Analysis of traffic conflicts with right-turning vehicles at unsignalized intersections in suburban areas

Abstract

Right-turn collisions at intersections are one of the most dominant crash types in suburban areas, especially at unsignalized intersections. There is, however, a lack of comprehensive research on the speed patterns of vehicles during right-turn manoeuvres and their impact on crashes. To provide an in-depth investigation of the factors determining the safety of right-turn manoeuvres, driving behaviour data were collected through an instrumented vehicle study. Using this data, binary logistic regression models were developed to identify the factors affecting the probability of Vehicle-Vehicle (V-V) and Vehicle-Pedestrian (V-P) conflicts at six suburban intersections in Babol, Iran, during right-turn stage manoeuvres. In total, 1,456 V-V and V-P conflicts were identified from the data analysis. The results from the logistic regression model showed that the vehicle speed, the distance between road users, as well as driver and pedestrian distractions were associated with higher risk for vehicle-to-vehicle or vehicle-to-pedestrian conflicts. To estimate the safe right-turn speeds to be selected by the drivers at different stages of the right turn, i.e., at the start, during, and end of the movement, linear regression models were developed. The results showed that participants adjust their driving behaviour the same way toward pedestrians as they do toward vehicles. The findings of this study can be leveraged for the development of a robust advanced driving assistance system, the use of which can further improve the safety performance of right-turn manoeuvres.