Calculating Driving Behaviour Score Based on Driving Background

Abstract

Improving driving behaviour can be a fruitful intervention to improve traffic safety. This paper proposes a method to determine a driving behaviour score (DBS) for each driver based on their driving history. For this purpose, a fuzzy inference system (FIS) was used to calculate DBS for every driver between 0 and 1. The input variables for this model are the frequency and severity of previous crashes, near-crash situations, and aberrant driving behaviours. The calculated DBS can then be applied in implementing usage-based insurance schemes. The proposed method is used for a case study in Kerman (Iran). For this purpose, 40 young drivers were recruited in an experiment to record their previous history of crashes, aberrant driving behaviours, as well as surrogate safety measures while driving on a specific route. According to the results, DBQ is a useful indicator to measure a driver's level of safe driving style since it considers the history of crashes, near-crash incidents and dangerous driving behaviours. In this study, DBQ was primarily affected by the frequency of previous crashes. In Iran, drivers with dangerous driving behaviours pay the same insurance premium as those with relatively safer driving habits. Due to the disregard of a complete driving history, the insurance premiums determination process is not fair. According to this paper, usage-based insurance pricing can become fair and dependent upon a driver's behaviour by using DBQ.