Abstract:Future vehicle warning systems needs a local (instead of global) analysis of real-time information transmitted between vehicles and infrastructures, to provide local warning information matching the instantaneous driving contexts. Spatial modeling techniques extracting the location information into the analysis fulfill the needs of local analysis. For truck-involved collisions at highway-rail crossings, the local warnings seem to matters more, than for traffic crashes at the normal highway segments and regular intersections. Crashes at rail grade crossings can result in severe injuries and fatalities to vehicle occupants, while truck-involved crashes at crossings can further result in serious damage to train, crossing and railway equipment. Truck-involved crashes at grade crossings have received limited attention compared with crashes involved with passenger cars. This study presents a methodology of improving safety of trucks at railroad crossings, by taking advantage of the big data containing location information. Unlike previous studies that constructed a direct relationship between the safety outcomes and associated factors, this study investigates direct relationships together with indirect relationships through the truck driver behaviors before collisions, using path analysis techniques. To sum up, this study applies a spatial approach fused with path analysis to uncover the local relationships between truck driver injury severity and crossing controls across the space. By doing so, the research is able to: i) provide a benchmark of identifying potentially risky vehicles on a real time basis; ii) evaluate current control devices at railroad crossing across the country and pin point the potentially problematic crossing sites. An empirical study was conducted by using a rich crash database from the Federal Railroad Administration (N= 4738 for 2004–2014). The results show that truck-involved crashes occurring at crossings without gate controls are generally associated with higher chance of injury, while the associations vary significantly across the space. In general, crashes in the Midwest and Great Lake regions are associated with an even higher chance of injury at crossings without gates, compared with other regions. More Results and implications are discussed in the paper.

Gate Violation Prediction at Highway-Rail Grade Crossings Using Tree-Based Ensemble Techniques

Gradient Boosting Regression Tree for Traffic Flow Prediction Considering Temporal and Spatial Correlations

Multi-step-ahead Traffic Speed Forecasting Using Multi-Output Gradient Boosting Regression Tree

Multi‐model Ensemble for Short‐term Traffic Flow Prediction under Normal and Abnormal Conditions

Accident Prediction Accuracy Assessment for Highway-Rail Grade Crossings Using Random Forest Algorithm Compared with Decision Tree

Exploring pre-crash gate violations behaviors of drivers at highway-rail grade crossings using a mixed multinomial logit model

Using hierarchical tree-based regression model to predict train-vehicle crashes at passive highway-rail grade crossings.

Analysis and Prediction of Pedestrians' Violation Behavior at the Intersection Based on a Markov Chain

Traffic Flow Prediction for Road Intersection Safety

Traffic Safety Assessment and Injury Severity Analysis for Undivided Two-Way Highway–Rail Grade Crossings

Analysis and Prediction of Spatiotemporal Distribution of Accidents at Highway-rail Grade Crossings Caused by Different Causes

How Big Data Serves for Freight Safety Management at Highway-Rail Grade Crossings? A Spatial Approach Fused with Path Analysis

GATR: A Road Network Traffic Violation Prediction Method Based on Graph Attention Network

Automated Detection of Grade-Crossing-trespassing Near Misses Based on Computer Vision Analysis of Surveillance Video Data

Intersection Traffic Prediction Using Decision Tree Models.

Predicting intersection crash frequency using connected vehicle data: A framework for geographical random forest

Artificial Intelligence-Aided Grade Crossing Safety Violation Detection Methodology and a Case Study in New Jersey

Computer vision-based approach for smart traffic condition assessment at the railroad grade crossing

Predicting Highway–Rail Grade Crossing Collision Risk by Neural Network Systems

Predicting and Analyzing Pedestrian Crossing Behavior at Unsignalized Crossings

A Novel Ensemble Learning Method for Crash Prediction Using Road Geometric Alignments and Traffic Data