An Automatic Detection of Drowsy or Distracted Driver Based on Brain Computer Interface

Driving requires sustained concentration, yet many road accidents are attributed to factors such as driver tiredness, drowsiness, inattention, and distraction. Electroencephalographic (EEG) recordings, which capture electrical brain activity via head-attached electrodes, have proven effective for drowsiness identification. This paper aims to utilize a brain-computer in-terface (BCI) that gathers EEG brain signals, classify driver's drowsiness using machine learning algorithms on mentioned EEG signals, and developing a mobile interface to alert drowsy classified drivers. The recorded brain signals from the BCI undergo several phases, including feature extraction and classifi-cation using various learning-based algorithms such as Logistic Regression, K-Nearest Neighbor, Naive Bayes, Support Vector Machines, Random Forest, Decision Tree, and Neural Networks. The accuracy of drowsiness detection using these algorithms is evaluated, with the Decision Tree achieving the highest accuracy of 78.08%. Detected instances of drowsiness are translated into visual and audible feedback on a mobile device. This research contributes to advancing drowsiness detection methods and intervention techniques, potentially enhancing road safety and mitigating the risks associated with driver fatigue.