Researchers at the University of Cambridge have developed a new AI model with the potential to help local councils identify ‘Hard-to-decarbonize’ (HtD) houses which are responsible for over a quarter of all direct housing emissions.

A new ‘deep learning’ model trained by researchers from Cambridge University’s Department of Architecture promises to make it far easier, faster and cheaper to identify these high priority problem properties and develop strategies to improve their green credentials. Houses can be ‘hard to decarbonize’ for various reasons including their age, structure, location, social-economic barriers and availability of data.  

The model developed by a team at Cambridge’s Sustainable Design Group, can classify HtD houses with 90% precision.

The researchers trained their AI model using open-source data from Cambridge, including Energy Performance Certificates, street view images, aerial view images and land surface. In total, their model identified 700 HtD houses in Cambridge. The researchers had previously worked with Cambridge City Council to assess council houses for heat loss and are set to discuss their latest findings with local representatives and local authority officers.  

The team has now used their model to develop a map of energy saving potential, and also received funding from the European Space Agency and UK Space Agency to advance this work.