Indoor environment control method for improving well-being using human thermal stress estimated by Yuragi learning

“Workstyle Reform” is promoted in Japan to improve the work environment surrounding workers. However, it is pointed out that the reduction of working hours by improving marginal productivity may increase workers’ stress on a per-unit hour basis. The realization of a system that quantifies stress and improves it accordingly will lead to our well-being, and in this paper, we propose methods that estimate the stress state of each individual and control the indoor environment in real time based on the estimated stress state to maintain or improve a well-being space. Estimation of the individual stress state is based on “Yuragi learning,” which is a method of making decisions from information including noise, based on a model of the cognitive process of a human brain. We implemented Yuragi learning to realize real-time stress estimation based on data streaming from biometric sensor devices. We also implemented the method to control devices that act in an indoor space on the basis of the estimation. In our experiment, we prepared two room conditions with different temperatures and humidity levels and confirmed that our method estimated the stress state of a subject and controlled the room with an actuator that directs the air flow to the subject in a real-time manner.