UAV-Derived Data Processing Framework for Enhancing Input Reliability to Create Accurate Community Building Energy Models

Reducing energy consumption of the urban building stock is a driving concern for meeting the global net-zero targets. It is commonly observed that the urban building energy modeling approaches that try to find suitable building retrofit strategies citywide have considerable uncertainty in the outputs generated, and most of the results are not validated because of the complexity of collecting a large data set of energy bills. In addition, the existing energy frameworks do not allow flexibility in studying the shading effect of trees on building energy consumption citywide and offer a limited application for investigating mixed-mode air-conditioned buildings typical in developing countries. To address these limitations, this study proposes a decentralized approach of clustering urban areas into building communities and developing an unmanned aerial vehicle (UAV) data-based building energy modeling schema [a community building energy model (CBEM)], particularly encompassing the building characteristics of tropical developing countries. The framework is implemented in a case study of a residential community in India, and the results indicate that an accuracy of greater than 80% was achieved for both communitywise and buildingwise energy assessments. The shading effect of trees was observed to result in an annual energy reduction of about 2.4%, and passive energy retrofit strategies, prescribed by India's energy conservation guidelines, were observed to bring down energy consumption to 14%.