Application of Tiling Theory for Path Planning Strategy in a Polyiamond Inspired Reconfigurable Robot

摘要

Commercial floor cleaning robots face significant challenges in accessing convex and narrow corners due to their fixed and regular morphologies. To overcome this, we develop a new class of self-reconfigurable floor cleaning robot, hTetrakis, which is composed of tetriamonds (four equilateral triangles aligned along the edges) that adopt three distinct forms (“I”, “A”, and “U” shapes). When on a flat and rigid platform, these forms have convex corners that help to cover narrow regions. This paper addresses the mechanical structural design, reconfiguration of the robot platform through the hinge mechanism, and the electronics and navigation module of hTetrakis. Based on finite element studies, we estimate the system’s natural frequency, stress, and deformation patterns developed in the structural components of the robot, and validate the proposed design to overcome structural failure and system resonance. In order to achieve maximum area coverage using the tetriamond forms, we formulate tiling theorems and apply them for path-planning techniques during the floor cleaning process. By using a robot prototype, we conduct experiments to validate the proposed tiling theorem based on the percentage of area coverage, and demonstrate that this platform is able to cover the floor area efficiently.