Analysis and modelling of crash severity of vulnerable road users through discrete methods: a case study approach

Road traffic injuries rank as the eighth most prevalent cause of mortality among individuals of various age groups while specifically representing the primary cause of death for individuals between the ages of 5 and 29. On a global scale, it is observed that pedestrians, bicyclists, and motorized 2- and 3-wheelers collectively account for 54% of the total number of fatalities in traffic incidents. Therefore, pedestrians, cyclists, and motorcyclists are the most vulnerable road users (VRU). In developing nations such as India, VRUs account for over 65% of fatalities. This high proportion can be attributed to factors such as fast urbanization, high population growth rates, and poor infrastructure, all of which have contributed to increased deaths and injuries among VRUs. The existing body of research about crashes involving VRUs in India primarily concentrates on large metropolitan cities. This study primarily examines tier-2 cities that are categorized as non-metropolitan regions. By analysing historical crash records (2014–2021) from "Visakhapatnam traffic police", India, and data from road safety audits, the current research attempts to fill the gap and identify the factors contributing significant risk to the most vulnerable road users. In this case study approach, three crash severity models were built using binary logistic regression, multinomial logistic regression, and ordered probit regression techniques to identify the significant risk factors associated with vulnerable road users. Based on the statistical analysis conducted in this study, it is evident that various factors, including segment length, driver sight clearance, land use, crash time, crash season, accused vehicle type, number of median openings, curves, and pedestrian crossings, exert a substantial influence on the safety of vulnerable road users. To decrease the probability of fatal crashes involving vulnerable road users, particularly on urban National Highways, specific planning and design features are employed and adjusted to mitigate risk, considering the unique risk factors associated with each location.