Can Machine Learning Automate Carbon Classification of Materials Within a BIM?

Abstract: Building environment accounts for 35% of the UK’s total greenhouse gas emissions, Contractors need carbon data to choose eco-friendly products. Existing tools provide data (Circular Ecology, 2020), but assigning it is hard and expensive. This paper fills a gap in automatically classifying assets by carbon impact. We use Machine Learning (ML) to predict and classify faulty or incomplete BIM data. Our ML method simplifies and speeds up carbon classification. This paves the way for more ML research on building data automation. It also helps practitioners manage data better in green construction, changing how asset classification is done for environmental assessment.