Construction-scale three-dimensional (3D) printing (C3DP) is reshaping building by enabling automated, low-cost and environmentally friendly construction. Yet it struggles with material variability, process control and limited real-time adaptability. This paper explores how machine learning (ML) can address these barriers. Through supervised, unsupervised, reinforcement and deep learning methods, ML strengthens quality control, robotic path planning, predictive maintenance and adaptive optimisation. Continuous sensing and feedback improve structural performance and reduce waste. Case studies from ICON, Apis Cor and WASP demonstrate practical gains from combining ML with large-scale 3D printing – such as better print reliability, smarter robotics and more sustainable materials. Critical enablers are also discussed in this paper, including sensor integration, edge artificial intelligence (AI) for low-latency decision making and ongoing regulatory challenges. Finally, emerging opportunities are identified in autonomous construction and generative AI–driven design. ML-enabled C3DP offers a promising route toward smarter, more sustainable and scalable building systems. This paper provides both a literature-based review and a conceptual framework outlining how these technologies can shape future adaptive construction.
Barbhuiya et al. (Tue,) studied this question.