High Risk Crash Analysis

US Transportation Collection2005PDFResearch PaperWashington, Simon P.Cheng, WenUniversity of ArizonaArizona. Department of TransportationUnited States. Department of Transportation. Federal Highway AdministrationArizona. Department of TransportationArizonaAccuracyCountermeasuresCrash analysisCrash dataEvaluationHigh risk locationsHighway safetyTraffic safetyRecommendationsSoftwareTraffic crashesHot Spot IdentificationHigh Risk SitesSites with PromiseMotor vehicle crashesFHWA-AZ-05-558SPR-PL-1-(63) 558In agencies with jurisdiction over extensive road infrastructure it is common practice to select and rectify hazardous locations. Improving hazardous locations may arise during safety management activities, during maintenance activities, or as a result of political pressures and/or public attention. Commonly a two-stage process is used. In the first stage the past accident history of all sites is reviewed to screen a limited number of high-risk locations for further examination. In the second stage the selected sites are studied in greater detail to devise cost-effective remedial actions or countermeasures for a subset of correctable sites. Due often to limited time and resources constraints and the extensive number of candidate sites typically considered in such endeavors, it is impractical for agencies to examine all sites in detail. The current Arizona Local Government Safety Project Analysis Model (ALGSP) is intended to facilitate conducting these procedures by providing an automated method for analysis and evaluation of motor vehicle crashes and subsequent remediation of 'hot spot' or 'high risk' locations. The software is user friendly and can save lots of time for local jurisdictions and governments such as Metropolitan Planning Organizations (MPOs), counties, cities, and towns. Some analytical improvements are possible, however. The objective of this study was to provide recommendations that will lead to improvement in the accuracy and reliability of the ALGSP software for identifying true 'hot spots' within the Arizona transportation system or network, be they road segments, ramps, or intersections. The research resulted in 1) a survey of past and current hot spot identification (HSID) approaches, 2) evaluation of HSID methods and exploration of optimum duration of before-period crash data under simulated scenarios, 3) development of safety performance functions (SPFs) for various functional road sections within Arizona, 4) extended comparisons of alternative HSID methods based on SPFs by using real crash data, and 5) recommendations for improving the identification ability of the current ALGSP model.127