Driver's arousal and workload under partial vehicle automation

Strayer, DL; Cooper, JM; Sanbonmatsu, DM; Erickson, GG; Simmons, TG; Erickson, Gus G. · 2020 · publications_jsonl

archive: archived pipeline: cataloged verified

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

Summary

This study investigates driver arousal and workload when operating vehicles with Level-2 partial automation compared to manual driving (Level-0). The research was motivated by the SAE requirement that drivers continuously monitor the environment while using automation, a shift from active control to passive monitoring that may erode attention, arousal, and readiness to respond. While prior research suggested potential for distraction or fatigue, little real-world data existed on how partial automation affects physiological arousal and cognitive workload. The study aimed to determine if Level-2 automation leads to under-arousal (disengagement) or over-arousal (stress) relative to manual driving. The researchers conducted an on-road evaluation involving 71 participants aged 21–64, divided into younger (21–42) and older (43–64) cohorts. Participants drove four different vehicles equipped with Level-2 automation (Cadillac CT6, Nissan Rogue, Tesla Model 3, and Volvo XC90) on Interstates 80 and 15. The experimental design was a 4 (Vehicle) x 2 (Age) x 2 (Automation Level) x 2 (Interstate) factorial. Data collection utilized a multi-modal approach: a Detection Response Task (DRT) measured reaction time and hit rate; electrocardiography (ECG) recorded heart rate and heart rate variability (RMSSD); electroencephalography (EEG) measured parietal alpha power; and surveys assessed subjective nervousness, inattention, and excitement. Linear mixed-effects models were used to analyze the data, accommodating the planned missing data design where participants drove varying numbers of vehicles. The results indicated that drivers exhibited slightly enhanced engagement with the driving task under Level-2 automation. Specifically, parietal alpha power and DRT hit rates were lower, while DRT reaction times were longer during automated driving compared to manual driving. These physiological and performance metrics suggest that participants directed more attention to the driving environment when automation was engaged. Subjectively, participants reported higher levels of excitement and nervousness during Level-2 automation. However, the magnitude of these differences was small, accounting for at most 2.7% of the variance across measures. Crucially, the study found no meaningful differences in overall arousal or workload between Level-0 and Level-2 conditions. The data did not support the hypothesis that partial automation leads to significant under-arousal or disengagement in this context. The findings suggest that drivers are better than expected at maintaining adequate attention and monitoring the environment during partial automation, at least in the short term. The slight increase in attentional focus and subjective arousal indicates that the transition to a monitoring role does not immediately induce complacency or fatigue. However, the authors note that the presence of a researcher and data equipment may have influenced behavior. They conclude that future research should examine long-term exposure to determine if drivers transition from an initial "novelty" phase of high vigilance to an "experienced" phase characterized by over-reliance and reduced arousal. The study implies that while current Level-2 systems do not significantly degrade driver performance or arousal in controlled on-road settings, sustained attention over longer periods remains a critical area for further investigation.

Key finding

Under Level-2 automation, parietal alpha power and DRT hit rate were lower and DRT reaction time was longer than under manual driving, and drivers reported more nervousness and excitement, consistent with slightly increased attention to the driving environment rather than disengagement. Effects were statistically significant but small (at most ~2.7% of variance), indicating no meaningful arousal or workload difference between Level-0 and Level-2 in this on-road sample.

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 author_sweep_intake on 2026-05-28 (5 acquisition events logged).

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.

• hands on hands off engagement
• situational awareness
• automation
• automation surprise
• mode awareness
• workload measurement

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: physiological data
• Theoretical Contribution: theory or model, conceptual framework