Advancing modular design generation: A review of modularization workflows and computational design approaches

Abstract

<p>Modular construction is an innovative building method that offers distinct advantages over on-site construction, including improved efficiency, safety, quality, and sustainability. The fundamental differences between modular and on-site construction demand corresponding shifts in design methods, and computational design approaches have increasingly been adopted to support and automate modular design generation. However, there is no systematic synthesis that specifically examines how computational design approaches are integrated into the design tasks and workflows unique to modular construction, and this study aims to fill the gap. Through a systematic literature review of 58 papers, this study identifies two predominant modularization workflows (i.e., the module-composition method and the decomposition-to-modules method), and three computational design approaches (i.e., boundary-guided, exploration-evolved, and hybrid approach). The study further develops the concept of “modularization lock-in point”, and analyzes the impact of its positioning on the production strategy, dominant cognitive mode, and the role of computation design tools. Based on the analysis, a three-step decision-making framework is proposed, and three future research directions are discussed, including advancing the integration of multi-layered building systems, incorporating general-purpose generative artificial intelligence (AI) into design workflows, and developing human-AI collaboration for design knowledge co-creation. The review provides a theoretical lens for understanding automated modular design, and offers a guide for practitioners to make more informed decisions when selecting a modularization workflow and its associated computational tools.</p>