Frequency and severity of crashes involving vulnerable road users: an integrated spatial and temporal analysis

Pedestrians and cyclists, often called vulnerable road users (VRUs), are more likely to be injured in road crashes as they are more exposed to risk. It is estimated that each year 1.2 million road users lose their lives on the world’s road crashes with half of them being VRUs. This situation has a dramatical impact in terms of health and economical development and costs to governments, when low- and middle-income countries lose approximately 3% of their GDP. The analysis of road crashes registrations and the development of predictive models to identify areas with higher risk could be a crucial step to improve road safety and sustainable urban mobility. The main objective of this paper is to find temporal and spatial patterns of crashes between motor vehicles-VRUs based on severity, in order to implement a model that estimates the probability of occurrence of a crash involving VRUs. For that purpose, crashes data from three cities in Portugal with different characteristics were examined. Crashes were georeferenced and blackspots were identified considering injury severity. Although georeferencing is often a method of identifying potential risk areas, it is not associated with time and injury severity. The proposed model is defined as a Multinomial logistic regression model (MLR) with pedestrians and cyclists as a response variable. The findings from this study highlighted target variables that may influence number and severity of crashes between motor vehicle and VRUs. The developed MLR models revealed that VRU gender and age, as well as weather conditions, are statistically significant