Design of an Event Data Recording System with Deep Learning Based Driver Inaction Detection

This project is a proposal of a Event data recording system with deep learning based driver inaction detection that is aimed at improving safety on the road, monitoring drivers, and preventing accidents based on intelligent sensing and in-time communication. The system is a miniature vehicle black box that continuously monitors important driving conditions, including speed of the vehicle, braking, acceleration, seatbelt, and driver state. The ESP32 microcontroller is the central processing unit that will connect to sensors, such as a gas sensor, brake button, seatbelt wire, accelerator potentiometer, and so on, to detect the behavior of the vehicle and the driver. In unstable conditions, e.g., fatigue detection, over-speeding, or intoxication, the system emits immediate notifications and sends event data to a remote monitoring (wifi or GSM SI900A) to a PHP-based cloud server. The data stored could be analyzed in the future to determine the performance of the driver or to analyze accidents. Also, there is an assisted vision system that is integrated with a camera that identifies inattention and drowsiness on the road to increase road safety. The proposed system guarantees the constant monitoring, automatic notification, and safe data recording by combining IoT connectivity and smart decision-making. The given innovation can improve the safety of the vehicles, encourage the creation of responsible drivers, and facilitate the trustworthy post-incident analysis, which is a step in the progress of smart and interconnected transportation systems.