We are pleased and humbled to share that iLab@HKU led by Prof. Wilson Lu has won a Hong Kong Research Grant Council (RGC) Collaborative Research Fund (CRF) in its 2022/23 exercise. Generative design, design for manufacturing and assemble (DfMA) and Design for Excellence (DfX) are all popular topics. This is perhaps the world’s first ‘generative design for excellence’ research project focusing on high-rise modular buildings! We adopt a combination of generative design and DfMA and DfX. We adopt many innovative methodological approaches such as Graph Learning (or machine learning with graphs), Generative Adversarial Network, heuristic optimization, and so on.
Funding Scheme: Research Grant Council (RGC) Collaborative Research Fund (CRF)
Funding Year: 2022/23
Project Title: Generative DfX in high-rise modular building: An expert-augmented cascade graph learning and optimisation approach
Project Coordinator (PC): Prof. Wilson Lu (REC)
Co-Principal-Investigators (Co-PI): Prof. Chris Webster (FoA), Dr. Kristof Crolla (DoA), Dr. Frank Xue (REC), Dr. Junjie Chen (REC), Prof. Geoffrey Shen (HKPolyU), Prof. Jianguo Dai (HKPolyU), Dr Jianxiang Zhu (CUHKU), Dr Jinying Xu (Cambridge U)
Co-Is: Dr Shang Gao (Melbourne U); Dr Tan Tan (TU Delft), Mr. Jinfeng Lou (HKU), Mr. Vikrom Laovisuttchichai (HKU)
Project start date: 1 April 2023
Project Duration: 36 months
Funded amount from RGC (without including on-cost) (HKD): 5.31 million
Project summary/abstract:
High-rise modular building (HRMB) is highly advocated to address the housing crisis in high-density cities around the world. In line with this advocate is the principle of design for excellence (DfX) that is vigorously explored to unlock the full potential of modular building. DfX encompasses ‘excellence’ criteria such as functionality, ease of manufacture and assembly, logistics, buildability, sustainability, and cost, which require multidisciplinary domain knowledge that is beyond the capability of any single designer. Computer-aided generative design seems to provide a promising strategy to handle the multifaceted knowledge requirement.
This project aims to develop a computer-aided, designer-oriented generative DfX methodology for HRMB. We employ graph learning in a top-down manner to generate a rough building (floor plan), then flat design, and finally detailed module design. It then leverages advanced heuristic algorithms to optimise the generated design options from the bottom up, i.e., from module to flat and ultimately to building (floor plan). The process will be augmented by design knowledge and includes human experts in the loop.
We will pilot the research in Hong Kong’s HRMBs, a rich context for considering DfX in relation to factors including user groups, available construction technologies, manufacturing capacity, logistics, and site conditions. The research will deepen our understanding of HRMB design by considering a wide range of excellence criteria, and may open up a new design paradigm through which humans and machines collaborate to deliver design value.
Highlights of the research project:
- Extending DfMA to DfX to consider more challenging and mysterious eXcellence factors.
- Deploying graph learning (or machine learning with graphs) in generative design, as many of our design and construction knowledge is represented as graphs;
- Using Generative Adversarial Network (GAN) instead of prevailing algorithms (e.g., auto-encoder, Shaper_GA) to more effectively generate design options;
- Respecting expert knowledge, be it explicit or tacit, e.g., to solve the local and global optimum problem and the “curse of high dimensionality” from DfMA to DfX;
- Organising the computational challenges in a cascade fashion – it is a typical “divide and conquer” arrangement.