Topology Design and Optimization of Modular Soft Robots (msoros) Capable of Homogenous and Heterogenous Reconfiguration

Abstract

Highly deformable Modular Soft Robots (MSoRos) have the unique ability to assemble into different topologies, e.g., planar and spherical, which display widely different locomotion modes (e.g., crawling or rolling). This research presents topology design and optimization methodology of MSoRos based on Archimedean solids capable of both heterogeneous and homogeneous spherical and planar reconfiguration. The sequential approach comprises of forward and inverse design processes. The forward design process, based on a selected polyhedron (either Archemedian or Platonic solid), models the robot topology in spherical configuration and generates the planar topology. The inverse design creates the spherical topology given using the planar topology. Reconfiguration alignment is ensured using the discussed polyhedron vertex alignment principle, and the planar and spherical distortion metrics are defined to characterize distortions due to reconfiguration. Optimal topology is obtained by minimizing a cost function that is a weighted sum of the two distortion metrics. As the design processes involve nonlinear projections, the starting base polyhedron for forward design is critical (Archemedian for heterogeneous and Platonic for homogeneous). The optimal trajectory is explored for both the scenarios and with varying weights in the cost function. The result is an MSoRo capable of distinct 1-D and 2-D planar configurations (both heterogeneous and homogeneous) and multiple 3-D spherical configurations of varying radii (both heterogeneous and homogeneous). The methodology is tested on an MSoRo system based on a cuboctahedron (Archimedean solid) and a cube and an octahedron (Platonic solids).