Electroencephalogram-Based Approaches for Driver Drowsiness Detection and Management: A Review

Area of Science:

• Neuroscience
• Automotive Engineering
• Human-Computer Interaction

Background:

• Driver drowsiness poses significant safety risks in traditional driving.
• Drowsiness is also a key challenge for the adoption of self-driving cars, manifesting as carsickness.
• Early detection of drowsiness is crucial for driver safety and vehicle system acceptance.

Purpose of the Study:

• To review and organize electroencephalogram (EEG)-based algorithms for driver drowsiness detection (DDD).
• To identify key questions and develop a superior EEG-based DDD system.
• To establish a taxonomy for EEG-based DDD approaches.

Main Methods:

• Systematic review of EEG-based driver drowsiness detection algorithms.
• Organization of algorithms into a tree structure taxonomy: 'detection only (open-loop)' and 'management (closed-loop)'.
• Addressing seven key questions to guide the development of improved DDD systems.

Main Results:

• A comprehensive review and classification of existing EEG-based DDD algorithms.
• Identification of pathways to enhance DDD systems for early detection, reliability, and practical utility.
• A foundational framework for developing advanced EEG-based drowsiness detection.

Conclusions:

• EEG-based DDD is essential for enhancing driving safety in both conventional and autonomous vehicles.
• A structured approach to reviewing DDD algorithms facilitates the development of more effective systems.
• The underlying brain network for drowsiness is consistent, regardless of the cause (e.g., fatigue vs. carsickness).