Multi-Sensor Vigilance Detection in Train Drivers – a Simulator Study

Kurzfassung

Our study investigated the reliability of multi-sensor systems for detecting low vigilance in train drivers operating under Grade of Automation Level 2 (GOA2) conditions. A high-fidelity train simulator study with 14 professional train drivers examined physiological indicators of mental fatigue using electrocardiography (ECG), electroencephalography (EEG), breathing rate, eye tracking, skin conductance, and self-reported sleepiness. While all physiological changes aligned with expected mental fatigue effects, heart rate variability showed the only statistically significant difference. However, blink duration, missed dead man’s handle reminders, and self-reported sleepiness showed large effect sizes, suggesting their potential for real-time driver monitoring. These findings will be compared with a real-world study using the same sensor system, informing the development of wearable monitoring solutions to enhance safety in automated rail systems.