Citation

Adane BY, Hou J, Peng B, Ullah H, Abate DY. Transp. Res. Rec. 2023; 2677(2): 648-660.

Copyright

(Copyright © 2023, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

10.1177/03611981221108150

PMID

unavailable

Abstract

Traffic violations are the leading cause of fatalities and injuries, and current trends indicate that this will proceed in the foreseeable future, especially alongside intersections. This study aims to examine the determinants of crash influencing factors alongside intersections using unweighted data points collected from the Michigan counties of Wayne, Genesee, and Macomb. Firstly, the random forest method identifies the significant predictor variables. Meanwhile, six predictor variables are considered: the speed limit, distance, light condition, crash location, train involvement, and weather conditions. Secondly, the multinomial logit approach is adopted to investigate the relationships among crash violation categories. Thus, the model predicts the likelihood of crash violation outcomes: angle, head-on, rear-end, rear-end right/left turn, same sideswipe, opposite direction, and single-vehicle crash. Likewise, the results confirmed that the predictor variables of the distance between 0 and 82?ft and train involvement increased the likelihood of a crash violation when driving at a speed limit of 26?50?mph. In addition, when the speed limit is between 0 and 25?mph, the odds ratios were greater than 1.0 (4.01, 3.85), indicating statistically significant positive relationships between the same sideswipe and opposite direction violations. Therefore, speeding is risky, but the safest speed is not always the slowest speed. Thus, the results provide evidence that the predictor variables are linked to various crash violations. Consequently, the findings of this study can be used to reduce the occurrence of crash violation and improve intelligent modes of transportation by transforming large datasets into knowledge and actionable intelligence considering the future roles of driving strategies.

Language: en