Proposal of Risk Estimation Index from Driver Behaviour for Approaching and Overtaking Vulnerable Road User

Abstract

To increase the acceptance and willingness of customers in using advanced driver assistance systems and autonomous vehicles, one of the best solutions is to reproduce the driving behavior of humans. In this regard, estimating risk based on driver behavior is a priority. An index of perceptual risk estimation by drivers toward pedestrians approaching and overtaking maneuvers is proposed herein. Risk estimation comprises longitudinal and lateral risk estimations. The index is verified using maneuver data obtained from a driving simulator. It is discovered that drivers with different driving styles have different risk index ranges. The index may be used to classify drivers for selecting the appropriate response from an assisting system.