%0 Generic %A Melman, Timo %A Weijerman, Mark %A de Winter, Joost %A Abbink, D.A. (David) %D 2022 %T Supplementary data for the paper 'Should steering settings be changed by the driver or by the vehicle itself?' %U https://data.4tu.nl/articles/dataset/Supplementary_data_for_the_paper_Should_steering_settings_be_changed_by_the_driver_or_by_the_vehicle_itself_/20484999/1 %R 10.4121/20484999.v1 %K Variable steering ratio %K Adaptive automation %K Adaptable automation %K Curve driving %K Driving simulator %K Function allocation %X
Objective. This paper aimed to investigate whether steering gain (SG) mode changes should be made by the driver or automatically. Introduction. Cars are increasingly computerized, and vehicle settings such as SG can now be altered during driving. However, it is unknown whether transitions in SG should be adaptable (i.e., triggered by driver input) or adaptive (i.e., triggered automatically). We examined this question for road segments expected to require different SG. Methods. Twenty-four participants drove under four conditions in a simulator: fixed low gain (FL), fixed high gain (FH), a machine-initiated steering system, which switched between the two SG levels at predetermined locations (MI), and a driver-initiated steering system, in which the SG could be changed by pressing a button on the steering wheel (DI). Results. Participants showed poorer lane-keeping and reported higher effort for FH compared to FL on straights, while the opposite held true on curved roads. On curved roads, the MI condition yielded better lane-keeping and lower subjective effort than the DI condition. However, a substantial portion of the drivers gave low preference rankings to the MI concept. Conclusion. Drivers prefer and benefit from a steering system with a variable rather than fixed gain. Furthermore, although automatic SG transitions reduce effort, some drivers reject this concept. Application. As the state of technology advances, MI transitions are becoming increasingly feasible, but whether drivers would want to delegate their decision-making authority to a machine remains a moot point.
%I 4TU.ResearchData