DOT/FAA/AM-02/5 Automatic Dependent Surveillance-

Broadcast / Cockpit Display of

Office of Aerospace Medicine
Traffic Information: Innovations in
Washington, DC 20591 Pilot-Managed Departures

O. Veronika Prinzo
Civil Aerospace Medical Institute
Federal Aviation Administration
Oklahoma City, OK 73125

April 2002

Final Report

This document is available to the public

through the National Technical Information
Service, Springfield, VA 22161.
 N O T I C E

This document is disseminated under the sponsorship of

the U.S. Department of Transportation in the interest of
information exchange. The United States Government
assumes no liability for the contents thereof.
 Technical Report Documentation Page
1. Report No. 2. Government Accession No. 3. Recipient's Catalog No.
DOT/FAA/AM-02/5

4. Title and Subtitle 5. Report Date

Automatic Dependent Surveillance -Broadcast Cockpit Display of Traffic April 2002

Information: Innovations in Pilot-Managed Departures
6. Performing Organization Code
7. Author(s) 8. Performing Organization Report
No.
Prinzo, O.V.

9. Performing Organization Name and Address 10. Work Unit No. (TRAIS)
FAA Civil Aerospace Medical Institute
P.O. Box 25082
Oklahoma City, OK 73125 11. Contract or Grant No.

12. Sponsoring Agency name and Address 13. Type of Report and Period
Covered
Office of Aerospace Medicine
Federal Aviation Administration
800 Independence Ave., S.W.
Washington, D.C. 20591 14. Sponsoring Agency Code

15. Supplemental Notes

This work was performed under Task AM-B-00-HRR-516
16. Abstract:
Avionics devices that provide a cockpit display of traffic information (CDTI) enable pilots to acquire, verify, and
maintain pre-defined spacing intervals from other aircraft. It is of interest to the Federal Aviation Administration to
determine how the use of these displays influences pilot/controller operational communications. The operational
evaluation of the CDTI in October 2000 provided an opportunity to examine some of these issues. Departure
profiles (13 without CDTI, 32 with CDTI) were established to evaluate the ability of flight crews and air traffic
controllers to manage long (6 nm) and short (4.5 nm) spacing intervals between departing aircraft during 3 day and
2 night operations. Subject-matter experts evaluated 15 hours of verbatim transcripts and audiotapes for benefits
that may have resulted from pilot use of the ADS-B/CDTI. They also evaluated these communications for
communication problems associated with misidentified aircraft, confusions, uncertainties, and operational
concerns. Computed for each departure was the time the aircraft was under local control, runway ownership time,
and the number and duration of messages. Approximately 4% of the departures conducted when CDTI was not
used, and 9% when it was, involved communication problems. In particular, for pilot messages during CDTI
departures, aircraft call signs were misstated or incorrect. However, when CDTI was not used, controller messages
included the correct flight identifier but the wrong company name. In addition, more messages were exchanged
during the short spacing interval. In contrast, when pilots executed CDTI departures, there was an overall increase
in time on frequency, more time was spent under local control during the day, and departures were completed in
less time when assigned the short spacing interval. Communication problems were associated with the call sign
procedure that was designed to distinguish between aircraft being talked to versus talked about. Fortunately,
controllers and pilots detected and corrected these problems in mid-stream, and statistically, communication
efficiency was not affected. In light of the findings and comments from the controllers and pilots, alternative call
sign procedures will be constructed and evaluated for the departure spacing application.
17. Key Words 18. Distribution Statement
Pilot Communication Document is available to the public through the
ATC Communication National Technical Information Service
Air Traffic Control Springfield, Virginia 22161
19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Price
Unclassified Unclassified 14
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized

i
 AUTOMATIC DEPENDENT SURVEILLANCE - BROADCAST / COCKPIT DISPLAY OF
TRAFFIC INFORMATION: INNOVATIONS IN PILOT-MANAGED DEPARTURES
“You on the cutting edge of technology
have already made yesterday’s impossibilities
the commonplace realities of today”
— President Ronald Reagan,
White House, February 12, 1985

INTRODUCTION and the tail of the immediately departing aircraft

using a CDTI. Once the spacing interval specified in
The Automatic Dependent Surveillance-Broadcast the PDC was achieved, the pilot would begin the take-
(ADS-B)1 is a data link application that will transmit off roll and the next aircraft would taxi into position
from aircraft and vehicles equipped to emit surveil- and hold. Under varying weather conditions, differ-
lance data (such as position and identification) by ent aircraft types or sequences (e.g., a B727 behind a
means of a broadcast-mode to users who can receive B757), and runway layouts, different spacing inter-
these data. The Federal Aviation Administration (FAA) vals might be desirable. The second operational evalu-
anticipates that ADS-B will provide many benefits — ation of CDTI (OpEval-2) occurred in October 2000
from extending the range and coverage of current and provided an opportunity to examine some of
ground-based secondary surveillance radar (especially these issues.
around airport surfaces) to increasing air-to-air situ- The stated purposes of OpEval-2 were to develop
ational awareness. The FAA envisions that enhanced and evaluate specific ADS-B air-air and air-ground
airborne and ground capabilities would provide for applications, to evaluate controller use of ADS-B, and
specific surveillance functions that would encourage to demonstrate ADS-B technology.2 Both the pilot and
co-operative air traffic management and collaborative controller based concepts of departure spacing/clear-
decision-making among its users (Prinzo 2001). ance were assessed. Before proceeding any further, it
ADS-B may also offer some innovative technology must be pointed out, as it was during OpEval-1, that
that could remove the delays attributed to low-ceil- OpEval-2 provided an opportunity to demonstrate
ing/reduced visibility weather that currently restricts new air- and ground-based capabilities and systems at
departure operations. For example, with improved an FAA-controlled airport tower and terminal radar
airport surface surveillance, ADS-B supported avion- approach control facility. With that in mind, it is
ics may provide synthetic or virtual digital-imaging of important to note that it was not possible to apply true
other aircraft, vehicles, or obstacles to the operator experimental and control conditions comparable to
that would not be visible otherwise. This capability that of laboratory-based experiments. In addition, the
could potentially enhance safety. training that pilots and controllers received was un-
Avionics devices that provide a cockpit display of even — making it impossible to exercise any statistical
traffic information (CDTI) enable pilots to acquire, control over the data (e.g., treating the number of
verify, and maintain pre-defined spacing intervals hours of training as a co-variant).
from other ADS-B equipped aircraft. It is of interest However, OpEval-2 provided an opportunity to
to the FAA to determine how the use of these displays gather field data that could be used to guide the
influences pilot/controller operational communica- development of ADS-B applications (e.g., approach
tions. For example, a pre-determined spacing interval spacing, departure spacing/clearance, runway and fi-
could be included as part of an aircraft’s (i.e., Ownship) nal approach occupancy awareness, and airport sur-
pre-departure clearance (PDC). Following route taxi face situational awareness). During that five-day event,
instructions, the pilot could taxi into position and objective and subjective data were collected from
monitor the distance between the nose of Ownship controllers and pilots, and a comprehensive report of

1
In 1998, RTCA SC-186 completed Minimum Aviation System Performance Standards for ADS-B. In addition to describing ADS-
B, this document also provides information on applications that may use ADS-B information.
2
This was cited in the Flight Crew Mission Guide that was developed for OpEval-2 by the Operational Evaluation Coordination
Group (OCG 2001).
1
the findings was prepared by the Operational Evalu- simulations, pilots were instructed on how to respond
ation Coordination Group (OCG, 2001). Joseph, to various types of ATC messages and listened to a
Domino, Battiste, and Bone (under review) provide a combination of ATC and pseudo-pilot communica-
summary of the subjective flightdeck observer data. tions over a party line. They were to abide by the same
Reported here is a summary of several analyses per- requirement for pilot-managed and ATC controlled
formed on the audiotaped communications between departure spacing. That is, at brake release, pilots were
pilots flying aircraft equipped with a CDTI device to say “MARK” on the tower radio frequency for data
and local controllers who provided them with air collection purposes.
traffic services. The objective of the voice tape analysis Experimental OpEval-2 Departures. During the pre-
was to identify any change in operational communica- flight briefings at the Air Guard (for pilots) and Tower
tions, workload, or both that resulted when pilots were (for controllers), all of the participants reviewed the
flying with and without the benefit of CDTI. This report flight scenarios and scripts. The facilitators reminded
provides a general description of the communication them to follow the established procedures and commu-
findings for the departure spacing application. nication protocols. Afterwards, the pilots proceeded to
their respective aircraft and the controllers went to their
METHOD air traffic control positions in the tower or radar room.
For each flight period, the departure profiles were
Participants set up to evaluate the ability of the flight crew and
Sixteen pilots, serving as a captain or first officer, controllers to manage a pre-determined spacing inter-
flew aircraft equipped with CDTI while a local and val between departing aircraft. Typically, pilots taxied
ground controller along with a coordinator served as their aircraft along the assigned routes, held short of
the OpEval-2 tower team who provided them with air the active runway, and, following the receipt of the
traffic control (ATC) services. The pilot participants departure clearance flew a pre-determined pattern,
were paid volunteers who received briefings and par- landed the aircraft, taxied to the runway and departed
ticipated in proficiency training exercises prior to the again. The sequence of aircraft in the taxi pattern
evaluation. The controllers, also volunteers, were on varied for each departure. Enumerated below are the
a temporary detail during training and on a regular procedures used to attain the OpEval-2 goals and
schedule during the evaluation. ensure compliance with standard ATC procedures.
For the purpose of this study, the departure spacing
Materials interval was the distance between a pair of aircraft
The Louisville (Standiford) International Airport when the leading aircraft began its take-off roll down
(SDF) Radar Approach Control (TRACON) facility the runway. Each departure scenario defined the spac-
provided five, digital audiotapes (DAT), one for each ing interval as 6 nautical miles (nm, long) or 4.5 nm
test period. Each DAT contained separate voice records (short) between each pair of departing aircraft. Pilots
of all the transmissions made to the radio frequency managed the spacing interval between all but the final
assigned to the Ground East, Local West, or Final aircraft in the departure sequence. The local control-
Radar West position on the left channel. The right ler managed that aircraft.
channel contained the Universal Time Coordinated Pilot-Managed Departure Spacing Interval. All par-
(UTC) time code expressed in date, hour (h), minute ticipants used scenario cards that defined the departure
(min), and whole second (s). The NiceLogger™ Digi- spacing interval to be achieved for the flight period. As
tal Voice Reproducer System (DVRS) decoded and shown in Figure 1, once the lead aircraft was 6000 ft
displayed time and correlated it with the voice stream above the runway (Point A), local control provided a
in real time. The data consisted of 15 hours of digi- take-off clearance to the aircraft that was holding short of
tized voice communications of which 6 hours were the active runway using standard phraseology and com-
from the Local West position. munication practices. The pilot positioned the aircraft
onto the runway. By using information displayed on the
Procedure CDTI (e.g., Ownship was 2.5 nm from the departing
Training on the Departure Spacing Application. aircraft and it was at R=6000 feet elevation), the pilot
Before OpEval-2, pilots and controllers participated determined when the scripted distance (either 4.5 nm or
in several pre-OpEval-2 simulations conducted at the 6 nm) would be achieved and then began the aircraft’s
Integration and Interaction Laboratory (I-Lab) of the take-off roll down the runway. Local control protected
MITRE Corporation Center for Advanced Aviation the runway during the time the flight crew delayed the
System Development (CAASD). During these take-off roll.
2
 CDTI Display

2.5 Interval Criteria (4.5 or 6 nm)

R = 6.0

✈
A

✈
2.5nm 6000 ft

Figure 1. Departure Spacing Interval Criteria between Participating Aircraft.

Speaker MESSAGE
N123 1. ACADEMY TOWER NOVEMBER ONE TWENTY-THREE' S HOLDING IN TURN
BEHIND NOVEMBER FOUR FIFTY-SIX
ATC 2. NOVEMBER ONE TWENTY-THREE {NAME} TOWER ROGER
ATC 3. NOVEMBER ONE TWENTY-THREE RUNWAY ONE SEVEN RIGHT TURN RIGHT
HEADING ONE NINER ZERO CLEARED FOR TAKE-OFF
N123 4. CLEARED FOR TAKE-OFF NOVEMBER ONE TWENTY-THREE
N123 5. NOVEMBER ONE TWENTY-THREE MARK
ATC 6. NOVEMBER ONE TWENTY-THREE CONTACT DEPARTURE
N123 7. NOVEMBER ONE TWENTY-THREE

Figure 2. Example of a Pilot-Managed Departure Communication Set.

Speaker MESSAGE
N123 1. ACADEMY TOWER NOVEMBER ONE TWENTY-THREE ZERO READY FOR TAKE-
OFF HERE ONE SEVEN RIGHT AT BRAVO
ATC 2. NOVEMBER ONE TWENTY-THREE {NAME} TOWER RUNWAY ONE SEVEN
RIGHT TAXI INTO POSITION AND HOLD TRAFFIC STARTING SIX MILE FINAL
N123 3. POSITION AND HOLD ONE TWENTY-THREE
ATC 4. NOVEMBER ONE TWENTY-THREE RUNWAY ONE SEVEN RIGHT CLEARED FOR
TAKE-OFF TURN RIGHT HEADING TWO FIVE ZERO WIND ONE EIGHT ZERO AT
FIVE
N123 5. CLEARED FOR TAKE-OFF NOVEMBER ONE TWENTY-THREE AND TWO FIVE
ZERO ON THE HEADING
ATC 6. NOVEMBER ONE TWENTY-THREE CONTACT DEPARTURE
N123 7. OVER TO DEPARTURE

Figure 3. Example of a Controller-Managed Departure Communication Set.

3
 As shown in Figure 2, a pilot-managed departure departing aircraft during 3 day (flight periods 1, 2,
communication set generally began when a pilot ini- and 4) and 2 night (flight periods 3 and 5) operations.
tiated the initial call up, as illustrated by message 1. The departure spacing application was performed by
When the runway was clear the controller issued the the pilots with the use of ADS-B distance information
take-off clearance, and in some instances, issued a displayed on a CDTI or by the Local West controllers
vector, or climb-out-instruction as was the case in without benefit of ADS-B information. Forty-five
message 3. The pilot taxied onto the runway and upon departures began in initial contact and ended in a
release of the break, responded with “MARK” (or hand-off to the departure controller. Analyses were
similar words such as “Rolling”) as in message 5. The performed on 32 pilot-executed departures conducted
inclusion of “MARK” indicated that the aircraft had with ADS-B/CDTI and 13 controller-managed de-
begun its acceleration down the runway. The commu- partures conducted without ADS-B/CDTI.
nication set ended when the local controller instructed Dependent Variables. To measure changes in
the pilot to contact departure as illustrated by message 6. workload and operational communication, (the pri-
Controller-Managed Departure Spacing Interval. mary variables of interest), the communications be-
Again, as with the pilot-managed departures, for the tween local control and each aircraft for each departure
controller-managed departures all of the participants were grouped into departure communication sets.
followed the scripted scenarios that defined the to-be- Presented in Figure 4 is an example of one of the pilot-
achieved departure spacing interval for the flight managed departure communication sets. Measures of
period. As shown in Figure 3, a controller-managed workload included efficiency of communications and
departure communication set generally began when duration measures for communication sets. Measures
the pilot initiated the initial call up, as illustrated by of operational communication included communica-
message 1. The local controller determined when to tion problems and operational concerns. These mea-
instruct the pilot to “taxi into position and hold.” sures were designed to provide the cost benefits
Then, when in the controller’s judgment the antici- subgroup of the OCG with a metric to help estimate
pated scripted distance of 4.5 or 6 nm would be the operational impact and the benefits of CDTI.
achieved, the take-off clearance was issued using stan- They are described below.
dard phraseology and communications procedures. Efficiency of Communications. Less time spent on
Similarly, the controller-managed communication set frequency coupled with fewer departure-related trans-
ended when the local controller instructed the pilot to missions to perform the departure spacing task may
contact departure, as illustrated by message 6. reflect improved efficiency in operational communi-
cation and a reduction in objective workload. Thus,
Experimental Design efficiency of communications included the number of
The TRACON and tower were divided into two messages in a communication set that contained de-
sections, with the West portion of the airspace dedi- parture information and the duration of each of those
cated to OpEval-2. In addition, a portion of the calls. The duration of individual calls was the time
airfield was set apart from normal operations and the spent on frequency (TOF) communicating. As shown
tower controllers limited access to the West runway to in the example in Figure 4, TOF for the first message
participating aircraft. The OpEval-2 flight periods was 2s.
were scheduled during normally low airport activity. Duration Measures for Communication Set. Addi-
The participating aircraft made 69 departures that tional measures of objective workload were frequency
resulted in 54 departure pairs, of which two were lost occupancy time (FOT), runway ownership time
to equipment failure and one to a loss of data. (ROT), and the amount of time the aircraft was under
This study used a two-factor within-subjects de- local control (TLC). They were computed for each
sign. The within-subjects factors were Time of Day communication set that began with initial contact
(Day, Night) and Spacing Interval (Short, Long). A and ended in transfer of communication to the depar-
between-groups comparison (Pilots, Controllers) was ture controller. Frequency occupancy time was the
not attempted since there were more pilot participants sum of all of the TOF for each communication set and
than controllers, and pilots had access to ADS-B in the example, FOT was 12s (FOT = S TOF =
information and controllers did not. 2+2+4+2+0+2+0).
Independent Variables. Departure profiles were ROT was computed as the time lapsed from the
established to evaluate the ability of flight crews and onset of a message by local control that included a
air traffic controllers to manage the 31 long (6 nm) take-off clearance to the onset of a message by the
and 20 short (4.5 nm) spacing intervals between controller for transfer of communication to the depar-
4
ture controller. In the example shown in Figure 4, communication” (Morrow, Lee, and Rodvold, 1990
151s lapsed from the issuance of the take-off clearance pp. 36). Communication problems included inaccu-
in message 3 (078s) to the transfer of communications racies, procedural deviations, and non-routine trans-
in message 6 (229s). During this time, the runway is actions that involved misunderstandings or other
active and, therefore, unavailable to local control for problems related to successful information transfer.
other vehicle movement or aircraft operations (run- The use of the traffic-flight identifier in traffic-
way crossing, landings, etc.). It is during this time that related messages by controllers during the Initial/
workload increases: For the local controller, additional Final Approach Spacing Application and Visual Ac-
effort is required to scan the airport surfaces for a quisition Evaluation may have encouraged pilots to
potential runway incursion when there is a delay in include it as part of their responses to those messages.
the take-off roll. For pilots, monitoring the supple- Pilots who received messages in the form of “NO-
mentary CDTI display may add workload while per- VEMBER ONE TWENTY-THREE TRAFFIC
forming routine station-keeping tasks and preparing TWELVE O’CLOCK … NOVEMBER FOUR
for take-off. Runway ownership time, as a by-product FIFTY-SIX” may have responded with … “NOVEM-
of ADS-B/CDTI use for pilot-managed departures BER FOUR FIFTY-SIX IN SIGHT. ” In fact, of the
task, was included as a measure of objective workload. 889 pilot responses to traffic-related messages, 45%
For each aircraft, the total time under local control included either the full (362) or partial (40) call sign
(TLC) was computed as the time lapsed from the of the aircraft called as traffic (Operational Evaluation
onset of the pilot’s initial call-up in message 1 (at Coordination Group, 2001). Accordingly, inclusion
007s) to the closing of the transaction in message 7 (at of the traffic-flight identifier may have inadvertently
232s). In the example TLC was 225s (232s-007s). migrated into pilots’ communications with the local
Communication Problems. SMEs identified com- controller during the departure spacing application.
munication problems as “any disturbance of routine Consequently, use of the traffic-flight identifier may
communication, where controllers and pilots do not have contributed to communication problems and
follow standard procedures, and/or where they must added workload.
interrupt information transfer in order to clarify the

Time (in seconds)

Speaker MESSAGE Start End TOF ROT TLC
N123 1. AND GOOD EVENING TOWER NOVEMBER ONE 007 009 2 151 225
TWENTY-THREE IS IN SEQUENCE
ATC 2. GOOD EVENING NOVEMBER ONE TWENTY-THREE 011 013 2
ACADEMY TOWER ROGER NUMBER TWO
ATC 3. NOVEMBER ONE TWENTY-THREE ACADEMY 078 082 4
TOWER RUNWAY THREE FIVE LEFT FLY HEADING
THREE THREE ZERO CLEARED FOR TAKE-OFF
N123 4. CLEARED FOR TAKE-OFF THREE FIVE LEFT THREE 083 085 2
THREE ZERO NOVEMBER ONE TWENTY-THREE
N123 5. MARK 176 176 0
ATC 6. NOVEMBER ONE TWENTY-THREE CONTACT 229 231 2
DEPARTURE
N123 7. NOVEMBER ONE TWENTY-THREE 232 232 0

LEGEND: TOF = Time on frequency per transmission

ROT = Runway occupancy time per aircraft
TLC = Time under local control per aircraft

Figure 4. Example of a Departure Communication Set.

5
 Operational Concerns. SMEs also identified and Hendrix, 1995). Each message was preceded by its
categorized operational concerns into the “Traffic onset and offset time represented in hour (HH) minute
Advisory”, “Position Operation”, or both categories. (MM) and seconds (SS) and was followed by a column
For example, when pilots reported seeing traffic on to record comments and a column to code operational
the CDTI but not out-the-window, SMEs encoded concerns. Unlike the identifiers “N123” and “N456”
these reports as a concern that the pilots may not be that are presented in Figure 5, true aircraft identifiers
maintaining compliance with traffic advisory proce- and flight numbers were included as part of the
dures. Similarly, SMEs encoded pilot self-separation OpEval-2 transcripts for all the aircraft that were
from traffic as a concern in flight deck operation. As present on the DAT.
SMEs listened to the audiotapes and read the tran- Training Subject Matter Experts. The lead SME
scripts, they encoded their concerns and provided provided the other SMEs with 16 hours of training on
brief comments. the data encoding process to achieve consistency and
In summary, message counts, contents, and dura- conformity in identifying communication transac-
tion were the objectively derived measures of workload tions and evaluating the accuracy of content. Since
and communication extracted from the time-stamped OpEval-2 imposed minor operational constraints,
voice tapes. They were used to compute descriptive SMEs received instruction on how to evaluate com-
statistics expressed as means (M) and standard devia- munications in light of those modifications. Further-
tions (SD) that summarized CDTI use versus non-use more, the lead SME encouraged the other SMEs to
on operational efficiency between ATC and the par- direct their attention to the detection and codifica-
ticipating flight crews. Operational communication tion of any benefits that may have occurred from the
in the form of phraseology, communication prob- pilots having access to a CDTI. In addition to com-
lems, and operational concerns provided some in- menting on positive outcomes, they also were to
sights and implications for future air traffic operations, comment on any situation involving a potential or
workload, and communications procedures. real loss of separation or situation awareness, misun-
derstanding, or communication problem (missed
Data Extraction, Training, and Data Encoding readback of the identifiers, routes, altitudes, etc.).
Procedures Finally, the SMEs received instruction on how to select
This section begins with a description of the quali- and enter their codes onto the OpEval-2 ATSAT-CF.
fications of the Subject Matter Experts (SMEs), con- Data Encoding. Once taught, each SME received a
tinues with the data extraction procedures and is complete set of audiocassettes, transcripts, and the
followed by the procedures used by the lead SME to code and instructional manual. The audiotapes and
train the other SMEs. The section ends with an transcripts aided the SMEs in the identification of
explanation of the data encoding process. departure clearance (DC) communication sets. Each
Qualifications of the Subject Matter Experts. The SME worked independently to identify and code the
lead Air Traffic Subject-Matter Expert was an instru- efficiency and accuracy of communications. However,
ment-rated pilot and former controller who had worked they met on a weekly basis to discuss their encoding and
as a FAA Academy instructor for 8 years and for 12 resolve any differences. Once the SMEs reached consen-
years in FAA supervision and management. Two sus, the data entry clerks received the final copy and it was
additional air traffic SMEs had been instructors (Ter- entered into the database for final analysis.
minal Option) at the FAA Academy in Oklahoma
City. The pilot subject matter expert was a recently RESULTS
retired airline pilot with 31 years of experience. Prior
to serving as a SME, the pilot’s duties included in- Changes in operational communication that may
structor on the B-727 and DC-8 aircraft; check airman have resulted from ADS-B/CDTI during OpEval-2
on the DC-9 aircraft; pilot of the CV-880, DC-8, B- were evaluated from verbatim transcripts and digi-
727, DC-9, L-1011, B-757, and B-767 aircraft. tized voice recordings provided by the TRACON
Data Extraction Procedures. Five sets of audiocas- facility. Although requests for the inclusion of baseline
sette tapes were dubbed from each digital audio tape circuits during OpEval-2 were made during the plan-
(DAT). The transcribers used one copy to generate ning of the event, none were conducted. Conse-
five sets of verbatim transcripts, and each message was quently, routine and OpEval-2 operational
typed onto an electronic copy of the Aviation Topic communications could not be compared since the
Speech Act Taxonomy-Coding Form (ATSAT-CF) baseline data necessary for comparison were not in-
like the one presented in Figure 5 (Prinzo, Britton, & cluded as part of OpEval-2.
6
LINE HH MM SS HH MM SS Speaker Receiver MESSAGE COMMENT CODE
1 16 5 46 16 5 49 N123 ATC ONE SEVEN RIGHT POSITION
HOLD FOR NOVEMBER ONE
TWENTY-THREE
2 16 6 38 16 6 40 ATC N456 NOVEMBER FOUR FIFTY-SIX
CONTACT DEPARTURE
3 16 6 40 16 6 41 N456 ATC FOUR FIFTY-SIX
4 16 7 3 16 7 6 ATC N123 NOVEMBER ONE TWENTY-
THREE RUNWAY ONE SEVEN
RIGHT TURN RIGHT HEADING
ONE NINER ZERO CLEARED
FOR TAKE-OFF
5 16 7 7 16 7 11 N123 ATC CLEARED FOR TAKE-OFF
RUNWAY ONE SEVEN RIGHT
HEADING ONE NINER ZERO
NOVEMBER ONE TWENTY-
THREE
6 16 8 28 16 8 30 ATC N123 NOVEMBER ONE TWENTY-
THREE CONTACT DEPARTURE
7 16 8 31 16 8 31 N123 ATC ONE TWENTY-THREE

Figure 5. Example of a Transcript.

Table 1. Pilot-Managed Departures Executed with ADS-B/CDTI

Time (in seconds)
Source (a) N Messages (b) TOF (c) FOT (d) ROT (e) TLC N Sets

Time of Day
Day 6.38 (2.83) 2.47 (.64) 15.44 (7.06)* 135.81 (22.04) 201.00 (097.38)* 16
Night 5.13 (1.31) 2.19 (.53) 10.75 (2.02) 146.00 (12.26) 168.50 (045.24) 16
Spacing Interval
Short Spacing 5.69 (1.70) 2.51 (.68) 14.12 (6.05) 128.81 (18.39)* 170.81 (080.46) 16
Long Spacing 5.81 (2.76) 2.15 (.44) 12.06 (5.17) 153.00 (06.43) 198.69 (072.14) 16

* statistically significant at p ≤ 05

Table 2. Controller-Managed Departures Executed without ADS-B/CDTI

Time (in seconds)
Source (a) N Messages (b) TOF (c) FOT (d) ROT (e) TLC N Sets

Time of Day
Day 7.57 (1.72) 2.34 (.11) 17.71 (4.07) 60.14 (10.29) 213.29 (118.42) 7
Night 6.83 (1.72) 2.21 (.58) 14.67 (3.20) 77.17 (39.89) 184.00 (050.97) 6
Spacing Interval
Short Spacing 8.67 (1.37)* 2.14 (.28) 18.67 (4.32) 58.17 (06.05) 224.50 (128.58) 6
Long Spacing 6.00 (0.58) 2.40 (.45) 14.29 (2.06) 76.43 (37.12) 178.57 (041.68) 7

* statistically significant at p ≤ 05
7
 The local controller and pilots on the flight deck time (FOT); (d) runway ownership time (ROT); and
exchanged 278 messages. These messages comprised (e) total time that the aircraft was under local control
45 departure clearance (DC) communication sets (TLC). As shown in Table 1 for the time of day factor,
(one for each departure) that began with the pilot the results indicated that for pilot-managed depar-
checking in (e.g., initial contact) and ended with a tures using ADS-B/CDTI, in addition to an increase
pilot acknowledging the hand-off (e.g., transfer of in radio frequency occupancy time per departure
communications to departure control). For all practi- (column c) [F (1,28)=4.95], longer periods of time
cal purposes, the number of DC communication sets were spent under local control (column e) during day
was approximately equal for day (n=23) and night but not night operations [F (1,28)=4.24]. For the
(n=22) departures using a short or long spacing inter- spacing interval factor, flight crews executed the take-
val. Analyses were performed for pilot executed depar- off clearance in less time (column d) when they were
tures conducted with ADS-B/CDTI (32 departures) on a short, rather than long, spacing interval [F (1,28)
and for the controller-managed departures that were = 19.26]. The absence of any significant interaction
conducted without ADS-B/CDTI (13 departures). between spacing interval and time of day for the pilot-
Efficiency of Communications. As mentioned ear- managed departures suggests that spacing interval
lier, the measures of communication efficiency for the alone accounted for the more than 20-second savings
departure communication sets were (a) number of in transfer from local to departure control.
messages (N messages) and (b) mean time on fre- Displayed in Table 2 are the means and standard
quency per message (TOF). To evaluate the effects of deviations of the duration measures for controller-
CDTI on communication efficiency during depar- managed departures conducted without ADS-B/
ture spacing, a Time of Day (Day, Night) by Spacing CDTI. The results of the ANOVA revealed no statis-
Interval (Short, Long) Analysis of Variance (ANOVA) tically significant differences in workload for the time
was performed on the local control communication- of day factor, the spacing interval factor, or their
data for Flight Periods 1-5. The results were evaluated interaction.
using a criterion level set to p =.05. Presented in Table Communication Problems. Using transcripts and
1 and Table 2 are the mean (M) and standard devia- cassette tapes, SMEs identified communication prob-
tion (SD) for each dependent variable. lems such as inaccuracies, procedural deviations, and
The results of the separate ANOVAs revealed that non-routine transactions involving misunderstand-
when pilots executed ADS-B/CDTI departures, it did ings or other problems related to information trans-
not matter whether they flew with a short or long fer. Illustrated in Figure 6 are several examples. While
spacing interval [F(3,28)=2.40] or if they flew during these types of problems can contribute to frequency
the day or night [F(3,28)=2.37] (see Table 1 columns congestion and increase workload, they do not neces-
a and b). When ADS-B/CDTI departures were con- sarily lead to operational errors or incidents.
ducted, pilots were instructed to say “Mark” as an An evaluation of the 278 messages involving the
indication that they were starting to move down the departure spacing application for ADS-B/CDTI re-
runway. These additional transmissions may have vealed that approximately 2% contained communica-
added to the number of messages transmitted and tion problems involving five departures. When ADS-B/
inflated the time data. However, upon re-analysis, the CDTI was in use, there were two departures in which the
results did not change and only decreased the M and aircraft call sign was either misstated or incorrect and one
SD by a fraction of a second. departure in which the pilot requested a “say again.”
When ADS-B/CDTI was not in use, as was the case As seen in Figure 6, the controller detected a
for controller-managed departures, the results revealed communication problem of either a misspoken call
an increase in workload for controllers as indicated by sign or the potential for a stolen clearance. The prob-
more messages exchanged between local controller lem was quickly resolved through the exchange of two
and the pilots on the flight deck during the short (but additional messages. In another departure clearance
not long) spacing interval [F(1,12)=16.54] (see Table readback, the pilot transposed the assigned heading
2 column a). Time of day did not exert a statistically with the numbers in the call sign, quickly discovered
significant effect on controller workload when mea- the problem and restarted the readback. “NOVEM-
sured by either the number of messages exchanged or BER THREE TEN OR UH NOVEMBER ONE
their mean duration (Table 2 columns a and b). TWO THREE TO THREE TEN ON A HEADING
Duration Measures for Departure Clearance Com- CLEARED FOR TAKE-OFF THREE FIVE LEFT.”
munication Set. The duration measures of DC Com- The last communication problem involved a pilot
munication Sets were c) total frequency occupancy request for the controller to repeat the preceding
8
 ADSB IN USE
Speaker Incorrect Call Sign
1. NOVEMBER ONE TWENTY-THREE … CLEARED FOR TAKEOFF
PILOT 2. … CLEARED FOR TAKEOFF … NOVEMBER ONE TWENTY-FIVE
CONTROLLER 3. AND VERIFY THAT WAS ONE TWENTY-THREE
PILOT 4. UH NOVEMBER ONE TWO THREE UH CLEARED FOR TAKEOFF …
Call Sign Midstream Correction
CONTROLLER 1. NOVEMBER ONE TWENTY-THREE … TAXI POSITION AND HOLD
PILOT 2. POSITION HOLD … NOVEMBER ONE TWENTY-THREE
CONTROLLER 3. NOVEMBER ONE TWENTY-THREE … TURN LEFT HEADING THREE ONE
ZERO CLEARED FOR TAKEOFF
PILOT 4. NOVEMBER THREE TEN OR THREE ONE THREE TO THREE TEN ON A
HEADING CLEARED FOR TAKEOFF …
Say Again
CONTROLLER 1. NOVEMBER ONE TWENTY-THREE RUNWAY
PILOT 2. SAY AGAIN FOR NOVEMBER ONE TWENTY-THREE
CONTROLLER 3. NOVEMBER ONE TWENTY-THREE DISREGARD HOLD SHORT …
PILOT 4. HOLDING SHORT …
ADS-B NOT IN USE
Speaker Call Sign Midstream Correction
CONTROLLER 1. NOVEMBER ONE TWENTY-THREE … CORRECTION DECEMBER ONE
TWENTY-THREE … TAXI INTO POSITION AND HOLD
PILOT 2. POSITION AND HOLD … DECEMBER ONE TWENTY-THREE
Call Sign Midstream Correction
CONTROLLER 1. DECEMBER ONE TWENTY-THREE … TAXI INTO POSITION AND HOLD
PILOT 2. POSITION AND HOLD … DECEMBER ONE TWENTY-THREE
CONTROLLER 3. NOVEMBER CORRECTION DECEMBER ONE TWENTY-THREE … CLEARED
FOR TAKEOFF
PILOT 4. CLEARED FOR TAKEOFF … DECEMBER ONE TWENTY-THREE

Figure 6. Examples of Communication Problems and Their Resolutions

9
transmission. When ADS-B/CDTI was not in use, of issuing take-off clearances that could specify vari-
there were several occasions when the controller referred able wake-vortex minimums between aircraft equipped
to the aircraft with the correct flight numbers but with with ADS-B/CDTI. With the availability of an ADS-
the wrong company name. Upon self-discovery of the B/CDTI system onboard their aircraft, pilots could
problem, the controllers restated these transmissions accept these clearances and begin their take-off roll at
using the correct call sign as shown in the lower half of these prescribed minimums. The distance (or time
Figure 6. No additional transmissions were required to interval) used to space aircraft for departures vary with
clarify the misstated call signs since they were identified the category of aircraft (Category I, Category II, or
and corrected midstream by the controllers. Category III), same runway versus intersecting or
parallel runways, the potential for wake turbulence,
DISCUSSION and other factors. 3 ADS-B/CDTI systems developers
could include these parameters as part of their algo-
The analysis of voice communications from the de- rithms and further optimize runway use with the
parture spacing application suggests that neither the added benefit of increased safety.
pilots who used the ADS-B/CDTI nor the controllers At noted earlier, controllers used standard phrase-
who applied current ATC procedures without the ben- ology to perform the departure spacing task (e.g.,
efit of ADS-B information experienced a notable in- cleared for take-off) without specifying the pre-deter-
crease in workload. In particular, when pilots executed mined spacing interval to be established or main-
the ADS-B/CDTI assisted departures, neither the num- tained between aircraft. For some of the other
ber nor duration of the messages exchanged between the applications that were evaluated, the controllers’ in-
flight deck and local control resulted in inefficient clusion of traffic flight identifiers was in accordance
communications. In fact, the procedural changes insti- with FAA Order 7110.65, where “additional infor-
tuted for the departure spacing application eliminated mation” is allowed at the end of traffic information
the instruction for the pilot to “taxi into position and messages.4 The controllers received flight strips that
hold, ” without a decrease in total radio frequency denoted the type of ADS-B equipage installed on the
occupancy time per departure. Unfortunately, OpEval- participating aircraft. Not surprisingly, on several
2 did not include conditions that would provide com- occasions, pilot-use of the traffic flight identifier
parisons between this procedure and those used currently migrated to their communications with the local
at the local control position. Consequently, a need exists controller. For example, “TOWER {Ownship’s}
for research to document the implications of eliminating HOLDING IN TURN BEHIND NOVEMBER
the “position and hold” procedure on aircraft movement ONE TWENTY-THREE, ” and “TOWER
and airport capacity. {Ownship’s} READY BEHIND ONE NINER.”
The findings from the duration measures indicated During preparation for this event, facility, regional and
that pilots succeeded in using ADS-B/CDTI infor- headquarters personnel focused on developing a phraseol-
mation to adjust their departure spacing interval from ogy and procedural environment that would address the
4.5 to 6.0nm. Importantly, these specific spacing issues noted in OpEval-1. This included a prototype phrase-
distances, as defined in the scripts and on pilot flight ology that included the call signs of participating aircraft in
cards, were well coordinated and orchestrated. In fact, all traffic calls and met the basic requirements of the
controllers managed the traffic in a manner that fully operational environment. Unfortunately, a comprehensive
supported the local operation and the flight period evaluation of the operational concept for the departure
spacing requirements to the extent possible. At times spacing application and phraseology was not attainable
the operational priorities restricted the ability of the during this operational event. Not surprisingly, significant
controllers to allow full conformance to scripted pat- work remains to address phraseology issues and safety
terns, planned spacing intervals, or both. Fortunately, concerns with flight crews occupying a runway for an
these occurrences were minimal. extended time.
Once the tower is provided with a display that The lack of standardization in voice communica-
includes ADS-B information, the departure spacing tions procedures was one of the operational concerns
application will provide local controllers with a means noted by the subject matter experts. To realize the full

3
See FAA Order 7110.65M Air Traffic Control Section 8, Spacing and Sequencing for FAA authorized air traffic control
procedures and phraseology for use by personnel providing air traffic control services.
4
FAA Order 7110.65M Air Traffic Control Section 4 Para. 2-4-20 Aircraft Identification NOTE: “Air carrier and other civil
aircraft having FAA authorized call signs may be pronounced using single digits if necessary for clarity.”
10
benefits from ADS-B/CDTI, a more concise lexicon Guidance for Initial Implementation of Cockpit Display of
of air traffic phraseology would be helpful. A consid- Traffic Information. Washington DC: RTCA, Inc.
eration of a data link for more routine communica- (February 1998). (Prepared by SC-186). RTCA/
tions, similar to what occurs for pre-departure RTCA/DO-243.
clearances, would help keep the voice channel avail-
Minimum Aviation System Performance Standards for
able should problems occur. Although pilots are en-
Automatic Dependent Surveillance Broadcast (ADS-
couraged to use the phraseology outlined in the
B). Washington DC: RTCA, Inc. (February 1998).
Aeronautical Information Manual (FAA 2001) and
(Prepared by SC-186). RTCA/RTCA/DO-242.
FAA Order 7110.65 The Handbook of Air Traffic
Control (FAA 2000), they are not required to use the Joseph, K.M., Domino, D. Battiste, V., and Bone, R.
phraseology. Air traffic controllers and pilots would (in progress). A Summary of Flightdeck Observer
benefit from precise, consistent, and standardized Data from SafeFlight 21 OpEval-2. Federal Avia-
communications. Communication capability, pilot tion Administration, Office of Aerospace Medi-
and controller workload, and system capacity all ben- cine Technical Report, Washington, DC.
efit from concise, standardized phraseology. Minimum Aviation System Performance Standards for
In addition to the phraseology and communication Automatic Dependent Surveillance Broadcast (ADS-
analysis, the subject matter experts expressed an op- B). Washington DC: RTCA, Inc. (February 1998).
erational concern that centered on the installation and (Prepared by SC-186). RTCA/DO-242.
commissioning of ADS-B at various airports. In par-
ticular, the pilot SME felt strongly that operational, Morrow, D.G., Lee, A., and Rodvold, M. (1990).
procedural, and human factors considerations involv- Analysis of routine pilot-controller communi-
ing ADS-B/CDTI use for pilot-managed departures cation. In: Managing the Modern Cockpit: Third
task and its affects on airport capacity, airport surface Human Error Avoidance Techniques Conference
movement, and safety need a thorough evaluation. In Proceedings. Warrendale PA: Society of Auto-
particular, research is needed to determine the affect motive Engineers, Inc.
of increased “runway ownership time” on system Operational Evaluation Coordination Group (2000).
safety and airport capacity. Specifically, runway own- CAA/FAA/Safeflight21 Phase 1 - Operational
ership time represents the time the flight crew expends Evaluation Final Report. www.faa.gov/safeflight21/
maneuvering the aircraft onto the runway, determin- orv/opeval_1/index.html
ing when the specified distance is achieved between
Ownship and the preceding aircraft, and executing Operational Evaluation Coordination Group (2001).
the take-off clearance. During this time, the runway is CAA/FAA/Safeflight21 Operational Evaluation-2
active and, therefore, unavailable to the controller for Final Report. www.faa.gov/safeflight21/orv/
other vehicle movement or aircraft operations (e.g., opeval_2/index.html
runway crossing and landing). Prinzo, O.V. (2001). Innovations in Pilot Visual Acquisi-
More research is required to determine the affect of tion of Traffic: New Phraseology for Air Traffic
increased “runway ownership time” on airport move- Control Operational Communication. Federal Avia-
ment, capacity, and safety during departure spacing. tion Administration, Office of Aerospace Medi-
If the ADS-B/CDTI departure spacing application is cine Technical Report DOT/FAA/AM-01/19,
implemented then a clear understanding is needed of Washington, DC.
the roles, responsibilities, procedures, and phraseol-
ogy for pilots and controllers who use it. Prinzo, O.V., Britton, T.W., and Hendrix, A.M. (1995).
Development of a Coding Form for Approach Con-
REFERENCES trol/Pilot Voice Communications. Federal Aviation
Federal Aviation Administration. (2000). Air Traffic Administration, Office of Aerospace Medicine
Control (7110.65M). Washington DC: U.S. Gov- Technical Report DOT/FAA/AM-95/15, Wash-
ernment Printing Office. ington, DC.

Federal Aviation Administration. (2001). Aeronautical

Information Manual. Washington DC. U.S. Gov-
ernment Printing Office.

11
12