Automated Design for Additive Manufacturing Optimization through Machine Learning

This project aims to leverage machine learning to automate Design for Additive Manufacturing (DfAM) with an initial focus on light weight design. An already developed training set of voxelized files will be put through autoencoders to create neural networks which recognize features associated with DfAM. These features will then be applied to user submitted parts, resulting will be a mechanical part that satisfies the performance attributes of the original input, and is significantly lighter.