웨어러블 디바이스를 이용한 사무실 근무자 각성 상태 예측

Abstract

Through physiological signals, detecting the changes of occupants’ physical and psychological aspects is possible, and wearable devices have enabled measurement in daily life. In this study, to see whether the wearable device could be used to interpret the office workers’ alert state, a field experiment was conducted. A wearable device was applied for monitoring the occupant, and productivity responses were collected inside a real office. As a result, when the productivity and alertness decreased, the room temperature was high, and the skin temperature and electrodermal activity were increased. A comparison between the group of alert and drowsy states was made, along with the prediction of alert and drowsy state of the workers using the machine learning algorithms. The results showed the insight of understanding the occupants’ alert state through wearable device measured data. Physiological factors were shown to provide sufficient accuracy when predicting the alert state of the office workers. The results of this study confirmed the possibility of using physiological signals from the wearable device to more accurately identify the alert status of office occupants.