The role of take-over cue informativity in interrupted take-over requests in a semi-automated driving scenario

Berger, Alexander; Damm, Nicole; Baumann, Martin; Kiefer, Markus · 2026 · Scientific Reports (Nature)

DOI: 10.1038/s41598-026-36614-y

URL: https://www.nature.com/articles/s41598-026-36614-y

archive: archived pipeline: cataloged verified

Summary

Scientific Reports study testing whether informative take-over-request (TOR) cues retain their advantage when an interrupting task arrives between cue and take-over action in semi-automated (SAE Level 3) driving. Forty-seven participants (45 retained after exclusions) watched videos of an ego-vehicle on a road; TOR cues were either task-specific informative or generic, and on half of trials a lane-decision interrupting task interposed between cue and take-over. Informative TOR cues improved take-over performance overall, but the interrupting task attenuated this advantage; the interruption cost peaked partway through the session and faded with practice. Practice and post-cue distraction therefore moderate informative-cue effectiveness.

Key finding

Informative TOR cues help take-over performance but their benefit shrinks when an interrupting task follows the cue; the cost is largest mid-session and dissipates with practice.

Methodology

Video-based experiment, within-subject TOR informativity x interruption factorial

Sample size: N=47 recruited; N=45 analyzed (1 excluded as outlier per preregistered criteria)

Quality score: 5 / 5

Topics