Research Utilizing SHRP2 Data to Improve Highway Safety: Development of Speed\u2014Safety Relationships

2020Final ReportMohamedshah, YusufPDFTech ReportFHWA-HRT-20-035DTFH61-16-C-00005Highway safetyStrategic planningCrash modification factorsCrash dataArterial highwaysHighway safetySuburban arterialSpeedCrash modification factorUrban arterialCrash dataNaturalistic drivingRoadway inventory dataUnited StatesUnited States. Federal Highway Administration. Office of Safety Research and DevelopmentHutton, Jessica M.Cook, Daniel J.Grotheer, J.Conn, M.United States. Federal Highway Administration. Office of Safety Research and DevelopmentMRIGlobal (Firm)Turner-Fairbank Highway Research CenterMohamedshah, Yusuf M.Federal Highway AdministrationThe objective of this research was to examine the link between driving speed and crash experience on roadway segments, taking into account roadway characteristics that influence speed and crash frequency and severity. A potential outcome of the research was improved safety performance functions that included a speed term or crash modification factors (CMFs) based on speed for use in a safety-prediction methodology for urban/suburban arterials. Using the second Strategic Highway Research Program Naturalistic Driving Study data and the Roadway Inventory Database (RID), the research examined individual drivers' speeds along 100 study segments. Variations within individual trips as well as among drivers on the same roadway segment over a few years were evaluated. Also, using the RID and roadway and roadside characteristics obtained from aerial and street-view imagery, relationships between speed choice and roadway characteristics were explored. The research found that a number of roadway characteristics are related to speed and crash experience for a roadway segment and that a higher measure of speed variance between trips was frequently correlated with higher crash frequency (especially for multivehicle crashes). Most other speed measures evaluated had no correlation with crash frequency or had a negative correlation. The research did not find that including a speed term into existing safety performance functions or developing a speed CMF would substantially improve the existing crash-prediction methodology for urban and suburban arterials presented in the Highway Safety Manual.89