The Bayesian Experimental Autonomous Researcher (BEAR) is a system that designs, prints, and tests components for mechanical performance without human intervention. Through the combination of extrusion-based 3D printing and uniaxial mechanical testing, along with characterization of parts including weighing and imaging, the system can print and test as many as 100 samples per day. This high throughput testing is coupled with advanced computational fabrication strategies for parametric design of mechanical structures, and an active learning loop that uses Bayesian optimization to iteratively select experiments. The overarching goal of this system is to allow users to take full advantage of the design flexibility inherent to additive manufacturing and mechanical metamaterials with superlative properties. This system specializes in developing systems for non-linear mechanical properties such as energy absorption.