In collaboration with Autodesk, we used artificial intelligence to predict how a passively actuated material would react to temperature. With the help of Hydra we ran thousands of simulations to understand how thermo-active laminated behave under varied heat conditions. We then used that data to train a deep neural network to tell us what the laminate layering should be given a particular deformation we required.

We used machine learning to teach a tool to approximate the complex analysis of visual and spatial connectivity, by training it with a generative dataset of tens of thousands of actual analysis. The predicted output is extremely comparable to the actual analysis and takes around 0.03 seconds to calculate, οffering vastly improved speeds.