Snapp, Kelsey L.Verdier, BenjaminGongora, AldairSilverman, SamuelAdesiji, Adedire D.Morgan, Elise F.Lawton, Timothy J.Whiting, EmilyBrown, Keith A.2023-09-122023-09-12https://hdl.handle.net/2144/46687Dataset.csv contains all of the metadata for each experiment, such as design parameters, filament information, and other important data. Importantly, the column labeled "Valid" signifies whether the experiment was considered successful and used to train the model. If the part was excluded, the reason for exclusion is notified by the subsequent columns. For each experiment, the Instron file which contains the time, force, and displacement for the physical test is included. NOTE that not every part will have an associated Instron file, as some parts were not tested. The format of the file is: ADTS#.csv where # is the ID number of the part that you are interested in.This dataset contains the experimental data for the research paper "Autonomous Discovery of Tough Structure" which is currently in the peer review process. The document contains the record of physical experiments performed over the course of two years by a self-driving lab called the BEAR (Bayesian Autonomous Experimental Researcher). The BEAR consists of five FFF 3d printers, a scale, and an Instron. A UR5 robot arm moves samples for testing.en-USAttribution-NonCommercial-ShareAlike 4.0 Internationalhttp://creativecommons.org/licenses/by-nc-sa/4.0/MechanicsSelf-driving labAutonomous experimentationEnergy absorbing efficiencyGeneralized cylindrical shellsMachine learningBayesian optimizationGaussian process regressionDataset