Toward autonomous additive manufacturing: Bayesian optimization on a 3D printer

Abstract Materials exploration and development for three-dimensional (3D) printing technologies is slow labor-intensive. Each 3D material developed requires unique print parameters be learned successful part fabrication, sub-optimal settings often result in defects or fabrication failure. To address this, we the Additive Manufacturing Autonomous Research System (AM ARES). As a preliminary test, tasked AM ARES with autonomously modulating four to direct-write single-layer features that matched target specifications. employed automated image analysis as closed-loop feedback an online Bayesian optimizer fewer than 100 experiments. In due course, this first-of-its-kind research robot will autonomous multi-dimensional optimization of accelerate materials discovery field AM. The combining open-source OS software low-cost hardware makes highly accessible, promoting mainstream adoption rapid technological advancement. Impact statement new processes hindered by labor-intensive trial-and-error processes. Coupled pervasive lack mechanisms printers, has inhibited advancement additive manufacturing (AM) approach. streamline process AM, have accessible employs machine learning planners, together our software, which release community open-source, rapidly effectively optimize complex, high-dimensional parameter sets associated printing. trials, robot, ARES), extrusion specimens closely targeted feature specifications under iterations. Delegating repetitive cognitive labor robots such frees researchers focus on more creative, insightful, fundamental scientific work reduces cost time required develop teaming human begets synergy exponentially propel progress