Assessing vulnerable road users (vrus) behavior from safety, mobility and policy perspectives

Abstract

Significant researches have been performed, and policies are adopted concerning Vulnerable Road User’s (VRU) safety. However, notable casualties are observed every year resulting from human factors to policy implications. Numerous VRU areas e.g., risk perception of vehicle-to-vehicle vendors alongside general pedestrians; behavioral differences of commercialized cyclists and motorcyclists than the general ones; safety issues of non-motorized vehicles from other road users’ perceptions; etc. still need to be explored. This study aims to contribute to existing VRU literature by risk perception analysis of vehicle-to-vehicle vendors and general pedestrians; study of the perceived safety behaviors of commercialized bicyclists and motorcyclists; and identification of key factors influencing risk perceptions of different road users on non-motorized vehicle movement e.g., cycle rickshaws. Data were collected from different residential, commercial areas of Dhaka city and the data collection process was conducted based on well-structured questionnaire. A Bayesian approach was opted to develop four separate models for four user groups e.g., pedestrians, bicyclists, NMVs and motorcyclists to unveil the relevant underlying factors with aid of conditional probability. From the self-reported data, risk categories were defined and classified for respective target variables. Results obtained from the analysis of the models showed that vendors have a 24% high risk perception than pedestrians where gender and education are the most significant variables that influenced risk taking tendency. Basic demographic factors influence the attributes of commercialized bicyclists most significantly. The NMV model predicted that location of NMV stops and existence of pavement hazards on the roads encourages higher risk perception towards NMVs on our roads. Risk perception of road users also varies among various socio-demographic segments like- education level, gender, age, etc. According to the sensitivity analysis of the motorcyclist model; age, gender, driving license status, and trip variations are the most important factors in driving behavior. In spite of the possibility for self-reporting bias, results from this study can be a useful resource for policy makers and law enforcement authorities to take necessary actions in increasing positive safety attitudes among the aforesaid road user group.