Real-time Driver Drowsiness Detection and Assistance System using Machine Learning and IoT

Road transport is often regarded as the most cost-convenient and preferred mode of transportation for both freight and passengers. Road transportation makes up about 87% of all passenger transport and 60% of all freight transportation in the nation. Some of the advantages of road transportation are its ease of accessibility, flexibility to individual demands, and cost savings. But all this comes with a cost. According to MoRTH., the number of people killed in traffic accidents caused by driving vehicles has consistently risen every year. Particularly, there are more fatalities and serious injuries than minor ones, and the harm caused by major accidents is growing. Heavy cargo truck accidents and high-speed bus crashes while driving at night has particularly become major social issues. These misadventures are caused by driver fatigue, microsleep, and drowsiness. Out of these, driver fatigue contributes more to freight accidents. This paper aims to implement an intelligent processing system that drastically reduces road accidents. This research will develop a system capable of monitoring the driver in real-time and m, ensuring he is in his active state, letting him in real-time in case sleep or fatigue is detected. Existing drowsiness detection systems just have accuracy for detection but are not up to the mark when it comes to prevention. We will design a system to alert the driver physically. Here, this research will make use of a machine learning algorithm called the Haar cascades algorithm and the Internet of Things for hardware setup. This technology will alert the driver if his or her eyes are maintained closed for an extended period by triggering an alarm and spraying water at a positioned spot to get back to his active state. Also, a cloud system will be involved to send an alert message to the concerned person to be aware of the driver’s condition.