Answers are grounded in this paper and cite their sources.
Explores how machine learning enhances building simulations, suggesting implications for sustainable designs.
Surrogate modelling involves training a machine learning (ML) regression model to emulate computationally expensive physics-based simulations. These can be evaluated very rapidly, essentially instantly, opening up use cases that are impractical with traditional simulations including interactive design and large-scale analyses. ML techniques can provide incredibly high accuracy, though care must be taken that the surrogate is fit for purpose. This paper explores recent developments and new directions in surrogate modelling. This is just the beginning of the Artifical Intelligence (AI) revolution, with generative AI tools poised to overhaul many fields; BPS will not be immune to this. These tools can bring incredible benefits but also introduce serious issues. The latter sections of this paper explore these issues. Moving forward, academic and practitioner communities must work together to provide guidance, leveraging the power of these tools to meet the demand for low carbon designs and retrofits at scale.
No takes yet. Share an insight, caveat, or question.
Ralph Evins (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: