Prevalence of Drowsy Driving Crashes: Estimates from a Large-Scale Naturalistic Driving Study
archive: archived pipeline: cataloged verified
Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)
Summary
This research brief addresses the significant discrepancy between official government statistics, which attribute only 1%–2% of motor vehicle crashes to drowsy driving, and expert estimates suggesting the true prevalence is much higher. The study aims to provide a more accurate estimate of drowsy driving involvement in crashes by analyzing data from a large-scale naturalistic driving study, thereby overcoming the limitations of police reports that often fail to identify drowsiness due to the lack of objective roadside testing and driver underreporting. The analysis utilized data from the Second Strategic Highway Research Program Naturalistic Driving Study (SHRP 2 NDS), which monitored 3,593 drivers across six U.S. sites between 2010 and 2013 using in-vehicle cameras. The study focused on 701 crashes (severe, moderate, and minor) for which video data allowed for the assessment of driver drowsiness. Drowsiness was quantified using the PERCLOS measure, a validated metric defining drowsiness as eyelids being closed for 12% or more of the time during the three minutes (or one minute, if three minutes were unavailable) preceding the crash. Trained analysts coded video frames to determine eyelid closure, excluding crashes where the driver’s eyes were not visible for at least 75% of the coding period. The results indicate that driver drowsiness was present in 8.8%–9.5% of all examined crashes and in 10.6%–10.8% of crashes severe enough to be reportable to police (involving injury, airbag deployment, or significant property damage). Drowsiness prevalence varied significantly by lighting conditions, with crashes occurring in darkness being more than three times as likely to involve drowsiness as those in daylight. More than half of all drowsy-driving crashes occurred in darkness. However, variation by driver age, sex, and crash severity was not statistically significant. These findings stand in stark contrast to National Highway Traffic Safety Administration (NHTSA) data, which reported drowsiness involvement in only 1.4%–2.4% of police-reported crashes. The study concludes that drowsy driving is a substantially larger traffic safety problem than official statistics suggest, corroborating previous research that characterizes police-reported data as a vast underestimate. The findings highlight the utility of naturalistic driving data and objective measures like PERCLOS in identifying impairment that is difficult to ascertain post-crash. While the study did not include fatal crashes, the strong association between darkness and drowsiness suggests that drowsiness likely plays a significant role in fatal crashes, which occur disproportionately at night. The authors note limitations, including the volunteer nature of the driver sample and the exclusion of crashes with poor video quality, but assert that the results provide a more reliable baseline for understanding the scope of drowsy driving.
Key finding
PERCLOS-coded pre-crash video indicates drowsiness in 8.8%–9.5% of crashes (10.6%–10.8% of police-reportable crashes), roughly five to seven times higher than NHTSA police-report estimates.
Methodology
naturalistic
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 bulk_ingest_aaa_foundation on 2026-05-23 (5 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | aaa_foundation | — | — | 2 | 2026-05-23 |
| archive | success | — | — | — | 1 | 2026-05-23 |
| extract | success | cached | — | — | 2 | 2026-06-10 |
| clean | success | — | — | — | 1 | 2026-06-01 |
| chunk | success | — | — | — | 1 | 2026-06-01 |
| embed | success | — | — | — | 1 | 2026-06-02 |
| enrich | success | — | — | — | 1 | 2026-05-23 |
| promote | success | — | — | — | 1 | 2026-05-23 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 19 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
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.
- drowsiness detection algorithms
- drowsiness
- incidence prevalence
- sleep deprivation
- naturalistic crash near crash
- drowsy as impairment
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: crash risk outcomes, observational prevalence
- Methodological Resource: dataset resource