EEG-based drowsiness alert system with IoT integration for safe driving

Arbain, Muhammad Azamuddin (2024) EEG-based drowsiness alert system with IoT integration for safe driving. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Text (Full Text) EEG-based drowsiness alert system with IoT integration for safe driving.pdf - Submitted Version Download (1MB)

Abstract

Driver drowsiness and fatigue are major factors in road accidents because they reduce focus, slow response times, and delay decision-making. In severe circumstances, drivers may suffer microsleeps, small moments of unconsciousness that can result in serious accidents if they occur while driving. To solve these problems, this research proposes an EEG-based drowsiness detection system that includes IoT integration for real-time monitoring and alert. Electroencephalogram (EEG) readings, which assess neural activity are used as physiological indicators to identify early signs of drowsiness. The system collects brainwave data using the MindLink brain sensor, a wireless EEG headband. The data is delivered to an ESP32 microcontroller over Bluetooth and evaluated to detect signs of tiredness. Real-time data is transmitted to the IoT platform ThingSpeak for storage, display and remote monitoring. This allows refined functionality like data analysis and application integration. The detection technique is based on assessing the power spectrum concentration of EEG signals, with particular focus on theta wave activity. The system detects drowsiness by observing an ongoing pattern of theta waves, as well as a drop in concentration and an increase in meditation. When the system detects drowsiness, it sends a multi-modal alert that consists an auditory alarm, a visual warning on an LCD display, and visual indicators to assure the driver's safety. This initiative aims to prevent drowsy driving accidents and increase road safety through the implementation of EEG technology and internet of things (IoT) connections.

Actions (login required)

View Item