Driving style recognition for co-operative driving: a survey

Abstract

Abstract—This paper serves as a critical survey for automatic driving style recognition approaches and presents “work in progress ” ideas that can be used for the development of intelligent context-adaptive driving assistance applications. Furthermore, a preliminary specification of a context-adaptive application that can be described by the following three steps is provided: at first, driving style is automatically classified into one out of a set of predefined classes that are learnt through historic driving and trip data; secondly, based on the driving style recognition a context-adaptive driving application is proposed; thirdly, eco-safe and co-operative driving behaviour can be rewarded by the system by introducing a serious game theoretic approach. While the focus of this paper lies on reviewing the state of the art for implementing the first step, providing the high-level specification of the two other steps offers valuable insight on the requirements of such collaborative driving application. Keywords- driving behaviour; vehicle dynamics; time-series analysis; supervised learning; classification; co-operative system. I.