Computer Vision-based Method to Energy Saving Retrofit: A Study of Improving Energy Efficiency in Existing Construction

Computer Vision-based Method to Energy Saving Retrofit: A Study of Improving Energy Efficiency in Existing Construction

연구 분야: Artificial Intelligence

학회: 2023 IEEE Conference on Technologies for Sustainability (SusTech)

This study presents a computer vision-based energysaving modeling method for identifying and addressing energy inefficiencies in existing buildings. The proposed method comprises three main phases: Inspection and Data Collection, Applying Techniques, and Model Building. The feasibility and effectiveness of the proposed method were evaluated using an energy performance simulation tool, Energy 3D. Various scenarios were simulated to assess the impact of different retrofit options on the building's energy efficiency, and the results were compared. The proposed method is an efficient and fast solution to the energy performance problems of existing buildings that are regularly challenging and time-consuming to simulate. The proposed method also has the potential to be used by policymakers for creating energy-saving policies and for homeowners to make informed and cost-efficient decisions through lifecycle cost analysis. This study demonstrates the potential of computer vision technology to improve the energy efficiency of existing buildings, reduce energy consumption, and lower costs, ultimately contributing to a more sustainable future.