Allocation of Drivers’ Visual Attention During Preliminary Uses of Automated Driving: A Wizard-of-Oz Study

Barré, Jessy; Sahai, Aisha; Bueno, Mercedes · 2022 · Crossref

DOI: 10.1177/03611981221108980

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates how drivers allocate visual attention during their initial experiences with automated driving (AD), specifically examining the influence of training methods and individual psychological factors. The research addresses the safety concern that drivers may lose situation awareness when transitioning from manual to automated control, potentially compromising their ability to intervene safely during system failures. The authors hypothesized that visual monitoring of the road and rear-view mirrors would decrease over time, particularly for drivers with higher trust in automation, and that practical training would reduce excessive monitoring compared to theoretical training methods. The experiment utilized a Wizard-of-Oz setup on a public road with 45 novice AD users. Participants were assigned to one of three training conditions: paper-based instructions, a video tutorial, or a hands-on practice session. Each participant completed two driving sessions: a forward path involving 10 minutes of AD with a mandatory non-driving-related task (NDRT) on an iPad, and a return path with 10 minutes of AD and an optional NDRT. Both sessions concluded with a request to intervene (RTI). Visual attention was measured via video analysis, counting the number and duration of gazes toward the road and rear-view mirrors. Self-assessment questionnaires measured trust, acceptability, and technophilia before and after the experiment. Results indicated that visual attention to the road and mirrors decreased significantly over time, particularly in the forward path. Drivers with low initial trust in the AD system glanced at the road and mirrors more frequently than those with high trust. Additionally, participants with high technophilia spent less time monitoring traffic during the forward path. Regarding training, the practice group exhibited fewer and shorter gazes toward the road and mirrors during the initial phase of the forward path compared to the video and paper groups. On the return path, the video training group monitored the environment more frequently than the practice and paper groups. During takeover requests, participants looked at rear-view mirrors an average of 10.05 seconds after an urgent RTI and 14.0 seconds after a nonurgent RTI. Most drivers gazed at the human-machine interface rather than the road immediately upon receiving an RTI. The findings suggest that while trust and technophilia influence monitoring behavior, practical training is more effective than theoretical methods in reducing excessive visual checks during early AD use. However, the delayed mirror checks during takeover requests raise concerns about drivers' situation awareness and their capacity to intervene safely in critical situations. The study highlights the need for training protocols that enhance trust and situational awareness to ensure safe human-automation cooperation.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• attention allocation
• visual
• temporal
• gaze based attention detection
• eye movements scanning
• automation

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: behavioral performance data
• Methodological Resource: tool software, measurement protocol