Smartphone-based modeling and detection of aggressiveness reactions in senior drivers

Abstract

Reckless driving is one of the leading causes of car accidents. In particular, reckless driving by senior drivers often results in serious consequences due to driver physical fragility. As the population in developed countries is aging, the number of elderly drivers is increasing rapidly. Thus, careless or reckless driving in the elderly has become an important research issue. To investigate driving behavior in the elderly, we used a smartphone because it is equipped with gyro sensors. We constructed driving tests for elderly people on two types of courses, and also performed the same test to young people for data comparison. We then analyzed the data through the classification of GMM(Gaussian Mixture Model) with Periodogram in the elderly group. Using our method, we can classify elderly people's driving style on a gradient from smooth to aggressive behavior. Our proposed method will be useful in building early warning systems for elderly drivers as part of Advanced Driver Assistance Systems(ADAS).