Real-Time Driver Drowsiness Alert System for Product Distribution Businesses in Phuket Using Android Devices

Abstract

Road accidents in product distribution businesses frequently occur during long weekends or festival periods. The cause of accidents usually comes from tired or fatigue drivers due to overwork or continuous long work. For some companies, the accident can cost up to 20% of the registered capital. This payment heavily affects the company's cash flow a nd account balance. Unfortunately, commercial devices for drowsiness detection or prevention are too expensive for SMEs. Therefore, SMEs required a low-cost, real-time, easy-to-use, and easy to install system. An Android smartphone based system to detect drowsy drivers was created. The application evaluates the drowsy drivers using ML Kit's face detection API and sends a message through Line Notify API. The results showed that the application can differentiate between normal and drowsy drivers as well as send an alarm to the drivers and manager. Moreover, the application is tolerated to the low FPS smartphone down to 3 FPS.