Enumerating Deformed Configurations of Reconfigurable Meshed Cellular Robots

Reconfigurable meshed cellular robots could be the future of next-generation adaptive structures and robots as they can undergo large shape changes while simultaneously bearing loads. This paper considers an embodiment of meshed cellular robots as a network combination of several Soft Two-State building blocks that can either be stiff or actuate like a muscle. Given a target deformed shape, it is required to determine the morphological state of each building block in the network (a stiff bone or a contractile muscle) that maximally meets the shape. The paper presents techniques to reduce the design space of possibilities by specifically isolating a bio-inspired two-muscle one-bone pattern for each triangular mesh. Enumeration of all designs that meet this pattern is shown to maximally capture all large deformation shapes. This significantly simplifies computations required for design evaluations and control with several design alternatives pre-sented to the designer. An example of an aquatic fin is used to illustrate the methodology where all prominent deformation modes of the fin can be captured through the pattern-based enumeration scheme. Future work can expand this methodology by considering more patterns and embedding the enumeration scheme within a generative topology design framework.