Researchers at ETH Zurich have developed an artificial intelligence (AI) tool to design metamaterials with specific mechanical properties, such as those found in bike helmets or running shoes. The AI deciphers essential microstructure features and predicts deformation behavior, bypassing time-consuming engineering simulations. By using variational autoencoders and large datasets, the AI efficiently learns design principles, allowing it to generate optimal metamaterial blueprints for desired properties. The researchers have made their AI tools available to the metamaterials community, opening new possibilities for applications in protective equipment, medical engineering, and soft robotics.

Read more here.