Non-visual Effects Driven Fatigue Level Recognition Method for Enclosed Space Workers

With the rapid development of underground space and manned tourist deep diver careers, it is of great research value to enhance the comfort and task performance of workers by improving the design of enclosed spaces. It has been well-acknowledged that vision is the primary factor that affects human perception and fatigue. On the one hand, lighting environment of an enclosed space can directly influence the visual perception of human through the human body’s third type of photoreceptor cells, intrinsically photosensitive retinal ganglion cells (ipRGCs), the human body’s pupil size, melatonin, body temperature, heart rate, and other non-visual physiological metrics. On the other hand, inappropriate lighting environments can reduce vigilance and even cause severe human errors. In this study, we designed four different lighting conditions to investigate the non-visual effects of the workers in completing the Psychomotor Vigilance Task (PVT). Task performances (i.e., Reaction Time - RT) and physiological indicators (i.e., electrocardiogram - ECG) are used to evaluate the effectiveness. In general, the results show that with an increased illuminance level, the workers are becoming more vigilant. We have collected the RT with different fatigue levels classified accordingly, and then analyzed the correlation between the RT and ECG indexes. Furthermore, we constructed the fatigue recognition model of workers based on ECG indexes by using the Support Vector Machine (SVM) algorithm and verified it. This study aims to optimizes the lighting environment and improve task perform efficiency by automatically identifying the fatigue level through monitoring the ECG indicators of the workers.