Measuring the Potential to Reduce Crash Frequency of Urban Segments by a Probabilistic Approach

Abstract

This paper presents a probabilistic approach to measure the potential to reduce crash frequencyof urban segments. This new approach leads to a hotspot definition and identification using aprobabilistic model defining the dependent variable as an indicator of a discrete choice. A binarychoice model is used considering a binary dependent variable that differentiates a hotspot from asafe site set by the number of crashes per kilometre. The explanatory variables to set similarsegments are based on average annual daily traffic, segment length, density of minorintersections. A threshold value for the number of crashes per kilometre is set to distinguishhotspots from safe sites. Based on this classification, a binary model is applied that allows theconstruction of an ordered site list using the probability of a site being a hotspot. Ademonstration of the proposed methodology is provided using simulated data. For the simulationdesign, urban segment data from Porto, Portugal, covering a five-year period are used. Theresults of the binary model show a good fit. To evaluate and compare the probabilistic methodwith other used methods described in the Highway Safety Manual, measures are used to test theperformance of each method in terms of its power to detect the "true" hotspots. As alreadydemonstrated through the application of the binary model to urban intersections, the test resultsindicate that the binary model performs better than the other two models. The gains of using thismethod are the simplicity, the reliability, and the efficiency.