Driver drowsiness is a major cause for highway accidents on Indian roads that leads to loss of human life and severe permanent injuries. To counterattack this situation a reliable driver drowsiness model which alerts and controls the driver before a mishap happens. We design this real-time project for Light motor vehicles and transportation trucks or cargo which are heavy-duty vehicles. This working system uses a Web camera that continuously monitors the driver in live mode which will keep a track to detect the driver’s condition while driving. Once Drowsiness is identified the system alerts the driver through a speaker placed in the system and the relay slows down the vehicle. The system also constitutes alcohol sensor to detect if the driver is drunk or not and the temperature of the driver also is monitored using temperature sensor and a vibration sensor detects accident with a GPS device to locate where the accident has taken place and to update GPS location to the nearby hospital. We also use an ultrasonic sensor to prevent collision between vehicles on highway. An alert message will be generated to the frontend page which is a web-based app when the driver is not in the position of continuing his journey to reach his destination when drowsiness conditions are observed by the Webcam which is done with the help of Haar Cascade Classifier which is an effective and reliable object detection approach used in this project. The alert message will be sent to the webserver interconnected with IoT. The Raspberry Pi 3 will be having Raspbian OS which will be installed by NOOBs. Raspbian is a Linux system distribution.

Keywords : Drowsiness Detection, Raspberry Pi 3b, Haar Cascade Classifier, GPS, Vibration sensor, Alcohol sensor and Temperature sensor, Ultrasonic sensor, OpenCV, Eye Aspect Ratio (EAR)