffic to HOV lanes under severely congested conditions. The desire of DART is to
maintain the credibility of the HOV lane and only use the HOV lane as a last resort in an incident
management plan. The respondents will attempt to use the shoulders or exit ramps before diverting traffic
onto the HOV lane. HOV lane diversion only occurs an average of five to six times a year on I-635 and I-
35E. The agencies responsible for incident management are DART and the local police. Incident detection
is determined by cameras placed at one-mile spacings along each freeway and surveillance performed by the
Mobility Assistance Patrol, Courtesy Patrol, transit police and local police. The transit police and local
police make the decision to open the HOV lane to the general traffic. DART does not have set criteria for
the respondents to follow to determine when the lane should be opened. However, it is understood that the
single occupant vehicles should not travel on the HOV lane unless the situation is unavoidable. Variable
message signs, cones, flags and media announcements are used to give motorists the proper instructions as
to diverting on the HOV lane. Once the single occupant vehicle (SOV) by-passes the incident on the I-635
or I-35E HOV lane the motorist is expected to merge back onto the general-purpose lane. Directional signs
are not used to direct SOVs back on to the general travel lanes and occupancy level is once again enforced
at a reasonable distance from the incident (5).

If a major incident significantly blocks the I-30 mainlanes, the superintendent of the DART crew stationed
on site recommends to DART administrators whether or not the HOV lane should be opened to general
traffic. Since there are not emergency access gates that can be opened on the non-peak side, the contraflow
facility cannot be reversed to divert non-peak mainlane traffic. Also, only the barrier transfer machine can
move the barrier to allow non-peak traffic into the HOV lane. According to officials the HOV lane diversion
has only occurred a couple of times on I-30 (1).

Houston, Texas

Facility Characteristics

Five barrier separated HOV lanes with an occupancy level of two or more are operated by the Houston
Metropolitan Transit Authority (METRO) and the Texas Department of Transportation (TxDOT). Each
facility is a one-lane facility except for a short two-lane section on US 290. I-45N, I-45S, US 59S, and US
290 allow inbound vehicles to travel between 5 a.m. and 11 a.m. and outbound traffic between 2 p.m. and
8 p.m during the weekdays. The I-45S lane is also open to inbound HOV traffic from 3 p.m. to 9 p.m. on
weekends in the summer (1).

The I-10 HOV lane is one of three congestion pricing demonstration projects currently functioning in the
United States. The I-10 HOV lane requires a fixed toll for vehicles with occupancy levels of two or more
during specified times of the day and are billed through transponders. The facility is open between 5 a.m.
Kevin P. Hoppers Page 212

and 6:45 a.m. and from 8 a.m. to 11 a.m. Vehicles with an occupancy level of three or more travel the HOV
lane for free between 6:45 a.m. and 8 a.m. while vehicles with only two occupants must pay a toll. The
outbound traffic travels on the HOV lane between 2 p.m. and 5 p.m. and 6 p.m. to 8 p.m. Vehicles are tolled
in the same manner as inbound HOVs between 5 p.m. and 6 p.m. Outbound vehicles with an occupancy of
2 or more are allowed to travel on Saturdays between 5 a.m. and 8 p.m. while on Sundays the same
restrictions are used for the inbound traffic (1,13).

Diversion Policies

Houston coordinates the ITS infrastructure through the TranStar traffic management center. TranStar is
jointly operated by Houston METRO, TxDOT, Harris County and the City of Houston. A network of loop
detectors, Automated Vehicle Identification (AVI) readers, VMSs, Closed Circuit Television (CCTV)
cameras and overhead lane control signals are used to monitor the traffic situation. The spacing of the
overhead lane control signals is approximately two miles apart throughout the HOV network (13).

HOV lane diversion is implemented approximately ten times per year. Houston does not have a specific set
of criteria that are followed to make a decision to lift the HOV restrictions. METRO’s police force can
decide to open the HOV lane to SOVs if the major incident causes extreme congestion and there is not an
available diversion route for traffic. VMSs upstream of the incident provide motorists with diversion
information. The messages are changed by METRO officers stationed at TranStar. The media is also used
extensively to give the public information on the traffic situation (13).

Maryland

Facility Characteristics

Two concurrent HOV lanes are operated by the Maryland State Highway Administration (SHA) along I-270.
The lanes are separated from the general-purpose lanes by a two-foot buffer. Vehicles with an occupancy
of two or greater are allowed access to the HOV lanes between 6 a.m. and 9 a.m. for southbound traffic and
from 3:30 p.m. to 6:30 p.m. for the northbound direction. All traffic is allowed to use the lanes outside of
these hours (1).

Figure 4. Concurrent HOV Lane Along I-270 in Maryland (14).

Kevin P. Hoppers Page 213

Diversion Policies

The incident management responsibilities are coordinated through the Maryland SHA, state police and
Montgomery County. Maryland does allow the diversion of general traffic for major incidents and severe
weather but there is no predetermined criteria as when to lower the restrictions on the HOV lane. It depends
generally on the extent of the incident, time of the incident and the length of the queue. A network of VMSs
and highway advisory radio along with media reports, coordinated through the Statewide Operations Center,
informs the drivers that restrictions will be lifted on the HOV lanes due to an incident on the mainlanes.
Once the restrictions on the HOV lane are lifted then the lane remains open to general traffic for the duration
of the peak period because it is difficult to clear the lanes of SOVs after diversion. Field personnel is
required to set up arrow boards and direct the diversion process in order to ensure drivers are aware of the
situation and are clearly told the proper action (6).

Minnesota

Facility Characteristics

The HOV facilities in Minneapolis-St. Paul, Minnesota opened in 1992. Currently, three facilities operate
within the freeway network around the Twin Cities. I-394 west of Highway 100 is a concurrent facility that
is reserved for HOVs with an occupancy level of two or more between 6 a.m. and 9 a.m. for eastbound traffic
and 3 p.m. to 6 p.m. for westbound traffic. This section of I-394 has two general-purpose lanes and two
concurrent HOV lanes. The section of I-394 east of Highway 100 is a reversible, barrier separated lane that
is open to eastbound HOVs between 6 a.m. and 1 p.m. and westbound traffic between 2 p.m. and midnight.
This section of I-394 consists of three general-purpose lanes and two reversible lanes in the middle of the
freeway. The third HOV facility on I-35W is a concurrent facility located between Highway 13 and I-494
and is available to HOVs between 6 a.m. and 9 a.m. and 3 p.m. and 6 p.m in each direction. I-35W has three
general-purpose lanes and two HOV lanes located on the inside of the freeway. SOVs are allowed to use the
concurrent facilities while restrictions are not in place (15).

Diversion Policies

The Minnesota Department of Transportation is responsible for the operation of HOV facilities. The three
HOV facilities operate within a surveillance network composed of loop detectors, CCTVs, regular police
patrol and citizen calls. The cameras on I-35W and I-394 west of Highway 100 are spaced at one- mile
intervals while the cameras placed on I-394 east of Highway 100 are spaced at one-half miles because of the
locations of bridges throughout the corridor. The loop detectors measure the speed and levels of congestion
on each facility. According to officials the most important and reliable indicators of the traffic situation are
calls from motorists. Motorist information is given to drivers by VMSs, highway advisory radio and through
the media (15).

The state police responding to the incident have the authority to implement a diversion strategy. However,
there is not a set policy regarding diversion of general traffic to HOV lanes. To date, the lanes have not been
used for diversion because congestion has generally not been considered high enough to warrant such a
strategy. The policy is that the Minnesota Department of Transportation has made a commitment to HOV
users that the HOV lane will be available during specified operating hours. The reliability of the HOV
facility is the incentive for drivers to use HOV lanes (15).
Kevin P. Hoppers Page 214

Seattle, Washington

Facility Characteristics

The Washington Department of Transportation (WSDOT) operates a number of different types of HOV
facilities. Concurrent facilities operate on I-405, SR 167 and SR 520 while reversible and concurrent HOV
lanes operate on I-5 and I-90. All of the existing I-405 lanes have an occupancy definition of two or greater
which is enforced 24 hours a day. SR 167 contains approximately eight HOV lane miles that are open to
traffic 24 hours a day with an occupancy greater than one between South Grady Way and 84th Avenue South.
Three more HOV projects are planned for the SR 167 corridor that will extend the HOV facility south to
Puyallup. A shoulder lane on westbound SR 520 requires vehicles with three or more occupants. SR 520 is
the only HOV lane on state highways with an occupancy level of three or more that is enforced 24 hours a
day. The existing HOV lane on I-90 includes a reversible HOV roadway with an occupancy requirement of
two or more between Seattle and Bellevue and 24-hour HOV lanes on the westbound and eastbound mainline
between Bellevue and Issaquah (9).

Figure 5. Reversible HOV Lane Along I-5 in Seattle, Washington (9).

The I-5 reversible lanes are open to HOVs with two or more occupants between 5 a.m. and 8:30 a.m. and
from 12 p.m. to 4 a.m. in the peak direction. The number of lanes provided by the facility varies from one
to four. Both facilities contain concurrent sections that are open to HOVs with an occupancy level of two
or more 24 hours a day (9).

Diversion Policies

The traffic conditions must prove to be extremely severe if the HOV lane is opened to SOVs. The WSDOT
try to provide efficient movement on the HOV lanes in order to satisfy carpoolers and bus riders. The
credibility of the HOV lanes with the public is a priority with WSDOT. Therefore, lifting the restrictions
on the HOV lane does not occur very often (16).
Kevin P. Hoppers Page 215

Figure 6. Concurrent HOV Lane Along I-5 in Seattle, Washington(9).

The state police have the authority to open the HOV lane to general traffic when a major incident occurs.
The state police do not have any type of set criteria to follow in order to make a decision. However, the
decision generally depends on the severity of the incident, time of day and the availability of diversion routes
other than the HOV lane. A network of VMSs and portable signs, along with police officers will direct the
traffic around the incident. The time of day that the incident occurs is an extremely important factor to the
decision of opening the HOV lane. WSDOT will normally not open the HOV lane to SOVs during the peak
period because they do not want to cause any delay to bus riders (16).

Virginia

Facility Characteristics

The Shirley Highway HOV lanes serve commuters traveling to and from Washington D.C. in Northern
Virginia. The HOV lane is part of I-95 and I-395 and has four general-purpose lanes in each direction. The
facility is a two-lane, reversible, barrier-separated system. The northbound direction of the Shirley Highway
is open to HOVs from 6 a.m. to 9 a.m. on Monday through Thursdays. It is open to all general traffic from
9 a.m. to 11 a.m. and 10 p.m. to 6 a.m. In the southbound direction, the HOV lane is open to HOVs from
3:30 p.m. to 6 p.m. and to all traffic from 1 p.m. to 3:30 p.m. and between 6 p.m. and 8 p.m. The schedule
on Fridays changes to allowing HOVs in the southbound to travel on the HOV lanes between 12 p.m. and
6 p.m. The lanes are opened to all southbound traffic from 6 p.m. on Friday to 8 a.m. on Sunday and then
reopened to all northbound traffic from 10 a.m. Sunday to 6 a.m. Monday (1,17).

Diversion Policies

The Virginia Department of Transportation (VDOT) and the state and local police are responsible for
incident management on the Shirley Highway. Diverting general traffic to the HOV lanes has been a very
successful and popular idea with the public and local media. It was estimated that the decision to divert
traffic onto the HOV lane occurs approximately 10 times per year. The decision is made jointly between the
state police on the scene of the incident and VDOT. The HOV lane volumes are so high that VDOT does
not want to allow general traffic to divert to the HOV lane often (18). Furthermore, VDOT has implemented
a policy that says if an incident should occur in the non-peak direction, non-peak traffic will not be diverted
to the HOV lane (1).
Kevin P. Hoppers Page 216

The criteria that are used to determine if the HOV lane should be opened to general traffic relates to the time
it takes to clear the incident and the percentage of reduced capacity caused by the incident. If the operation
of clearing a major incident lasts longer than two hours then the restrictions on the HOV lane will be lifted.
Also, when an incident blocks 50 percent of the mainlanes in the peak direction then traffic will be diverted
to the HOV lane. VDOT maintains a network of VMSs to alert drivers of any changes in the HOV lane
restrictions. The VMSs will inform the driver of no restriction if the HOV lane is opened to general traffic.
The restriction will last until the end of the peak period. Weather has rarely caused the opening of HOV
lanes to general traffic. Only in the case of an extremely severe snowstorm where several feet of snow cover
the freeway would VDOT consider lifting the restrictions on the HOV lanes. However, this situation has
rarely occurrs (17).

Table 2. Summary of Facility Characteristics

Location Facility Type Occupancy Time of Operation

I-30 1 lane; contraflow 2+ peak hours in peak

direction
Dallas, Texas
I-35E, I- 1 lane; concurrent 2+ 24 hour
635

I-45N 1 lane; reversible, barrier 2+ Peak hours in peak

I-45S separated (US 290 direction; Weekends on
US59S includes a short two-way I-45S
US290 section)

Houston, Texas I-10 1 lane; reversible, barrier 2+ Peak hours in peak

separated direction; outbound on
(congestion pricing Saturdays; inbound on
project) Sundays

Toll for 2+; 6:45. to 8 a.m. and 5 to

free for 3+ 6 p.m.

Maryland I-270 1 lane; concurrent; buffer 2+; SOVs Peak hours in peak
other times direction

East of HW 100-1 lane; 2+ Peak hours in peak

reversible, barrier direction
separated
Minneapolis-St. Paul, I-394
Minnesota West of HW 100-1 lane; 2+; SOVs Peak hours in peak
concurrent other times direction

I-35W 1 lane; concurrent 2+; SOVs Peak hours in peak

other times direction
Kevin P. Hoppers Page 217

Table 2. Continued

Location Facility Type Occupancy Time of Operation

I-5N, I- 1 lane; concurrent 2+ 24 hours

5S
I-5N 1 to 4 lanes with 2+ Peak hours in peak
ramps; reversible; direction
barrier separated
I-90 2 lanes; reversible, 2+ 24 hours except for
Seattle, Washington barrier separated reversal periods
1 lane; concurrent 2+ 24 hours
SR 167 1 lane; concurrent 2+ 24 hours
SR 520 WB shoulder lane; 3+ 24 hours
concurrent
I-405 1 lane; concurrent 2+ 24 hours
Virginia Shirley 2 lane; reversible, 2+ Peak hours in peak
Highway barrier separated direction

Table 3. Summary of Facility Diversion Policies

Location Agencies Involved Criteria Used to Motorist Information Surveillance

Divert Techniques Techniques

Dallas, Texas DART No specific criteria VMSs, cones, flags and loop detectors,
media announcements cameras, Courtesy
Patrol, citizen reports

Houston, Metro, TxDOT, No specific criteria VMSs, overhead lane loop detectors,
Texas Harris County and control signals, media cameras, AVI, police
City of Houston announcements patrol, citizen reports

Maryland Maryland SHA, No specific criteria VMSs, highway loop detectors,

state police and advisory radio and cameras, police
Montgomery media announcements patrol, citizen reports
County

Minneapolis- Minnesota DOT, No specific criteria VMSs, highway loop detectors,

St. Paul state and local advisory radio and cameras, police
Minnesota police media announcements patrol, citizen
reports,

Seattle, Washington DOT, No specific criteria VMSs, portable signs, loop detectors,
Washington state police police direction cameras, police
patrol, citizen reports

Virginia Virginia DOT, state 50 percent of VMSs, highway cameras, police

and local police mainlanes closed and advisory radio, police patrol, citizen reports
time to clear incident direction
more than 2 hours
Kevin P. Hoppers Page 218

ISSUES TO IMPLEMENTING AN HOV DIVERSION PLAN

Agency interviews and the literature review identified a number of barriers to implementing an HOV
diversion plan. These include agency policy, motorist information, public acceptance and development of
HOV lane diversion criteria.

Agency Policy

A problem must exist on the freeway before a diversion plan that would divert general traffic onto the HOV
lane is implemented. The problems that might exist are heavy volumes on the freeway or a lack of alternate
routes for mainlane diversion during incidents (1). It is important to understand agency policies on lifting
the restrictions on HOV lanes during operating hours. Minnesota has a very strict policy to keep the HOV
operating hours. Due to relatively low congestion levels, most corridors can handle the daily traffic load.
As discussed in the Study Results section of the paper, Minnesota is not opposed to opening the HOV lane
to general traffic during extremely severe situations but the congestion levels have never been high enough
to warrant such an action. However, due to heavy congestion in Virginia, VDOT’s policy allows the opening
of the HOV lane approximately ten times per year. Similarly, due to the high traffic volumes on the Dallas
freeways, DART allows opening HOV lanes approximately five to six times per year. The agency policies
are very dependant upon the mainlane congestion levels of the freeways. Policy also depends on the traffic
load on the HOV lane.

Motorist Confusion

Some agencies choose to operate the concurrent flow lanes 24 hours a day and do not allow general traffic
on HOV lanes outside of peak periods. The reason for this policy is to minimize motorist confusion as to
when the HOV restrictions are in place. Agencies such as DART where the HOV lanes are relatively new
and might be unfamiliar to drivers are afraid too much HOV lane information might cause confusion. Also,
the 24 hour operation time protects infrequent users from misunderstanding the HOV operating periods.

Areas with peak period operation are also concerned with the driver understanding the instructions to divert
to the HOV lane. Clear and concise information must be presented to reduce any confusion for the driver.
A person who is accustomed to driving in the mainlanes might not be familiar with the HOV lane. The
following key pieces of information must be given to a driver in order to provide understanding during a
diversion process.

• Reason for diversion;

• Length of time or distance that the motorist be allowed to continue to drive on the HOV lane;
• Ability to exit on the desired ramps for barrier separated facilities.

In order to implement a diversion strategy in areas with 24 hour a day or peak period operation, careful
attention must be given to informing drivers of current situation. The driver must understand the steps that
they should follow in order to avoid the incident or divert into the HOV lane.

Public Acceptance

HOV lanes were created to increase the travel time savings and provide vehicles with multiple occupants a
less congested environment. Maintaining HOV credibility might mean operating a lane at a specific level
of service or as in the case of Seattle, keeping the average HOV lane speed above 45 miles per hour. The
users depend on the HOV lane to provide significant travel time savings. Otherwise, drivers would be
discouraged from traveling the HOV lane and not see the benefit of carpooling.

If the HOV lane restrictions are lifted too often then the public may question the value of the lane. Negative
public perception may lead to the abolishment of the HOV lane. Also, if restrictions are modified often, then
Kevin P. Hoppers Page 219

SOVs may become so used to traveling in the HOV lane that violations may increase and enforcement
becomes difficult. The SOVs would also not see the benefit of carpooling if the occupancy level is lifted
often.

Development of HOV Lane Diversion Criteria

None of the agencies except one interviewed for this paper had predetermined criteria used to decide if the
HOV lane should be opened for general traffic during major incidents or severe weather. Virginia opens the
HOV lane if an incident blocks 50 percent of the mainlanes in the peak direction and is expected to take two
hours to clear. The majority of the agencies interviewed for this report give the responsibility of the decision
to divert to the response team that arrive on the scene. However, the response team often consults with the
traffic management center or the state department of transportation before making a decision. It is important
for the officers on the scene to have a clear understanding of what type of conditions to look for before
opening the HOV lane to general traffic. A general set of criteria given to the officers or incident
management team would allow a decision to be made quickly and decisively.

The set of general criteria must be a list of what type of conditions should be analyzed before a decision is
reached. Issues including the severity of the situation, time of day, impact on the mainlanes and availability
of diversion routes. The severity of the impact on traffic would account for the percentage of the capacity
reduced on the mainlanes and the time to clear the incident. The availability of alternative routes would
involve using shoulders, ramps or available mainlanes to divert traffic. In order to maintain the credibility
of the HOV lane, diversion to the HOV lane should only take place as a last option. The time of day would
give an indication of how the current and expected traffic would be affected. Information on the current
operating travel volumes on the HOV lane is important in order to determine the type of impact diverting
traffic would have on the existing HOV traffic.

The criteria can be a set of guidelines for the response team so that an informed decision can be made. It
must train the officers to know the road conditions that are vital in determining if HOV restrictions should
be lifted. The decision will ultimately be a judgement call based on the situation of the specific location.
Criteria cannot be developed for every situation and every location. Therefore, the experience and instinct
of the officers on the scene plays an extremely important part in the decision to open the HOV lane to SOVs.

DEVELOPMENT OF GENERAL CRITERIA FOR DIVERSION TO THE HOV LANE

The following section contains guidelines developed by the author that could be used when determining if
the restrictions on the HOV lane should be lifted during a severe situation. It is helpful for response teams
to have a set of criteria of what they should look for in order to make a good decision. Also, a set of general
criteria provides a level of consistency for each decision that is made to open the HOV lane. The criteria is
considered general because it must be tailored to each location. As the interviews revealed, every situation
is different and a set of general criteria is not able to account for each and every situation. These criteria
include determining the severity of the incident, time of day, impact on the mainlane traffic and availability
of alternative routes.

Severity of Situation

As was discussed in an earlier section, a major freeway incident includes downed power lines, infrastructure
failures and crashes involving fatalities, vehicles on fire, spilled cargo, hazardous cargo and overturned
vehicles. The severity of the incident determines the type of response team that is needed. A quick response
is needed for any type of severe accident especially when a person’s life is in jeopardy. A path needs to be
cleared for the emergency response team. The officer on the scene or traffic management team must decide
if diverting traffic to the HOV lane would clear a needed path for the response team. The first responsibility
that the officer or incident management team has is the safety of the victims involved in the incident. The
Kevin P. Hoppers Page 220

weather situation can also make the conditions on the roadway dangerous. Weather conditions can cause
severely congested conditions. It is important for the correct assessment of the accident or weather situation
to be made immediately, before any action is taken.

The severity of the situation determines the length of time that is needed to clear the roadway for oncoming
traffic. The time needed to clear the incident will provide a good estimation of the queue length caused by
the situation. As was mentioned earlier, when one lane on a three lane freeway is blocked during an incident
the resulting back-up grows at a rate of 8.5 miles per hour. Therefore, it is vital to make a quick decision in
order to prevent severe congestion.

Time of Day

The time of day of the incident is extremely important to the decision of opening the HOV lane to SOVs.
The peak period time is the most heavily traveled time of the day and when most commuters are most
dependant on the operation of the system. Motorists use the HOV lanes because they have been convinced
of the significant travel time savings. Citizens that use the transit system have also been told of the savings
in travel time that they will receive. Opening the HOV facility to SOVs during peak periods must be
carefully considered because of the detrimental effect that will occur to the credibility of the HOV lane if
there is a significant loss in travel time savings.

Impact on Mainlane Traffic

An assessment of the impact that the incident is having on the mainlane traffic is important in determining
if opening the HOV lane to general traffic is warranted. Developing performance measures that will assist
the management team in determining the impact that the incident or weather situation has on the freeway
mainlanes is very helpful in the decision making process. The primary impact that a major incident or severe
weather conditions have on the freeway mainlanes is impeding traffic flow by reducing capacity. A reduced
capacity can cause long queues and severe delays.

Only one of the agencies interviewed for this paper provides criteria to quantify the impact the incident or
severe weather has on the mainlanes. Other agencies make a judgement call on the movement of traffic with
no specific guidelines. The majority of the agencies decide to divert general traffic to the HOV lane if a
major incident produces “prolonged congestion.” A set of performance measures used as a guideline for
incident management teams would provide consistency in the decision making process.

The response team should be in communication with the traffic management center to determine the correct
course of action. A set of measures such as volume-capacity ratio, percentage of reduced capacity and
estimated length to clear the accident should be used in the decision making process. The traffic
management center could provide information on the volume-capacity ratio to the response team from
surveillance cameras and loop detectors. The response team would be able to assess the seriousness of the
situation and estimate the length of time needed to clear the accident and the percentage of the capacity
reduced.

The author suggests that a certain level of traffic on the freeway be present before HOV diversion is
considered. A volume-capacity ratio of 0.7, which is a level of service D according to the 1997 Highway
Capacity Manual, is a reasonable threshold value that should exist before the general traffic is diverted to
the HOV lane. Once an incident occurs, the goal of the diversion plan is to maintain a volume-capacity ratio
of 0.95 throughout the duration of the incident. A freeway can theoretically sustain a volume-capacity ratio
of 1.0. However, the volume-capacity ratio of 0.95 was chosen because 1.0 is potentially unstable and
anything above this value could cause the freeway to breakdown. The threshold value for reduction in
capacity for the freeway is suggested by the author to be 50 percent for an estimated time to clear the incident
of 1 hour. Fifty percent was chosen because a single lane blockage on a three-lane freeway results in a
reduction of fifty percent of the capacity (19). A duration of the incident of 1 hour was chosen because the
Kevin P. Hoppers Page 221

average incident takes around 45 minutes to clear (20). If the capacity is reduced by 50 percent on a freeway
for more than an hour with a volume-capacity ratio of 0.7 then the alternative routes must be able to handle
33 percent of the traffic on the freeway in order to maintain the desired volume-capacity ratio of 0.95. If the
alternative routes off of the freeway and the remaining mainlanes cannot handle this amount of traffic then
the HOV lane should be opened.

Availability of Alternative Routes

Decision makers have a responsibility to the HOV users to maintain a facility that reduces the travel time
for multi-occupant vehicles. A commitment has been made that the lane will be available during the
specified operating hours. Reliability is part of the incentive for the HOV lanes. The lane becomes especially
valuable to users when incidents or bad weather causes additional congestion on the general purpose lanes.
Therefore, alternative routes other than the HOV lanes should be the first option in a diversion plan. In a
diversion situation, the availability of shoulders and ramps should be determined before a decision to
lowering the restrictions on the HOV lanes. The criteria presented in the previous section states that if the
route off of the freeway and the remaining mainlanes is able to handle at least 33 percent of the freeway
traffic then the HOV lane does not need to be opened to general traffic. Coordination with the traffic
management center would give the response team the needed information to determine if the alternative
routes can handle the desired traffic. It should be understood that the specific numbers could be changed to
accommodate a specific location or traffic situation but the theory would remain the same.

GUIDELINES FOR TYPICAL HOV DIVERSION PLAN

The following section includes guidelines that should be followed when developing a typical HOV diversion
plan. Several steps must be performed in developing a plan. These include identifying the corridor, forming
agency partnerships, defining criteria used to make a decision to divert, determining alternate routes,
developing effective signing and other traffic control devices, coordinating with the media and implementing
the plan.

Identify the Corridor

The corridor that will be used in the diversion plan must be established. For plans using HOV lanes as a
diversionary strategy, a facility must be located within the corridor. The corridor must also have a history
of traffic problems due to severe accidents. The traffic problems may include congestion as well as
secondary accidents due to the freeway conditions (21).

Form Agency Partnerships

It is vital to develop strong agency partnerships in any type of incident management plan. The plan presented
in this report is not a full incident management plan but only one aspect of a plan. However, in order for any
type of management plan to be successful it is important to establish good coordination between agencies.
Only through strong partnerships can the diversion plan between jurisdictions run smoothly. Good
communication and partnership must also be maintained between all of the groups that are involved in the
incident management team. Often agencies such as the police, state department of transportation and transit
agency work together to keep traffic running efficiently during an accident.

Determine Alternative Routes

Once the diversion corridor is identified it is important to develop a diversion route for the corridor.
Developing a diversion route involves examining the major roadways in the area to determine which ones
would be able to handle the traffic load from the freeway. Also, the location of where traffic should be
diverted need to be analyzed. Areas with high incident rates will need to be taken into consideration during
Kevin P. Hoppers Page 222

the selection process. The current location of existing static or VMSs should be considered when
determining a possible route.

The diversion routes should be considered before opening the HOV lane to general traffic. In order to
maintain the credibility of the HOV lane it is important keep HOV users satisfied with the level of service
on the HOV facility. Therefore, the HOV lane could be used if the traffic congestion is severe and planned
diversion routes are not available.

Through communication with the traffic management center, the response team is able to determine the
conditions of the diversion route. If the diversion route is considered heavily congested then the officers
should consider lifting the restrictions of the HOV lane. Through loop detectors and cameras placed in
strategic locations, the traffic management center can decide the impact the traffic from the freeway will have
on the diversion route.

Define Criteria

Criteria that determine the warrants of opening an HOV lane to divert general traffic are important to
establish. The response team must have an idea of what to look for in order to make a good decision on
whether traffic should be diverted. A general set of criteria was presented in an earlier section. It is
important to note that every accident situation is different so at times the traffic situation might not meet the
general criteria but still should be diverted to the HOV lanes.

Development of Signing and Other Traffic Devices

Once the HOV lane is opened it is essential to inform the drivers of the traffic situation and the state of the
HOV facility. The decision points on the freeway will include entrance points on the freeway, entrance
points to the HOV lane, exit points to the freeway and exit points to the HOV lane. The signs should be in
the form of VMSs in order to provide the motorist with enough information to make a decision. The message
should be concise while still providing all of the information necessary. An example of a message that might
be used would be as follows:

MAJOR ACCIDENT
1 MILE AHEAD
USE HOV LANE

Other technology that could be used to inform the motorists of a severe traffic situation is flashers. Flashers
that are blinking would signify that traffic in the HOV lane is restricted. However, if the flashers are turned
off then SOV traffic is allowed to enter the HOV lane. If drivers are familiar with the flashers, then they can
be a very effective tool to provide information. Flashers would be located before the entrance to the HOV
lane.

It is important to determine the policy if the HOV lanes will remain open for the duration of the operating
period or if they will be available to all traffic while the accident is being cleared. For HOV facilities with
24 hour operating periods, the author suggests that the traffic only be diverted around the accident in the
HOV lane and then return to their proper lane as soon as possible. However, so as not to create too much
confusion, for facilities with specific operating periods it is suggested that once the restrictions are lifted then
the facility be open for the duration of the peak period.

Coordination With Media

The media is a very useful tool in providing the public with information of the traffic situation. The incident
management team should use the media to give information on freeway conditions. The media should be
used in providing the following type of information.
Kevin P. Hoppers Page 223

• Level of congestion on the freeway

• Suggest alternate routes other than the freeway
• Inform the public if general traffic is being diverted to the HOV lane in order to avoid an accident location
• Inform the public of where they will be allowed to exit if they choose to divert to the HOV lane.

The television media or radio media could provide this information to motorists.

Final Implementation

The HOV lane diversion strategy is ready to be implemented once the necessary steps have been taken. Prior
to implementing the diversion strategy, all message signs should be inspected in order to ensure that they will
provide the correct information to the motorist. Once implemented, it would be beneficial to the safety and
efficiency of the diversion plan if there was a continuous evaluation of the plan to ensure that the freeway
operation improves once the HOV lane is opened. Also, an evaluation to conclude that the HOV lane is able
to handle the extra traffic without limiting its effectiveness should be conducted. The cooperating agencies
should perform the evaluation to ensure that the plan meets the proposed objectives. The main criteria in
evaluating the implementation plan should be driver safety and the level of service of the mainlanes. A
record of secondary incidents should be logged in order to determine how the implementation plan improved
the safety of the motorists. The traffic management center should be able to determine the operating
conditions of the general-purpose lanes and HOV lane from the information gained by the network of
cameras and loop detectors. A third measure of the effectiveness of the plan is the extent of the public
support of the decision. The satisfaction of the public is an extremely important factor in evaluating the
implementation plan. A survey of users and non-users of the HOV facility should be performed to determine
the public’s opinion on whether the HOV lane should have been opened to general traffic.

CASE STUDY: I-270 IN MONTGOMERY COUNTY, MARYLAND

The guidelines presented in the previous section will be applied to the HOV facility on I-270 in Montgomery
County, Maryland to illustrate how to implement a plan that allows SOVs to divert to the HOV lanes during
major incidents and severe weather conditions. The area was selected because of available information on
traffic patterns, lane configurations and surveillance equipment used to detect the traffic situation. The
following figure shows the current HOV lane sequencing and intelligent transportation systems used on I-
270.

Identify the Corridor

The need for a diversion plan on I-270 is obvious after examining the traffic patterns along the freeway. The
Maryland SHA operates two concurrent HOV lanes that are restricted to vehicles of an occupancy of two
or greater from 6 a.m. to 9 a.m. for southbound traffic and between 3:30 p.m. and 6:30 p.m. for the
northbound direction. An incident would have a major impact on the traffic traveling on I-270. I-270
between the Capital Beltway and Clarksburg Road or MD 121 is a heavily traveled route for commuters.
The following table includes information on traffic volumes on I-270 in the Rockville area.
Kevin P. Hoppers Page 224

Table 4. Traffic Volumes For I-270 at Montrose Road in Rockville, Maryland (14)

AM Peak Period (6:00-9:00) PM Peak Period (3:30-6:30)

Average Volume in the HOV 1225 1600

Lane(Vehicles/Lane/Hour)

Average Volumes in the SOV 2030 1985

Lanes(Vehicles/Lane/Hour)

Average # of People in the HOV 2735 3830

Lane(People/Lane/Hour)

Average # of People in the SOV 2225 2035

Lane(People/Lane/Hour)

Total Volume(Vehicles/Hour) 9360 9540

The average volumes in the SOV lanes are high enough to cause severe delays and long queues if impeded
by a severe incident.

Agency Partnerships

The agencies involved in the HOV diversion plan would include the state police, county police and
Maryland SHA. The local agencies will also be involved because some of the diversion routes selected
include local roadways. The state agencies will be in close contact with the cities directly involved which
are Rockville, Gaithersburg, Germantown and Clarksburg. Even though the state agencies will handle the
majority of coordination for the diversion plan, it is important to establish good relationships with the local
agencies in case any future problems occur that would require a closer partnership. The state agencies will
be responsible for detecting the incident and operating the VMS in order to provide diversion information
and providing details to the media that will assist motorists.

Determine Alternative Routes

The alternative routes used to avoid the incident should be considered before the HOV lane restrictions are
lifted. The availability of the 14-foot shoulders along I-270 should be analyzed by the response team to
determine if they would be adequate to divert traffic around the incident. Emergency vehicles must be
allowed access to the incident. If using the shoulder as a diversion route for the general traffic would impede
the emergency vehicles then another route must be chosen. The available lanes on the freeway should also
be used for traffic to travel past the incident if possible.

If the queue length and delay are so severe that traffic is not moving efficiently, then a route off of the
freeway should be determined. In the case of I-270, it was decided that MD 355 would be the best alternative
route to divert the traffic through the corridor. MD 355 runs parallel to I-270 throughout Montgomery
County. MD 355 is located east of I-270.

If the condition of the traffic situation on MD 355 is severe then the HOV lane should be considered. The
criteria set in the following section should be used to determine if the HOV lane should be opened to general
traffic. Since the HOV lanes on I-270 are concurrent lanes separated from the mainlanes only by a buffer
they are very accessible to the general traffic. The HOV lanes are only restricted during peak hours.
Therefore, in order to avoid driver confusion it is suggested that if the restrictions are lifted then the HOV
lanes should remain open to all traffic the duration of the peak operating period.
Kevin P. Hoppers Page 225

Figure 7. I-270 HOV Lane Sequencing Map (14)

Kevin P. Hoppers Page 226

Define Criteria

The criteria used to decide if the HOV lane should be opened needs to be determined. The decision should
be made based on the severity of the incident, time of day, impact on mainlane traffic and availability of
alternate routes. As stated earlier in the report, the HOV lane should only be used as a last option. It is
important to provide reliable travel time savings for the HOV users. Therefore, the incident must be
considered a major incident or severe weather condition that has the potential to incur significant delay on
the mainlanes. If the capacity is reduced by 50 percent on a freeway for more than an hour with a volume-
capacity ratio of 0.7 then the alternative routes must be able to handle 33 percent of the traffic on the freeway
in order to maintain the desired volume-capacity ratio of 0.95. If the alternative routes off of the freeway
and the remaining mainlanes cannot handle this amount of traffic then the HOV lane should be opened. The
availability of the alteranitve routes would be determined by the TMC. These criteria need to be taken into
account but the decision is ultimately a judgement call by the response team on the scene and in the traffic
management center.

Development of Signing and Other Traffic Devices

Motorists need to be well informed of any traffic situation which might cause them significant delay and
informed how to avoid the problem. According to Figure 7, Maryland State Highway Administration has
three variable message signs on I-270 located in the vicinity of Rockville. They are located in the following
locations:

• Prior to the I-495 Split

• Prior to Exit 4
• South of Middlebrook Road

The messages on the signs need to be carefully worded so that the motorist will receive the proper
information in order to avoid any serious delays. The message would need to contain a problem statement,
location and an action statement. The following is an example of what should be included on the variable
message sign if the HOV lane has been opened to all traffic.

MAJOR ACCIDENT
2 MILES AHEAD
USE HOV LANE

Coordination With the Media

The most successful way to reach the public with traffic information is through the media. The radio and
television media are very effective ways to provide information. The incident management team should
inform the media to give information on I-270. If the HOV lane is being opened to general traffic then the
following information should be provided to the public:

• Level of congestion on the freeway

• Location of the accident
• Suggested alternate routes other than the freeway
• HOV restrictions have been lowered for the duration of the peak period

This information can be broadcast on newscasts, radio stations and on the internet. Traveler’s Advisory
Radio (TAR) should also be used extensively. I-270 has a network of towers used for sending the signal for
the TAR which is shown in Figure 7.
Kevin P. Hoppers Page 227

Final Implementation

Once all of the agencies have formed partnerships, diversion routes have been chosen, criteria has been
selected, variable message signs are in place and the communication with the media has been established,
it is important to test the equipment before the first incident or major weather storm. The signs should be
controlled at one central location. The Traffic Management Center (TMC) is located in Rockville as shown
in Figure 7. The TMC has the capability to operate traffic devices. The catalog of possible messages to be
shown on the variable message signs should be entered into the computer system.

An evaluation of the HOV diversion plan should be constantly performed. The improvement of traffic
movement on the mainlanes should be determined by analyzing length of queue and the volume-capacity
ratio on the freeway. The cameras and detectors should be used to determine this information. The diversion
routes should be evaluated to determine if they were overloaded and if they caused significant delay to
motorists and bus riders. Public opinion should also be logged to determine if the public supported the
decision to lift the HOV restrictions.

CONCLUSIONS

Providing specific criteria for a response team in order to make a good decision is extremely useful.
However, the developed criteria cannot account for every situation and location. Therefore, the decision is
heavily based on the good judgement and experience of the response team. One of the information contacts
used in this report said that making a decision to open the HOV lane to general traffic during a severe
situation “is more of an art than a science” (21 ). None of the agencies that were interviewed except Virginia
even cited specific guidelines that are followed to determine if the HOV lane restrictions should be lifted.
However, it is the belief of the author that more specific criteria would benefit the response team because
they would have more reliable information upon which to base their decision.

Another problem in opening the HOV lane to general traffic is the political aspect of the subject. Agencies
do not want to disrupt the flow of the HOV lanes because of the commitment to travel time savings that has
been made to the users. The users expect a specific level of service each day on the HOV lane. However,
in order to benefit the total freeway system it is sometimes necessary to open the HOV lane up to all traffic.
Obviously, this will limit the effectiveness of the HOV lane. The general traffic would not accept severely
congested conditions on the mainlanes due to a severe situation while the HOV lane is moving at free flow
speed. Agencies are forced to walk a very fine line in order to please both the HOV users and general traffic,
some of whom are against HOV lanes. It is the belief of the author that specific criteria will benefit the HOV
diversion process because by following certain guidelines for each decision a level of consistency will be
established. It is believed that consistent decision making will ease the tension of the users of the HOV lane
and general purpose lanes because they will better understand the reason for opening the HOV lane to
general traffic. Therefore, good decisions must be made in the field that are justified by specified criteria.
Kevin P. Hoppers Page 228

ACKNOWLEDGMENTS

This paper was prepared for Advanced Surface Transportation Systems, a graduate course in Civil
Engineering at Texas A&M University. The author would like to extend a special thanks to Thomas Hicks
for his guidance and contribution of time throughout the program. Also, the author like to express gratitude
to Dr. Conrad Dudek for organizing the program and the remaining mentors, Patrick Irwin, Wayne Kittelson,
H. Douglas Robertson, William Spreitzer and Carol Zimmerman for their participation in this program. By
sacrificing their time and sharing their knowledge in the transportation field they have provided an invaluable
experience to the graduate students involved. Finally, the author would like to thank the following
transportation professionals contacted throughout this research for sharing their thoughts and opinions
regarding the issues of this paper:

Mahesh Kuimil, Dallas Area Rapid Transit

John Gaynor, Texas Department of Transportation, Houston
Michelle Hoffman, Maryland State Highway Administration
Saed Rahwanji, Maryland State Highway Administration
Joel Katz, Minnesota Department of Transportation
Les Jacobson, Washington Department of Transportation
Jimmy Chu, Virginia Department of Transportation
Bob Marburg, WTOP Radio.

REFERENCES

1. Blume, Kelly. “Implementation of a Dynamic HOV Lane.” Compendium: Graduate Student Papers on
Advanced Surface Transportation Systems. Texas Transportation Institute, Texas A&M University,
College Station, Texas (August 1998).

2. Judycki, D.C., and J.R. Robinson. “Managing Traffic During Nonrecurring Congestion.” ITE Journal,
Volume 62, Number 3. Institute of Transportation Engineers, Washington D.C. (March 1992).

3. Court, T., S. Prosi, A. Alonzi, And W. Stoeckert. “Using Freeway Incident Management to Reduce
Congestion.” Public Works, Volume 125, Number 9. Public Works Journal Corporation, Ridgewood,
New Jersey (August 1994).

4. National Highway Traffic Safety Administration. Highway Safety Desk Book. From
http://www.nhtsa.dot.gov (June 1999).

5. Conversation with Mahesh Kuimil of Dallas Area Rapid Transit (June 1999).

6. Conversation with Saed Rahwanji of the Maryland State Highway Administration (June 1999).

7. Turnbull, Katherine F., R.W. Stokes, and R.H.Henk. “Current Practices in Evaluating Freeway HOV
Facilities.” Transportation Research Record 1299. Transportation Research Board, Washington D.C.
(1991).

8. Texas Transportation Institute, Parsons Brinkerhoff, and Pacific Rim Resources. HOV Systems
Manual. National Cooperative Highway Research Program Report 414. National Cooperative Highway
Research Program, Washington D.C. (1998).

9. Washington Department of Transportation. http://www.wsdot.wa.gov

Kevin P. Hoppers Page 229

10. Minnesota Department of Transportation. http://www.dot.state.mn.us/tmc/program/hov.fhtml

11. Lomax, T.J. Fort Worth HOV System Study: Overview of National Experience and Local Issues. Texas
Transportation INSTITUTE, Texas A&M University, College Station, Texas (September 1995).

12. Massachusetts Highway Department. http://www.ma.us/mhd/hov/hovmain.htm

13. Conversation with John Gaynor of the Texas Department of Transportation, Houston.

14. Montgomery County Department of Public Works and Transportation. http://www.dwpt.com

15. Conversation with Joel Katz of the Minnesota Department of Transportation.

16. Conversation with Les Jacobson of the Washington Department of Transportation

17. Conversation with Jimmy Chu of the Virginia Department of Transporatation.

18. Turnbull, Katherine, R.H. Henk, and D.L. Christiansen. “Suggested Procedures for Evaluating the
Effectiveness of Freeway HOV Facilities.” Texas Transportation Institute, Texas A&M University,
College Station, Texas (February 1991).

19. “Freeway Systems (Chapter 6),” Special Report 209: Highway Capacity Manual (third edition),
Transportation Research Board, Washington D.C. (1997).

20. Kaan, Ozbay and Kachroo Pushkin. “Incident Management in Intelligent Transportation Systems.”
Artech House, Boston, Massachusetts (1999).

21. Durkop, Brooke. “The Use of Dynamic Message Signs for Diverting Traffic During Freeway Incidents.”
Compendium: Graduate Student Papers on Advanced Surface Transportation Systems. Texas
Transportation Institute, Texas A&M University, College Station, Texas (August 1998).

22. Conversation with Bob Marburg of WTOP Radio.

APPENDIX

Guideline Questions for Interviews

1. Describe your general operating characteristics, i.e., hours of operation and occupancy requirements.

2. What are the indicators of the traffic situation, i.e., loop detectors, radar, video, regular police patrol?

3. Which agencies are responsible for the incident management and HOV lane operations in your area?

4. Do you allow general traffic to divert to HOV lanes during major incidents and/or severe weather
conditions?

5. Do you allow the use of the HOV lane by general traffic for holiday traffic?

6. What criteria are used to determine if the HOV lane should be opened to the general traffic during
serious unplanned incidents?
Kevin P. Hoppers Page 230

7. Who makes the decision to open the HOV lane?

8. What is the diversion plan that is used to divert general traffic into the HOV lane?

9. What type of technology is used in the diversion plan?

10. What specific information does the driver need in order to successfully divert? How do you
communicate with them?

11. What assumptions are made about the motorists?

12. How long is the general traffic allowed to travel on the HOV lane?

13. What is your experience with the success of your diversion strategies? What specific problems have you
encountered?

14. Are there performance measures that you use to determine the effectiveness of the plan, i.e., motorist
delay, queue length, estimated secondary accidents or estimated flow rate before and during the incident?

15. How many motorists choose to divert? Are they forced to divert?

16. How do the different agencies coordinate in the event of a major incident or severe weather conditions?

17. Is there an operations center where information from the police and/or other highway operating personnel
to the other agencies involved is relayed?

18. For 24-hour HOV lanes, when are the general traffic forced back on the HOV lane?

KEVIN P. HOPPERS

Kevin P. Hoppers received his B.S. in Civil Engineering from

Texas A&M University in May 1998. Kevin is currently pursuing
his M.S. in Civil Engineering with an emphasis on transportation
from Texas A&M University and plans to graduate in December
1999. During the summers of 1996 and 1997, Kevin worked for
Parsons Transportation Group in Dallas, Texas. Currently, Kevin
is employed as a Graduate Research Assistant at the Texas
Transportation Institute. Kevin has been actively involved with
the student chapters of the Intelligent Transportation Society of
America and the Institute of Transportation Engineers where he
currently serves as the New Student Coordinator. His areas of
interest include transportation operations and design.
 APPLICATION OF ITS TECHNOLOGY
TO HURRICANE EVACUATION ROUTES

by

Renee Hulett

Professional Mentor
Douglas Robertson, Ph.D., P.E.
Science Applications International Corporation

Course Instructor
Conrad L. Dudek, Ph.D., P.E.

Prepared for
CVEN 677
Advanced Surface Transportation Systems

Department of Civil Engineering

Texas A&M University
College Station, Texas

August 1999
Renee H ulett Page 232

SUMMARY

This study investigated the enhancement of hurricane evacuation procedures through the application of ITS
technologies. The objectives inherent to accomplishing this goal include identification of areas with
hurricane evacuation practices, assessment of the effectiveness of current practices, identification of
problems and voids of current practices, identification of feasible ITS technologies for application,
development of hurricane evacuation guidelines, and application of the guidelines to a case study.

The development of a set of hurricane evacuation guidelines included an extensive literature review and
numerous interviews. The literature review consisted of an assessment of current evacuation practices,
recommended evacuation practices, and available ITS technologies. Current evacuation practices were
further assessed in interviews with various state emergency management agencies along the Gulf and Atlantic
coasts.

An assessment of current evacuation practices included the following topics: training, evacuation route
development, coordination agreements, equipment and personnel resources, deciding when to evacuate,
public notification, communications, traffic control devices, traffic management, and agency experience.

Both existing and developing ITS technologies were considered for use during evacuations. While existing
technologies may be applied at present, developing technologies may be included in the planning process
for the future. ITS technologies of particular concern include geographic information systems, advanced
traveler information systems, real-time traffic detection, real-time information displays, in-vehicle traveler
information systems, and laptop communication systems.

Based on the literature review and interviews, the author established a general set of hurricane evacuation
guidelines pertaining to practices and technologies that have proven to be effective and may also potentially
solve some of the problems associated with existing practices. These guidelines encompass the areas of
planning and preparation, making the evacuation decision, public notification, traffic management during
the evacuation, and preparation for the future. A case study of Galveston, Texas illustrated the application
of the evacuation guidelines.

Overall, hurricane evacuation procedures must be continually assessed and updated to reflect changes in the
transportation system and to further improve safety and efficiency. The author’s guidelines should be applied
to specific regions when feasible. Additionally, developing ITS technologies should be incorporated into
the planning process to prepare for future implementation.
Renee H ulett Page 233

TABLE OF CONTENTS

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Research Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
Organization of Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236

BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Transportation Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
Agency Involvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
Hurricanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
Hurricane Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
Hurricane Intensity Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
Hurricane Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
Hurricane Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239

EVACUATION PRACTICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239

Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Evacuation Route Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Clearance Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Coordination Agreements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Equipment and Personnel Resource Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Equipment Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Personnel Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Deciding When to Evacuate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
National Weather Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
NOAA Weather Radio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
Emergency Alert System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
Emergency Managers Weather Information Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
HURREVAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
SLOSH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
The Public . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Public Awareness Campaigns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Public Notification of Evacuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Communications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Transportation System Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
Traffic Control Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
Channelizing Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
Static Signs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
Traffic Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Emergency Operations Center . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Law Enforcement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Signal Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Reversible Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
High Occupancy Vehicle Lane Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
Shoulders as Temporary Travel Lanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
Suspension of Tolls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
Tow Truck Usage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
Renee H ulett Page 234

On-Street Parking Restrictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246

Minimization of Left-Turning Movements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Vehicle Restrictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247

AGENCY EXPERIENCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247

Common State Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
Florida . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
Georgia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Maryland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
Mississippi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
North Carolina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
South Carolina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
Texas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
Problems Identified by State Agencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Suggested Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251

ITS TECHNOLOGIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251

Geographic Information Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Advanced Traveler Information Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
Automated Public Information System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
World Wide Web . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
Traveler Information Kiosks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Real-Time Traffic Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Closed Circuit Television Cameras . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Real-Time Information Displays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Changeable Message Signs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Lane Control Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
In-Vehicle Traveler Information Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Highway Advisory Radio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Automatic Vehicle Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
In-Vehicle Computer Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
Laptop Communications System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254

HURRICANE EVACUATION GUIDELINES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254

Planning and Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Making the Evacuation Decision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Public Notification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Traffic Management During the Evacuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Preparation for the Future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256

CASE STUDY: GALVESTON, TEXAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256

Evaluation of Current Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
Validation of Practices Not in Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
Good Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Recommendations for Galveston . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257

CONCLUSIONS AND RECOMMENDATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259

Future Research Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
Renee H ulett Page 235

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260

APPENDIX A: SAMPLE INTERVIEW QUESTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262

Renee H ulett Page 236

INTRODUCTION

Hurricane season lasts from June through November and presents great concern to residents of barrier islands
and coastal regions of the United States. Areas along the Gulf and Atlantic Coasts are of particular concern
due to frequent and severe hurricanes in past years. Due to the severity of many hurricanes, sizable coastal
areas must be evacuated to ensure the safety of residents and visitors alike. Such large-scale evacuations
cause extensive traffic problems due to massive numbers of people using the street and highway networks
at the same time. Not all transportation systems are equipped to handle such evacuation situations.

Increasing efforts to prepare for and respond to hurricanes have led to the formation of numerous government
agencies at all levels to specifically handle emergency situations. These agencies vary in methodology and
approach, but all maintain the same common goal: safe and efficient evacuation of people during emergency
situations such as hurricanes. Much can be learned from the emergency management systems developed by
these agencies with regard to handling evacuation situations. However, more research can help to improve
hurricane evacuation procedures.

Problem Statement

Although many government agencies have emergency management systems in place, the question remains:
Are they working effectively? Hurricane evacuations still produce heavy congestion and raise safety
concerns as motorists try to evacuate. Evacuations must be conducted in a timely manner to ensure evacuees
reach a safe destination before the hurricane arrives. Further documentation of transportation operations
during hurricane evacuations is needed to assess safety and efficiency issues. In addition, the application
of ITS technology may help increase safety and efficiency and decrease the time required for evacuation.
Documentation of these issues will provide the basis for a set of hurricane evacuation guidelines.

Research Objectives

The main purpose of this research project is to assess current hurricane evacuation procedures and make
suggestions for improvement with the application of ITS technology. The objectives inherent to attaining
this goal include to:

• Identify areas with hurricane evacuation procedures;

• Assess the effectiveness of current practices;
• Identify problems and voids of current practices;
• Identify feasible ITS technologies for use in hurricane evacuation procedures;
• Develop hurricane evacuation guidelines using effective existing practices and ITS technologies; and
• Apply these guidelines to an area with hurricane evacuation needs.

Scope

This study is limited to analysis of the evacuation process within a specific region with regard to preparation
and the actual evacuation. It does not focus on coordination with destination cities of evacuees and directing
traffic to shelter areas, but may touch on these areas as part of preparation. The task of managing traffic
returning to the region once the hurricane no longer poses a threat is not examined.

Organization of Report

The literature review for this paper extends through the Background, Evacuation Practices, and ITS
Technologies sections. The Background section includes transportation emergencies and the agencies that
handle them, as well as a general overview of hurricanes in general. Evacuation procedures utilized by
various agencies, as well as practices that have been recommended by experts, are summarized in the
Renee H ulett Page 237

Evacuation Practices Section. The Agency Experience section contains the results of interviews about
evacuation practices used by various state emergency management agencies. The ITS Technologies section
includes available ITS technologies that could be applied to evacuation procedures. Based on the Evacuation
Practices, Agency Experience, and ITS Technologies, the section on Hurricane Evacuation Guidelines
outlines recommended practices available for enhancement of current evacuation systems. These guidelines
are applied to Galveston in the Case Study: Galveston, Texas section. Finally, Conclusions and
Recommendations are set forth in the last section.

BACKGROUND

Most levels of government set up agencies specifically for the management of all transportation emergencies.
These agencies include hurricane evacuations in their scope, and hurricanes must be understood in general
before one can develop evacuation procedures.

Transportation Emergencies

Transportation emergencies occur when an extraordinary event affects the transportation system and causes
congestion, delay, confusion, and/or general disruption of one or more modes of transportation in such a way
that the impacts are not part of day-to-day transportation operations (1). Traffic impacts may be severe, of
long duration, or require special motorist reactions. Transportation emergencies may result as one of three
types of hazards: 1) natural or weather related, 2) technological, and 3) civil/political. Table 1 summarizes
the possible hazards in each category (2,3). Some emergencies, such as tornadoes or hazardous materials
spills, are reactionary and cannot provoke response until they occur. Other emergencies, such as hurricanes
and nuclear/chemical plant disasters, can be planned for proactively.

Table 1. Possible Hazards that Cause Transportation Emergencies (2,3).

Natural Technological Civil/Political

Hurricanes Hazardous Materials Incidents Terrorism
Flooding Radiological Incidents Civil Disorder/Riots
Tornadoes Nuclear Power Plant Incidents Weapons of Mass
Volcanic Eruptions Chemical Plant Incidents Destruction
Wildfires Utility or Telecommunications Bridge or Pavement Failure
Fog Failures Major Accidents
Ice and Snow Storms Computer Viruses
Earthquakes
Asteroid and Comet Impacts

The use of forecasting technology allows prediction of hurricane arrivals 24 to 36 hours prior to landfall,
which provides key agencies with time to prevent the loss of lives through the implementation of evacuation
measures. In this regard, evacuation planning can be conducted far in advance of hurricane season to prepare
evacuation plans and procedures for use should the need arise. Advance planning includes identification of
hurricane frequency, anticipated hurricane impacts on a region, and mitigation procedures to minimize the
impacts and return the environment to a state of normalcy (4).
Renee H ulett Page 238

Agency Involvement

With respect to hurricane evacuations, the federal, state, and local levels conduct transportation planning.
Originally, the duties of the Federal Highway Administration included the development of a national
emergency highway traffic regulation plan in 1969 (1). In 1979, the Federal Emergency Management
Agency was created to assist state and local governments with natural disaster preparedness planning. Their
mission is to reduce the loss of lives and property and to use an emergency management program of
mitigation, preparedness, response, and recovery to protect the nation’s infrastructure (5). In addition, the
Office of Emergency Transportation, a branch of the United States Department of Transportation, has
responsibility for the emergency preparedness of the transportation system (1). They operate a 24-hour Crisis
Management Center which serves as a central location for emergency transportation response (6).

Apart from the federal agencies involved, each state possesses some type of emergency management agency
to handle major emergencies. The federal government provided a Guide for Emergency Highway Traffic
Regulation in 1985 to provide guidance and encourage each state to develop specific emergency plans for
their particular political, geographic, and traffic characteristics (1). Each state emergency management
agency serves as the lead agency and works with other supporting state agencies, private organizations, and
volunteer organizations. They also encourage each local region to develop emergency plans of their own.
When the threat of a hurricane approaches, the local government usually makes the official decision of
whether or not evacuation is needed.

Hurricanes

Hurricanes are defined as tropical storms with surface winds that reach a constant speed of 74 miles per hour
or more (7). In some parts of the world they are known as typhoons or tropical cyclones. They occur when
low pressure disturbances form in the atmosphere between the 5/ and 30/ latitudes in both the northern and
southern hemispheres (8). These storms spiral around a calm center known as the eye of the storm, and
effects of a hurricane may extend as far as 400 miles away from the eye of the storm. One hurricane can
potentially last for more than two weeks over open water and could run a path across the entire eastern
seaboard (9).

Hurricane Effects

As a hurricane approaches land, the combination of low atmospheric pressure and wind draws ocean water
inland and causes waves and tides to raise the water level in a phenomenon known as storm surge (8). The
storm surge may rise up to 20 feet, range from 50 to 100 miles wide, and last for several hours (7). This
causes many fatalities associated with hurricanes and has the potential to devastate coastal communities.

In addition to storm surge, hurricanes may produce additional weather effects. A fully developed hurricane
may produce heavy rainfall of approximately six to ten inches as it passes, with a possibility of amounting
to as much as 40 inches (8). Excessive rain and flooding of inland streams and rivers may cause landslides
(7). Also, hurricanes may spawn tornadoes at landfall because of the existence of ambient conditions for
tornado formation (10).

Hurricane Intensity Classification

The Saffir-Simpson Scale classifies hurricane intensity from Category 1 to 5 based on rotational wind speeds
and expected damage (11). Most people term Categories 3, 4, and 5 as “major” hurricanes. Table 2 outlines
the Saffir-Simpson Scale with regard to maximum wind speed, storm surge, damage, and examples of North
American hurricanes (10). Emergency management officials often use the category of a hurricane as a
criteria for determining whether or not to issue an evacuation.
Renee H ulett Page 239

Table 2. Saffir-Simpson Hurricane Intensity Scale (10).

Category Maximum Storm Damage North American

Wind Speed Surge (ft) Level Examples
(mph)
1 74 - 96 3-5 Minimal Hurricane Joey (1989)
2 97 - 111 6-8 Moderate Hurricane Bob (1991)
3 112 -131 9 - 12 Extensive Hurricane Gloria (1985)
4 132 - 155 13 - 18 Extreme Hurricane Andrew (1992)
5 156+ 19+ Catastrophic Hurricane Camille (1969)

Hurricane Statistics

Each year, approximately ten tropical cyclones develop in the North Atlantic (7). Of these ten cyclones, six
are likely to strengthen to hurricane proportion, of which two will probably strike the United States’ coast.
Approximately two major hurricanes (Category 3, 4, or 5) hit the United States every three years.

Advances in technology and severe weather detection have decreased the number of hurricane fatalities in
the United States. In 1900, the Galveston, Texas hurricane took over 6,000 lives (maybe as many as 10,000),
which amounts to the most fatalities in United States history for a natural disaster (7, 10). After the first
decade of this century, hurricane fatalities dropped off to approximately 1,000 to 2,000 per decade up
through the 1960’s (12). The 1970’s saw about 250 hurricane-related deaths and approximately 100 occurred
during the 1980’s. Up through 1995, the 1990’s have experienced less than 100 deaths. The population at
greatest risk during a hurricane, also termed the vulnerable population, includes those who live in areas
subject to storm surge and those who reside in manufactured homes (11).

Hurricane Tracking

The National Weather Service is the lead US agency for tracking severe weather (13). Meteorologists have
made advances in detecting tropical cyclone formation from distant waters with the help of earth-orbiting
and geostationary satellites (8). The National Weather Service issues a hurricane watch when the threat of
hurricane conditions in a specific area is 24 to 36 hours away (7). They increase the watch to a hurricane
warning when hurricane conditions are expected in less than 24 hours. In a 24-hour forecast, cyclone tracks
can be predicted within an error of 10 degrees, which relates to the hurricane striking within 115 miles of
its estimated landfall point (8). Once hurricane watches and warnings have been issued for an area, the
appropriate officials must decide whether or not a precautionary or mandatory evacuation are in order.

EVACUATION PRACTICES

Current evacuation procedures are assessed to identify good practices, problems, and voids. Additional
procedures are identified in the next section on agency experience. Reliable and effective methods are
included later in the evacuation guidelines.
Renee H ulett Page 240

Training

Training of personnel and volunteers plays an important part of preparation for hurricane evacuations. Some
common forms of training include workshops, short courses, conferences, video training tapes, informal staff
meetings, and mock disaster exercises (3). These types of exercises, especially mock disaster exercises, help
improve coordination within and among agencies and helps pinpoint strengths and weaknesses in existing
evacuation plans (14). Training also promotes safe practices and helps minimize liability (3).

Evacuation Route Development

Evacuation routes must be developed in advance and brought to the public’s attention in order for
evacuations to run smoothly. Feasible evacuation routes should be selected based on their location, capacity,
survivability, ease of restoration, functional service, and strategic characteristics (1). In addition to planning
routes, evacuation zones are established to determine the vulnerabilities of the area’s storm surge, evacuation
routes, and shelter locations (12). The natural, demographic, and political features of an area help determine
the shape of the zones and the identification of zones is made readily available to the public. This section
includes an analysis of the use of demand, capacity, and human behavior to help determine evacuation routes
and clearance time. The methods presented are extremely generalized and only represent one form of
modeling. Numerous types of modeling techniques could be applied to hurricane evacuation systems.

Demand

Demand is the number of vehicles, or people, that wish to travel past a point during a specific time period.
Demand calculations vary, but the following summarizes a basic process for demand estimation based on
demographic (number of residents and tourists in area), vehicle (number of people to evacuate per vehicle),
and response factors (15). Response factors can be estimated two ways: 1) everybody will evacuate, or 2)
the people most likely to leave (determined by behavioral analysis) will evacuate. The percentage of people
likely to evacuate ranges from 30 – 100 percent. Once the demographic, vehicle, and response factors have
been determined, the following procedure estimates demand:

• Estimate the population using US Census data;

• Determine current per capita vehicle rates for study area;
• Determine the maximum number of vehicles by multiplying the population by the per capita rate; and
• Reduce the demand by the appropriate response factor (0.30 to 1.00).

Capacity

The Highway Capacity Manual defines capacity as “the maximum hourly rate at which persons or vehicles
can reasonably be expected to traverse a point or uniform section of a lane or roadway during a given time
period under prevailing roadway, traffic, and control conditions” (16). If capacity exceeds demand, some
of the possible roadway capacity is lost over time (15). The roadway section with the least capacity
represents the entire evacuation route’s capacity. Generalized capacity values may be developed for various
facility types by applying ambient condition and poor roadway condition adjustment factors to ideal capacity
values as set forth in the Highway Capacity Manual (15, 16).

Clearance Time

Clearance time represents the total time elapsed from an evacuation issuance until all vehicles have cleared
the roads in an evacuation area (11). The actual clearance point varies among agencies. Calculation of
clearance time depends highly on human behavior, demand, capacity, time of day, and hurricane category.
A behavioral-based simulation model includes three components of clearance time: 1) initial warning time,
2) individuals’ preparation time, and 3) network clearance time (17). The initial warning time varies for each
Renee H ulett Page 241

individual as they decide whether or not to evacuate and as they search for confirmation of the evacuation
recommendation. Once an individual decides to evacuate, preparation time is needed to secure homes and
property, contact family and friends, pack a few belongings, and other such matters. Network clearance time
can be estimated by the demand (total number of vehicles) divided by the generalized capacity for each link
of evacuation routes. The overall clearance time helps determine how far in advance to issue an evacuation.

Coordination Agreements

A single agency cannot handle a major emergency alone. Cooperation and teamwork is needed from all
affected agencies (14). Agreements among agencies can be either informal or formal. Important items for
consideration prior to an emergency that should be part of an agreement include (1):

• Chain of command within an agency and among agencies;

• Address and phone list of key personnel in each agency;
• Agency lists of available manpower and equipment capabilities;
• Method and sequence of alerting each agency of an emergency; and
• Mutual-aid agreements.

Equipment and Personnel Resource Assessment

During an emergency such as a hurricane evacuation, responses to the situation must be made quickly.
Precious time can be saved if each agency involved draws up equipment and personnel lists during their
preparation efforts (1). This also helps the response efforts run more efficiently.

Equipment Resources

The Guide for Emergency Highway Transportation Regulation suggests the development of a jurisdictional
map and a listing of storage locations, including the materials and equipment stored at each facility (1).
Developing equipment resource lists should be an easy task because many agencies keep track of their
equipment and supplies as a part of their normal daily activities (14). In addition, officials involved in
Hurricane Hugo in 1989 suggest real-time assessment of resources throughout the emergency (1). Computer
technology also aids resource management, but hard copies of the equipment lists should be kept on hand
in case of electrical power loss or computer networking failure.

Personnel Resources

Personnel management is also essential for the smooth flow of evacuation operations. Agency actions to
manage personnel during an emergency include (3):

• Providing duplicity in key positions;

• Establishing relief schedules to rotate personnel and allow sleep during emergencies lasting longer than
24 hours; and
• Establishing mechanisms to allow personnel to communicate with their families.

Deciding When to Evacuate

Deciding whether or not to evacuate a region is a tough decision, usually made on the local level. If an
evacuation is ordered too early and the hurricane makes landfall outside of the error range, the public is less
likely to respond next time. Several agencies and tools help those in charge with their ultimate decision.
Renee H ulett Page 242

National Weather Service

The National Weather Service (NWS) is the only official United States agency to issue warnings during life-
threatening weather situations (13). The NWS stores data in a national database available to government
agencies, the private sector, the public, and the global community. Their web site also provides an interactive
weather information network, which provides real-time information, forecasts, watches, and warnings for
all regions of the United States.

NOAA Weather Radio

National Oceanic and Atmospheric Administration (NOAA) Weather Radio is a nationwide network of radio
stations that broadcasts National Weather Service forecasts, watches, warnings, and hazard information 24
hours a day as a public service (18). NOAA Weather Radios are equipped with an alarm that sounds when
life-threatening situations arise. Other features such as strobe lights, pagers, bed-shakers, personal
computers, and text printers are available for the hearing- and visually-impaired. A new digital technology
called Specific Area Message Encoding (SAME) allows broadcasts to be targeted to specific areas, such as
a county or region. The SAME technology also allows listeners to pre-select the National Weather Service
alerts they want to receive based on their county. To receive NOAA Weather Radio, special receivers must
be purchased ranging in price from about 25 to 100 dollars. Public safety experts would like to one day see
NOAA Weather Radio receivers as standard equipment in every home. Several areas have received funding
and implemented programs to install NOAA Weather Radio in public facilities.

Emergency Alert System

The Emergency Alert System (EAS) is a tool that allows the President and others to warn the public about
emergency situations (19). The EAS interrupts regular television and radio programming to provide
emergency guidance to specific viewing and listening areas. The EAS is implemented by the Federal
Communications Commission in conjunction with the National Weather Service and Federal Emergency
Management Agency. Currently, the EAS reaches most citizens. The digital system used by the EAS allows
signals to travel over non-broadcast frequencies and telephone lines. For a small fee, new specially equipped
consumer products, such as cellular phones, pagers, and other devices, can receive EAS messages.

Emergency Managers Weather Information Network

The Emergency Managers Weather Information Network (EMWIN) provides a service that allows users to
obtain almost real-time information, such as weather forecasts, warnings, and other information, from the
National Weather Service (20). The use of EMWIN is primarily intended for use by emergency managers
and public safety officials. EMWIN can be obtained by direct satellite broadcast and repeat radio broadcast,
both of which require the purchase of special hardware. For those who do not wish to purchase the hardware,
several methods exist for obtaining EMWIN feed directly over the internet.

HURREVAC

HURREVAC (HURRicane EVACuation program) is an US government computer program that is restricted

to use by official government emergency managers (21). The program tracks hurricanes and helps users
reach evacuation decisions for their communities. In 1989, government officials in Georgia, North Carolina,
and South Carolina successfully used HURREVAC to help them decide upon an evacuation order for
Hurricane Hugo (1). HURREVAC is only one of many software tools available.
Renee H ulett Page 243

SLOSH

The National Weather Service developed the Sea, Lake, and Overland Surges from Hurricanes (SLOSH)
model to calculate potential storm surge heights produced by hurricanes (11). In selected Gulf and Atlantic
basins, the SLOSH model provides real-time surge forecasts for approaching hurricanes. In addition to surge
height calculation, the SLOSH model can simulate storm surge routing into sounds, bays, estuaries, and
coastal river basins. In addition to the use of HURREVAC in 1989, Georgia, North Carolina, and South
Carolina government officials also successfully used the SLOSH model to aid them in their Hurricane Hugo
evacuation decision (1).

The Public

Positive public response to evacuation warnings is essential for the successful, safe, and efficient evacuation
of an area. Prior to hurricane season, public awareness campaigns remind the public to prepare for
hurricanes. As a hurricane approaches, the appropriate notification procedures and messages must be used
to convince the public that evacuation is in their best interest.

Public Awareness Campaigns

Several effective methods to promote public awareness include: (3)

• Educational campaigns;
• Public service announcements;
• Brochures illustrating designated evacuation routes; and
• Contact information during emergencies (i.e., internet, phone number).

Public Notification of Evacuation

Apart from hurricane warnings provided by the Emergency Alert System, the public needs supplemental
notification concerning what actions should be taken. Behavioral analysis indicates most people respond to
an evacuation directive with denial or the need for confirmation (22). To promote evacuation, the public
must be convinced they are in danger. The display of past hurricane destruction on television just prior to
making an evacuation notice has been effective (23).

An important aspect of public notification is timely and consistent information. For this reason, a public
information coordinator should be assigned to disseminate transportation information directly to the media
and public (14). Key personnel involved in the evacuation should direct all questions and comments to the
public information coordinator to reduce the potential for the distribution of inaccurate or conflicting
information. The public information coordinator should also maintain close ties with the media.

Communications

Communications abilities, both within and among agencies, is of critical importance during an evacuation.
Within many transportation agencies, good communication systems already exist and can be used during an
emergency. However, systems are needed to communicate among agencies because the telephone system
is often inoperable during large-scale emergencies. The telephone system may be down due to electrical
outages, damage, or an overwhelming number of calls (24). The Federal Communications Commission has
recognized the Public Information Emergency System (PIES) as an exemplary radio communications network
that should be modeled by other areas (1). PIES is currently in use in the Houston/Galveston area of Texas
and utilizes a special radio frequency that links Emergency Operations Centers in the area with one another
and with media outlets. The use of Amateur Radio Emergency Service (ARES) and Radio Amateur Civil
Renee H ulett Page 244

Emergency Services (RACES) have proven useful in past emergencies (14). Both organizations provide
volunteer radio communication services to all levels of government and non-profit organizations.

In the event of a breakdown in the main communication system, back-up communication systems should be
available (25). This is necessary in case the regular communication system loses power or problems arise.
In addition, back-up communication systems can supplement the regular communication system by allowing
more communication than normally possible.

Transportation System Assessment

The most fundamental preparation measure a transportation agency can take is to assess the existing
transportation system and analyze its ability to function during a large-scale hurricane evacuation (14). At
a minimum, the assessment should include:

• Traffic operations analysis of critical signalized and unsignalized intersections;

• Capacity analysis of key evacuation routes;
• Identification of “sensitive” locations within the system (i.e., schools, hospitals, etc.); and
• Assessment of manpower and traffic control device needs for real-time traffic management if needed.

Assessment of the transportation system should be a continuous process to accommodate changes in the
system. Temporary conditions, such as construction or malfunctioning traffic control devices, should also
be kept up-to-date and included in the assessment (3).

Traffic Control Devices

Traffic control devices provide direction during an evacuation. The public recognizes commonly used
devices and follows them well.

Channelizing Devices

Channelizing devices such as cones, tubes, barrels, and barricades are useful traffic control devices during
hurricane evacuations. Some of their potential uses include: (1)

• Roadway or ramp closure;

• Splitting a ramp into two narrow lanes to increase capacity; and
• Supplementation of law enforcement directives.

All channelizing devices should be adequately weighted down because of potential weather hazards (26).

Static Signs

The main static sign used for hurricane evacuations is the evacuation route marker, which is included in the
Manual on Uniform Traffic Control Devices (MUTCD) (27). The circular sign measures 18 inches in
diameter and reads “EVACUATION ROUTE” in white legend on a blue background. The legend also
includes a white reflectorized arrow, which varies in direction. The MUTCD suggests the placement of
evacuation route markers 150 to 300 feet in advance of and at any turn in an approved evacuation route and
as straight-ahead confirmation where needed. A public awareness campaign should include the meaning and
existence of the evacuation routes and markers to prevent confusion (3).

Stop signs, detour signs, arrow signs, and road closed signs may also be erected (24). These signs can be
mounted on barricades as typically required or simply mounted on or leaned against anything, such as an
agency vehicle. Sometimes stop signs successfully replace a signal when power has been lost (1).
Renee H ulett Page 245

Traffic Management

Many traffic management options exist to help expedite the evacuation process. Negative impacts must also
be taken into account when considering many of the following procedures.

Emergency Operations Center

Emergency Operations Centers (EOC’s) are commonly used to control all emergency activities (i.e., traffic
control, resident evacuation, restriction of access to the emergency area, etc.) in a coordinated manner (1).
EOC’s exist at the federal, state, and local level. The local emergency management coordinator or law
enforcement commander typically controls a regional EOC and representatives from other support agencies
often report to the EOC.

Law Enforcement

Law enforcement officers play a critical role in hurricane evacuations, especially with respect to traffic
management. Often, active traffic control is needed to aid the flow of traffic at critical intersections or to
take care of traffic accidents.

Signal Issues

Increasing green time and/or the progression band for a given direction has proven to be an effective method
for increasing flow on signalized roadways (3, 25). This method should be applied to the direction of the
traffic heading out of the evacuation zone in such a manner to ensure the cross-streets are not too heavily
penalized.. This is a practical technique for agencies with computerized control systems with the capability
to change timing plans from a central location.

Agencies may consider the temporary removal of signal heads before a hurricane arrives along with
placement of law enforcement officers at intersections to direct traffic (3). This preventative measure can
prevent signal head destruction during a hurricane and allows an agency to quickly replace the signal heads
after the hurricane to reestablish control of the intersection as evacuees begin returning. During Hurricane
Hugo in 1989, approximately 275 signal heads were completely ripped off spanwire and destroyed and about
500 signal heads were damaged to some extent in South Carolina (26).

Reversible Flow

The conversion of roadways to reversible, or one-way, flow has been considered by several agencies and
mostly negative feedback resulted. The negative aspects of reversible flow include: (3, 14, 25)

• Violation of driver expectancy;

• Safety issues;
• Extensive manpower for implementation;
• Problems in converting the roadway back to two-way flow without creating bottlenecks; and
• Dangerous geometric implications (i.e., adverse superelevation, limited sight distance, etc.).

Some agencies feel reversible flow may have some benefits for certain situations, particularly for short
sections of roadway (3, 25). This might also be effective when evacuation needs to occur in a shorter time
frame than possible. For this technique to work, a special lane would have to be set aside to allow emergency
vehicles to travel in the opposite direction if needed.
Renee H ulett Page 246

High Occupancy Vehicle Lane Usage

During an evacuation, officials may opt to allow high occupancy vehicle (HOV) lanes to be opened to all
traffic regardless of occupancy restrictions (3, 25). Several issues need to be examined when considering
this option. Bottlenecks may form at the terminus of the HOV lane, which may reduce capacity and offset
any potential benefits. Confusion may result because not all motorists may be familiar with HOV facilities.
Public awareness prior to evacuation is needed to ease confusion. Furthermore, dropping occupancy
limitations sets precedents for similar actions in the future and this may not be desired.

An alternative to lifting occupancy requirements on HOV lanes would be to encourage normal HOV lane
usage during evacuations. This might encourage people to travel in groups. Furthermore, “evacuation buses”
could be provided to get people to leave their personal vehicles behind, which reduces capacity. The ability
of such buses to use the HOV lanes to bypass evacuation congestion might provide strong incentives for their
usage.

Shoulders as Temporary Travel Lanes

Some agencies have suggested the use of shoulders as temporary travel lanes during hurricane evacuations
(3). This would only prove effective if downstream bottlenecks can be avoided. Also, shoulders on both
sides of the roadway would be necessary so the outer shoulder could be utilized by emergency vehicles and
disabled vehicles. Measures should be taken to ease motorist confusion because driving on the shoulder
violates driver expectancies. This could be achieved through the use of temporary signing and law
enforcement personnel on site to direct traffic to the shoulder.

Suspension of Tolls

Some agencies have suspended toll collection during past emergencies to increase capacity and reduce
bottlenecks created by the toll collection process (3, 25). Ideal situations for using this policy include:

• When the evacuation of an area needs to take place in a shorter time frame than possible; and
• No suitable alternative evacuation routes exist and citizens would be forced to pay tolls.

To invoke this practice, cooperation agreements must be established between public agencies and the toll
authority prior to implementation. Motorists would need to be notified through the use of special signing,
real-time information displays, or public service announcements. Any types of special vehicle restrictions
(i.e., trailers and wide loads) would need to be clearly disseminated. A negative effect of toll suspension is
that it sets precedents for future actions.

Tow Truck Usage

Minor vehicular incidents, such as stalled vehicles or fender-benders, may create problems on major
evacuation routes, especially if they occur at bottleneck locations( 3, 25). During an evacuation, tow trucks
could be stationed at bottleneck locations to help detect and clear incidents in order to maintain a roadway’s
capacity. Suspension of normal tow truck rotations and special arrangements and location assignments
would need to be established. The well-being of the tow truck operators also needs to be considered. In
addition, the use of tow trucks along evacuation routes is costly because it may be necessary to pay for the
time they provide services on the roadway.

On-Street Parking Restrictions

The restriction of on-street parking along an arterial on an evacuation route could provide an additional travel
lane (3, 25). This would only be useful if the increased capacity could be maintained downstream at the end
of the route. Also, cars already parked along the street would have to be moved somehow, either by notifying
Renee H ulett Page 247

vehicle owners or using tow trucks. The use of tow trucks could lead to liability and cost issues. In addition,
a strong public awareness campaign prior to hurricane season should inform the public of the restrictions.

Minimization of Left-Turning Movements

The restriction of left-turning movements at critical intersections along evacuation routes has the benefits
of increasing green time for through movements at signals and reducing blockages in through lanes when left-
turn bays are not provided (3, 25). Evacuation routes can also be assigned to minimize left-turn conflicts.
This has proven effective in Duval County, Florida, where coastal evacuation zones are set up to feed
motorists to the nearest major east-west arterial south of the zone so that right turns are the predominant
movement to access evacuation routes (26). Public awareness, special signs, or law enforcement direction
of traffic would be needed to uphold this practice.

Vehicle Restrictions

Consideration should be given to the restriction of the movement of some vehicles such as oversize cargoes
and mobile homes. Oversize and overweight cargoes can decrease the capacity and safety along evacuation
routes (1). In the Florida Keys, the movement of mobile homes is restricted after a hurricane warning is
issued because just one disabled mobile home could block the only escape route (26). Mobile homes may
also have a tendency to blow over when hurricane strength winds approach.

AGENCY EXPERIENCE

Emergency management and transportation agencies were contacted for interviews in order to learn what
practices are used in each state and to assess their effectiveness. State agencies in states along the Gulf Coast
and most of the Atlantic Coast were targeted due to their high hurricane frequency. Of ten agencies
contacted, seven responded for interviews. Table 3 contains a list of the people interviewed and Appendix
A lists sample interview questions asked of each agency. This section summarizes the information collected
from interviews and other literature sources.

Table 3. Interview Contacts.

Agency Contact Position

Florida Department of Transportation Jack Brown State Traffic Operations Engineer
Georgia Emergency Management Agency Chuck Gregg Regional Planner and Hurricane
Specialist
Gulf Regional Planning Commission Ken Holland GIS Specialist
Maryland Emergency Management Carol Thiel Hurricane Planner
Agency
Mississippi Emergency Management Leon Schaifer Deputy Director
Agency
Andy Crawford Hurricane Program Manager
North Carolina Emergency Management Tom Hegele Chief
South Carolina Emergency Preparedness Charmel Mendez Hurricane Program Manager
Division
Governor’s Division of Emergency Mike Peacock Hurricane Planner
Management (Texas)
Renee H ulett Page 248

Common State Practices

Common practices are used by many of the state agencies surveyed. All seven of the agencies work closely
with other transportation-related state agencies such as the Department of Transportation, Department of
Public Safety, local emergency management agencies, and state, county, and local law enforcement. They
all use the NOAA Weather Radio and the Emergency Alert System to obtain some of their hurricane
information. All of them also conduct training programs that range from workshops to table-top and mock
exercises. Most of the agencies educate the public about hurricanes through campaigns and Hurricane
Awareness Week. The news media (i.e., television, radio, and newspaper) is mainly used to notify the public
of hurricane evacuations in the form of press conferences, public service announcements, and statements
from state and local officials. Under extreme circumstances, law enforcement personnel drive through key
residential and tourist areas with public announcement systems and sirens. Table 4 includes a summary of
many of the common traffic control and management practices.

Table 4. Common State Evacuation Practices.

Mississippi
Maryland

Carolina
Carolina
Georgia
Florida

Evacuation Practices

North

Texas
South
Evacuation Route Markers U U U U U U U
Static Evacuation Signs U U U U
Changeable Message Signs U U U U U U
Closed Circuit Television Cameras U1 U1
Lane Control Signals U1 U1
Reversible Flow2 U3 U U
Highway Advisory Radio U U U U U U
1
Only used if part of an existing regional system.
2
Mostly used on sho rt sections of roadway, bridges, and roadways to barrier islands.
3
Contingency plan for extreme cases that has yet to be used.

Florida

The Florida Division of Emergency Management (FDEM) controls emergency management in the state of
Florida. Evacuation clearance times range from 3.5 to 51 hours depending on the location and the severity
of the approaching storm. Decisions to evacuate are made with the help of the Emergency Managers
Emergency Information Network (EMWIN) and Florida’s own Florida Warning and Information Network
(FWIN). Some areas provide a local 1-800 number to provide the public with information.

The FDEM developed an Emergency Satellite Communications (ESATCOM) system for state-wide
communications among the State Warning Point (the state emergency communications center), federal and
local government agencies, and other entities (11). The system does not rely on commercial electrical power
or overhead terrestrial lines and provides high-speed data transmissions of facsimile, video, and voice.
Renee H ulett Page 249

Traffic control procedures vary throughout the state because of Florida’s extensive coastline. Static flip-
down signs are used throughout the state to provide information. For instance, a “Wear Seat Belt” sign might
flip down to reveal a sign with shelter information. Changeable message signs are utilized when they are
already in place in the existing transportation system and are sometimes temporarily moved from
construction sites to hurricane evacuation routes. Traffic counters are used on interstates to provide the state
Emergency Operations Center with real-time traffic flow data.

Georgia

In Georgia, the lead agency for emergency management is the Georgia Emergency Management Agency
(GEMA) (28). Recent hurricane evacuations include Hurricane Hugo (1989), Hurricane Bertha (1996), and
Hurricane Fran (1996). Evacuation clearance times range from 9 to 28 hours depending on the size of the
county. The evacuations in 1996 were voluntary and only five to ten percent of the affected areas evacuated.

Teleconferences are used to communicate among the agencies involved in the evacuation process.
Communication is also made with visitor centers along the major evacuation routes so that motorists stopping
at the visitor centers can receive information.

Along major evacuation routes, temporary and manual counters are used to estimate traffic volumes. A
contingency plan exists for one-way operations along a segment of IH-16 as a means of last resort. This has
never been implemented due to logistics and the number of personnel needed for safe operation.

GEMA recently developed a new emergency management plan that focuses on phased evacuations. Under
this plan, evacuations would be ordered in succession of four phases: 1) voluntary evacuation of barrier
islands and storm surge zones, 2) mandatory evacuation of barrier islands and storm surge zones, 3) voluntary
evacuation of coastal regions, and 4) mandatory evacuation of coastal regions. These phases will be
implemented as needed.

Maryland

The Maryland Emergency Management Agency (MEMA), a division of the Maryland Military Department,
coordinates emergency management in Maryland (29). During the past two decades, evacuations have only
been ordered for Hurricane Gloria (1985) and Hurricane Emily (1993). These evacuations have been
voluntary because the state of Maryland does not mandate evacuations.

Evacuation of barrier islands, such as Ocean City, involves conversion of major freeways to one-way
operations (30). This is accomplished using changeable message signs and arrow boards along the routes
to direct motorists. Shoulders are kept open for disabled vehicles and emergency services vehicles. State
Police and barricades and/or barrels are placed at all closed ramps to maintain one-way flow in the
appropriate direction. A telecommunications network has been set up between the state Emergency
Operations Center and local governments, the federal government, weather services, and other facilities (29).
These methods have proven to be effective and reliable. A new evacuation plan includes phased evacuations
similar to the methods mentioned by Georgia.

Currently, the DelMarVa Emergency Task Force is constructing a regional evacuation plan for the eastern
shores of Delaware, Maryland, and Virginia. This alliance aims to eliminate communication problems across
jurisdictional boundaries. A similar regional plan is being developed for several Maryland counties along
the western shore of the Chesapeake Bay.

Mississippi

The Mississippi Emergency Management Agency (MEMA) is the lead agency for emergency management
in the state of Mississippi (31). Other than several precautionary evacuations issued during the 1990’s, only
Renee H ulett Page 250

two other hurricanes have warranted full-scale evacuation in the past two decades: Hurricane Elena (1985)
and Hurricane Georges (1998). Evacuation clearance times range from 12 to 24 hours dependin g on the
location and the time of day the evacuation is issued.

Evacuation routes are developed with the aid of Geographic Information Systems (GIS). These routes are
developed for each evacuation zone and are based on a factored version of the 1990 Census of Population.
Brochures containing evacuation maps are distributed in tourist locations such as hotels and casinos along
the coast because tourists and out-of-town visitors are often unfamiliar with the area and do not fully realize
the implications of an impending hurricane. The casinos in Mississippi are very cooperative and often help
provide funding for the distribution of brochures.

Large static evacuation signs are used to provide information regarding radio station frequencies, shelters,
and other general information. Currently, the Mississippi Department of Transportation is in the process of
installing traveler information kiosks at 12 rest areas to alert motorists of construction zones and emergency
information such as hurricane evacuation routes (32). These kiosks will be free of charge and have been
recognized by the Federal Highway Administration as a good practice. Video conference calls are also
conducted twice daily between MEMA and the regional Federal Emergency Management Agency,
neighboring states and regions, and the National Hurricane Center as a hurricane approaches.

North Carolina

North Carolina Emergency Management (NCEM) is a division of the North Carolina Department of Crime
Control and Public Safety (33). Typically North Carolina experiences at least one hurricane or nor’easter
per year, but in the last two decades has only evacuated for Hurricane Bob (1991), Hurricane Emily (1993),
Hurricane Bertha (1996), Hurricane Fran (1996), and Hurricane Bonnie (1998). Evacuation clearance times
for the barrier islands range from 8 to 30 hours, depending on the island.

Bridges and parts of some roadways that connect barrier islands to the mainland are converted to one-way
operations and many law enforcement officers are used in this process. Eventually some bridges are closed
completely prior to hurricane landfall.

South Carolina

In South Carolina, the state Military Department houses the South Carolina Emergency Preparedness
Division (SCEPD) to handle the state’s emergency management needs (2). In past years, part or all of the
South Carolina coast was evacuated for Hurricane Hugo (1989), Hurricane Bertha (1996), Hurricane Fran
(1996), and Hurricane Bonnie (1998). Evacuation clearance times range from 12 to 30 hours depending on
the location, month, and tourist activity.

Static signs are used to direct motorists into the appropriate lanes for evacuation routes. The South Carolina
Department of Transportation uses counters on the evacuation routes to monitor traffic flow. If the volumes
do not indicate that most people have evacuated, more focus is shifted to notifying the public.

Texas

In Texas, the Governor’s Division of Emergency Management (DEM) is a part of the Texas Department of
Public Safety (9). Hurricane evacuations are not mandatory in Texas, but are strongly encouraged when
public safety is at risk. The last hurricane evacuation was recommended in 1992 for Hurricane Andrew.
Evacuation clearance times range from two hours for a Category 1 hurricane to 30 hours for a Category 5
hurricane. Clearance times are highly dependent on the coastal area, transportation network of that area, and
the category of the hurricane.
Renee H ulett Page 251

Major cities along evacuation routes, such as Houston and San Antonio, use lane control signals and closed
circuit television cameras already in place to monitor traffic flow. Conversion of Interstate Highway 37 to
reversible flow coming out of Corpus Christi has been planned, but has yet to be implemented.

Problems Identified by State Agencies

All of the agency personnel interviewed addressed various problems associated with hurricane evacuations.
The main problem cited by most agencies is convincing the public that it is in their best interest to evacuate.
Most people want confirmation of an evacuation order, whether it is mandatory or precautionary. The public
is also confused by conflicting information, such as media reports that do not reflect evacuation orders, but
instead show clear and sunny conditions along the coast.

Other problems identified deal with communications, special needs, vehicles, accidents, and construction.
Many agencies indicated that although communication systems within each agency work well, improvements
can still be made to communication systems among agencies. Also, more communication is needed among
the personnel of all the participating agencies during evacuations. Special needs citizens (i.e., elderly and
handicapped) are not always taken care of during evacuations. Many families evacuate in all of the vehicles
they own and try to take part of their property using trailers. Accidents are difficult to take care of during
an evacuation because law enforcement personnel are tied up and tow trucks do not operate as they usually
do. Lastly, construction work often occurs heavily in the summer months, which provides the potential to
slow down traffic along major evacuation routes during hurricane season.

Suggested Solutions

In addition to identifying evacuation problems, the interviews also revealed feasible solutions and other
suggestions. Increased public education through various media might help convince people to evacuate. This
should include pre-planning suggestions so that the public is not taken by surprise when evacuations are
issued. Restrictions on boats and trailers would help increase the capacity on roadways. More evacuation
information on static and changeable message signs would help keep the public informe d and ease traffic
flow. Throughout an evacuation procedure, information could be tracked using emergency management
software for later analysis and to share with other emergency management agencies. This would help
agencies learn from past experiences and to better prepare for future evacuations.

ITS TECHNOLOGIES

According to the Highway Capacity Manual, Intelligent Transportation Systems (ITS) are “considered to
include any technology that allows real-time information to be gathered and used by drivers and traffic
control system operators to provide better vehicle navigation, roadway system control, or both” (16).
Although some of the ITS technologies presented in this section will not be available for several years, others
can currently be applied to hurricane evacuations.

Geographic Information Systems

A Geographic Information System (GIS) “is a specialized information management system that can collect,
store, retrieve, analyze, and visualize geographic data within a spatial environment” (34). GIS software
contains a database and usually contains tools for database management, graphic processing, and spatial
analysis. GIS can allow emergency managers quick access to critical information pertaining to a particular
location (35). It can also be used to help manage the evacuation process.

Recently, researchers developed a flexible evacuation management system that ties together GIS software
and a database management system (DBMS) that integrates dynamic behavioral patterns, the transportation
network, and regional land-use information (34). DBMS computer applications help store, retrieve, and
Renee H ulett Page 252

analyze data. Together, DBMS sets up the data management environment, and GIS provides spatial analysis
and graphic processing.

Based on local characteristics, such an evacuation management system could be applied to a region based
on the time of day an evacuation is issued so that appropriate traffic control measures can be made to reflect
the location of the evacuating public.

Advanced Traveler Information Systems

Advanced Traveler Information Systems (ATIS) provide motorists with information prior to a trip. Such a
system could be applied to hurricane evacuations in the form of an automated public information system, the
world wide web, or multimedia kiosks to provide the general public with general and real-time information.

Automated Public Information System

An automated public information system can be very useful in relaying a consistent and trusted message to
the public. A good example of an automated public information system exists in the Dallas district of the
Texas Department of Transportation (3). They have implemented a voice-mail telephone system with the
capability of handling numerous calls simultaneously that has proven to be effective. Information on such
a system must be as complete and up-to-date as possible because the system does not allow callers to ask
questions.

World Wide Web

The Florida Division of Emergency Management’s (FDEM) web site provides an excellent example of ATIS
already at work. The FDEM web site contains numerous links to hurricane, weather, and disaster agency
sites and contains an entire hurricane section (11). On a general information level, the site contains the
Regional Evacuation Procedure, hurricane arrival time calculations, evacuation decision-making criteria,
preparedness information (i.e., checklist, supply kit, home protection, etc.), and statewide evacuation routes.
Real-time information is also provided in the form of current weather and updates about tropical storm
threats. Additional real-time information could also be provided in the form of real-time traffic data and
travel time information as posted by Houston TranStar (36).

Traveler Information Kiosks

Multimedia kiosks provide travelers with an interactive medium to provide them with information prior to
or during a trip (37). Kiosks have the capability to display information in various formats including text,
graphics, images, and live video. Public places such as car rental agencies and highway rest areas provide
ideal locations for hurricane evacuees to obtain information on hurricane updates, evacuation routes, shelter
areas, and construction zones.

Real-Time Traffic Detection

Numerous methods are available for the detection of traffic. Many of these methods include inductance loop,
radar, infrared, microwave, audio, and video detection (38). Accuracy and reliability varies for each
detection method. Inductive loop detectors are most commonly used by transportation agencies, but two
major drawbacks exist in their use during hurricane evacuations (39). One drawback is that a loops
sometimes resets itself when a vehicle is stopped over the loop for an extended time period. This is likely
to happen during an evacuation. The other drawback is that emergency management officials cannot “see”
the roadway conditions with the use of inductive loops. For this reason, some agencies use closed circuit
television cameras.
Renee H ulett Page 253

Closed Circuit Television Cameras

Placement of closed circuit television (CCTV) cameras along critical roadways provides real-time video
footage of traffic. This allows for fast detection of incidents and congestion. During an evacuation, such
real-time visuals could be crucial and possibly life-saving, while also helping to reduce evacuation clearance
times (39). Used with real-time information displays and in-vehicle information systems, information from
CCTV could be used to re-route traffic around troubled sections of roadway.

Real-Time Information Displays

Real-time information displays convey useful information when motorists need to be kept up-to-date on the
hurricane evacuation. This can be achieved by displaying information on changeable message signs and by
directing traffic with lane control signals.

Changeable Message Signs

Changeable message signs (also called dynamic or variable message signs by some agencies) are the main
traffic control device used for displaying real-time information. For instance, information concerning routes,
shelters, upcoming detours, and radio frequencies can be displayed. Messages displayed on changeable
message signs should be comprehensible to motorists and supply timely and useful information (25).

Lane Control Signals

Lane control signals display “X”s and arrows over individual travel lanes or alongside the roadway to show
whether a lane is open or closed (38). Green arrows, yellow “X”s (or diagonal arrows), and red “X”s
indicate if a lane is open, about to close, or closed, respectively. These signals provide motorists with
advance warning of lane closures due to incidents and construction, which in effect increases traffic flow
during evacuations.

In-Vehicle Traveler Information Systems

In-Vehicle Traveler Information Systems allow motorists to obtain general and real-time information during
a trip. With regard to hurricane evacuations, critical information pertaining to weather, evacuation orders,
and evacuation routes may be relayed to the motorist. Highway advisory radio is already in use and new
developments are being made with automatic vehicle identification and in-vehicle computer systems.

Highway Advisory Radio

Highway Advisory Radio (HAR), a portable, low-power system, broadcasts area-specific traffic information
over an AM radio band (38). In the past, HAR has proven to be an effective means of disseminating
information on subjects such as incidents, construction, and airport parking. This can also be applied to
hurricane evacuations. Often, static signs, portable signs, and changeable message signs display the radio
frequency for motorists to use.

Automatic Vehicle Identification

Automatic Vehicle Identification (AVI) technology is advancing with regard to in-vehicle traveler
information (36). Developments are currently being made in the area of two-way communication between
the roadside and AVI equipment in a vehicle. Ways are currently being explored to broadcast information
to vehicles with AVI equipment using commercial radio or highway advisory radio. This type of information
would be similar to the information broadcast on changeable message signs and would be disseminated on
a location-specific basis. This too, however, has not fully been developed yet. Once completed, real-time
weather and traffic information pertaining to hurricane evacuations could be transmitted to motorists with
Renee H ulett Page 254

AVI tags when they pass through specific areas. Agencies are currently trying to get AVI tags placed in more
vehicles.

In-Vehicle Computer Systems

In-vehicle computer-based applications should increase dramatically over the next five to ten years with
respect to number and variety (40). Preliminary models have been utilized by various automobile
manufacturers and rental agencies, but have yet to be perfected and completely implemented by the private
sector. On a simplistic level, route guidance systems are intended to provide drivers with basic instructions
on how to best reach their destinations (41). On a larger scale, the goal is to provide in-vehicle systems that
provide Driver Assistance (DA) systems and Mobile Services and Information (MSI) systems. DA systems
include such features as adaptive cruise control, and obstacle warning systems. MSI systems include features
such as navigation, roadside assistance, electronic fee payment, real-time traffic information, weather reports,
driver-specific comfort and convenience adjustments, and office-in-the-car services. The MSI system could
feasibly be extended to provide evacuation routes and real-time weather and traffic updates during a
hurricane evacuation.

Although companies are beginning to market in-vehicle computing systems, much work still lies ahead. The
DA and MSI systems must be incorporated somehow, standard guidelines must be established, and systems
must be designed for compatibility with human information processing capabilities. In addition, a reliable
wireless technology has yet to be developed to support the delivery of real-time information. Wireless
communication technologies are currently being experimented with. Until these obstacles and others are
overcome, route guidance systems will not be applicable to hurricane evacuations. Once it is possible, it
could prove very useful in the dissemination of real-time evacuation information.

Laptop Communications System

Recently, the US Department of Justice’s National Institute for Justice granted money to Monroe County in
Florida for the Two-Eyes Intelligence Project which enables field deputies to access a wide area network
using desktop machines or laptop computers (42). To date, only 13 laptops are in use and each cost about
$9,000 to bring into operation. Each laptop utilizes a touch screen or pen-based screen and connections are
made using either an UHF radio band or cellular digital packet data connections (CDPD). The CDPD has
proven much faster than the UHF radio band. Each laptop has access to the Internet, the Miami-Dade Police
Department network, and the Florida Criminal Information Center and National Criminal Information Center
computer networks.

In 1998, the wide area network was put to the test as Hurricane Georges approached the Florida Keys and
proved very successful. More deputies were put on patrol because of the evacuation, which meant too many
users for the traditional law enforcement radio, which causes officers to have to wait turns to use the radio.
The officers with the laptops, however, did not have to wait and helped relieve the backlog of radio users and
lightened the dispatchers’ load. Not only are the computers useful for coordinating officers during the
evacuation and allowing access to dispatch databases, the laptops also proved useful after the hurricane had
passed. Only residents were allowed back into the Keys to prevent looters and con-artists from swarming
the area as they did in 1992 after Hurricane Andrew. The laptops were used to check the identification and
automobile information of everybody wanting to return to the Keys.

HURRICANE EVACUATION GUIDELINES

The author developed the following guidelines to enhance and improve existing evacuation procedures
currently in use. Not all guidelines can be applied by every agency. Many of them highly depend on
location, the existing transportation network, and state policy. Funding and cost-benefit analysis also play
a role in the process. The guidelines can be broken down into the following categories: Planning and
Renee H ulett Page 255

Preparation, Making the Evacuation Decision, Public Notification, Traffic Management During the
Evacuation, and Preparation for the Future.

Planning and Preparation

1. Develop Coordination Agreements to establish a chain of command, lists of resources, and mutual-aid
agreements.
2. Use Equipment and Personnel Resource Lists to keep tabs on available equipment and personnel and
update the lists in real-time.
3. Conduct Training to adequately prepare personnel and volunteers. Conduct tabletop exercises as a part
of the training to improve coordination and to pinpoint strengths and weaknesses.
4. Plan Evacuation Routes based on the demand, capacity, and clearance time for a particular region.
5. Consider Phased Evacuation Plans using both voluntary and mandatory evacuation orders starting with
the most critical areas and moving inland.
6. Conduct Transportation System Assessments periodically and include temporary system conditions, such
as construction activity.
7. Plan for How to Deal with Construction Along Evacuation Routes when construction is scheduled to take
place on major evacuation routes during hurricane season. Where feasible, alternate evacuation routes
may be developed and provided to the public. Detours around the construction area may also be planned
in advance. If the evacuation route is needed despite the delay due to construction, the use of shoulders
as temporary travel lanes might be considered with the aid of law enforcement direction.
8. Plan for Use of Shelters in case evacuation is not possible or is not mandatory.
9. Conduct Public Awareness Campaigns to encourage the public to pre-plan and to bring awareness to out-
of-town visitors.

Making the Evacuation Decision

10. Use Available Tools to Aid the Decision to Evacuate. Examples of such tools include NOAA Weather
Radio, EAS, SLOSH, EMWIN, and HURREVAC.
11. Consider Use of a GIS-DBMS Model to enhance evacuation operations by predicting the spatial and
temporal characteristics of a region depending on the time of day an evacuation is issued.

Public Notification

12. Use a Public Information Coordinator who provides information to the media. Use law enforcement
officers with sirens and public announcement systems as needed.
13. Consider Use of a World Wide Web Site that contains both general and real-time information.
14. Consider Use of an Automated Public Information System to provide a voice-mail telephone system to
relay consistent and trusted messages to the public.
15. Consider Use of Traveler Information Kiosks at key places such as rest areas along evacuation routes.

Traffic Management During the Evacuation

16. Operate an Emergency Operations Center to coordinate all of the evacuation activities.
17. Establish and Maintain Communications among agencies. Use systems such as PIES in Texas and
ESATCOM in Florida which have proven to be reliable. Have a back-up system available.
18. Consider a Laptop Communications System to help law enforcement personnel monitor the outgoing
traffic and to respond quickly to incidents.
19. Use Traffic Counters to assess traffic flow and the volume of vehicles evacuating.
20. Use Closed Circuit Television Cameras to monitor traffic flow and to manage incidents.
21. Use Channelizing Devices as a part of traffic control.
22. Use Static Signs to provide general information. Consider the use of flip-down signs.
23. Use Changeable Message Signs to help convey timely and useful information.
Renee H ulett Page 256

24. Use Lane Control Signals in critical areas to help direct traffic.
25. Use Highway Advisory Radio to provide information pertaining to the weather, evacuation routes,
incidents, and construction.
26. Consider Vehicle Restrictions for oversize cargoes and mobile homes.
27. Adjust Signal Timings to increase the green time and/or progression band for through movements leading
out of an evacuation zone.
28. Minimize Left-Turning Movements along evacuation routes and on roads leading to evacuation routes.
29. Promote HOV Lane Usage to encourage motorists to evacuate in groups.
30. Consider Reversible Flow for short sections of roadways under extreme situations, such as the approach
of a Category 4 or 5 hurricane with limited time to evacuate the public.
31. Consider Suspension of Tolls to encourage people to use toll roads, which in turn reduces bottlenecks
at toll collection booths.
32. Consider Use of a Bus System to provide transportation for special needs members of the community
(i.e., elderly, handicapped, etc.) and take them to shelters that provide special needs.
33. Consider Tow Truck Usage at key bottleneck locations along evacuation routes to help detect and clear
minor accidents and help maintain traffic flow and capacity.

Preparation for the Future

34. Develop a Database for In-Vehicle Computer Systems that contains existing road networks and
evacuation routes at a minimum. Include directions to shelter areas and alternate routes when
construction takes place. Prepare to feed real-time information into the database when technicians
develop a reliable communications system for delivering real-time information.
35. Campaign to Install Automatic Vehicle Identification Tags in All Vehicles so that traffic flow may be
monitored during an evacuation and vehicles will be ready once technology has the capability to transmit
messages through the AVI equipment.

CASE STUDY: GALVESTON, TEXAS

To illustrate an application of the evacuation guidelines presented in the previous section, the evacuation
plans of Galveston, Texas were analyzed and the results are presented in this section. Galveston is located
on a barrier island off of the coast of Texas, and Interstate Highway 45 (IH-45) is the only roadway that
provides access to the island. The island is populated by approximately 60,000 inhabitants. An interview
with William Zagorski, the Emergency Management Coordinator for the City of Galveston’s Emergency
Management Division (EMD), produced information about the practices in Galveston. The last evacuation
conducted in Galveston was in 1989 for Hurricane Gilbert. Again, the state of Texas only issues voluntary
evacuations. The current system in place has proved to be highly effective in the past. Although effective,
no system is perfect and improvement is always possible.

Evaluation of Current Practices

Table 5 contains a summary of which evacuation practices recommended in the guidelines are used in the
city of Galveston. Valid reasons exist for some of the practices not in use and improvements can be made
to some of the practices already in use.

Validation of Practices Not in Use

Some of the practices not in use are not feasible for Galveston. The City of Galveston controls all activities
on the island, and therefore, does not need formal agreements with other agencies. However, they do
coordinate with the Galveston County Office of Emergency Management and Houston TranStar on an
informal basis. Training is not currently a part of Galveston’s EMD practices because personnel are
continuously in action responding to various emergencies and activities such as tropical storms, flooding,
Renee H ulett Page 257

and island celebrations (i.e., Mardi Gras and Beach Party). Due to the small size of Galveston, evacuation
route markers adequately mark evacuation routes and additional static signs are not needed. Lane control
signals are not recommended because of the wind loading experienced during severe weather. Too few
evacuations have been conducted to assess the feasibility of minimizing left-turning conflicts on evacuation
routes. Although reversible flow has been considered on IH-45, it has been deemed impossible to implement
because it requires more law enforcement officers than can be provided. No HOV lanes or toll roads exist,
which eliminates the need to address them. Tow trucks have been used to clear disabled vehicles from
bottleneck locations during special events but have never been used for hurricanes because of the danger
placed upon the tow truck drivers.

Good Practices

Several exemplary practices are already in place in Galveston. These practices include the following:

• Extensive public awareness campaigns are conducted. Lectures are conducted at schools, the University
of Texas Medical Branch (the island’s largest employer), other large employers, and at other functions.
Hurricane information is also presented at hotels to help keep tourists informed.
• Public notification is far-reaching. The Emergency Management Coordinator has access to the only
cable provider on the island. He has the power to override all programming to alert the public of a
recommended evacuation. Additionally, police cruise through assigned districts with public
announcement systems to alert residents of an impending hurricane. Police are very familiar with their
districts and even go door-to-door to alert the elderly and hard-of-hearing.
• As mentioned earlier, the Galveston area utilizes the Public Information Emergency System (PIES),
which is a radio network that has proven itself reliable and effective.
• Although a few old mobile homes are left in Galveston, mobile homes are no longer allowed on the
island because of their vulnerability to severe weather.

Recommendations for Galveston

The author recommends the following to enhance the existing hurricane evacuation procedures in
Galveston:

• Plan for construction along IH-45;

• Consider the use of a GIS-DBMS model to aid traffic management;
• Update the City of Galveston Emergency Management Division web site by providing links to the
Galveston County Evacuation Map, Houston TranStar’s real-time traffic information, and the National
Weather Service’s Interactive Weather Information Network’s real-time local weather (43,44,45);
• Consider the use of traveler information kiosks at rest areas along IH-45;
• Consider obtaining a laptop communications system for local law enforcement personnel;
• Consider the development of a bus system to service the special needs population.
• Develop a database with stationary evacuation route maps for in-vehicle navigation computer systems
and prepare the database for use in real-time for the future.
• Convince the residents of Galveston to install automatic vehicle identification tags in their vehicles
Renee H ulett Page 258

Table 5. Evacuation Practices Used in Galveston.

Evacuation Practice Yes No

Coordination Agreements W
Equipment/Personnel Resource Lists U
Training and Tabletop Exercises W
Evacuation Route Plans U
Phased Evacuation Plans U
Periodic Transportation System Assessments U
Plan for Shelter Use U
Public Awareness Campaigns U
Evacuation Decision Tools U
Public Information Coordinator U
World Wide Web Site U
Automated Public Information System U
Traveler Information Kiosks W
Emergency Operations Center U
Intraagency Communication System U
Traffic Counters U
Closed Circuit Television Cameras U
Channelizing Devices U
Static Signs (other than Evacuation Route Markers) W
Changeable Message Signs U
Lane Control Signals W
Highway Advisory Radio U
Vehicle Restrictions (No Mobile Homes) U
Signal Timing Adjustments U
Minimal Left-Turning Movements W
HOV Lane Usage W
Reversible Flow W
Suspension of Tolls W
Tow Truck Usage W
Renee H ulett Page 259

CONCLUSIONS AND RECOMMENDATIONS

Due to the dangerous nature of hurricanes, emergency management agencies must continue to assess and
improve hurricane evacuation plans. Successful plans help save lives by evacuating hurricane-prone areas
in a timely and efficient manner. Currently, a multitude of hurricane evacuation procedures are in use or are
being planned for use. All of the suggestions summarized in the Hurricane Evacuation Guidelines section
should be considered for use by emergency management agencies. The feasibility of the guidelines presented
will vary greatly by location, transportation network setup, cost, and hurricane frequency.

An analysis of Galveston, Texas illustrates that a solid evacuation plan is in place and that the City of
Galveston has most areas of concern under control. Several minor suggestions were made toward the
improvement of the Galveston evacuation process. All agency plans can use fine-tuning on a periodic basis.

Future Research Areas

As ITS technologies develop, emergency management agencies should include their possible use in the
planning process. Geographic Information Systems and In-Vehicle Traveler Information Systems show
promise for emergency management usage. Although these systems still have many obstacles to overcome,
it is only a matter of time before all of the bugs are worked out. Emergency management agencies should
be developing plans for integrating these ITS technologies into their emergency management plans, so that
when the technologies do become fully developed, they will be ready for implementation.

ACKNOWLEDGMENTS

This report was prepared for a course entitled Civil Engineering 677: Advanced Surface Transportation
Systems at Texas A&M University. The author expresses her sincere thanks to Doug Robertson for his
guidance and time contribution throughout the duration of this course.

Thanks is extended to Dr. Conrad Dudek for organizing this course and allowing the opportunity to interact
with highly distinguished professionals. The author would also like to thank Thomas Hicks, Pat Irwin,
Wayne Kittleson, William Spreitzer, and Carol Zimmerman for their insight, support, and encouragement
in this program.

The author would like to thank the following professionals for their time and involvement in this report.
Their input and patience is much appreciated:

• Gerald Ullman, Texas Transportation Institute

• Nada Trout, Texas Transportation Institute
• Jack Brown, Florida Department of Transportation
• Chuck Gregg, Georgia Emergency Management Agency
• Carol Thiel, Maryland Emergency Management Agency of the Maryland Military Department
• Leon Schaifer, Mississippi Emergency Management Agency
• Andy Crawford, Mississippi Emergency Management Agency
• Ken Holland, Gulf Regional Planning Commission
• Tom Hegele, North Carolina Emergency Management of the North Carolina Department of Crime
Control and Public Safety
• Charmel Mendez, South Carolina Emergency Preparedness Division of the South Carolina Military
Department
• Mike Peacock, Governor’s Division of Emergency Management of the Texas Department of Public
Safety
• William Zagorski, City of Galveston Emergency Management Division
• Doug Jonas, ITS Weather Information Applications Task Force of ITS America
• Ed Boselly, ITS Weather Information Applications Task Force of ITS America
Renee H ulett Page 260

REFERENCES

1. Ullman, G.L., N.D. Trout and T. Urbanik II. Synthesis of Traffic Management for Major Emergencies.
Research Report 1231-1. Texas Transportation Institute, Texas A&M University, College Station, 1991.

2. South Carolina Emergency Preparedness Division. Jan. 18, 1999. http://www.state.sc.us/epd. Accessed
June 1999.

3. Ullman, G.L., N.D. Trout and R. Collins. Transportation Management for Major Emergencies. Texas
Transportation Institute, Texas A&M University, College Station, 1993.

4. Signor, S.C. Benefits of Advance Planning to Meet Transportation Emergencies. In Transportation

Research Circular 280, TRB, National Research Council, Washington, D.C., June 1984.

5. About FEMA. http://www.fema.gov/about. Accessed June 1999.

6. Office of Emergency Transportation. April 1999. http://www.rspa.dot.gov/oet/maintel.html.

Accessed June 1999.

7. Hurricane Fact Sheet. Federal Emergency Management Agency, Washington, D.C., Sept. 1993.

8. Maybury, R.H. Violent Forces of Nature. Lomond Publications, Inc., Mt. Airy, 1986.

9. Governor’s Division of Emergency Management. May 3,1999. http://www.txdps.state.tx.us/dem.

Accessed June 1999.

10. National Hurricane Center. http://www.nhc.noaa.gov. Accessed June 1999.

11. Florida Division of Emergency Management. http://www.dca.state.fl.us/fdem/index.htm. Accessed June

1999.

12. Pielke, Jr., R.A. and R.A. Pielke, Sr. Hurricanes Their Nature and Impacts on Society. John Wiley and
Sons, Inc., West Sussex, 1997.

13. The National Weather Service. June 7, 1999. http://www.nws.noaa.gov. Accessed June 1999.

14. Ullman, G.L. and N.D. Trout. Planning Guidelines for Major Transportation Emergencies. Research
Report 1231-3F. Texas Transportation Institute, Texas A&M University, College Station, 1991.

15. Urbanik II, T. Texas Hurricane Evacuation Study. Texas Transportation Institute, Texas A&M
University, College Station, 1978.

16. Highway Capacity Manual, Special Report 209. TRB, National Research Council, Washington, D.C.,
1997.

17. Stern, E. and Z.Sinuany-Stern. A Behavioral-Based Simulation Model for Urban Evacuation. Papers
of the Regional Science Association, Vol. 66, 1989, pp. 87-103.

18. NOAA National Weather Radio. http://tgsv5.nws.noaa.gov/nwr. Accessed July 1999.

19. Emergency Alert System. http://www.fcc.gov/cib/eas. Accessed June 1999.

Renee H ulett Page 261

20. Emergency Managers Weather Information Network. Jan. 14, 1999.

http://iwin.nws.noaa.gov/emwin/index.htm. Accessed June 1999.

21. HURREVAC Support Site. June 22, 1999. http://www.hurrevac.com. Accessed June 1999.

22. Drabek, T.E. Disaster Evacuation and the Tourist Industry. Institute of Behavioral Science, University
of Colorado, Boulder, 1994.

23. Ruch, C.E. and L.B. Christensen. Hurricane Message Enhancement. Research Report TAMU-SG-80-
202, Sea Grant College Program, Texas A&M University, College Station, 1981.

24. Kruger, G.E. Emergency Traffic Operations during the 1989 Earthquake. Compendium of Papers,
Institute of Transportation Engineers, 1990, pp. 346-350.

25. Wohlschlaeger, S.D. and G.L. Ullman. Catalog of Transportation Management Activities for Major
Emergencies. Research Report 1231-2. Texas Transportation Institute, Texas A&M University, College
Station, 1991.

26. Lewis, D.C. Transportation Planning for Hurricane Evacuations. ITE Journal, Vol. 55, No. 8, Aug.
1985, pp. 31-35.

27. Manual on Uniform Traffic Control Devices. FHWA, United States Department of Transportation,
Washington, D.C., 1988.

28. Georgia Emergency Management Agency. 1998. http://www2.state.ga.us/gema. Accessed June 1999.

29. Maryland Emergency Management Agency. June 18, 1999. http://www.mema.state.md.us. Accessed
June 1999.

30. Hurricane Evacuation Information. Maryland Department of Transportation, State Highway

Administration, Aug. 1993.

31. Mississippi Emergency Management Agency. http://www.memaorg.com/index2.htm. Accessed

June 1999.

32. FHWA Quality Journey – Best Practices. http://www.fhwa.dot.gov/quality/HP-MS7.htm. Accessed

June 1999.

33. North Carolina Emergency Management. June 7, 1999. http://www.ncem.org/main4.htm. Accessed

June 1999.

34. Alam, S.B. and K.G. Goulias. Dynamic Emergency Evacuation Management System Using GIS and
Spatio-Temporal Models of Behavior. In Transportation Research Board 78th Annual Meeting (CD-
ROM), TRB, National Research Council, Washington, D.C., 1999.

35. ESRI – Public Safety – GIS for Disaster Management. June 17, 1998.
http://www.esri.com/industries/public_safety/disaster.html. Accessed June 1999.

36. Goolsby, M.E. Houston ITS Priority Corridor Program Plan – 1996 Update. Research Report 2931-3,
Texas Transportation Institute, Texas A&M University, College Station, 1997.
Renee H ulett Page 262

37. Jackson, D.D. An Evaluation of Public-Private Partnership Strategies within the Traveler Information
Services Market. Compendium: Graduate Student Papers on Advanced Surface Transportation Systems,
1998. Southwest Region University Transportation Center, Report SWUTC/98/72840-00003-3, Texas
Transportation Institute, Texas A&M University, College Station, 1998.

38. Blume, K.L. Implementation of a Dynamic HOV Lane. Compendium: Graduate Student Papers on
Advanced Surface Transportation Systems, 1998. Southwest Region University Transportation Center,
Report SWUTC/98/72840-00003-3, Texas Transportation Institute, Texas A&M University, College
Station, 1998.

39. Zaragoza, D.P., M.W. Burris and E.A. Mierzejewski. Hurricane Evacuation Traffic Analysis and
Operational Measures. Center for Urban Transportation Research, University of South Florida, Tampa,
1998.

40. Research Directions for In-Vehicle Computing. In Transportation Research Circular E-C009, TRB,
National Research Council, Washington, D.C., 1991.

41. Stockton, W.R. and S.P. Venglar. Texas Department of Transportation Intelligent Transportation
Systems Deployment Strategy. Texas Department of Transportation, Austin, 1996.

42. Dussault, R. Looking Through the Eye of a Storm.

http://govt-tech.govtech.net/gtmag/1999/mar/jandt/jandt.shtm. Accessed June 1999.

43. City of Galveston Emergency Management Division. 1999.

http://www.cityofgalveston.org/cityorg/emergency. Accessed June 1999.

44. Galveston County Emergency Management. Dec. 20, 1998. http://www.phoenix.net/~gceoc. Accessed
June 1999.

45. Houston TranStar. http://traffic.tamu.edu. Accessed June 1999.

APPENDIX A: SAMPLE INTERVIEW QUESTIONS

• How frequently are hurricane evacuations conducted?

• What other transportation agencies are involved in the evacuation process?
• Do you conduct a training program to prepare for an evacuation?
• How is the public alerted that they must evacuate?
• How long does it typically take to evacuate the area?
• What types of devices/media/techniques are used in the evacuation process? (ie. static signs, variable
message signs, radio, television, internet, wireless communication, lane control signals, closed circuit
television cameras, in-vehicle navigation systems, one-way roadway operations, etc…)
• How reliable and effective are these devices and procedures?
• Are there any problems with the current process?
• If no constraints were present, how would you describe your ideal evacuation system?
Renee H ulett Page 263

RENEE HULETT

Renee Hulett earned her Bachelor of Science degree in Civil

Engineering at Texas A&M University in December 1998. As
an undergraduate junior and senior, the Texas Transportation
Institute employed her as a student worker in the Transport
Operations Program. Renee spent two summers working for
Traffic Engineers, Inc. in Houston, Texas where she worked in
a variety of areas including traffic impact studies, traffic
control plans, signing and striping, drafting, traffic counts, and
field studies. She currently works at the Texas Transportation
Institute as a graduate research assistant in the Traffic Analysis
Program and is pursuing a Masters of Science degree in Civil
Engineering. Renee is active in the Texas A&M Student
Chapter of the Institute of Transportation Engineers and
currently holds the membership secretary position.
 SUITABILITY OF OPTICAL, RADIO AND RADAR TECHNOLOGIES
AS POTENTIAL EMERGENCY VEHICLE SIREN SYSTEM ENHANCEMENTS

by

Jacqueline Jenkins, P.Eng.

Professional Mentors
Carol A. Zimmerman, Ph.D.
Battelle

and

H. Douglas Robertson, Ph.D., P.E.

Science Applications International Corporation

Course Instructor
Conrad L. Dudek, Ph.D., P.E.

Prepared for
CVEN 677
Advanced Surface Transportation Systems

Department of Civil Engineering

Texas A&M University
College Station, TX

August 1999
Jacqueline Jenkins Page 265

SUMMARY

Emergency vehicle warning systems are used by police, fire and emergency medical services personnel to
alert drivers sharing the roadway to yield the right-of-way to the approaching emergency vehicle. The
warning to the drivers is given as a combination of visual and audible signals. The effectiveness of the siren
system or audible signal is debatable. Some people believe the siren results in faster response times while
others believe the siren is not significantly beneficial.

This paper presents an evaluation of the emergency siren system as an audible warning for passenger car
drivers and investigates the suitability of optical, radio and radar technologies to enhance the current system.
Basic knowledge of acoustics is required to evaluate the siren system. For a sound to be heard, the sound
must be produced by a source, travel from the sound source to a receiver and the receiver must have the
capability to perceive the sound. If any of the three components are missing, there is no sound. The details
of the sound source, sound propagation, and the receiver are presented and discussed with respect to the
emergency siren system.

Five criteria, identified as properties of a good audible warning system, are used to evaluate the emergency
siren system. These criteria require that the siren be heard by the driver without being startling or disruptive.
The siren must have meaning and must not harm the operator. The emergency siren system satisfies only
two of the five evaluation criteria, which suggests that the audible signal used by the emergency siren system
is inadequate.

When the passenger car is idling, the driver will likely hear the siren from up to 70 meters away. When the
passenger car is traveling at a cruising speed of 70 mph (112 km/h), the driver may be able to hear the siren
from up to 14 meters away. If the driver hears the siren when the emergency vehicle is very close, the siren
will be loud, have a rapid onset and will startle the driver. The driver’s attention may be shifted from driving
to searching for the location of the emergency vehicle. This attention shift can be disruptive, and the driver
may lose control of the vehicle.

The operators of the emergency siren are protected by the sound exposure limits published by the
Occupational Safety and Health Association. If the guidance provided by the state driver manuals are
indicative of the drivers’ understanding of the emergency siren, each driver should have a general
understanding of the required response to the approach of an emergency vehicle using emergency lights and
siren.

A remedy for the deficiency of the siren system may include an enhancement of the sound source, or the
sound receiver. Perhaps a different means of communication could be used. These alternatives are
investigated.

Enhancing the sound source by increasing the sound output of the siren is not feasible. Alternatively,
enhancing the sound receiver is feasible. There are several sound-based detection devices currently available
which detect the siren and in turn notify the driver to the approach of the emergency vehicle.

Emergency vehicles have used sound, optical, radio and radar transmission for signal preemption or for the
surveillance of speeding vehicles. The suitability of using an optical, radio or radar-based system is
investigated. Considering the current system uses sound transmissions, the suitability of the other systems
is based on a comparison with a sound-based system.

Optical, radio and radar-based systems have operating ranges greater than sound-based systems. The optical-
based system provides no other benefit and is not able to warn pedestrians and other non-vehicle roadway
users to the approach of an emergency vehicle. Both radio and radar-based systems provide additional
benefits. Radio-based systems do not require the source to be in direct line of sight with the receiver and
Jacqueline Jenkins Page 266

radar-based systems have precise signal direction ability. Neither radio or radar transmissions are sensitive
to weather. Therefore, the use of sound, radio, or radar transmissions is suitable when considering an
enhancement to the current emergency siren system.

The author recommends the continued use of the siren system and the application of radio or radar
transmissions as an enhancement. Study into the effectiveness of the siren system , with respect to response
time is recommended for future study. As always, efforts to educate the operators of the emergency warning
system and the drivers and the other roadway users should continue.
Jacqueline Jenkins Page 267

TABLE OF CONTENTS

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268

THE EMERGENCY SIREN SYSTEM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268

The source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
Sound transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
The receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270

EVALUATION OF THE SIREN SYSTEM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271

Evaluation criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
The siren must be heard by the driver of the passenger car . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
The sound level of the siren should not endanger the operator . . . . . . . . . . . . . . . . . . . . . . . . 274
The siren should not be overly startling to the driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
The siren should not disrupt the driver from maintaining control of the vehicle . . . . . . . . . . 275
The siren should have meaning for the driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Evaluation findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Evaluation results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276

ENHANCEMENTS TO THE EMERGENCY SIREN SYSTEM . . . . . . . . . . . . . . . . . . . . . . . . . . 276

Enhance the sound source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Enhance the sound receiver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Means of communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Sound transmissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Light and strobe transmissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
One-way radio transmissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Radar transmissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
Additional consideration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
Enhancement summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
Enhancement findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Enhancement results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279

RECOMMENDATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
Jacqueline Jenkins Page 268

INTRODUCTION

Emergency warning systems are used by police, fire and emergency medical services personnel to alert
drivers sharing the roadway to yield the right-of-way to the approaching emergency vehicle. The warning
is comprised of visual and audible signals, commonly referred to as lights and sirens.

The effectiveness of the emergency warning system is debatable. The National Highway Traffic Safety
Administration supports the use of emergency warning systems and encourages drivers to take the sirens
seriously. A brochure is available which describes how drivers can safely yield the right-of-way to an
emergency vehicle(1). In contrast, in the April 1997 edition of the National Safety Council First Aid
Newsletter, the American Academy of Orthopedic Surgeons was quoted as saying "the siren is probably the
most overused piece of equipment on an ambulance. It does not really help other motorists in closed cars.
Motorists driving at the speed limit with the windows rolled up, and the radio on and/or air conditioner or
heater set on high, cannot hear the siren until the ambulance is only a short distance away. If the radio is
particularly loud, the other driver may not hear the siren at all"(2). The newsletter also refers to a study
conducted in North Carolina which found that ambulances using lights and sirens reached the hospital
emergency department 43.5 seconds faster than those not using the emergency warning systems(2). Although
the statements of the American Academy of Orthopedic Surgeons and the findings of the North Carolina
study were limited to the use of lights and sirens on ambulances, they still suggest that emergency warning
systems in general are not significantly effective.

Purpose

The purpose of this paper is to evaluate the emergency vehicle siren system as an audible warning for
passenger car drivers and to determine whether optical, radio or radar transmissions should be used to
enhance the siren system.

Scope

The emergency vehicle siren system is used to warn all types of drivers of all types of vehicles in a plethora
of traffic and environmental conditions. The evaluation presented in this paper is limited to a passenger car
driver, traveling directly in front of the emergency vehicle on a straight, level roadway. The siren is assumed
to be a Class A siren capable of emitting a sound level of 120 dBA at 10 feet. Meteorological conditions and
the effects of traffic, buildings, berms and barriers are neglected.

THE EMERGENCY SIREN SYSTEM

Emergency vehicles, such as police, fire and emergency medical services vehicles are permitted to have a
siren. The purpose of the siren is to alert drivers of the approaching emergency vehicle. Upon hearing the
emergency siren, each driver is expected to maneuver his or her vehicle to the right edge or curb of the
roadway and stop. The driver should remain stopped until the emergency vehicle has passed, and it is safe
to merge into traffic.

The siren system is a means of communication. Three main components of the emergency siren system must
be present for the communication to be effective. The message must be created, relayed and received. In
other words, the siren sound must be produced by the siren, travel through the atmosphere, and be perceived
by the driver of the passenger car.
Jacqueline Jenkins Page 269

The source

The emergency siren is an electronic or electromechanical device for producing standardized acoustical
signals.(3) The sound emitted by a signal propagates through the atmosphere as vibrational sound waves.
The frequency and amplitude of the sound wave can be used to describe the perceived sound. The frequency
is the measurement of the number of wave cycles occurring in a second and corresponds to the pitch of the
sound. The frequency is expressed in cycles per second or hertz (Hz). The amplitude of the sound is a
measurement of the height of the sound wave and corresponds to the sound intensity or loudness. The
amplitude of the sound pressure P1 is usually expressed in terms of the reference sound pressure (Po). P0
relates to the threshold of human hearing. This expression is shown below.

Sound intensity (decibels) = 20 log (P1/Po) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . equation 1

To provide a basis for comparison, the absolute sound intensity for some common sound sources are shown
in Table 1. The sound intensity is relative to the threshold for human hearing. Therefore, a sound intensity
of zero refers to a sound which is barely audible and does not mean there is no sound.

Different decibel scales can be used to measure the intensity of a sound. The A scale differentially weights
sounds to reflect the characteristics of human hearing, providing greatest weighting at those frequencies
where humans are most sensitive(4). A-weighted sound levels are most often used for estimating the effects
of noise on people. For most transient sounds, such as a passenger car passing by, the A-weighted scale is
preferred(5).

The Society of Automotive Engineers publishes the test procedures, requirements and the guidelines for the
use of electronic and electromechanical emergency sirens. SAE J1849 recommends the minimum on-axis
sound level produced by the emergency siren at 3.0 m ± 0.1 m to be 118 dBA(3). The Federal Signal
Corporation manufactures several models of sirens, amplifiers and speakers for emergency vehicle use.
These systems have an on-axis sound output of approximately 120 dBA at 10 ft and comply with SAE J1849
and CCR, Title 13, Article 22 (Class A) requirements(6,7).

Table 1. Absolute Intensity of Some Common Sound Sources. (4)

Sound Level (dBA) Common Sound Source

140 ear damage possible - jet at take-off
130 painful sound
120 propeller plane take-off
110 loud thunder
100 subway train
90 truck or bus
75 average passenger car; loud radio
60 normal conversation
50 quiet restaurant
40 quiet office; household sounds
20 whisper
0 threshold of hearing
Jacqueline Jenkins Page 270

There are three basic operation modes for a siren. The high-low mode produces a pattern of alternating high
and low tones with a cycle rate of 40 to 60 cycles per minute. The wail mode produces a pattern of slow
increases and decreases in frequency with a cycle rate of 15 to 30 cycles per minute. According to SAE J
1849, the frequency range should be greater than 850 Hz and lie within 650 and 2,000 Hz. The yelp mode
is very similar to the wail mode except the cycle rate is 150 to 250 cycles per minute. The recommended
frequency range is the same as the wail mode(3).

The mode of the siren does not affect the relative sound level at a distance from the siren. The frequency
spectrum of the siren, whether in yelp, wail or high-low mode is essentially the same(8). Therefore, at any
location from the siren, the relative sound level is independent of the mode of operation of the siren.

Sound transmission

As the sound wave travels through the atmosphere, from the source to the receiver, the intensity of the sound
wave decreases. For a point source such as a siren, the drop-off rate of the intensity is approximately 6 dBA
per doubling of distance. This relationship is referred to as the mean-square sound pressure and represents
the geometric or spherical spread of the sound emanating from a single point source.

If the sound level (SL1) in a particular direction from the source at a distance D1 is known, the sound level
at any distance (D2) in the same direction can be calculated using the mean-square sound pressure equation:

SL2 = SL1+10log10(D1/D2)2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . equation 2

If the distance of point 2 is farther from the source than point 1, the sound level at point 2 is lower than the
sound level at point 1. This relationship is expected. The sound level you experience is strongest when you
are near the source and weakens as you move away.

This equation does have limitations. When the siren emits a sound, the particles in the atmosphere are
disrupted. Close to the siren the particles may not necessarily travel in the same direction as the sound wave.
If measured with a sound measuring device, the sound levels would be greater than that expected by the
mean-square sound pressure equation. Farther from the siren, the particles tend to move in the same direction
of the sound wave. This area is considered the free space and the sound intensity is represented the
mean-square sound pressure relationship.

Equation 2 does not account for the excess attenuation of the sound caused by geological and meteorological
factors. The terrain, ground cover, foliage, berms and barriers as well as the air turbulence, temperature,
wind gradients and precipitation affect the amplitude or loudness of the sound.

The receiver

The sound is the stimulus needed for human hearing. The healthy human auditory system is capable of
perceiving sound frequencies between 20 Hz and 20,000 Hz, although it is most sensitive to frequencies
between 1,000 and 5,000 Hz(4,9). The auditory system includes the outer, middle and inner ear. The outer
ear captures the sound waves and funnels them into the middle ear. The sound waves are then amplified and
conducted into the inner ear. The cochlea is located in the inner ear and contains the basilar membrane
which translates the physical sound wave into electrical nerve energy. The auditory nerve then transmits this
energy to the brain(4).

Some people experience hearing losses or the complete loss of hearing. The human auditory system can be
damaged from exposures to loud sounds or long exposures to particular sounds. As people get older, they
can experience a gradual decline in hearing capability. If the hearing loss is caused by a breakdown within
the middle ear, hearing aids can be beneficial. The hearing aid can replace the function of the middle ear by
Jacqueline Jenkins Page 271

conducting and amplifying the sound wave. If the hearing is caused by a breakdown within the inner ear,
hearing aids are not effective.

EVALUATION OF THE SIREN SYSTEM

To determine if the emergency vehicle siren system is a good audible warning for passenger car drivers, the
siren must be evaluated. The author uses five criteria which have been identified as properties of a good
warning system for use in a working environment(4). The use of these criteria is justified through a
comparison of a worker in a workplace and a driver in a vehicle.

In the workplace, a worker is required to use tools to complete a task. A warning system alerts the worker
of a particular situation requiring immediate attention. Similarly, a driver is required to use controls (steering
wheel, accelerator, brake) to safely maneuver a vehicle. The emergency siren system alerts the driver of a
particular situation which requires immediate attention. Since the task of driving has the same basic
components of a workplace task, and the function of the warning system is the comparable to the function
of the emergency siren system, the criteria used to evaluate a workplace warning system can be applied to
the emergency siren system.

Evaluation criteria

The five criteria used to evaluate the emergency siren system are as follows:

1. The siren must be heard by the driver of the passenger car;

2. The sound level of the siren should not endanger the operator;
3. The siren should not be overly startling to the driver;
4. The siren should not disrupt the driver from maintaining control of the vehicle; and
5. The siren should have meaning for the driver.

Each of these criteria are discussed in the following sections.

The siren must be heard by the driver of the passenger car

To investigate whether the driver can hear the emergency siren, an analysis of the sound emanating from the
siren is required. The analysis considers the source of the sound, the sound propagation, the sound
transmission loss, and the resulting sound level perceived by the driver. In addition, competing sounds need
to be discussed and compared to the sound level of the siren.

Sound level at the source. The Class A emergency siren emits sound pressures which measure 120 dBA
on-axis at 10 feet, (3.084 m)(6,7) fulfilling the SAE J1849 requirements as published by the Society of
Automotive Engineers(3).

Sound propagation. As the sound wave propagates through the atmosphere, the intensity of the wave is
reduced. The reduction corresponds to the distance the wave has traveled such that a doubling of distance
will result in a 6 dBA reduction in intensity. The upper line in Figure 1, labeled as the exterior siren sound
level, represents the sound level of the siren within a range of 3m to 200 m from the siren.
Jacqueline Jenkins Page 272

140

Sire n S ound Le ve l - Ex te rior

120

S ound Level (dB A)

S ire n Sound Le ve l - Inte rior
100

80 Inte rior Sound Le ve l Ra nge

Ope ra tion M ode - Cruise
60

40 Inte rior Sound Le ve l Ra nge Ope ra tion M ode - Idle

20

0
0 50 100 150 200

D istance (m)

Figure 1. Siren sound levels and interior sound level ranges

The propagation of the sound wave can be affected by the environment. Meteorological factors such as
temperature, humidity, precipitation and wind can cause fluctuations in the frequency spectrum of the sound
wave. These factors are interrelated to some extent. Noise barriers, buildings, and other solid surfaces
reflect the sound waves, and large areas of grass, shrubs or trees can absorb or dampen the sound waves.

Sound transmission loss. The exterior of the passenger car can be considered as a collection of sound
reflective surfaces. When the person is seated in the drivers seat, their ears are in an area protected by the
greenhouse of the vehicle. The greenhouse is comprised mainly of glass, which insulates the passenger
compartment from noise.

The effectiveness of the glass depends on the type and thickness. A typical ¼ inch plane of safety glass can
reduce the sound level by 27 dBA. In comparison, a ¼ inch 2 ply plane of laminated glass can reduce the
sound by 34 dBA.(10) As the thickness increases the sound reduction increases. By introducing layers of
air or a laminating interlayer such as polyvinyl butyral, the effectiveness of the glass can be improved
resulting in the increase of sound transmission loss.

The lower line in Figure 1, labeled as the interior siren sound level, represents the sound transmission loss
associated with the typical ¼ inch safety glass. This type of glass is commonly used for side and rear
windows and laminated glass is used for windshields, although some vehicle manufacturers are also using
laminated glass in the side and rear windows.

Interior sound level. When a vehicle is in operation, the engine, transmission, tires, cooling fan, air inlet and
exhaust create sounds and vibrations. The vibrational waves travel through the air and through the structure
of the vehicle and combine to produce an ambient interior sound level. The sound level is dependent on the
mode of operation. At low speeds, the engine sounds are the primary contributors. At high speeds the tire
noise becomes more prevalent (11).
Jacqueline Jenkins Page 273

Although there is no standard regarding the allowable sound level in the passenger compartment, automotive
manufacturers regard the mitigation of these sounds as an important marketing concept. The quality of sound
can be used to describe the quality of the vehicle and the comfort of the ride. Controlling the interior sound
level also has functional considerations. Lower interior sound levels ease voice communication, decrease
fatigue, reduce risk of hearing damage, and improve the ability to hear warning signals that originate outside
of the vehicle(5).

In the 1999 Ultimate Comparison edition of the Road and Track magazine, interior sound levels were
measured for different models of vehicles (12). A range of 35 dBA to 51 dBA was measured in 43 vehicles
while the vehicles were idling. A range of 65 dBA to 93 dBA was measured in 182 vehicles while the
vehicles were cruising at 70 mph. These interior sound level ranges are represented in Figure 1 and will be
used to compare with the siren sound levels.

There are additional sound sources which can contribute to the interior sound level, such as compact disc
players, tape players, radios, cellular phones, and air conditioners. Since the driver controls the operation
of these systems and can easily turn them down or shut them off, these noise sources are not considered in
the evaluation.

Masking. When two sound sources compete, one sound can hide or mask the other sound such that only one
sound is perceived. For instance, if the radio is playing loudly, the driver may not notice the sound of the
siren. This phenomenon is referred to as masking. To ensure the siren is perceived, it needs to be 15 dBA
louder than the music being played on the radio (4).

Comparing the interior siren sound level to the ambient interior sound level. To account for the masking
phenomenon, the interior sound level ranges presented on Figure 1 have been adjusted by 15 dBA and
reproduced on Figure 2.

140
S ire n Sound Le ve l - Ex te rior
120
S ire n Sound Le ve l - Inte rior
S ound Level (dB A)

100 Adjuste d Inte rior Sound Le ve l Ra nge

Ope ra tion M ode - Cruise
80

60
Adjuste d Inte rior Sound Le ve l Ra nge Ope ra tion M ode -

40

20

0
0 50 100 150 200

D istance (m)

Figure 2. Adjusted interior sound level ranges.

Jacqueline Jenkins Page 274

By interpreting Figure 2, statements can be made regarding when the emergency siren can be heard. Within
the adjusted interior sound level range for the idle mode, the area below the interior siren sound level line
represents the conditions when the siren can be heard by the driver of the passenger car. Therefore, if the
emergency vehicle is within 70 meters of a stopped passenger car, the driver of the car can most likely hear
the siren. If the emergency vehicle is further than 70 meters, the driver of the car may be able to hear the
siren.

Similarly, within the adjusted interior sound level range for the cruise mode, the area below the interior siren
sound level line represents the conditions when the siren can be heard by the passenger car driver. This area
is quite small. If the emergency vehicle is within 14 meters of a car traveling at 70 mph (112 km/h), there
is a possibility that the driver can hear the siren.

Factors affecting response. Some drivers have permanent sensory impairments which affect their ability to
perceive the emergency siren. Others drivers suffer from temporary impairments. Sleepiness, inattention,
emotional state, intoxication, or the influence of drugs can affect a person's ability to perceive and respond
to the emergency siren (13).

The response to the sound is not only dependent on the capabilities of the auditory system but also on the
attitudes, memory and expectation of the human (14). The message may be perceived as unimportant or not
requiring a response. The inappropriate use of the emergency siren may lead the driver to believe the siren
is not important and can be ignored. If the message is not familiar, the driver may not be able to recall what
the appropriate action is to take. The driver may have forgotten what to do when or may lack the knowledge
of the appropriate response. It is also possible that the driver believes the siren has importance and recalls
the appropriate response but performs another action. This action may be influenced by the driver’s
anticipation of the actions of the emergency vehicle.

The sound level of the siren should not endanger the operator

Long or repeated exposure to high sound levels can cause damage to the human auditory system. The
Occupational Safety and Health Administration (OSHA) provides protection for workers by regulating the
acceptable noise exposure. For instance OSHA limits the exposure of a 90 dBA continuous sound to eight
hours.(4) If the exposure is greater, hearing protection is required to ensure hearing damage does not occur.

The siren should not be overly startling to the driver

When a person is startled by a loud, unexpected sound, several involuntary reactions occur. The person
experiences a tightening of the muscles and a temporary increase in heart rate and blood pressure. As the
person tries to appraise the situation a brief period of disorientation occurs. Once the threat passes or the
person responds to the situation, the disorientation subsides and the muscles begin to relax. This is referred
to as the acoustic startle effect. The strength of the acoustic startle effect is determined by the familiarity,
loudness and the speed of onset of the sound, and the preparedness of the driver (14).

Just how familiar a driver is with the sound of an emergency siren is probably dependent on the frequency
of exposure. It is reasonable to assume that the more times the sound is experienced the more quickly it can
be distinguished. The preparedness of the driver is likely related to the driver's knowledge about the area.
If the driver knows the road which he is traveling on is the major route for emergency vehicles, he is likely
to expect to hear a siren. If he is traveling on a country road, the expectation of hearing a siren is probably
quite low.

As the emergency vehicle approaches a passenger car, the siren gradually increases in intensity with respect
to the drivers’ position. If the driver is unable to hear the siren until the emergency vehicle has become quite
Jacqueline Jenkins Page 275

close, the sudden onset may startle the driver, and the threat will require an immediate response. The
disorientation experienced by the driver may contribute to an inappropriate response.

The siren should not disrupt the driver from maintaining control of the vehicle

If the siren is disruptive, the drivers ability to process other signals will be affected. The driver will feel
overloaded and will involuntarily neglect information. If the information is important, such as the position
of other vehicles, a collision could result. However, the same result could be a manifestation of the
effectiveness of an individuals driving behavior.

The March, 1998 edition of the Human Factors journal included an article about the factors that influence
response times to audible stimuli. In three separate experiments, participants were required to respond to
sounds emanating from different locations. Wallace and Fisher concluded that the participants were quicker
to respond to sounds which emanated from a location directly in front of the participant position. The
response times increased as the information content of the auditory stimulus became more complex (15).

The emergency siren is complex. It can emanate from any direction with respect to the driver of a passenger
car. The siren can emanate from behind, in front, or to either side of the driver. Perhaps, the complexity of
the siren disrupts the driver from maintaining control of the vehicle as the driver's attention becomes focused
on identifying the location of the emergency vehicle.

The siren should have meaning for the driver

To evaluate whether drivers understand the meaning of the siren, the author reviewed the driver manuals
from the 50 U.S. states. The author assumes the direction provided by these manuals are indicative of a
driver’s knowledge of the required actions with respect to emergency sirens.

In each of the 50 U.S. state manuals, there is a section containing the expected response of a passenger car
driver upon the approach of an emergency vehicle whose lights and siren are in operation. All of the driver
manuals instruct the driver to maneuver his or her vehicle to the side of the roadway.

Several inconsistencies were noted during the review of the driver manuals. Thirty-one state manuals
specifically stated that the driver should not block the intersection. The remaining 19 manuals did not
provide any details about clearing the intersection.

Thirty-three state manuals specified that the driver must yield regardless of the direction of the emergency
vehicle. The Minnesota driver manual specifies for an emergency approaching from behind. The remaining
manuals use general terms, such as "approaching" and do not specify direction.

Forty-nine state manuals direct the driver to pull over to the right and the Idaho manual specifies that the
driver should pull over to the right side of the road where it is legal to park.

In addition, the Hawaii and Minnesota state manuals included don't panic statements and the Arizona and
Indiana manuals included a statement about being alert and keeping the radio low.

Evaluation findings

In the evaluation of the siren as an audible warning for passenger car drivers, the author considered the driver
of a passenger car, traveling directly in front of the emergency vehicle on a straight, level roadway. The siren
was assumed to be a Class A siren, capable of emitting a sound level of 120 dBA at 10 feet. The effects of
unimpeded sound propagation, and the sound dampening properties of the passenger vehicle were discussed.
Jacqueline Jenkins Page 276

Meteorological factors and the effects of traffic, buildings, berms and barriers on the propagation of the
sound wave were not considered.

Five criteria, identified as properties of a good audible warning system, were used to evaluate the emergency
siren system. These criteria require that the siren be heard by the driver without being startling or disruptive.
And that the siren must have meaning and must not harm the operator. The findings of the evaluation are
summarized in Table 2.

Table 2. Summary of evaluation findings

Evaluation criteria Does the siren satisfy the criteria?

The siren must be heard by the driver of the passenger car Sometimes
The sound level of the siren should not endanger the operator Yes
The siren should not be overly startling to the driver Sometimes
The siren should not disrupt the driver from maintaining Sometimes
control of the vehicle
The siren should have meaning for the driver Yes

Evaluation results

Only two of the five criteria are satisfied by the siren system. The operator is protected from the sound level
of the siren by the exposure limits published by the Occupational Safety and Health Association. A review
of 50 state manuals suggests that the siren should be understood by drivers. The remaining three criteria are
related to the sound level of the siren and when the sound is perceived by a driver.

When a passenger car is at rest, it is likely that the driver can hear the siren from 70 meters away. Beyond
70 meters the driver may hear the siren, depending on the ambient sound level inside the vehicle. When the
passenger car is traveling at 70 mph (112 km/h), the driver may be able to hear the siren from 14 meters
away.

The evaluation considered only one particular situation. It was shown that even in such a limited situation
the siren would not be heard by all types of drivers, in all types of passenger cars. The siren may be overly
startling or disrupt the driver from maintaining control of the vehicle. It is concluded that the emergency
siren system is deficient as an audible warning for passenger car drivers.

ENHANCEMENTS TO THE EMERGENCY SIREN SYSTEM

A remedy for the deficiency of the emergency siren system may include an enhancement of the sound source,
or the sound receiver. Perhaps a different means of communication could be explored. These potential
remedies will be discussed in the following sections.
Jacqueline Jenkins Page 277

Enhance the sound source

Increasing the sound level output of the siren would most likely increase the occurrence of the siren being
heard. More drivers would hear the siren from farther distances. Unfortunately, this simple approach to
remedy the deficiency of the siren system has several major drawbacks. An increase in the siren output level
could potential endanger the operator. At current operating sound output, the siren can startle and/or disrupt
a passenger car driver. Should the siren become louder, drivers may experience the startle effect more often
and with greater severity. Simply increasing the sound level output of the siren would not remedy the present
deficiencies of the emergency siren system.

Enhance the sound receiver

Alternately, the receiver could be enhanced. Instead of relying on the driver’s ability to detect the siren, an
artificially intelligent receiver could detect the sound of the siren and in turn notify the driver to the approach
of the emergency vehicle. There are several sound-based detection devices currently available, which are
typically used by the hard of hearing and the deaf. The device is installed inside the vehicle. When it detects
the siren sound it provides either an amplification of the siren or a visual cue such as a flashing red light.
Recognizing the needs of those suffering with hearing loss, General Motors uses its mobility fund to provide
siren detectors. These detectors are small units mounted to the dashboard. A flashing red light warns the
driver that an emergency vehicle is near (16).

Means of communication

Emergency vehicles have used sound, optical, radio and radar transmissions either for signal preemption or
for the surveillance of speeding vehicles. Such technology may be suitable in enhancing the siren system.

Sound transmissions

Sound-based technology has been used for signal preemption systems. The sound of the siren is detected
by a receiver connected to the signal controller. These systems have been found to have a short operating
range of approximately 1200 feet (17). The detection is dependent on the strength of the sound signal and
is sensitive to ambient noise(18). Meteorological conditions can influence the propagation of the sound
waves thereby limiting the use of sound-based technology. Physical barriers such as buildings, berms, and
trees can obstruct the sound wave, thereby requiring the source and the receiver to have a direct line of sight.

Light and strobe transmissions

Light and strobe-based technology have been applied to signal preemption systems. An infrared, visual light,
or xenon light is transmitted from the emergency vehicle and detected by a receiver connected to the signal
controller. These systems have been found to have an operating range of approximately 2500 feet(19). The
detection of the optical signal is based on the strength of the signal. It is influenced by meteorological
conditions and requires that the source be in direct line of sight with the receiver(18). The accuracy of the
signal is imprecise. Other light sources can interfere with the system and cause false detection.

One-way radio transmissions

The radio-based preemption systems use a one-way radio communication from the emergency vehicle to a
receiver connected to the signal controller. Although this technology is not influenced by weather
conditions, and does not require the source to be in direct line of sight with the receiver, the detection of the
radio signal is still dependent on the signal strength. The operating range of approximately 3500 feet, is
reduced when the signal is obstructed by objects such as buildings (18). The focus of the radio signal is
imprecise. All detectors within the operating range will receive the signal.
Jacqueline Jenkins Page 278

Radar transmissions

Traditionally radar and laser technology used by emergency personnel has been aimed at detecting vehicles
exceeding the posted speed limit of the roadway. This technology is now being used to alert drivers to the
approach of emergency vehicles. There are two companies which applied the radar and laser technology for
this new use.

Cobra Electronics Corporation has developed Safety Alert, a radar-based detection system. The transmitter
is designed to be linked to the operation of the emergency vehicle's light system. When the emergency lights
are on, the Safety Alert system transmits a radar signal. Older radar detectors can receive the signal and alert
the driver by means of an LED display and tone. A new generation of radar detectors can also receive text
messages. There are three different messages which may be transmitted, one for an approaching emergency
vehicle, another for a road hazard and a the third message to warn of an approaching train. This system has
a range of ¾ mile (20).

The Radio Association Defending Airwave Rights, Inc. conceived the concept of the Safety Warning System
(SWS), another radar-based warning system. Through assistance from industry partners and the Georgia
Tech Research Institute, SWS was developed. SWS is referred to as " the most technologically advanced
FCC approved system of its kind" (21).

The Safety Warning System uses a transmitter to emit a radar signal which can be detected by both
traditional radar detectors and the new SWS-aware radar detectors. The transmitter can be used in police,
fire, and emergency medical services vehicles, but are not limited to such use. It is designed to operate at
from both a moving vehicle and a stationary location. It can be mounted in conjunction with changeable text
message boards and even be linked to traffic management centers. The signal has a range of 1 ½ miles and
transmitters are presently in use in 33 U.S. states by transportation departments, police, sheriff and fire
departments, and emergency medical units.

When a traditional radar detector receives the SWS signal, it warns the driver by means of an LED display
and a tone. The SWS-aware detectors feature text messaging and synthesized voice messaging in addition
to the traditional display and tone. There are over 60 messages reserved for highway construction and
maintenance, highway hazard zone advisory, weather related hazards, travel information/convenience, and
fast/slow moving vehicles. The system distinguishes between a stationary police vehicle, a stationary
emergency vehicle, an emergency vehicle in transit, and a police vehicle in pursuit (21).

Additional consideration

When evaluating potential enhancements of the siren system, the ability to warn non-vehicle roadway users
must be considered. The current siren system, which produces an audible signal, not only warns vehicle
drivers of the approaching emergency vehicles but also warns non-vehicle roadway users such as pedestrians
and bicyclists. Optical, radio and radar-based systems do not have the ability to warn non-vehicle roadway
users.

Enhancement summary

Optical, radio or radar transmissions may be used to enhance the existing emergency siren system. Table
3 highlights the operating characteristics of the technologies used by emergency vehicles for signal
preemption and the surveillance of speeding vehicles.
Jacqueline Jenkins Page 279

Table 3. Summary of operating characteristics

Type of Transmission
Characteristic Sound Optical Radio Radar
Operating range 1200 ft 2500 ft 3500 ft 3/4 to 1 ½ miles
Sensitive to signal strength Yes Yes Yes Yes
Requires line of sight Yes Yes No Yes
Precise signal direction No No No Yes
Weather sensitive Yes Yes No No
Sound sensitive Yes No No No
Light sensitive No Yes No No
Warns drivers Yes Yes Yes Yes
Warns pedestrians Yes No No No

Enhancement findings

The current emergency siren system is a sound-based system. To determine the suitability of the alternate
means of communication as potential remedies of the siren system deficiencies, the characteristics of the
optical, radio and radar-based systems are compared to the characteristics of a sound-based system. The use
of an alternate means of communication must provide an additional benefit to be considered an improvement
or enhancement of the existing siren system.

Sound-base systems have a unique quality. Such systems are able to warn drivers, pedestrians, cyclists and
other potential roadway user groups. No other technology presented above is capable of warning roadway
users other than vehicles drivers.

Although, optical-based systems have an operating range more than double that of sound-based systems, both
systems require that the source be in direct line of sight with the receiver. These systems are weather
sensitive and the receivers are dependent on the strength of the signal. Optical-based systems do not provide
any benefit not already provided by sound-based systems.

Radio-based systems have an operating range almost three times that of sound-based systems. Unlike sound-
based systems, radio-based systems do not require a direct line of sight between the source and the receiver
and are not affected by weather conditions.

Radar-based systems have a few advantages over sound-based systems. Radar-based systems are not affected
by weather and have much greater operating ranges. The radar signal can be directed with some precision
unlike sound signals.

Enhancement results

Since the current emergency vehicle siren system uses sound transmissions for communication, the alternate
means of communication were evaluated through a comparison of operating characteristic with those of
sound-based systems. The use of an alternate type of transmission must provide an additional benefit to be
Jacqueline Jenkins Page 280

considered suitable as an enhancement of the existing siren system. The benefits and shortcomings of the
transmission types are summarized in Table 4.

Table 4. Summary of the benefits and shortcomings of the transmission types.

Transportation Benefits Shortcomings

Types
Optical transmissions Greater operating range Cannot warn pedestrians
Not sensitive to ambient sound Sensitive to light sources
Radio transmissions Greater operating range Cannot warn pedestrians
Not sensitive to weather
Not sensitive to ambient sound
Does not require line of sight
Radar transmissions Greater operating range Cannot warn pedestrians
Not sensitive to weather
Not sensitive to ambient sound
Signal direction is precise

Although the use optical transmissions will provide additional benefit, its suitability as an enhancement of
the existing emergency siren system is outweighed by its shortcomings. Therefore, the author feels that the
use of optical transmissions should not be considered as a potential enhancement of the existing siren system.
The operation of the emergency siren system would benefit from the inclusion of either radio or radar
transmissions. These technologies must be used as enhancements and not as replacements, otherwise the
ability to warn pedestrians and other non-vehicle roadway users would be lost.

RECOMMENDATIONS

Although the emergency vehicle siren system was found to meet only two of the five evaluation criteria, this
system has the unique ability to warn not only vehicle drivers but also pedestrians and other non-vehicle
roadway users. Systems which use means of communication other than sound transmission do not exhibit
this characteristic. Therefore, maintaining the use of the emergency vehicle siren system is recommended.

A comparison of alternate means of communication was made. Optical-based systems do not exhibit
characteristics beyond that of sound-based systems with the exception of the operating range. Optical
technology is not recommended as an enhancement as it does not remedy the deficiency of the emergency
siren system.

Radio and radar-based systems provide greater operating ranges and are not sensitive to weather conditions.
Radio transmissions do not require line of sight and do not have specific signal direction. Radar
transmissions required broad line of sight but the signal is direction specific. Both of these alternate means
of communication provide benefits not provided by sound-based systems. Although these systems are not
able to warn pedestrians and other non-vehicle roadway users, they are recommended as enhancements to
remedy the deficiency of the current emergency siren system.

This research paper was limited to the evaluation of the emergency siren system as an audible warning
system for passenger car drivers. The effectiveness of the emergency siren in reducing the response time of
emergency vehicles was not evaluated. Such an evaluation is recommended for future study.
Jacqueline Jenkins Page 281

Evaluation criteria 5 stated that the siren must have meaning. The author assumed that the information in
the 50 U.S. state driving manuals was indicative of drivers knowledge, it is possible that the driver
understanding associated with the emergency siren system is slightly different. It is also possible that driver
reaction to the siren is inconsistent with driver understanding due to the startle effect. This is an area that
requires future investigation.

To ensure drivers understand the meaning of the emergency siren and to give the driver the confidence to
act appropriately, continued driver training is required. To ensure the siren has a consistent meaning,
operator training is essential.

ACKNOWLEDGMENTS

This paper was prepared for the graduate course, Advanced Surface Transportation Systems, offered at
Texas A&M University. The author would like to express her thanks to Dr. Conrad Dudek for his efforts
in putting together such an excellent course and for providing an opportunity to work closely with a number
of transportation professionals. The author recognizes the contributions of all the mentors including Thomas
Hicks, Patrick Irwin, Wayne Kittelson, Douglas Robertson, William Spreitzer, and Carol Zimmerman and
thanks them for their time, advice, and encouragement.

REFERENCES

1. Let them through… It could be you. National Highway Traffic Safety Administration, Hanson Industries,
New York, N.Y.

2. National Safety Council First Aid Newsletter. April 1997.

http://bedford.jbpub.com/nsc/firstaidnews/april1997/fad.htm Accessed July 12, 1999.

3. Emergency Vehicle Sirens. J1849. Society of Automotive Engineers, Warrendale, P.A. Rev. August
1995.

4. C. D. Wickens, S. E. Gordon, and Yili Liu. (1988). An Introduction to Human Factors Engineering.
Addison-Wesley Educational Publishers Inc., pp. 113-144., 513-518.

5. Noise and Vibration Control. Edited by L. L. Beranek. Published by the Institute of Noise Control
Engineering, Washington, D.C. 1988.

6. Federal Signal Corporation. http;//www.fedsig.com/products/ Accessed May 13, 1999.

7. Vehicle Speakers. http//www.wfrfire.com/lighting/speakers/ Accessed May 13, 1999.

8. National Institute of Justice Standard for Emergency Vehicle Sirens. Department of Justice, National
Institute of Justice. December 1981.

9. Technical Noise Supplement.

http:www.dat.ca.gov/hq/Environmental/offdocs/hazdocs/tens/tens2000.htm October 1998 Accessed
June 24, 1999.

10. DuPont Glass Laminating Products. http://www.dupont.com/safetyglass/ Accessed July 12, 1999.
Jacqueline Jenkins Page 282

11. Transportation Noise Reference Book. Edited by P.M. Nelson, BSc, PhD, FIOA. Butterworths & Co.
Ltd., 1987.

12. Car and Driver Road Test Annual '99Ultimate Comparisons. Hachette Filipacchi Magazines, Inc.,
New York, N.Y. 1999

13. Rubin, A. I., G. L. Howett Emergency Vehicle Warning Systems. U.S. Department of Commerce,
National Bureau of Standards, U.S. Government Printing Office, Washington D.C. 1981.

14. Fundamentals of Noise and Vibration. Edited by F.J. Fahy and J.G. Walker, Routledge, New York,
N.Y. 1998.

15. Wallace, J. S., D. L. Fisher. Sound Localization: Information Theory Analysis. The Journal of the
Human Factors and Ergonomics Society, Volume 20, Number 1, March 1998.

16. GM Mobility Program. http://www.gm.com/vehicles/us/owners/mobility.html Accessed July 12, 1999.

17. Emergency Preemption Systems Inc. http://www.preemption.com/superior.htm. Accessed July 30, 1999.

18. ASKCA Inc. http://askca.com/cat 01201/040001.htm Accessed July 30, 1999.

19. Tomar Electronics Inc. http://www.tomar.com/products/sectionj/p075.htm Accessed July 30, 1999.

20. Scaduto Teri, Driving Intelligence. http://popularmechanics.com/popmech/elect/9610EFMCM.html

October 1996, Accessed July 12, 1999.

21. Safety Warning System. http://www.swslc.com/ Accessed July 12, 1999.

JACQUELINE JENKINS

Jacqueline Jenkins is a 1996 graduate of the Bachelor of Science,

Civil Engineering program at the University of Waterloo and a
licensed Professional Engineer in the province of Ontario,
Canada. During her undergraduate studies, she was employed by
the Toronto Transit Commission to analyze proposed public
transit infrastructure improvements. Prior to graduation, she
joined the Transportation Planning Department of the consulting
firm Marshall Macklin Monaghan where she conducted traffic
impact and infiltration studies. In the fall of 1997, Jacqueline
accepted the position of Collision Investigator with the Vehicle
Safety Research Center at Ryerson Polytechnic University. As
part of a national study funded by Transport Canada, she
investigated vehicle collisions to evaluate existing vehicle safety
features. Her interest in traffic safety and occupant protection brought her to Texas A&M to pursue a
Masters of Engineering degree, followed by a doctorate degree in Traffic Safety. Currently, she is a Research
Assistant with the Texas Transportation Institute and a member of both the Society of Automotive Engineers
and the Institute of Transportation Engineers.
APPLICATION OF AUTOMATED ENFORCEMENT FOR
HIGHWAY-RAILROAD GRADE CROSSINGS

by

Jonathan K. Lammert

Professional Mentors
Wayne K. Kittelson
Kittelson & Associates, Inc.

and

William M. Spreitzer
General Motors

Course Instructor
Conrad L. Dudek, Ph.D., P.E.

Prepared for
CVEN 677
Advanced Surface Transportation Systems

Department of Civil Engineering

Texas A&M University
College Station, Texas

August 1999
Jonathan K. Lammert Page 284

SUMMARY

With the hundreds of thousands of highway-railroad grade crossings in the U.S., it is impossible, both from
a standpoint of funds and practicality, to enforce violation laws with traditional methods. Several techniques
which utilize state-of-the-art technology have been developed to solve this problem. These enhancements
to highway-railroad grade crossings could be used at locations with the largest traffic and train volumes.
Automated enforcement is one of these enhancements that could prove to be extremely valuable.

Automated enforcement typically consists of detection equipment and cameras to capture images of drivers
violating the law. New York City and Los Angeles were among the first locations in the U.S. to use this type
of law enforcement. After the favorable results that these programs produced, other areas such as San
Francisco, California; Howard County, Maryland; Jonesboro, Arkansas; and Ames, Iowa have initiated
automated enforcement programs of their own. These programs are ideal for examination in the installation
of new automated enforcement systems.

The primary objectives of this paper were to evaluate the use of automated enforcement at railroad crossings
already in place in the United States, including the observed results in both safety and public acceptance;
examine the automated enforcement products, such as sensors and cameras, now available, including their
cost and capabilities; identify the concerns of railroad companies in regard to automated enforcement; apply
the principles and data gathered to develop a procedure for installing an automated enforcement system; and
formulate an automated enforcement system for a highway-railroad grade crossing in College Station, TX.

These objectives were accomplished by performing a literature review on the issues involved with accidents
and violations at highway-railroad grade crossings. In addition, telephone conversations were conducted
with officials involved with automated enforcement programs in the U.S.

Based on the research, the six following guidelines were formulated to follow in the implementation of an
automated enforcement program:

• Review applications in the U.S.;

• Establish good relations with the public;
• Purchase equipment;
• Install system;
• Execute trial runs; and
• Evaluate and modify.

All officials that were contacted for this report expressed confidence that automated enforcement could
become an extremely valuable tool in the future of law enforcement. However, in comparison with other
transportation technologies and advances, automated enforcement is still relatively new. It was concluded
that additional research and study are needed to identify problems and their corresponding answers
concerning this technology.
Jonathan K. Lammert Page 285

TABLE OF CONTENTS

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
Research Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
Study Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
Study Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Organization of the Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289

VIOLATIONS OF HIGHWAY-RAILROAD GRADE CROSSINGS . . . . . . . . . . . . . . . . . . . . . . 289

Accident History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Definition of Crossing Violations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
Traditional Enforcement Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
Enhancements to Highway-Railroad Grade Crossings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292

TECHNOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Automated Enforcement Subsystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Violation Detection Subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Violation Recording Subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
Citation Processing Subsystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
Available AE Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
American Traffic Systems - SafeTrax RC-200 Railroad Crossing Violation Camera . . . . . . . 295
Econolite - AUTOSCOPE Video Detection System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
SAIC - Traffic Violator Grade Crossing Monitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
USPT - TRAXGUARD Automated Railroad Grade Crossing Enforcement System . . . . . . . . . 296
Other Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297

HIGHWAY-RAILROAD GRADE CROSSING AE PROGRAMS IN U.S. . . . . . . . . . . . . . . . . . 297

Los Angeles, California . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
Chosen Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
Citations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
Effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
Public Awareness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
Jonesboro, Arkansas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
Chosen Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
Citations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
Effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
Public Awareness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Ames, Iowa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Chosen Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Citations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Public Awareness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
Norfolk Southern Corporation, NC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
Texas Department of Transportation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
Miami, Oklahoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304

INFORMATION ACQUIRED FROM AE PROGRAM OFFICIALS . . . . . . . . . . . . . . . . . . . . . 304

Jonesboro, Arkansas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
Ames, Iowa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
Jonathan K. Lammert Page 286

Texas Department of Transportation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305

GENERAL GUIDELINES FOR INSTALLATION OF AE SYSTEM . . . . . . . . . . . . . . . . . . . . . 305

Review of Applications in the U.S. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
Establishment of Good Relations with Public . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
Purchase of Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
System Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
Execution of Trial Runs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
Evaluation and Modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307

APPLICATION OF AE PROGRAM TO CASE STUDY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307

Review of Applications in the U.S. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Establishment of Good Relations With Public . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Purchase of Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
System Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Execution of Trial Runs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Evaluation and Modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311

CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312

ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313

APPENDIX: SAMPLE INTERVIEW QUESTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315

Jonathan K. Lammert Page 287

INTRODUCTION

Automated enforcement typically consists of detection equipment and cameras, either 35-mm or digital
models, to capture images of drivers violating the law. This method of law enforcement has been in use in
some form in the United States since the 1980s. The most popular applications of automated enforcement
(AE) are for violations of traffic signals, HOV and bus lanes, speed limits, highway-railroad grade crossings,
and electronic toll collection (1).

The applications that have undergone the most studying and testing are for violations of traffic lights and
highway-railroad grade crossings. Large cities such as New York City and Los Angeles were at the forefront
of the use of AE, mostly due to their agencies’ ability to fund the programs (2). Because of the success these
two agencies have had, automated enforcement has been expanded to several other cities such as San
Francisco; Howard County, Maryland; Jonesboro, Arkansas, and Ames, Iowa. Other states have had
agencies conduct studies and tests using automated enforcement for these two applications, but have held
off implementing it the programs until legal questions have been answered. Table 1 summarizes the use of
AE for traffic signals and highway-railroad grade crossings.

Table 1. Automated Enforcement Programs in Operation and Demonstration

Projects of Traffic Signals and Highway-Railroad Grade Crossings (2)

Application Types Traffic Signals Highway-Railroad Grade

Crossings
Programs in Operation Oxnard, CA Jonesboro, AR
San Francisco, CA Los Angeles, CA
Howard County, MD Ames, IA
New York City, NY
Victoria, Australia
Demonstration Projects Polk County, FL Norfolk Southern Corporation, NC
Jackson, MI Department of Transportation, TX
Lincoln, NE Miami, OK

AE at traffic signals, where a violation occurs when a vehicle crosses the stop bar after the traffic signal has
changed to the red phase, has been the application that has been tested and studied the most (2). As of
December 1998, AE systems have been installed at over 130 intersections in Victoria, Australia, and at nearly
50 intersections in the United States. The Netherlands has also used AE for this purpose.

Violations at highway-railroad grade crossings occur when a vehicle is driven around, under, or through a
crossing gate that is closed, or is in the process of being raised or lowered (3). AE for highway-railroad grade
crossings has been in place in three areas in the United States: Los Angeles, California; Jonesboro, Arkansas;
and Ames, Iowa. AE has also been extensively tested by several state agencies.

Highway-railroad grade crossings can use either an active or passive warning system. Active warning
systems use flashing lights or signals, bells, and lowering gates to convey the message that a train is
approaching to drivers. A passive warning system, on the other hand, does not feature any of these. Since
a passive system will typically be in a location with a low amount of traffic and train volume, the drivers will
use the sight distance provided to watch for an oncoming train (1).
Jonathan K. Lammert Page 288

In the past few years, several tragic accidents have occurred at highway-railroad grade crossings in the U.S.
which have made the national news and attracted public attention. These collisions between trains and
vehicles usually result in serious injury or death for the vehicle operator, and always cause extensive property
damage. Some accidents have also resulted in the loss of life for passengers on trains that have derailed after
the collision (4). The fact that a collision between a train and a vehicle contains a high amount of potential
for injury and loss of life is evidence that steps must be taken to increase the safety at highway-railroad grade
crossings.

The traditional method for a violator of a highway-railroad grade crossing to be ticketed was if a nearby
police officer saw them. The number of crossings that have histories of violations makes the deployment
of an officer at each crossing both economically infeasible and impossible in regards to the limited resources
at a police department’s disposal. Even if this plan could be implemented, the police presence might cause
distractions along the roadway (5). Therefore, an enforcement plan must be both cost-friendly to the tax-
paying public and keep the intrusion on the natural traffic flow as low as possible. Such a plan could be used
on railroad crossings that experience the highest volume of traffic and/or observed violations. AE has the
potential to fulfill these requirements. Further testing is required to determine if this potential can be
realized.

There are two reasons why enforcement systems such as AE have first been installed at highway-railroad
grade crossings with active warning systems. The first reason is that active warning systems typically
operate in areas where both train and vehicle volumes are high and the necessary funds are available to install
these types of warning systems. Conversely, passive warning systems remain in place at highway-railroad
grade crossings for one of two scenarios: (1) The crossing does not have a history of accidents, or (2) The
crossing is maintained by agencies that do not have the resources to upgrade it to an active warning system.
The second reason is that highway-railroad grade crossings with passive warning systems tend to be in
locations where power is not readily or easily available for advanced systems such as AE (6).

Research Objectives

The primary objectives of this paper were to:

• Evaluate the use of automated enforcement at highway-railroad grade crossings already in place in the
United States, including the observed results in both safety and public acceptance;
• Examine the automated enforcement products, sensors and cameras, now available, including their costs
and capabilities;
• Identify the concerns of railroad companies in regard to automated enforcement;
• Apply the principles and data gathered to develop a procedure for installing an automated enforcement
system at a highway-railroad grade crossing; and
• Formulate an AE system for a highway-railroad grade crossing in College Station, TX.

Study Approach

The objectives of this paper were accomplished with several methods. Literature was reviewed regarding
the recent history of accidents at highway-railroad grade crossings and the types of motorist violations that
have caused them. Enforcement efforts to curtail these violations were identified. The components of an
AE system were discussed, including what products currently exist.

Locations that have used AE for highway-railroad grade crossings were identified. Information concerning
the accident problems that these crossings experienced was acquired, as were the results of having the AE
system in place. Officials involved in the AE systems were contacted, and their opinions and the information
pertaining to the crossings was analyzed. Guidelines for installing an AE system were then prepared.
Jonathan K. Lammert Page 289

Study Scope

The scope of this paper was limited to highway-railroad grade crossings that have active warning devices.
While violations of both active and passive warning systems can lead to accidents, active warning systems
were chosen as the focus for reasons mentioned in the Introduction. Further research is needed on identifying
techniques that improve safety at highway-railroad grade crossings with passive warning systems.

Organization of the Report

Following the introduction, this research report is divided into six sections. The first section contains an
overview of the recent history of accidents at highway-railroad grade crossings. In addition, the violations
that caused these collisions and enforcement efforts are discussed. The following section is composed of
a review and summary of the types of AE products and technology that are presently available. Following
the technology section is a detailed account of the locations in the United States that either have an AE
program in place, or have conducted tests or demonstrations with AE. The following section contains a
summary of discussions that were made between the author and officials at three of these locations. The next
section includes a discussion of the general procedures that must be done to install an AE system. After the
procedures section, recommendations are made regarding the application of an AE system to a highway-
railroad grade crossing in College Station, Texas. Finally, there are conclusions based on the previous
sections.

VIOLATIONS OF HIGHWAY-RAILROAD GRADE CROSSINGS

Accident History

Highway-railroad grade crossings have been locations of accidents for decades. The statistics indicate a
disturbing trend. In the U.S., a collision between a train and a vehicle or pedestrian occurs every 100 minutes
(4). If only vehicles are included in these collisions with trains, the frequency of an occurrence is once every
90 minutes. These collisions occur on the 280,000 highway-railroad grade crossings, which are either
publicly or privately owned (7). While the number of collisions at highway-railroad grade crossings has
decreased steadily over the last eighteen years (Table 2), the number of fatalities these collisions have caused
has decreased at a much slower rate.

Table 2 indicates that the number of collisions at crossings has decreased from 9295 in 1981 to 3446 in 1998,
or a reduction of nearly 63 percent. During this time frame, the number of fatalities at these crossings has
decreased from 728 to 422, a reduction of only 42 percent. The fact that the number of collisions has fallen
could be attributed to several reasons. First, significant improvement has been made over the last two
decades in the quality and quantity of advance warnings that are provided for drivers approaching highway-
railroad grade crossings. The Highway Safety Act and the Federal Railroad Safety Act of 1970 initialized
procedures and allocated federal expenditures for safety improvements at highway-railroad grade crossings.
More than 25,000 improvement projects resulted from this Act, which have saved an estimated 6,400 lives
(8). Second, public awareness has been heightened with the use of television commercials, educational
literature, and the media attention given to previous accidents. The number of highway-railroad grade
crossings might have also decreased due to railroad deregulation (6).

While the numbers of collisions and fatalities have decreased, a closer examination of the data reveals a
growing problem. The number of fatalities per collision has risen from 0.078 in 1981 to 0.122 in 1998, an
increase of over 50 percent. Furthermore, the increase was relatively steady from 1981 until 1989, where
the number remained near 0.12 fatalities per collision in the following years (4). It is the opinion of the
author that three conclusions can be drawn from this trend.
Jonathan K. Lammert Page 290

Table 2. National Highway-Rail Grade Incidents at Public and

Private Crossings for All Highway Users (4)

Year Collisions Fatalities Number of

Fatalities
per Collision
1998 3446 422 0.122
1997 3865 461 0.119
1996 4257 488 0.115
1995 4633 579 0.125
1994 4979 615 0.124
1993 4892 626 0.128
1992 4910 579 0.118
1991 5386 608 0.113
1990 5713 698 0.122
1989 6525 801 0.123
1988 6615 689 0.104
1987 6391 624 0.098
1986 6396 616 0.096
1985 6919 582 0.084
1984 7281 649 0.089
1983 7161 575 0.080
1982 7748 607 0.078
1981 9295 728 0.078

First, it is possible that the accidents that have been eliminated are not as severe as the accidents that are still
occurring. A collision where a vehicle is driven into the side of a train at a crossing is not as severe as a
collision where a train strikes a vehicle that is traveling across a crossing. With the upgrading of warning
systems from passive to active, the likelihood of a vehicle being driven into the side of a train has decreased
(4). Second, the overall average speed of trains might have increased. Researchers have found that the large
size of a train causes it to appear to be traveling at a much slower speed to a driver. A driver can
miscalculate the amount of time until the train arrives at the crossing. If the driver decides to violate the
crossing, a collision with the train will be even more severe with the train’s increased speed (9). Finally,
there has been an increase in the percentage of small, compact cars on the road. The vehicles would not offer
as much protection in the event of a collision as larger vehicles would (10). While these are potentially
plausible reasons, they and their corresponding solutions lie outside the scope of this report.

In 1994, 53 percent of the 4,979 highway-railroad grade crossing accidents occurred at sites equipped with
an active warning system. It has been estimated that 18 percent of the nation’s highway-railroad grade
crossings can be classified as using an active warning system (5). Therefore, even though active warning
systems account for a small minority of the highway-railroad grade crossings, they are the location for a
majority of the collisions. As was mentioned in this report’s introduction, crossings with active warning
systems typically experience a higher volume of train and traffic volume than crossings with passive warning
systems. This higher volume leads to higher number of possible collisions. Studies can be conducted on the
causes of these accidents because crossings with active warning systems are usually located in areas with
agencies that can fund these activities.
Jonathan K. Lammert Page 291

Definition of Crossing Violations

The focus of this report was on highway-railroad grade crossings with active warning systems. To narrow
the scope even further, only active warning systems with automatic gates was discussed. The methods of
violating this type of active warning system are as follows (5):

• Driving under the gates as they are descending;

• Driving around the gates after they are in the horizontal position;
• Driving through the gates after they are in the horizontal position; and
• Driving under the gates as they are ascending.

These types of violations can lead directly to collisions that can result in injuries and fatalities for the
occupants of both the train and the vehicle. It has been the responsibility of police departments to apprehend
violators of highway-railroad grade crossings.

Traditional Enforcement Methods

Before AE and other recent advancements, the only way for a violator of a highway-railroad grade crossing
to be ticketed was if a nearby police officer saw them. The police officer would make a traffic stop and issue
a citation. Two problems stem from attempting to use this traffic stop technique exclusively.

First, the observing police officer must be able to pursue the violator to administer a traffic stop. If the
officer is in the lane that is traveling in the opposite direction of the violator, this poses no problem. The
officer can perform a U-turn maneuver and stop the violator. In this case, it is doubtful that a driver will
commit a violation if an officer is in view in front of the driver. However, if the officer is traveling behind
the driver in the same lane, the officer will also have to cross over the railroad tracks and through or between
the gates to pursue the violator. This places the officer in an extremely dangerous position. In this case, it
is not worth catching an offender for a fine if additional lives are placed in danger. Also, it is possible that
the train will arrive before the violator can be pursued. Unless the officer has recorded the license plate
number of the vehicle, it is nearly impossible to apprehend the violator.

To combat this potentially dangerous exposure to police, team enforcement has been attempted, which is the
process of using multiple officers. One police officer is stationed near the highway-railroad grade crossing,
usually in an unmarked vehicle. When a violation is observed, the officer radios a police officer downstream
from the crossing with information about the vehicle. The second officer then apprehends the violator.
Another team enforcement method that has been tried is called the “Trooper of the Train” plan. In this plan,
a police officer who is equipped with a radio rides in the locomotive cab of a train. When the officer
observes a crossing violation, the officer notifies another nearby police officer to stop the vehicle (10).

These methods eliminate the danger of exposure that would occur if an officer was to follow the violator
through the crossing. However, this is an extremely costly use of resources for any police department, and
few departments can continue using team enforcement for a long period of time (1).

The second problem with using traffic stop methods, both single officers and teams, is related to the large
number of crossings. In 1997, approximately 225,000 public and 140,000 private highway-railroad grade
crossings were in existence in the U.S. (5). Enforcement of these crossings would require a very large
amount of officers, police vehicles, and time, and would be practically impossible. Even if only the crossings
with the largest histories of violators are manned, they would still require a large amount of resources and
time from the police departments. Also, much of the time spent at a crossing would be wasted due to the
low frequency and unpredictability of violations. Also, if these traffic stop methods were implemented, the
police presence might cause distractions along the roadway.
Jonathan K. Lammert Page 292

Furthermore, it is always possible that the violation will be challenged in court. For the violation to be
upheld, the police officer would have to attend the trial, spending time which would be better served
elsewhere. Also, there is always the chance that the violation will not be upheld by the court. Violations
have been dismissed due to reasons such as the observing police officer not showing up for the trial, poor
field notes being written about the violation, and lawyers being able to establish that the officer did not see
if the crossing’s warning system was operational (1). Therefore, an enforcement plan must be cost-friendly
to the tax-paying public and keep the intrusion on the natural traffic flow as low as possible. Such a plan
could be used on railroad crossings that experience the highest volume of traffic and/or observed violations.

Enhancements to Highway-Railroad Grade Crossings

Several new enhancement plans have been developed in recent years that are designed to prevent violations
from occurring at highway-railroad grade crossings. A few of these enhancements are (5):

• Four-quadrant gates;
• Highway traffic signals;
• Raised medians; and
• Automated enforcement.

Four-quadrant gates are installed at highway-railroad grade crossings with active warning systems already
in place. When the gates are in the horizontal position, both directions of roadway on both sides of the
railroad tracks are blocked by the gates. This enhancement is designed to prevent violations of vehicles
being driven around the gate arms. Four-quadrant gates do not, however, prevent violations where the
vehicle is driven through the gates in the horizontal position or under the gates as they are ascending or
descending (5).

The flashing red lights of a typical active warning system can be replaced with highway traffic signals. This
method is based on the assumption that drivers will recognize and obey a traffic signal more often than they
will a system of flashing red lights. However, this enhancement plan does not directly prevent the four
violations of gate arms previously mentioned (5).

Another enhancement that can be used to prevent vehicles from traveling between gate arms in the horizontal
position are raised medians. These barriers separate the opposing traffic lanes and keep vehicles from
performing the zigzag maneuver. Like four-quadrant gates, however, raised medians can not prevent
violations where a vehicle travels through the lowered gates or under the gate arms as they are ascending or
descending. In addition, the construction work that is needed to install median barriers will cause disruptions
to the normal flow of traffic (6).

While these three enhancements have the potential to be effective at enforcing highway-railroad grade
crossings, each possesses a weakness that violators can exploit. Further testing must be carried out on each
of the three enhancements before wide-spread implementation takes place.

Each of these three systems attempts to prevent the physical act of violating a highway-railroad grade
crossing. Automated enforcement, on the other hand, is designed to enforce crossings by automatically
recording the violation and enabling police officers to issue citations. The components of an AE system are
identified in the following section.
Jonathan K. Lammert Page 293

TECHNOLOGY

Automated enforcement systems can record photographs of vehicles that commit violations at highway-
railroad grade crossings. These photos can be used as evidence in issuing the violator a citation. This section
contains an extensive review of how automated enforcement systems operate.

Many requirements need to be met by AE technology for an enforcement system to be effective. According
to a 1996 article, AE technology must be capable of attaining the following ten qualities (11):

• Easily capture, transmit, process, store, and recover images;

• Provide resolution of the image that is extremely high in detail;
• Minimize the effect that overexposed portions (vehicle headlights and sunlight) have on the image as a
whole;
• Provide a high amount of detail through the use of light and dark contrasting;
• Capture blur-free images of moving vehicles;
• Detect violations in varying levels of light (which occur at different times during the day);
• Eliminate sensor defects such as bright or dark columns on the image through the use of camera
electronics such as image enhancement circuitry;
• Capture images at frame rate high enough to identify multiple vehicles committing violations;
• Possess installation flexibility so it can be mounted into permanent, temporary, and mobile settings; and
• Be constructed of materials that are environmentally friendly.

In addition, AE systems must be durable, cost-effective, and accurate.

AE technology will be exposed to the elements of nature. It must be able to operate in variable conditions
such as rain, high wind, ice, and hot and cold temperatures. AE systems are also frequent targets of
vandalism, especially in remote and economically depressed areas (6).

The AE system must be inexpensive enough to warrant the large capital cost that is required to install it. If
the perceived benefits do not out-weigh the cost of purchasing and setting up the equipment, other methods
of enforcement of highway-railroad grade crossings should be explored (2).

Finally, an AE system must be able to consistently provide accurate photographs that can be admissible in
courts of law as evidence of a violation. It is a waste of time and resources if the system produces images
from false alarms. If the photographe is taken either a second too soon or too late, it might be impossible
to identify the violator (12). The components of an AE system must all work together efficiently to qualify
the entire system a success.

Automated Enforcement Subsystems

AE systems are typically made up of three main subsystems: a violation detection subsystem, a violation-
recording subsystem, and a central computer subsystem (12). The operations of the AE subsystems are
described in the following paragraphs.

Violation Detection Subsystem

The violation detection subsystem can be of three different types: inductive loops , infrared beams, and video
imaging cameras. All three types employ different technologies to accomplish the same basic task.
Typically, the device is activated along with the active warning system components (gates, bells, and flashing
red lights). This is accomplished by the violation detector subsystem being connected with the normal
circuitry of the highway-railroad grade crossing. Either a singular detector or multiple detectors are used to
cover both highway approaches. Both highway approaches must be covered so that the four types of
Jonathan K. Lammert Page 294

violations can be detected (12). After the train has passed and the features of the active warning system have
ceased operation, the violation detection equipment will power down and remain in stand-by mode.

Violation Recording Subsystem

For a citation to be mailed to the violator, evidence of the violation is required, therefore each AE system
features a violation recording subsystem. For a citation to be issued and upheld in court, there must be a
photograph of the vehicle crossing the tracks while the crossing’s warning system is active. The four
products that will be later discussed feature the ability to capture two photographs of the violation. These
photos or images are taken when the violation detection subsystem relays notification that a violation is
taking place. The images must meet three requirements to be admissible in court (2):

• The license plate of the vehicle is visible and can be identified;

• The vehicle is shown to be committing one of the four types violations mentioned in the previous section;
and
• The warning system of the crossing must be shown to be operating correctly.

The subsystem will typically imprint violation information such as location, time, and date on the image that
is taken. Both violation detection and violation recording subsystems can be housed in or on structures such
as controller cabinets and signal and sign poles. This characteristic is designed to provide safety from the
elements and vandalism (12).

Citation Processing Subsystem

After the previous subsystems have obtained the evidence that is needed to cite a violator, equipment must
be available that will store and process the photographs. Earlier AE systems kept the images stored in the
housing structures that contained the violation recording devices. An officer would periodically stop at each
crossing and retrieve the stored information. Newer systems have the ability to relay information from each
crossing by way of telephone line, fiber-optic wire, or other advanced methods (13). The data can be sent
to police departments for storage in a computer database.

In addition to the three requirements of images that were addressed in the previous paragraphs, the National
Cooperative Highway Research Program Synthesis 219 identifies six requirements that must be met in order
for a citation given through the use of AE to be acceptable (14).

• Identification of the date, time, and location of the violation;

• Existence of an enabling stature that meets the legal and constitutional standards of the courts;
• Periodic certification that the products meet the specifications set forth by the NHTSA and the
corresponding state agency;
• Evidence that the instrument was working properly at the time of the offense;
• Proof that the operator/monitor of the AE instruments are properly trained; and
• Proof of the scientific reliability of the AE instrument.

If the police officers at the department can verify that these requirements have been met, they can continue
with the process of issuing a citation. The officers can record the license plate number from the images, and
match the vehicle with its owner. A computer database at the police department will contain information
regarding the owners of registered vehicles. Citations can then be sent out to the car owners. Certain AE
programs have allowed owners to identify the driver of the vehicle if they were not responsible for the
violation. The citation is then sent by the police department to the driver responsible (2).
Jonathan K. Lammert Page 295

Available AE Products

Due to the fact that AE for highway-railroad grade crossings has only been installed in three locations in the
U.S., and has been demonstrated by a few other states, there are a very limited number of results that can be
analyzed with regard to the different product manufacturers. Therefore, the focus of this section will be on
the four manufacturers whose products have been installed. The companies are American Traffic Systems,
Econolite, SAIC, and US Public Technologies. Their corresponding products are the SafeTrax RC-200
Railroad Crossing Violation Camera, the AUTOSCOPE Video Vehicle Detection System, the Traffic
Violator Grade Crossing Monitor, and the TRAXGUARD Automated Railroad Grade Crossing Enforcement
System, respectively (1). These products will be discussed in detail, and other potential companies that
produce technology which could be useful for railroad crossing AE will be identified.

American Traffic Systems - SafeTrax RC-200 Railroad Crossing Violation Camera

In regard to violation detection subsystem, SafeTrax offers a choice between standard inductive loops and
piezoelectric sensors. Inductive loops are used for basic vehicle detection, without any special features.
Piezoelectric sensors, however, have the ability to classify the vehicle based on its length and number of
axles and measure the vehicle’s speed. The violation recording subsystem will use this speed to determine
when the second image is recorded. This feature of the sensors can also serve another purpose. In addition
to violation data, SafeTrax can record traffic data for all vehicles. This data could be used in traffic studies
and other transportation projects in the vicinity.

Both types of detectors are placed into the pavement between the warning gates and the railroad tracks, and
across both lanes of each approach. This placement should detect the common act of violation where a
vehicle travels around the gates, and the occasional violation where a vehicle travels through the gates.

The camera, which uses digital technology, imprints the time, date, and location of the violation, lane
number, vehicle type, amount of time that the warning system has been active, and frame sequence numbers
on the images it captures. There is also the option of including an integrated video monitoring system. This
is used for 24-hour recording of a crossing with a history of violations in a day. Two SafeTrax cameras can
also be installed in conjunction with each other at a crossing, for the purpose of obtaining photographs of
both the front and rear of the violating vehicles. There are areas in the U.S. where many vehicles have
license plates on one end (2).

SafeTrax relays the stored information through remote access. The system can be accessed by telephone line,
cellular communication, fiber-optic wire, or satellite link (13). Power is supplied by a 110/220 volt battery.
In case of an interruption of power, a backup battery maintains violation data for 30 days.

Econolite - AUTOSCOPE Video Vehicle Detection System

The AUTOSCOPE Video Vehicle Detection System is used in conjunction with cameras and a processing
computer system. The computer system, or “supervisor” computer, is equipped with an AUTOSCOPE
Supervisor Digitizer Board. In the application that will be discussed later in the report, these products were
supplied by U.S. Public Technologies (15).

The AUTOSCOPE system can support up to four video cameras at each crossing. To set the areas of
detection of a violation, a user can draw the detector shapes graphically on a video monitor with the use of
a mouse. The graphical images of the detectors change colors when a vehicle crosses over them, providing
a visual verification of the detection. Different types of detectors can be selected, and the placement of the
detection zones is limited by the boundaries of the field of view of the cameras (12). Changing the locations
of the detection areas involves redrawing the detector shapes on the screen. Once again, the probable
Jonathan K. Lammert Page 296

placement of the detectors will be across both lanes of each approach, and between the railroad tracks and
warning gates (11).

The cameras are connected to an input/output management device which is stored in a traffic cabinet. This
traffic cabinet is connected, typically by phone line, to the supervisor computer, which handles the citation
processing. In certain instances, an environmentally hardened supervisor computer can be installed along
with the I/O management device inside the traffic cabinet. The AUTOSCOPE detection system can
additionally perform the task of traffic data collection. Such factors that can be recorded are volume, vehicle
classification, flow rate, headway, and average vehicle speed (15). While the AUTOSCOPE system was
initially designed for intersection control, it has since been adapted for use with highway-railroad grade
crossings in the U.S.

SAIC - Traffic Violator Grade Crossing Monitor

The violation detection sub-system of the Traffic Violator system uses another method which is different
from the two products previously mentioned. When the railroad’s warning system is activated, mounted
devices which send out infrared beams are alerted. These beams are targeted across the violation area. When
the beams detect a vehicle, the high resolution video system is activated (12).

The remote video system captures the image of the vehicle, time stamps it, and compresses and transmits the
digitized form over basic phone lines to the central monitoring location (16). SAIC supplies housing and
zoom lens for the high-resolution cameras, as well as a steel pole with brackets and mounts (1).

The Traffic Violator system uses two different modes in its capturing of digitally-recorded images. The first
mode takes a wide-angle image of the crossing and a close-up image of the license plate. The second mode
takes a time-lapse wide angle view of the crossing (1). This mode is designed for crossings that have a
history of multiple violations occurring at a time, since multiple license plates can be identified on the high
resolution images. Another feature, which is designed to enhance the effectiveness of the citation, is the
inclusion of how close the train was to the crossing, or the number of seconds until impact (16). This
demonstrates the true speed of the train, which drivers tend to underestimate (8).

The computer terminal is equipped with a customizable screen (16). The system requires AC power, and
additional lighting might be needed (1). Since the video systems are accessed over basic telephone lines,
railroad officials can use the system to remotely view the condition of the grade crossing equipment. Also,
the presence of the video cameras is designed to deter vandals (17).

USPT - TRAXGUARD Automated Railroad Grade Crossing Enforcement System

The TRAXGUARD system has been used extensively throughout the world. TRAXGUARD was first used
in the Netherlands, and when the number of violations decreased dramatically, its use was expanded to other
locations in Europe (18).

Violation detection is accomplished by inductive loops in the pavement, which are placed in similar locations
to the previous loops and sensors that have been discussed. Since the loops can not measure speed, the
second photo is automatically taken after a predetermined amount of time. This amount of time has been
usually set as two seconds after the gate arms have began their descent. Both photos are superimposed with
violation data (1).

Since TRAXGUARD has been around since the late 1980's, it uses an older method of recording violations.
It uses high-speed 35-mm cameras, which come equipped with zoom lenses. A flash system is also included,
if it is deemed necessary. These cameras are housed in an environmental, bullet-proof cabinet on pre-wired,
steel observation poles. (12, 19). Field tests in Europe have typically employed the use of five empty
Jonathan K. Lammert Page 297

cabinets for every cabinet with a TRAXGUARD camera. The TRAXGUARD cabinet can be randomly
rotated among the involved crossings. This method is meant to cause uncertainty among potential crossing
violators as to which crossing is actually being monitored (18).

Unlike the other three products, the TRAXGUARD data is stored in camera itself. Film retrieval is typically
done daily during the work week. This system does not require telephone lines to be nearby. The inclusion
of early technology in the system translates to a low cost. The TRAXGUARD system’s cost has led to it
being installed at numerous crossings around the world.

Table 3 summarizes the features of the four major products.

Table 3. Features of Four AE Products

Product Violation Detection Violation Access of Data

Recording
SafeTrax RC-200 (13) Piezoelectric/ Inductive Loops Digital Remote
AUTOSCOPE (15) Video Imaging Video Imaging Remote
Traffic Violator (16) Infrared Beam Digital Remote
TRAXGUARD (12, 19) Inductive Loops 35-mm Manual

Other Products

With the increase in use of AE in the U.S. over the last decade, other companies are developing AE devices
for use at highway-railroad grade crossings. Products such as the Harmon Highway Crossing Analyzer are
new and have not been extensively tested (17). Other companies started their products in the field of traffic
light enforcement, and are adjusting them to be usable in railroad crossing applications. One of the U.S.
manufacturers that is attempting this action is Kodak. Their Digital Traffic Camera has yet to be used in a
highway-railroad grade crossing project, but it has experience in the areas of speed, toll road, and border
crossing enforcement, as well as traffic light enforcement. The camera uses special circuitry to handle
difficult lighting conditions and a 100 microsecond shuttering designed to eliminate blurry images caused
by motion (20). European companies include Gatsometer B.V. in the Netherlands, Zellweger Uster AG in
Switzerland, and Proof Digitalsystemer A/S in Denmark (14).

HIGHWAY-RAILROAD GRADE CROSSING AE PROGRAMS IN U.S.

This section contains a review of the applications of AE of highway-railroad grade crossings in the U.S. The
areas which have installed AE for the field are Los Angeles, California; Jonesboro, Arkansas; and Ames,
Iowa. Covered in this review are the reasons for the initiation of their respective AE program, technology
that was used, public information campaigns, enforcement and citation procedures, and the overall
effectiveness of the AE program. This information was gathered from an extensive literature review.

In addition, the demonstration projects in the U.S. are discussed. These were initiated by transportation
agencies in North Carolina and Texas, and the city of Miami, Oklahoma. The factors that were considered
in the tests and the results are also reviewed.
Jonathan K. Lammert Page 298

Los Angeles, California

In a survey that was conducted by the Los Angeles County Metropolitan Transportation Authority (MTA)
on residents along the 22-mile Metro Blue Line (MBL), which connects downtown Los Angeles to Long
Beach, many problems were identified (21). The identification of these problems was the first step to a major
enforcement program of the area’s crossings. The line contains more than 100 highway-railroad grade
crossings. For twelve miles, the light rail line runs through heavily-populated downtown city streets and
adjacent to Southern Pacific freight lines. Between July 1990 and January 1995 more than 250 train-vehicle
and train-pedestrian accidents occurred, causing 27 fatalities (5). It was discovered that 76 percent of the
residents who used the crossings at least once a week also stated that drivers that tried to beat trains by
driving around the gates were a serious problem.

Other responses which indicated the causes of the collisions include that it was not known that Blue Line
trains arrive at the crossing within 20 seconds of the warning lights flashing (80 percent of responding
residents), Southern Pacific’s freight trains were too slow (70 percent), and that it was not widely understood
that multiple trains could go through the intersection at the same time (21). This combination of slow and
fast-moving trains was a major contributing factor in these accidents. Reports suggested that motorists,
seeing a slow oncoming Southern Pacific train, attempted to cross the gates and would be struck by an
obscured MBL train (12).

Other factors, in regard to geometry, which contributed to the high number of accidents were reported as
(22):

• Streets running parallel to the tracks made it easier for motorists making left turns from these streets to
drive around lowered crossing gates;
• The width of most of the crossings were increased from the presence of three or four tracks, which made
it easier for motorists to drive around the lowered gates;
• Motorists’ confusion over traffic signals and signing at the intersection; and
• Unusual crossing configurations.

In response to these problems, the MTA’s Board of Directors initiated the Metro Blue Line Grade Crossing
Safety Program in March 1993. The program was formalized to implement new plans to discourage
violations at highway-railroad grade crossings. The previous year, tests had been conducted concerning
enforcement with traffic stops and AE.

For a three-month period in the summer of 1992, the MTA used a system of increased police presence and
enforcement. While the program was very successful, with 3503 citations being written for violations where
drivers had traveled around the gate arms, the costs of the demonstration was too large. The MTA needed
an easier and cheaper method of enforcement (21).

Beginning in 1992, the MTA conducted four demonstration projects involving the installation of photo
enforcement systems at four highway-railroad grade crossings along the MBL, which had proved to be an
effective tool at deterring speed and traffic light violations in the U.S. and Europe. Two of these projects
were at crossings with gates, and two were at crossings without gates. For the scope of this report, only the
results of the gated crossings will be discussed (21).

Chosen Technology

After seven vendors responded to the MTA’s advertisement, four were chosen based on their experience with
AE for traffic lights and crossings. Three accepted the invitation to participate. U.S. Public Technologies
was chosen for the two sites that featured gated crossings - Compton Boulevard and Alondra Boulevard.
Jonathan K. Lammert Page 299

Both had three tracks passing through the crossing and a parallel street on both sides. Cameras on the
southeast corner of the crossing monitored traffic in the eastbound traffic lanes (21).

The MTA absorbed all the funding for the projects related to the equipment acquisition and installation,
which was $50,000 for a camera and $10,000 for the installation onto a pole at one intersection. Signs
alerting motorists to the presence of photo enforcement were also installed. The U.S. Department of
Transportation funded an evaluation on the effectiveness of the project (1).

Citations

The dollar amounts of the citations were in accordance with the provisions of the Rail Transit Safety
Enforcement Act. This stated that the fine for not stopping at a gated crossing with signals flashing or
driving around closed gates was $104. The court was further authorized to levy an additional $100 for a first-
time violation and an additional $200 for any subsequent offense (23). Figure 1 is a violation photo taken
by the system that was used to issue a citation (21).

Effectiveness

In regard to Compton Boulevard, the first informational press conference was held on January 19, 1993.
Between this date and March 19, 1993, warning letters were sent to motorists that were detected by the
system as committing violations. From March 19, 1993 until July 19, 1993, when the study was completed,
548 violations were recorded and 232 citations were issued. Reasons for the large number of non-cited
violators were the effect of the sun’s glare in the photos, cars without frontal license plates, and license plate
numbers without a match in the Department of Motor Vehicles’ database (12).

Figure 1. Photo From AE Program in Los Angeles (21)

Jonathan K. Lammert Page 300

The project resulted in a 92 percent drop in violations, with 0.15 violations occurring per hour for the last
two months of the study. When the system was restarted for the last 21 days of September, the violation rate
had fallen to 0.07 per hour (21).

The three-month Alondra Boulevard study was completed on September 9, 1993. The equipment had been
in place for six months before the study began. Crossing violation rates fell from 0.5 per hour in December
1992 to 0.16 per hour when the project was completed. When the study had first began, the rate had fallen
to 0.28 per hour. The MTA attributed this decline to the visibility of the signs and cameras to the driving
public. During the study period, 254 violations occurred at the crossing, with 142 resulting in citations (12).

Due to the success of the Los Angeles demonstration program, AE has been expanded to seventeen crossings
along the MBL. At each crossing, poles, cabinets, and detector loops were installed. A total of ten camera
systems were purchased by the MLA, and these cameras were rotated from one crossing to another at
different intervals. In the two years since this program was started, 5,000 citations have been issued, with
a 95 percent rate of conviction. Train/vehicle collisions at gated crossings have been reduced by over 70
percent (24).

Public Awareness

In addition to the major press conference that was held on January 19, 1993, photo enforcement signs, public
service announcements, and posters were distributed before the study began.

Jonesboro, Arkansas

In 1991, officials of the city of Jonesboro, Arkansas and Burlington Northern and Santa Fe Railroad
combined efforts to install and test the first AE highway-railroad grade crossing system in the U.S. The main
reason Arkansas was selected was because the state treated highway-railroad grade crossing violations
similarly to parking violations, where a citation is mailed to the owner of the vehicle (1). A crossing on East
Highland Drive was selected. The crossing had experienced three fatal collisions during 1990, and five other
collisions in the previous five years. In addition, the gates at the crossing had to be repaired frequently. Prior
to the installation of the AE system, the average number of times the gates had to be repaired had risen to
three per week (5).

Chosen Technology

The company SAIC, which now produces the Traffic Violator system, was chosen to provide the equipment.
The Video Masters/Traffic Monitor system was purchased for $50,000 from SAIC. Video cameras were
mounted on pole brackets and encased in protective housings. Information that was collected was sent to
the police department via a standard telephone line (1,12). The infrared beam detected vehicles, and the
computer at the Jonesboro police station had storage capacity for data on 500 violators (25).

Citations

Violators received a cover letter which outlined the Arkansas statute that authorizes citations to be issued
to registered vehicle owners, the citation, and two photographs showing the offense. If convicted, the
violators were fined $100 plus the costs of the court (12).

Effectiveness

Twelve months after the system had been in place, officials reported that the gates had been knocked down
six times and ten citations had been issued (25). The violation rate was estimated to be two per month during
the time period the system was functioning (1).
Jonathan K. Lammert Page 301

City officials determined that lightning had apparently struck the system eighteen months after it was
installed. The system had been disabled by the electric surge. After the railroad repaired the system, another
malfunction occurred in 1995 which caused the system to transmit blank images. The railroad and city
officials agreed that the system should be repaired by the railroads (1). The author contacted the officials
involved and their comments on the status of the system are included in the next section.

Public Awareness

Extensive local newspaper coverage was done during and after the plan’s installation. Television companies
did report on the ribbon-cutting ceremonies. These activities were designed to heighten public awareness.
Unlike other installations, there were no warning signs to inform the motorists of the presence of the AE
system.

Ames, Iowa

Ames officials met with members of the Iowa Department of Transportation and the Federal Highway
Administration after 11 accidents occurred in 17 years at the highway-railroad grade crossing at Duff
Avenue. This crossing was close in proximity to the intersection of Duff and Main Street, and thus
experienced a heavy volume of traffic (20,000 vehicles per day). The train volume was 50 to 60 trains per
day over two tracks, with speeds of 30 to 40 mph. Three accidents over a five-year period were directly
attributed to violators driving around lowered gates. During this period of time, the AE projects in Los
Angeles and Jonesboro were reviewed (1).

Chosen Technology

The city examined the abilities of the USPT, which was in place in Los Angeles, and the SAIC system, which
was in Jonesboro. Officials used the characteristic of remote access to choose the SAIC system, and SAIC
performed most of the installation. The project cost an estimated $50,000, and the City of Ames assisted in
installing cabinets and erected utility poles to provide power to the AE system (5, 26). The SAIC Traffic
Violator system provided high-resolution images of the drivers, which was required by state law, and used
infrared beams for violation detection (5).

After the system began operation on May 1996, city officials, who had conducted demonstration tests during
the previous winter, determined the detection systems were not going to be effective. The cold weather that
Iowa experiences during these months convinced them to plan an upgrade. The infrared sensors were not
operating correctly, and the city officials decided on loop detectors. During July 1997, the system was
temporarily removed to perform the upgrade, which was completed in September 1998 (6). Figure 2
demonstrates the Duff Avenue crossing layout.

Citations

A certain traffic law negated some of the efficiency and cost-effectiveness of the AE program. An Iowa state
law requires that citations be hand-delivered to the violator’s home, instead of mailing. In addition to this
inconvenience and the fact that the gate arms had a tendency to malfunction, city officials decided that
citations would be issued only for drivers who travel under and around the gate arms after the first few
seconds of them being lowered. If the gates have been down for an excessive period of time, drivers could
travel around the gates. Also, violations were only recorded during certain time intervals (1).

Effectiveness

Due to the fact that the photographs lacked a high level of resolution, in many cases the driver was
unidentifiable. Also, it was difficult to identify nighttime violators due to the retroreflective material of the
Jonathan K. Lammert Page 302

license plates. This was the case for the 92 violations which were committed between May 1996 and June
1997. Instead of issuing citations, the city delivered warning letters to registered owners of violating vehicles
during the day. Since June 1997, 37 violations were recorded and four citations were issued, which includes
the time the system was offline (1).

Public Awareness

A public information campaign began toward the end of 1995. The campaign featured articles in local
newspapers and advertisements on television. Warning signs were also installed in locations in advance of
the crossing (12).

Figure 2. Layout of AE system at Duff Avenue (26)

Jonathan K. Lammert Page 303

Table 4 summarizes the characteristics of the three AE programs in the U.S.

Table 4. AE Programs in U.S. (1, 12)

Location No. of Date Program Subsystem Citations/Warnings

Crossings Began
Detector Recorder Access
Los Angeles 17 1992 Loops 35-mm Manual 5000+ Cit.
Jonesboro 1 1991 Infrared Digital Remote 10 Cit.
Ames 1 1996 Loops Digital Remote 4 Cit. / 92 Warn.

Norfolk Southern Corporation, NC

In 1992, the Washington D.C.-Raleigh-Charlotte rail corridor was selected as one of the five nationally-
assigned high-speed rail corridors. With the increased speed of the trains along the line, the likelihood of
tragic collisions also increased (16). Therefore, the North Carolina Department of Transportation (NCDOT)
received federal funds to develop highway-railroad grade crossing safety improvements for this corridor.
The Norfolk Southern Corporation was chosen to demonstrate video monitoring at several crossings (1).

After a 20-week study was completed to acquire an average violation rate, the data was used to formulate
improvements to the crossings. The Norfolk researchers looked at strategies that would reduce the number
of drivers that intentionally ignore crossing gates. Possible solutions included building pavement-mounted
median barriers and quad-gates, which covered all lanes in each direction of travel (27). The data was also
used to start a couple of AE programs. A SAIC video camera system has been installed at the Henderson
Road crossing in Salisbury NC to be used in conjunction with these other improvements. Violators had the
choice of either paying a fine or participating in a university study on violators (1).

Texas Department of Transportation

In 1995, the Texas Department of Transportation (TxDOT) was authorized by the state legislature to conduct
demonstration projects to install and operate AE systems at highway-railroad grade crossings. By August
1997, three crossings in Texas had been reviewed, with the product manufacturers being SAIC, USPT, and
American Traffic Systems. At one of the sites, the vendors mailed the violators educational letters. At the
other two sites, the information was provided to the local police department for further processing (1).

The researchers concluded that AE equipment could clearly be used at highway-railroad grade crossings.
However, the researchers did at times have problems identifying the license plate of the violator. This was
the cause of 40 to 50 percent of the captured images being deemed unusable. A majority of the unused
images had been captured while the front or rear plate was still out of view. They determined that the
equipment involved would need adjustments before it could be applied on a permanent basis (5).

An AE issue that was not addressed by the project researchers was the process of issuing citations and the
efforts required to implement the AE program on a large scale (5).
Jonathan K. Lammert Page 304

Miami, Oklahoma

This city modeled its plan after that of Jonesboro, Arkansas, and chose SAIC to provide the AE systems.
The characteristics of the site include a low train volume of five trains per day and 14,000 vehicles per day.
As of July 1997, the equipment has not been installed (1). The author was not able to contact the city for an
update on the project.

INFORMATION ACQUIRED FROM AE PROGRAM OFFICIALS

The author attempted to contact officials that were involved in with the AE programs and demonstration
projects in the U.S., and was successful for three of the six locations discussed in the previous section. The
officials who were contacted were:

• Allen Kuhn - Burlington/Santa Fe Railroad;

• John Waldrip - Arkansas State Highway and Transportation Department;
• Rick Brown - Iowa Department of Transportation; and
• Paul Carlson - Texas Transportation Institute.

In addition to answering a few questions concerning the AE location they were familiar with, the officials
also provided items such as crossing sketches, memos, and newspaper and magazine articles. Sample
questions that were asked are included in the Appendix.

Jonesboro, Arkansas

Burlington/Santa Fe Railroad was one of the leading proponents for installing AE at the East Highland Drive
crossing. Allen Kuhn, a Burlington public works engineer, negotiated the agreement with the City of
Jonesboro. It was decided that the railroad company would pay for and install the system. Jonesboro
provided the power source (28). Kuhn, the leading railroad official involved with the project, initially had
no concerns that the AE system would hinder the normal operation of the railroad in any way. Kuhn stated
that the AE system had never been intrusive, and believed that the railroad employees felt safer knowing the
system was in place (29).

The present status of the AE system was also determined. John Waldrip, an employee of the Arkansas
Transportation Department, stated that the system was never repaired after blank images began to be
recorded in 1995. Disagreements arose as to how the system should be handled, and it has remained offline
for the past four years (30). Kuhn noted that there had been ten violations of the crossing when the system
was operational, and was in favor of it being brought back online. Kuhn also said that the violations had all
occurred shortly after the AE system was first implemented (29). Other concerns have required the city
officials’ attention in the time since the system has gone offline.

Ames, Iowa

Rick Brown, an official of the City of Ames, provided insight to the installation at Duff Avenue. Union
Pacific was the railroad that was involved in the AE project and was in favor of the installation of the system.
Brown stated that Union Pacific did not have any major concerns or questions. A federal safety grant funded
a majority of the project, and Union Pacific contributed 10 percent of the funding. Brown confirmed that
four citations had been issued to drivers for violation of the Duff Crossing (26).

Cold weather was blamed for the malfunction of the infrared beams. This problem with the infrared beam
system caused the officials to switch to inductive loops for violation detection. The lack of high resolution
Jonathan K. Lammert Page 305

on the images was the reason for only four citations being issued, a very low rate. Brown stated that this flaw
of the system has been repaired and the AE program has been performing well (26).

Texas Department of Transportation

Paul Carlson, a research scientist for the Texas Transportation Institute, stated that there were several
problems with the TxDOT project producing photos that were precise. In many photographs, the gate arms
of the crossing were not visible. The presence of the arms was needed in order for the photo to be admissible
in court. As was mentioned in the Technology section, any photo captured with AE must demonstrate that
the warning system of the crossing was operating correctly. To correct this problem, the cameras were
reprogrammed to begin recording the violations earlier after detection. It was also necessary to reposition
several cameras (31).

More than 80 companies expressed interest in the demonstration project, and thirteen companies were
represented at the pre-proposal conference. Carlson interpreted this as demonstrating the broad market for
AE technology, and the bright future for AE applications in the U.S (31).

GENERAL GUIDELINES FOR INSTALLATION OF AE SYSTEM

This section contains an outline of the basic steps that an agency would take for an AE system to be set up
at a highway-railroad grade crossing. The focus will be on the crossing’s needs as related to geometry,
climate, and advantages pertaining to certain equipment. It was assumed that public institutions involved
in the project (city and state transportation departments, railroad companies, police departments, etc.) will
have already aligned in favor of the program and have made basic preparations (1). Also, for this example,
enabling legislation will already be in place (32).

Review of Applications in the U.S.

An in-depth study of AE programs that were either tried or that are still in place should be one of the first
actions taken. In addition to the information that is provided in this report, copies of the community meeting
discussions, AE proposals and blueprints, environmental impact statements, and estimations of costs should
be obtained. With AE programs having existed in the United States for nearly a decade, there is a large
amount of information that is available for each site.

The existing programs can be examined to find similarities between them and the crossing that is to be
targeted for installation. This can save money in a variety of ways. The right type of technology can be
chosen to best match the situation. A large amount of violations and an absence of available phone lines
could make the choice of 35-mm cameras attractive. It is possible that the cost of manually retrieving the
data could be offset by the savings of not having to install transmission wires for remote access. A large
budget and workers that are well-trained could make the newer and more expensive digital cameras the ideal
choice. If the accident history is extensive, and the agencies funding the project have enough resources, the
expensive video imaging devices could be chosen.

Substantial money can also be saved if the problems that the previous programs experienced could be
avoided. An agency planning to install AE at a highway-railroad grade crossing located in a particularly
cold climate could benefit from the knowledge that infrared beams did not function well in Ames, Iowa. A
costly system upgrade after a few years could be avoided (30). It is important to know what state laws exist
that could threaten the normal operation of the AE program. The legality of AE citations might require both
images of the vehicle’s license plate and the face of the driver (12).
Jonathan K. Lammert Page 306

Establishment of Good Relations with Public

It is important that the citizens are aware that the proposed program is in the best interests of the community,
and not a program whose goal is revenue generation. This can be accomplished by either setting the citation
amount low enough so that the program is funded through the money received in citations, or by using the
money generated for a high-profile transportation improvement project. Also, it is fundamental that the
public is made aware of the situation as soon as possible. Sudden changes in the normal flow of driving can
be irksome to motorists, so early print and TV ads are important. Advance signs warning of photo
enforcement are important.

The railroad companies that were affected by the AE programs mentioned earlier in this report were highly
involved in the installation process. Efforts by both the agency and the railroad company can be combined
to increase public awareness.

Purchase of Equipment

Advertisements are distributed to manufacturers, preferably to as many as possible, to further the chance that
the location’s exact needs can be met. In drawing up the proposals and contracts, the agency involved must
be very specific in what type of system it is looking for, and what special options and features are or are not
important. Even if care has been done in reviewing previous applications so as to not repeat mistakes, large
amounts of money can be wasted on technology that is either lacking or extraneous.

The agency must keep the nature of the community in mind. If the population tends to skew older, it could
be possible to save money in buying material that is not bullet-proof, or contains an expensive anti-vandalism
feature. If costs are intended to be cut by rotating cameras over several crossings, it is paramount to choose
a system that can be easily dismantled and moved.

System Installation

Several factors must be considered once the needed technology has been purchased and the actual physical
set-up is to begin.

The placements of the detection cameras need to grant an unobstructed view of the crossing. In other words,
the cameras need to be high enough off the ground so that they can capture photographs over the traffic not
involved in the violation. This is especially important if the crossing is in close proximity to a major
intersection.

If suitable, mountable structures exist nearby, like traffic signal posts or telephone poles, it is important that
the agency utilizes them. Money can be saved in taking advantage of previous work. The agency must be
careful, however, in not creating a large distraction for the natural flow of traffic. If no such structures exist,
it will be necessary to erect them. Steel posts have been the normal choice of past AE programs. A height
of over 12 feet should make it possible to clear most obstructions (30). Several sample photographs must
be taken to ensure that the camera location is correct if secondary images will be captured based on the
vehicle’s measured speed. Last but not least, the images must be able to demonstrate that the crossing gates
are either in, going toward, or coming up from the lowered position.

For AE systems that require piezoelectric sensors or inductive loops to be placed in the pavement, more work
is required. An effective work crew must be used up for its installation. Since the loops will usually be
placed across both lanes of traffic, proper traffic control is necessary.
Jonathan K. Lammert Page 307

Lastly, all of the components need to be connected. To save power, the violation detection subsystem will
likely be using the same train detection technology that the crossing’s warning system uses. If the subsystem
houses all of the data in a structure along the roadway, the job is slightly simplified. However, if it employs
the latest technology for remote access, a little more work is needed in coordinating the violation recording
subsystem and the citation processing subsystem.

Execution of Trial Runs

It is usually a good idea to begin the AE project as a warning mechanism, at least for the first couple of
months after installation has been completed. The agency should use the citation processing computer to
send out warning letters or educational literature to violators. This alerts the violators to a new but very
important concern, and prepares them for when the time comes to start assessing actual fines.

Since only warnings are being handed out, this is also a good time to check the system for bugs and glitches
in its operation. Since weather is one of the most important and variable factors, it would be advisable to
have the trial run cover a length of time where the crossing will see different extremes in temperature and
moisture.

Evaluation and Modification

After a certain length of time (usually about a year so that all expected variables will be experienced), it is
important to measure and quantify what effect the AE system has had on the safety of the crossing in regards
to its cost. In all of the present-day applications of AE that were studied, the numbers of violations and
collisions went down soon after the system was installed. What is undesirable is if the program is requiring
too large of a cost in relation to the light traffic volumes involved. It is also possible that other, more
pressing transportation needs have surfaced since the AE program’s inception. In such a case, a reduction
in the amount of equipment (through the rotation plan) or an increase in the fine for citations might be in
order.

APPLICATION OF AE PROGRAM TO CASE STUDY

The following section includes an analysis of the characteristics of a highway-railroad grade crossing in the
College Station, TX area. Following this analysis is a list of the guidelines that were listed in the previous
section to set up a hypothetical AE program. Choices were made based on what information was available,
and a simple sketch was made of the layout.

The twin Texas cities of Bryan and College Station have a major rail line that runs north and south through
their western half. This line runs in close proximity to Wellborn Road, a major arterial. Because of this
location, the rail line experiences a large volume of crossing traffic each day. In addition, a major rail
company runs several trains along the line each day. The highway-railroad grade crossings contain active
warning systems, and the principal devices of each system are gate arms. While there is not an extensive
history of collisions along this segment of track, the growing population of the surrounding area will continue
to increase the volume these crossings experience, and thus the number of possible accidents.

The Holleman Avenue crossing in southwestern College Station was chosen for this application. Reasons
for its selection include the presence of a major intersection nearby, Wellborn Road and Holleman Avenue,
which is similar in physical layout to the crossings in Los Angeles and Ames, Iowa, and the train volume that
it carries, which is about 25 trains a day (33). Its close proximity also granted the author convenience in its
analysis. Holleman Avenue runs east-west, and the Wellborn intersection begins about 25 feet east of the
rail line. The eastern leg of the avenue has four lanes, two in each direction, and after passing through the
Jonathan K. Lammert Page 308

intersection, the roadway is narrowed to two large lanes, one in each direction. Wellborn Road is a four lane
roadway with left turn bays at the intersection. Figure 3 illustrates the crossing (34).

Figure 3. Photograph of Holleman Avenue Railroad Crossing (34)

As can be seen from the figure, there is a short segment of roadway between the intersection and the crossing.
For the eastbound approach at the intersection, there is room for the storage of one vehicle. For additional
vehicles on this lane, it is illegal to stop on the railroad tracks (9). These vehicles must remain on the
opposite side of the railroad tracks until the traffic signal changes and they can proceed through the
intersection. Vehicles becoming “trapped” on the railroad tracks when trains arrive should not be a problem
since the intersection uses signal preemption. Signal preemption uses sensors on the railroad tracks to detect
trains approaching the crossing. These sensors alert the signal control equipment and the signal timing plan
is adjusted. Vehicles on the approach with the crossing will be granted a green light at the signal so that the
vehicles will have enough time to clear the railroad tracks. This characteristic is important to consider in the
implementation of an AE system. Whenever the violation detection subsystem is activated, it is expected
that there will not be any vehicles that are traveling away from the crossing that would be detected as
committing a violation (33).

The traffic signals at this intersection are operated under actuated signal timing plans. Actuated signals use
sensors such as inductive loops in the pavement to detect vehicles and alter the signal phases (35). There
is an inductive loop on the eastbound approach between the intersection and the crossing. It is approximately
ten feet in length, which is about half the length of the eastbound approach.

The traffic signal mast arms have been installed in each of the four corners of the intersection. Also, there
is a telephone line which runs north-south across the westbound approach.
Jonathan K. Lammert Page 309

Review of Applications in the U.S.

Since the Holleman crossing does not have a history of accidents, comparisons in this respect can not be
made with existing AE programs. The factors that can be considered are the crossing’s geometry and the
area’s climate.

After reviewing the AE programs in the U. S., the author determined that the crossing that resembled the
Holleman crossing the closest was the East Highland Drive crossing in Jonesboro, Arkansas. Both have a
single track and a similar volume of traffic, both train and vehicle. Los Angeles’ crossings all feature
multiple rail lines, and Ames’ Duff Avenue carries a volume of 20,000 vehicles a day (5). Also, Texas and
Arkansas experience similar climate due to their close proximity to one another. Therefore, equipment could
be expected to perform similarly with respect to weather in both locations.

Establishment of Good Relations With Public

The Bryan-College Station local newspaper and the Texas A&M University newspaper reach thousands of
readers a day, and news stories of the program would cause word of mouth to spread quickly. The local
television station could be invited for a broadcasted interview of the primary officials involved. A graphical
display of how the system would work would also be useful. Signs would be installed alerting the public of
the presence of photo enforcement. Figure 4 is an example of a photo enforcement sign which is used by the
Los Angeles MTA (5).

Figure 4. Photo Enforcement Warning Sign in Use in Los Angeles (5)

This sign is bilingual in nature due to the large Hispanic-speaking population of Los Angeles. It would be
determined if a similar sign could be used at the Holleman crossing, based on whether the population makeup
of College Station is similar to Los Angeles. The warning signs would be installed on each of the four
approaches, in advance of the intersection. The sign on the eastbound approach would be placed about a
hundred feet west of the crossing. The last action to garner good public relations would be to hold several
Jonathan K. Lammert Page 310

town meetings in the months leading up to the system installation to gauge the public’s reaction and answer
their questions.

Purchase of Equipment

Detectors that are placed into the pavement should be avoided, since the moderate amount of rain in the area
could increase erosion of the roadway (14). The presence of inductive loops for the purpose of traffic signal
actuation complicates the issue. One possible idea that came from a transportation professional was to use
these inductive loops for violation detection. In this scenario, the violation detection subsystem would
remain active due to the traffic signal controller requiring information pertaining to a vehicle presence.
Instead of the violation detection subsystem becoming activated with the crossing’s warning system, the
violation recording subsystem would be the device that becomes operational (6). After the traffic signal
controller is notified of a vehicle presence, this subsystem would capture the photos required.

If not for the size and placement of the inductive loop on the eastbound approach, this could be a plausible
system. A site visit was made by the author to gather information on the geometry of the crossing and the
intersection. It was discovered that the inductive loop does not extend far enough toward the crossing. It
was determined that a vehicle committing a violation by driving between the gate arms could do so without
traveling over the inductive loops. A violation could take place at the crossing without the vehicle being
detected (36). This would defeat the purpose of installing an AE system at the crossing.

Another option for using this type of violation detection would be to install a second inductive loop between
the crossing and the first inductive loop. This could, however, cause an undesirable side effect. Having two
inductive loops placed into the pavement in close proximity could decrease the strength of the roadway. It
might also hasten the rate of erosion, as was mentioned earlier.

Infrared beams would not cause any of these problems. They also should not be affected by the winter
temperature in Texas as they were in Iowa (30). For that reason, and because video imaging is too expensive
for a roadway that does not have an extremely large volume of train and vehicle traffic, the author chose
infrared beams for the violation detection subsystem and digital cameras for the violation recording
subsystem. A similar system has performed well in Ames, Iowa, and eliminates the need for visitation of
the site. Also, there is a telephone line that runs north-south across the eastern leg of Holleman which could
be used for sending the images. Two cameras would be purchased, and the equipment could be expected to
be similar to the $55,000 that Ames, Iowa spent on their initial system. The portion of the cost could also
be paid by the involved railroad company, which was the case with the Jonesboro AE system.

Both the violation detection subsystem and the violation recording subsystem could be housed in the same
cabinet. The cabinet would be made of bullet-proof metal.

System Installation

There is a traffic signal mast arm on each corner of the intersection. The mast arm on the southwestern
corner could be used to mount one of the AE system cabinets. This would detect and record violators on
the eastbound lane. Utilizing structures already in-place and not erecting a new mountable structure would
save money. The cabinet would have to be placed as far off the ground as possible, which would be near the
mast arm’s joint. This height is needed so that vehicles traveling through the intersection would not be
detected by the infrared beams.

The second camera would monitor violators in the westbound lane and could be placed on the northern side
of the Holleman leg that is located west of the crossing. This would make the erection of a steel post
necessary. The city of Ames, Iowa uses posts that are 14 feet in height, and a similar post could be
purchased. The height is designed to grant the camera clearance. Care would have to be taken in the
Jonathan K. Lammert Page 311

placement of the post. It would be necessary to consult city’s regulations in regards to sign placement and
right-of-way.

Finally, the wires that transmit the information could be ran across the length of the traffic signal arm and
be attached to the telephone wires on the poles across the intersection. These wires would eventually lead
to the College Station Police Department. Figure 5 demonstrates the set-up.

Figure 5. AE Installation for Holleman Avenue

Execution of Trial Runs

A three-month period would be used for the “warning-phase” of the program. During this time, violators of
the highway-railroad grade crossing would be mailed warnings. The police department could also include
educational literature in the form of brochures which illustrate the dangers of crossing violations. The period
would also allow the system’s engineers to identify and solve any technical problems in the system. The city
could also answer any new questions from the public. In order for the weather extremes of the area to be
experienced, the trial period would run from either February until April or September until November.

After the three-month trial period, citations would be issued to violators of the crossing.

Evaluation and Modification

After a one-year period in which citations were issued to violators, the costs and benefits of the program
would be considered. The change in safety of the crossing would be determined by comparing violation and
Jonathan K. Lammert Page 312

accident rates of the crossing before and after the AE system was installed. If the safety of the crossing has
not increased, the city should consider discontinuing the operation. It is the opinion of the author that the
system would be deemed unnecessary. The light volumes of traffic and the lack of an accident history at the
crossing would render AE an extraneous choice. Table 5 summarizes the recommendations for the AE
system at the Holleman crossing.

Table 5. Summary of Subsystems Chosen for Holleman Crossing

Subsystem Subsystem Type

Violation Detection Infrared Beams
Violation Recording Digital Cameras
Citation Processing Remote Access Via Telephone Wire

CONCLUSIONS

In conclusion, for the Holleman case study, the author reviewed the AE systems that were in place in the U.S.
In particular, the Jonesboro and Ames systems were examined due to their similarities to the crossing and
problems they experienced. It was determined that for violation detection, violation recording, and citation
processing, the agencies involved should select infrared beams, digital cameras, and remote access,
respectively. It would also be necessary to purchase and erect a 14-foot steel poll. The author speculated that
the system would eventually prove to not be cost effective. The question of cost effectiveness is the main
issue facing automated enforcement proponents in the future. Practically all tests and demonstration projects
have shown that AE is effective in reducing traffic violations. In today’s world, however, transportation
officials must make difficult choices in the allocation of funds.

Keeping the public involved in the process is essential to the support and success of an AE system. The
public must be convinced that public servants are making an effective use of an already limited supply of
time and resources.

In comparison with other transportation technologies and advances, AE is still relatively new. Continued
research and study is needed to not only validate the present uses, but to develop new and innovative
applications for AE in the future.

ACKNOWLEDGMENTS

This paper was prepared for the graduate summer course CVEN 677 Advanced Surface Transportation
Systems at Texas A&M University. Special thanks are extended to Wayne K. Kittelson and William M.
Spreitzer for their guidance and contribution of personal time throughout the program.

Also, the author would like to thank Dr. Conrad Dudek for bestowing his organizational skills and
leadership to CVEN 677. Thanks are also in order to the remaining mentors, Tom Hicks, Patrick L.
Irwin, H. Douglas Robertson, and Carol A. Zimmerman for their participation in the program. By
donating your time and knowledge, you have truly improved the learning experience of many future
transportation professionals.
Thanks also go to the following professionals who provided direction and information for this paper:
Jonathan K. Lammert Page 313

1. Paul Carlson, Texas Transportation Institute

2. John Waldrip, Arkansas State Highway and Transportation Department
3. Rick Brown, Iowa Department of Transportation
4. Allen Kuhn, Burlington/Sante Fe Railroad

Finally, I would like to thank my parents, Danny and Helen Lammert. While helping me grow as an
individual, you always remind me where my roots are. Instant support and encouragement are always one
long-distance phone call away.

REFERENCES

1. Institute of Transportation Engineers Automated Enforcement Technical Committee 96-5. Automated

Enforcement in Transportation. Informational Report, December 1998.

2. Passetti, Karl A., “Use of Automated Enforcement for Red Light Violations.” Compendium: Graduate
Student Papers on Advanced Surface Transportation Systems. Southwest Region University Center,
Texas Transportation Institute, Texas A&M University System, College Station, TX. August 1997.

3. Uniform Vehicle Code and Model Traffic Ordinances. National Committee on Uniform Traffic Laws
and Ordinances, 1992.

4. Operation Lifesaver, Inc., WWW page, [http://www.oli.org] July 1, 1999.

5. Fitzpatrick, K., R. Bartoskewitz, J. Bean, and P. Carlson. Traffic Violations at Gated Highway-Railroad
Grade Crossings. Texas Transportation Institute Research Report 2987-1, College Station, Texas,
October 1997.

6. Kittelson, Wayne K. Kittelson & Associates, Inc., OR. Meeting July 1999.

7. “Highway-Rail Crossing Consolidation and Elimination: A Public Safety Initiative.” Operation Lifesaver
Educational Literature. 1996.

8. Bartoskewitz, R.T., D.B. Fambro, and H. A. Richards, Texas Highway-Rail Intersection Field Reference
Guide. FHWA/TX-94/1273-2F. Texas Transportation Institute, Texas Department of Transportation.
May 1994.

9. “Information and Key Safety Tips at Highway-Rail Grade Crossings.” Operation Lifesaver Educational
Literature. 1997.

10. “Funding Crossroad for Grade Crossing.” Modern Railroads. February 1991, pp. 5-12.

11. Erickson, G. “Digital Traffic Cameras for Violation Enforcement.” Traffic Technology International,
Annual 1996, pp. 332-333.

12. Fitzpatrick, K., R. Bartoskewitz, and P. Carlson. Demonstration of Automated Enforcement Systems at
Selected Highway-Railroad Grade Crossings in Texas. Texas Transportation Institute Research Report
2987-2F, College Station, Texas, October 1997.

13. American Traffic Systems WWW page.[http://www.rl-200.org] June 23, 1999.

Jonathan K. Lammert Page 314

14. Blackburn, R.R., and Daniel T. Gilbert. Photographic Enforcement of Traffic Laws. National
Cooperative Highway Research Program, Synthesis 219, Transportation Research Board, 1995.

15. Econolite Control Products Inc. promotional literature, as of December 1994.

16. Science Application International Corporation promotional literature, as of August 1991.

17. “Video System Helps Norfolk Southern Evaluate Grade Crossing Barrier Devices” Progressive
Railroading. September 1995, pg. 39.

18. U.S. Public Technologies Inc. promotional literature, as of 1990.

19. U. S. Public Technologies Inc., “TRAXGUARD Rail Crossing Enforcement System.” Presentation for
the Texas Department of Transportation. 1994.

20. Kodak promotional literature, as of 1995.

21. Hubbard, L., and Linda Meadow. “Los Angeles Blue Line Enforcement Program.” 1995.

22. “Enforcement Up; Collisions Down,” Railway Age, Vol. 195, No. 4, April 1994, pp. 71-72.

23. L. Meadow and Lt. A. Frisch, “The Los Angeles Metro Blue Line Grade Crossing Demonstration
Program,” Second International Symposium on Railroad-Highway Grade Crossing Safety,
Knoxville, Tennessee, December 8-10, 1992.

24. Turner, S. and A. E. Polk. “Overview of Automated Enforcement in Transportation,” ITE Journal. Vol.
68, No. 6, June 1998, pp. 20-29.

25. LeVere, J.G., and Gary Storbeck. “Traffic Monitor System.” 1991 National Conference on Highway-
Rail Safety. 1991.

26. Telephone Interview with Rick A. Brown from Iowa Department of Transportation, July 1999.

27. Edwards, G., “NS Stops Gate Runners.” Roanoke Times, November 8, 1995.

28. “Traffic Violator System.” BN Team: Technology, Engineering, and Maintenance - Technical Bulletin.
Burlington Northen’s Technology, Engineering, and Maintenance Department, Springfield, MO. May
1991.

29. Telephone Interview with J. Allen Kuhn from Burlington Northern/Santa Fe Railroad, July 1999.

30. Telephone Interview with John Waldrip from Arkansas Department of Transportation, July 1999.

31. Interview with Paul Carlson from Texas Transportation Institute, July 1999.

32. Hoff, C., “Legal Issues Surrounding Photo-Radar Vehicle Speed Enforcement.” University of Texas
at Arlington. Spring 1995.

33. Personal interview with Matt Estes from Texas Transportation Institute, July 1999.

34. Personal photograph by Jonathan Lammert, July 1999.

Jonathan K. Lammert Page 315

35. Roess, R. P., W. R. McShane, and E. S. Prassas. Traffic Engineering. Prentice-Hall, Inc., Belmont,
California, 1993.

36. Personal interview with Jason Pierce from Powerware Solutions in College Station Texas, August 1999.

APPENDIX: SAMPLE INTERVIEW QUESTIONS

Transportation Department contacts

• How long, and where, have you been involved with the use of automated enforcement in general, and for
railroad crossings in particular?
• What was the process that was carried out in selecting AE as a potential solution?
• What were some of the problems that arose in implementing the AE system?
• In addition to history of accidents, what other factors were considered when selecting crossings for
demonstration or implementation of the AE system?
• What issues did the railroad professionals raise in you discussions with them?
• How has the public accepted the use of enforcement through the use of digital cameras?
• What changes (if any) would you like to see be done in your AE program?

Railroad contacts

• What were some of the concerns you had when the issue of AE came up?
• In what way (if any) has the use of AE affected the normal operation of the railroad system?

JONATHAN K. LAMMERT

Jonathan K. Lammert received his B.S. in Civil Engineering

from Texas A&M University in December 1998. Jonathan is
currently pursuing his M.S. in Civil Engineering at Texas A&M
University. He is currently employed as a Graduate Research
Assistant in the Systems Implementation program at the Texas
Transportation Institute (TTI). He also worked for TTI as an
undergraduate student worker in the Traffic Management and
Information Systems program. Jonathan worked three summers
for the Texas Department of Transportation, performing duties
such as inspecting construction projects and surveying. Jonathan
has been active in the Texas A&M student chapters of the
American Society of Civil Engineers and the Institute of
Transportation Engineers, where he is currently serving as the
Chapter Librarian and Webmaster. His areas of interest include traffic operation and safety.