At the outset of the 1990s, leading transportation professionals and policy makers jointly determined that the future of transportation is necessarily one in which traffic operations will receive increasing attention, given imminent completion of the massive Interstate Highway System road construction program initiated four decades ago. For the foreseeable future, the focus increasingly will be on ensuring that existing facilities are operated safely and efficiently. As traffic continues to grow, with little increase in available road space, traffic problems are mounting. There is growing interest in electronic devices and systems that can sense traffic conditions in near real-time, communicate current traffic situations to highway facility and fleet operators, including public transit providers; convey travel planning information to travelers both in advance and during trips; enhance fleet dispatch and management capabilities; and enable `non-stop' toll, transit fare, and regulatory transactions. Collectively known as Intelligent Vehicle - Highway Systems (IVHS), these technologies are expected to dramatically alter the future of transportation, as well as create a market opportunity in excess of $200 billion over the next two decades.

Traffic management can be thought of as a stochastic queuing process where the serving time at one of its control points is dynamically linked to the global traffic pattern, which is, in turn, dynamically linked to the control point. For this closed-loop system to be effective, the traffic management system must sense and interpret a large spatial projection of data originating from multiple sensor suites. This concept is the basis for the development of a traffic flow wide-area surveillance (TFWAS) system. This paper presents the results of a study by Oak Ridge National Laboratory to define the operational specifications and characteristics, to determine the constraints, and to examine the state of technology of a TFWAS system in terms of traffic management and control. In doing so, the functions and attributes of a TFWAS system are mapped into an operational structure consistent with the Intelligent Vehicle Highway System (IVHS) concept and the existing highway infrastructure. This mapping includes identifying candidate sensor suites and establishing criteria, requirements, and performance measures by which these systems can be graded in their ability and practicality to meet the operational requirements of a TFWAS system. In light of this, issues such as system integration, applicable technologies, impact on traffic management and control, and public acceptance are addressed.

Traffic data collection studies and management systems require accurate and cost-effective sensors for roadway volume, vehicle speed, and vehicle type classification. In this study, we compare the performance of inexpensive and non-intrusive sensing using video cameras, passive acoustic microphone arrays, active ultrasonic acoustic ranging and Doppler sensors, Doppler radar, and passive infrared sensors in these applications. Using data taken on a highway near Boston, Mass. the various sensor types are coupled with appropriate signal processing for counting, speed estimation, and vehicle type classification. Emphasis is placed on understanding the fundamental limitations of each sensor type, and an interactive system for ground-truth database construction and comparison with the detected outputs and intermediate signal processor products is described. This system is essential for understanding the reasons for the observed performance.

A measurement system using 3-D imaging laser radars and real-time image processing to quantitatively measure and characterize inter-vehicular motion (i.e., the vehicle motion environment, VME) in real-world traffic settings is being developed by the Environmental Research Institute of Michigan (ERIM) under a subcontract with the University of Michigan Transportation Research Institute (UMTRI), who has a cooperative agreement with the National Highway Traffic Safety Administration. The system will be deployed at the roadside to acquire data on vehicle centroid and heading trajectories that will be used to study vehicle dynamics and accident causation. This effort is a direct response to the recognized need for a fundamental understanding of the vehicle motion environment to aid in the development of automotive active safety technologies for Intelligent Vehicle Highway System (IVHS) applications. This paper presents overviews of the national initiative for crash avoidance research, the role of the VME Program and the vehicle motion environment measurement system (i.e., VME-MS). Simulation studies were conducted in support of the image and data processing algorithm development. The motivation for and the implementation, capabilities and results of these simulation studies are presented.

One freeway and one surface street arterial site were chosen in each of three states to test and evaluate alternative traffic detector technologies. The states were selected to be representative of extremes in climatic conditions. Accordingly, Minnesota was chosen for its cold winter environment; Florida for its summer thunderstorms, lightning, and humidity; and Arizona for its dry desert summer heat. Sites were located on roadways that had high traffic density and suitable structures for mounting the overhead detectors. The detector technologies evaluated were ultrasonic, microwave radar, infrared laser radar, imaging and nonimaging passive infrared, video image processing with visible spectrum imagery, acoustic array, high sampling rate inductive loop, conventional inductive loop, microloop, and magnetometers. Approximately 5.9 GBytes of digital and analog vehicle detection data and more than three hundred video tapes of the corresponding traffic flow were recorded. The detector outputs were time tagged and recorded on 88 MByte replaceable magnetic cartridges by using a data logger specifically designed and built for this project. Data analysis software was written to convert the data into an easily accessible Paradox data base format compatible with a Windows personal computer operating system. Traffic volume ground truth data, obtained by counting vehicles in the recorded video imagery, were compared with the counts from the detector outputs. Speed ground truth data, obtained by driving probe vehicles through the field of view of the detectors and noting the vehicle speed as measured by the vehicle instrumentation, were compared with the speed measurement from the detectors.

Single vehicle roadway departures (SVRD) are serious vehicular accidents that may be avoided through the use of advanced technology. In this paper we briefly describe our Vehicle Lateral Position Measurement System designed to alert the driver of possible SVRD situations. We then describe our testing of the system under a contract awarded by the National Highway Traffic Safety Administration to Rockwell International.

Sensor technologies of collision avoidance systems are identified based on a literature search about available products, prototypes, and experimental systems. These sensors constitute the front-end functional element of potential countermeasure systems to rear-end, backing, lane change, roadway departure, opposite direction, intersection, and reduced visibility crashes. The characteristics and capabilities of alternative sensor technologies are described based on published literature. Crash avoidance sensor technologies encompass microwave, millimeter- wave, and near-infrared radars; ultrasonic transducers, charge-coupled device cameras operating in near-infrared and visible bands; uncooled, passive far-infrared detectors; millimeter-wave imaging radar; and near-infrared communications.

This paper discusses traffic signal control strategies that are suitable for advanced traffic management within IVHS (Intelligent Vehicle Highway Systems). The strategies consist of a multi-level design for the real-time, traffic-adaptive control of the urban signal network system. Each control level has different response characteristics, with the more advanced levels incorporating in a nested fashion the capabilities of the lower levels. A principal goal of the new multi-level design is to invoke a selected control strategy when it can provide the greatest benefit.

This paper describes a diode-laser-based vehicle detector/classifier (VDC) presently being developed by Schwartz Electro-Optics (SEO) under an IVHS-IDEA program for the National Academy of Sciences. The VDC employs a scanning laser rangefinder to measure three- dimensional vehicle profiles that can be used for very accurate vehicle classification. The narrow laser beam width permits the detection of closely spaced vehicles moving at high speed; even a two-inch-wide tow bar can be detected. The VDC shows great promise for applications involving electronic toll collection from vehicles at freeway speeds, where very high detection and classification accuracy is mandatory.

Vision-based sensing can be used for lane sensing, adaptive cruise control, collision warning, and driver performance monitoring functions of intelligent vehicles. Current computer vision algorithms are not robust for handling multiple vehicles in highway scenarios. Several new algorithms are proposed for multi-lane sensing, near-host object detection, vehicle cut-in situations, and specifying regions of interest for object tracking. These algorithms were tested successfully on more than 6000 images taken from real-highway scenes under different daytime lighting conditions.

An analog VLSI implementation of a smart microsensor that mimics the early visual processing stage in insects is described with an emphasis on the overall concept and the front- end detection. The system employs the `smart sensor' paradigm in that the detectors and processing circuitry are integrated on the one chip. The integrated circuit is composed of sixty channels of photodetectors and parallel processing elements. The photodetection circuitry includes p-well junction diodes on a 2 micrometers CMOS process and a logarithmic compression to increase the dynamic range of the system. The future possibility of gallium arsenide implementation is discussed. The processing elements behind each photodetector contain a low frequency differentiator where subthreshold design methods have been used. The completed IC is ideal for motion detection, particularly collision avoidance tasks, as it essentially detects distance, speed & bearing of an object. The Horridge Template Model for insect vision has been directly mapped into VLSI and therefore the IC truly exploits the beauty of nature in that the insect eye is so compact with parallel processing, enabling compact motion detection without the computational overhead of intensive imaging, full image extraction and interpretation. This world-first has exciting applications in the areas of automobile anti- collision, IVHS, autonomous robot guidance, aids for the blind, continuous process monitoring/web inspection and automated welding, for example.

Optical sensors are used for several ITS applications, including lateral control of vehicles, traffic sign recognition, car following, autonomous vehicle navigation, and obstacle detection. This paper treats the performance assessment of a sensor/image processor used as part of an on-board countermeasure system to prevent single vehicle roadway departure crashes. Sufficient image contrast between objects of interest and backgrounds is an essential factor influencing overall system performance. Contrast is determined by material properties affecting reflected/radiated intensities, as well as weather and visibility conditions. This paper discusses the modeling of these parameters and characterizes the contrast performance effects due to reduced visibility. The analysis process first involves generation of inherent road/off- road contrasts, followed by weather effects as a contrast modification. The sensor is modeled as a charge coupled device (CCD), with variable parameters. The results of the sensor/weather modeling are used to predict the performance on an in-vehicle warning system under various levels of adverse weather. Software employed in this effort was previously developed for the U.S. Air Force Wright Laboratory to determine target/background detection and recognition ranges for different sensor systems operating under various mission scenarios.

This paper examines markets, competing technologies, and required production costs of 77 GHz automotive radars. These products will be offered to the market a few years from now and represent the largest opportunity ever offered to the millimeter-wave (MMW) industry. To succeed in this business, an entire industry, primarily focused in the past on expensive small volume military applications, has to be re-engineered to successfully design and manufacture low cost, large volume parts.

This document presents the CO concentration simulator design. It may be used to study the performance of sensors with ventilation and traffic control systems. The model is general enough to be used with one way or two way tunnel networks. Sensor locations and performance can be varied and the resulting changes in driver exposure can be measured.

A frequency modulated/continuous wave (FM/cw) radar developed for automotive applications is described in this paper. The objective of this effort was to design a low-cost automotive collision warning radar that could be operated under Part 15 of the U.S. Federal Communications Commission regulations regarding intentional radiators including proximity sensors. In this paper, we describe a forward looking homodyne FM/cw 24.125 GHz radar with a digital signal processor (DSP). The data is collected while the FM/cw transceiver is modulated with several linear chirps of differing bandwidths and modulation slopes. The processor uses this data to calculate the range and Doppler velocity of multiple targets for the purpose of finding the safe following distance that should be kept between the host vehicle and targets. The system specifications and the effects of power and bandwidth on radar performance are shown. The ambiguity function of the homodyne transceiver and FFT spectral processor are shown along with the method of resolving these ambiguities for multiple targets. Data are presented showing radar measurements of a conducting sphere and test vehicle.

Detection of vehicle emissions and other gases will be an important part of the future ITS infrastructure, especially for enclosed roadways. The recently passed Intermodal Surface Transportation Efficiency Act and the Clean Air Amendments provide great incentive for the collection of emissions data. Enclosed roadways can use such data to control ventilation systems and detect hazardous substances. This paper examines the feasibility of using optical gas sensors for real-time control of enclosed roadways as well as regional environmental monitoring of mobile source emissions.

This paper should help those who are new to Intelligent Vehicle Highway Systems (IVHS) understand what the current IVHS needs are and how they relate to the Intermodal Surface Transportation Efficiency Act (ISTEA) of 1991 and the current planning process. The paper outlines the relationship between the current set of IVHS user services, the national program goals, and the technologies that can help meet those goals. It provides a framework for understanding how the deployment of IVHS technologies fits into the planning process by outlining the relationship between technology providers, the IVHS product/service provider, and the private consumer. The term `user' in this paper refers to individual travelers, commercial and public transportation fleet managers, local traffic and public transportation management agencies, and/or commercial vehicle drivers.

The Central Artery/Tunnel (CA/T) project provides an excellent opportunity to implement and test the application of Intelligent Transportation Systems (ITS) technology to improve traffic safety in a congested urban corridor. As currently designed, the CA/T project will have the most extensive ITS capabilities of any highway in the U.S.A., and numerous opportunities exist to build upon these capabilities. Advances in the state-of-the-art of both ITS technology and applications can improve the safety and convenience of the travelling public in Eastern Massachusetts and provide significant regional economic benefits as well.

Many recent IVHS deployments feature vehicle probes as prominent sources of information about traffic conditions either in addition to or in place of traditional fixed location measurement devices such as inductive loop detectors. A number of other technologies now being considered for traffic surveillance that are deployed in fixed locations are reviewed and are found to provide the same kinds of information as more complex probe vehicle data collection systems. Recent experience with both fixed location and vehicle-based incident detection systems is summarized. Several incident detection algorithms calibrated with both fixed location and vehicle-based traffic measurements from a simulation of a signalized arterial street for a soon to be launched IVHS demonstration are introduced. The performance of these algorithms, some of which use data from only one source while others use data from both sources, is compared. While the fixed location measurements appear to be superior to the probe reports for detecting incidents when used alone, incidents detected by the respective algorithms do not completely overlap. Algorithms using both sources of data perform substantially better than either single-source algorithm, indicating that there is value in collecting information using both methods.

Development of advanced, electronic automotive equipment as part of intelligent vehicle highway systems (IVHS) is a cause for concern. These new electronic systems may fundamentally change the driving task as we know it, making current display technology obsolete. As an example, a possible change in utility of a speedometer is discussed. In addition to changes in driving, new controls and displays associated with IVHS will increase the amount of information available to be processed by drivers, perhaps leading to overload. Overload, however, is a multidimensional construct and is not simply related to the amount of information presented. This paper attempts to address these issues and presents a model for the development of IVHS controls and displays based on human factors, ergonomic methodology. The model requires the use of task analysis to match human capabilities to the demands of using the new automotive systems.

Measurement of travel times provides a critical component, along with traffic volumes, in the calculation of traffic congestion. This paper describes the results of extensive field tests of several methods used to calculate travel times. In particular, the paper examines the use of Hi- 8 mm video cameras (camcorders) and an automatic license plate reading system, based on machine vision technology, to perform travel time calculations. The paper presents the preliminary results of field trials conducted in Boston, Mass., Seattle, Wash., and Lexington, Kentucky. Besides the results of the surveys, the paper examines issues related to transferring this new technology to this application. The paper looks at the logistics of conducting extensive surveys, lessons applicable to survey design, and the quantity and statistical significance of the data required. A comparison is made between the application of machine vision to traffic surveys with other applications of machine vision to the transportation industry. It also presents future plans for the use of machine vision for the measurement of various traffic parameters.

Autonomous intelligent cruise control (AICC) systems are not only controlling vehicles' speed but acting on the throttle and eventually on the brakes they could automatically maintain the relative speed and distance between two vehicles in the same lane. And more than just for comfort it appears that these new systems should improve the safety on highways. By applying a technique issued from the space research carried out by MATRA, a sensor based on a charge coupled device (CCD) was designed to acquire the reflected light on standard-mounted car reflectors of pulsed laser diodes emission. The CCD is working in a unique mode called flash during transfer (FDT) which allows identification of target patterns in severe optical environments. It provides high accuracy for distance and angular position of targets. The absence of moving mechanical parts ensures high reliability for this sensor. The large field of view and the high measurement rate give a global situation assessment and a short reaction time. Then, tracking and filtering algorithms have been developed in order to select the target, on which the equipped vehicle determines its safety distance and speed, taking into account its maneuvering and the behaviors of other vehicles.

We take advantage of recent technological advances in the field of ferroelectric liquid crystal silicon back plane optoelectronic devices. These are well suited to perform massively parallel processing tasks. That choice enables the design of low cost vision systems and allows the implementation of an on-board system. We focus on transport applications such as road sign recognition. Preliminary in-car experimental results are presented.

The complexity and congestion of current transportation systems often produce traffic situations that jeopardize the safety of the people involved. These situations vary from maintaining a safe distance behind a leading vehicle to safely allowing a pedestrian to cross a busy street. Environmental sensing plays a critical role in virtually all of these situations. Of the sensors available, vision sensors provide information that is richer and more complete than other sensors, making them a logical choice for a multisensor transportation system. In this paper we present robust techniques for intelligent vehicle-highway applications where computer vision plays a crucial role. In particular, we demonstrate that the controlled active vision framework can be utilized to provide a visual sensing modality to a traffic advisory system in order to increase the overall safety margin in a variety of common traffic situations. We have selected two application examples, vehicle tracking and pedestrian tracking, to demonstrate that the framework can provide precisely the type of information required to effectively manage the given situation.

This paper presents a high dynamic logarithmic photoreceptor compatible with standard CMOS process. This photoreceptor uses a bulk diode in the photovoltaic mode. Thanks to its embedded structure, it gives smaller offset noise. Some test results are included and commented on. The first-hand experiments show a dynamic range of 7 decades. Most high contrast road scenes could be sensed directly by using this photoreceptor without any external control.

This paper describes a motion-analysis system, applied to the problem of vehicle tracking in real-world highway scenes. The system is structured in two stages. In the first one, a motion- detection algorithm performs a figure/ground segmentation, providing binary masks of the moving objects. In the second stage, vehicles are tracked for the rest of the sequence, by using Kalman filters on two state vectors, which represent each target's position and velocity. A vehicle's motion is represented by an affine model, taking into account translations and scale changes. Three types of features have been used for the vehicle's description state vectors. Two of them are contour-based: the bounding box and the centroid of the convex polygon approximating the vehicles contour. The third one is region-based and consists of the 2-D pattern of the vehicle in the image. For each of these features, the performance of the tracking algorithm has been tested, in terms of the position error, stability of the estimated motion parameters, trace of the motion model's covariance matrix, as well as computing time. A comparison of these results appears in favor of the use of the bounding box features.

While there is a lot of recent development in the entire IVHS field, very few have had the opportunity to combine the many areas of development into a single integrated `intelligent' vehicle. This is the story of a currently deployed, commercially developed and sold, integrated autonomous automobile. This system was developed specifically to serve a major automobile manufacturer's need for an automated vehicle chassis durability test facility. Due to the severity of the road surface human drivers could not be used. A totally automated robotic vehicle driver and guidance system was necessary. The goals of the project were to create a combination of robotic vehicle driver systems and a base traffic control system, that was capable of testing up to 20 vehicles at a time on a 1.3 mile oval test track at speeds up to 50 MPH or greater. As a fixed price commercial project, system and component costs were of paramount importance. As a result, the greater part of the design effort was not technology development, but evaluation, modification, and integration of proven, existing technology in new and often novel ways.

One of the most difficult areas in the IVHS field, much effort is being spent on vehicle guidance systems. Recent developments in technology have allowed major upgrades in capabilities and performance of guidance systems originally developed for AGVs. Advances yield a robust, cost effective single wire system. This system has inherent advantages in system simplicity. A Cyplex developed variant of a single wire system has demonstrated high enough immunity to rain and light snow cover that normal (human) safety reductions in speed are more than adequate to stay within the required system performance envelope. This system and its antenna interface have shown the ability to guide the vehicle at slow speeds (10 MPH) with a tracking repeatability of plus or minus 1/8 of an inch. The basic guide and antenna system have been tested at speeds up to 80 MPH. The system has inherently superior abilities for lane changes and precision vehicle placement, such as stopping precisely on location for fully automated refueling operations. The operation of this Cyplex refined or developed system is described and the impact of a system that is commercially viable now for highway and off road use is discussed. A case for single wire systems as a primary guide is made. Single wire systems as a complement to machine vision based and other systems also are discussed.

8Early deployment of Intelligent Vehicle Highway Systems would necessitate the internal instrumentation of infrastructure for emergency preparedness. Existing quantitative analysis and visual analysis techniques are time consuming, cost prohibitive, and are often unreliable. Fiber optic sensors are rapidly replacing conventional instrumentation because of their small size, light weight, immunity to electromagnetic interference, and extremely high information carrying capability. In this paper research on novel optical fiber sensing techniques for health monitoring of civil infrastructure such as highways and bridges is reported. Design, fabrication, and implementation of fiber optic sensor configurations used for measurements of strain are discussed. Results from field tests conducted to demonstrate the effectiveness of fiber sensors at determining quantitative strain vector components near crack locations in bridges are presented. Emerging applications of fiber sensors for vehicle flow, vehicle speed, and weigh-in-motion measurements are also discussed.

The objective of this paper is to highlight recommendations made by fifteen experts from industry and State Departments of Transportation (DOT) regarding the design, implementation, operations, and maintenance of fiber optic communication links presently being used in their transportation management systems. This paper also brings forth the problems faced during the deployment of these systems. The procedure followed for this research was to review the specifications and design guidelines for various Federal Highway Administration (FHWA) projects which have implemented, or are in the process of implementing, fiber optic communication links for their traffic management systems. DOT officials and industry design consultants who were directly involved in the implementation of the FHWA projects were then interviewed on issues concerning system design, operations and management, bidding, and other institutional aspects. The result of these interviews is a set of recommendations ranging from increased use of the latest standards to suggestions for more efficient planning of the traffic management center. These problems and recommendations are presented in this paper. This paper thus offers valuable guidelines for the design and implementation of fiber optic communication systems for future IVHS and transportation management applications.

In this report we present some initial ideas about possible uses of the principally new advanced sensor technology based on the photoelectric structures with memory (PESM) in intelligent vehicle highway systems (IVHS). PESM is an integral multilayer solid-state device capable of registering, storing, converting, and processing of images. The information is registered and stored in the form of two-dimensional charge and potential patterns in the plane of the layers, and is transferred and transformed in a perpendicular direction due to interaction of these patterns correspondingly with electrical and optical input signals. That is why PESM provides high operation potential (to 10¹⁴ operation/bit/s) and some microsecond to millisecond operating time for such integral operations with images such as addition, subtraction, contouring, correlation, and so on. It seems to be very suitable for solving such widespread tasks of technical vision and image processing as object counting and recognition, moving object selection with recording of its trajectory, object tracking, and so on in real time and without complicated computer processing. We endeavor to use PESM, combining in itself the capabilities of photosensor and preprocessor devices, in intelligent observation and control systems on highways and for vehicle guidance applications.

This paper introduces our recent research on vehicle lateral control, a component of an automated highway system (AHS). We have developed a laser radar based automatic steering control system to study the performance and feasibility of such a system in an actual highway environment. Our approach does not require significant infrastructure investment because it is based on the retroreflective road markers currently installed on many roads. The system consists of a laser radar sensor, steering actuator, steering controller, and data processing system. The newly developed laser radar system measures not only the range to a target, but also the azimuth angle of the target in the sensor field of view. The azimuth angle measurement, which is essential for the lateral control system, has been implemented without any moving parts in the sensor system. A line scan CCD is used as the azimuth detector while an APD is used for the range measurement. The vehicle control algorithm was studied using computer simulation and a full-scale test vehicle was developed. Some previous related projects are reviewed and our unique sensor concept is described. The total system performance was demonstrated and evaluated on a California highway. Several issues remaining for further improvement of the system are also discussed.