A Proactive In-Car Alarm System for Driver Fatigue Detection Using Machine Learning

In numerous research studies, it has been consistently established that driver fatigue represents a predominant factor contributing to a significant portion of traffic accidents, incidents that could have been averted if measures were in place to rouse drowsy drivers. In the scope of this paper, we delve into the development of an affordable and highly efficient in-car alarm system, meticulously engineered to mitigate the risks associated with driver fatigue. Our primary focus centers on the identification of exhaustion and drowsiness indicators, thus laying the foundation for a proactive approach to road safety. Within this study, we employ a drowsiness detection methodology, harnessing the capabilities of machine learning, which serves as the bedrock of the prototype presented herein. Furthermore, our investigations underscore the critical importance of incorporating drivers' EEG signals for optimal sleepiness detection. Beyond this, we introduce a dedicated machine learning model designed to analyze drivers' EEG data, promptly activating alerts, when necessary, thereby further enhancing the safety features integrated into our prototype.