Generative Design with Quality Function Deployment for Architectural Layout Design Optimization

Abstract

Architectural design problems are inherently complex due to their non-linear nature and the interdependence of many variables. Traditional design methods struggle with the speed and breadth of exploration, often failing to adapt quickly to changing client needs and project timelines. This research addresses these challenges by developing a framework that integrates Quality Function Deployment with generative design to automate and optimize the architectural layout process. The methodology is segmented into three stages: pre-generative design, generative design, and post-generative design. In the pre-generative design stage, client needs are translated into specific design requirements using Quality Function Deployment. In the generative design stage, generation algorithms are developed, namely bottom-up and top-down method, that are applicable to different problem settings. The generated design solutions are evaluated and optimized through NSGA-II genetic algorithms. The final stage, post-generative design, focuses on refining the chosen designs and converting them into functional Building Information Modelling models. A case study of designing a single detached house with 3 bedrooms and 2.5 bathrooms is conducted following this methodology. The effectiveness of the integrated Quality Function Deployment and generative design approach is exemplified through the case study. The use of advanced computational tools enables the exploration of numerous design alternatives, efficiently narrowing down to the best solutions that meet predefined criteria.