A Before-and-After Empirical Bayes Evaluation of Automated Mobile Speed

1 This study evaluates the safety effects of automated mobile enforcement on urban arterial roads. 2 The before-and-after Empirical Bayes (EB) method was used to account for the regression-to3 the-mean effects and other confounding factors. Locally developed safety performance functions 4 and yearly calibration factors for different collision severities were obtained by using a reference 5 group of urban arterial roads. Eight years of data was collected to perform the evaluation, 6 including information on collision records, deployment information, traffic counts, and 7 geometric road data. The results showed consistent reductions in different collision severities, 8 ranging from 14% to 20%, with the highest reductions observed for severe collisions. The 9 enforced segments were further categorized according to site selection criteria and deployment 10 hours to examine their effects on collision reduction. The study also compared the safety effects 11 of continuous and discontinuous enforcement strategies on different arterials, and the analysis 12 revealed that continuous enforcement had a stronger influence on all severity/type of collisions. 13 Moreover, the study also investigated the spillover effects on adjacent approaches, and the 14 findings were discussed with regard to the general and specific deterrence of the enforcement. 15