Automated Real-World Sustainability Data Generation from Images of Buildings

TL;DR

This paper presents a novel method using large language models and images to estimate building features relevant for sustainability, enabling data-driven improvements without prior data access.

Contribution

It introduces an image-to-data approach leveraging language models for building feature estimation, outperforming human accuracy and aiding sustainability efforts.

Findings

Achieved higher accuracy than humans in estimating building features.

Successfully generated tailored sustainability improvement recommendations.

Demonstrated scalability of the image-to-data method.

Abstract

When data on building features is unavailable, the task of determining how to improve that building in terms of carbon emissions becomes infeasible. We show that from only a set of images, a Large Language Model with appropriate prompt engineering and domain knowledge can successfully estimate a range of building features relevant for sustainability calculations. We compare our novel image-to-data method with a ground truth comprising real building data for 47 apartments and achieve accuracy better than a human performing the same task. We also demonstrate that the method can generate tailored recommendations to the owner on how best to improve their properties and discuss methods to scale the approach.