Thermovote

Abstract

Thermal comfort has traditionally been measured solely by temperature. While other methods such as Predicted Mean Vote (PMV) are available for measuring thermal comfort, the parameters required for an accurate value are overly complicated to obtain and require a great deal of sensory input. This paper proposes to bypass overly cumbersome or simplistic measures thermal comfort by bringing humans in the loop. By using humans as sensors, we can accurately adjust temperatures to improve occupant comfort. We show that occupants are more comfortable with a system that continually adjusts to thermal preference than a system that attempts to predict user comfort based on environmental factors. In addition, we also show that such a system is able to save 10.1% energy while improving the quality of service.