A Bayesian experimental autonomous researcher for mechanical design

TLDR

Additive manufacturing enables complex structures, yet finding the optimal design remains difficult, and while numerical methods aid optimization, experiments are still the gold standard for nonlinear mechanical properties such as toughness. The authors develop a Bayesian experimental autonomous researcher (BEAR) that merges Bayesian optimization with high‑throughput automated experimentation to navigate the vast additive‑manufacturing design space. BEAR iteratively selects experiments based on accumulated results, enabling rapid, high‑throughput testing. Using BEAR, the authors achieve an almost 60‑fold reduction in experiments needed to identify high‑performing structures compared to a grid search, demonstrating the value of machine learning in data‑sparse experimental domains.

Abstract

While additive manufacturing (AM) has facilitated the production of complex structures, it has also highlighted the immense challenge inherent in identifying the optimum AM structure for a given application. Numerical methods are important tools for optimization, but experiment remains the gold standard for studying nonlinear, but critical, mechanical properties such as toughness. To address the vastness of AM design space and the need for experiment, we develop a Bayesian experimental autonomous researcher (BEAR) that combines Bayesian optimization and high-throughput automated experimentation. In addition to rapidly performing experiments, the BEAR leverages iterative experimentation by selecting experiments based on all available results. Using the BEAR, we explore the toughness of a parametric family of structures and observe an almost 60-fold reduction in the number of experiments needed to identify high-performing structures relative to a grid-based search. These results show the value of machine learning in experimental fields where data are sparse.