Vigilance Decrement During On-Road Partially Automated Driving Across Four Systems

Strayer, David L. · 2023 · Human Factors The Journal of the Human Factors and Ergonomics Society

DOI: 10.1177/00187208231189658

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 whether driver vigilance declines during real-world operation of partially automated (SAE Level 2) driving systems and how this decrement varies across different vehicle manufacturers. While previous research established vigilance decrements in manual and fully automated driving using simulators, the validity of detection tasks for measuring vigilance in on-road partial automation remained unproven. Additionally, prior studies typically examined single vehicle systems, failing to isolate the impact of unique system designs on driver behavior. The authors aimed to validate detection tasks as a metric for vigilance in real-world Level 2 driving and to determine if specific vehicle systems induce distinct vigilance patterns. The researchers conducted an on-road experiment with 71 participants who drove four vehicles equipped with Level 2 systems: a Tesla Model 3 (Autopilot), Cadillac CT6 (Super Cruise), Volvo XC90 (Pilot Assist), and Nissan Rogue (ProPILOT Assist). Each participant drove each vehicle twice—once in manual mode and once in partially automated mode—on a mountainous section of Interstate 80. During approximately 20-minute drives, participants performed a detection response task, pressing a button upon feeling a vibration stimulus presented every 3–5 seconds. Response times were recorded and analyzed using Bayesian statistics to assess the effects of driving mode, vehicle type, and time period (split into eight intervals) on performance. Results indicated a significant main effect of time period, with response times increasing from an average of 477 ms in the first period to 545 ms in the eighth, confirming a vigilance decrement in both driving modes. Crucially, a significant interaction between driving mode and time period revealed that the decline in performance was more severe during partially automated driving than manual driving. Vehicle-specific analysis showed significant differences in performance across systems. The Volvo XC90 exhibited the slowest response times, performing significantly worse than the Tesla, Cadillac, and Nissan. No significant differences were found among the Tesla, Cadillac, and Nissan vehicles. The lack of a vehicle-by-mode interaction suggested that vehicle-specific effects were additive rather than unique to the automated mode. The findings demonstrate that detection tasks are effective for measuring vigilance decrements in real-world partially automated driving and that Level 2 systems exacerbate the decline in driver attention compared to manual driving. This suggests a potential safety risk associated with automation-induced complacency or under-arousal. Furthermore, the study highlights that vehicle-specific interface designs significantly influence driver vigilance, with the Volvo’s interface associated with greater cognitive demand or distraction. These results imply that human factors practitioners and regulators must consider unique system designs when evaluating the safety of automated driving features, as performance risks are not uniform across all Level 2 systems.

Key finding

Drivers exhibited a significantly greater decline in detection task performance over time during partially automated driving compared to manual driving, with the Volvo system showing the slowest response times among the four tested vehicles.

Methodology

on_road

Sample size: 71

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. Discovered via unpaywall on 2026-05-07 (8 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

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

• vigilance
• automation
• automation complacency bias
• situational awareness
• mode awareness
• automation surprise

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: measurement protocol
• Theoretical Contribution: conceptual framework