Effects of automation trust in drivers’ visual distraction during automation

Zhang, Yijing; Ma, Jinfei; Pan, Chunyang; Chang, Ruosong · 2021 · Crossref

DOI: 10.1371/journal.pone.0257201

archive: archived pipeline: cataloged

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

Summary

This study investigates how automation trust influences drivers’ visual distraction and supervisory performance during Level 2 automated driving. The research addresses the concern that drivers who trust automated systems may become overly relaxed, leading to increased engagement in non-driving-related tasks (NDRTs) and reduced readiness to take over. The authors hypothesized that drivers with high automation trust would exhibit lower mental workload and greater attention to visual NDRTs, and that more entertaining videos (variety shows) would capture more attention than less interesting ones (news). The experiment involved 48 drivers (aged 20–35) with no prior automated vehicle experience, recruited in Dalian, China. Participants were divided into high and low trust groups based on their scores on a 21-item automation trust scale. They performed a 30-minute simulated automated driving task on a monotonous highway using a driving simulator. The design was a 2 (trust type: high vs. low) × 2 (video type: variety-show vs. news) × 3 (measurement stage: 1–3) mixed design. Drivers completed detection-response tasks (DRTs) where they had to brake upon seeing pedestrian images. Eye movements were recorded using Tobii Pro Glasses II, and mental workload was assessed via the NASA-TLX. Visual NDRTs were presented on a tablet mounted on the steering wheel. Results indicated that drivers in the high trust group reported significantly lower cognitive load and effort levels on the NASA-TLX compared to the low trust group. High-trust drivers also spent significantly more time fixating on the NDRT area, particularly in the early stages of the drive. DRT accuracy was significantly lower for high-trust drivers, especially when watching variety-show videos in the first stage. Reaction times to DRTs were longer for high-trust drivers in the first stage when watching variety-show videos. Eye movement analysis showed that total fixation duration in the front road area was a significant predictor of DRT accuracy; drivers who looked more at the road performed better. Pupil diameter was larger for high-trust drivers and those watching variety-show videos in the first stage, suggesting higher arousal or attentional engagement with the entertainment. The findings suggest that high automation trust leads to reduced mental workload and increased visual distraction by non-driving-related tasks, which can impair drivers’ ability to respond to takeover requests. The type of entertainment matters, with more interesting content causing greater distraction. These results highlight the need for designing automated systems that manage driver trust to maintain appropriate supervisory vigilance, particularly by monitoring and mitigating the risks associated with high-trust drivers engaging in visually demanding NDRTs.

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 qwen3.8-27b-gittensor on 2026-08-23; verification: pending re-verification.

Topics

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

• trust calibration
• automation
• automation surprise
• visual
• trust in automation foundations
• situational awareness

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: self report data, behavioral performance data
• Theoretical Contribution: conceptual framework