A Comprehensive Review of Vehicle–Pedestrian Interactions: Crash Analysis and Conflict Assessment Approaches

Abstract

Pedestrian–vehicle collisions remain a critical issue in transportation safety, contributing disproportionately to global traffic-related fatalities and injuries. Unlike vehicle occupants, pedestrians lack physical protection and are therefore more susceptible to severe or fatal outcomes when involved in crashes. Understanding the mechanisms of vehicle–pedestrian interactions and contributing factors to pedestrian crashes is essential for uncovering the crash nature and informing the development of effective safety countermeasures and technologies. This review synthesizes recent advancements in the study of pedestrian safety research at intersections, covering historical crash data–based modeling and analysis, as well as conflict-based studies using field observations and simulation. Key topics include the application of statistical and machine-learning models in crash likelihood and severity analysis, the use of surrogate safety measures, and the integration of conflict analysis frameworks such as extreme value theory. Critical challenges related to pedestrian safety modeling methodologies and evaluation metrics, the evolving safety implications in connected and automated vehicle environments, and the practical applications of these insights for policy and infrastructure design are discussed in depth. By reviewing methodological innovations and highlighting emerging research directions, this paper offers a comprehensive foundation for advancing pedestrian safety research and guiding the development of data-driven, context-sensitive policy, operation, and infrastructure solutions.