Duke Materials Database to Power New AI-Driven Cloud Laboratory

Over a decade of work pursued by Duke Engineering researchers to computationally predict recipes for useful materials will soon help fuel a new AI-driven cloud laboratory that will transform materials innovation.

Led by Georgia Tech, a new Programmable Cloud Laboratory will leverage AI, simulation and autonomous experimentation to dramatically accelerate the process of materials discovery. Built upon Georgia Tech’s Advanced Manufacturing Pilot Facility (AMPF) and Duke’s materials discovery computational ecosystem, the cloud lab will allow researchers across the country to direct work remotely, refine experiments based on results and AI recommendations, and tap into advanced manufacturing capabilities without spending weeks on-site. In effect, the cloud lab will bring the facility to the researcher.

Duke’s part in this ambitious project is led by Stefano Curtarolo, the Edmund T. Pratt Jr. School Distinguished Professor of Mechanical Engineering and Materials Science. His group has spent over a decade building and maintaining the Automatic-FLOW for Materials Discovery database, or aflow.org. The expansive work provides information on more than 3.5 million material compounds with over 730 million calculated properties and state-of the-art software, called aflow++, that automatically discovers and parameterizes new materials through thermodynamic and quantum calculations. This rich platform will integrate computational discovery with physical experimentation to help researchers more rapidly identify, evaluate and manufacture promising new materials.

The integration of AI into manufacturing will enable rapid optimization of ceramics and metallic alloys for achieving the necessary performance required by modern technologies,” Curtarolo said. “The project will also function as a potential incubator for collaborations with companies and businesses working in the Duke-Georgia Tech area along the I-85 corridor.”

The cloud lab is supported by $18.1 million from the National Science Foundation (NSF) and is part of a broader effort to build a national network of 20 AI-enabled cloud laboratories. The labs are designed to work together, eventually allowing researchers to combine capabilities and workflows across the network. The new research program will connect advanced scientific infrastructure with expertise across the country.

Read more about the Programmable Cloud Laboratory in Georgia Tech’s announcement.