Search

The San Pedro Bay Ports (SPBP) complex of Los Angeles and Long Beach in Southern California is one of the largest container port complexes in the world. This complex contributes significantly to both regional and national economies in California and the United States, respectively. However, the ongoing growth and economic benefits of the SPBP are threatened by negative externalities associated with port operations, particularly increasing congestion and air pollution. The objective of this paper is to explore a new approach to estimating vehicle emission impacts of freight corridor operations related to the port area, particularly those associated with heavy-duty diesel trucks. The paper combines a microscopic traffic simulation model to capture detailed vehicle trajectories (speeds and accelerations) and congestion effects, with an emission model and a spatial dispersion model to facilitate the estimation of the health and environmental justice impacts of freight corridor operations. Focus is on the I-710 freeway in the Alameda Corridor from the SPBP area up to downtown Los Angeles, California, some 20 mi north. Several scenarios were evaluated in addition to the 2005 base scenario: replacement of the current fleet of port heavy-duty diesel trucks with zero-emission vehicles, elimination of port heavy-duty diesel truck trips that would correspond to shifting more containers to other modes such as rail, and implementation of a truck-restricted lane, preventing trucks from using the leftmost lanes. The results show that fleet replacement with cleaner trucks yields the most emission reductions both quantitatively and spatially.

Vehicle class is an important characteristic of traffic measurement, and classification information can contribute to many important applications in various transportation fields. For instance, vehicle classification is helpful in monitoring heavy vehicle traffic for road maintenance and safety, modeling traffic flow, and obtaining performance measurements based on each vehicle class for traffic surveillance. A real-time vehicle classification model was introduced. A heuristic method combined with decision tree and K-means clustering approaches was proposed to develop the vehicle classification model. The features used in the proposed model were extracted from piecewise slope rate values, which were obtained from single-loop inductive signature data. Three vehicle classification schemes–FHWA, FHWA-I, and Real-time Traffic Performance Measurement System–and a data set obtained from square single-loop detectors were used for model development. A data set obtained from round single-loop detectors was applied to test the transferability of the proposed model. The results demonstrated that the proposed real-time vehicle classification model is not only capable of categorizing vehicle types on the basis of the FHWA scheme but also is capable of grouping vehicles into more detailed classes. The classification model can successfully classify vehicles into 15 classes using single-loop detector data without any explicit axle information. In addition, the advantages of the proposed vehicle classification model are its simplicity, its use of the current detection infrastructure, and its enhancement of the use of single-loop detectors for vehicle classification. The initial results also suggest the potential for transferability of the vehicle classification approach and are very encouraging.

This study presents a warning information system based on an innovate methodology to estimate accident likelihood in real time. Bayesian modeling approach implemented by the probabilistic neural network (PNN) is conducted to identify hazardous traffic conditions leading to potential accident occurrence. The proposed system displays warning signs to call drivers' attention for safer and careful driving once hazardous traffic conditions are observed by evaluating accident likelihood. It is believed that the proposed system to produce effective warning information for real-time safety enhancement could be a valuable tool to highway users and operators.

The Highway Capacity Manual provides a method for determining the level of service (LOS) on freeways to evaluate freeway performance. Apart from being essentially an off-line decision support tool for planning and design, it is also based on point measurements from loop detectors, which may not provide an accurate assessment of freeway section performance. In order to meet user requirements of advanced traffic management and information systems, new LOS criteria based on section measures are required for real-time freeway analysis. The main aim of this research was to demonstrate a technique for development of such LOS criteria. The study uses a new measure of effectiveness, called reidentified median section speed (RMSS), derived from analysis of inductive vehicle signatures and reidentification of vehicles traveling through a major section of freeway in the City of Irvine, CA. Two main issues regarding real-time LOS criteria were addressed. The first was how to determine the threshold values partitioning the LOS categories. To provide reliable real-time traffic information, the threshold values should be decided such that RMSSs within the same LOS category represent similar traffic conditions as much as possible. In addition, RMSSs in different LOS categories should represent dissimilar traffic conditions. The second issue concerned the aggregation interval to use for deriving LOS categories. Two clustering techniques were then employed to derive LOS categories, namely, k-means and fuzzy approaches. Wilk's Lambda analysis and LOS stability analysis were performed to design new LOS criteria. Six LOS categories defined in terms of RMSS over a fixed 240-s interval were identified as the best solution to meet two major considerations described above. The procedures used in this study are readily transferable to other similarly equipped freeway sections for the derivation of real-time LOS.

Unlike conventional traffic safety studies that focused on histrionic data analyses, this study attempts to identify traffic conditions that might lead to a traffic accident from real-time freeway traffic data. An innovative feature of the study is to apply the concept, real-time and preaccident, to accident studies by integrating real-time capabilities in advanced traffic management and information systems (ATMIS). In this study, the traffic conditions leading to more accidents are defined as real-time accident likelihood, and the accident likelihood is estimated by employing a nonparametric Bayesian model. The main goal of the study is to remove hazardous traffic condition prior to accident occurrence by incorporating the real-time accident likelihood into ATMIS. This study estimates real-time accident likelihood from empirical data on I-880 freeway in California, and shows its applicability as an accident precursor.

The innovative feature of this study is first to develop a methodology for designing advanced traffic surveillance systems based on microscopic traffic simulations. The developed methodology makes it possible ultimately to obtain algorithm structures and associated parameter values. The methodology consists of two main simulation experiments. The first experiment includes a parametric evaluation based on Monte Carlo simulation to identify the required performance of the surveillance system. The second experiment involves performing a microscopic traffic simulation with actual surveillance algorithms and synthetic algorithm inputs. Paramics capable of adding various external modeling routines realized by advanced programming interfaces was used for both simulation experiments. As an application of the proposed methodology, an inductive signature–based anonymous vehicle tracking system was designed via origin–destination flow estimation problem. The outcomes of this study can serve as a logical and necessary precursor to field implementation of a variety of traffic surveillance systems.

Providing reliable predictive traffic information is a crucial element for successful operation of intelligent transportation systems. However, there are difficulties in providing accurate predictions mainly because of limitations in processing data associated with existing traffic surveillance systems and the lack of suitable prediction techniques. This study examines different aggregation intervals to characterize various levels of traffic dynamic representations and to investigate their effects on prediction accuracy. The relationship between data aggregation and predictability is explored by predicting travel times obtained from the inductive signature–based vehicle reidentification system on the I-405 freeway detector test bed in Irvine, California. For travel time prediction, this study employs three techniques: adaptive exponential smoothing, adaptive autoregressive model using Kalman filtering, and recurrent neural network with genetically optimized parameters. Finally, findings are discussed on suggestions for applying prediction techniques effectively.

Vehicle reidentification is the process of matching vehicles from one point on the roadway (one field of view) to the next. By performing vehicle reidentification, important traffic parameters including travel time, travel time variability, section density, and partial dynamic origin/destination demands can be obtained. Field traffic data were collected in Alton Parkway in Southern California for training and testing of the multidetector vehicle reidentification algorithm. These data consisted of inductive loop signatures of vehicles that traversed two detector stations spanning a section of an arterial and the corresponding video of these signatures. Even though the video collected was not optimized for pattern-recognition purposes, an investigation into the feasibility of fusing inductive vehicle signatures with video for anonymous vehicle reidentification was conducted. The resulting reidentification rate of over 90% shows that this approach merits further investigation. The results also show that the use of detector fusion provides system redundancy and yields slightly better results than the use of a single detector.

One of the fundamental requirements for facilitating implementation of any advanced transportation management and information system (ATMIS) is the development of a real-time traffic surveillance system able to produce reliable and accurate traffic performance measures. This study presents a new framework for anonymous vehicle tracking capable of tracing individual vehicles by the vehicle features. The core part of the proposed vehicle tracking method is a vehicle reidentification algorithm for signalized intersections based on inductive vehicle signatures. The new vehicle reidentification system consists of two major components: search space reduction and probabilistic pattern recognition. Not only real-time intersection performance but also intersection origin–destination information can be obtained as the algorithm’s basic output. A systematic simulation investigation was conducted of the performance and feasibility of anonymous vehicle tracking on signalized arterials using the Paramics simulation model. Extensive research experience with vehicle reidentification techniques on single roadway segments was the basis for investigating the performance that could be obtained from tracking individual vehicles across multiple detector stations. The findings of this study serve as a logical and necessary precursor to possible field implementation of vehicle reidentification techniques. The proposed anonymous vehicle tracking methodology with existing traffic surveillance infrastructure would be an invaluable tool for operating agencies in support of ATMIS strategies for congestion monitoring, adaptive traffic control, system evaluation, and provision of real-time traveler information.

This paper describes Cartesius , an innovative multi-agent architecture for the provision of real-time decision support to Traffic Operations Center personnel for coordinated, inter-jurisdictional traffic congestion management on freeway and surface street (arterial) networks. Cartesius is composed of two interacting knowledge-based systems that perform cooperative reasoning and resolve conflicts, for the analysis of non-recurring congestion and the on-line formulation of integrated control plans. The two agents support incident management operations for a freeway and an adjacent arterial subnetwork and interact with human operators, determining control recommendations in response to the occurrence of incidents. The multi-decision maker approach adopted by Cartesius reflects the spatial and administrative organization of traffic management agencies in US cities, providing a cooperative solution that exploits the agencies’ willingness to cooperate and unify their problem-solving capabilities, yet preserves the different levels of authority and the inherent distribution of data and expertise. The interaction between the agents is based on the functionally accurate, cooperative paradigm, a distributed problem solving approach aimed at producing consistent solutions without requiring the agents to have shared access to all globally available information. The cornerstone of this approach is the assumption that effective solutions can be efficiently obtained even when complete and up-to-date information is not directly available to the agents, thus reducing the need for complex data communication networks and synchronization time delays. The simulation-based evaluation of the system performance validates this assumption. The paper focuses on the distributed architecture of the agents and on their communication and decision making characteristics.

Accurate traffic data acquisition is essential for effective traffic surveillance, which is the backbone of advanced transportation management and information systems (ATMIS). Inductive loop detectors (ILDs) are still widely used for traffic data collection in the United States and many other countries. Three fundamental traffic parameters—speed, volume, and occupancy—are obtainable via single or double (speed-trap) ILDs. Real-time knowledge of such traffic parameters typically is required for use in ATMIS from a single loop detector station, which is the most commonly used. However, vehicle speeds cannot be obtained directly. Hence, the ability to estimate vehicle speeds accurately from single loop detectors is of considerable interest. In addition, operating agencies report that conventional loop detectors are unable to achieve volume count accuracies of more than 90% to 95%. The improved derivation of fundamental real-time traffic parameters, such as speed, volume, occupancy, and vehicle class, from single loop detectors and inductive signatures is demonstrated.

In this paper, we develop a coordinated traffic responsive ramp control strategy based on feedback control and artificial neural networks. The proposed feedback control law is nonlinear and realized by a series of neural networks. The parameters of the neural networks are obtained through a nonlinear optimization procedure. Traffic simulations show that the proposed nonlinear ramp control strategy compares favorably against the well-known linear quadratic (LQ) control strategy in reducing total travel times, particularly at situations where drastic changes in traffic demand and road capacity occur.

Time-varying lane-changing fractions and queue lengths are important lane traffic characteristics which may exhibit significant changes in the presence of a lane-blocking incident. This paper describes a stochastic system modeling approach to estimate time-varying lane-changing fractions and queue lengths for real-time incident management on surface streets. A discrete-time nonlinear stochastic model, which consists of recursive equations, measurement equations, and boundary constraints, is proposed to characterize inter-lane and intra-lane traffic state variables during incidents. To estimate lane-changing fractions and other state variables of the model, a recursive estimation algorithm is developed which primarily involves an extended Kalman filter, truncation, normalization, and a queue-updating procedure. Lane traffic counts are the sole input data used in this method. These data can be readily collected from conventional point detectors. The proposed model was calibrated using video-based data, then tested using simulated data from the TRAF-NETSIM simulation model, Version 5.0, as well as real video-based data sets. Preliminary test results indicate the feasibility of employing the proposed approach to estimate time-varying mandatory lane-changing fractions as well as queue lengths during incidents. The estimated lane-changing fractions and queue lengths can be used not only in better understanding the phenomena of incident-related inter-lane and intra-lane traffic characteristics, but also in developing real-time incident management technologies. Moreover, it is hoped that the results of this study might contribute to future research in related areas such as incident traffic prediction, incident-responsive traffic control and management, and automatic road congestion warning systems for further use in advanced transportation management and information systems.

The initial testing and evaluation of one of the key components of the Test-bed Real-Time Integrated Control and Evaluation Prototype System (TRICEPS) are described. TRICEPS is a software platform that facilitates the implementation and evaluation of a wide range of algorithms for traffic control and advanced transportation management systems. TRICEPS supports research activities by providing consistent interfaces for transportation management modules to both simulated and real-world environments. One of the key components of the TRICEPS platform is a distributed architecture for real-time decision support to transportation management center operators that provides a set of core transportation management applications for multijurisdictional traffic control and incident management on freeway and arterial networks. The architecture hosts algorithms for the estimation of current traffic conditions, the analysis of incident characteristics, and the formulation of multiple-decision-maker traffic control plans using advanced methodologies for cooperation and conflict resolution. Although the process of evaluation of such methodologies using the TRICEPS platform is aimed at demonstrating the effectiveness of the cooperative approach, it also provides a demonstration of platform functionality for range of related applications.

Intelligent transportation systems (ITS) hold great promise for public works professionals seeking to optimize those public investment strategies that deal with traffic congestion and other growth pressures. Advanced traffic and fleet management systems as well as traveler information and vehicle-based systems can take advantage of information technology advances and private market products to substantially improve the productivity, connectivity, and safety of transportation. And as the new federal transportation act further spurs deployment of these systems, ITS research programs will continue to play a vital role in supporting such deployment and the overall evolution of ITS. Such research includes developing technology tools, providing testing and evaluation environments, and helping advance the state of the practice.

Vehicle classification is the process of separating vehicles according to various predefined classes. Vehicle-classification information can be used in many transportation applications, including road maintenance, emissions/pollution estimation, traffic modeling and simulation, traffic safety, and toll setting. An example of a classification scheme using the following seven vehicle classes is presented: cars, sport-utility vehicles/ pickups, vans, limousines, buses, two-axle trucks, and trucks with more than two axles. This system uses vehicle inductive signatures collected from existing loop-detector infrastructure. It also uses a heuristic-discriminant algorithm for classification and a multi-objective optimization for training the heuristic algorithm. Feature vectors obtained by processing inductive signatures are used as inputs into the classification algorithm. Three different heuristic algorithms were developed, yielding encouraging results of 81 to 91 percent overall classification rates. The results demonstrate the potential of collecting network-wide vehicle-classification data from inductive loops. The availability of vehicle-classification data helps to improve traffic surveillance and better defines dynamic traffic networks.

Travel time is the reciprocal of speed and is a useful measure of road congestion and traffic system performance. Travel time is also a basic traffic variable that is used in many intelligent transportation system strategies such as route guidance, incident detection, and traveler information systems. Previously, speeds were mainly acquired from double inductive loops configured as speed traps, because single loop speed estimates based on assumptions of a constant vehicle length were inaccurate. However, more accurate measurements of speed can now be accomplished with single loops by utilizing inductive waveforms of vehicles that are output from newer detector cards. An algorithm using signal processing and statistical methods was developed to extract speeds from inductive waveforms. The results show that the proposed algorithm performs better than conventional single loop estimation methods. The results also show that the algorithm is robust under different traffic conditions and is transferrable across surveillance sites without the need for recalibration. The use of the extensive single loop surveillance infrastructure is a cost-effective way of obtaining more accurate networkwide travel time information.

Development of a universal freeway incident detection algorithm is a task that remains unfulfilled despite the promising approaches that have been recently explored. Incident detection researchers are realizing that an operationally successful detection framework needs to fulfill a full set of recognized needs. In this paper we attempt to define one possible set of universality requirements. Among the set of requirements, a freeway incident detection algorithm needs to be operationally accurate and transferable. Guided by the envisioned requirements, we introduce a new algorithm with potential for enhanced performance. The algorithm is a modified form of the Bayesian-based Probabilistic Neural Network (PNN) that utilizes the concept of statistical distance. The paper is divided into three main sections. The first section is a detailed definition of the attributes and capabilities that a potentially universal freeway incident detection framework should possess. The second section discusses the training and testing of the PNN. In the third section, we evaluate the PNN relative to the universality template previously defined. In addition to a large set of simulated incidents, we utilize a fairly large real incident databases from the I-880 freeway in California and the I-35W in Minnesota to comparatively evaluate the performance and transferability of different algorithms, including the PNN. Experimental results indicate that the new PNN-based algorithm is competitive with the Multi Layer Feed Forward (MLF) architecture, which was found in previous studies to yield superior incident detection performance, while being significantly faster to train. In addition, results also point to the possibility of utilizing the real-time learning capability of this new architecture to produce a transferable incident detection algorithm without the need for explicit off-line retraining in the new site. In this respect, and unlike existing algorithms, the PNN has been found to markedly improve in performance with time in service as it retrains itself on captured incident data, verified by the Traffic Management Center (TMC) operator. Moreover, the overall PNN-based framework promises potential enhancements towards the envisioned universality requirements.

Automated incident detection algorithms usually utilize direct measurements of traffic volumes and occupancies as input and output binary indications of the traffic status for every time step. The transferability of such algorithms is often poor because of the extensive overlap between the incident-related cluster and the incident-free cluster of patterns in the input space. The observed overlap is not only significant but also location specific. The extent of the overlap is shown, its causes are discussed, and an input preprocessor is proposed that relies on modifying the input feature space in order to segregate the two clusters of patterns and reduce their overlap. The result is a more readily separable set of classes; easier classification of new patterns, which leads to better performance; and most important, less site-dependent performance and hence significantly better direct transferability of the detection algorithm with less need for retraining at new sites. On the output side, isolated binary algorithm outputs also have several drawbacks that are identified. An output postprocessor is proposed that links the isolated outputs using a Bayesian update process and converts the isolated outputs into a continuous probabilistic measure that is updated for every time step. The output at one interval is therefore used as a prior probability for the next. The preprocessor and the postprocessor proposed in this paper were developed as integrated components of a universally transferable freeway incident detection framework, UNITID, but are generally applicable to any similar incident detection system.

In this paper, a new methodology is presented for real-time detection and characterization of incidents on surface streets. The proposed automatic incident detection approach is capable of detecting incidents promptly as well as characterizing incidents in terms of time-varying lane-changing fractions and queue lengths in blocked lanes, lanes blocked due to incidents, and incident duration. The architecture of the proposed incident detection approach consists of three sequential procedures: (1) Symptom Identification for identification of incident symptoms, (2) Signal Processing for real-time prediction of incident-related lane traffic characteristics and (3) Pattern Recognition for incident recognition. Lane traffic counts and occupancy are the only two major types of input data, which can be readily collected from point detectors. The primary techniques utilized in this paper include: (1) a discrete-time, nonlinear, stochastic system with boundary constraints to predict real-time lane-changing fractions and queue lengths and (2) a pattern-recognition-based algorithm employing modified sequential probability ratio tests (MSPRT) to detect incidents. Off-line tests based on simulated as well as video-based real data were conducted to assess the performance of the proposed algorithm. The test results have indicated the feasibility of achieving real-time incident detection using the proposed methodology.

This paper proposes using artificial neural networks in a modular architecture to assist in detecting different types of operational problems on signalized urban arterials. A traffic surveillance infrastructure that includes inductive loop detectors on intersection approaches as well as mid-block system loops for traffic monitoring is used. For arterials, problems that require the attention of a traffic management center operator include lane-blocking incidents, special event conditions, and detector malfunctioning. Problem detection depends on factors such as operating conditions, configuration of sensors within the network, and block or link length. The feasibility of training and testing neural network models as components of a modular architecture, with an appropriate model for each sub-problem of pattern recognition, is demonstrated. The performance of this modular architecture exceeded that of any single architecture applied to the detection of different types of operational problems. The paper also reports on the performance of each type of model considered and the effect traffic flow levels and detector configuration have on the performance of the incident detection model. The results show that with the selection of a suitable architecture, the modular neural classifier outperforms alternative discriminant analysis-based classifiers. This is demonstrated using cyclic data from microscopic simulation and field data from Urban Traffic Control System (UTCS) implementations in the Cities of Los Angeles and Anaheim, California.

This paper proposes a nonlinear approach for designing local traffic-responsive ramp controls using artificial neural networks. The problem is formulated as a nonlinear feedback control problem, where the system model is the well known hydrodynamic model developed by Lighthill and Whitham (1955), and Richards (1956), the model's flow-density relationship is nonlinear, and the feedback nonlinear controllers are composed of one or a number of feed-forward neural networks. These neural network controllers are of integral (I) or proportional-plus-integral (PI) type, and can be tuned on-line to achieve prescribed performance. Initial simulation results show that such an approach is promising.

This paper describes the process and the results or the evaluation of Traffic Congestion Manager (TCM), a knowledge-based system for Traffic Management Center operator assistance in real-time congestion management The process of evaluation included the validation of the correctness of the underlying knowledge base and the assessment of the system’s performance, through the definition of several simulation-based scenarios. The results demonstrate the ability of the system to effectively reduce driver travel time and stop time in the network under congested conditions, through the implementation of real-time traveler information and signal control. The results also provide insights into some of the weaknesses of the TCM approach, that will be addressed to fine-tune its reasoning capabilities and improve its performance.

Traffic flow on freeways is a complex process that often is described by a set of highly nonlinear, dynamic equations in the form of a macroscopic traffic flow model. However, some of the existing macroscopic models have been found to exhibit instabilities in their behavior and often do not track real traffic data correctly. On the other hand, microscopic traffic flow models can yield more detailed and accurate representations of traffic flow but are computationally intensive and typically not suitable for real-time implementation. Nevertheless, such implementations are likely to be necessary for development and application of advanced traffic control concepts in intelligent vehicle-highway systems. The development of a multilayer feed-forward artificial neural network model to address the freeway traffic system identification problem is presented. The solution of this problem is viewed as an essential element of an effort to build an improved freeway traffic flow model for the purpose of developing real-time predictive control strategies for dynamic traffic systems. To study the initial feasibility of the proposed neural network approach for traffic system identification, a three-layer feed-forward neural network model has been developed to emulate an improved version of a well-known higher-order continuum traffic model. Simulation results show that the neural network model can capture the traffic dynamics of this model quite closely. Future research will attempt to attain similar levels of performance using real traffic data.

The creation and progress of OAK-TREE (One-of-a-Kind Traffic Research and Education Experiment) are chronicled. OAK-TREE is a traffic educational laboratory experiment that was developed and conducted at the University of California at Irvine (UCI) during the spring quarter of 1996. This project involved a cooperative effort between the academic community and public-sector transportation operating agencies in developing a comprehensive field and laboratory educational experience for undergraduate students in transportation engineering. The agencies involved in this effort were the Department of Civil and Environmental Engineering at the University of California at Irvine, the Advanced Traffic Surveillance and Control Center of the city of Los Angeles, the Transportation Management Center of the city of Anaheim, and the Irvine Traffic Research and Control Center of the city of Irvine. These agencies were instrumental in creating an innovative laboratory experience for academic training in the use of state-of-the-practice resources and methods for traffic engineering. The results were the development of a state-of-the-art traffic-control educational laboratory at UCI and the genesis of a unique traffic-control course that fulfilled the requirements of both fundamental academic education and rigorous professional training.

In an effort to relieve peak hour congestion on freeways, various ramp metering algorithms have been employed to regulate the inputs to freeways from entry ramps. In this paper, we consider a freeway system comprised of a freeway section and its entry/exit ramps, and formulate the ramp control problem as a dynamic optimal process to minimize the total time spent in this system. Within this framework, we are able to show when ramp metering is beneficial to the system in terms of total time savings, and when it is not, under the restriction that the controlled freeway has to serve all of its ramp demand, and the traffic flow process follows the rules prescribed by the LWR theory with a triangular flow-density relationship. We also provide solution techniques to the problem and present some preliminary numerical results and empirical validation.

A major source of urban freeway delay in the U.S. is non-recurring congestion caused by incidents. The automated detection of incidents is an important function of a freeway traffic management center. A number of incident detection algorithms, using inductive loop data as input, have been developed over the past several decades, and a few of them are being deployed at urban freeway systems in major cities. These algorithms have shown varying degrees of success in their detection performance. In this paper, we present a new incident detection technique based on artificial neural networks (ANNs). Three types of neural network models, namely the multi-layer feedforward (MLF), the self-organizing feature map (SOFM) and adaptive resonance theory 2 (ART2), were developed to classify traffic surveillance data obtained from loop detectors, with the objective of using the classified output to detect lane-blocking freeway incidents. The models were developed with simulation data from a study site and tested with both simulation and field data at the same site. The MLF was found to have the highest potential, among the three ANNs, to achieve a better incident detection performance. The MLF was also tested with limited field data collected from three other freeway locations to explore its transferability. Our results and analyzes with data from the study site as well as the three test sites have shown that the MLF consistently detected most of the lane-blocking incidents and typically gave a false alarm rate lower than the California, McMaster and Minnesota algorithms currently in use.

A feedback neural net controller for an isolated freeway ramp is developed to locally stabilize freeway traffic. The neural net controller is of the integral type and has a nonlinear gain. Simulation results show that the neural net controller can stabilize traffic in a wider range of traffic conditions and with faster response than one of the leading existing ramp control algorithms.

This paper presents a comparative evaluation of traditional and neural‐network classifiers to detect cracks in video images of asphalt‐concrete pavement surfaces. The traditional classifiers used are the Bayes classifier and the k‐nearest neighbor (k‐NN) decision rule. The neural classifiers are the multilayer feed‐forward (MLF) neural‐network classifier and a two‐stage piecewise linear neural‐network classifier. Included in the paper is a theoretical background of the classifiers, their implementation procedures, and a case study to evaluate their performance in detection and classification of crack segements in pavement images. The results are presented and compared, and the relative merits of these techniques are discussed. The research reported in this paper is part of an ongoing research project, the objective of which is to develop a neural‐network‐based methodology for the processing of video images for automated detection, classification, and quantification of cracking on pavement surfaces.

Traffic incidents are believed to cause about 60% of the congestion delay in the United States. Accordingly, a number of U.S. cities maintain Traffic Operations Centers (TOCs) for incident detectio...

This paper presents a methodology for automating the processingof highway pavement video images using an integration of artificial neural network models with conventional image-processing techniques. The methodology developed is able to classify pavement surface cracking by the type, severity, and extent of cracks detected in video images. The approach is divided into five major steps: (1) image segmentation, which involves reduction of a raw gray-scale pavement image into a binary image, (2) feature extraction, (3) decomposition of the image into tiles and identification of tiles with cracking, (4) integration of the results from step (3) and classification of the type of cracking in each image, and (5) computation of the severities and extents of cracking detected in each image. In this methodology, artificial neural network models are used in automatic thresholding of the images in stage (1) and in the classification stages (3) and (4). The results obtained in each stage of the process are presented and discussed in this paper. The research results demonstrate the feasibility of this new approach for the detection, classification, and quantification of highway pavement surface cracking.

A major source of traffic delay in many large urban areas in the United States is non-recurring congestion caused by incidents. In the last several decades, a number of incident detection algorithms have been developed for freeway surveillance and control systems. However, conventional algorithms have generally met with mixed success in terms of performance criteria, such as detection rate, false alarm rate, and the mean time to detect incidents. The need for improved techniques is pressing, particularly with the advent of intelligent vehicle-highway system concepts. These systems will rely heavily on the ability to detect non-recurring traffic congestion automatically. In this paper, we hypothesize that spatial and temporal traffic patterns can be recognized and classified by an artificial neural network, and we present an investigation of such models for the automated detection of lane blocking incidents in a one-mile section of urban freeway. The artificial neural network was trained with data obtained from a microscopic freeway traffic simulation model that was specially calibrated for the actual freeway test section. The neural network first classifies the traffic state of the freeway section into either “incident-free” or “incident” conditions in every 30-second interval. The change in traffic state from incident-free to incident conditions is then used to trigger an incident alarm. Based on the results of an off-line test using simulated data, and comparisons with the well known California incident detection algorithm and the recently developed modified McMaster algorithm, the results suggest that neural network models have the potential to achieve significant improvements in incident-detection performance.

The need for accurate, comprehensive and timely traffic surveillance information is critical to ensure optimal traffic operations and management for advanced traffic management systems (ATMS). This paper describes an on-going study that involves the design and implementation of the section-based freeway real-time traffic performance measurement system (RTPMS)-an advanced surveillance system based on inductive vehicle signature technologies. Unlike traffic performance measurement systems that depend on point measures, RTPMS provides section-based travel time measures via matching of inductive vehicle signatures obtained at two adjacent detector station locations. Hence, the performance measures account for traffic conditions spanning an entire section, not just at a local detector station. In addition, each re-identified vehicle is also classified in RTPMS, yielding detailed section-based performance measures of different vehicle classes. This gives the ability to obtain more accurate travel statistics and vehicle exposure rates, such as those of commercial vehicles.

Complete and accurate traffic information is becoming more and more available with the advance in transportation surveillance technology. Especially, vehicle classification information can contribute to many transportation related fields such as road pavement management, estimation of polluted emission etc. In previous sections and chapters, it was shown that vehicle signature is function of vehicle type and traffic conditions. By exploiting this concept, the algorithm development in vehicle classification is investigated and corresponding results are presented.

The potential application of image-processing methods to automate the manual data collection and visual rating of pavement surface conditions has not been widely recognized until relatively recently. A cost-effective automated system to capture and extract pavement-surface distress from video images would improve safety and efficiency, and could offer consistency and uniformity of data and data quality, both locally and nationally. The purpose of this paper is to provide an introduction for transportation engineers, especially pavement engineers, to digital image-processing concepts and applications in pavement management. This includes pavement-surface-distress data of concern, basic machine-vision and digital-image-processing concepts, video-system characteristics for automated distress-data collection, and existing systems for digital imaging in pavement management in the United States, Japan, and France. It is concluded that in a relatively short time substantial progress has been made in the development of automated systems for distress-data acquisition and interpretation, and enhanced capabilities can be expected in the near future. Digital imaging technology is playing a significant role in these efforts.

An innovative feature of this paper is the demonstration of the feasibility of real-time vehicle reidentification algorithm development at a signalized intersection, where different traffic detection technologies were employed at upstream and downstream locations. Previous research by the authors on vehicle reidentification has utilized the same traffic sensors (e.g., conventional square inductive loops) and detectors (e.g., high-speed scanning detector cards) at both locations. In this paper, an opportunity arose for the first time to collect a downstream data set from a temporary installation of a prototype innovative inductive loop sensor known as a ldquobladerdquo sensor in conjunction with conventional inductive loops upstream. At both locations, advanced high-speed scanning detector cards were used. Although the number of vehicles for which data could be collected was small, encouraging results were obtained for vehicle reidentification performance in this system of mixed traffic detection technologies. In future large-scale applications of vehicle reidentification approaches for real-time traffic performance measurement, management, and control, it would be most beneficial and practical if heterogeneous and homogeneous detection systems could be supported. This initial paper yielded many useful insights about this important issue and demonstrated on a small scale the feasibility of vehicle reidentification in a system with heterogeneous detection technologies.

An innovative feature of this study is to firstly attempt to capture rear-end collision potentials from the analysis of inductive loop detector data. Signals collected from loops are applied for monitoring individual vehicle information on freeways to estimate safe stopping distances in car-following situations. An index to quantify the potential of rear-end collisions is derived, and further employed for developing criteria to evaluate levels of rear-end collision risks. The proposed methodology based on loop detector data enables to identify collision potentials in real time. It is believed that the index would be a valuable tool for operating agencies in developing various strategies and policies toward enhancements of traffic safety.

With the advent of Advanced Transportation Management and Information System (ATMIS), much attention has been paid to the estimation of dynamic or time-variant OD matrices, as development of improved methods for the derivation of OD-based real time traffic information is vital for analysis of transportation systems and various ATMIS strategies involving traveler information systems and route guidance, dynamic traffic assignment, and adaptive traffic signal control, among others. This study performs a systematic simulation investigation of the performance and feasibility of anonymous vehicle tracking in signalized networks using the Paramics simulation model. Previous research experience with vehicle reidentification techniques on single roadway segments is used to investigate the performance obtainable from tracking individual vehicles across multiple detector stations through a network to obtain real-time OD path flow information such as travel time and volume. The findings of this and subsequent studies serve as a logical and necessary precursor to possible field implementation in signalized networks of vehicle reidentification techniques.

With the advances in traffic sensor technology, more useful traffic information can be generated directly or indirectly. This study introduces an innovative single-loop speed estimation model using recent traffic sensor technology. Proposed speed estimation model can be easily integrated with existing loop systems without requiring any roadside construction work. With accurately estimated speed, accurate travel time can also be derived. This will help disseminate accurate real-time traffic information, and therefore improve overall network performance, and traffic control/operation. The spatial and temporal transferability of the developed model was tested and encouraging results were obtained. Automated vehicle grouping is also one of the main contributions of the proposed model.

Fundamental to the operation of most currently envisioned Intelligent Vehicle-Roadway System (IVRS) projects are advanced systems for surveillance, control and management of integrated freeway and arterial networks. A major concern in the development of such Smart Roads, and the focus of this paper, is the provision of decision support for traffic management center personnel, particularly for addressing nonrecurring congestion in large or complex networks. Decision support for control room staff is necessary to effectively detect, verify and develop response strategies for traffic incidents. The purpose of this paper is to suggest a novel artificial intelligence-based solution approach to the problem of providing operator decision support in integrated freeway and arterial traffic management systems, as part of a more general IVRS. A conceptual design is presented that is based on multiple real-time knowledge-based expert systems (KBES) integrated by a distributed blackboard problem-solving architecture. The paper expands on the notions of artificial intelligence and Smart Roads, and in particular the role, characteristics and requirements of KBES for real-time decision support. The overall concept of a decision support architecture is discussed and the blackboard approach is defined. A conceptual design for the proposed distributed blackboard architecture is presented, and discussed in terms of the component KBES functions at an areawide level, as well as the subnetwork or individual traffic control center level.

The papers in this Special Issue exemplify the current state-of-the-art of knowledge-based expert system research in transportation

This paper discusses the potential application of expert systems, a new information technology derived from artificial intelligence research, to pavement management. The principal focus concerns the development of such knowledge-based or expert systems for project-level analysis and design of pavement rehabilitation strategies. Particular emphasis is placed on the needs of locally managed highway systems, although many of the concepts are also more broadly applicable. The paper provides a brief overview of the evolution and application of expert systems. Several major issues in pavement management are discussed which strongly suggest the appropriateness of expert systems approaches. These issues include the role of computerized pavement management systems, the acquisition of pavement performance data and current procedures for pavement structural evaluation and design. Current research on PARADIGM, an integrated set of expert systems for pavement rehabilitation decision-making, is described. Finally, a PARADIGM component system consisting of an operational prototype expert system for pavement surface condition evaluation, is presented as an example.

Pavement surface condition data play a pivotal role in the analysis and design of pavement rehabilitation strategies. In project design, this information is usually used by a limited number of pavement engineering specialists who apply experience and judgment to formulate design and investment decisions. The nature of this process resists computerization using conventional computer tools. The development of a knowledge-based expert system that is based on programming concepts from the field of artificial intelligence is described. The system is called SCEPTRE, and assists highway engineers in planning cost-effective flexible pavement rehabilitation strategies at the project level. The use of an expert-systems approach permits SCEPTRE to make its specialized body of knowledge accessible to a much broader range of potential engineering users.

This paper describes recent experience with the development and performance of OVERDRIVE 1.2. a central component of PARADIGM (the Pavement Rehabilitation Analysis and Design Mentor). PARADIGM is a microcomputer-based tool to integrate a set of knowledge-based expert systems (KBES) and algorithmic models for pavement rehabilitation decision-making, particularly at the local level. Because the analysis and design of pavement rehabilitation strategies relies so heavily on experts, and many of the tasks involved are both complex and ill-defined, conventional computer tools are of limited use in this domain. A basic premise of this research is that KBES can play a significant role in addressing these problems. OVERDRIVE is the Overlay Design Heuristic Adviser, and is a research prototype KBES to assist local engineers evaluate the structural adequacy of existing flexible pavement and design the structural thickness of an asphalt concrete overlay. In this paper the general domain, the domain tasks add...

Understanding truck activity is an essential component of strategic freight planning and programming. However, recent studies have revealed a significant void in the availability of detailed truck activity data. Although some existing detectors are capable of providing truck counts by axle configuration, higher-resolution data that indicate truck body configuration, industry served, and commodity carried cannot be obtained from existing sensors. This paper presents the newly developed Truck Activity Monitoring System, which leverages existing in-pavement traffic sensors to provide truck activity data in California. Existing inductive loop detector sites were updated with inductive signature technology and advanced truck classification models were implemented to provide detailed truck count data with more than 40 truck body configurations. The system has been deployed to more than 90 detector locations in California to provide coverage at state borders, regional cordons, and significant metropolitan truck corridors. An interactive geographic information system website provides users with advanced visual analytics and access to archived data across all deployed locations. The case studies presented in this paper demonstrate the potential of the data obtained from this system in analyzing and understanding current and historical industry-specific truck activity.

The "simulated field test of an open architecture transportation controller (OATC)" accomplished an initial demonstration in a laboratory environment of the feasibility and capability of an OATC as a vital component of an advanced transportation management system (ATMS). This simulation involved the integration of the OATC with expert system software named "FRED" developed at the Institute of Transportation Studies (ITS) for real-time traffic management and control as part of a future ATMS. A cross platform interface was built for the expert system to communicate with the OATC. Software was implemented on the OATC to test the TCBLKS (traffic control blocks) software model and structure. The results of this field test demonstrated that using the TCBLKS was simple and intuitive. This field test also revealed many of the benefits of integrating the OATC and FRED such as data sharing and the potential for distributed processing.

As transportation surveillance technology continues to advance, the measurement of more complete traffic information is becoming increasingly feasible. ICAN stands for Inductive Classifying Artificial Network and is used to conveniently describe a self–organizing feature map (SOFM) for vehicle type categorization using inductive signatures as input. Vehicle type categorization is the separation of vehicles into predefined classes and can be useful for improving transportation efficiency, cost, environmental sustainability, enforcement, safety, and education. ICAN mainly focuses on the challenging task of differentiating between two–axle vehicles such as passenger car, sports utility vehicle (SUV), van, truck, and bus. This is in contrast to systems that classify according to the number of axles. One characteristic of ICAN is the simplicity of the 13–neuron one–dimensional neural network, and the employment of a small training set of 13 signatures. The overall classification results of 87% (data set 1) and 82% (data set 2) for seven categories coupled with consistent performance across all vehicle categories was significant and encouraging. Field freeway data was used for testing ICAN; representative signatures and video images of different vehicle types are presented.

The vehicle reidentification problem is the task of matching a vehicle detected at one location with the same vehicle detected at another location from a feasible set of candidate vehicles detected at the other location. This paper formulates and solves the vehicle reidentification problem as a lexicographic optimization problem. Lexicographic optimization is a preemptive multi-objective formulation, and this lexicographic optimization formulation combines lexicographic goal programming, classification, and Bayesian analysis techniques. The solution of the vehicle reidentification problem has the potential to yield reliable section measures such as travel times and densities, and enables the measurement of partial dynamic origin/destination demands. Implementation of this approach using conventional surveillance infrastructure permits the development of new algorithms for ATMIS (Advanced Transportation Management and Information Systems). Freeway inductive loop data from SR-24 in Lafayette, California, demonstrates that robust results can be obtained under different traffic flow conditions.

Highway resurfacing, restoration, rehabilitation, and reconstruction can extend the life of an existing facility. At the same time, enhanced safety can be engineered into a project by making improvements to other features that relate to, for example, bridge improvements, minor widening of lanes and shoulders, and minor alterations to horizontal and vertical alignment, in addition to pavement resurfacing and repair. Recent rehabilitation projects in California that were expanded to include safety enhancements resulted in original projct-cost estimates underestimating costs, often by millions of dollars. This paper presents an integrated database-management expert-system computer program (4RSCOPE), which is a design tool to address this problem. The program has been developed to allow Caltrans engineers to gather data from both office records and field assessments, determine the design features to be included, and thus determine the project scope. The program is expected to become the standard system used by district engineers statewide in California to develop rehabilitation projects.