Abstract:Metamaterials are designed to realize exotic physical properties through the geometric arrangement of their underlying structural layout<a href="/articles/s41586-020-03123-5#ref-CR1">1</a>,<a href="/articles/s41586-020-03123-5#ref-CR2">2</a>. Traditional mechanical metamaterials achieve functionalities such as a target Poisson's ratio<a href="/articles/s41586-020-03123-5#ref-CR3">3</a> or shape transformation<a href="#ref-CR4">4</a>,<a href="#ref-CR5">5</a>,<a href="/articles/s41586-020-03123-5#ref-CR6">6</a> through unit-cell optimization<a href="#ref-CR7">7</a>,<a href="#ref-CR8">8</a>,<a href="/articles/s41586-020-03123-5#ref-CR9">9</a>, often with spatial heterogeneity<a href="#ref-CR10">10</a>,<a href="#ref-CR11">11</a>,<a href="/articles/s41586-020-03123-5#ref-CR12">12</a>. These functionalities are programmed into the layout of the metamaterial in a way that cannot be altered. Although recent efforts have produced means of tuning such properties post-fabrication<a href="#ref-CR13">13</a>,<a href="#ref-CR14">14</a>,<a href="#ref-CR15">15</a>,<a href="#ref-CR16">16</a>,<a href="#ref-CR17">17</a>,<a href="#ref-CR18">18</a>,<a href="/articles/s41586-020-03123-5#ref-CR19">19</a>, they have not demonstrated mechanical reprogrammability analogous to that of digital devices, such as hard disk drives, in which each unit can be written to or read from in real time as required. Here we overcome this challenge by using a design framework for a tileable mechanical metamaterial with stable memory at the unit-cell level. Our design comprises an array of physical binary elements (m-bits), analogous to digital bits, with clearly delineated writing and reading phases. Each m-bit can be independently and reversibly switched between two stable states (acting as memory) using magnetic actuation to move between the equilibria of a bistable shell<a href="#ref-CR20">20</a>,<a href="#ref-CR21">21</a>,<a href="#ref-CR22">22</a>,<a href="#ref-CR23">23</a>,<a href="#ref-CR24">24</a>,<a href="/articles/s41586-020-03123-5#ref-CR25">25</a>. Under deformation, each state is associated with a distinctly different mechanical response that is fully elastic and can be reversibly cycled until the system is reprogrammed. Encoding a set of binary instructions onto the tiled array yields markedly different mechanical properties; specifically, the stiffness and strength can be made to range over an order of magnitude. We expect that the stable memory and on-demand reprogrammability of mechanical properties in this design paradigm will facilitate the development of advanced forms of mechanical metamaterials.

Reprogrammable, in-materia matrix-vector multiplication with floppy modes

A reprogrammable mechanical metamaterial with stable memory

Rubik's cube as in-situ programmable matter and a reconfigurable mechanical metamaterial

Reprogrammable MetamaterialProcessors for Soft Machines

Reprogrammable Digital Metamaterials for Interactive Devices

A mechanical metamaterial with reprogrammable logical functions

A Reconfigurable Photonic Computing Chip with Easily Scalable Topology

Reprogrammable Logic in Memristive Crossbar for In-Memory Computing

Reprogrammable and reconfigurable mechanical computing metastructures with stable and high-density memory

Reconfigurable Matrix Multiplier with On-Site Reinforcement Learning.

Large-Scale Integrated Vector-Matrix Multiplication Processor Based on Monolayer MoS2

Multifunctional and reprogrammable 4D pixel mechanical metamaterials

A dynamically reprogrammable surface with self-evolving shape morphing

Flexible Communication Avoiding Matrix Multiplication on FPGA with High-Level Synthesis

Mechanical intelligence via fully reconfigurable elastic neuromorphic metasurfaces

Low-depth, compact and error-tolerant photonic matrix-vector multiplication beyond the unitary group

Fully reconfigurable coherent optical vector-matrix multiplication

Digital Mechanical Metamaterial: Encoding Mechanical Information with Graphical Stiffness Pattern for Adaptive Soft Machines

Analogue Signal and Image Processing with Large Memristor Crossbars

MatPIM: Accelerating Matrix Operations with Memristive Stateful Logic

Learning Arbitrary Complex Matrices by Interlacing Amplitude and Phase Masks with Fixed Unitary Operations