Safety Analysis and Crash Modification Factors of an Adaptive Signal Control Technology along a Corridor

2019Safety Report, 1/1/13 \u2013 12/1/18PDFTech ReportFHWA-ICT-19-001R27-127Adaptive controlBayes' theoremBefore and after studiesCalibrationCrash modification factorsCrash ratesCrash severityTraffic signal control systemsCrash Modification FactorSafety Performance FunctionsHighway Safety ManualAdaptive Signal Control TechnologySynchroGreenIllinoisUnited StatesIllinois Center for TransportationOsorio, Jesus J.Benekohal, Rahim F.University of Illinois at Urbana-Champaign. Dept. of Civil and Environmental EngineeringIllinois Center for TransportationIllinois. Department of TransportationUnited States. Department of TransportationUnited States. Federal Highway AdministrationUS Transportation CollectionThe main objective of this study is to determine the safety effectiveness of the adaptive signal control technology (ASCT) SynchroGreen using an observational before and after study applying the Empirical Bayes (EB) method. Both national (HSM) and state specific (Illinois) safety performance functions (SPF) were selected and calibrated for the local conditions. A total of 14 SPFs from the HSM and 3 additional from Illinois were calibrated and crash modification factors (CMF) were developed. For multiple-vehicle fatal and injury (FI) crashes at all intersections (four-legged and three-legged combined), the CMF was 0.67, which was not statistically significant at 95 percent confidence level (it was significant at 87 percent). For four-legged-only intersections the CMF was 0.67 as well, which was not significant with 95 percent confidence (it was significant at 85 percent). The 87 and 85 percent are not confidence levels used in practice, however they clearly indicate a decreasing trend in FI crashes due to the implementation of ASCT. For PDO and total crashes, all CMF computed were very close to one indicating no crash reduction due to the implementation of ASCT. The CMF for Illinois KAB crashes (fatal, type A injury, and type B injury crashes combined) was found to be 0.68, which was not significant at 95 percent confidence level (it was significant at 71 percent indicating a decreasing trend in these types of crashes). Paired t-test results showed no reduction in sideswipe same direction, turning, and type B injury crashes. However, angle, rear end, type A and type C injury crashes showed slight decreases that were not significant.80