Fatigue Detection Systems
Drowsy driving does not announce itself with a blood alcohol reading or a positive drug test. It leaves no definitive chemical trace at a crash scene, and drivers rarely self-report fatigue after a collision. For decades, the evidentiary gap meant that fatigue’s role in crashes was systematically undercounted. As a result, carriers had little incentive to invest in proactive detection.
The research, however, is unambiguous about the scale of the problem. Not only does driver fatigue harm the general public, driver fatigue is also deadly to the drivers themselves. The National Transportation Safety Board (NTSB) estimated that driver fatigue was the principal cause in 31 percent of fatal-to-the-driver crashes, a finding drawn from an in-depth 1990 examination of heavy truck crashes across eight different states.¹ Official federal statistics have historically estimated fatigue’s share far lower, but those figures are drawn from police reports where investigators cannot measure how long a driver has been awake or observe the gradual degradation of alertness the way investigators can test for alcohol. NHTSA acknowledges that “determining a precise number of drowsy-driving crashes, injuries, and fatalities is not yet possible. Crash investigators can look for clues that drowsiness contributed to a crash, but these clues are not always identifiable or conclusive.”² There is an apparent gap between what police reports show and what the research suggests.
Fortunately, the human body produces measurable physiological signals before fatigue becomes severe enough to cause a crash. In-cab fatigue monitoring technology has been developed to detect these signs and act on them in real time before a crash occurs. Understanding what these systems do, what data they generate, and what it means when a carrier ignores or suppresses that data is essential context for any crash investigation involving a fatigued driver.
PERCLOS and Driver Monitoring
Modern in-cab fatigue detection is measured by the percentage of time the eyes are at least 80 percent closed over a given interval or PERCLOS.³ Slow eyelid droops, as opposed to full eye blinks, are the key indicator. A blink is quick and involuntary. A droop, on the other hand, descends slowly and the eyelids stays down for a fraction of a second longer than normal. Signs of droop indicate that the central nervous system is struggling to stay awake.
PERCLOS is one of the most validated indicators of general drowsiness. “PERCLOS increases with sleep deprivation, after partial sleep restriction, at nighttime, and by other drowsiness manipulations during vigilance tests, simulated driving, and on-road driving.”³ And while PERCLOS is not the only metric these systems use, it is the most scientifically grounded single indicator of fatigue onset that can be captured non-invasively through a camera.⁴
Federal Motor Carrier Safety Association (FMCSA)-funded research into driver fatigue and distraction monitoring has developed systems that combine a vision module with a driving style module. Vision models are capable of measuring driver pose and other psychophysiological measures of alertness including PERCLOS and yawning. Driving style modules are designed for detecting erratic inter- and intra-lane behavior and speed variations based on yaw rate sensor and controller area network (CAN) bus signals.⁵ The integration of multiple data streams is intentional because no single measure is fully reliable across all drivers, lighting conditions, and operating environments. Commercial systems fuse multiple behavioral indicators into a composite fatigue assessment.⁶
In-Cab Alert Systems
When a driver monitoring system detects fatigue, it responds through one or more warning modalities. When an “in-vehicle system detects a drowsy or distracted driving event, it intervenes in real time through audio, visual, and vibration alerts. Instant feedback encourages drivers to avoid risky behaviors and allows them to recognize and develop strategies to mitigate these risks.”⁷
Audio alerts are loud, distinct tones designed to interrupt the slide towards microsleep and are typically the most prominent. Visual alerts appear on a dashboard display or through a light indicator. Haptic alerts, delivered through seat or steering wheel vibration, add a physical stimulus that is harder to ignore than a sound a drowsy driver may not fully register.
The generation of an alert is a data event. Real-time data via secure web portals and regular reports provide fleet managers with valuable information and rich insights within minutes of a confirmed fatigue event.
As a result, every alert is timestamped and logged, capturing:
- When the system detected fatigue
- How severe the detection was
- Whether the driver responded
- How the vehicle was behaving at the moment of detection
Carriers with active Driver Monitoring Systems (DMS) have access to a continuous record of driver fatigue events across their fleet.
Lane Departure Warning Systems
While DMS watches the driver’s face, lane departure warning systems watch the road. Together, they create a two-channel detection framework, measuring what the driver is doing and what the truck is doing as a response.
Lane departure warning systems are designed to warn the driver when the vehicle begins to move out of its lane without a turn signal active, addressing the main causes of collisions including driver error, distractions, and drowsiness. These systems use a forward-facing camera mounted on the windshield to track lane markings.
The investigative significance of lane departure data is distinct from facial monitoring data. Facial monitoring detects the driver’s internal state. Lane departure data records the physical consequence of that state, which is the loss of vehicle control. “[Lane departure warning] activations could have a dual benefit of alerting drivers to unintentional lane departures and notifying safety managers of compliance issues with turn signal usage.”⁸
Research using National Highway Traffic Safety Administration (NHTSA) Fatality Analysis Reporting System data from 2016 to 2022 found that lane keeping assist-equipped vehicles were 24 percent less likely to be involved in fatal road departure crashes compared to non-equipped vehicles.⁹ That finding cuts two ways: it confirms the technology prevents crashes, and it means that carriers choosing not to deploy it on their fleets are making a documented, quantifiable safety tradeoff.
Pre-Crash Warnings
When a crash involves a potentially fatigued driver and the carrier’s vehicle was equipped with a DMS or lane departure warning system, the data those systems generated before and during the crash becomes a central piece of the evidentiary record.
The alert log from a DMS tells a story with timestamps. If the system recorded multiple fatigue alerts in the hour before the crash the log documents that the driver was impaired and that the carrier’s system was aware of it. When a carrier receives those alerts via its fleet management platform and does not intervene (by contacting the driver, routing the driver to a rest stop, or removing the driver from the load), the carrier has failed to prevent a crash by intentionally choosing not to act.
Fatigue management systems map fatigue across stages, from early cognitive decline to advanced drowsiness, using multiple behavioral cues alongside contextual driving data such as lane positioning and steering patterns.¹⁰ The records establish a timeline, documenting when impairment began, how it progressed, and at what point the crash occurred relative to the system’s prior warnings.
For carriers that have DMS systems but suppress, disable, or fail to maintain them, the absence of data is itself significant. A system that should have been generating alerts and was not raises questions about whether the carrier was managing its technology responsibly or avoiding a record it knew would be unfavorable.
Post-Crash Investigations & Discovery
In any crash investigation where driver fatigue is a potential factor, the inquiry should extend beyond hours-of-service logs and Electronic Logging Device data to the full range of monitoring technology the carrier had deployed.
The first step is identifying what systems the vehicle was equipped with. Carriers are not currently required by federal regulation to install DMS technology, so the presence and configuration of these systems will vary from carrier to carrier. The vehicle’s onboard systems should be documented immediately through inspection and through the manufacturer’s build records. If the vehicle had an integrated DMS, the system’s alert logs should be preserved before they are overwritten.
Preservation requests should specifically name:
Many systems retain data locally on the device and also transmit it to a carrier-hosted or vendor-hosted cloud platform. Both copies should be demanded.
The carrier’s internal records about its DMS program are also relevant.
These records include:
- A carrier’s policies on fatigue alert response
- Documentation of training provided to safety managers on interpreting alert data
- Records showing how prior fatigue alerts across its fleet were handled
A carrier that had a DMS policy on paper but no evidence of ever acting on an alert differs from a carrier that had no system at all. Timestamped fatigue monitoring alerts, coaching records, and intervention logs demonstrate a carrier’s safety culture (or the lack thereof) to fact-finders evaluating the carrier’s conduct.¹¹
Sources
- [1] Transportation Research Board and National Research Council, "Fatigue, Hours of Service, and Highway Safety," in Commercial Motor Vehicle Driver Fatigue, Long-Term Health, and Highway Safety: Research Needs (National Academies Press, 2016).
- [2] National Highway Traffic Safety Administration, "Drowsy Driving".
- [3] "PERCLOS-Based Technologies for Detecting Drowsiness: Current Evidence and Future Directions," National Center for Biotechnology Information.
- [4] Id.
- [5] Federal Motor Carrier Safety Administration, "Driver Fatigue and Distraction Monitoring and Warning System, Phase I," updated February 11, 2016.
- [6] Federal Motor Carrier Safety Administration, "Multi-Modal Driver Distraction and Fatigue Detection and Warning".
- [7] Seeing Machines, "Fleet" (Guardian product page).
- [8] National Highway Traffic Safety Administration, "Field Study of Heavy-Vehicle Crash Avoidance Systems" (Report No. DOT HS 812 280, June 2016).
- [9] National Highway Traffic Safety Administration, "Estimating Effectiveness of Lane Keeping Assist Systems in Fatal Road Departure Crashes" (Report No. DOT HS 813 663, December 2024).
- [10] Id.
- [11] FleetRabbit, "Driver Fatigue Monitoring Systems FAQs for Trucking Fleets (2026)".