Driving Behavior Monitoring of Older Drivers in Daily Life and Analyzing Driving Characteristics at Stop-Controlled Intersections

Abstract

In this study, we installed vision-based driver monitoring sensors to older drivers' private vehicles for measuring their driving behavior (e.g. scanning behavior, vehicle speed control, etc.) in real traffic environment in their daily life, and collected large-scale driving behavior data for 737 days, 14,233 kilo meters in total. Since the most frequent traffic accident pattern by older drivers is broadside collision, we focused on stop-controlled intersections and analyzed older drivers' behavior from the viewpoint of active safety. As a result of our analysis, we found out that 22.8% of all cases of driving straight through an intersection, older drivers insufficiently performed scanning behavior to the both sides of the intersection for checking safety. We also found out that 46.8% of all cases of turning right/left an intersection, older drivers performed scanning behavior only to the one side of the intersection, which could lead to increase collision risks to other vehicles, cyclists, and pedestrians.