Feasibility of a Mobile Electroencephalogram (EEG) Sensor-Based Stress Type Classification for Construction Workers: Fid-Bau Portal

Feasibility of a Mobile Electroencephalogram (EEG) Sensor-Based Stress Type Classification for Construction Workers

Lee, Gaang / Lee, SangHyun

Wearable biosensors have been applied for continuously monitoring construction workers’ stress during fieldwork in minimally invasive ways. However, little effort has been made to differentiate between stress types [e.g., eustress (positive stress) and distress (negative stress)] even though the impact of stress on workers differs significantly depending on the stress type. For example, while eustress can improve overall worker performance, distress is a root cause of detrimental “stress-induced” consequences (e.g., lack of task-focus and depression). Since emotional and physiological responses to different stress types manifest in varied brain activity, the authors propose an electroencephalogram (EEG)-based stress type classification method. The proposed method was tested using EEG signals labeled as three types of stress: low-stress, eustress, and distress. Test accuracy was 0.842, indicating that the proposed method is promising toward understanding different stress types. This finding can contribute to advancing stress management in construction fields by enabling selective stress interventions that alleviate distress while keeping the benefits from eustress.