Drivers' collision avoidance behavior: Timing of pedestrian detection when turning right at signalized intersections

This study evaluated drivers' collision avoidance behavior by focusing on the timing of their visual detection of pedestrians when turning right at signalized intersections. A headset-based virtual reality driving simulator was used to monitor drivers' behavior. The experiment showed that early detection of pedestrians can reduce the potential for severe conflicts. Drivers found it difficult to detect pedestrians early when using a gap between oncoming vehicles to turn right. Based on gaze tracking, we propose and validate a safety level that could potentially evaluate the potential for conflict between pedestrians and a right-turning vehicle. To increase safety level reliability, the speed of the right-turning vehicle should be included in addition to the eye measurements.