Challenges and opportunities in quantitative aerial thermography of building envelopes

Abstract

The use of small unmanned aerial vehicles (UAVs) equipped with an infrared (IR) camera has created opportunities for qualitative and quantitative thermal assessments of building envelopes. However, it is currently unclear how successful this method is at quantifying heat losses through the building envelope. The inconsistency in findings indicates that sources of error that affect the accuracy of surface temperature measurements are not well understood by thermographers and that recommendations to minimize the errors in aerial thermography measurements have yet to be determined. In this study, laboratory and field experiments were conducted to investigate the scientific reasons for discrepancies between aerial (i.e. dynamic) and conventional stationary infrared thermography (IRT) observations. A set of practical approaches were presented to minimize the sources of error in UAV-IRT measurement. The results of field experiments showed a non-linear and dramatic variation of measured surface temperature during flight, where the deviations of results between dynamic and stationary thermography were beyond the manufacturer-reported accuracy of ±5 °C. Additionally, the results indicated that induced convection from drone propellers affects microbolometer stabilization with surrounding environmental conditions, which significantly influences the IR camera measurement accuracy. Finally, subsequent investigations utilized a shield around the camera to minimize the convective turbulence on the lens and thermal sensors, successfully achieving temperature deviations of less than 1 °C. These findings may inform manufacturers about the limitations of current IR camera technology during aerial surveys and therefore provide opportunities to define a more robust thermal imaging protocol for the quantification of building envelope thermal performance.