Car-following crash risk analysis in a connected environment: A Bayesian non-stationary generalised extreme value model

A connected environment provides driving aids to assist drivers in decision-making and aims to make driving manoeuvres safer by minimising uncertainty associated with deci-sions. The role of a connected environment becomes vital for car-following manoeuvres in a safety-critical event, whereby drivers follow a lead vehicle, and if timely action is not taken, it is likely to lead to a rear-end collision. Moreover, how different drivers per-ceive and react to the same information needs to be explored to understand the differential effects of a connected environment on car-following behaviour. As such, this study inves-tigated the effects of the traditional and connected environments on car-following crash risk using traffic conflict techniques. Data were collected using the CARRS-Q advanced driv-ing simulator, whereby 78 participants performed a car-following task in two randomised driving conditions: baseline (without driving aids) and connected environment (with driv-ing aids). The safety-critical event in the car-following scenario was the leader's hard brak-ing, for which participants received advance information, besides several other driving aids. Modified time-to-collision was used as the traffic conflict measure for characterising rear-end crash risk and modelled using a generalised extreme value (GEV) model in the Bayesian framework. This model incorporated driving-related factors and driver demo-graphics to address the non-stationarity issue of traffic extremes. Results reveal that the car-following crash risk is significantly reduced in the connected environment. Further, using the developed model, separate GEV distributions were estimated for each individual driver, providing insights into the heterogeneous effects of the connected environment on crash risk. The developed model was employed to understand the crash risk across differ-ent driver characteristics, and results suggest that crash risk decreases for all age groups and gender, with the maximum safety benefits obtained by young and female drivers. The findings of this study shed light on the efficacy of the connected environment in improving car-following behaviour and drivers' ability to make safer decisions. (c) 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Automated summary

This study investigates the effects of traditional and connected environments on car-following crash risk using traffic conflict techniques. The results show that the connected environment significantly reduces car-following crash risk.