IOT-Enabled Real-Time Driver Drowsiness Detection System

Abstract

Road safety has remained a world issue, and driver fatigue has been one of the major road accident causative factors when engaged in long-distance journeys and driving late at night. Conventional safety measures, such as lane departure warnings and seatbelt alarms, cannot address the problem of drowsiness before accidents occur. The proposed paper introduces an IoT-based real-time driver drowsiness detection system utilising computer vision technology, integrated with the Internet of Things (IoT). It features a low-cost integrated camera that tracks the driver's eye movements and estimates the Eye Aspect Ratio (EAR) to assess their level of alertness. When this eye closure happens over an extended period, a local alarm is sent immediately to alert the driver, and a warning is also transmitted to remote monitoring equipment or cloud platforms through an IoT module. Experimental trials conducted on experimental subjects demonstrate that the system can reliably and early detect drowsiness with low false alarm rates and low latency of communication. The provided model suggests that road safety can be enhanced by preventing fatigue-related accidents, which is a cost-effective measure; both individual drivers and the organisations of commercial fleets can be improved. The models can be enhanced by the addition of future enhancements through machine learning to categorise better and utilise multimodal sensors, which will make the models more reliable.