Longitudinal jerk and celeration as measures of safety in bus rapid transit drivers in Tehran

Abstract

Traditionally, analysis of individual differences in road traffic crash risk has relied on after the fact crash data. Nowadays with the help of technologies like GPS, new measures are developed to assess driving risk, inferred from naturalistic driving behaviour of drivers. In this study, two main ways of modelling driver behaviour in naturalistic driving research were discussed and compared to each other. For this purpose, 176 Bus Rapid Transit (BRT) drivers were investigated during their normal driving on Tehran BRT routes. Their speed was continuously recorded by a smart phone app from which was derived deceleration, jerk and celeration measures. Analysis showed that all of the proposed measures had positive correlations with culpable crashes and that one of the jerk variables yielded the highest correlation. Measurements of speed derivatives like jerk and celeration can help to identify dangerous driving styles in public transportation and reduce the number of crashes. Although the results of this study are encouraging, further studies for longer periods are needed to improve the reliability of the measures.