Citation

Maag C, Schömig N, Naujoks F, Karl I, Keinath A, Neukum A. Transp. Res. F Traffic Psychol. Behav. 2023; 92: 201-219.

Copyright

(Copyright © 2023, Elsevier Publishing)

DOI

10.1016/j.trf.2022.11.010

PMID

unavailable

Abstract

Augmented reality (AR) technology could establish direct relationships between displayed information and objects in the real driving environment, e.g. by highlighting relevant objects in the traffic environment. However, it is unclear how these potential benefits of augmentation affect drivers' distraction from the driving task and their level of workload. Since gaze detection cannot separate glances at AR-elements from glances at the driving scene, new evaluation criteria are needed for measuring the distraction potential of AR-elements. The paper evaluates a methodology based on the Detection Response Task (DRT) for measuring the distraction and workload effects of three different display technologies: Head-Down Display (HDD), Head-Up Display (HUD), and Augmented-Reality HUD (AR-HUD). The DRT measures cognitive and visual workload based on a recognition-response task using the two parameters, reaction time and misses. Two experimental studies using driving simulation (N = 24) and production vehicles in real traffic (N = 24) are reported. As use cases, the subjects had to navigate through an urban traffic environment and pass different intersections. During that they were assisted by navigation systems based on the above mentioned display technologies. The resulting workload is compared to a baseline drive without the need to navigate and a reference drive in which the subjects had to work on a secondary task while driving. The results of the DRT are compared with subjective measures, gaze behaviour, and driving performance. In addition, drivers' preference and user experience regarding the three display technologies and their feedback concerning the DRT method were assessed. The results of both experimental studies show that drivers prefer the AR-HUD display technology. Reasons are mainly the hedonic aspects of the AR-HUD such as the novelty of the technology. Nevertheless, drivers report that AR-elements could obstruct important areas of the road and distract more from the driving task than HUDs. The subjective and visual-cognitive workload measured by the DRT do not differ between the HMI conditions. An analysis for each type of intersection yields most misses for the AR-HUD condition and fewest misses in the HDD condition. However, visual workload (i.e. eyes-off-road time) is higher for HDD compared to (AR-)HUD. Feedback by participants reveals that the DRT as a method for workload assessment is well accepted. All in all, the two DRT measures, misses and reaction time, allow a valid and objective workload measurement and could serve as a supplement or substitute for gaze analysis. The discussion gives indications for the measurement of workload effects caused by AR-HUD using DRT.

Language: en

Keywords

(Head-up display; Augmented reality; Distraction; Human-machine interface); Workload measurement