IRG1: Learning Metamaterials

IRG Leaders: John C. Crocker & Eleni Katifori

Senior Investigators: Douglas J. Durian, David A. Issadore, Paul A. Janmey, Andrea J. Liu, Marc Z. Miskin, Yoichiro Mori, James H. Pikul, Kevin T. Turner

Collaborators: Yi-Wei Chang, Daeyeon Lee

Local adaptive learning in networks – edge weight remodeling that uses only locally available information – has been proposed for living systems. This IRG will study locally learning networks, and in particular advance the theory of coupled learning, a recently introduced framework, as well as produce physical realizations of such systems i.e. learning metamaterials and a VLSI electrical coupled-learning circuit. These learning platforms will be used to solve engineering and materials problems by designing multi-functional large-scale microfluidic networks, realize highly capable soft robots and shape camouflage materials. The team will explore biopolymer networks that are implementing local learning rules via biopolymer-associated proteins, and functional mechanical metamaterials with the ability to learn. Last, this IRG will deepen our understanding of physical learning processes by studying structure-property (or structure-function) relations and the link between learning and information, explore new learning strategies that exploit dynamics and nonlinearities, as well as different ways of coupling local rules to create future learning metamaterials to promote the Data Revolution.

Eleni Katifori presenting IRG1 science at a Science Café on Networks.

Highlights for IRG1