Scalable Thermoset Informatics and On-Demand Discovery through Autonomous Experimentation

Abstract

As photo-polymer additive manufacturing matures, there is an increasing need for on-demand, application-specific materials. A major bottleneck is the limited exploration of the vast design space for new thermoset polymer formulations, which must satisfy both material property requirements and processing constraints. This challenge stems from the combinatorial complexity of possible formulations and the high cost of experimental evaluation. In this work, we present an end-to-end pipeline that integrates a self-driving laboratory (SDL) with informatics-driven optimization to discover and validate novel thermoset acrylate formulations. The pipeline starts with a database focusing on thermoset acrylates that consolidates multiple data sources (including manual and SDL experiments, molecular dynamics simulations, and literature data). Selected datasets are used to train both single-task and multitask predictive models, which interface directly with our SDL platform to measure mechanical and processing-related properties. Each component of the pipeline is validated independently and then deployed together to optimize a target elastomeric material in significantly less time and with substantially reduced human effort. This work establishes a scalable framework for embedding multi-fidelity machine learning tools within laboratory workflows, paving the way for the next generation of autonomous, data-driven thermoset formulation discovery for additive manufacturing.