Using Fully Bayesian Approaches to Evaluate the Impact of Section Speed Enforcement System on Expressways

Effective speed enforcement is essential for reducing speeding-related crashes and improving road safety. Among enforcement methods, the Section Speed Enforcement System (SSES) addresses limitations of point-speed enforcement system by capturing both instantaneous and average travel speeds. Despite its recent installation on Korean expressways, previous studies of its effectiveness evaluation have limited. As such, this paper quantitatively assessed the safety effects of SSES using Empirical Bayes (EB) and Full Bayes (FB) methods with small sample data. The analysis covered four crash types, and the overall Crash Modification Factor (CMF) for total crashes was estimated at 0.653, indicating a 34.7% reduction in crash frequency. Additionally, heavy vehicle-related and adverse weather crashes showed safety improvements of 23.1% and 12.8%, respectively, after treatment. The Poisson-Gamma distribution provided the best model fit for crash frequency, while the Poisson-Lognormal distribution was more appropriate for crash severity. Moreover, downhill slopes were found to significantly influence crash risk, suggesting the importance of road geometry in speed management strategies. While this study incorporated spatial heterogeneity, further research including temporal heterogeneity and traffic flow analysis is recommended to enhance the reliability and applicability of safety evaluations for section speed enforcement systems.