Citation

Wörle J, Metz B. Transp. Res. F Traffic Psychol. Behav. 2023; 99: 460-472.

Copyright

(Copyright © 2023, Elsevier Publishing)

DOI

10.1016/j.trf.2023.10.023

PMID

unavailable

Abstract

In a driving simulator study on behavioural adaptation to a level 3 automated driving system (ADS), half of the participants slept during the study. Sleep was found to impair driver's take-over performance and lead to critical situations in previous studies. That is why in a follow-up interview study, we tried to understand if and why users would show this potentially dangerous behaviour when using a level 3 ADS in reality. The majority of participants who stated their willingness to sleep during automated driving in reality had a correct mental model of the level 3 ADS and understood that sleep was not allowed. Participants with willingness to sleep had higher trust levels and showed more trust-related behaviours. They slept more frequently during the study and experienced more take-over situations after sleep which they solved with less errors than participants without the willingness to sleep. Semi-structured interviews on the willingness to sleep with a subsample revealed that most of the participants with the willingness to sleep would sleep only under certain circumstances and when they had gained some experience with the ADS. The findings suggest that after some experience with the ADS, drivers are at risk of becoming complacent and they might sleep. Vehicle designers must consider automation complacency in the design of automation designing for appropriate trust levels or by using driver monitoring systems to detect sleep onset.

Language: en

Keywords

Behavioural adaptation; Conditionally automated driving; Misuse of automation; Sleep