Real-Time Driver Fatigue Detection Device

Driver fatigue is one of the leading causes of accidents in long-haul transport, but fleet operators had no way to detect it in real time. Tracking driving hours and relying on supervisor judgment wasn't enough to catch a drowsy driver before something went wrong. What needed was a device inside the cab that could spot fatigue the moment it started, day or night, and warn the driver before it turned dangerous.

We built a compact dashboard-mounted device with an IR camera that monitors the driver's face continuously day and night using IR LED illumination. A YOLO-based model running on the embedded processor analyzes eye closure duration, yawning frequency, and head pose deviation in real time. The moment early fatigue signs are detected, a multi-stage alert escalates from a soft warning to a strong in-cab buzzer and LED alert. Every fatigue event is logged with an annotated snapshot and video clip, stored locally, and pushed via 4G to a fleet management portal where supervisors can review events, track driver fatigue scores, and run coaching sessions.