Identification of differential risk hotspots for collision and vehicle type in a directed linear network

In this study, a spatial approach is followed to estimate the risk that different types of collisions and vehicles present in the central area of Valencia (Spain), considering the accidents observed in this city during the period 2014–2017. A directed spatial linear network representing the non-pedestrian road structure of the area of interest was employed to guarantee an accurate analysis of the point pattern.A kernel density estimation technique was used to approximate the probability of risk along the network for each collision and vehicle type. A procedure based on these estimates and the sample size locally available within the network was designed and tested to determine a set of differential risk hotspots for each typology of accident considered. A Monte Carlo based simulation process was then defined to assess the statistical significance of each of the differential risk hotspots found, allowing the elaboration of rankings of importance and the possible rejection of the least significant ones.
Source: Accident Analysis and Prevention - Category: Accident Prevention Source Type: research