Applying different analytic methods to determine black spots in two-lane highways

Abstract

Various analytic methods have been proposed to determine sections with the highest crash risk. Each method has unique specifications and tries to model the crash risk from a different viewpoint. The main objective of this article is to benefit the strengths of three methods that rely on accident data, road safety inspection, and traffic conflict technique to determine black spots for two-lane highways simultaneously. Fuzzy inference system (FIS) is considered as the method to combine the results of these methods and report one number (R MI) as the crash risk for each section. For comparative evaluations, a case study with 20 sections for two consecutive periods was considered in the roads of southeast of Iran. We have tried to select sections with various conditions from crash data, road condition, and surrogate safety measures viewpoint. First, the black spots are determined with the help of previous criteria such as crash frequency (CF), crash rate (CR), empirical Bayes (EB), and equivalent property damage only (EPDO). Then the black spots are specified by the new proposed criteria (R MI). Three tests are applied to compare the efficiency of these five methods. The results indicate that the proposed method is a powerful tool to identify black spots. R MI considers the frequency and severity of observed crashes and at the same time frequency and severity of predicted crashes based on road deficiencies and near crashes. Therefore, it has a more realistic attitude in black spot identification.