A latent class modeling approach for identifying vehicle driver injury severity factors at highway-railway crossings

Unraveling the relation between cycling accidents and built environment typologies: Capturing spatial heterogeneity through a latent class discrete outcome model

Today, cities seek to transition to more sustainable transportation modes. Cycling is critical in this shift, promoting a more beneficial lifestyle for most. However, cyclists are exposed to many hazardous circumstances or environments, resulting in accidents, injuries, and even death. Transport authorities must understand why accidents occur, to reduce the risk of those who cycle. This study applies a new modeling framework to analyze cycling accident severities. We employ a latent class discrete outcome model, where classes are derived from a Gaussian-Bernoulli mixture, applied to data from Berlin, and augmented with volunteered geographic information. We jointly estimate model components, combining machine learning and econometric approaches, allowing for more intricate and flexible representations while maintaining interpretability. Results show the potential of our approach. Risk factors are indexed depending on where accidents occurred and their contribution. We can discover complex relations between specific built environments and accident characteristics and uncover differences in the impact of certain accident factors on one environment typology but not others. Using multiple data sources also proves helpful as an additional layer of knowledge, providing unique value to understand and model cycling accidents. Another critical aspect of our approach is the potential for simulation, where locations can be examined through simulated accident features to understand the inherent risk of various locations. These findings highlight the ability to capture heterogeneity in accidents and their relation to the built environment. Capturing such relations allows for more direct countermeasures to risky situations or policies to be designed, simulated, and targeted.

Investigating the role of preference variation in the perceptions of railway passengers in Great Britain

This study explores the factors associated with passenger satisfaction on the UK railways. To uncover taste variation, the data was segmented into three homogeneous groups of passengers through a latent class ordered logit model, whereby the class allocation was based on observed personal and trip characteristics. The findings suggest that there is significant variation in the impact of service attributes on overall satisfaction across the segments, 'class a', 'class b' and 'class c'. Class a (15% of the sample) consists of moderately dissatisfied to highly dissatisfied passengers, for whom 'punctuality/reliability' is most impactful on overall satisfaction. Respondents in this class are much more likely to experience adverse service conditions such as delays or crowding conditions. Class b (32% of the sample) consists of passenger who are quite critical and moderately satisfied, for whom 'hedonic' factors such as 'upkeep and repair of the train' and 'seat comfort' were most impactful. Finally, class c (53% of the sample) consists of passengers that are generally satisfied, and for whom the 'value for money of the ticket price' is most impactful on overall satisfaction. Interestingly, for both 'class b' and 'class c', 'punctuality/reliability' plays a more limited role in determining overall satisfaction compared to 'class a'. This suggests that the role of 'punctuality/reliability' in determining overall satisfaction is more complex than presented in the literature thus far. Finally, unobserved taste variation plays an important role in the model, as the class allocation is not always easily linked to observed groups in the data. This paper thus highlights the importance of accounting for unobserved and systematic sources of heterogeneity in the data and could provide useful insights for analysts, policy makers and practitioners, to provide more targeted strategies to improve passenger satisfaction.

Learning about injury severity from no-injury crashes: A random parameters with heterogeneity in means and variances approach

The traditional approach to injury-severity analyses does not allow in-depth understanding of no-injury crashes, as crash factors found to contribute to the various injury severities may have similar effects on the severity of vehicle damage even if no injury is recorded. Viewing no-injury crashes using the vehicle damage severities as sub-categories and bases for potential injuries can improve understanding of future injury crashes. To better understand the mechanism of no-injury crashes and the crash factors that contribute to the extent of vehicle damage beyond the single categorization of these crashes in injury severity analysis, this study presents a vehicle damage severity analysis for no-injury crashes. To compare the effects of crash contributing factors on crash outcomes, two injury severity models were also estimated. Random parameters multinomial logit models with heterogeneity in means and variances were developed to account for unobserved heterogeneity. Model estimation results revealed that several common factors (e.g., unsafe speed, distracted driving, driving under influence, vehicle age, and run-off-road) are correlated with both injury severity in injury crashes and vehicle damage severity in no-injury crashes. Therefore, the sub-categorization of no-injury crashes by vehicle damage severity can potentially improve estimates of injury severity considered in resource allocation decisions for traffic safety.

Relationship between Vehicle Safety Ratings and Drivers' Injury Severity in the Context of Gender Disparity

Previous studies have analyzed the relationship between vehicle safety ratings from impact tests and actual crash injury severity. Nevertheless, no study has investigated the relationship in the context of gender disparity. The main objective of this paper is to explore the validity of the 5-star ratings of the U.S. National Highway Traffic Safety Administration, which describes vehicles’ protectiveness, using actual traffic crash data by gender. Random parameter models are developed using 2015–2020 two-vehicle crash data from Maryland, United States. According to the data, over 90% of vehicles have 4–5 stars in overall, front-impact, and side-impact 5-star ratings. After controlling other factors, it is shown that woman drivers are more likely to be seriously injured in two-vehicle crashes than men drivers when using vehicles with the same 5-star safety ratings. Moreover, there is significant individual heterogeneity in the effect of vehicles with different 5-star safety ratings on driver injury severity. Using vehicles with more stars can reduce the risk of being seriously injured for most man drivers. However, the probability of woman drivers being seriously injured is reduced by approximately 5% on average by using vehicles with higher star ratings in the overall and front-impact 5-star rating, and individual heterogeneity shows a difference of nearly 50% in positive and negative effects. The overall and front-impact 5-star ratings of vehicles could not provide reasonable information as the safety performance of vehicles in traffic crashes for woman drivers. On the other hand, drivers’ residence, driving characteristics, crash types, and environmental characteristics are significantly associated with the injury severity. It is expected that the results from this study will contribute to guide a better vehicle safety design for both men and women.

Bayesian spatial analysis of crash severity data with the INLA approach: Assessment of different identification constraints

Multinomial logit models have been widely used in the analysis of categorical crash data. When the regional information of the data is available, the dependence structure needs to be incorporated into the model to accommodate for spatial heterogeneity. We consider a Bayesian multinomial structured additive regression model to analyze categorical spatial crash data and compare its performance with a fractional split multinomial model. We use the multinomial-Poisson transformation to apply the integrated nested Laplace approximation method for fitting the proposed model efficiently and fast. Moreover, we consider two different types of identifiability constraints to deal with the inherent identifiability problem of the multinomial models. The proposed models are studied through simulated examples and applied to a road traffic crash dataset from Mazandaran province, Iran.

Exploring the impacts of built environment on pedestrian injury severity involving distracted driving

INTRODUCTION

This study develops an injury severity model that demonstrates level of pedestrians' injury severity during pedestrian-vehicle collisions, specifically those involving distracted driving.

METHOD

It uses data from a police-reported collision database that contains pedestrian-vehicle collision information in Nova Scotia, Canada. A latent segmentation-based ordered logit (LSOL) model is developed in this paper that comprehensively examines the influence of built environment characteristics on pedestrian injury severity. It estimates a latent segment allocation model within LSOL modeling framework to capture unobserved heterogeneity across pedestrians. Two latent segments, high- and low-risk segments, are identified probabilistically based on pedestrian characteristics and action, driver action, and collision attributes.

RESULTS

Results suggest that higher mixed land-use, transit stop density, length of sidewalk in the collision locations, and collisions occurring near schools yield lower pedestrian injury severity. In contrast, pedestrian-vehicle collisions in arterial roads, curved roads, sloped roads, and roundabouts tend to result in severe injuries. Interactions between distracted driving type and built environment characteristics are examined in this study. For example, using a communication device while driving on straight roads increases likelihood of higher pedestrian injury severity. This study also confirms the existence of heterogeneity across latent segments. For instance, higher percentage of people commuting by walking in the collision areas yield severe pedestrian injury in high-risk segments and lower injury severity in low-risk segments. Practical applications: The findings of this study will assist transportation planners and road safety stakeholders in developing effective and prioritized policies to reduce pedestrian injury severity involving distracted driving incidents.

How did the COVID-19 pandemic affect road crashes and crash outcomes in Alabama?

With the rising number of cases and deaths from the COVID-19 pandemic, nations and local governments, including many across the U.S., imposed travel restrictions on their citizens. This travel restriction order led to a significant reduction in traffic volumes and a generally lower exposure to crashes. However, recent preliminary statistics in the US suggest an increase in fatal crashes over the period of lockdown in comparison to the same period in previous years. This study sought to investigate how the pandemic affected road crashes and crash outcomes in Alabama. Daily vehicle miles traveled and crashes were obtained and explored. To understand the factors associated with crash outcomes, four crash-severity models were developed: (1) Single-vehicle (SV) crashes prior to lockdown order (Normal times SV); (2) multi-vehicle (MV) crashes prior to lockdown order (Normal times MV); (3) Single-vehicle crashes after lockdown order (COVID times SV); and (4) Multi-vehicle crashes after lockdown order (COVID times MV). The models were developed using the first 28 weeks of crashes recorded in 2020. The findings of the study reveal that although traffic volumes and vehicle miles traveled had significantly dropped during the lockdown, there was an increase in the total number of crashes and major injury crashes compared to the period prior to the lockdown order, with speeding, DUI, and weekends accounting for a significant proportion of these crashes. These observations provide useful lessons for road safety improvements during extreme events that may require statewide lockdown, as has been done with the COVID-19 pandemic. Traffic management around shopping areas and other areas that may experience increased traffic volumes provide opportunities for road safety stakeholders to reduce the occurrence of crashes in the weeks leading to an announcement of any future statewide or local lockdowns. Additionally, increased law enforcement efforts can help to reduce risky driving activities as traffic volumes decrease.

Highway-rail grade crossings accident prediction using Zero Inflated Negative Binomial and Empirical Bayes method

INTRODUCTION

Recently the Federal Railroad Administration (FRA) released a new model for accident prediction at railroad grade crossings using a Zero Inflated Negative Binomial (ZINB) model with Empirical Bayes (EB) adjustments for accident history (2). This new model is adopted from the work that was conducted by the authors (3-6). The unique feature of the new FRA model is that it has a single equation for all three warning devices (crossbuck, flashing light, and gates) and uses the same variables regardless of the warning devices at the crossing. Since the New FRA model incorporates the warning device category as one of the variables in its model equation, the predicted accident frequency is higher when a crossing has crossbucks than flashing lights, and higher when it has flashing lights than gates. While this model is significantly better than the old USDOT model (7), its shortcoming is that the single equation does not accurately represent the field condition.

METHOD

This paper presents the ZINEBS model (Zero Inflated Negative binomial with Empirical Bayes adjustment System). The ZINEBS model gives three different equations depending on the type of warning device used at the crossings (gates, flashing lights, and crossbucks). The three equations use variables, some of which are common across all warning devices, while other variables are specific to a warning device. The predicted values for the ZINEBS model show a closer agreement with the field data than the new FRA model. This observation was true for all three warning device types analyzed. Practical Applications: Based on the results of this study, the ZINEBS compliments the new FRA model and should be used when the single equation is not adequately representing the role of traffic control device types and relevant variables associated with that device type.

Factors affecting injury severity and the number of vehicles involved in a freeway traffic accident: investigating their heterogeneous effects by facility type using a latent class approach

The number of vehicles involved in a traffic accident can be representative of the severity of the accident and provide profound insight into the diverse factors affecting severity, which cannot be identified through the victim fatality rate. This paper presents an analysis and comparison between the effects of factors affecting injury severity and the number of involved vehicles. In this study, a latent class model was used to investigate the unobserved heterogeneity of the accident factors. Freeway facility types are latent factors that affect the heterogeneity of the effects of accident factors. The class mainly including accidents at the freeway mainline sections included more injury/fatal accidents and multiple-vehicle accidents and more significant accident factor estimation results than the other class including accidents at the tollgates or ramps. Among these factors, night-time, faults made by the driver, and heavy vehicle accidents were found to increase the accident severity. Investigating accident factors affecting both the injury severity and number of involved vehicles is important as the number of people who are injured or dead is likely to increase when multiple vehicles are involved in the accident.

Injury-severity analysis of intercity bus crashes in Ghana: A random parameters multinomial logit with heterogeneity in means and variances approach

Travel by bus is an efficient, cost-effective, safe and preferred means of intercity transport in many advanced countries. On the contrary, there is huge public sentiment about the safety records of intercity buses in low- and middle-income countries given the increasing bus-involved road traffic crashes and high fatality rates. This study sought to model the injury severity of intercity bus transport in Ghana using the random parameters multinomial logit with heterogeneity in means and variances modelling technique to account for unobserved heterogeneity in the dataset. The dataset involves crash data from the 575 km long Accra-Kumasi-Sunyani-Gonokrom highway in Ghana. Four discrete crash outcome categories were considered in this study: fatal injury, hospitalized injury, minor injury, and no injury. The study observed that crashes involving pedestrians, unlicensed drivers, and drivers and passengers aged more than 60 years have a higher probability of sustaining fatal injuries. Also, speeding, wrong overtaking, careless driving and inexperienced drivers were associated with fatal injury outcomes on the highway. The incidence of intercity bus transport crashes involving larger buses and minibuses were also found to more likely result in fatalities. The probability of hospitalized injury increased for crashes that occurred in a village setting. Given these findings, the study proposed improvement of the road infrastructure, enforcing seatbelt availability and use in intercity buses, increased enforcement of the traffic rules and regulations to deter driver recklessness and speeding as well as improving the luminance of the highways. Additionally, apps that have features for customers to rate intercity bus operators, the quality of services provided, and also have the option to report reckless driving activities can be developed to promote safe and inclusive public transport in the country.

Geographically weighted poisson regression under linear model of coregionalization assistance: Application to a bicycle crash study

While cycling benefits individuals and society, cyclists are vulnerable road users, and their safety concerns arouse more macro-level spatial crash studies. Our study intends to investigate the spatial effects of population, land use, and bicycle lane infrastructures on bicycle crashes. This was done by developing a semi-parametric Geographically Weighted Poisson Regression (sGWPR) model which deals with the issue of spatial correlation and spatial non-stationarity simultaneously. It is a model that combines both constant and geographically varying parameters. To determine which parameter is fixed or non-stationary, previous studies suggest monitoring the Akaike Information Criterion (AICc). Yet, relying only on AICc might bury some spatial associations. So, in this study, we propose a Linear Model of Coregionalization (LMC) to assist the decision. Here, we use bicycle crash data across the metropolitan area of Greater Melbourne to establish sGWPR models suggested by AICc and LMC, respectively. Comparing the two sGWPR models, we found the sGWPR model under LMC results performs as well as sGWPR models suggested by AICc from the AICc perspective, and a 22.5% improvement in the mean squared error (MSE). It also uncovers more details about the spatial relationship between bicycle crashes and bicycle lane intersection density (BLID), an effect not suggested under AICc results. The parameters of BLID, a new measurement of bicycle lane facilities proposed by us, vary over space across analysis zones in Greater Melbourne.

Correlated mixed logit modeling with heterogeneity in means for crash severity and surrogate measure with temporal instability

This study employs the correlated mixed logit models with heterogeneity in means by accounting for temporal instability to estimate both injury severity and vehicle damage. Two years of intersection crash data from Connecticut were analyzed based on driver characteristics, highway and traffic attributes, environmental variables, vehicle and crash types. These elements were used as independent variables to explore the contributing factors to crash outcome. The likelihood ratio test highlights that the temporal instability exists in both injury severity and vehicle damage models. The model estimation results illustrate that the means of some random parameters are different among crashes. The correlation coefficients of random parameters verify that these random parameters are not always independent, and their correlations should be considered and accounted for in crash severity estimation models. The investigation and comparison between injury severity models and vehicle damage models verify that the injury severity and vehicle damage are highly correlated, and the effects of contributing factors on vehicle damage are consistent with the results of injury severity models. This finding demonstrates that vehicle damage can be used as a potential surrogate measure to injury severity when suffering from a low sample of severe injury crashes in crash severity prediction models. It is anticipated that this study can shed light on selecting appropriate statistical models in crash severity estimation, identifying intersection crash contributing factors, and help develop effective countermeasures to improve intersection safety.

Investigating the effectiveness of safety countermeasures at highway-rail at-grade crossings using a competing risk model

INTRODUCTION

Highway-rail at-grade crossings (HRGCs) are critical locations where a railway and a roadway intersect with one another. Crashes at those locations often result in fatalities and economic and social damages due to the impacts on both road and rail users. The main purpose of countermeasures at HRGCs is to permit safe and efficient rail and highway operations.

METHOD

Countermeasures at highway-rail grade crossings (HRGCs) considered in this study include all traffic control devices and other warning and barrier devices at or on approaches to crossings. In general, active devices are commonly accepted as more effective countermeasures than passive devices. However, many of the previous effectiveness studies are either at the project level or were conducted without considering the before-improvement condition. This study focuses on the network-level marginal effectiveness of countermeasures on crash rate and severity levels during the 29-year study period from 1990 to 2018 by fully considering before-improvement control levels. A competing risk model (CRM) is able to accommodate the competing nature of crash severities as multiple outcomes from the same event of interest, which is crash occurrence in this study. Subsequently, CRM is used in this study as an integrated one-step estimation approach that investigates both crash frequency and severity likelihood over time.

RESULTS

The study findings indicate that adding audible devices to crossings already equipped with gates will result in a considerable annual decline in crash occurrence likelihood (0.25%). The same device installed at crossings already controlled by gates and flashing lights results in less reduction in crash occurrence likelihood of 0.14%. Moreover, adding a stop sign to the active crossing controls of gates, standard flashing lights, and audible devices will lead to a decrease in the probability of crash occurrence and severe crashes (injury and fatal). However, adding stop signs to crossings equipped only with crossbucks will increase the crash occurrence.

Injury severity analysis of pedestrian and bicyclist trespassing crashes at non-crossings: A hybrid predictive text analytics and heterogeneity-based statistical modeling approach

Non-motorists involved in rail-trespassing crashes are usually more vulnerable to receiving major or fatal injuries. Previous research has used traditional quantitative crash data for understanding factors contributing to injury outcomes of non-motorists in train involved collisions. However, usually overlooked crash narratives can provide useful and unique contextual crash-specific information regarding factors associated with injury outcomes. The main objective of this study is to harness the rapid advancements in more sophisticated qualitative analysis procedures for identifying thematic concepts in unstructured crash narrative data. A two-staged hybrid approach is proposed where text mining is applied first to extract valuable information from crash narratives followed by inclusion of the new variables derived from text mining in formulation of advanced statistical models for injury outcomes. By using ten-year (2006-2015) non-motorist non-crossing trespassing injury data obtained from the Federal Railroad Administration, statistical procedures and advanced machine learning text analytics are applied to extract unique information on contributory factors of trespassers' injury outcomes. The key concepts are systematically categorized into trespasser, injury, train, medical, and location related factors. A total of 13 unique variables are extracted from the thematic concepts that are not present in traditional tabular crash data. The analysis reveals a positive statistically significant association between presence of crash narrative and trespasser's injury outcome (coded as minor, major, and fatal injury). Compared to crashes with minor injuries, crashes involving major and fatal injuries are more likely to be reported with crash narratives. A crosstabulation of new variables derived from text mining with injury outcomes revealed that trespassers with confirmed suicide attempts, trespassers wearing headphones, or talking on cell phones are more likely to receive fatal injuries. Among other factors identified, trespassers under alcohol influence, trespasser hit by commuter train, and advance warnings by engineer are associated with more severe (major and fatal) trespasser injury outcomes. Accounting for unobserved heterogeneity and controlling for other factors, fixed and random parameter discrete outcome models are developed to understand the heterogeneous correlations between trespasser injury outcomes and the new crash specific explanatory variables derived from text mining - providing deeper insights. Practical implications and future research directions are discussed.

Investigating hierarchical effects of adaptive signal control system on crash severity using random-parameter ordered regression models incorporating observed heterogeneity

By handling conflicting traffic movements and establishing dynamic coordination between intersections in real-time, the Adaptive Signal Control System (ASCS) can potentially improve the operation and safety of signalized intersections on a corridor. This study identifies the hierarchical effects of ASCS on the crash severity by exploring the heterogeneous effect of ASCS on the crash severity. Four different random-parameter ordered regression models (two ordered probit models, and two ordered logit models) are developed and compared. The analysis reveals that the random-parameter ordered probit and logit models (ROP and ROL) with observed heterogeneity perform better than the random-parameter ordered probit and logit models (RP and RL) without observed heterogeneity in terms of the Akaike information criteria and the goodness of fit of the model. The ROP model performs better than the ROL model in terms of classification model performance measures. The ROP model enables parameters (i.e., the coefficients of the explanatory variables) to vary as a function of explanatory variables as well as across observations, thus accounting for both observed (captured by available explanatory variables) and unobserved (not captured by available explanatory variables) heterogeneity. The analysis reveals that the presence of ASCS is associated with lower crash severity. In this study, observed heterogeneity of ASCS effects on the crash severity is captured by variables related to the intersection and corridor features. Other contributing factors besides ASCS, such as annual average daily traffic, speed limit, lighting, peak period, crash type (rear-end, angle), and pedestrian involvements, are also associated with the probability of crash severity. Unobserved heterogeneity of the effect of angle crash type on the crash severity is found to exist across the observations. The findings of this research have practical implications for establishing ASCS implementation guidelines in lowering the probability of higher crash severity.

Analysis of the Railway Accident-Related Damages in South Korea

Railway accidents are critical issues characterized by a large number of injuries and fatalities per accident due to massive public transport systems. This study proposes a new approach for evaluating the damages resulting from railway accidents using the two-part models (TPMs) such as the zero-inflated Poisson regression model (ZIP model) and the zero-inflated negative-binomial regression model (ZINB model) for the non-negative count measurements and the zero-inflated gamma regression model (ZIG model) and the zero-inflated log-normal regression model (ZILN model) for the semi-continuous measurements. The models are employed for the evaluation of the railway accidents on Korea Railroad, considering the accident damages, such as the train delay time, the number of trains delayed and the cost of considering the accident count responses, for the period 2008 to 2016. From the results obtained, we found that the human-related factors, the high-speed railway system or the Korea Train Express (KTX) and the number of casualties, are the main cost-escalating factors. The number of trains delayed and the amount of delay time tend to increase both the probability of incurring costs and the amount of cost. For better evaluation, the railway accident data should contain accurate information with less recurrence of zeros.

Investigating occupant injury severity of truck-involved crashes based on vehicle types on a mountainous freeway: A hierarchical Bayesian random intercept approach

Earlier research on injury severity of truck-involved crashes focused primarily on single-truck and multi-vehicle crashes with truck involvement, or investigated truck-involved injury severity based on rural and urban locations, time of day variations, lighting conditions, roadway classification, and weather conditions. However, the impact of different vehicle-truck collisions on corresponding occupant injury severity is lacking. Therefore, this paper advances the current research by undertaking an extensive assessment of the occupant injury severity in truck-involved crashes based on vehicle types (i.e., single-truck, truck-car, truck-SUV/pickup, and truck-truck), and identifies the major occupant-, crash-, and geometric-related contributing factors. A series of log-likelihood ratio tests were conducted to justify that separate model by vehicle and occupant types are warranted. Injury severity models were developed using 10 years of crash data (2007-2016) on I-80 in Wyoming through binary logistic modeling with a Bayesian inference approach. The modeling results indicated that there were significant differences between the influences of a variety of variables on the injury severities when the truck-involved crashes are broken down by vehicle types and separated by occupant types. The age and gender of occupants, truck driver occupation, driver residency, sideswipes, presence of junctions, downgrades, curves, and weather conditions were found to have significantly different impacts on the occupant injury severity in different vehicle-truck crashes. Finally, with the incorporation of the random intercept in the modeling procedure, the presence of intra-crash and intra-vehicle correlations (effects of the common crash- and vehicle-specific unobserved factors) in injury severities were identified among persons within the same crash and same vehicle.

A crash severity analysis at highway-rail grade crossings: The random survival forest method

This paper proposes a machine learning approach, the random survival forest (RSF) for competing risks, to investigate highway-rail grade crossing (HRGC) crash severity during a 29-year analysis period. The benefits of the RSF approach are that it (1) is a special type of survival analysis able to accommodate the competing nature of multiple-event outcomes to the same event of interest (here the competing multiple events are crash severities), (2) is able to conduct an event-specific selection of risk factors, (3) has the capability to determine long-term cumulative effects of contributors with the cumulative incidence function (CIF), (4) provides high prediction performance, and (5) is effective in high-dimensional settings. The RSF approach is able to consider complexities in HRGC safety analysis, e.g., non-linear relationships between HRGCs crash severities and the contributing factors and heterogeneity in data. Variable importance (VIMP) technique is adopted in this research for selecting the most predictive contributors for each crash-severity level. Moreover, marginal effect analysis results real several HRGC countermeasures' effectiveness. Several insightful findings are discovered. For examples, adding stop signs to HRGCs that already have a combination of gate, standard flashing lights, and audible devices will reduce the likelihood of property damage only (PDO) crashes for up to seven years; but after the seventh year, the crossings are more likely to have PDO crashes. Adding audible devices to crossing with gates and standard flashing lights will reduce crash likelihood, PDO, injury, and fatal crashes by 49 %, 52 %, 46 %, and 50 %, respectively.

The novel approaches to classify cyclist accident injury-severity: Hybrid fuzzy decision mechanisms

In this study, two novel fuzzy decision approaches, where the fuzzy logic (FL) model was revised with the C4.5 decision tree (DT) algorithm, were applied to the classification of cyclist injury-severity in bicycle-vehicle accidents. The study aims to evaluate two main research topics. The first one is investigation of the effect of road infrastructure, road geometry, street, accident, atmospheric and cyclist related parameters on the classification of cyclist injury-severity similarly to other studies in the literature. The second one is examination of the performance of the new fuzzy decision approaches described in detail in this study for the classification of cyclist injury-severity. For this purpose, the data set containing bicycle-vehicle accidents in 2013-2017 was analyzed with the classic C4.5 algorithm and two different hybrid fuzzy decision mechanisms, namely DT-based converted FL (DT-CFL) and novel DT-based revised FL (DT-RFL). The model performances were compared according to their accuracy, precision, recall, and F-measure values. The results indicated that the parameters that have the greatest effect on the injury-severity in bicycle-vehicle accidents are gender, vehicle damage-extent, road-type as well as the highly effective parameters such as pavement type, accident type, and vehicle-movement. The most successful classification performance among the three models was achieved by the DT-RFL model with 72.0 % F-measure and 69.96 % Accuracy. With 59.22 % accuracy and %57.5 F-measure values, the DT-CFL model, rules of which were created according to the splitting criteria of C4.5 algorithm, gave worse results in the classification of the injury-severity in bicycle-vehicle accidents than the classical C4.5 algorithm. In light of these results, the use of fuzzy decision mechanism models presented in this study on more comprehensive datasets is recommended for further studies.

Bicyclist injury severity in traffic crashes: A spatial approach for geo-referenced crash data to uncover non-stationary correlates

INTRODUCTION

Bicyclists are among vulnerable road users with their safety a key concern. This study generates new knowledge about their safety by applying a spatial modeling approach to uncover non-stationary correlates of bicyclist injury severity in traffic crashes.

METHOD

The approach is Geographically Weighted Ordinal Logistic Regression (GWOLR), extended from the regular Ordered Logistic Regression (OLR) by incorporating the spatial perspective of traffic crashes. The GWOLR modeling approach allows the relationships between injury severity and its contributing factors to vary across the spatial domain, to account for the spatial heterogeneity. This approach makes use of geo-referenced data. This study explored more than 7,000 geo-referenced bicycle--motor-vehicle crashes in North Carolina.

RESULTS

This study performed a series of non-stationarity tests to identify local relationships that vary substantially across the spatial domain. These local relationships are related to the bicyclist (bicyclist age, bicyclist behavior, bicyclist intoxication, bicycle direction, bicycle position), motorist (driver age, driver intoxication, driver behavior, vehicle speed, vehicle type) and traffic (traffic volume).

CONCLUSIONS

Results from the regular OLR are in general consistent with previous findings. For example, an increased bicyclist injury severity is associated with older bicyclists, bicyclist being intoxicated, and higher motor-vehicle speeds. Results from the GWOLR show local (rather than global) relationships between contributing factors and bicyclist injury severity. Practical Applications: Researchers and practitioners may use GWOLR to prioritize cycling safety countermeasures for specific regions. For example, GWOLR modeling estimates in the study highlighted the west part (from Charlotte to Asheville) of North Carolina for increased bicyclist injury severity due to the intoxication of road users including both bicyclists and drivers. Therefore, if a countermeasure is concerned with the road user intoxication, there may be a priority for the region from Charlotte to Asheville (relative to other areas in North Carolina).

Geometric effect analysis of highway-rail grade crossing safety performance

Highway-rail grade crossings (HRGCs) are where a roadway and railway intersect at the same level. Safety at HRGCs has been identified as a high-priority concern among transportation agencies, but there has been little research on the effects of HRGC geometric parameters on their safety performance. This paper evaluates the effects of HRGC geometric parameters on crash occurrence and severity likelihoods. The competing risk algorithm is selected to simultaneously analyze crash occurrence and severities. Four main HRGC geometric factors, along with other contributors, are investigated at 3,194 public HRGCs in North Dakota. This study focuses primarily on four geometric features of an HRGC: (1) acute crossing angle, (2) number of tracks (indicator of crossing width), (3) the roadway distance between the HRGC and the signalized intersection, and (4) number of highway lanes. Distance to the nearest roadway intersections and highway-railway crossing angles are map-based calculations drawn from geographic information systems (GIS). The findings are: (1) all contributors tested in this study, including highway characteristics, traffic exposures from both railway and highway, and the four geometric features, significantly affect at least one crash severity level; (2) all contributors significantly impact crash frequency except for the distance between crossings and the nearest roadway intersection; and (3) geometric parameters' long-term effects on cumulative probability of crash severity and occurrence over 30 years is also evaluated. Crossings with three main tracks contribute the highest long-term crash probabilities.

Analysis of the effects of visibility and warning devices on driver injury severity at highway-rail grade crossings considering temporal transferability of data

This study explores the joint effect of visibility and warning devices on driver injury severity at the highway-rail grade crossings (HRGCs), while also considering other contributing factors. For this purpose, four mixed logit models are developed to estimate the determinants of driver injury severity considering the combinations of visibility conditions (daylight vs. no daylight) and type of warning devices (active vs. passive warning). The models were calibrated using the data obtained from the USDOT Federal Railroad Administration for HRGC crashes that occurred over a ten-year period 2008-2017 across the United States. A temporal transferability test was conducted and confirmed the stability of model specifications considering a ten-year span of collected data. The pseudo-elasticity analysis was conducted to ascertain marginal impact of the contributing factors on driver injury severity in each model. While the vehicle speed, train speed, time of day and driver age are found to be common significant factors among the four models, there are marked differences between parameters associated with various crash factors. The study provides new insight into the driver injury severity in train-vehicle collisions considering visibility and type of warning devices, which can help in setting up proper policies to improve safety at HRGCs.

Investigation of injury severity in urban expressway crashes: A case study from Beijing

Urban expressway is the main artery of traffic network, and an in-depth analysis of the crashes is crucial for improving the traffic safety level of expressways. This study intended to address the injury severity of expressways in Beijing by proposing Bayesian ordered logistic regression model. Crash data were collected from urban express rings and expressways in 2015 and 2016. The results showed that crash location, time and crash season are significant variables influencing injury severity. The findings revealed that the proposed model can address the ordinal feature of injury severity, while accommodating the data with small sample sizes that may not adequately represent population characteristics. The conclusions can provide the management departments with valuable suggestions for the injury prevention and safety improvement on the urban expressways.

Segmentation of Severe Occupational Incidents in Agribusiness Industries Using Latent Class Clustering

One of the principle objectives in occupational safety analysis is to identify the key factors that affect the severity of an incident. To identify risk groups of occupational incidents and the factors associated with them, statistical analysis of workers’ compensation claims data is performed using latent class clustering, for the segmentation of 1031 severe occupational incidents in agribusiness industries in the Midwest region of the United States between 2008–2016. In this study, severe incidents are those with workers’ compensation costs equal to or greater than $100,000 (USD). Based on the latent class clustering results, three risk groups are identified with injury nature as the most statistically distinctive classifier. The highest cost injuries include strain, tear, fracture, contusion, amputation, laceration, burn, concussion, and crushing. The most prevalent and statistically significant injury type is permanent partial disability. The study introduces a novel application of latent class clustering in the segmentation of high severity occupational incidents. The analytical approach and results of this study will aid safety practitioners in identifying occupational risk groups and analyzing injury patterns, and inform safety intervention plans to avoid the occurrence of similar incidents in agribusiness industries.

Using latent class analysis and mixed logit model to explore risk factors on driver injury severity in single-vehicle crashes

The single-vehicle crash has been recognized as a critical crash type due to its high fatality rate. In this study, a two-year crash dataset including all single-vehicle crashes in New Mexico is adopted to analyze the impact of contributing factors on driver injury severity. In order to capture the across-class heterogeneous effects, a latent class approach is designed to classify the whole dataset by maximizing the homogeneous effects within each cluster. The mixed logit model is subsequently developed on each cluster to account for the within-class unobserved heterogeneity and to further analyze the dataset. According to the estimation results, several variables including overturn, fixed object, and snowing, are found to be normally distributed in the observations in the overall sample, indicating there exist some heterogeneous effects in the dataset. Some fixed parameters, including rural, wet, overtaking, seatbelt used, 65 years old or older, etc., are also found to significantly influence driver injury severity. This study provides an insightful understanding of the impacts of these variables on driver injury severity in single-vehicle crashes, and a beneficial reference for developing effective countermeasures and strategies for mitigating driver injury severity.

Crash severity effects of adaptive signal control technology: An empirical assessment with insights from Pennsylvania and Virginia

Adaptive signal control technology (ASCT) is an intelligent transportation systems (ITS) technology that optimizes signal timings in real time to improve corridor flow. While a few past studies have examined the impact of ASCT on crash frequency, little is known about its effect on injury severity outcomes. Similarly, the impact of different types of ASCTs deployed across different states is also uncertain. This paper therefore, used ordered probit models with random parameters to estimate the injury severity outcomes resulting from ASCT deployment across Pennsylvania and Virginia. Two disparate systems deployed across the two different states were analyzed to assess whether they had similar impacts on injury severity, although signal timings are optimized using different algorithms by both systems. The estimation results revealed that both ASCT systems were associated with reductions in injury severity levels. Marginal effects showed that Type A ASCT systems reduced the propensity of severe plus moderate and minor injury crashes by 11.70% and 10.36% while type B ASCT reduced the propensity of severe plus moderate and minor injury crashes by 4.39% and 6.92%. Similarly, the ASCTs deployed across the two states were also observed to reduce injury severities. The combined best fit model also revealed a similar trend towards reductions in severe plus moderate and minor injury crashes by 5.24% and 9.91%. This model performed well on validation data with a low forecast error of 0.301 and was also observed to be spatially transferable. These results encourage the consideration of ASCT deployments at intersections with high crash severities and have practical implications for aiding agencies in making future deployment decisions about ASCT.

Accommodating exogenous variable and decision rule heterogeneity in discrete choice models: Application to bicyclist route choice

The proposed research contributes to our understanding of incorporating heterogeneity in discrete choice models with respect to exogenous variables and decision rules. Specifically, the proposed latent segmentation based mixed models segment population to different classes with their own decision rules while also incorporating unobserved heterogeneity within the segment level models. In our analysis, we choose to consider both random utility and random regret theories. Further, instead of assuming the number of segments (as 2), we conduct an exhaustive exploration with multiple segments across the two decision rules. The model estimation is conducted using a stated preference data from 695 commuter cyclists compiled through a web-based survey. The probabilistic allocation of respondents to different segments indicates that female commuter cyclists are more utility oriented; however, the majority of the commuter cyclist’s choice pattern is consistent with regret minimization mechanism. Overall, cyclists’ route choice decisions are influenced by roadway attributes, cycling infrastructure availability, pollution exposure, and travel time. The analysis approach also allows us to investigate time based trade-offs across cyclists belonging to different classes. Interestingly, we observe that the trade-off values in regret and utility based segments for roadway attributes are similar in magnitude; but the values differ greatly for cycling infrastructure and pollution exposure attributes, particularly for maximum exposure levels.

Examining driver injury severity in intersection-related crashes using cluster analysis and hierarchical Bayesian models

Traffic crashes are more likely to occur at intersections where the traffic environment is complicated. In this study, a hybrid approach combining cluster analysis and hierarchical Bayesian models is developed to examine driver injury severity patterns in intersection-related crashes based on two-year crash data in New Mexico. Three clusters are defined by K-means cluster analysis based on weather and roadway environmental conditions in order to reveal drivers' risk compensation instability under diverse external environment. Hierarchical Bayesian random intercept models are developed for each of the three clusters as well as the whole dataset to identify the contributing factors on multilevel driver injury outcomes: property damage only (Level I), complaint of injury and visible injury (Level II), and incapacitating injury and fatality (Level III). Model comparison with an ordinary multinomial logistic model omitting crash data hierarchical features and cross-level interactions verifies the suitability and effectiveness of the proposed hybrid approach. Results show that a number of crash-level variables (time period, weather, light condition, area, and road grade), vehicle/driver-level variables (traffic controls, vehicle action, vehicle type, seatbelt used, driver age, drug/alcohol impaired, and driver age) along with some cross-level interactions (i.e., left turn and night, drug and dark) impose significantly influence driver injury severity. This study provides insightful understandings of the effects of these variables on driver injury severity in intersection-related crashes and beneficial references for developing effective countermeasures for severe crash prevention.

An investigation on fatality of drivers in vehicle-fixed object accidents on expressways in China: Using multinomial logistic regression model

This study aims to identify the effects of characteristics of vehicle, roadway, driver, and environment on fatality of drivers in vehicle-fixed object accidents on expressways in Changsha-Zhuzhou-Xiangtan district of Hunan province in China by developing multinomial logistic regression models. For this purpose, 121 vehicle-fixed object accidents from 2011-2017 are included in the modeling process. First, descriptive statistical analysis is made to understand the main characteristics of the vehicle-fixed object crashes. Then, 19 explanatory variables are selected, and correlation analysis of each two variables is conducted to choose the variables to be concluded. Finally, five multinomial logistic regression models including different independent variables are compared, and the model with best fitting and prediction capability is chosen as the final model. The results showed that the turning direction in avoiding fixed objects raised the possibility that drivers would die. About 64% of drivers died in the accident were found being ejected out of the car, of which 50% did not use a seatbelt before the fatal accidents. Drivers are likely to die when they encounter bad weather on the expressway. Drivers with less than 10 years of driving experience are more likely to die in these accidents. Fatigue or distracted driving is also a significant factor in fatality of drivers. Findings from this research provide an insight into reducing fatality of drivers in vehicle-fixed object accidents.

Latent class analysis of accident risks in usage-based insurance: Evidence from Beijing

Car insurance is quickly becoming a big data industry, with usage-based insurance (UBI) poised to potentially change the business of insurance. Telematics data, which are transmitted from wireless devices in car, are widely used in UBI to obtain individual-level travel and driving characteristics. While most existing studies have introduced telematics data into car insurance pricing, the telematics-related characteristics are directly obtained from the raw data. In this study, we propose to quantify drivers' familiarity with their driving routes and develop models to quantify drivers' accident risks using the telematics data. In addition, we build a latent class model to study the heterogeneity in travel and driving styles based on the telematics data, which has not been investigated in literature. Our main results include: (1) the improvement to the model fit is statistically significant by adding telematics-related characteristics; (2) drivers' familiarity with their driving trips is critical to identify high risk drivers, and the relationship between drivers' familiarity and accident risks is non-linear; (3) the drivers can be classified into two classes, where the first class is the low risk class with 0.54% of its drivers reporting accidents, and the second class is the high risk class with 20.66% of its drivers reporting accidents; and (4) for the low risk class, drivers with high probability of reporting accidents can be identified by travel-behavior-related characteristics, while for the high risk class, they can be identified by driving-behavior-related characteristics. The driver's familiarity will affect the probability of reporting accidents for both classes.

A semi-nonparametric Poisson regression model for analyzing motor vehicle crash data

This paper develops a semi-nonparametric Poisson regression model to analyze motor vehicle crash frequency data collected from rural multilane highway segments in California, US. Motor vehicle crash frequency on rural highway is a topic of interest in the area of transportation safety due to higher driving speeds and the resultant severity level. Unlike the traditional Negative Binomial (NB) model, the semi-nonparametric Poisson regression model can accommodate an unobserved heterogeneity following a highly flexible semi-nonparametric (SNP) distribution. Simulation experiments are conducted to demonstrate that the SNP distribution can well mimic a large family of distributions, including normal distributions, log-gamma distributions, bimodal and trimodal distributions. Empirical estimation results show that such flexibility offered by the SNP distribution can greatly improve model precision and the overall goodness-of-fit. The semi-nonparametric distribution can provide a better understanding of crash data structure through its ability to capture potential multimodality in the distribution of unobserved heterogeneity. When estimated coefficients in empirical models are compared, SNP and NB models are found to have a substantially different coefficient for the dummy variable indicating the lane width. The SNP model with better statistical performance suggests that the NB model overestimates the effect of lane width on crash frequency reduction by 83.1%.

Latent class analysis of factors that influence weekday and weekend single-vehicle crash severities

This paper investigates factors that influence the severity of single-vehicle crashes that happen on weekdays and weekends. Crash data from 2012 to 2016 for the State of Alabama was used for this study. Latent class logit models were developed as alternative to the frequently used random parameters models to account for unobserved heterogeneity across crash-severity observations. Exploration of the data revealed that a high proportion of severe injury injury crashes happened on weekends. The study examined whether single-vehicle crash contributing factors differ between weekdays and weekends. The model estimation results indicate a significant association of severe injury crashes to risk factors such as driver unemployment, driving with invalid license, no seatbelt use, fatigue, driving under influence, old age, and driving on county roads for both weekdays and weekends. Research findings show a strong link between human factors and the occurrence of severe injury single-vehicle crashes, as it has been observed that many of the factors associated with severe-injury outcome are driver behavior related. To illustrate the significance of the findings of this study, a third model using the combined data was developed to explore the merit of using sub-populations of the data for improved and detailed segmentation of the crash-severity factors. It has also been shown that generally, the factors that influence single-vehicle crash injury outcomes were not very different between weekdays and weekends. The findings of this study show the importance of investigating sub-populations of data to reveal complex relationships that should be understood as a necessary step in targeted countermeasure application.

Application of Fractal theory for crash rate prediction: Insights from random parameters and latent class tobit models

The repercussions from congestion and accidents on major highways can have significant negative impacts on the economy and environment. It is a primary objective of transport authorities to minimize the likelihood of these phenomena taking place, to improve safety and overall network performance. In this study, we use the Hurst Exponent metric from Fractal Theory, as a congestion indicator for crash-rate modeling. We analyze one month of traffic speed data at several monitor sites along the M4 motorway in Sydney, Australia and assess congestion patterns with the Hurst Exponent of speed (H_speed). Random Parameters and Latent Class Tobit models were estimated, to examine the effect of congestion on historical crash rates, while accounting for unobserved heterogeneity. Using a latent class modeling approach, the motorway sections were probabilistically classified into two segments, based on the presence of entry and exit ramps. This will allow transportation agencies to implement appropriate safety/traffic countermeasures when addressing accident hotspots or inadequately managed sections of motorway.

Impact of roadway geometric features on crash severity on rural two-lane highways

This study examines the impact of a wide range of roadway geometric features on the severity outcomes of crashes occurred on rural two-lane highways. We argue that crash data have a hierarchical structure which needs to be addressed in modeling procedure. Moreover, most of previous studies ignored the impact of geometric features on crash types when developing crash severity models. We hypothesis that geometric features are more likely to determine crash type, and crash type together with other occupant, environmental and vehicle characteristics determine crash severity outcome. This paper presents an application of multilevel models to successfully capture both hierarchical structure of crash data and indirect impact of geometric features on crash severity. Using data collected in Illinois from 2007 to 2009, multilevel ordered logit model is developed to quantify the impact of geometric features and environmental conditions on crash severity outcome. Analysis results revealed that there is a significant variation in severity outcomes of crashes occurred across segments which verifies the presence of hierarchical structure. Lower risk of severe crashes is found to be associated with the presence of 10-ft lane and/or narrow shoulders, lower roadside hazard rate, higher driveway density, longer barrier length, and shorter barrier offset. The developed multilevel model offers greater consistency with data generating mechanism and can be utilized to evaluate safety effects of geometric design improvement projects.

Barrier-relevant crash modification factors and average costs of crashes on arterial roads in Indiana

The objective of this study was to develop crash modification factors (CMFs) and estimate the average crash costs applicable to a wide range of road-barrier scenarios that involved three types of road barriers (concrete barriers, W-beam guardrails, and high-tension cable barriers) to produce a suitable basis for comparing barrier-oriented design alternatives and road improvements. The intention was to perform the most comprehensive and in-depth analysis allowed by the cross-sectional method and the crash data available in Indiana. To accomplish this objective and to use the available data efficiently, the effects of barrier were estimated on the frequency of barrier-relevant (BR) crashes, the types of harmful events and their occurrence during a BR crash, and the severity of BR crash outcomes. The harmful events component added depth to the analysis by connecting the crash onset with its outcome. Further improvement of the analysis was accomplished by considering the crash outcome severity of all the individuals involved in a crash and not just drivers, utilizing hospital data, and pairing the observations with and without road barriers along same or similar road segments to better control the unobserved heterogeneity. This study confirmed that the total number of BR crashes tended to be higher where medians had installed barriers, mainly due to collisions with barriers and, in some cases, with other vehicles after redirecting vehicles back to traffic. These undesirable effects of barriers were surpassed by the positive results of reducing cross-median crashes, rollover events, and collisions with roadside hazards. The average cost of a crash (unit cost) was reduced by 50% with cable barriers installed in medians wider than 50ft. A similar effect was concluded for concrete barriers and guardrails installed in medians narrower than 50ft. The studied roadside guardrails also reduced the unit cost by 20%-30%. Median cable barriers were found to be the most effective among all the studied barriers due to the smaller increase in the crash frequency caused by these barriers and the less severe injury outcomes. More specifically, the occupants of vehicles colliding with near-side cable barriers tended to have less severe injuries than occupants of vehicles entering the median from median's farther side. The near-side cable barriers provided protection against rollover inside the median and against a potentially dangerous collision with or running over the median drain; therefore, the greatest safety benefit can be expected where cable barriers are installed at both edges of the median. The CMFs and unit crash costs for 48 road-barrier scenarios produced in this study are included in this paper.

Gate-violation behavior at highway-rail grade crossings and the consequences: Using geo-Spatial modeling integrated with path analysis

Drivers undertaking risky behaviors at highway-rail grade crossings are often severely injured in collisions with trains. Among these behaviors, gate-violation (referring to driving around or through the gates that were activated and lowered by an approaching train) seems to be one of the most dangerous actions a driver might take at a gated crossing; it may compromise the intended safety improvement made by adding gates at crossings. This study develops a nuanced conceptual framework that uses path analysis to explore the contributing factors to gate-violation behaviors and the correlation between gate-violation behaviors and the crash consequence - the driver injury severity. Further, using geo-spatial modeling techniques, this study explores whether the correlates of gate-violation behaviors and their associations with injury severity are stationary across diverse geographic contexts of the United States. Geo-spatial modeling shows that the correlates of gate-violation and its associations with injury severity vary substantially across the United States. Spatial variations in correlates of gate-violation and injury severity are mapped by estimating geographically weighted regressions; the maps can serve as an instrument for screening safety improvements and help identify regions that need safety improvements. For example, the results show that two-quadrant gates are more likely to have gate-violation crashes than four-quadrant gates in Iowa, Illinois, Wisconsin and Minnesota. These states may need to receive more attentions on the enforcement of inhibiting gate-violation at crossings with two-quadrant gates or have the priority over other states to upgrade these crossings to four-quadrant gates if financially feasible.

Comparison of four statistical and machine learning methods for crash severity prediction

Crash severity prediction models enable different agencies to predict the severity of a reported crash with unknown severity or the severity of crashes that may be expected to occur sometime in the future. This paper had three main objectives: comparison of the performance of four statistical and machine learning methods including Multinomial Logit (MNL), Nearest Neighbor Classification (NNC), Support Vector Machines (SVM) and Random Forests (RF), in predicting traffic crash severity; developing a crash costs-based approach for comparison of crash severity prediction methods; and investigating the effects of data clustering methods comprising K-means Clustering (KC) and Latent Class Clustering (LCC), on the performance of crash severity prediction models. The 2012-2015 reported crash data from Nebraska, United States was obtained and two-vehicle crashes were extracted as the analysis data. The dataset was split into training/estimation (2012-2014) and validation (2015) subsets. The four prediction methods were trained/estimated using the training/estimation dataset and the correct prediction rates for each crash severity level, overall correct prediction rate and a proposed crash costs-based accuracy measure were obtained for the validation dataset. The correct prediction rates and the proposed approach showed NNC had the best prediction performance in overall and in more severe crashes. RF and SVM had the next two sufficient performances and MNL was the weakest method. Data clustering did not affect the prediction results of SVM, but KC improved the prediction performance of MNL, NNC and RF, while LCC caused improvement in MNL and RF but weakened the performance of NNC. Overall correct prediction rate had almost the exact opposite results compared to the proposed approach, showing that neglecting the crash costs can lead to misjudgment in choosing the right prediction method.

To stop or not to stop: Contrasting compliant and non-compliant driver behaviour at rural rail level crossings

Many rail level crossings (RLXs) have only passive protection, such as static signs instructing road users to stop, yield, or look for trains. Stop signs have been suggested as a low-cost option to improve safety at passive RLXs, as requiring drivers to stop should encourage safe behaviour. However, field observations have noted high rates of non-compliance at stop-controlled RLXs. To explore this further, we conducted an on-road study to identify factors that influence compliance at stop-controlled RLXs. Twenty-two drivers drove a 30.5km route in rural Australia, encompassing three stop-controlled RLXs. In over half of all cases (59%) drivers stopped completely at the RLX; on 27% of crossings drivers executed a rolling stop, and on 14% of crossings drivers violated the stop controls. Rolling stops were defined as a continuous deceleration to <10km/h, but remaining above 0km/h, before accelerating to >10km/h. Behavioural patterns, including visual checks and decision-making, were similar when comparing drivers who made complete versus rolling stops. Non-compliant drivers did not differ from compliant drivers in approach speeds, but spent less time visually checking for trains. Post-drive interviews revealed some drivers wilfully disregarded the stop sign, whereas others did not notice the stop sign. Those who intentionally violated noted trains were infrequent and suggested sight distance was good enough (even though all crossings had been formally assessed as having inadequate sight distance). Overall the results suggest most drivers exhibit safe behaviour at passive RLXs, but a notable minority disregard or fail to notice signs. Potential avenues for redesigning passive RLXs to improve safety are discussed.

Using fixed-parameter and random-parameter ordered regression models to identify significant factors that affect the severity of drivers' injuries in vehicle-train collisions

This study attempts to identify significant factors that affect the severity of drivers' injuries when colliding with trains at railroad-grade crossings by analyzing the individual-specific heterogeneity related to those factors over a period of 15 years. Both fixed-parameter and random-parameter ordered regression models were used to analyze records of all vehicle-train collisions that occurred in the United States from January 1, 2001 to December 31, 2015. For fixed-parameter ordered models, both probit and negative log-log link functions were used. The latter function accounts for the fact that lower injury severity levels are more probable than higher ones. Separate models were developed for heavy and light-duty vehicles. Higher train and vehicle speeds, female, and young drivers (below the age of 21 years) were found to be consistently associated with higher severity of drivers' injuries for both heavy and light-duty vehicles. Furthermore, favorable weather, light-duty trucks (including pickup trucks, panel trucks, mini-vans, vans, and sports-utility vehicles), and senior drivers (above the age of 65 years) were found be consistently associated with higher severity of drivers' injuries for light-duty vehicles only. All other factors (e.g. air temperature, the type of warning devices, darkness conditions, and highway pavement type) were found to be temporally unstable, which may explain the conflicting findings of previous studies related to those factors.

Incorporating real-time traffic and weather data to explore road accident likelihood and severity in urban arterials

INTRODUCTION

The effective treatment of road accidents and thus the enhancement of road safety is a major concern to societies due to the losses in human lives and the economic and social costs. The investigation of road accident likelihood and severity by utilizing real-time traffic and weather data has recently received significant attention by researchers. However, collected data mainly stem from freeways and expressways. Consequently, the aim of the present paper is to add to the current knowledge by investigating accident likelihood and severity by exploiting real-time traffic and weather data collected from urban arterials in Athens, Greece.

METHOD

Random Forests (RF) are firstly applied for preliminary analysis purposes. More specifically, it is aimed to rank candidate variables according to their relevant importance and provide a first insight on the potential significant variables. Then, Bayesian logistic regression as well finite mixture and mixed effects logit models are applied to further explore factors associated with accident likelihood and severity respectively.

RESULTS

Regarding accident likelihood, the Bayesian logistic regression showed that variations in traffic significantly influence accident occurrence. On the other hand, accident severity analysis revealed a generally mixed influence of traffic variations on accident severity, although international literature states that traffic variations increase severity. Lastly, weather parameters did not find to have a direct influence on accident likelihood or severity.

CONCLUSIONS

The study added to the current knowledge by incorporating real-time traffic and weather data from urban arterials to investigate accident occurrence and accident severity mechanisms.

PRACTICAL APPLICATION

The identification of risk factors can lead to the development of effective traffic management strategies to reduce accident occurrence and severity of injuries in urban arterials.

A Hybrid Latent Class Analysis Modeling Approach to Analyze Urban Expressway Crash Risk

Crash risk analysis is rising as a hot research topic as it could reveal the relationships between traffic flow characteristics and crash occurrence risk, which is beneficial to understand crash mechanisms which would further refine the design of Active Traffic Management System (ATMS). However, the majority of the current crash risk analysis studies have ignored the impact of geometric characteristics on crash risk estimation while recent studies proved that crash occurrence risk was affected by the various alignment features. In this study, a hybrid Latent Class Analysis (LCA) modeling approach was proposed to account for the heterogeneous effects of geometric characteristics. Crashes were first segmented into homogenous subgroups, where the optimal number of latent classes was identified based on bootstrap likelihood ratio tests. Then, separate crash risk analysis models were developed using Bayesian random parameter logistic regression technique; data from Shanghai urban expressway system were employed to conduct the empirical study. Different crash risk contributing factors were unveiled by the hybrid LCA approach and better model goodness-of-fit was obtained while comparing to an overall total crash model. Finally, benefits of the proposed hybrid LCA approach were discussed.

Characteristics of cyclist crashes in Italy using latent class analysis and association rule mining

The factors associated with severity of the bicycle crashes may differ across different bicycle crash patterns. Therefore, it is important to identify distinct bicycle crash patterns with homogeneous attributes. The current study aimed at identifying subgroups of bicycle crashes in Italy and analyzing separately the different bicycle crash types. The present study focused on bicycle crashes that occurred in Italy during the period between 2011 and 2013. We analyzed categorical indicators corresponding to the characteristics of infrastructure (road type, road signage, and location type), road user (i.e., opponent vehicle and cyclist's maneuver, type of collision, age and gender of the cyclist), vehicle (type of opponent vehicle), and the environmental and time period variables (time of the day, day of the week, season, pavement condition, and weather). To identify homogenous subgroups of bicycle crashes, we used latent class analysis. Using latent class analysis, the bicycle crash data set was segmented into 19 classes, which represents 19 different bicycle crash types. Logistic regression analysis was used to identify the association between class membership and severity of the bicycle crashes. Finally, association rules were conducted for each of the latent classes to uncover the factors associated with an increased likelihood of severity. Association rules highlighted different crash characteristics associated with an increased likelihood of severity for each of the 19 bicycle crash types.

Analyzing injury severity factors at highway railway grade crossing accidents involving vulnerable road users: A comparative study

OBJECTIVE

The main objective of this study is to identify the main factors associated with injury severity of vulnerable road users (VRUs) involved in accidents at highway railroad grade crossings (HRGCs) using data mining techniques.

METHODS

This article applies an ordered probit model, association rules, and classification and regression tree (CART) algorithms to the U.S. Federal Railroad Administration's (FRA) HRGC accident database for the period 2007-2013 to identify VRU injury severity factors at HRGCs.

RESULTS

The results show that train speed is a key factor influencing injury severity. Further analysis illustrated that the presence of illumination does not reduce the severity of accidents for high-speed trains. In addition, there is a greater propensity toward fatal accidents for elderly road users compared to younger individuals. Interestingly, at night, injury accidents involving female road users are more severe compared to those involving males.

CONCLUSIONS

The ordered probit model was the primary technique, and CART and association rules act as the supporter and identifier of interactions between variables. All 3 algorithms' results consistently show that the most influential accident factors are train speed, VRU age, and gender. The findings of this research could be applied for identifying high-risk hotspots and developing cost-effective countermeasures targeting VRUs at HRGCs.

Examining driver injury severity outcomes in rural non-interstate roadway crashes using a hierarchical ordered logit model

Rural non-interstate crashes induce a significant amount of severe injuries and fatalities. Examination of such injury patterns and the associated contributing factors is of practical importance. Taking into account the ordinal nature of injury severity levels and the hierarchical feature of crash data, this study employs a hierarchical ordered logit model to examine the significant factors in predicting driver injury severities in rural non-interstate crashes based on two-year New Mexico crash records. Bayesian inference is utilized in model estimation procedure and 95% Bayesian Credible Interval (BCI) is applied to testing variable significance. An ordinary ordered logit model omitting the between-crash variance effect is evaluated as well for model performance comparison. Results indicate that the model employed in this study outperforms ordinary ordered logit model in model fit and parameter estimation. Variables regarding crash features, environment conditions, and driver and vehicle characteristics are found to have significant influence on the predictions of driver injury severities in rural non-interstate crashes. Factors such as road segments far from intersection, wet road surface condition, collision with animals, heavy vehicle drivers, male drivers and driver seatbelt used tend to induce less severe driver injury outcomes than the factors such as multiple-vehicle crashes, severe vehicle damage in a crash, motorcyclists, females, senior drivers, driver with alcohol or drug impairment, and other major collision types. Research limitations regarding crash data and model assumptions are also discussed. Overall, this research provides reasonable results and insight in developing effective road safety measures for crash injury severity reduction and prevention.

Latent segmentation based count models: Analysis of bicycle safety in Montreal and Toronto

The study contributes to literature on bicycle safety by building on the traditional count regression models to investigate factors affecting bicycle crashes at the Traffic Analysis Zone (TAZ) level. TAZ is a traffic related geographic entity which is most frequently used as spatial unit for macroscopic crash risk analysis. In conventional count models, the impact of exogenous factors is restricted to be the same across the entire region. However, it is possible that the influence of exogenous factors might vary across different TAZs. To accommodate for the potential variation in the impact of exogenous factors we formulate latent segmentation based count models. Specifically, we formulate and estimate latent segmentation based Poisson (LP) and latent segmentation based Negative Binomial (LNB) models to study bicycle crash counts. In our latent segmentation approach, we allow for more than two segments and also consider a large set of variables in segmentation and segment specific models. The formulated models are estimated using bicycle-motor vehicle crash data from the Island of Montreal and City of Toronto for the years 2006 through 2010. The TAZ level variables considered in our analysis include accessibility measures, exposure measures, sociodemographic characteristics, socioeconomic characteristics, road network characteristics and built environment. A policy analysis is also conducted to illustrate the applicability of the proposed model for planning purposes. This macro-level research would assist decision makers, transportation officials and community planners to make informed decisions to proactively improve bicycle safety - a prerequisite to promoting a culture of active transportation.

Investigating risk factors of traffic casualties at private highway-railroad grade crossings in the United States

Private highway-railroad grade crossings (HRGCs) are intersections of highways and railroads on roadways that are not maintained by a public authority. Since no public authority maintains private HRGCs, fatal and injury crashes at these locations are of concern. However, no study has been conducted at private HRGCs to identify the safety issues that might exist and how to alleviate them. This study identifies the significant predictors of traffic casualties (including both injuries and fatalities) at private HRGCs in the U.S. using six years of nationwide crashes from 2009 to 2014. Two levels of injury severity were considered, injury (including fatalities and injuries) and no injury. The study investigates multiple predictors, e.g., temporal crash characteristics, geometry, railroad, traffic, vehicle, and environment. The study applies both the mixed logit and binary logit models. The mixed logit model was found to outperform the binary logit model. The mixed logit model revealed that drivers who did not stop, railroad equipment that struck highway users, higher train speeds, non-presence of advance warning signs, concrete road surface type, and cloudy weather were associated with an increase in injuries and fatalities. For example, a one-mile-per-hour higher train speed increases the probability of fatality by 22%. On the contrary, male drivers, PM peak periods, and presence of warning devices at both approaches were associated with a fatality reduction. Potential strategies are recommended to alleviate injuries and fatalities at private HRGCs.