Optimize Any Topology: A Foundation Model for Shape- and Resolution-Free Structural Topology Optimization

2025-10-26

AI Summary

Existing topology optimization methods suffer from high computational cost, while mainstream deep learning approaches are constrained by fixed square grids, limited boundary conditions, and reliance on post-processing—resulting in poor generalization. This paper introduces the first generative foundation model framework for topology optimization that eliminates post-processing and supports arbitrary aspect ratios and resolutions. Key contributions include: (i) construction of OpenTO, a large-scale dataset comprising 2.2 million samples; (ii) design of a shape- and resolution-invariant autoencoder, an implicit neural field decoder, and a physics-aware conditional latent diffusion model. Evaluated on multiple public benchmarks, our method reduces compliance by up to 90% relative to optimal baselines. It enables sub-second inference on a single GPU across resolutions from 64×64 to 256×256 and aspect ratios up to 10:1, significantly enhancing cross-configuration generalization and engineering practicality.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Search and Optimization: Sampling/Simulation-based Search
• Computer Vision: Diffusion Models for Vision

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Structural topology optimization (TO) is central to engineering design but remains computationally intensive due to complex physics and hard constraints. Existing deep-learning methods are limited to fixed square grids, a few hand-coded boundary conditions, and post-hoc optimization, preventing general deployment. We introduce Optimize Any Topology (OAT), a foundation-model framework that directly predicts minimum-compliance layouts for arbitrary aspect ratios, resolutions, volume fractions, loads, and fixtures. OAT combines a resolution- and shape-agnostic autoencoder with an implicit neural-field decoder and a conditional latent-diffusion model trained on OpenTO, a new corpus of 2.2 million optimized structures covering 2 million unique boundary-condition configurations. On four public benchmarks and two challenging unseen tests, OAT lowers mean compliance up to 90% relative to the best prior models and delivers sub-1 second inference on a single GPU across resolutions from 64 x 64 to 256 x 256 and aspect ratios as high as 10:1. These results establish OAT as a general, fast, and resolution-free framework for physics-aware topology optimization and provide a large-scale dataset to spur further research in generative modeling for inverse design. Code & data can be found at https://github.com/ahnobari/OptimizeAnyTopology.

Problem

• Overcoming fixed grid limitations in structural topology optimization methods
• Enabling arbitrary resolution and boundary condition handling in topology optimization
• Reducing computational intensity while maintaining physics-aware optimization performance

Innovation

• Uses shape-agnostic autoencoder with implicit neural-field decoder
• Employs conditional latent-diffusion model for structure prediction
• Trains on large dataset of 2.2 million optimized structures