Métodos de identificação de zonas de acumulação de acidentes: Revisão e aplicação a um caso de estudo

Abstract

This paper presents a description of the various methods proposed by the Highway Safety Manual (HSM) to identify hotspots. In fact, this manual has become a tool for the professional in road safety all over the world. The HSM suggest and described 13 distinct methods in terms of application context and method complexity. Therefore, the main goal of this paper is to describe and organize those methods to help safety professionals to select the method(s) to apply to the study case. Thus, the various methods are organized based on 3 application criteria, which may be crucial to the decision process. In addition to the HSM methods, a recent method entitled probit binary model is also shown. An application to the Porto city is presented in order to better analyze each method. In fact, based on this study case, it was concluded that not all of the methods were possible to apply to this study case. A comparison of the various methods is developed using tests to assess the performance of each method. The methods named excess predicted average crash frequency proposed by the manual and the binary probit model, recently developed, present the best results based on those tests. The results of the case study application may be the basis for future applications in other regions and/or countries.