Driver Distraction System and Vehicle Speed Controlling Using AI and IOT

Clearly, the road safety business requires more oversight, as over 50 million automobiles are sold annually and over 1.3 million people die in car accidents. India, which accounts for 11% of all road fatalities globally, is a developing country where driver conduct needs special attention. Redirections account for 78% of driver-involved incidents, according to the report. As a result of drivers being preoccupied with other things, such as texting or chatting on the phone, road safety is significantly diminished. A cutting-edge ML model that employs computer vision techniques to reliably detect various forms of driver distractions in real-time is the primary focus of this endeavor. We aim to address this important matter. Using a convolutional neural network (CNN) ensemble, such as the state-of-the-art YOLO model. In order to proactively intervene and avoid accidents, we want to use deep learning and image recognition to categorize distractions. Studying the model’s overall performance, scalability, and classification accuracy is crucial for deployment on edge devices. We assess the approach’s viability and generalizability by looking at performance indicators like inference time and resource use. We propose a model that, with the help of an IoT model and a microcontroller, can do more than only detect distractions; it can also alert the driver and make the vehicle slow down.