Intelligent algorithm-based generative modular design for building façades in off-site construction
Jiawei Dou, Jianing LuoPurpose
This paper develops a generative modular design approach, supported by dynamic programming (DP) and genetic algorithm (GA), for architectural dividing design (ADD) of building façades in off-site construction (OSC). It aims to optimize modular design deliverables by improving adherence to modular design principles, including reducing the number of component types and enhancing standardization and modular coordination.
Design/methodology/approach
A sequential mixed-methods strategy was applied to a real-world OSC project across five phases. A hierarchical formalization translates modular design principles into modular dividing rules and constraints, which are implemented using a DP- and GA-based design approach.
Findings
The approach reduces the number of component types from 15 to 12, increases standardization from 23.08% to 35.71% and achieves 100% modular coordination across all façade components.
Research limitations/implications
The study focuses on rectilinear façades and requires manual preprocessing of modular dividing constraints. Applicability to irregular geometries remains to be validated across various façade typologies.
Practical implications
The proposed approach automates the optimization of modular design deliverables, improves adherence to modular design principles and transforms façade ADD into an algorithmic, data-driven process.
Social implications
By bridging the knowledge gap between architectural design and industrialized construction, this research fosters a more efficient and sustainable OSC.
Originality/value
This study develops a computational approach by integrating DP and GA to implement modular design principles. It contributes by formalizing the façade ADD process and improving the consistency and alignment of deliverables between architectural design and OSC.