Roadway safety management consists of network screening, diagnosis, countermeasure selection, economic appraisal, prioritization, and safety effectiveness. Application of the safety management process is limited in small municipalities because of data, statistical expertise, and resources required. This paper addresses the challenges faced by small jurisdictions and illustrates the implementation of the safety management process for local agencies in the Madison metropolitan area in Wisconsin. Jurisdiction-specific crash prediction models were developed by intersection type using data from over 4,000 intersections. Performance measures included the Equivalent Property Damage Only (EPDO) average crash frequency with Empirical Bayes adjustments and the Level of Service of Safety (LOSS). Wisconsin Crash Outcome Data Evaluation System (CODES) data was used to estimate local crash costs by severity and type. Sites were provisionally ranked in network screening, and diagnosis was automated based on intersection-observed crash types and distributions. Potential treatments were selected for each intersection and costs of treatments were obtained from local estimates and available literature. Crash cost benefit and treatment cost were used to estimate benefit–cost ratio (B/C) by site. A combination of sites that had the greatest overall cost-effective safety benefit on the network were selected through an incremental optimization process. This paper contributes to existing practice by illustrating the development of jurisdiction-specific crash prediction models, integration of pedestrian and cyclist crashes, application of EPDO and LOSS performance measures, and selection of sites with promise through an incremental optimization process for a given budget and resources available at the local level.
Comparative Analysis of Performance Measures for Network Screening: A Case Study of Brazilian Urban Areas
