Factor analysis of pedestrian accidents based on driver’s eye movement and the time that drivers attract pedestrian at right turn in an intersection with traffic lights

Abstract

The characteristic scenario of the Hiyari-Hatto event (i.e., a near-miss incident) wherein a driver is making a right turn is extracted from the Near-Miss Incident Database created by the Smart Mobility Research Center of the Tokyo University of Agriculture and Technology. It is confirmed that in this scenario, there exists a human error: the driver did not notice the pedestrian although the pedestrian is within the driver's field of view. Closer observation of the driver's eye movements reveals that the driver's line of sight is fixed once the driver starts making a turn. Using a novel and unprecedented measurement method, the time at which the driver recognizes the pedestrian can be pinpointed. This method is tested through the driving simulation experiment at an intersection with traffic lights, and the results confirm that there is a high possibility that the driver can miss the pedestrian after the driver starts making a turn.

Keywords

• safety
• accident investigation and analysis
• drive recorder
• near-miss analysis
• driving simulator
• human error
• driver attention