Identification of Pedestrian Risk Factors Using Negative Binomial Model

Abstract

Pedestrian deaths causing from road traffic crashes are a foremost and rising problem in urban India. However, pedestrian safety assessment using precise statistical analysis and modeling techniques is still in an early stage in urban India owing to the absence of crash data collection and maintenance, difficulties in obtaining appropriate road inventory data and traffic volume data. The present study composes an effort to fill the research gap by analyzing past crash data (2011–2016) collected from Kolkata police and identifies the hazardous corridors and intersections posing the high risk to the pedestrian. Consequently, to develop an understanding of the risk factors associated with pedestrian fatalities, a set of fixed-parameter Negative Binomial models is established. In these models, several attempts are made to identify the risk factors linked to traffic exposures and operational parameters, roadway characteristics and infrastructure, land use and road network planning. Based on the statistical analyses carried out in this study, there is evidence that factors such as approaching speed of the vehicles, vehicular traffic and pedestrian volume of the intersection and their interaction, disorderly movement of traffic (i.e., overtaking tendency of the vehicles), the presence of specific land-use type (i.e., commercial hubs), inefficient planning and design (i.e., inaccessibility of the pedestrian crosswalk, the absence of pedestrian signal head, etc.), the presence of wider carriageway, the encroachment of the footpath, restricted visibility (i.e., inadequate sight distance) are found to be major risk factors with regards to the fatal pedestrian crash occurrence in Kolkata City, India.