Exploiting Geometrical Frustration in Multistable Soft Robots Part 2: Finite-State Mechanologic

Mechanologic and information processing in elastic metamaterials has focused on realizations of Boolean logic following a von Neumann architecture. This implies the separation of memory and computation and the use of mechanical binary logic as the fundamental operation for information processing. Recently, hierarchical multistable metasheets have been shown to display collocated memory and computation capabilities. Concretely, such multistable metastructures use energy minimization into an encoded stable state as the fundamental operation for computation. We extend this principle to realize embodied finite-state mechanologic in multistable soft robots, whose kinematic configuration depends on a sequence of state transitions. We show that constructive deformation of parts of the robot leads to particular kinematic finite states. These finite states can be designed as desired robotic configurations, for example, for adopting different manipulation poses. Our results demonstrate finite-state mechanologic in soft multistable robotics, thus departing from von-Neumann architectures and binary logic for mechanical computation.