Researchers at Australia's Edith Cowan University (ECU) have unveiled an artificial intelligence system that analyzes facial dynamics to detect drunk driving, fatigue, and emotional states linked to road accidents. Announced on March 11, 2026, the innovation uses 3D facial scanning to achieve 90% accuracy in alcohol detection and 95% accuracy in drowsiness identification.
How It Works
The AI model tracks subtle facial movements – including eye blinking patterns, micro-expressions, and progressive feature changes – to classify intoxication levels as sober, moderate, or severe. Unlike traditional breathalyzers, the system operates continuously without requiring driver participation, enabling real-time intervention opportunities.
Practical Applications
A companion study demonstrates enhanced low-light performance through combined infrared and color video analysis. Lead researcher Abdullah Tariq states this technology could integrate with vehicle safety systems or roadside monitoring infrastructure. "This non-invasive approach addresses multiple accident factors simultaneously," Tariq explained.
With road safety remaining a global priority, this development positions Australia at the forefront of preventive AI solutions as nations work toward 2030 traffic mortality reduction targets.
