Optimize Any Topology: A Foundation Model for Shape- and Resolution-Free Structural Topology Optimization
Amin Heyrani Nobari, Lyle Regenwetter, Cyril Picard, Ligong Han, Faez Ahmed
2025-10-29
Summary
This paper introduces a new system called Optimize Any Topology, or OAT, which uses artificial intelligence to quickly and efficiently design strong and lightweight structures for engineering purposes.
What's the problem?
Designing the best shape for a structure to withstand forces, a process called topology optimization, is usually very slow and requires a lot of computing power. Existing AI methods for this task were limited because they only worked with simple square shapes, couldn't handle many different types of supports or loads, and needed further tweaking after the AI made its initial design.
What's the solution?
The researchers created OAT, a system built on a 'foundation model' – think of it as a pre-trained AI that can be adapted to many different situations. OAT uses a special type of AI called a diffusion model, trained on a massive dataset of over two million optimized structures with varying shapes, loads, and supports. This allows OAT to directly predict the best structure layout for almost any situation, regardless of its shape or size, and do so very quickly.
Why it matters?
OAT is a big step forward because it's a general-purpose, fast, and accurate tool for structural design. It can significantly improve designs compared to previous methods and can generate solutions in under a second on a standard computer. Furthermore, the researchers also released the large dataset they used to train OAT, which will help other researchers develop even better AI-powered design tools in the future.
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.