Capturing drivers' Aggressiveness from inter-vehicular dynamics data

Abstract

Understanding driver's aggressiveness can be useful for many aspects ranging from negotiating insurance premium to determining appropriate perception-reaction time for driver specific warnings. Most studies have used (i) survey questionnaires asking the crash experience, citations, and driving behaviors, and/or (ii) analyzed individual vehicular data such as maximum deceleration rate and highest speed. This study proposed a new index that captures drivers' aggressiveness based on inter-vehicular dynamics data. The proposed index based on the time-to-collision and deceleration rate difference was applied to real world vehicular dynamics data obtained from drivers participated in driving tests. The proposed index clearly showed that younger female drivers were the most aggressive and followed by younger male drivers. Responses of survey questionnaire data from the participants were compared with the proposed index. The results indicated that the index does not appear to match well with some drivers, implying that questionnaire responses might not be the best indicator for capturing drivers' aggressiveness. However, responses on unintentional violation and ideal safe driving speed supported the proposed aggressiveness index.