Real-Time Vigilance Monitoring via Biosignals: A Scoping Review

This scoping review aims to identify types of computational models that use physiological signals for real-time vigilance monitoring. Vigilance, or the ability to sustain attention over time, is essential in domains such as driving, aviation, and industrial operations, where lapses can cause serious harm. Following PRISMA guidelines, the review searched IEEE, ProQuest, PubMed, and ScienceDirect and identified 16 studies that examined vigilance and real-time monitoring systems. Most studies applied supervised machine learning, mainly for binary classification of vigilance states such as alert versus drowsy, and were concentrated in automotive contexts. Electroencephalography (EEG) was the most commonly used sensor, while heart rate variability (HRV), electrooculography (EOG), photoplethysmography (PPG), and galvanic skin response (GSR) appeared less frequently. Overall, current systems still prioritize threshold-based detection over continuous tracking of vigilance changes. The review highlights the need for multimodal sensing, real-world validation, and adaptive modeling frameworks to support reliable deployment across diverse operational environments.