RAKOMO: Reachability-Aware K-Order Markov Path Optimization for Quadrupedal Loco-Manipulation
🤖 AI Summary
Quadrupedal mobile manipulation faces challenges including hybrid dynamical modeling difficulties due to contact switching, stringent kinematic constraints, and limited leg workspace. Method: This paper proposes a reachability-aware whole-body motion planning framework. Its core innovation integrates neural-network-predicted reachability margins into the KOMO (K-order Markov Optimization) framework, enabling gradient-based optimization to explicitly account for limb reachability boundaries and contact-state transitions. The method jointly optimizes base pose, foot contact sequences, and manipulator trajectories. Results: Evaluated on a HyQReal+Kinova Gen3 simulation platform, the approach achieves safe and continuous loco-manipulation planning. In pick-and-place tasks, it improves planning success rate by 32% and reduces collision rate by 57% compared to standard KOMO, demonstrating superior effectiveness and robustness in complex, dynamic contact-switching scenarios.
Application Category
📝 Abstract
Legged manipulators, such as quadrupeds equipped with robotic arms, require motion planning techniques that account for their complex kinematic constraints in order to perform manipulation tasks both safely and effectively. However, trajectory optimization methods often face challenges due to the hybrid dynamics introduced by contact discontinuities, and tend to neglect leg limitations during planning for computational reasons. In this work, we propose RAKOMO, a path optimization technique that integrates the strengths of K-Order Markov Optimization (KOMO) with a kinematically-aware criterion based on the reachable region defined as reachability margin. We leverage a neural-network to predict the margin and optimize it by incorporating it in the standard KOMO formulation. This approach enables rapid convergence of gradient-based motion planning -- commonly tailored for continuous systems -- while adapting it effectively to legged manipulators, successfully executing loco-manipulation tasks. We benchmark RAKOMO against a baseline KOMO approach through a set of simulations for pick-and-place tasks with the HyQReal quadruped robot equipped with a Kinova Gen3 robotic arm.
Problem
Research questions and friction points this paper is trying to address.
Motion planning for legged manipulators with kinematic constraints
Addressing hybrid dynamics from contact discontinuities in trajectory optimization
Incorporating leg limitations and reachability in loco-manipulation tasks
Innovation
Methods, ideas, or system contributions that make the work stand out.
Reachability-Aware K-Order Markov Optimization
Neural-network predicted reachability margin
Gradient-based motion planning adaptation
🔎 Similar Papers
No similar papers found.
