A Study for Elderly Driver Retraining Using Real-time Safety Advice System

Abstract

In this study, we have developed a real-time safety driving advice system based on automated method of driver’s skill. Our system measures drivers scanning behavior with head motion by wearable sensors and detects his/her risky behavior (e.g. a driver enters a blind inter section with excessive speed without scanning around) from the view point of active safety. By evaluating measured driving behavior based on driving instructors’ safety driving knowledge, our system provides safety driving advice in real time. In Japan, elderly drivers aged over 70 years are required to take the mandatory driving retraining course when renewing their driver’s licenses. In this paper, we experimentally take our system into the retraining course for elderly drivers and verified the effectiveness of our system. Through experiments with 72 subjects (36 subjects were given safety driving advice in real time, and the others were not), we confirmed that our system can improve elderly drivers’ behavior by providing safety driving advice in real time.