Real-time Alarm Monitoring System for Detecting Driver Fatigue in Wireless Areas

Fu, Rongrong; Wang, Shutao; Wang, Shiwei · 2017 · Crossref

DOI: 10.7307/ptt.v29i2.2058

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the critical safety issue of driver fatigue, a primary contributor to highway traffic accidents, by developing a real-time alarm monitoring system. The research is motivated by the need for an objective, non-intrusive, and reliable method to detect fatigue onset, particularly among young drivers who are prone to underestimating risks and overestimating their skills. While previous studies utilized multiple EEG channels or combined visual and biological features, this work aims to simplify the system by using fewer electrodes and wireless communication to enhance driver comfort and practical applicability. The methodology involved eleven healthy participants aged 22 to 31 performing a two-hour monotonous driving simulation. Electroencephalogram (EEG) signals were recorded wirelessly from two occipital electrodes (O1 and O2) using a portable biofeedback system. The signals were processed to extract seven frequency bands: gamma, high beta, beta, sigma, alpha, theta, and delta. To identify the optimal fatigue indicator, the researchers employed Relative Operating Characteristic (ROC) curves and Grey Relational Analysis (GRA). Subjective fatigue levels were assessed using the Karinska Sleepiness Scale (KSS) to classify states as alert, mild fatigue, or fatigue. The system design utilized the first ten minutes of driving to establish a baseline threshold for alertness. The results indicated that theta and delta waves performed best in distinguishing between alert and fatigue states, with Area Under the Curve (AUC) values of 0.935 and 0.938, respectively. Although the delta wave had a slightly higher AUC, the theta wave’s ROC curve was closer to the ideal upper-left corner, suggesting better classification performance. Grey Relational Analysis confirmed this, yielding a higher relational grade for theta (0.9402) compared to delta (0.7544). Consequently, the theta wave was selected as the optimal fatigue feature. The study demonstrated that theta wave amplitude decreases significantly as drivers become drowsy. The developed system successfully detected this decrease and triggered an automatic audio alarm when the signal dropped below the established threshold. The significance of this work lies in the validation of a simplified, wireless EEG-based system for real-time fatigue detection. By relying on only two occipital electrodes and a single frequency band (theta), the system reduces hardware complexity and user discomfort compared to multi-channel setups. This approach offers a practical solution for preventing fatigue-related accidents by providing immediate warnings, thereby allowing drivers to take corrective measures before performance degrades critically. The findings support the use of theta wave analysis as a robust, objective metric for monitoring driver alertness in real-world applications.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• drowsiness detection algorithms
• dms validation
• truck driver fatigue
• drowsiness
• vigilance

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: physiological data
• Methodological Resource: validation psychometrics, tool software