What a Comfortable World: Ergonomic Principles Guided Apartment Layout Generation

2026-04-09

AI Summary

This work addresses a critical limitation in existing data-driven floorplan generation methods, which often reproduce ergonomic deficiencies present in real-world datasets, thereby compromising habitability. To overcome this, the authors propose a novel approach that integrates architectural ergonomic design principles into a Transformer-based generative model in a differentiable manner. By formulating building-code-informed, differentiable loss functions, the method enables end-to-end optimization of spatial adjacency and distance relationships among rooms. This is the first layout generation framework to explicitly incorporate ergonomic priors during synthesis, achieving significantly improved compliance with human-centered design standards while maintaining high structural validity. Experimental results demonstrate that the proposed method outperforms current baselines in generating ergonomically sound residential layouts.

Abstract

Current data-driven floor plan generation methods often reproduce the ergonomic inefficiencies found in real-world training datasets. To address this, we propose a novel approach that integrates architectural design principles directly into a transformer-based generative process. We formulate differentiable loss functions based on established architectural standards from literature to optimize room adjacency and proximity. By guiding the model with these ergonomic priors during training, our method produces layouts with significantly improved livability metrics. Comparative evaluations show that our approach outperforms baselines in ergonomic compliance while maintaining high structural validity.

Problem

• ergonomic inefficiencies
• floor plan generation
• apartment layout
• livability metrics
• architectural design

Innovation

• ergonomic design
• layout generation
• differentiable loss
• transformer-based generative model
• architectural priors