Controlling 3D Printers with Artificial Neural Networks

With the increased availability of 3D printers, the access to such fabrication methods has outpaced the education in the skills needed to design objects for printing. One approach to solving this problem is to create a system capable of designing these objects independently of human interaction. The EvoFab Project is a system whose long term goal is to solve this problem. Throughout the EvoFab project, both open-loop and closed-loop instruction sets were used to evolve solutions to various fabrication problems, but it is unclear as to which control set performs better under the noisy environment of the real world. Our study explores the design and implementation of a 3D printing system controlled via Artificial Neural Networks to determine the robustness of this control technique and to further work toward the overarching goal of the EvoFab Project.