Anuj Sharma

Multivariate Poisson-lognormal model for analysis of crashes on urban signalized intersections approach

ABSTRACT Many studies investigate contributing factors of intersection crashes, but very limited studies focus on cashes on the intersection approach. It is important to address the characteristics of intersection-approach crashes to better understand intersection safety. This article analyzes the crashes on signalized intersection approach on urban arterials with a focus on traffic and geometric elements. The intersection approach is defined as the segment between stop bar and the location 200 ft upstream from the stop bar. The multivariate Poisson log-normal (MVPLN) model is used to model crash counts by severity. Ten-year crash data collected from 643 signalized intersections in Nebraska are used for analysis. One-way road is found to be negatively related to all three severity levels (light crash, moderate crash, and severe crash) of crashes. Compared to the 12 ft lane width, narrower lane widths generally lead to fewer crashes. The intersection approaches on urban arterials are expected to have more crashes than collector roads. The numbers of right-turn, left-turn, and through lanes, as well as the annual average daily traffic on the intersection approach and its crossing approach are statistically significant factors increasing crash frequency. The MVPLN model is compared to univariate and zero-inflated Poisson models. The results reveal that the MVPLN model provides a superior fit over the univariate Poisson model.

Driver decision-making in the dilemma zone - examining the influences of clearance intervals, enforcement cameras and the provision of advance warning through a panel data random parameters probit model

In recent years, there have been a series of innovations in the field of vehicle detection at intersection approaches. Modern radar-based smart sensors make it possible to track individual vehicles in close proximity to an intersection. These advancements in technology potentially enable the provision of vehicle- and site-specific decision dilemma zone protection at the onset of the yellow indication at signalized intersections. To exploit this opportunity, it is critical to develop an in-depth understanding of those factors influencing a drivers decision to stop or go at the onset of yellow. This study investigates how signal timing strategies such as yellow interval durations, all-red clearance intervals, advance warning flashers, and automated camera enforcement affect driver decision-making. Data from 87 intersection approaches across five regions of the United States are used to develop a series of decision (i.e., probability of stopping) curves using vehicle trajectory and signal phasing data. A panel data random parameters probit model is used to account for heterogeneity across locations, as well as correlation in driver decision-making, due to unobserved factors that are unique to each signalized intersection. The results demonstrate drivers are more likely to stop at locations where enforcement cameras or flashers are present. Stopping was also more prevalent at intersections with lower speed limits, longer crossing distances, and where pedestrian crosswalks were present.

Reliability of Probe Speed Data for Detecting Congestion Trends

This paper presents a framework for evaluating the reliability of probe-sourced traffic speed data for congestion detection and general infrastructure performance assessment. The methodology outlined employs pattern recognition and time-series analysis to accurately quantify the similarity and dissimilarities between probe-sourced and benchmarked local sensor data. First, an adaptive and multiscale pattern recognition algorithm called Empirical Mode Decomposition (EMD) is used to define short, medium and long-term trends for the probe-sourced and infrastructure mounted local sensor datasets. The reliability of the probe data is then estimated based on the similarity or synchrony between corresponding trends. The synchrony between long-term trends are used as a measure of accuracy for general performance assessment, whereas short and medium term trends are used for testing the accuracy of congestion detection with probe-sourced data. Using one-month of high-resolution speed data, the authors were able to use probe data to detect on average 74% and 63% of the short-term events (events lasting for at most 30 minutes), 95% and 68% of the medium-term events (events lasting between 1 and 3 hours) on freeways and non - freeways respectively. Significant latencies do however exist between both datasets. On non - freeways, the benchmarked data detected events, on average, 12 minutes earlier than the probe data. On freeways, the latency between the datasets was reduced to 8 minutes. The resulting framework can serve as a guide for state DOTs when outsourcing or supplementing traffic data collection to probe-based services.

Performance comparison of two model based schemes for estimation of queue and delay at signalized intersections

Reliable estimation of performance measures such as queue and delay at intersections is important for the proper management of traffic. The information about these variables is valuable for the development of various traffic control strategies. The spatial nature of queue and delay makes their direct measurement a challenging task. The present study estimated these performance measures for the scenario when the queue ends within the advance detector using the data obtained from loop detectors installed at the entry and the exit of the intersection. A detailed analysis of the data obtained from loop detectors revealed that there were errors in the data. Two model based schemes, namely the occupancy based method and the queue clearance based method, were used for estimation of queue and delay using the erroneous data obtained from loop detectors. The results showed that the queue clearance based method was performing better while estimating queue and delay compared to the occupancy based method. Thus, the queue clearance based method would be valuable for the estimation of queues and delays while implementing with erroneous field data.

Distributed traffic control for reduced fuel consumption and travel time in transportation networks

This paper proposes a distributed framework for optimal control of vehicles in transportation networks. The objective is to reduce the balanced fuel consumption and travel time through hybrid control on speed limit and ramp metering rate. The dual decomposition theory associated with the subgradient method is then applied in order to decompose the optimal control problem into a series of suboptimal problems and then solve them individually via networked road infrastructures (RIs). Coordination among connected RIs is followed in each iteration to update the individual controls. An example is demonstrated to verify the reduction in terms of fuel consumption and travel time using the proposed approach.

Framework for Evaluating the Reliability of Wide-Area Probe Data

This paper presents a framework for evaluating the reliability of probe-sourced traffic speed data for detection of congestion and assessment of roadway performance. The methodology outlined uses pattern recognition to quantify accurately the similarities and dissimilarities of probe-sourced and benchmarked local sensor data. First, a pattern recognition algorithm called empirical mode decomposition was used to define short-, medium-, and long-term trends for the probe-sourced and infrastructure-mounted local sensor data sets. The reliability of the probe data was then estimated on the basis of the similarity or synchrony between corresponding trends. The synchrony between long-term trends was used as a measure of accuracy for general performance assessment, whereas short- and medium-term trends were used for testing the accuracy of congestion detection with probe-sourced data. By using 1 month of high-resolution speed data, the authors were able to use probe data to detect, on average, 74% and 63% of the short-term events (events lasting for at most 30 min) and 95% and 68% of the medium-term events (events lasting between 1 and 3 h) on freeways and nonfreeways, respectively. Significant latencies do, however, exist between the data sets. On nonfreeways, the benchmarked data detected events, on average, 12 min earlier than the probe data. On freeways, the latency between the data sets was reduced to 8 min. The resulting framework can serve as a guide for state departments of transportation when they outsource collection of traffic data to probe-based services or supplement their data with data from such services.

Data driven exploration of traffic network system dynamics using high resolution probe data

This paper, for the first time, explores the use of data-driven method to model the traffic dynamics of an interstate highway system with high resolution probe data. The dynamic mode decomposition and spatio-temporal pattern network were used to analyze traffic patterns on a 290 mile interstate highway across Iowa. The results show the data-driven methods can effectively detect the changes in traffic system dynamics during different time of day. Traffic dynamics during morning peak hours, evening peak hours and off-peak were very different on the studied road. In contrast, the trends over multiple months were similar during the same time periods. The study also found that inclement weather had a significant impact on the system dynamics. In future, the proposed methodology can be used to gain insights in the system dynamics of a traffic network. These models will be instrumental in optimal traffic control, traffic sensor placement and other policy decisions affecting the capacity of the network.

Evaluating the Reliability, Coverage, and Added Value of Crowdsourced Traffic Incident Reports from Waze

Traffic managers strive to have the most accurate information on road conditions, normally by using sensors and cameras, to act effectively in response to incidents. The prevalence of crowdsourced traffic information that has become available to traffic managers brings hope and yet raises important questions about the proper strategy for allocating resources to monitoring methods. Although many researchers have indicated the potential value in crowdsourced data, it is crucial to quantitatively explore its validity and coverage as a new source of data. This research studied crowdsourced data from a smartphone navigation application called Waze to identify the characteristics of this social sensor and provide a comparison with some of the common sources of data in traffic management. Moreover, this work quantifies the potential additional coverage that Waze can provide to existing sources of the advanced traffic management system (ATMS). One year of Waze data was compared with the recorded incidents in the Iowa’s ATMS in the same timeframe. Overall, the findings indicated that the crowdsourced data stream from Waze is an invaluable source of information for traffic monitoring with broad coverage (covering 43.2% of ATMS crash and congestion reports), timely reporting (on average 9.8 minutes earlier than a probe-based alternative), and reasonable geographic accuracy. Waze reports currently make significant contributions to incident detection and were found to have potential for further complementing the ATMS coverage of traffic conditions. In addition to these findings, the crowdsourced data evaluation procedure in this work provides researchers with a flexible framework for data evaluation.

Traffic Congestion Detection from Camera Images using Deep Convolution Neural Networks

Recent improvements in machine vision algorithms have led to closed-circuit television (CCTV) cameras emerging as an important data source for determining of the state of traffic congestion. In this study we used two different deep learning techniques, you only look once (YOLO) and deep convolution neural network (DCNN), to detect traffic congestion from camera images. The support vector machine (SVM), a shallow algorithm, was also used as a comparison to determine the improvements obtained using deep learning algorithms. Occupancy data from nearby radar sensors were used to label congested images in the dataset and for training the models. YOLO and DCCN achieved 91.5% and 90.2% accuracy, respectively, whereas SVM’s accuracy was 85.2%. Receiver operating characteristic curves were used to determine the sensitivity of the models with regard to different camera configurations, light conditions, and so forth. Although poor camera conditions at night affected the accuracy of the models, the areas under the curve from the deep models were found to be greater than 0.9 for all conditions. This shows that the models can perform well in challenging conditions as well.

Building Intelligence in the Automated Traffic Signal Performance Measures with Advanced Data Analytics

Automated traffic signal performance measures (ATSPMs) are designed to equip traffic signal controllers with high-resolution data-logging capabilities which may be used to generate performance measures. These measures allow practitioners to improve operations as well as to maintain and operate their systems in a safe and efficient manner. While they have changed the way that operators manage their systems, several shortcomings of ATSPMs, as identified by signal operators, include a lack of data quality control and the extent of resources required to use the tool properly for system-wide management. To address these shortcomings, intelligent traffic signal performance measurements (ITSPMs) are presented in this paper, using the concepts of machine learning, traffic flow theory, and data visualization to reduce the operator resources needed for overseeing data-driven ATSPMs. In applying these concepts, ITSPMs provide graphical tools to identify and remove logging errors and data from bad sensors, to determine trends in demand intelligently, and to address the question of whether or not coordination may be needed at an intersection. The focus of ATSPMs and ITSPMs on performance measures for multimodal users is identified as a pressing need for future research.