NISTIR 5825 NIST Construction Automation Program Report No. 1 Non-Line-of-Sight (NLS) Construction Metrology William C, Stone February 1996 Building and Fire Research Laboratory National Institute of Standards and Technology Gaithersburg, Maryland 20899 ws we U.S. Department of Commerce Ronald H. Brown, Secretary Technology Administration Mary L. Good, Under Secretary for Technology National Institute of Standards and Technology Arati Prabhakar, DirectorABSTRACT This paper addresses the subject of automated metrology (surveying) for use on construction sites. Specifically, the research is directed to the development of a novel Non-Line-of-Sight (NLS) system with which the real-time position and orientation (attitude) of any object on a construction jobsite may be determined, irrespective of the presence of intervening obstacles that would otherwise render optical and/or electro-optical techniques useless. Tests were conducted using a specially configured broad-band, low-frequency spread-spectrum radar. The transmission and receiving antennae, which in normal radar are typically one and the same, were physically separated so as to create a system with a fixed broadcast unit and a “roving” receiver, whose range was to be determined relative to the transmission antenna by means of time-of-arrival measurements. Time domain response was synthesized by means of fourier theory from a broad spectrum of data sampled in the frequency domain. Numerous field experiments were performed in which typical construction site obstacles were placed between the transmitter and receiver with separation distances of up to 80 meters. The obstacles included a half-meter thick, heavily reinforced concrete wall, varying combinations of masonry block and brick up to more than a meter in thickness, and metal pre- fabricated wall panels. In all but the latter case, repeatable distances were obtained . Range detection was lost in the presence of extensive metal panels that contained no windows. However, the presence of even small openings permitted range acquisition. Sources of error, limits of resolution and accuracy, and factors affecting time of flight measurement are discussed. KEYWORDS: Construction automation, dielectric constant, diffraction, metrology, multipath, NLS, non-line-of-sight, penetration capacity, positioning system, propagation delay, spread spectrum radar, surveying.Table of Contents Chapter One : Introduction: ‘The Automated Construction Site 1.1.1 Real Time Metrology and Sensing Needs 2 Data Telemetry. 3 Virtual Site / Simulator Standards 3 User-in-the-Loop Feedback... 3 Construction Robotics. LL. LL. ae LL Chapter Two: A Review of Present Electronic Positioning Technologies. 2:1 Elestronie Distance Measurement (EDM). 2.2 GPS-Based Metrology 2.2.1 GPS Ranging Modes . 2.2.2 Kinematic GPS... 2.23 Brief Summary of GPS Field Tests Conducted at NIST 2.3 Pseudolites .. 2.4 Fanning Lasers 2.5 Summary .. Chapter 3: Impulse Radar Tutorial: 3.1 Basic Radar Principles 3.2 Why single frequency systems cannot be used for Distncing 3.3 Dual-Frequency Radars .. 3.4 Pulsed Radar 3.5 Pulse compression techniques 3.6 Operating Principles from Fourier Theory. 3.7 Developing a Practical NLS Survey System Chapter 4: Surveying Through a 500 mm Reinforced Concrete Wall. 4.1 Test Description. 4.1.1 Radar Hard) 4.1.2 Test Wall. 4.2 Test Procedure 4.3 Resulls.. Chapter 5: Surveying Through Building Wall: 5.1 Test Description (Building 202) 5.2 Test Procedure . 5.3 Results . 5.3.1 Masonry Biock and Brick Wall Tests. 5.3.3 Mixed Obstacle Tests... 5.34 Reaction Wall Penetration Tests Chapter 6: Electromagnetic Wave Propagation Through Engineering Materials 6.1 Introduction .. 6.2 Propagation Through Nonconducting 633 Discussion of Building 226 “Wall” Tests.. 6:4 Discussion of Building 202 Tests .. 6.5 Recommended Future ResearchChapter One: Introduction: Surveying has traditionally been defined as the science and art of determining relative positions of points on, above, or beneath the surface of the earth, or establishing such points. It is one of the oldest and most important arts, stemming from the need to mark boundaries and divide land. In modern structural and civil engineering, surveying is an essential element required to plan, construct, and maintain highways, railroads, rapid-transit systems, buildings, bridges, tunnels, canals, dams, drainage works, and land subdivisions, to name but a few. Such surveys will generally fall under categories of establishing local topography and land boundaries prior to commencement of construction; in-progress assessments conducted during actual construction to provide control elevations, horizontal positions, dimensions, and configurations; and as-built data to provide exact final location and layout of engineering works, their positional verification, and records that include design changes and retrofits. The surveying process, as practiced at a typical construction site generally involves the following steps: + Research and analysis: selecting the survey method and equipment. + Field work or data acquisition: making measurements and recording data in the field. + Computing or data processing: performing calculations based on the recorded data to determine locations, areas, volumes, and so on. + Mapping or data representation: plotting measurements or computed values to produce a map, or chart, or portraying the data in a numerical or computer format. * Stakeout: setting monuments and stakes to delineate boundaries or guide construction operations. The past 15 years have seen significant improvements in surveying technology, the most important being the development of electronic distancing equipment and the integration of these devices into "total stations", which contain onboard microcontrollers which automatically reduce inclination, azimuth, and distance to provide cartesian coordinates of a survey target relative to the survey instrument. New technologies that are just now beginning to see limited use on construction sites include Differential GPS (satellite based positioning relative to an orbiting constellation of time-encoded transmitters) and "fanning" laser systems operating in the ultraviolet range of the electromagnetic spectrum. These latter technologies possess the important attribute of being capable of providing nearly instantaneous, or "real time’, position information. 1HThe advantages and disadvantages of each of these systems will be described below. However, all share the common limitation of requiring a direct line of sight between the instrument and the point for which the position is desired. At a typical construction site, where optically opaque walls, beams, girders, slabs, and ceilings are constructed as a matter of course, component obstructions will constitute a severe impediment to real-time surveying. In the context of present day construction, line of sight surveying instruments meet the existing needs. However, the traditionally conservative construction industry has long been identified as one which is overdue to reap the benefits of “information age” technology. Global competition is now forcing a reconsideration of this status quo. Unlike the manufacturing and shipping industries, construction has been unsuccessful at delivering improved value over time. The potential benefits from improvement in performance of the construction process are enormous, given the scale of the industry and its impact ‘on national and global economies. The construction process itself must be re- designed as a system if technology innovation is to be effectively exploited by the US. construction industry. Construction automation has been proposed as a central enabler to the betier use of limited global resources (capital, human and material) and therefore has a key role in achieving development which is environmentally sustainable. Innovation through construction automation is also central to competitiveness of the industry. Construction automation addresses the following problems: decreasing worker productivity; poor working conditions; injuries to workers; decreasing desirability of building and construction jobs; declining quality of the constructed project; over-run schedules; high project costs; decreasing skill level of workers; hazardous materials or site conditions (both known and unexpected); manual handling of heavy and awkward assemblies; and time consuming inspection, repair and maintenance. Construction automation embodies a systems engineering approach to hardware, software and interface technologies, as well as to condition assessment and real-time site metrology. Accurate and flexible simulation tools for the construction process, in all its dimensions, are central to solving the problem of performance of the constructed facility over its entire life cycle. In the design stage , simulations of constructability and maintainability, as well as performance under natural and manmade environments during the facility's use, must be undertaken. Simulation includes performance of construction equipment; environmental factors and impacts; and behavior of human participants. Thus, simulations of construction and maintenance or renovation activities are essential to design as they are to selection of optimal alternative construction processes. 12Within the above construction automation context the problem distills to one of knowing where everything is at all times. This requirement for global position knowledge when components, machines, and people move about outside and within the constructed facility poses metrology problems which exceed present day capabilities. What is needed is a metrology system -- a “Non-Line-of Sight” (NLS) survey system -- in which position at a construction site may be determined irrespective of the presence of intervening obstacles. It is the objective of the research described below to enable the development of such a metrology system. The Automated Construction Site In 1994, NIST began a research initiative in construction automation. The broad- based objectives of this program are to develop standards to permit the construction industry to achieve the benefits of large scale automation and systems integration that have heretofore been possible only in controlled factory environments. Several architectures have been proposed for automated integration of construction site tasks (see Figs. 1.1.1, 1.1.2, and 1.13). Common features depicted in the above architectures which are necessary to permit one-of-a-kind, large-scale manufacturing at highly varied construction sites include: * Sensors for Real-Time Construction Site Metrology © Wide Band Telemetry and Data Acquisition © Virtual Site Simulation and Object Representation Standards * Person-in-Loop Systems Feedback * Construction Robotics As suggested in Fig. 1.1.4 (BFRL pyramid slide), these technologies may be viewed of as a pyramid, with each successively more sophisticated capability being dependent on those beneath it. Described below are these five topical areas, identified by NIST as necessary to achieve a global closure of the construction automation process from real-time data acquisition at the job site through the return of that information, in a processed format, in numerous useful ways on- demand. 13AUTOMATION IN CONSTRUCTION - BFRL ROLE Tasks: Sheet [saa = tdontity || Chips + Location + atrbutes obo Central Management & ide + Ment a Control 1 Gapubitis + Assignments "Simulation ‘Models Central Database -———= Electronic Representation Intelligent Retrieval Figure 1.1.1: Early BERL vision of open architecture for construction automation. Key facets include metrology; wireless communications; global interactive databases; worker feedback, training, and teleoperation systems; and semi-automated and fully automated machinery.Virtual Robot Tools Trucks Robotic Intelligence ‘As Built Database Modifications: Other Equipment Supervisory Control Safety Coordination Communications) orners: Material Open Architecture Robotics Figure 1.1.2: Construction automation open architecture proposed by Cybo Robotics.Figure 1.1.9: NIST Vision for construction automation. Jobsite wireless communications links workers and 10 the loca contruction shes which isin tur inked tothe information superhighway ‘Via‘a minimum Tt data ink (15 Mbits/sec), Automated routers permit design office engineers and lanes, maniac ob shops and fabricators and reguaory ators susstbe project {o various portions ofthis dataflow. Workers and foremen at he construction ste wil be able to ‘ascertain n reav-time the status and location of raw materials and fabricated components.Uietts el ACE ier tae Beng Decry Construction Automation Hierarchy Figure 1.1.4: Technological underpinnings of construction automation. Realtime measurement metrology) forms the foundation of al automation. Wireless jobsite communications involving potentialy hundreds, to thousands of data channel is required to uplink information in a fashion which ls unintrusive toon site operations. Viral modeling and the develonment ofa real-ime global database is needed to store all project data, including that uplinked from the jobsite as well as tht from design offices and fabricator, {nd fo make this availabe to all autnorized subscribers ina format which provides realistic assessment ata glance. User feedback systems and traning systems include helmetmounted displays (HMDs) and hhead-up displays (HUDs) both for data interrogation, machinery operation, and machinery simulation ‘Semi-automated and eventually fuly autorrated (robotic) machinery will bgin to see effective use at |b stas only after tne fst four foundation levels have been effecively implemented.11.1 Real Time Metrology and Sensing Needs The fundamental problem with any construction site is that it is continually in a state of semi-controlled chaos. One reason for this is the lack of instantaneous knowledge about what needs to be done next, where it needs to be done, and where the people and/or machines, materials and tools are that are needed to do it. Thus the underpinning, as suggested by Fig. 1.1.4, is real-time metrology. Before addressing how one might conduct such surveying, it will be useful to establish how much accuracy is needed. The required accuracy depends primarily on the purpose of the survey. For horizontal and vertical control surveys standards have been established by the Federal Geodetic Control Committee [Berry, 1976], as shown in Table 1.1.1. Accuracy is expressed as a relative fraction, for example, 1 part in 100,000 for a First Order survey. At a typical construction site, where survey distances realistically are less than 100 m, First Order accuracy amounts toa +/- 1mm error band ; Third Order, Class I accuracy would be +/-10 mm over the same distance. Table 1.1.1: Horizontal Control Survey Accuracy Standards “Order and Relative Accuracy ‘Accuracy on 100 mi Class Between Adjacent Survey Line Points First Order 1 part in 100,000 1mm Second Order: Class 1 1 part in 50,000 2mm Class Ti 1 part in 20,000 5mm Third Order Class I 1 part in 10,000 10 mm Class I 1 part in 5,000 20 mm Typically, Third Order, Class II accuracy is specified for local construction site work and is sufficient for grade line control for earth moving, where +/- 20 mm over 100 m (1 part in 5,000) is commonly specified for plane control of, for example, paved parking lots [Allegheny Excavating, 1995]. Building stakeouts typically involve accuracies of 1 part in 10,000 [Dewberry & Davis, 1995]. First order surveys are typically commissioned only for the setting of control points where position and/or alignment are critical. For example, the Washington DC Metro requires First Order accuracy for all rail lines. 14For the placement of steel framework the American Institute of Steel Construction [AISC, 1980] provides the position and alignment tolerances shown in Table 1.1.2. Table 1.1.2: Steel Construction Position and Alignment Tolerances Task Relative Accuracy ‘Accuracy on 100m Between Adjacent Points Survey Line Horizontal Alignment of 1 part in 2400 42 mm Exterior Columns Vertical Alignment of 1 part in 2000 50 mm Exterior Columns (inter-story slope) Vertical Alignment of 1 part in 864 116 mm Exterior Columns (total deviation from column line) More stringent accuracies are required by AISC for the installation of anchor bolts and embedded items. Table 1.1.3 lists typical requirements for these types of connections. Table 1.1.3: Installation of Anchor Bolts and Embedded Items Category ‘Absolute Error Tolerance Center-to-Center_of any 32 mm ‘two bolts within a group Center -to-Center of 64 mm adjacent bolt groups Elevation of tops of bolts +/-12.7 mm Deviation of bolt group Centerline 6.4 mm from Column Line The absolute error limits established in Table 1.1.3 represent a class of common construction related assembly operations, and similar values could be prescribed, for example, in the error tolerance that might be acceptable for an automated robot designed to place structural steel in a high rise structure. From the above discussion it is apparent that a graded accuracy scale might be developed for general automation at a construction site. Many items that will require position monitoring, such as the location of supplies, equipment, and personnel, do not require as great a level of accuracy. Table 1.1.4 presents what the author feels to 1sbe a representative first cut at establishing standard metrology needs for the automated construction workplace: Table 1.1.4: NIST Proposed Working Model for Automated Construction Site Metrology Acura: Category ‘Absolute Error Tolerance (over 100 m radius) Vehicle position +7 500 mm (remote reporting) Personnel (remote reporting) +/- 500 mm Material stockpiles and discrete +/- 500 mm components within those stockpiles (remote reporting) General earthmoving +/- 100 mm (coarse machine control) Structural alignment +/-50 mm (remote reporting) Final Grade earthmoving +/-20mm (machine control) Final component position +/-10 mm (remote reporting) Component Placement +/-5 mm Full Registration Coarse Alignment Component Placement +/- 2mm Full Registration ‘ine Alignment There are several emerging technologies which are poised to address portions of the accuracy spectrum described in the above table. All of these technologies, described in greater detail in Section 1.2, have the common requirement of an unobstructed line-of-sight between the survey instrument and the target, whose position is desired. In addition to position acquisition (three degrees of translational freedom), there will also be a requirement in the automated construction workplace for several 16Sensors For Real-Time Construction Site Metrology Figure 1.1.5: Sensors for Real-Time Construction Site Metrology. In order to track and eventually place machinery under semi-autonomous and eventually fully autonomous control the position and attitude of the vehicle must be known at ail times. Additionally, for collision avoidance, the status of ail other kinematic degrees of freedom must also be known, such as the angles associated with ariiculated arms. The fist six degrees of freecom (vehicle position and attitude) can be determined by several approaches including differential phase GPS (which requires no additional jobsite metrology equipment) or by an array of onsite pseudolites, fanning laser, of NLS (nor-tine-of-sight) surveying systems now under developement, These extemal reference metrology systems may be aided by auxiliary onboard IMUs {inertial measurement units). The development ot generic, low cost “strap-down' units for vehicle tracking will be essential to fully realize the Potential of construction automation.other sensed characteristics, including attitude (three degrees of rotational freedom); joint rotations, actuator and/or cable extensions, and various strains, loads, pressures, and temperatures which are needed to completely describe the state of machinery, vechicles, and robotic hardware (Figure 1.1.5). Furthermore, a means for uniquely identifying each item that can move about, including machines, people, components and raw materials must be developed. Acquisition of this identification code, which might take the form of a bar code or “smart chip,” will be part of the sensing technology that will one-day be integral with the positioning system. From the view point of “real-time” operations, such position, state, and identification actions will need to take place at sufficient update frequencies to permit remote operation (teleoperation) of vehicles and realistic virtual representations of the site to be created at remote offices. ‘A simple example of the identification task is: suppose one has a pile of rebar on the construction site. A construction foreman, or worker, needs to know + what that rebar is... is it a #4 bar, a # 6 bar etc. © where is it located right now and © where is it supposed to go. The engineer and planner, on the other hand, would like to know © is it Grade 60? © what mill did it come from? © what is its stress-strain data © when was it manufactured © is it supposed to be epoxy coated or bare? ete. It is assumed that there will be some type of bar code or smart chip identifier for every significant item entering a construction site (Figure 1.1.6). It would not be unreasonable for a deformed bar manufacturer to automatically stamp a bar code in each bar as it comes off the mill. What that code would mean, and what its format should be remains to be defined. The same approach could be adapted to wide flange sections and other standard construction materials and components. Although initially an inspector will be responsible for keeping track of scanning codes, counting components and acquiring positions, this task will eventually be done by erection crews as they do their job. NIST has chosen to concentrate initially on development of standards for identification systems for construction site components, including familiar items like rebar and rolled sections and eventually proceeding onward to complete kinematic descriptions of vehicles that would allow manufacturers to easily enter their machinery into a standard “7SIRI EA Cutlie-late| Sedilelllsie cect lelaiele| Figure 1.1.6: Real-Time Material and Component Positioning: In addition to vehicle tracking, the ‘position of all construction materials and constructed components must be known. This permits near instantaneous knowledge of the as-built status of the structure as well as provides information to construction foremen and workers of the location of hardware and materials. Unilke vehicle tracking, Q/A workers will be tasked with tracking onsite movements of these items. Bar codes ane matt chips will permit component recognition but the location mus be spatially derived by touching key points (pre-marked) on components which will be referenced in the glok database, so that the location and orientation of each component (once installed) is known. ‘As shown above, one possibility for acquiring key point positions is though the use of an articulated seanning arm connected to a roving backpack receiver. The backpack will use the same positioning systems described earlier for vehicle tracking and the articulated arm will account for the offset relative to a sensing element directed by the worker to the component keypoints.identification system that would permit menu-select entry of that item into a virtual world model of a construction site. 1.1.2 Data Telemetry It is one thing to acquire bar codes and positions. One can presume that for a large construction site that will eventually see controlled vehicles and tracked positions of every moving object, that there will be hundreds if not thousands of such position and identification transactions occurring each second. In order for any sense to be made of this data it must reach a global database to which remote and local queries can be made. The transfer of data must be accomplished via telemetry (wireless communications), although for a few specific cases fiber optic lines can be considered. This massive transfer of data from independent sources will require novel approaches to prevent radio frequency chaos and the intervention of the FCC (Federal Communications Commission) that might prevent or hamper implementation. Key issues to be resolved include: + Interfacing hundreds of on site positioning systems with a global jobsite database * Maximizing real time data reliability (inter-city construction will involve transmission from within a congested location and competetion with similar transmitters at nearby jobsites). + Federal communications laws. © Data security © Cost It is anticipated that many techniques developed during the past ten years for battlefield command and control will be brought to bear on this issue. 1.1.3 Virtual Site / Simulator Standards The third step in closing the automation loop is the development of simulations and simulation tools to model the performance of individual pieces of construction equipment, eg,, earth moving equipment or cranes, and other objects which move about the construction site either under their own means of locomotion (eg. human workers) or materials and components which may, from time to time, be moved by either of the afforementioned agents. These simulations and software tools will permit users to determine key characteristics of construction equipment and processes, eg. the productive output of the equipment and the response, or failure, of the equipment to its environmental factors such as weather and soil conditions. Manufacturers and commercial software companies will develop the "commercial" simulation capabilities, but 18Virtual Site Simulation and Object Representation Standards Figure 1.1.7: Virtual Site Simulation: Data telemetered from the construction site will be used to update full three-dimensional kinematically realistic representations of the items and machine: which exist at the at the construction site, in addition to the present site geometry (which may be modified by the actions of machines and workers). Any node connected to the information superhighway and possessing the proper authorization (see Figure 1.1.3) can subscribe to the global database and view either numerical or three dimensional renderings of the status of the site from any perspective. Standard formats for storing and displaying component and machine: representations remains to be developed.will make use of common standards for data exchange, thus facilitating inter- operability and flexibility. Federal R&D will focus on fundamental and enabling research needed, eg. dynamic properties of soil, nonlinear response of mechanical systems, vision/user interface technologies, control algorithms, automated site metrology/telemetry, data exchange standards (e.g. for kinematic virtual models), and planning/reasoning techniques. Simulations will be developed of the construction process at various levels of detail. Critical path scheduling is simulated at a high level of abstraction, whereas simulation of mechanical tasks (for example, back hoe loading of a truck) is at a very detailed level. Eventually, there should be capability to move through a hierarchy of planning detail for the construction process. Commercial organizations are expected to provide the "commercial" versions of these simulations, but construction contractors, equipment manufacturers and academic and federal researchers will be principals in the development and verification of algorithms, database standards, and proof-of-concept demonstrations through virtual construction test-beds. Construction process simulation (Figure 117) includes simulations of interactions with the environment and people; static and dynamic, linear and non-linear response of materials and systems to forces, displacements and temperatures; behavior of equipment considering its position and attitude, its flexibilities and degrees of freedom, component inertias, environment, control systems and operator interventions; and gaming-type models for effects of operations with time. Simulations of maintenance, renovation, and demolition are within this scope provided that removals as well as placement of materials and components are simulated. Federal R&D can facilitate development of effective, neutral standards in the form of guide specifications to manufacturers of the physical systems and equipment that will utilize automation, advanced metrology and sensing capabilities and the software/data interface standards that will link all these disparate sub-systems. These standards and guides will act as a framework, evolving under industry consensus and ‘technological innovation, that equipment manufacturers can use to ensure acceptance of new and innovative equipment and systems. Standards will support robust and reliable equipment which will yield attractive production rates and high quality construction. Three dimensional graphics (including virtual reality technologies) will be a key aspect of the simulation tools, so that construction workers, managers and engineers can understand the reasons for problems in the construction process and easily identify improvements. Non-graphic analytical and reasoning techniques also will be needed to identify opportunities and needs for improvements and recommend changes in construction processes. Commercial software companies will be involved in developing and marketing construction 19simulations, with research activities involving research organizations and contractors. 1.1.4 User-in-the-Loop Feedback Once one has position and identification, what can be done with the information? One promising use would be to enter it into a 3D immersive simulator that would permit a management level project planner to assess what is happening at the site and where things are. Establishing such a virtual environment is a big job and will, by nature, involve many iterations in which concepts shown to work on testbed simulators are put before industry for common concensus of practicality. However, once agreed upon, the impact of this system will carry all the way to the indivdual hardhat worker on the jobsite who will, on demand, have interactive access to this information.. Human factors and display research will be essential to close the loop and bring the database information back to the construction site worker and foreman where it can be put to practical use in the form of selected access to registered 3D as well as alphanumeric data, updated in realtime, on the jobsite. The development of lightweight, robust, inexpensive helmet mounted displays (HMDs) and head-up displays (HUDs) are crucial to this process [Chinnock, 1995], as are human factors design to determine methods of displaying data, while avoiding information overload, eye strain, and disorientation typical of such systems today. Voice activation and recognition technologies, high capacity, compact, local storage media, and inter-worker communications will also be crucial technologies to this effort. It is presumed that Federal R&D will develop the initial proof-of-concept technologies, with the private sector subsequently manufacturing commercial versions of these systems. One variant of this technology envisioned at NIST is that of a Head-up Construction Database Interrogator: a flip down, see-through visor that mounts to a standard construction helmet (Figures 1.1.8 and 1.1.9). A hands-off voice activated menu system would come up which would permit the worker to identify the section of the jobsite he is working at and call up a blueprint if there is a question concerning what goes there. If the question really is: "where are the 135-degree bent #8 bars that are supposed to go here?" the database can be accessed and the answer displayed on the visor. It may be appreciated that this system could extend far beyond this level; supervisors could request shipment status, locate and call meetings with key workers, interactively flag problems to management etc. Many of the required base technologies to achieve this already exist; assembling them in a fashion that meets the “effective, cheap, and robust” criterion will prove the measure of the problem. 110Retractable HUD (Heads Up Display) for Database Interrogator Figure 1.1.8: One possible manifestation of an individual helmet-mounted database interrogator. An on-site worker is able to querry, for example, regarding the location of a particular component. More advanced versions will permit full registration of the virtual world database three dimensional site representation model with that of the real world in real-time. This will permit placement of components without the need for on site surveying or measuring._ Retractable HUD (Heads Up Display) : for Database Interrogator Figure 1.1.9: Using the portable Database Interrogator to identify the intended final destination for an on site component. Each component will have a bar stripe or smart chip identifier installed at the fabricators. Erection sequencing programs assigned at the design office will indicate when and where a particular component is called for, thus preventing mistaken installation.Vehicle-Based Heads Up Display Ly Figure 1.1.10: Vehicle-Based Heads Up Display (HUD). Real-time information concerning the desired site geometry may be downlinked to a see-through projection HUD to a machine operator who can then guide the earthmoving Speration to match the registered geometry of the final desired landscaped Surface. ‘These actions can be semi-automated to the extent that the operator should be able to request “auto-pilot” control through a particular maneuver, after which the machine will carry out the necessary actions. Operators will be able to over-ride the automated system at any time. Since complete kinematic models are kept in real time for each vehicle, each piece of machinery in effect becomes a real-time surveying instrument, relaying the present state of the site geometry local to the vehicle to the global database.The feedback from the global database may come on a hierarchy of levels, beginning with the above mentioned construction worker interface. The next level would be dedicated head-up displays mounted in construction vehicles. But the concept can be taken much further, effectively turning the real vehicle cab into a virtual receiving port (Figure 1.1.10) that would make it indistinguishable from a virtual training simulator for the same piece of machinery. The person operating the actual vehicle can call up the virtual image of the site indicating what needs to be done and providing recommended maneuvers. This is easy to envision with earth moving, but can be extended to all forms of construction machinery. By building realistic virtual simulators to match the actual cabs and by equipping the cabs with virtual feedback systems one immediately obtains the spinoff product of tele-operation, if desired, since an operator working in a high fidelity virtual cab should be able to operate the actual machinery without actually being in it. The fidelity will be limited by the time-of-flight of the transmission signals and will be a factor on some very long distance control operations. 1.1.5 Construction Robotics The final level of feedback is autonomous and/or semi-autonomous control, in essence “robotics.” The slow pace of development in the robotics world during the last decade, as opposed to the early, and wildly optimistic, expectations of the degree to which robots would penetrate everyday life, has given way to a cautious re-assessment of what is achievable and what is “futuristic” with respect to construction robots. It is now generally accepted among researchers that turning loose a 1000 horsepower bulldozer or a 50 ton crane on a construction site without human supervision is not likely in the foreseeable future. More likely it will come to pass that a full time operator does the setup, fixturing, initialization, and choice of process to be performed and then the machine performs the task. In effect, the strategy is to let the human operator do what is easy and natural for a human, and let the computer do what is easy and natural for the computer. The resulting machine-operator team may be 5-10 times more productive than present, conventional methods. Initial, tentative steps in this direction will take place during the next five years. NIST is presently studying various aspects of construction operations in order to logically downselect the most desirable task candidates for automation. Initial demonstration projects in semi-automated robot operations are anticipated to involve on-site setup by construction personnel, followed by autonomous task execution by the robot. The operator must, at all times, have the ability to over- ride and stop the robotic process. One example expected to see early implementation is semi-automated three dimensional terrain profiling within a Fucontrolled area (Figure 1.1.11) and semi-automated grade level controls on earth moving and paving machinery (PWRI, 1995). References: Berry, RM, (1976), “History of Geodetic Leveling in the United States,” Surveying and Mapping 36 (no. 2):137 Allegheny Excavating, (1995) personal communication, Gary Sippel, President, Allegheny Excavating, Inc., Butler, PA Dewberry & Davis, (1995) personal communication Barry Kendel, surveying division, Dewberry & Davis, Gaithersburg, MD. AISC, (1980), Manual of Steel Construction, Eighth Edition, American Institute of Steel Construction, Inc., Chicago, ILL. Chinnock, (1995), “Flat Panels Launch Helmet-Mounted Display Market,” Military & Aerospace Electronics, Feb. 1995, pp. 11-13. PWRI (1995), “Start of Experiments on ETC and ARTS/AHS at PWRI Test Course,” PWRI Newsletter, No. 62, Oct. 1995, Public Works Research Institute, Ministry of Construction, Tsukuba, Japan, pp. 1-4 12Construction Robotics Figure 1.1.11: Ultimately, fully autonomous machinery will see use on construction sites. Much earlier, however, will come semi- autonomous vehicles in which an operator will have the ability to make on site judgements and override the operation of the machine if an unstructured event occurs for which the machine is not programmed. Also, remote operation (tele-operation) of un-manned vehicles (with remote operator override capability) will begin to see wide use in hazardous environments such as toxic waste and nuclear cleanup.Chapter 2: A Review of Present Electronic Positioning Technologies 2.1 Electronic Distance Measurement (EDM) During the past 15 years, electronic distance measuring (EDM) instruments have seen wide acceptance among surveyors and have almost completely surplanted the use of stadia and tapes for this purpose. These devices determine lengths based on phase changes that occur as electromagnetic energy of known wavelength travels from one end of a line to the other and returns. Early forms of these instruments tranmitted visible light and were capable of measuring distances up to 40 km at night; they were degraded during daylight hours. These were followed with microwave-based systems capable of 80 km range in daylight. Figure 2.1.1 [Brinker & Wolf, 1984] shows frequency and wavelength values for the electromagnetic spectrum along with identification of bands commonly used for the above distancing technology as well as some other common electronic instruments. The chief advantages of electronic surveying are the speed and accuracy with which distances can be measured. If a line of sight is available, long or short lengths can be measured over bodies of water or terrain that is inaccessible or rugged (e.g. a virgin construction site). Present EDM systems can be classified by the wavelength of the transmitted electromagnetic energy and fall into two general categories: a) Electro-optical systems which transmit either modulated laser light between 400-900 nm (visible spectrum through far ultraviolet: 4.5-7.5 (10) Fy). b) Microwave systems which transmit microwaves. Although microwaves occupy a large portion of the electromagnetic spectrum (from 3(10)* - 1(10)° nm), most measurement geodetic survey instruments fall in the 8.6 to 100 (10)$ nm range (3 to 35 GHz). In general, EDM equipment measures distances by comparing a line of unknown Iength to the known wavelength of modulated electromagnetic energy. Electromagnetic energy propagates through the atmosphere as: Vain eq(1.1) Where V = velocity of electromagnetic radiation, in meters per second; f = the frequency of modulation, in hertz (cycles/second); and 2 = the wavelength, in 2Frequency f, in hertz to 10°_10°_107_10°_10°_10"_10"_10" 10" 10" 10% 10% 107 10" 10” Short TAM | Wave Band Radio Wave favelength 2, in Meters 3x10 10° 10? 10° 10’ 107 107 10° 10* 10° 10* 107 10* 10° 10% 10” po 5 = hal a ae Ss — Figure 2. Frequency spectrum and Wavelengths for electromagnetic radiationmeters. In a vacuum, V is equal to the relativistic constant c = 299,792,500 m/second. In the earth’s atmosphere this velocity is reduced in accordance with the local index of refraction (n) as: Vee/n eq (1.2) Where n is dependent on the pressure, temperature, and humidity of the air column and has a sea level value of approximately 1.0003. Because of this, and the interdependence of equations 1.1 and 12, accurate electronic distance measurement using phase information requires that the atmosphere be sampled local to the measurements and corrections made either to the instrument or to the results. In general, the higher the frequency of operation, the finer the range accuracy. But with higher frequency, atmospheric effects also become greater. Most instruments are now specified with both an instrumental accuracy (a function of the physics of the device) and an atmospheric error, usually expressed in parts-per-million (ppm) of the surveyed distance. The general EDM distancing procedure is shown in Figure 2.1.2 (from Brinker & Wolf, 1984). An EDM device, centered by means of a plumb bob or optical plummet over the reference transmitting station, transmits a carrier signal or electromagnetic energy upon which a reference frequency has been superimposed, or modulated. The signal is returned from the receiving station to the receiver, located on the EDM instrument, so that the travel path is double the slope distance. In Figure 2.1.2 the modulated electromagnetic energy is represented by a series of sine waves having wavelength lambda. Any position along a given wave can be specified by its phase angle, which is O-degrees at its beginning, 180-degrees at the midpoint, and 360-degrees at its end. EDM devices used in surveying operate by measuring phase shift. In this procedure, the returned energy undergoes a complete 360-degree phase change for each even multiple of exactly one-half the wavelength separating the line's endpoints. When a line is not exactly an even multiple of the half wavelength (which would usually be the case), the fractional part is measured by the instrument as a nonzero phase angle. If the precise length of a wave is known (via eq(1.1) with atmospheric compensation) then the fractional part can be converted to distance. There is a problem, however, in that EDMs using this technique can directly resolve only a fractional part of a wavelength, but not the number of full wavelengths over the full transit path to and from the reflector system. In commercial systems this ambiguity is resolved by transmitting additional signals of lower frequency and longer wavelength, so that the true distance, within the design range of the device, can be uniquely determined. 22Modulated electromagnetic Figure 2.1.2: Standard EDM surveying system in which a carrier frequency is transmitted from the survey instrument toa retroreflector target. The phase of the returned signal is compared with that of the transmitted signal. The phase difference is used to determine the total transit distance to and from the retroreflector. The true line-of-sight survey distance is half the round trip distance.As an example, if a system used four separate frequencies of 50 MHz, 5 MHz, 500 kHz, and 50 KHz (with associated wavelengths of 6, 60, 600, and 6000 m, respectively) then the total range of the instrument over which the distance could be ascertained unambiguously is 6,666 m. The accuracy of the instrument is determined by the shortest wavelength (highest frequency) and the resolution of the digital phase detectors used to measure the fractional wavelength. For example, if the phase detector was only accurate to I-degree , then the accuracy would be given by 6m /360-degrees = 0.0167 m = 16.7 mm, not accounting for atmospheric effects. Most practical instruments involve the use of retro- reflectors (corner cube mirrors) which return the transmitted signal to the EDM instrument, and so the actual unambiguous range would be half the above amount or 3,333 m, which is beyond that typically required of most site- surveying total stations. Some present-day instruments use only two frequencies (eg. 75 kHz and 15 MHz), with a commensurate shortening of the unambiguous range. Typical accuracies and ranges for various EDM units are given in Table 13.1, Figures 2.13 and 2.1.4 show on site use of a typical two-frequency “total station” and its associated target retroreflector. Table 1.2.1: Typical EDM Instrument Accuracies Instrument Range ‘Accurai Wild TC500 400-1300 m mm + Sppm Wild TCO10 400-1300 m +/-3mm +2 ppm Wild TC 1610 400-1300 m +/-2mm +2 ppm MicroFix 100C___ 60,000 m +/-15 mm +3 ppm References: Brinker, R.C., and Wolf, P.R, (1984), “Elementary Surveying,” Seventh Edition, Harper & Row, Publishers, New York. 2.2 GPS-Based Metrology Background The GPS (Global Positioning) system presently consists of a constellation of 24 satellites, four in each of six circular 20,321 km orbits with inclinations of 55 degrees. The fundamental navigation technique for GPS is to use one way ranging from the GPS satellites which are also broadcasting their estimated positions. Ranges are measured to four (or more) satellites simultaneously in view by matching (correlating) the incoming signal with a user generated replica signal and measuring the received phase against the user’s (relatively crude) crystal clock. 23Figure 2.1.3a: Survey crew setting up a typical Total Station survey shot. Azimuth and vertical angles are are alighed using the optical sight and digitally displayed on an integral LCD panel. EDM distance is obtained using the same sighting arrangement. The received signal from a retroreflector located at the target station is processed onboard and displayed as well on the LCD.Figure 2.1.3b: Closeup photo of a typical Total Station transmitter/ receiver unit. Optical alignment with a retroreflector target is achieved via the central lens; digital values for azimuth, inclination, and computed distance are accessed sequentially in the LCD display located just below the lens body.Figure 2.1.4a: Survey crew member setting up a typical retroreflector to be used in a Total Station survey shot. The optical center of the retroreflector is aligned, using an optical plumb, over the target station.Figure 2.1.4b: Closeup photo of typical retroreflector unit set up on station for an EDM shot from a Total Station. Accuracies using this technology can achieve +/-2 mm survey error and 2 ppm error due to propagation delays resulting from variance in atmospheric conditions (humidity, temperature, barometric pressure. )With four satellites and appropriate geometry, four unknowns can be determined; typically, they are: latitude, longitude, altitude, and a correction to the user's clock. If altitude or time are already known, a lesser number of satellites can be used. Each satellite's future position is estimated from ranging measurements taken at world wide monitoring stations. These stations calculate, and upload, future satellite locations and future satellite clock corrections, Excellent summaries of various aspects of general GPS technology are presented in [ Parkinson et.al., 1995; Lichtenegger & Collins, 1992; Stein et.al., 1994; Trimble, 1994; Hurn, 1989; ION, 1980; ION, 1984; ION, 1986; and ION, 1993 ]. A key feature of the GPS design is that the satellites need not be continuously monitored and controlled. To achieve this autonomy, the satellites must be preditable in four dimensions: three of position and one of time. Predictability, in the orbital position, is aided because the high altitude orbits are virtually unaffected by atmospheric drag. In order to insure that all satellite clocks remained synchronized, provisions were developed to fly accurate timing standards. The only type of clock sufficiently accurate to achieve the error tolerance specifications for GPS (5 feet accumulated error per day) is an atomic standard. GPS has traditionally used two types of atomic clocks: Rubidium and Cesium. The GPS ranging signal is broadcast at two frequencies: a primary signal at 1575 MHz (called L1) and a secondary broadcast at 1227 MHz (called L2). These signals are generated synchronously, so that a user who receives both signals can directly calibrate the ionospheric group delay and apply appropriate corrections. Civilian users are limited to the Li band. Both the Li and L2 frequencies can carry two independent modulations through the use of phase quadrature. The present GPS system permits two modulations for the L1 signal and one, known as protected mode, on L2. These modulations are defined in the following sections. 2.2.1 GPS Ranging Modes Code Solution The primary intent of the GPS system was to provide 5 to 10 meter accuracy absolute point positions for the U.S. Depariment of Defense using coded information on two carrier frequencies, the Ll frequency at 1575.42 MHz and the L2 frequency at 1227.60 MHz. The high-accuracy service is call the Precise Positioning Service (PPS) and uses what is called P-code (Precise-code). The use of PPS is restricted and is not available for civilian use when Selective Availability (S/A) is turned on. A lower level of precision is available at all times and is called the Standard Positioning Service (SPS) which uses the Coarse 24Acquisition or C/A-code. In this a short pseudo-random noise code is broadcast ata rate of 1.023 megabits/second and contains satellite position and time. The P-code is broadcast at ten times the rate of C/A, 10.23 megabits/second. Because of its higher modulation bandwidth, the P-code ranging signal is more precise. This code , when encrypted, becomes the Y code. The military uses this encryption capability in such a way as to prevent the more precise positioning service from being used by an unauthorized user. During S/A the satellite frequency is dithered, limiting the point position to an accuracy of 100 m in the horizontal and 150 m in the vertical components. An example of the variation in position using SPS with S/A on is shown for 25 hours of data collected at UNAVCO using a Trimble SSe receiver (Figures 2.2.1.1 and 2.2.1.2). During this period the position varied up to 50 meters horizontally and 100 meters vertically. Code Differential Solution Considerable improvement is obtained by combining observations from two receivers; the second unit comprises the “reference” receiver (Figure 2.2.1.3). Both receivers, because of their proximity, see essentially the same range error to each satellite and therefore the same corresponding error in position. With one receiver at a known position, the range errors can be determined and transmitted to the roving receiver. The roving receiver applies these corrections to the observed ranges in real-time [Hurn, 1993]. The standard format for code differential corrections is RTCM. Almost all GPS receivers with a serial interface are capable of accepting RTCM corrections. For small inexpensive receivers ($300) the accuracy is limited by the noise level of the code measurement, which is typically 2 to 10 meters. A newer class of enhanced C/A code tracking receivers such as the Trimble 4000 SSe and Ashtech Z12 have noise levels at the 0.5 meter level and advertise 1 meter level differential position accuracy. A pair of Trimble 4000 SSe’s were connected using a radio link with one set as a reference station and the other to accept RTCM corrections. The majority of the horizontal positions differ by less than +- 50 cm with the exception of a nearly 2 meter horizontal excursion near the start of the time series (Figure 2.2.1.4). The vertical solution variation is up to +- 5 meters but more typically less than +- 1 meter. The precision of the differential solution using only C/A code will degrade with increasing distance due to ionospheric effects, tropospheric effects and errors in the broadcast orbit ephemeris. Current development is directed toward increasing the range of code differential GPS (DGPS) beyond about 100 km, From a software point of view, the techniques are relatively straightforward and the reliability is high. 25