Drowsy driving is the quiet epidemic, and why detection beats willpower

Drunk driving has a number you can blow into a tube. Drowsy driving has nothing, no breathalyser, no roadside test, no easy way to prove it after the wreck. That is precisely why it is so badly under-counted, and so dangerous.

NHTSA records around 91,000 crashes, 50,000 injuries and 795 deaths a year from drowsy driving, and explicitly notes these are underestimates (NHTSA). When researchers used in-vehicle cameras instead of police reports, the true picture was far worse: the AAA Foundation found drowsiness present in 8.8% to 9.5% of all crashes, around eight times the official rate (AAA Foundation, 2018).

Tired is a kind of drunk

The most useful way to think about fatigue is as an impairment with a dose-response curve, just like alcohol. Classic research found that being awake for about 17 hours degrades performance to roughly a 0.05% blood-alcohol level, and 24 hours awake is comparable to 0.10% (Dawson & Reid, 1997). We would never let a driver start a shift at 0.05%. We do it constantly with hours-awake, because nobody is measuring it.

Why willpower is the wrong tool

The cruel twist of fatigue is that it attacks the very faculty you would use to judge it. As people get drowsier, they get worse at noticing they are drowsy, and the "I'll just push through" decision is itself a symptom. Microsleeps, brief involuntary lapses of a few seconds, happen with no warning the person can feel. You cannot will your way out of a state that disables your judgement.

What a camera sees that the driver can't

This is the case for detection. A driver-monitoring camera does not rely on the driver's self-assessment, it watches the physical signs, prolonged eye-closure, a falling blink pattern, the head-nod, and reacts in real time with an in-cab alert. Independent research is still maturing and the strongest fleet studies are small, one 2025 Monash study of twelve trucks found camera-based monitoring roughly halved fatigue events when active, though drivers habituated to the alerts over time (JOEM, 2025). The lesson is not "the camera fixes it," it is "the camera plus a coaching loop fixes it," because the alert that wakes you tonight only changes tomorrow if someone follows up.

You can't coach willpower into a tired brain. You can give it a second pair of eyes that never blinks.

Drowsy driving stays a quiet epidemic for as long as we rely on tired people to police their own tiredness. The way out is measurement: see it, alert on it, and act on it.