A comprehensive framework for AI-driven chair design integrating finite element analysis, ergonomics, and manufacturing constraints

This study, developed at Central Saint Martins, created an integrated AI-driven framework for chair design that combines generative modeling, finite element analysis (FEA), ergonomic evaluation, and manufacturing constraints to ensure innovation, structural safety, ergonomic comfort, and production feasibility at the early design stage. A multimodal database of chair images, manufacturing process parameters, and ergonomic standards was first constructed. A Generative Adversarial Network (GAN) then generated 1,000 chair designs, followed by automated FEA simulations and ergonomic screening to eliminate unsafe or uncomfortable models. Reinforcement learning was applied to optimize geometry, material distributions, and structural performance iteratively. Finally, a questionnaire survey in Beijing’s Fengtai District evaluated aesthetics, functionality, user experience, and market acceptance. After several rounds of optimization, only four designs met all structural, ergonomic, and manufacturability criteria. Survey results showed consistently high ratings for innovation (5.1–5.6) and aesthetic–functional balance (4.4–4.8), demonstrating the framework’s effectiveness in enhancing design feasibility, safety, and creativity.