Evaluation of hot spot identification methods for municipal roads

AbstractEstimating crash prediction models and applying the Empirical Bayesian approach in identifying hotspots for roads under municipal jurisdiction is often challenging due to the lack of traffic count data. This study presents five hotspot identification (HSID) methods in which annual average daily traffic (AADT) information is not required (i.e., crash frequency [CF], equivalent property damage only, relative severity index, excess predicted average crash frequency using method of moments [MOM], and cross sectional analysis [CSA]), to identify hotspots for road segments under municipal jurisdiction in Connecticut. The segments were categorized into 11 homogenous groups based on the roadway geometric characteristics. The five HSID methods were applied to all segments in each roadway group separately and across the entire State for a systemic analysis. Four quantitative tests (i.e., site consistency test, method consistency test, total rank difference test, and total score test) were used to compare th...