A comparative study of predicting individual thermal sensation and satisfaction using wrist-worn temperature sensor, thermal camera and ambient temperature sensor

Abstract

Recent advancements in Internet of Things and Machine Learning have opened the possibility of deploying sensors at a large scale to monitor the environment and to model and predict thermal comfort at an individual level. There has been a growing interest to use physiological information obtained from wearable devices or thermal imaging to improve individual thermal comfort prediction. In this study, we compared the accuracies of using environmental sensing with an air temperature sensor, physiological sensing with a wrist-worn device to monitor wrist skin temperature or thermal camera to monitor facial skin temperatures for predicting individual thermal sensation and satisfaction. The experiment was conducted in a controlled environment without any radiant heat sources or local comfort devices; solely the air temperature was changed. For the conditions studied, our results indicate that using data from an environmental sensor for predicting thermal comfort results in a higher accuracy compared to using physiological sensors (either wearable device or thermal camera) alone. Combining data from both environmental and physiological sensors leads to about 3%–4% higher accuracy than using environmental sensors only. Slight improvement in accuracy from the physiological sensors might not be sufficient to justify the privacy concerns and additional costs of using physiological sensors at a large scale for predicting thermal comfort in environments without radiant heat sources or local comfort devices. Future studies under different environmental conditions with a larger population are needed to better understand the tradeoffs between different sensing methods for predicting thermal comfort at an individual level.