Changqing Chen is a Professor and the Head of the Department of Engineering Mechanics at Tsinghua University in China. He obtained his B.Eng. and Ph.D. degrees from Xi’an Jiaotong University in China in 1993 and 1997, respectively. He also completed a Postdoctoral program at Cambridge University in the UK. Before joining Tsinghua University in 2009, he was a professor of mechanics at Xi’an Jiaotong University. He has received numerous awards and honors, including the National Distinguished Young Scientist Award from the NSFC of China and the China Youth Science and Technology Award from the MOST of China. His research interests primarily involve multifunctional materials and structures, with a particular focus on mechanical metamaterials in recent years.

 

Abstract: Cogitative/intelligent materials in which information computing plays a pivot role have been imagined but not realized for decades. In this talk, we discuss two cogitative mechanical metamaterials. The first is a reprogrammable mechanological metamaterial (ReMM) that is capable of universal computing within the von Neumann architecture. Its feasibility for basic functions such as mechanical computation, deformation signal transmission, combinational logic, and sequential logic is demonstrated experimentally and numerically. The second is an in-memory mechanical computing system that mimics neuromorphic computing. In this system, computing occurs within the interaction network of mechanical memory units. The architecture that is suitable for in-memory mechanical computing, together with the basic interactions needed to fulfill function-complete computing, is proposed. Several mechanical neural networks, including the binary neural network, self-learning perceptron, convolutional neural network, and recurrent neural network with long short-term memory, are demonstrated. The obtained results show that by delicately designing their microstructure, mechanical metamaterials can be developed to possess cogitative capabilities.