Towards IoT-assisted Non-Intrusive Monitoring of Thermal Variation on Building Envelope

Abstract

Miscellaneous surface damages and air leakages through building components significantly affect energy consumption as well as the thermal comfort of inhabitants in building. Formal thermal inspection of built environment by professionals is expensive and intrusive, often inconclusive and inconvenient, which also requires a lot of parameter tuning. Besides, non-intrusive monitoring with low-cost sensors and thermal images can provide data-driven knowledge of the thermal properties of built surfaces and also helps in accelerating the process of thermal inspection by professionals. We introduce novel quantitative approaches based on non-intrusive temperature and humidity sensor data as well as thermal images which simultaneously learn spatial and temporal thermal characteristics from different places in the inside built environment and provide data-driven thermal variation analysis for identifying potential damage-prone and air-leaking areas in order to reduce energy loss through building components.