A novel network-level integrated crash frequency, exposure and severity approach to assessing and mitigating pedestrian crash risk.

Abstract

One of the main challenges in urban mobility is improving pedestrian safety. Despite recent reductions in road crashes and fatalities, current progress remains insufficient to achieve international road safety targets. Existing studies typically examine crash frequency or severity separately or estimate them jointly without integrating both dimensions into a comprehensive risk assessment. In addition, most analyses focus on specific locations rather than evaluating the safety performance of the entire urban road network. This study proposes an integrated framework for pedestrian crash risk built upon the Road Network Screening (RNS) approach. The framework segments the urban road network and estimates pedestrian crash risk by integrating crash frequency (including exposure) and severity. It provides thematic risk maps to prioritise high-risk road segments and simulate potential safety interventions. The framework is applied to the urban road network of Brescia, Italy, using nine years of pedestrian crash data. Results show that crash frequency is mainly influenced by road class and crossing design, while crash severity is associated with infrastructure characteristics, vehicle type, and environmental conditions. The integrated risk metric highlights critical road segments that may not emerge when considering frequency or severity independently. The proposed framework provides road safety practitioners with an operational tool for systematic, evidence-based prioritisation of interventions and supports more effective strategies to improve pedestrian safety across urban road networks.