Estimating energy consumption of residential buildings at scale with drive-by image capture. (15th April 2023)

• Record Type: Journal Article
• Authors: Ward, W.O.C.; Li, X.; Sun, Y.; Dai, M.; Arbabi, H.; Tingley, D. Densley; Mayfield, M.
• Is Part Of: Building and environment. Volume 234(2023)
• Publication Date: 2023-04-15
• Subjects: Building energy modelling -- Residential buildings -- Mobile sensing -- Data-driven methods -- Retrofit -- Artificial Intelligence; Buildings -- Environmental engineering -- Periodicals; Building -- Research -- Periodicals; Constructions -- Technique de l'environnement -- Périodiques; Electronic journals; 696
• DOI: 10.1016/j.buildenv.2023.110188
• Languages: English
• ISSNs: 0360-1323

Abstract

Abstract: Data-driven approaches to addressing climate change are increasingly becoming a necessary solution to deal with the scope and scale of interventions required to reach net zero. In the UK, housing contributes to over 30% of the national energy consumption, and a massive rollout of retrofit is needed to meet government targets for net zero by 2050. This paper introduces a modular framework for quantifying building features using drive-by image capture and utilising them to estimate energy consumption. The framework is demonstrated on a case study of houses in a UK neighbourhood, showing that it can perform comparatively with gold standard datasets. The paper reflects on the modularity of the proposed framework, potential extensions and applications, and highlights the need for robust data collection in the pursuit of efficient, large-scale interventions. Highlights: Proposes a modular framework to estimate energy consumption from mobile sensing data. Uses computer vision techniques to detect and measure features of buildings. Combines features with other data sources to produce energy models. Demonstrates efficacy of framework with results comparable to EPC estimates. Reflects on the limitations and potential extensions of the proposed framework.