J. Wu, E. Radwan, H. Abou-Senna

Pages: 33-46

Abstract
Pedestrian safety has become more prevalent for governmental agencies to address the safety of public. The tabulation of total numbers of conflicts is usually used as a surrogate safety measure to indicate the safety issues. However, the severity of the conflicts is another element of the safety issue. This study aimed to assess pedestrian-vehicle conflicts under different potential risk factors at unsignalized midblock crossings. Using the driving simulator, a full factorial experiment is designed to study the pedestrian-vehicle conflicts, using four potential risk factors which included time of day, crosswalk marking, roadway type, and pedestrian dressing color. Performance measures such as the maximum deceleration, post-encroachment time (PET), and the minimum time to collision (TTC) are adopted as the surrogate safety measures to assess the pedestrian-vehicle conflicts. A mixed effect model is developed to associate the four potential risk factors and two human factors with the surrogate measures. According to the results, driver’s gender and age are the two significant factors that affect the the maximum deceleration and the minimum TTC. Time of day and pedestrian dressing color are found to have a significant impact on the maximum deceleration, PET, and the minimum TTC. Crosswalk marking and roadway type only impact the maximum deceleration and the minimum TTC. This study proposes a new method that can be used for evaluating pedestrian-vehicle conflicts at mid-block crossings. Further studies are needed to examine more potential risk factors that are related to pedestrian safety.
Keywords: driving simulator; laboratory experiment; driving behavior; mixed model; pedestrian conflicts