Fatigue Detection for Accident Avoidance using Mobile Application

Drowsiness and fatigue are primary reasons for road accidents worldwide, and hence, efficient safety measures are necessary. This paper presents a Driver Fatigue Detection and Accident-Avoidance System as an android app based on the MediaPipe framework and Kotlin. This app utilizes real-time face detection to calculate the facial features and blink rate of the driver, enabling early detection of drowsiness. On detection of fatigue symptoms, an audio alarm is triggered to remind the driver to regain attention. In extreme cases where the driver is unresponsive, emergency alerts with accurate location are sent to family members, enabling early intervention. To promote safe and sustainable driving habits, the app monitors extended driving hours and suggests resting breaks following extended hours of driving. A hardware module consisting of an IR sensor, phone holder mount, buzzer and adapter, improves the system's efficiency and reliability. The IR sensor prevents the system from being triggered unless the device is mounted, and the adapter provides continuous power during extended usage. The combined system combines state-of-the-art AI-based fatigue detection with real-time emergency response and active monitoring, providing an effective and pragmatic solution to road accident prevention due to driver fatigue, thereby enhancing road safety and driver health.