Detection of Driver Fatigue Caused by Sleep Deprivation
DOI: 10.1109/tsmca.2009.2018634
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 the detection of driver fatigue caused by sleep deprivation by analyzing driver–vehicle interaction characteristics rather than relying on physiological signals like eye closure or EEG. The research is motivated by the "camouflage" nature of drowsiness, where sleep-deprived drivers may appear alert and maintain performance in routine tasks but suffer significant degradation in their ability to handle unexpected disturbances. The authors aim to identify specific driving metrics that reliably indicate drowsiness and propose a probabilistic framework for detection. The experimental design involved 12 male subjects participating in two sessions: one under non-sleep-deprived conditions (7–8 hours of sleep per day for a week) and one under partial sleep-deprivation conditions (less than 7 hours two days prior and less than 4 hours on the eve of the test). Sleep was objectively measured using actigraphy watches. Subjects performed simulated driving tasks in a laboratory test bed, including routine lane-tracking tasks (straight road, following a lead vehicle) and tasks involving disturbances (wind gusts, altered steering dynamics, curved roads). Additionally, subjects completed stimulus-response tasks requiring immediate reactions to visual or auditory cues. Performance was evaluated using metrics such as Root-Mean-Square error with threshold (RMT) for tracking accuracy and reaction time for stimulus responses. The results revealed that sleep deprivation had a differential effect on cognitive functions. Drivers under partial sleep deprivation showed statistically significant performance degradation in tasks requiring adaptation to unexpected disturbances, such as wind gusts, altered steering dynamics, and curved-lane tracking. However, their performance in routine, rule-based tasks like straight-lane tracking and vehicle following did not differ significantly from the non-sleep-deprived group. This confirms the hypothesis that drowsy drivers can mask their impairment during routine driving but fail when faced with novel or complex situations. Furthermore, the study demonstrated that reaction times to unexpected stimuli were significantly slower in the sleep-deprived group. The significance of this work lies in its identification of specific driver–vehicle interaction metrics that can serve as reliable indicators of drowsiness, particularly those related to handling unexpected disturbances. The authors introduce a probabilistic framework based on Bayesian networks to infer driver drowsiness states, addressing temporal aspects and individual differences. This approach offers a viable alternative to physiological monitoring systems, which may fail to detect drowsiness when drivers are physically awake but cognitively impaired. The findings provide qualitative and quantitative guidelines for designing active safety systems that can detect fatigue before it leads to accidents, emphasizing the importance of monitoring performance in complex driving scenarios rather than just routine lane keeping.
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.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | OpenAlex-citations | — | — | 1 | 2026-06-19 |
| archive | success | unpaywall | — | — | 2 | 2026-06-25 |
| extract | success | cached | — | — | 2 | 2026-06-26 |
| clean | success | clean | — | — | 1 | 2026-06-19 |
| chunk | success | chunk | — | — | 1 | 2026-06-19 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-19 |
| promote | success | — | — | — | 1 | 2026-06-19 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-26 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-19 |
| verify | success | — | — | — | 1 | 2026-06-26 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
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, behavioral performance data
- Theoretical Contribution: theory or model