Image © Shutterstock
InnovateUK has awarded £240k to Kestrix – and partners Peabody Trust and United Living Property Services Ltd – to use AI to measure home heat loss at scale.
The project, lasting two years, will support the development, demonstration, and commercialisation of Rapid Thermal Performance Assessment Algorithms (RaThPAs) – algorithms that interpret thermal images of buildings for retrofit insights.
AI predicts U-values (heat transfer) of varying materials, identifies faulty insulation, and estimates overall space heating demand intensity (the rate at which the whole building leaks heat). When applied at portfolio-scale, these allow a housing provider, city, or energy utility to stack-rank subsets of buildings by level of thermal efficiency with more confidence than ever before, helping prioritise work and plan for net zero over the long term.
Lucy Lyons, CEO of Kestrix, said:
This project brings together the machine learning and computer vision expertise of Kestrix, the extensive experience of Peabody Trust as one of the UK’s largest social housing providers, and the retrofit delivery leadership of United Living. Together, we aim to demonstrate a game-changing technology that will transform the way the industry approaches retrofit, enabling smarter, smoother retrofit planning – and ultimately the more efficient allocation of scarce retrofit resources.
Richard Ellis, Director of Sustainability for Peabody, said:
It’s clear that improving the energy efficiency of homes is crucial to reducing carbon emissions and lowering costs for residents. We also know that warmer and more cost-efficient homes vastly improve residents’ lives. But we don’t yet have a quick, completely accurate and efficient way of planning and prioritising our retrofit works. Our hope is that this project and others like it will unlock the untapped potential of advancing technology, ultimately helping us to retrofit homes faster and more effectively than ever before.
