Journal of Transportation Engineering, Part A: Systems

Autonomous intersection management (AIM) is a type of intersection control for autonomous vehicles which eliminates the need for a traffic signal by using vehicle-to-infrastructure communication. Vehicles communicate information to an intersection manager which determines vehicle ordering and spacing so that vehicles can pass safely through the intersection. Reservation-based AIM, which gives vehicles space-time path reservations through an intersection, has the potential to greatly increase the capacity of intersections by allowing an intersection controller to optimize the path that each vehicle takes. A mixed-integer linear program is proposed which gives the intersection manager more flexibility through optimizing vehicle acceleration and velocity through the intersection. This model was integrated with microsimulation software and various scenarios were simulated, including fluctuating the vehicle demands, altering the permitted vehicle accelerations and speeds, and modifying the safety buffer between vehicles. The results indicate that the model proposed in this study has the capability to reduce delay and increase average speed experienced by vehicles compared with the existing reservation-based intersection control formulations and conventional signal controls.

This paper modeled rail track segments subject to multiple competitive failure modes and shock deterioration processes caused by passing trains. To determine the optimal inspection interval and intervention level of each failure mode, competitive failures were modeled by a series system, the shock arrivals were modeled by a Poisson process, and the deterioration caused by each shock was modeled by a gamma process. A maintenance improvement factor was considered as a function to model the decreasing effect of maintenance and aging on the system. The proposed model was applied to a real-world rail data set in which three geometry failure modes were the possible system deterioration mechanisms, and the condition-based maintenance strategy was determined by a Monte Carlo simulation algorithm for two classes of rail. The effects of reliability level and improvement factor on maintenance strategies were discussed numerically.

Signal control devices have been continuously evolving to make green assignments more responsive to traffic. Recent advances in connected and automated vehicles (CAVs) provide new opportunities to achieve higher performance levels for signalized intersections through an increased coordination level between vehicles and control devices. This study compares two state-of-the-art intersection management algorithms (IMAs) for CAVs and conventional vehicles (CNVs) to an actuated signal control system (ASCS). The two IMAs, the intelligent intersection control algorithm (IICA) and hybrid autonomous intersection management (H-AIM), are designed to enhance the efficiency of intersections by leveraging vehicle automation and connectivity. Our results show that the performance of both IICA and H-AIM improves as the CAV penetration rate increases. H-AIM attains lower average travel times than the state-of-the-practice ASCS only at a CAV penetration rate of 90% and greater. IICA, which jointly optimizes signal phase and timing (SPaT) and CAV trajectories, achieves the best average travel times and throughput for a wide range of CAV ratios. H-AIM yields lower average travel time and higher throughput compared to IICA at penetration rates close to 100%.

This paper presents the field implementation results of directional rumble strips (DRS), a low-cost traffic control device (TCD) designed to deter wrong-way driving (WWD) on freeways. Southbound off-ramps at Exits 208 and 284 on I-65 in Alabama were selected for implementation because they were ranked as high-risk locations by a network screening tool. A previous study developed and tested five DRS designs (A–E) with various configurations (e.g., A1, A2, and so forth). Three patterns (D3, C, and E2) were recommended for field implementation according to the previous test results. Pattern D3 was installed at the off-ramp terminal near the stop bar or yield line. Pattern C was implemented at the segment between the terminal and ramp curve. Pattern E2 was placed on the tangent part before the ramp curve. WWD incidents and distances before and after the implementation were collected using cameras. Field driving tests were conducted to collect sound and vibration data at various speed categories for both right-way (RW) and wrong-way (WW) directions. Before-and-after studies evaluated the effectiveness of DRS patterns in deterring WWD incidents. Sound and vibration analyses quantified the differences between RW and WW drivers’ perceptions. Results showed that the number of WWD incidents and average driving distances significantly decreased, by roughly 90% and 85%, respectively, after implementing all three DRS patterns. The results confirmed that WW drivers can perceive elevated sound and vibrations when passing the DRS. A general guideline was developed for implementing three different DRS on freeway off-ramps to deter WWD.

The total cost of road construction comprises direct costs paid by agencies and indirect costs paid by road users. In densely populated urban areas, work zone instigated user costs could outweigh direct costs and therefore need to be considered in project alternative selection. This paper develops and applies a probabilistic methodology for quantifying user delay cost (UDC) for urban arterial work zones. The methodology incorporates traffic microsimulation and Monte Carlo simulation to establish a distribution of UDCs, which supports risk-based optimization of work zone configuration, justification of accelerated construction methods, and establishment of contractual incentives and disincentives. The paper demonstrates the development of the methodology through a bridge rehabilitation case study in Calgary, Alberta, Canada. The results revealed that every hour of work zone operation during the morning peak resulted in 169.2 h of network-wide vehicle delay and a mean UDC of CAD 2,816 (in 2016 Canadian dollars). To further demonstrate the applicability of the methodology, a second case study examined three work zone configurations and concluded that the traditional work zone configuration instigated the lowest UDC.

One of the crucial areas in road safety is that of passing-related crashes on two-lane highways. Studies have shown that passing-related crashes that occurred due to inadequate passing sight distances were uncommon, indicating that the Manual on Uniform Traffic Control Devices constitutes safe designs. However, a proportion of such crashes were severe. Past research was conducted to assess passing sight distances and passing maneuvers. Yet, modeling passing-related crashes by type while considering unobserved heterogeneities, particularly of data that could not be collected such as driver aggressiveness, was not implemented. In this research, the severities of single- and multiple-vehicle passing-related crashes on two-lane highways were modeled using the mixed ordinal probit structure. As per the results of this research, insights were attained regarding the impacts of loss-of-control single-vehicle crashes, rollover single-vehicle crashes, single-motorcycle crashes, weather conditions, opposite-direction multiple-vehicle crashes, multiple-vehicle crashes involving motorcycles, high traffic volumes, and seatbelt use on the risk of sustaining severe injuries. Finally, passing-related crash countermeasures were suggested.

Drivers have an increased crash or incident risk when driving on freeway interchange ramps compared with the other interchange-related segments. Directional rumble strips (DRS), a pattern similar to the traditional traversable rumble strip, were developed as a low-cost traffic control device to deter wrong-way driving, meanwhile reducing right-way traffic speeds and changing driver behavior on off-ramps. This paper presents the impact of DRS on vehicle speed and driver behavior based on the two case studies at southbound off-ramps at Exits 208 and 284 on Interstate-65 (I-65) in Alabama. Three DRS patterns (D3, C, and E2) were implemented at different locations on off-ramps. Pattern D3 was installed at the off-ramp terminal near the stop bar or yield line. Pattern C was implemented at the segment between the terminal and ramp curve. Pattern E2 was placed on the tangent part before the ramp curve. A total of 1,344 h traffic speed data before and after the implementation were collected using magnetic sensors. Driver behavior was monitored for 576 h using video cameras. Before-and-after studies evaluated the impact of three DRS patterns on traffic speed on these two off-ramps. The results revealed that DRS can significantly reduce the mean, 85th percentile, and standard deviations (SDs) of off-ramp traffic speeds. In addition, DRS can help to mitigate aggressive driver behavior (e.g., exceeding the ramp speed limit) and guide turning traffics at ramp terminals.

Florida’s Road Rangers monitor the freeways for incidents to minimize incident clearance duration. The objective of this study was to evaluate the extent to which Road Rangers reduce incident clearance duration. Because incident clearance duration distributions are often right-skewed, the study applied quantile regression to relate incident clearance duration to influencing factors. Data skewed to the right are usually a result of lower bounds in a data set being extremely low relative to the rest of the data. Data from 28,000 incidents that occurred on freeways in Jacksonville, Florida, for the years 2014–2017 were analyzed. Of the factors analyzed, crash events, incident severity, shoulder blockage, weekends, nighttime, number of responding agencies, and towing were found to significantly increase incident clearance durations. Road Rangers were found to reduce incident clearance duration by 25.3%. In other words, shorter incident clearance durations were observed when Road Rangers responded to incidents compared to other agencies. The results of this study can, in general, provide researchers and practitioners with an effective way of evaluating the mobility benefits of the Road Ranger program.

Few studies have focused on the impacts of fog conditions on lane-based speeds on freeways. To explore speed differences and to predict speeds under fog conditions, this paper analyzed the historical traffic data and meteorological data at two freeway sections in Shanghai, China. The Kolmogorov–Smirnov test was used to compare speed distributions between the different levels of visibility. Multiple linear regression models for two sections with different numbers of lanes were established to predict speeds, in which lane number was used as an explanatory variable. It was found that the impacts of reduced visibility on speed distributions and speeds vary with visibility level, lane number, type of vehicle, day of the week, time of day, and the number of lanes. More importantly, the impacts of reduced visibility on mean speeds and speed distributions are different at sites with different numbers of lanes. The results can assist traffic management agencies to better implement variable speed limits for the fine management of lane levels under fog conditions.

Traffic data collection is the fundamental step in most applications of intelligent transportation systems (ITS). Recently, traffic data collection methods have become more robust and diversified, yet still have some limitations in their flexibility and coverage. Onboard monocular cameras have considerable potential to be turned into cost-effective moving traffic sensors combining the low cost and ego-vehicles’ high mobility. Existing studies have explored the feasibility of onboard cameras for scene understanding, etc. However, few studies have been conducted to utilize onboard monocular cameras for traffic flow data collection. To this end, this paper puts forward a method using the onboard monocular camera to collect traffic data. The basic structure is composed of a you-only-look-once (YOLO) model and spatial transformer network (STN) to detect vehicles in real-time. Then the traffic flow parameters are computed via fundamental optic and traffic flow theories. The experiment results show its reliability and similar sensing accuracy with inductive loop detectors on the road segment detection. In addition, the STN-YOLO model has a higher vehicle detection accuracy than the original YOLO model under complicated conditions.

Computer-driven vehicles will behave differently from human-driven vehicles due to changes in perception abilities, precision control, and reaction times. These changes are expected to have profound impacts on capacity, yet few models of automated driving are based on empirical measurements of computer-driven vehicles in real traffic. To this end, this paper investigates characteristics of an early form of longitudinal control automation, a commercially available adaptive cruise control (ACC) system driven in real traffic. Two car-following models were calibrated to a vehicle with ACC. First, the Intelligent Driver Model was reformulated to comply with ACC design standards then calibrated to match speed and range data from the test vehicle. The vehicle with ACC was found to decelerate less severely than predicted by the model when tested in severe braking and unimpeded acceleration scenarios. Second, the Wiedemann 99 model was calibrated because it is the default car-following model in the traffic microsimulation software program Vissim and can therefore be implemented cheaply and quickly in sophisticated models of roadways worldwide. Four parameters of the Wiedemann 99 model were measured directly from field observations of the test vehicle: standstill distance, start-up time, unimpeded acceleration profile, and maximum desired deceleration. Simulation results in Vissim were found to match the adaptive cruise control in unimpeded acceleration tests. These findings will benefit researchers and modelers seeking more accurate models of car-following behavior with adaptive cruise control and automated longitudinal control.

With the aid of advanced information technology, car parking space management is evolving dramatically toward an automatic way. The most efficient approach for parking-spot detection is based on embedded sensors, which can cause a significant cost of equipment purchasing, installation, and maintenance. Therefore, a growing number of studies have been done on vision-based detection methods using cameras. This paper aims to develop a parking-spot detection method based on images captured by existing surveillance cameras at car parks. Such images are used for recognition of parking lines, parking-spot positioning, and vehicle feature extraction. The issue of vehicle occlusion due to the limited installation height of surveillance cameras in car parks is handled. With the proposed detection approach, vacant and occupied parking spots could be distinguished to provide useful information of parking-space statuses to drivers. Various experiments have been conducted with promising results in different environmental conditions, like daytime and evening, sunny and rainy days, indoor and outdoor, and low and high camera positions. The proposed approach is applicable to large-scale car parks based on an extended multicamera system.

Turnouts are important components of railway infrastructure that require more attention as they must be frequently maintained. To transfer the resultingly high investment costs into a correspondingly long service life, the effects of all maintenance decisions must be identified. It is necessary to objectively weigh the impact of all maintenance activities and the optimum point in time for their execution. To make these decisions, information about the condition of the whole system as well as of the individual components must be available. This paper presents a model for describing the current ballast condition based on track measurement data collected by a track recording car. These data include longitudinal level measurements, whereby information on the changes observed in various wavelength ranges can be inferred by means of a power density spectra analysis. Time series analyses of these spectra allow conclusions to be drawn regarding the current condition of the ballast. By applying this method, new information can be collected on the component condition using existing data. In this study, statements could be derived about the ballast conditions at 45 turnouts, and they could be divided into three distinct ballast condition classes.

Determining the location of submain tracks (SMTs) or sidings to accommodate flexible operation is a critical issue in supporting construction investments and meeting passenger preferences. This project considered the problem of calculating the optimal location of SMTs. The objective was to minimize the sum of the construction costs for the SMTs and the total delay cost for passengers. We introduced properties for two train types. Additionally, we developed a mixed-integer model for multiple train types. We propose a procedure for using our model to determine the locations of SMTs for various headways and traffic mixes, and conducted numerical experiments to verify its performance. The results confirmed that our procedure is practical for an urban railway because the average total delay does not increase by more than about 55% compared with the optimal solution for any operational scheme.

There has been a growing interest to improve urban rail services for increasing a city’s economic competitiveness. While accessibility provided by rail services has been the focus of many studies, determining accessibility to rail stations is less investigated. The present study examines how accessibility to rail stations via walking has social justice implications. Accessibility can be measured by socioeconomic background, trip characteristics, self-selection, and neighborhood characteristics. Commuters from two rail stations in Auckland, New Zealand—a transit-oriented development (TOD) and a local station—participated in an interactive online survey. Spatial and statistical measures were used to assess the ease of access to the two stations. Under the developed framework, which evaluates social justice in transit, minimum accessibility should be provided (sufficientarianism) and accessibility should disproportionately benefit the less well-off (egalitarianism). The findings show that minimum accessibility was not met for either station, but the TOD station provides better access for low-income individuals. This study contributes by demonstrating the use of a developed framework to evaluate accessibility to rail stations via walking from a social justice perspective. This can be adapted to analyze the accessibility to rail stations in an urban environment of any city.

On-demand mobility services provided by transport network companies (TNCs) have experienced significant growth in their adoption and diversification of services in major metropolitan cities around the world. This study analyzed primary data from Singapore to explore the sociodemographics of TNC users and determine who among TNC users is more likely to pool their trips and what modes these services are replacing. We compared these results with a comprehensive literature review of similar studies of TNC users in the metropolitan US. We found that the sociodemographics of TNC users in general are similar in Singapore and the US: younger, highly educated, and higher income individuals are more likely to have used TNC services. On the other hand, when differentiating by type of TNC service, we found that younger individuals from households that do not own a car are more likely to have pooled in Singapore, whereas employment is an important predictor in the US. We also found differences in mode substitution; whereas TNC trips in the US primarily induce additional trips or replace trips by public and non-motorized transport, in Singapore they primarily replace personal/private vehicle trips. In Singapore, we explored mode substitution by exclusive and pooled TNC services separately, and found that pooled trips draw more from public and nonmotorized transport, whereas exclusive trips replace more personal/private vehicle trips. These results suggest that people in Singapore view exclusive and pooled TNC services as distinct travel options that may be more closely related to other private or public transport, respectively. Differences between Singapore and the US highlight the importance of accounting for local context and suggest that the quality of all travel alternatives in the urban area will affect the mode substitution of TNC trips.

Recurrent bottlenecks in the urban rail transit (URT) system during rush hours have become a significant concern for both passengers and URT operators. In order to alleviate the resulting peak-hour congestion in urban rail transit networks in a more effective and efficient manner, a systematic reliability measure-based data analytics method is developed to identify and rank recurrent bottlenecks at the network level. The reliability measure applied in this study is the frequency of congestion (FoC) measure, which is specially redefined for use in the URT system. The Shanghai Metro system is used to conduct a real-world case study to validate the proposed approach. Based on the passenger trip data collected at $5\text{-}\min$ time intervals by the automatic fare collection (AFC) system and train diagrams on workdays in the Shanghai Metro system, the values of newly defined FoC are computed using two different thresholds for both morning and evening peak periods. Research results of the Shanghai Metro system indicate that it is a feasible and effective way to apply FoC values in identifying and ranking bottlenecks in the URT network. Impacts of using different threshold values in defining FoC is studied, and comparison results between FoC and the peak-hour load factor (PLF), which is the other commonly-used identification indicator, are also provided. The methodology developed and detailed information on bottlenecks presented in this study can greatly help decision makers and operators in the URT system to evaluate the crowding conditions along rail segments and develop targeted bottleneck mitigation solutions in a more effective and efficient manner.

A car-following (CF) model can reproduce various micro traffic phenomena and plays a crucial role in traffic theory. In this study, we combine Markov theory and a gated recurrent unit (GRU) neural network (NN) to propose a new CF model. Next-generation simulation (NGSIM) data were used to generate the Markov chain and train the GRU-NN. Considering the memory effects, we predicted each vehicle’s state at the next time step by the headways and speeds in the last several time steps. Simulations were used to test the merits of the proposed CF model under some given scenarios. The results indicate that the proposed CF model has high accuracy and can enhance the stability of trajectory prediction in simulation, which provides a new approach for micro traffic simulation.

China’s rapid development of high-speed railways has imposed increasing requirements for safety and reliability of signal systems, especially the critical part: turnouts. In this paper, we propose an intelligent fault diagnosis approach that can effectively detect turnout faults based on self-generated fault samples. First, the action mechanism of a switch machine is analyzed and we establish a turnout action model to simulate the turnout operation current curves, thus considerable samples for a following diagnosis can be obtained. Second, we develop a turnout fault diagnosis model based on convolutional neural networks (CNNs). The networks can be trained by those simulated samples. Our experiments verify that the turnout action model can accurately simulate turnout fault curves and the diagnosis model can effectively identify faults through various formats of curve pictures.

Bus bunching deteriorates transit service quality and passengers’ experience. The modeling and prediction of bus operation states are essential for improving the quality of transit service. Due to the nature of traffic evolution and state transition, bunching-oriented modeling based on bus operation state is more intuitive when compared with the headway-based modeling approach. This work explicitly predicted bus operation state by modeling the dynamic evolution of different states. Five different bus operation states were defined and classified by the K-means algorithm, and the dynamic state evolution was formulated as a Markov chain model. Finally, a multinomial logistic model was developed to predict the bus operation state. A case study was designed to test the performance of the proposed model based on the Global Positioning System (GPS) trajectory data collected from four bus routes in Xi’an, China. The results showed that the proposed model was able to accurately predict the bus operation states.

The accurate determination of vehicle delays is crucial for effective intersection management. Because the optimization of signal duration in a signalized intersection is based on the reduction of delays, the rapid and accurate determination of the upcoming delay becomes an important step toward the solution. The delay times can be determined by field studies, but this approach is time-consuming and costly. An estimation of delays using analytical methods is also an approach used by transportation agencies. However, inaccurate predictions of this approach, especially in oversaturated traffic flows, is a significant disadvantage. In order to overcome these problems, estimation models based on artificial intelligence have been developed in recent years, and examples of their application are presented in the literature in which successful results have been achieved in the estimation of vehicle delays. Starting from this point, gene expression programming, an artificial intelligence technique, was used in this study to obtain models that can estimate vehicle delays rapidly and quite accurately. For the selection of this method, the high success rate of the gene expression programming technique for different engineering problems in the literature was crucial. In this study, three delay estimation models were developed by using different parameters. These models were tested with the data collected from 18 different signalized intersections in the Kayseri and Konya provinces in Turkey. When the results obtained were evaluated, it was seen that the models acquired using the gene expression programming technique were very successful in vehicle delay estimation.

Analyzing the influencing factors in expressway traffic crashes is an important aspect of traffic safety analysis. This study examines detailed data from 217 crashes that took place from 2015 to 2016 on the Shaoyang–Xinhuang section of the Shanghai–Kunming Expressway. The data were recorded by the Expressway Administration of Hunan Province, China, which is the only officially available and reliable source of traffic accident data. Descriptive statistics for crash characteristics, vehicle conditions, and road environmental condition data are provided. This paper studies the characteristics of expressway traffic accidents and their influencing factors via the association rule data mining method. The 21 rules obtained from the association rules were analyzed, the accident characteristics of various types of vehicles were studied, and the causes of injury/fatal accidents in various situations were identified. The results of this study will contribute to the targeted enforcement of traffic regulations and the improvement of road facilities to reduce traffic casualties and promote road safety in other areas.