SomniScan is a standalone, low-cost driver drowsiness detection and alert system built on a Raspberry Pi 4. Using computer vision and multi-feature sensor fusion, it monitors a driver's eye closure, yawning, and head pose in real time — including in complete darkness via an infrared night-vision subsystem. When fatigue is detected, the system escalates through a multi-stage alert process using haptic vibration and audio warnings. SomniScan requires no vehicle integration or wearable sensors, making it accessible to any driver at a total cost of approximately $243.
Team Members: Tanbir Badhan, Narayan Vaid, Jaskeerat Sandhu
Faculty Advisor: Dr. Xun Li
Program: ELEC ENG 4OI6 — Electrical & Computer Engineering Capstone, McMaster University