Before–After Safety Evaluation of Coordinated Ramp Metering System Using Empirical Bayes Approach: A Case Study on I-80 in California

Abstract

Coordinated ramp metering (CRM) systems are implemented on freeways primarily to improve operational conditions. However, smoother traffic flow resulting from CRM may also have significant safety benefits. The main objective of this research is to evaluate the safety performance of CRM systems on the Interstate 80 (I-80) corridor in California by using an empirical Bayes before–after approach. We collected geometric features, traffic volume, and historical crash data from I-80 in the San Francisco Bay area (Caltrans District 4). Then, the freeway safety prediction methodology implemented using the Enhanced Interchange Safety Analysis Tool (ISATe, developed as part of a National Cooperative Highway Research Program Project 17-45) was utilized to estimate the counterfactual, i.e., the number of crashes if no CRM system was implemented on the corridor. Based on a comparison of this counterfactual with the actual crash counts in the post-CRM period, CRM implementation led to a decrease in the number of fatal and injury crashes on I-80 but an increase in minor property damage only crashes. Disaggregate analysis of the results was used to gain further understanding of the CRM safety performance. The differences in the resulting safety performances were contextualized based on the different settings where the systems were implemented. As expected, CRM systems were more effective for segments in the vicinity of ramps. Safety performance functions for shorter durations (e.g., for peak hours), the subject of the ongoing National Cooperative Highway Research Program Project 22-48, will help more precisely estimate the safety impact of CRMs.