Crash Risk Models Research Articles

Studying the predictability of crash risk caused by manual takeover of autonomous vehicles in mixed traffic flow

ABSTRACT This study explores how to reduce the cost of prediction as much as possible while ensuring the prediction accuracy of a real-time crash risk model. The extreme gradient enhancement (XGBoost) algorithm was used to predict the crash risk of autonomous vehicles in different sections of highway. The results show that the prediction performance of the model is the best when the threshold value is 0.05. Choosing two variables to predict can ensure high accuracy and simultaneously reduce the cost of prediction when the accuracy of crash risk prediction of the three sections can reach 73%, 62%, and 70%. However, when only one variable can be selected due to sensor or system failure, the speed difference between the takeover car and the front car can be chosen to achieve the greatest benefit. These findings could provide a reference for technicians to design safer and more economical autonomous vehicles.

Treating Noise and Anomalies in Vehicle Trajectories From an Experiment With a Swarm of Drones

Unmanned aerial systems, known as “drones”, are relatively new in collecting traffic data. Data from drone videography can have potential applications for traffic research. Drones can record the vehicles from their aerial point-of-view and provide their naturalistic driving behavior. Processing raw data from drones to remove noise and anomalies is crucial to ensure that the data are fit for subsequent applications, e.g., the development of traffic flow or crash risk models. This study uses a part of the pNEUMA dataset, a large dataset with almost half a million trajectories captured by a swarm of drones over Athens, Greece. This novel dataset offers an opportunity to analyze the data attributes and treat the noise and outliers in the data. We use a combination of smoothing filters and Extreme Gradient Boosting with adaptive regularization to process the speed and acceleration profiles of the vehicle trajectories in the dataset. Our approach can help prospective data users treat this or similar trajectory datasets alternatively to applying manual thresholds and assist in accelerating research in microscopic traffic analysis.

Assessing the crash risk of mixed traffic on multilane rural highways using a proactive safety approach

The crash risk assessment of mixed traffic has been understudied due to a lack of pertinent data. In recent years, proactive methods have gained significant attention in transportation safety analysis because of their numerous advantages. In this study, we model and evaluate the effect of speed difference on the sideswipe crash risk of mixed traffic using a novel proactive safety indicator called Anticipated Collision Time (ACT). For analysis, detailed trajectory data were collected using an unmanned aerial vehicle from four-lane and six-lane rural highways. The crash risk was derived from the observed conflict risk and then used to assess the highway system's safety performance under study. Using the Extreme Value Theory (EVT), the conflict risk was mapped to the crash risk. The extreme events were identified using the Block Maxima (BM) approach. Later, the Generalized Extreme Value Distribution (GEV) models were developed for each location by extracting the sideswipe conflicts from the vehicle trajectories. The findings suggest that sideswipe conflicts are more of a safety concern than rear-end conflicts due to the frequent nature of lane changes or passing operations. Also, significant speed difference exists among different vehicle modes present in mixed traffic, and the sideswipe crash risk increases with the increase in the maximum speed difference. The analyses of speed differences also indicate that compared to the four-lane highway, the available safety margin is limited on a six-lane highway due to the higher maximum speed difference. Hence, any driver error may lead to sideswipe crashes. In light of the findings of this study, we strongly suggest the need to provide speed control strategies and restrict the frequent risky lane-changing or passing operations that mainly give rise to sideswipe conflicts on a six-lane highway. Further, this study found that the sideswipe crash risk decreases with the increase in vehicle size on both four-lane and six-lane highways. Hence, we suggest developing separate crash risk models for different vehicle modes in mixed traffic on multilane rural highways. Otherwise, the crash risk mitigation strategies may not be appropriate for mixed traffic conditions.

Identification of adequate sample size for conflict-based crash risk evaluation: An investigation using Bayesian hierarchical extreme value theory models

The use of traffic conflict-based models to estimate crash risk and evaluate the safety of road locations is a popular direction for road safety analysis. However, a challenging issue of traffic conflict-based crash risk modeling is the selection of an appropriate sample size. Reliable conflict-based crash risk models typically require a large sample size which is always very difficult to collect. Further, when choosing a sample size, the bias-variance trade-off of model estimation is a constant concern. This study proposes an approach for identifying an adequate sample size for conflict-based crash risk estimation models. The appropriate sample size is determined by checking the model convergence and its goodness-of-fit. A quantitative approach for objectively testing the model goodness-of-fit is developed. Both the trace plots and the variation tendencies of Brooks-Gelman-Rubin statistics of parameter simulation chains are examined to inspect the model convergence. A graphical method is also used to check the model goodness of fit. If the model has not converged or fits poorly, then additional samples are required. The proposed method was applied to identify the adequate sample size for a Bayesian hierarchical extreme value theory (EVT) block maxima (BM) model using traffic conflict data from four signalized intersections in the city of Surrey, British Columbia. The indicator, modified time to collision (MTTC), was used to delineate traffic conflicts. A series of stationary and non-stationary Bayesian hierarchical BM models were developed using the cycle-level maximums of negated MTTC. The adequate sample sizes of stationary and non-stationary Bayesian hierarchical BM models were determined separately. Further, two methods of increasing the sample size (i.e., extending the observation period and combining data from different sites) were compared in terms of goodness-of-fit as well as crash estimate accuracy and precision. The results show that for both stationary and non-stationary models, the sample size used is adequate for model convergence and goodness-of-fit. Moreover, adding covariates to the stationary Bayesian hierarchical BM model does not affect the size of the required sample. Extending the observation period outperforms combining data from different sites in terms of goodness-of-fit as well as crash estimation accuracy and precision of non-stationary models. This is likely related to the existence of unmeasured factors that could impair model estimation and inference when merging data from several sites to augment the number of samples. Overall, the findings of this study can be applied to examine whether available data is adequate and the amount of additional data required for producing reliable statistical inference.

Effects of design consistency on run-off-road crashes: An application of a Random Parameters Negative Binomial Lindley model

Run-off-road crashes are one of the most common crash types, especially in rural roadway environments contributing significantly to fatalities and severe injuries. These crashes are complex and multi-dimensional events, and factors like road geometry, driver behaviour, traffic characteristics and roadside features contribute to their occurrence, separately or interactively. Sudden changes in road geometry, in particular, can influence driver behaviour, and therefore, in developing a micro-level crash risk model for run-off-road crashes, one of the challenges is incorporating the effects of driver behaviour (disaggregated information) that may arise from the variations in road geometry (aggregated information). This study aims to examine the interaction between road geometry and driver behaviour through a set of measures for design consistency on two-lane rural roads. Multiple data sources, including crash data for 2014–18, traffic data, probe speed data and roadway geometric data, for twenty-three highways in Queensland, Australia, have been fused for this study. Seventeen types of design consistency measures with regard to alignment consistency, operating speed consistency and driving dynamics are tested. A run-off-road crash risk model is estimated by employing the Random Parameters Negative Binomial Lindley regression framework, which accounts for excess zeros in the crash counts and captures the effects of unobserved heterogeneity in the parameter estimates. Results indicate that the geometric design consistency capturing the interaction between driver behaviour and operational factors better predicts run-off-road crashes along rural highways. In addition, roadside attributes like clear zone width, infrastructures, terrain, and roadway remoteness also contribute to run-off-road crashes. The findings of the study provide a comprehensive understanding of the influence of variations in roadway geometry on driver behaviour and run-off-road crashes along rural highways.

Utilizing angle-based outlier detection method with sliding window mechanism to identify real-time crash risk

Developing real-time crash risk models has been a hot research topic as it could identify crash precursors and thus triggering active traffic management strategies. Currently, crash risk identification models were mainly developed based upon supervised learning techniques, which requires large sample size of historical crash data. However, crashes are rare events in the real world, where the performance of supervised learning methods can be severely degraded to deal with the imbalanced sample. Besides, the data heterogeneity issue is another critical challenge. In this study, the unsupervised learning approach has been introduced to address unbalanced samples and data heterogeneity issues, and the experimental results has verified the effectiveness of the method. Data from the Shanghai urban expressway system were utilized for the empirical analyses. Several unsupervised learning methods were tested, among which, Angle-Based Outlier Detection (ABOD) model showed the best performance with 80.4% sensitivity and 25.4% false alarm rate (FAR). Considering the varying traffic flow distribution, dynamic ABOD with sliding window is further proposed, which improves the sensitivity by 6.3% and reduces the FAR by 8.1%. Finally, the proposed model is used to construct personalized road-level models, which achieve good performance despite the small sample size and severe sample imbalance.

A Hierarchical Multinomial Logit model to examine the effects of signal strategies on right-turn crash risks by crash movement configuration

Crash risk models relying on total crash counts are limited in their ability to extract meaningful insights regarding the context of crashes and to identify effective remedial measures. In addition to the typical classification of collisions noted in the literature (e.g., angle, head-on and rear-end), crashes can also be categorised according to vehicle movement configurations (Definitions for Coding Accidents or DCA codes in Australia). This classification presents an opportunity to extract useful insights into road traffic collision causes and contributing factors that are highly contextual. With this aim, this study develops crash-type models by DCA crash movement, with a focus on right-turn crashes (equivalent to left-turn crashes for right-hand traffic) at signalised intersections using a novel approach for linking crashes with signal control strategies. The modelling approach with contextual data enables quantification of the effect of signal control strategies on right-turn crashes, offering potentially unique and novel insights into right-turn crash causes and contributing factors.Crash-type models are estimated with the crash data of 218 signalised intersections in Queensland from 2012 to 2018. Multilevel (Hierarchical) Multinomial Logit Models with random intercepts are employed to capture the hierarchical influence of factors on crashes and unobserved heterogeneities. These models capture upper-level influences on crashes from intersection characteristics and lower-level influences from individual crash characteristics. The models specified in this way account for the correlation among crashes within intersections and influences on crashes across spatial scales. The model results reveal that the probabilities of the opposite approach crash type are significantly higher than the same direction and adjacent approach crash types for all right-turn signal control strategies at intersections except the split approach, for which the opposite is true. The results also suggest that the number of right-turning lanes and occupancy in conflicting lanes are positively associated with the likelihood of crashes for the same direction crash type.

Non-stationary crash risk modelling of powered two-wheelers using extreme value analysis of surrogate crash events

Evaluating the impact of evasive actions such as braking and steering on the crash risk assessment of vehicles is a scarce endeavor due to the lack of relevant data. This study uses Extreme Value Theory to investigate and model the effect of evasive actions on the sideswipe crash risk of powered two-wheelers (PTWs) moving on multilane rural highways. The crash risk was projected from the observed sideswipe conflicts that were quantified using a surrogate safety indicator called anticipated collision time (ACT). The vehicle trajectory data extracted from traffic videos, collected using an unmanned aerial vehicle, was used as the input for the analysis. The data was denoised using a state-of-the-art trajectory reconstruction method called recursively ensembled low pass filtering. Once the conflicts were identified from the trajectory data, the crash risk models were developed considering five covariates: maximum deceleration rate, maximum yaw rate, and the times spent in decelerating, accelerating, and steering during a sideswipe conflict. These covariates were used to capture the non-stationarity in the traffic conflict extremes. The best performing non-stationary model was selected by comparing the negative log-likelihood values with the stationary-one. The findings suggest that the PTWs experience significant sideswipe crash risk on four-lane (crash risk 0.09%) and six-lane (crash risk 0.17%) highways. The sideswipe crash risk of PTWs increases with the increase in the intensity of braking and steering actions measured in terms of maximum deceleration and yaw rates. Further, this study emphasizes that incorporating the effects of evasive actions in the crash risk estimation and developing non-stationary models could significantly improve the precision of crash frequency estimates. Based on the findings it can be concluded that for the safety improvement of PTWs on multilane highways, lane-restriction should be imposed which can increase the safety margin during sideswipe conflicts.

A Bayesian generalised extreme value model to estimate real-time pedestrian crash risks at signalised intersections using artificial intelligence-based video analytics

Pedestrians represent a vulnerable road user group at signalised intersections. As such, properly estimating pedestrian crash risk at discrete short intervals is important for real-time safety management. This study proposes a novel real-time vehicle-pedestrian crash risk modelling framework for signalised intersections. At the core of this framework, a Bayesian Generalised Extreme Value modelling approach is employed to estimate crash risk in real-time from traffic conflicts captured by post encroachment time. A Block Maxima sampling approach, corresponding to a Generalised Extreme Value distribution, is used to identify pedestrian conflicts at the traffic signal cycle level. Several signal-level covariates are used to capture the time-varying heterogeneity of traffic extremes, and the crash risk of different signal cycles is also addressed within the Bayesian framework. The proposed framework is operationalised using a total of 144 hours of traffic movement video data from three signalised intersections in Queensland, Australia. To obtain signal cycle-level covariates, an automated covariate extraction algorithm is used that fuses three data sources (trajectory database from the video feed, traffic conflict database, and signal timing database) to obtain various covariates to explain time-varying crash risk across different cycles. Results show that the model provides a reasonable estimate of historical crash records at the study sites. Utilising the fitted generalised extreme value distribution, the proposed model provides real-time crash estimates at a signal cycle level and can differentiate between safe and risky signal cycles. The real-time crash risk model also helps understand the differential crash risk of pedestrians at a signalised intersection across different periods of the day. The findings of this study demonstrate the potential for the proposed real-time framework in estimating the vehicle-pedestrian crash risk at the signal cycle level, allowing proactive safety management and the development of real-time risk mitigation strategies for pedestrians.

Understanding Crash Risk Using a Multi-Level Random Parameter Binary Logit Model: Application to Naturalistic Driving Study Data

This study presents a framework to employ naturalistic driving study (NDS) data to understand and predict crash risk at a disaggregate trip level accommodating for the influence of trip characteristics (such as trip distance, trip proportion by speed limit, trip proportion on urban/rural facilities) in addition to the traditional crash factors. Recognizing the rarity of crash occurrence in NDS data, the research employs a matched case-control approach for preparing the estimation sample. The study also conducts an extensive comparison of different case-to-control ratios including 1:4, 1:9, 1:14, 1:19, and 1:29. The model parameters estimated with these control ratios are reasonably similar (except for the constant). Employing the 1:9 sample, a multi-level random parameters binary logit model is estimated where multiple forms of unobserved variables are tested including (a) common unobserved effects for each case-control panel, (b) common unobserved factors affecting the error margin in the trip distance variable, and (c) random effects for all independent variables. The estimated model is calibrated by modifying the constant parameter to generate a population conforming crash risk model. The calibrated model is employed to predict crash risk of trips not considered in model estimation. This study is a proof of concept that NDS data can be used to predict trip-level crash risk and can be used by future researchers to develop crash risk models.

Transferability of multivariate extreme value models for safety assessment by applying artificial intelligence-based video analytics

Traffic conflict techniques represent the state-of-the-art for road safety assessments. However, the lack of research on transferability of conflict-based crash risk models, which refers to applying the developed crash risk estimation models to a set of external sites, can reduce their appeal for large-scale traffic safety evaluations. Therefore, this study investigates the transferability of multivariate peak-over threshold models for estimating crash frequency-by-severity. In particular, the study proposes two transferability approaches: (i) an uncalibrated approach involving a direct application of the uncalibrated base model to the target sites and (ii) a threshold calibration approach involving calibration of conflict thresholds of the conflict indicators. In the latter approach, the conflict thresholds of the Modified Time-To-Collision (MTTC) and Delta-V indicators were calibrated using local data from the target sites. Finally, the two transferability approaches were compared with a complete re-estimation approach where all the model parameters were estimated using local data. All three approaches were tested for a target set of signalized intersections in Southeast Queensland, Australia. Traffic movements at the target intersections were observed using video cameras for two days (12 h each day). The road user trajectories and rear-end conflicts were extracted using an automated artificial intelligence-based algorithm utilizing state-of-the-art Computer Vision methods. The base models developed in an earlier study were then transferred to the target sites using the two transferability approaches and the local data from the target sites. Results show that the threshold calibration approach provides the most accurate and precise predictions of crash frequency-by-severity for target sites. Thus, for peak-over threshold models, the threshold parameter is the most important, and its calibration improves the performance of the base models. The complete re-estimation of models for individual target sites yields inferior fits and less precise crash estimates than the two transferability approaches since they utilize fewer traffic conflict extremes in their development than the larger dataset utilized in base model development. Therefore, the study results can significantly advance the applicability of traffic conflict models for crash risk estimation at transport facilities.

Improved Driveway Design for Superblocks to Reduce the Crash Risk

The superblock has become a typical land use in China and many growing Asian cities. Superblock access points generate traffic congestion and many conflicts among all road users. Driveway design is a critical process and has a major impact on traffic conditions around superblocks. There are various guidelines for the two key factors for driveway design, driveway width and curb radius, but they provide reference values corresponding to traffic volume and speed; these are not sufficient for managing the complex traffic environment around superblocks. To improve driveway design, we develop detailed access point design models that account for conflicts between turning and through motorized vehicles, conflicts between motorized and nonmotorized traffic, speed differential larger than 10 km/h, and lane encroachments of entering and exiting vehicles. The crash risk models evaluate and optimize the combination of driveway width and curb radius with respect to three traffic safety indexes: traffic conflict, lane encroachment, and speed differential. A case study evaluation shows that the updated driveway design models produce a lower crash risk; seven of the ten driveways improved by 16.14% or more. The updated driveway design for superblocks would be beneficial for analyzing, permitting, and managing traffic operations at superblocks and oversized development with many and complex driveways.

Transfer learning-based highway crash risk evaluation considering manifold characteristics of traffic flow

The main objective of this study is to evaluate highway crash risk and improve the spatial and temporal transferability of crash risk models. The predictive performance is affected by the difficulty of existing models to quantify crash risk from historical traffic flow data at different locations and times and may not fully capture the complex nonlinear relationships between high-dimensional factors in traffic flow states. Oregon US26W freeway data from 2016 and 2017 and I5N freeway data from 2017 were used. Raw detector data collected from two consecutive detector stations upstream-downstream detector stations were converted into 30 traffic variables. The averages, standard deviations, and coefficients of variation were obtained by aggregating traffic values using each lane. Candidate variables for traffic flow were extracted, and the importance of each variable was calculated using LightGBM, which reveals that variable differences between lanes contributed more. The manifold distance was then applied to quantify the crash risk and classify traffic crashes or not. When the manifold distance is 0.4, it could effectively distinguish traffic crashes. TransferBoost was further employed to build a crash risk model. Modeling using 2016 and 2017 data from the US26W freeway revealed a significant decrease in AUC and a gradual decrease in the model’s sensitivity. However, the crash risk prediction performance of TransferBoost improved by 5.2% when modeling using 2017 data from US26W and I5N freeways. The results show that the model developed for one time period cannot be directly used to predict crash risk for another period on the same freeway. However, the model developed for one highway cannot be directly used to predict the crash risk of another highway either, maintaining some transferability at low false alarm rates. TransferBoost provides a fresh perspective on the transferability of the model. The findings of this study could facilitate more accurate proactive safety management and improvement countermeasure development.

Examining Bayesian network modeling in identification of dangerous driving behavior

Traffic safety problems are still very serious and human factor is the one of most important factors affecting traffic crashes. Taking Next Generation Simulation (NGSIM) data as the research object, this study defines six control indicators and uses principal component analysis and K-means++ clustering methods to get the driving style of different drivers. Then use the Bayesian Networks Toolbox (BNT) and MCMC algorithm to realize the structure learning of Bayesian network. and parameter learning was completed through Netica software. Finally, the vehicle-based traffic crash risk model was created to conduct sensitivity analysis, posterior probability inference, and simulation data was used to detect the feasibility of the model. The results show that the Bayesian network modeling can not only express the relationship between the crash risk and various driving behaviors, but also dig out the inherent relationship between different influencing factors and investigate the causes of driving risks. The results will be beneficial to accurately identify and prevent risky driving behavior.

Improve adaptive cruise control under driver distraction: Time headway compensation based on a random‐effect crash risk model

Human drivers conduct compensatory behaviour to counteract the increased risk while being distracted. This kind of compensation strategy should be learned for a safer and smart design of adaptive cruise control system (ACC). Hence, a simulator study was conducted, requiring performance visual, cognitive, and combined secondary tasks during a car following scenario. An increased time headway (THW) was found in all of the three distraction conditions, which confirmed that drivers compensated an extra THW to counteract the increased crash risk. Furthermore, crash probability models using binary logistic regression with random intercept were constructed where driver distraction and dynamic traffic situations were embodied as inputs. Results showed that crash risk increased with reduced THW, increased lead vehicle deceleration, and unopened brake light of the lead vehicle. Besides, visual-related distractions increased crash risk, while pure cognitive distraction lowered crash risk in low THW (lower than 1.8 s) condition and increased crash risk in high THW (larger than 1.8 s) condition. Based on the authors' proposed models, theoretical compensation in THW to fully counteract the increased crash risk by distraction was derived, which could be used for the design of a human-like ACC with automatic adjustment of THW setting considering driver distraction.

Investigating the predictability of crashes on different freeway segments using the real-time crash risk models

Improvement of the prediction efficiency of crash risks has attracted the attention of numerous studies. Nevertheless, one of the most important factors, crash precursors, were neglected. This study mainly focuses on identifying optimal crash precursors for different freeway section types, as well as providing a threshold selection method for real-time crash risk models. Freeway sections are divided into four types, i.e. basic sections, weaving areas, merging areas, and diverging areas. Bayesian logistic regression (BLR) models were established for each type of segment, and significant factors were distinguished. A threshold selection method was proposed based on cost-benefit theory, and the threshold is determined as the value when the number of proactive safety interventions to prevent a crash is 5000 in this study. BLR models with one, two and three optimal variables were developed. Then the sensitivity and false alarm rate of the models were obtained and compared. Comparison results show that the minimum amount of parameters which can achieve the ideal prediction effectiveness is two. In this situation, 25 %, 50 %, 20 % and 20 % of the crashes occurring at basic sections, weaving areas, merging areas and diverging areas can be accurately predicted respectively. Downstream average speed was recommended as the best crash precursor variable for all the segment types. Support Vector Machine and Random Forest were applied to confirm the conclusion. The conclusion of this paper has the possibility to help reduce crash risk to a relatively economical level in practical applications.

Reinforcement Learning-Based Variable Speed Limits Control to Reduce Crash Risks Near Traffic Oscillations on Freeways

Traffic oscillation on freeways results in abrupt and frequent vehicle deceleration, which remarkably increases crash risks. Artificial Intelligence based traffic control provides a new opportunity for addressing the safety issues. This study aims at proposing a Reinforcement Learning (RL)-based variable speed limits (VSL) control algorithm to reduce crash risks associated with oscillations. The state, action and reward, which are the critical components in the RL, were designed carefully for safety improvement. The RL was trained to learn the optimal speed limit for various traffic states to achieve a goal of safety optimization. A rear-end crash risk model was applied to assess crash risks associated with oscillations near freeway bottlenecks. The cell transmission model was modified as the simulation platform for evaluating the control effects. The results showed that after the training process, the proposed RL-based VSL control successfully reduced the crash risks by 19.4%. A continuous online learning function was developed in RL to enhance the robustness of our strategy. The results showed that with continuous learning, the RL-based VSL control performed reasonably well under lower driver compliance situations.

Application of Connected and Automated Vehicles in a Large-Scale Network by Considering Vehicle-to-Vehicle and Vehicle-to-Infrastructure Technology

Application of connected and automated vehicles (CAVs) is expected to have a significant impact on traffic safety and mobility. Although several studies evaluated the effectiveness of CAVs in a small roadway segment, there is a lack of studies analyzing the impact of CAVs in a large-scale network by considering both freeways and arterials. Therefore, the objective of this study is to analyze the effectiveness of CAVs at the network level by utilizing both vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication technologies. Also, the study proposed a new signal control algorithm through V2I technology to elevate the performance of CAVs at intersections. A car-following model named cooperative adaptive cruise control was utilized to approximate the driving behavior of CAVs in the Aimsun Next microsimulation environment. For the testbed, the research team selected Orlando central business district area in Florida, U.S. To this end, the impacts of CAVs were evaluated based on traffic efficiency (e.g., travel time rate [TTR], speed, and average approach delay, etc.) and safety surrogates (e.g., standard deviation of speed, real-time crash-risk models for freeways and arterials, time exposed time-to-collision). The results showed that the application of CAVs reduced TTR significantly compared with the base condition even with the low market penetration level. Also, the proposed signal control algorithm reduced the approach delay for 94% of the total intersections present in the network. Moreover, safety evaluation results showed a significant improvement of traffic safety in the freeways and arterials under CAV conditions with different market penetration rates.

Modeling crash risk of horizontal curves using large-scale auto-extracted roadway geometry data

Highway horizontal curves (H-curves) provide a smooth transition between two tangent sections of roadways. They allow vehicles to adjust their travel directions gradually. However, the geometry changes of the highway sections with H-curves often raise safety concerns. In order to deploy effective safety countermeasures, a critical task is to understand the risk factors associated with H-curves. Existing studies have made efforts to probe the safety issues associated with H-curves, whereas they were limited to relatively small-scale examinations because of the challenges in identifying H-curves from large road networks. In addition, due to the lack of well-archived traffic and roadway information, gathering other data associated with the H-curves is also difficult. Regarding to these gaps, this study aims to leverage open-source data to analyze the crash risk of highway sections with H-curves. In particular, the present study highlights itself from two main aspects: (i) a H-curve extraction tool was developed to facilitate large-scale curve data collection through the analytics of different open source data; and (ii) a crash modeling framework was developed to quantify H-curve crash risk. A case study based on a statewide road network was performed to test the developed crash risk models with the collected curve data. The results show the opportunities of using the developed tool for large-scale data collection and analyze the safety impacts of H-curve geometric properties, elevation change, traffic exposure, among others. Findings of this study provide insights into the improvement of H-curve data collection and safety evaluation.

Cyclist crash rates and risk factors in a prospective cohort in seven European cities

Increased cycling uptake can improve population health, but barriers include real and perceived risks. Crash risk factors are important to understand in order to improve safety and increase cycling uptake. Many studies of cycling crash risk are based on combining diverse sources of crash and exposure data, such as police databases (crashes) and travel surveys (exposure), based on shared geography and time. When conflating crash and exposure data from different sources, the risk factors that can be quantified are only those variables common to both datasets, which tend to be limited to geography (e.g. countries, provinces, municipalities) and a few general road user characteristics (e.g. gender and age strata). The Physical Activity through Sustainable Transport Approaches (PASTA) project was a prospective cohort study that collected both crash and exposure data from seven European cities (Antwerp, Barcelona, London, Örebro, Rome, Vienna and Zürich). The goal of this research was to use data from the PASTA project to quantify exposure-adjusted crash rates and model adjusted crash risk factors, including detailed sociodemographic characteristics, attitudes about transportation, neighbourhood built environment features and location by city. We used negative binomial regression to model the influence of risk factors independent of exposure. Of the 4,180 cyclists, 10.2 % reported 535 crashes. We found that overall crash rates were 6.7 times higher in London, the city with the highest crash rate, relative to Örebro, the city with the lowest rate. Differences in overall crash rates between cities are driven largely by crashes that did not require medical treatment and that involved motor-vehicles. In a parsimonious crash risk model, we found higher crash risks for less frequent cyclists, men, those who perceive cycling to not be well regarded in their neighbourhood, and those who live in areas of very high building density. Longitudinal collection of crash and exposure data can provide important insights into individual differences in crash risk. Substantial differences in crash risks between cities, neighbourhoods and population groups suggest there is great potential for improvement in cycling safety.