Materials Innovation at Argonne

 New looks at innovation with AI and HPC at Argonne.

Automatic Database Creation for Materials Discovery: Innovation From Frustration

Argonne National Laboratory  by John Spizzirri

Scientists at the U.S. Department of Energy's Argonne National Laboratory and the U.K.'s University of Cambridge have developed a method that generates automatic databases to support specific scientific disciplines, using artificial intelligence and high-performance computing (HPC). The technique can assemble databases via natural language processing (NLP) and HPC, most of which was performed at the Argonne Leadership Computing Facility. The team built a database on both material structures and material properties, using the NLP ChemDataExtractor data-mining application. Cambridge's Jacqueline Cole said, "It's probably the first such compilation of a database on such a massive scale, with 5,380 like-for-like pairs of experimental and calculated data. And because it's such a large amount, it serves as a repository in its own right and really opens the door to predicting new materials."