Motion and Morphology Planning for Autonomous Manipulation in Clutter With a Shape-Shifting, Variable-Stiffness Robot

Abstract

Autonomous operation in confined spaces demands robots that can simultaneously navigate tight passages and manipulate objects. However, conventional mobile manipulators are often too large, while compliant soft manipulators are typically limited by tethers and a small operational workspace. This article introduces a novel hybrid robot designed to address this trade-off. The robot achieves mobile manipulation without dedicated end effectors by leveraging its entire body. It comprises two autonomous locomotion units linked by a Variable-Stiffness Bridge (VSB). This design enables transitions between a rigid state for efficient locomotion and a flexible state for full-body enclosure of objects. After enclosure, the robot re-stiffens to enable push-coupled planar transport with orientation regulation. We present the second iteration of this platform, featuring a modular VSB design that achieves a 2.5-fold increase in state transition speed. Furthermore, we introduce a comprehensive planning framework, the Voronoi-based Motion and Morphology Planner (VMMP), to enable autonomous navigation in highly cluttered planar environments. VMMP addresses the high-dimensional planning challenge through a hierarchical decomposition: it first uses Voronoi diagrams to generate traversable paths for three key control points on the robot's body and then reconstructs the kinematically feasible robot configuration sequence via constrained optimization. The source code and simulations are available at https://github.com/Louashka/2sr-robot.