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

Buildings account for an estimated 35% of the UK’s total greenhouse gas emissions. Assigning carbon data to building designs can aid contractors in sustainable material selection. Authoring inconsistencies in building data mean addition of environmental data is difficult and expensive. Machine learning enables classification of materials at a scale manual approaches cannot match. A machine learning approach is documented for classifying building products that enables automatic augmentation of environmental data from carbon databases. Our findings provide foundational research on automating data authoring, that can reduce costs and simplify processes associated with adding environmental assessment data to building designs.