Application of adaptive AI in the biophilic dwelling design process: a review

The built environment significantly impacts human health and wellbeing, but urbanisation has led to disconnection from nature, causing adverse psychological and physiological effects. Biophilic design aims to bridge this gap, but its implementation is limited by discrepancies between design intent and realisation, context dependency, and challenges in quantifying outcomes. Adaptive AI offers a transformative solution, enabling responsive, personalised, and data-driven biophilic environments. This review synthesises research on Adaptive AI applications in biophilic dwelling design, focusing on methods, capabilities, and empirical evidence. AI techniques, such as generative models and neural networks, accelerate design exploration (up to 90–95%), predict environmental performance, and optimise balanced solutions. Preliminary evidence reports positive outcomes, including energy savings (up to 30%) and enhanced thermal comfort prediction accuracy (>96%). However, research remains fragmented, with limited long-term evidence and real-world implementation. Gaps include a narrow focus on visual elements, inconsistent environmental reasoning, scarce datasets, and underdeveloped ethical considerations. Integrating Adaptive AI with biophilic design offers a promising frontier for creating healthier, sustainable, and responsive dwellings. Realising its potential requires interdisciplinary collaboration, empirical validation, standardised datasets, and an ethical approach prioritising occupant agency and equity.