Assessment of Surrogate Safety Benefits of an Adaptive Traffic Control System

When Adaptive Traffic Control Systems are evaluated their safety benefits are rarely investigated. This ‘lack of interest’ in safety performance may be attributed to two major factors: 1) field assessments of safety improvements are impractical (i.e. time consuming and costly), and 2) methods for obtaining safety metrics from microscopic simulation models are uncommon. While collecting and analyzing field safety metrics (e.g. crash rates) remains operationally and analytically challenging, new tools are becoming available to provide surrogate safety measures based on microscopic simulation program outputs. The framework of this study uses a microsimulation model connected to Sydney Coordinated Adaptive Traffic System (SCATS) to generate vehicular trajectories which are fed into a Surrogate Safety Assessment Model. Realworld data from two Utah state routes are collected to reflect crashes before and after the adaptive system was deployed in 2006. The crash data are analyzed and compared to relevant surrogate safety measures to detect potential correlations between the two data sets. Results show that SCATS generates fewer rear-end and total conflicts than traditional traffic control while traditional control generates fewer crossing and lane-changing conflicts. Field crashes have increased in the last few years likely as a consequence of more numerous road construction activities. Correlation indicators between field crashes and simulated conflicts are not very strong but follow trends reported by other studies.