Analysing the interaction between driver and pedestrian at unsignalized crosswalks

Abstract

This study aims to improve pedestrian safety at unsignalized crosswalks by analyzing driver-yielding behaviour. Data from two-lane, two-way divided roads was used to capture vehicle trajectories and pedestrian characteristics. Driver-yielding patterns were determined using TRACKER Software and the factors that influence drivers’ yielding willingness were determined using logistic regression models. The study found that yielding likelihood is influenced by factors such as female pedestrians, median pedestrians, injured pedestrians, and vehicle position. Further revealed that most vehicle types yield similarly, with a decreasing average rate, and no significant difference was observed among the six vehicle types considered. This research provides insights into key factors affecting pedestrian safety at unsignalized crosswalks, guiding targeted interventions and traffic management strategies.