Scientists at MIT CSAIL (MIT Computer Science and Artificial Intelligence Laboratory) have developed a new method to more easily personalize 3D models for 3D printing. Using artificial intelligence (AI), the models are automatically broken down into functional and design parts. This allows users to customize the design without compromising function.
Previously, it was difficult to modify models designed for 3D printing by others. If the models were edited globally, it was often impossible to ensure function when printed. The MIT researchers therefore analyzed 1000 models from the online platform Thingiverse and found out which parts were important for function.
From these findings, they developed a semi-automatic approach to divide 3D models into functional and design segments. External-functional parts that interact with the environment are detected, as are internal-functional connections between components.
In tests, the function-based partitioning was found to help preserve function during personalization, especially for more complex models. The researchers presented Style2Fab, a system that can adjust design elements based on the new method without compromising function.
The full paper, titled “Style2Fab: Functionality-Aware Segmentation for Fabricating Personalized 3D Models with Generative AI” can be accessed in PDF format here.
Subscribe to our Newsletter
3DPresso is a weekly newsletter that links to the most exciting global stories from the 3D printing and additive manufacturing industry.
