With the increased adoption of connected vehicle (CV) technologies, safety information is becoming increasingly available to drivers. This study investigates three main questions (1) Do CV-based traffic management applications improve safety on roadways with existing infrastructure-based traffic management systems? (2) Can combining two CV technologies have a greater impact on safety than a single CV technology? and (3) Do geometric and traffic composition factors impact the efficiency of CV technologies? We applied a rarely-used CV dataset and conducted a comprehensive simulation analysis of varying conditions and CV penetration rates that studies have not considered. Two CV applications (queue warning and speed harmonization) implemented in the Intelligent Network Flow Optimization experiment in Seattle, WA were evaluated. Results showed that driver safety performance, based on speed metrics (standard deviation and percentage of extreme values) improved under the CV driving conditions. Combining conventional variable speed limit systems with queue warnings also improved safety for CV drivers. Furthermore, the implementation of a single CV application (queue warning) showed positive changes in the aforementioned speed metrics, congestion mitigation, and reduced conflicts. With the two CV applications combined, no significant differences were observed. Additional tests investigated the impacts of lane changes and roadway attributes on safety in the CV environment.

At the intersection of artificial intelligence and urban development, this paper unveils the pivotal role of Foundation Models (FMs) in revolutionizing Intelligent Transportation Systems (ITS). Against the backdrop of escalating urbanization and environmental concerns, we rigorously assess how FMs—spanning large language models, vision-language models, large multimodal models, etc.—can redefine urban mobility paradigms. Our discussion extends to the potential of modular, scalable models and strategic public-private partnerships in facilitating seamless integration. Through a comprehensive literature review and theoretical framework, this paper underscores the significant role of FMs in steering the future of transportation towards unprecedented levels of intelligence and responsiveness. The insights offered aim to guide policymakers, engineers, and researchers in the ethical and effective adoption of FMs, paving the way for a new era in transportation systems.

While many studies have already suggested that CV warnings can positively improve driver safety, a gap remains in the literature on how the interfaces through which these warnings are provided to drivers will influence their behavior and in turn their effectiveness. To fill this gap, we conducted a study to develop a driving simulator-based connected vehicle (CV) environment to investigate the impacts of CV safety warning interfaces on driver behavior. In this study, we reviewed existing studies and guidelines to determine the various driver-vehicle interfaces (DVIs) that have been used to date and developed three modalities of interfaces including auditory, visual, and hybrid modalities. Two scenarios are designated to investigate when: (1) these warnings are active to provide dynamic safety warnings depending on the drivers’ real-time conditions, and (2) they are passive to provide static warnings regardless of the driver’s condition. The DVIs used in those two scenarios are assessed in the application cases of providing curve speed warning and incident zone warning, respectively. Results show that the hybrid conditions (for both passive and active warning modes) encourage safer behavior in terms of the deceleration response time, driving speed, acceleration noise, maximum steering speed, lane change distance, and the time spent above the speed limit, compared to the baseline and the other driving conditions (in which different warning modalities were used). Compared to the baseline, for example, the active hybrid DVIs are most effective in ensuring drivers respond to CV warnings on time, with the potential to reduce the deceleration response time by up to 47% in some cases. Stratifying the data by gender, the active hybrid DVI modality was found to encourage earlier deceleration response, compared to the other DVI modalities for both male and female genders. Female drivers were also found to need shorter exposure to respond to the warnings compared to male drivers. Lastly, the subjective data showed a preference for the visual and hybrid interfaces, with most participants rating the auditory interfaces as distracting or too frequent.

The characteristics of intersection crashes are not only affected by the subject intersection where the crash occurs but also are correlated with environmental conditions of neighboring analysis zones. There are few studies on intersection crash analysis to solve certain spatial effects on microscopic safety issues by proactively incorporating highway safety improvement measures into the long-term transportation planning process. The objective of this paper is to develop a heuristic traffic safety analysis system where spatial spillovers analysis is integrated into roadway safety assessment to incorporate micro variables and macro variables. With K-means clustering technique in a GIS environment, 8 hotspot counties are identified from 88 counties in Ohio, which have high intersection crash propensity. The rest of counties are identified as general counties. Then, an innovative integrated Generalized Linear Model is adopted to identify 11 and 20 significant variables that contribute to the intersection crash propensity in hotspot counties and general counties, respectively. To verify compatibility of intersection crash frequency models with macro-level and micro-level measurement, Reading Road in Cincinnati, Hamilton County (hotspot county) and I-71 in Mason City and Lebanon City of Warren County (general county) are used as examples for the test, and the results show a good consistence.

This study provides insights on enhancing snowplow drivers' training using simulation technologies. Existing literature was reviewed to understand the present snow and ice removal practices of different state Departments of Transportation (DOTs) and corporations, as well as their effectiveness. Two alternatives (in-house capability or contractor training) for setting up a simulation training platform were evaluated using Ohio snowplow drivers as a case study. Based on these, a benefit-cost (BC) analysis was conducted to identify the best alternative. The state-of-the-practice review shows that many DOTs are already adopting simulation training technologies, but with varying degrees of complexities in implementation with reported benefits, including crash reductions, fuel savings, and training cost reductions. Also, a survey of snowplow experts showed that to set up an effective simulation training platform, a driving simulator with powerful motion and visual systems is needed. Finally, a BC analysis showed that the best alternative for agencies is to develop an in-house simulation training capability.

Travel times are an important measure for quantifying travel quality across different modes. However, collecting travel time data is a non-trivial and expensive task. Current practice involves using separate data collection methods for each mode. This paper presents a cost-effective and simple way to collect travel time data across multiple modes using media access control (MAC) matching detected by the mobile unit for sensing traffic (MUST) sensor. This technology detects personal electronic devices to determine people's movement instead of traditional methods which detect singular modes, such as induction loops. This paper proposes a new travel-time calculation method for pedestrian, bicycle, and automobile travelers. A linear model distributes the travel time between different modes by weighting the travel time based on highest, lowest, and most likely speeds. Comparing estimated results for the modal distribution, an accuracy of approximately 83% is achieved, which is acceptable for most applications in transportation engineering.

Freeway capacity analyses are critical for transportation professionals to both assess the current state of a facility’s operations and plan for construction. Such analyses involve converting the demand volume into a demand flow rate in units of passenger car per lane per hour through an adjustment that uses a passenger car equivalent (PCE) factor. This PCE factor represents the equivalent effect heavy vehicles have on capacity in terms of a representative passenger car. Estimation of PCEs for basic freeway segments has been a topic of research for decades. Existing studies have often sought to estimate PCEs based on either small sets of data that are not readily available (eg, individual vehicle headway measurements) or use of traffic microsimulation. Indeed, PCE values used in the 2016 Highway Capacity Manual (HCM) are based on the results of VISSIM simulation. While these approaches have their benefits (for example one can control and adjust many environmental and vehicular factors in a simulation to a very precise degree), they are not without their limitations. Further, such studies have often found results that do not directly agree with those in the HCM or other studies. In order to overcome some of the drawbacks of existing studies, a data-driven approach to estimate PCEs for basic freeway segments operating at capacity is developed herein. The method makes use of large quantities of readily available real-world data, namely, volume and other information obtained from dual loop detectors, in order to estimate PCEs via an equal impedance method.

Predicting traffic crashes has been an important topic of traffic safety research for the past many years. This paper investigates the data from police crash reports provided by the Washington State Department of Transportation. The data consists of records of four years from January 2011 to December 2014 for three main interstate highways (including I-5, I-90, and I-405). A deep learning model using a recurrent neural network (RNN) combined with particle swarm optimization (PSO) is developed and employed to predict the crash density in different severity levels such as property damage only (PDO) and fatal-injury crashes, based on 48,154 crash records that have occurred. All the crash records are randomly divided into training set, validation set, and test set with the proportion ratio of 70, 15, and 15% . The cross-validation is employed to prevent the model from over fit during the training period. A normalized probability-based PSO is designed for optimizing the identified significant factors which can improve the prediction accuracy. The weighted mean squared error (MSE) of the prediction result is employed to measure the performance of the developed model. Nine explanatory variables are selected from fifteen contributing factors. The proposed model is compared with generalized nonlinear model-based mixed multinomial logit approach (GNM-based mixed MNL). The results show that the new model has lower fatal-injury and PDO MSEs. Sensitivity analysis on the selected variables demonstrates the capability of the new model for generating interpretable parameters. The findings of this study provide new insights into the prediction of crash density and severity from the perspective of using roadway segment-based crash records.

A major focus for transportation safety analysts is the development of crash prediction models, a task for which an extremely wide selection of model types is available. Perhaps the most common crash prediction model is the negative binomial (NB) regression model. The NB model gained popularity due to its relative ease of implementation and its ability to handle overdispersion in crash data. Recently, many new models including the Poisson-Inverse-Gaussian, Sichel, Poisson-Lognormal, and Poisson-Weibull models have been introduced as they can also accommodate overdispersion and could potentially replace the NB model, because many have been found to perform better. All five of the aforementioned models, including the NB model, can be classified as mixed-Poisson models. A mixed-Poisson model arises when an error term, following a chosen mixture distribution, enters the functional form for the Poisson parameter. For the NB model, the mixture distribution is selected as gamma, hence the alternate model name of Poisson-Gamma model. In this paper, confidence intervals (CIs) for the Poisson mean ( mu ) as well as prediction intervals (PIs) for the Poisson parameter ( m , alternately referred to as the safety), and the predicted number of crashes at a new site ( y ) are derived for each of the aforementioned types of mixed-Poisson models. After the derivations, the theory is put into practice when CIs and PIs are estimated for mixed-Poisson models developed from an animal-vehicle collision data set. Ultimately, this study provides safety analysts with tools to express levels of uncertainty associated with estimates from safety-modeling efforts instead of simply providing point estimates.

This paper focuses on developing a framework of a vehicle-to-device (V2X) communication system for enhancing vehicle and pedestrian safety at un-signalized intersections. A comprehensive review of the literature has been made to investigate existing V2X safety applications. A cost-effective, solar-energy driven, small, and lightweight communication node device is developed to communicate with connected vehicles (CVs) via LoRa and dedicated short range communications (DSRC), and with pedestrians and unconnected vehicle through cell phones and other mobile devices via Bluetooth. A mobile application that allows pedestrians and drivers of unconnected vehicles to communicate with the communication node device and vice versa is also designed. A crash prediction algorithm is developed to identify unsafe conditions and determine appropriate CV-based safety countermeasures to be presented to system users. Finally, a CV simulation test bed is established in VISSIM to evaluate the safety benefits of the proposed methodology under various traffic and landscape conditions. The simulation results indicate that the number of conflicts increases when the penetration rate of connected devices decreases.

Unmanned aerial vehicles (UAVs) are gaining popularity in traffic monitoring due to their low cost, high flexibility, and wide view range. Traffic flow parameters such as speed, density, and volume extracted from UAV-based traffic videos are critical for traffic state estimation and traffic control and have recently received much attention from researchers. However, different from stationary surveillance videos, the camera platforms move with UAVs, and the background motion in aerial videos makes it very challenging to process for data extraction. To address this problem, a novel framework for real-time traffic flow parameter estimation from aerial videos is proposed. The proposed system identifies the directions of traffic streams and extracts traffic flow parameters of each traffic stream separately. Our method incorporates four steps that make use of the Kanade-Lucas-Tomasi (KLT) tracker, k-means clustering, connected graphs, and traffic flow theory. The KLT tracker and k-means clustering are used for interest-point-based motion analysis; then, four constraints are proposed to further determine the connectivity of interest points belonging to one traffic stream cluster. Finally, the average speed of a traffic stream as well as density and volume can be estimated using outputs from previous steps and reference markings. Our method was tested on five videos taken in very different scenarios. The experimental results show that in our case studies, the proposed method achieves about 96% and 87% accuracy in estimating average traffic stream speed and vehicle count, respectively. The method also achieves a fast processing speed that enables real-time traffic information estimation.

A two-lane highway lane closure work zone is a unique work zone type due to its traffic impact. As one lane of traffic is blocked, it is necessary to implement a traffic control strategy to effectively serve bi-directional traffic. In the sense that the right of way is allocated between two directions sequentially, traffic control at two-lane highway work zones is similar to signalized intersection traffic control. In order to analyze the problem, this study developed two methods: a mathematical capacity and delay model with calculations based on signalized intersection theory, and a VISSIM micro-simulation model calibrated using field observed data. After fine tuning the parameters, the mathematical model was able to make reasonably accurate delay estimates. The study also recommended a smaller vehicle random arrival adjustment in the stochastic delay model compared to Highway Capacity Manual (HCM) 2010 recommend value for signalized intersections. The developed models were applied to optimize two-lane highway lane closure work zone control management. The delay-capacity diagrams indicate that in order to minimize delay, the roadway capacity should be maintained slightly higher than the traffic demand (specifically, the greater of 1.2 times demand or 200 veh/h higher than the traffic demand). Apart from pre-timed traffic control, a dynamic (actuated) traffic control algorithm is also developed in the micro-simulation model to deal with stochastic vehicle arrivals. In the studied traffic scenario,dynamic traffic control is able to achieve lower delay results than the optimal pre-timed signal control.

The mixed multinomial logit (MNL) approach, which can account for unobserved heterogeneity, is a promising unordered model that has been employed in analyzing the effect of factors contributing to crash severity. However, its basic assumption of using a linear function to explore the relationship between the probability of crash severity and its contributing factors can be violated in reality. This paper develops a generalized nonlinear model-based mixed MNL approach which is capable of capturing non-monotonic relationships by developing nonlinear predictors for the contributing factors in the context of unobserved heterogeneity. The crash data on seven Interstate freeways in Washington between January 2011 and December 2014 are collected to develop the nonlinear predictors in the model. Thirteen contributing factors in terms of traffic characteristics, roadway geometric characteristics, and weather conditions are identified to have significant mixed (fixed or random) effects on the crash density in three crash severity levels: fatal, injury, and property damage only. The proposed model is compared with the standard mixed MNL model. The comparison results suggest a slight superiority of the new approach in terms of model fit measured by the Akaike Information Criterion (12.06 percent decrease) and Bayesian Information Criterion (9.11 percent decrease). The predicted crash densities for all three levels of crash severities of the new approach are also closer (on average) to the observations than the ones predicted by the standard mixed MNL model. Finally, the significance and impacts of the contributing factors are analyzed.

Hotspot identification (HSID) is a critical part of network-wide safety evaluations. Typical methods for ranking sites are often rooted in using the Empirical Bayes (EB) method to estimate safety from both observed crash records and predicted crash frequency based on similar sites. The performance of the EB method is highly related to the selection of a reference group of sites (i.e., roadway segments or intersections) similar to the target site from which safety performance functions (SPF) used to predict crash frequency will be developed. As crash data often contain underlying heterogeneity that, in essence, can make them appear to be generated from distinct subpopulations, methods are needed to select similar sites in a principled manner. To overcome this possible heterogeneity problem, EB-based HSID methods that use common clustering methodologies (e.g., mixture models, K-means, and hierarchical clustering) to select “similar” sites for building SPFs are developed. Performance of the clustering-based EB methods is then compared using real crash data. Here, HSID results, when computed on Texas undivided rural highway cash data, suggest that all three clustering-based EB analysis methods are preferred over the conventional statistical methods. Thus, properly classifying the road segments for heterogeneous crash data can further improve HSID accuracy.

This paper aims to develop a crash counts by severity based hotspot identification method by extending the traditional empirical Bayes method to a generalized nonlinear model-based mixed multinomial logit approach. A new safety performance index and a new potential safety improvement index are developed by introducing the risk weight factor and compared with traditional indexes by employing four hotspot identification evaluating methods. The comparison results reveal that the new safety performance index derived by the generalized nonlinear model-based mixed multinomial logit approach is the most consistent and reliable method for identifying hotspots. Finally, a regional map based analytical platform is developed by expanding the safety performance module with the new safety performance index and potential safety improvement functions.

Adjustment of parking price has long been considered an effective way to control parking demand and demand has often been shown to be affected by spatial factors. The primary objective of this study is to investigate the spatial heterogeneity in the sensitivity of parking occupancy to price change using data obtained in downtown San Francisco between 2011 and 2014. The performance-based pricing implemented in the study area allows parking rate to increase, decrease or remain unchanged in neighborhoods with parking occupancy levels higher than, lower than, or within a desired range. As such, the relationship between change in occupancy and change in parking rate is explored. The geographically weighted regression (GWR) method was used to capture the spatial heterogeneity in sensitivity in different blocks and modeling results showed that there is a significant negative correlation between occupancy change and parking rate change. Thus, sensitivity of on-street parking occupancy to price change has an obvious trend of spatial variation. By capturing the spatial heterogeneity in the dataset, the GWR model achieved higher prediction accuracy than a global model. Variables including time of day, block-level features, and socio-demographic characteristics were also found to be correlated with occupancy change. Based on the GWR outputs, a generalized linear model was estimated to further identify how various factors affect sensitivity in different block areas. Findings of this study can be used to help parking authorities with tasks such as identifying which blocks are suitable for balancing parking demand and supply by adjusting price and designing optimal parking rate schemes to achieve desired on-stteet parking occupancy levels. (C) 2017 Elsevier Ltd. All rights reserved.

ABSTRACT The empirical Bayes (EB) method is commonly used by transportation safety analysts for conducting different types of safety analyses, such as before–after studies and hotspot analyses. To date, most implementations of the EB method have been applied using a negative binomial (NB) model, as it can easily accommodate the overdispersion commonly observed in crash data. Recent studies have shown that a generalized finite mixture of NB models with K mixture components (GFMNB-K) can also be used to model crash data subjected to overdispersion and generally offers better statistical performance than the traditional NB model. So far, nobody has developed how the EB method could be used with finite mixtures of NB models. The main objective of this study is therefore to use a GFMNB-K model in the calculation of EB estimates. Specifically, GFMNB-K models with varying weight parameters are developed to analyze crash data from Indiana and Texas. The main finding shows that the rankings produced by the NB and GFMNB-2 models for hotspot identification are often quite different, and this was especially noticeable with the Texas dataset. Finally, a simulation study designed to examine which model formulation can better identify the hotspot is recommended as our future research.