PARAMTR - Enhanced generative design tools for prefabricating large-scale residential developments

Designers are encountering greater issues with residential projects, which are increasing in complexity, scale, and performance requirements. Despite significant advancements in technology and the AEC industry, large-scale residential developments are still designed and built at scale as if they were singular projects. Variable and increased construction time, cost, and material waste at scale are all issues with existing design and construction methodologies for construction at scale. Prefabrication and generative design tools have the potential to significantly reduce these issues.

This paper investigates how collaborative, human-generative design tools can optimise building performance and make prefabricated housing at scale feasible, whilst still encouraging design variance. In this context, collaborative human-generative tools refer to a partially algorithmic design tool that facilitates an open-box approach to design. Using a mixture of research-based design and design-based research, a new tool (PARAMTR) was created to improve feasibility whilst reducing time, complexity, and cost of designing and building residential projects using prefabrication at scale.

The research demonstrates eight unique designs produced using the new human-generative tool. Despite their individuality, these designs have 8-10 times fewer unique components when compared to existing residential projects. Designs produced using PARAMTR could reduce construction/design time by up to 50%, reduce construction costs by up to 26% and share no design commonality, enabling unique designs across an entire development. This research paper could therefore fundamentally change how the AEC industry builds at scale, using algorithms and human-generative design tools.