Introduction
Urban populations in the UK continue to grow, increasing pressure on housing supply while intensifying the need to reduce carbon emissions and improve the environmental performance of the built environment.
Between 2011 and 2019, the UK's urban population increased by 6.2% [1]. Addressing these challenges requires architectural design approaches capable of balancing multiple, often competing, performance objectives, including operational energy, embodied carbon, occupant health, daylight availability, thermal comfort, natural ventilation, etc.
Recent advances in artificial intelligence provide new opportunities to support early-stage architectural design through data-driven optimisation and decision-making [2][3]. This research proposes an AI-assisted design framework that integrates generative design algorithms, multi-objective optimisation, and machine learning-based performance prediction with environmental simulation to develop and evaluate sustainable housing prototypes.
The novelty of this research lies in the integration of AI-driven generative design with multi-criteria performance optimisation to produce housing solutions that balance environmental sustainability, occupant well-being, and architectural quality. While previous studies have explored generative design primarily for energy or structural optimisation [4], relatively limited research has developed an integrated framework that simultaneously considers carbon reduction, indoor environmental quality, spatial performance, and health-related design outcomes [5] [6]. The proposed framework will combine iterative design generation, environmental simulation, and AI-guided feedback to support evidence-based decision-making during the early design stages.
Research Aim and Questions
The research will address the following questions:
How can AI-based generative design methods improve the environmental and spatial performance of residential housing during the early design stage?
To what extent can AI-assisted design workflows support healthier and more resilient housing compared with conventional design approaches?
Methodology
The research methodology will combine computational design, environmental performance simulation, and multi-objective optimisation within an iterative workflow. Candidate housing layouts will be automatically generated using parametric modelling and AI-based optimisation techniques, evaluated through environmental simulation, and refined through performance feedback. Comparative analysis against conventionally designed housing schemes will be used to assess improvements in environmental performance, occupant comfort, and design quality.
Outcomes
The outcomes of this research will contribute both methodological and practical advances to AI-assisted architectural design. The project will not only establish an integrated workflow that combines generative AI, optimisation algorithms, and building performance simulation, but also demonstrate how AI can support architects in designing housing that is more sustainable, healthier, and better adapted to future environmental and societal challenges in the UK.
References:
[1] Government Office for Science (2021) Trend Deck 2021: Urbanisation. GOV.UK. Available at: https://www.gov.uk/government/publications/trend-deck-2021-urbanisation/trend-deck-2021-urbanisation (Accessed: 5 August 2026).
[2] Sadiku, M.N.O. (2025). Generative Design in Architecture. In: Emerging Technologies in Architecture. Springer, Cham. https://doi.org/10.1007/978-3-031-99509-5_4
[3] Chew, Z. X., Wong, J. Y., Tang, Y. H., Yip, C. C., & Maul, T. (2024). Generative Design in the Built Environment. Automation in Construction, 166, 105638.
[4] Peckham, O., Raines, J., Bulsink, E., Goudswaard, M., Gopsill, J., Barton, D., ... & Hicks, B. (2025). Artificial intelligence in generative design: a structured review of trends and opportunities in techniques and applications. Designs, 9(4), 79.
[5] Elnagar, E., Düvier, C., Batra, Z., Christoffersen, J., Mandin, C., Schweiker, M., & Wargocki, P. (2024). Creating a Comprehensive Framework for Design, Construction and Management of Healthy Buildings. Energy and Buildings, 324, 114883.
[6] Che, W., & Zhuang, W. (2026). AI-enabled Pre-design Programming for Low-carbon Buildings: Toward Spatial Form Following Carbon Flow. Energy and Buildings.