Examining the driver-pedestrian interaction at pedestrian crossings in the connected environment

A Hazard-based duration modelling approach

Journal Article (2021)
Author(s)

Mohammed Mazharul Haque (Queensland University of Technology)

O. Oviedo-Trespalacios (Queensland University of Technology)

Anshuman Sharma (Indian Institute of Science)

Zuduo Zheng (University of Queensland)

Affiliation
External organisation
DOI related publication
https://doi.org/10.1016/j.tra.2021.05.014
More Info
expand_more
Publication Year
2021
Language
English
Affiliation
External organisation
Volume number
150
Pages (from-to)
33-48

Abstract

The availability of advisory warnings via Vehicle-to-Vehicle and Vehicle-to-Infrastructure communication in the connected environments is expected to gradually increase over the next few years. Much of the research on advisory warning systems have examined driving behaviour in response to unexpected driving hazards; however, very little research has been conducted on common driving interactions such as interacting with pedestrians at pedestrian crossings. Therefore, the aim of this study is to investigate the effects of the connected environment on driving behaviour at pedestrian crossings. The connected environment was designed within the CARRS-Q advanced driving simulator. A combination of auditory (beep sound) and imagery message was simultaneously displayed on the windscreen to advise the driver on the presence of a pedestrian entering from a sidewalk. Seventy-eight licensed drivers drove the simulator in two driving conditions, namely, baseline and connected environment. The participants were 18–65 years old, and a third of them were females. Drivers' response to the driving aids and the braking behaviour were analysed in the latent response phase and the observable response phase, and the corresponding response times were modelled using the hazard-based duration modelling approach. In particular, this study applied the Weibull accelerated failure time model with shared frailty accounting for multiple observations from the same driver. Results showed that the time taken to respond to the pedestrian in the latent response phase was longer when the advisory warning was provided to the drivers, but the corresponding time in the observable response phase was shorter, indicating that drivers take an informed decision in the connected environment. Moreover, the safety margin—measured in terms of time-to-collision—was higher in the connected environment than the traditional driving environment, indicating a safer behavioural adaptation towards the connected environment.

No files available

Metadata only record. There are no files for this record.