A Context-Aware EEG Headset System for Early Detection of Driver Drowsiness

Area of Science:

• Biomedical Engineering
• Neuroscience
• Human-Computer Interaction

Background:

• Driver drowsiness is a significant global cause of traffic fatalities.
• Electroencephalographic (EEG) signals offer direct insight into brain activity related to drowsiness.
• Existing Brain-Machine-Interface (BMI) systems face challenges in detecting early-stage driver drowsiness.

Purpose of the Study:

• To propose a context-aware BMI system for early detection of driver drowsiness.
• To enhance EEG data with head-movement intensity for improved drowsiness detection.
• To develop a low-power, mobile-based system for real-time analysis.

Main Methods:

• Developed a context-aware BMI system integrating EEG and head-movement data.
• Implemented on-chip feature extraction and low-energy Bluetooth for power efficiency.
• Created a JAVA-based mobile application for online drowsiness analysis.
• Evaluated the system using 266 datasets from a driving simulation experiment.

Main Results:

• EEG data alone achieved 82.71% accuracy in detecting alert vs. slightly drowsy states.
• The hybrid system combining EEG and head-movement data reached 96.24% accuracy.
• The system demonstrated robust performance in early drowsiness detection.

Conclusions:

• Combining EEG signals with head-movement data provides a robust method for early driver drowsiness detection.
• The developed context-aware BMI system offers a practical solution for enhancing road safety.
• The system's low-power design and mobile implementation facilitate real-world application.