Ancient Art Meets AI

 Reported on this before,   unexpected connection.

Ancient Art Meets AI for Better Materials Design

Argonne National Laboratory, John Spizzirri, April 7, 2022

University of Southern California (USC) researchers combined kirigami, the ancient Japanese art of paper cutting, with autonomous reinforcement learning to help improve materials design. In an effort to create a two-dimensional molybdenum disulfide structure embedded with electronics that can stretch while remaining stable, the researchers determined that a series of precise cuts could enable the thin material to stretch up to 40%. To determine the correct combination of cuts, the researchers performed simulations on the Theta supercomputer at the U.S. Department of Energy's Argonne National Laboratory. The model was trained on 98,500 simulations of kirigami design strategies involving one to six cuts; even without additional training data, it determined in a matter of seconds that 10 cuts would provide more than 40% stretchability. USC's Pankaj Rajak said, "It learned something the way a human learns, and used its knowledge to do something different."