A Framework for Deploying Learning-based Quadruped Loco-Manipulation

2025-12-21

AI Summary

Sim-to-real transfer for quadrupedal mobile manipulators faces challenges including unreliable control and proprietary, closed frameworks. This paper introduces the first open-source, end-to-end framework enabling training, evaluation, and deployment of PPO-based whole-body pose-manipulation coordinated controllers on the Unitree B1 quadruped equipped with the Z1 manipulator, establishing a cross-engine sim-to-sim and sim-to-real pipeline spanning Isaac Gym → MuJoCo → real hardware. Key contributions include: (1) the first seamless policy reuse and unified hardware-abstracted deployment between Isaac Gym and MuJoCo; (2) an empirical analysis revealing how contact model discrepancies critically impact policy generalization across simulators; and (3) experimental validation demonstrating that whole-body coordination expands the manipulator’s reachable workspace by 32% and improves grasping success rate by 27% over a floating-base baseline. The framework is fully open-sourced to ensure reproducibility.

Technology Category

• Intelligent Robots: Manipulation
• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Quadruped mobile manipulators offer strong potential for agile loco-manipulation but remain difficult to control and transfer reliably from simulation to reality. Reinforcement learning (RL) shows promise for whole-body control, yet most frameworks are proprietary and hard to reproduce on real hardware. We present an open pipeline for training, benchmarking, and deploying RL-based controllers on the Unitree B1 quadruped with a Z1 arm. The framework unifies sim-to-sim and sim-to-real transfer through ROS, re-implementing a policy trained in Isaac Gym, extending it to MuJoCo via a hardware abstraction layer, and deploying the same controller on physical hardware. Sim-to-sim experiments expose discrepancies between Isaac Gym and MuJoCo contact models that influence policy behavior, while real-world teleoperated object-picking trials show that coordinated whole-body control extends reach and improves manipulation over floating-base baselines. The pipeline provides a transparent, reproducible foundation for developing and analyzing RL-based loco-manipulation controllers and will be released open source to support future research.

Problem

• Develops an open pipeline for RL-based quadruped loco-manipulation control
• Addresses sim-to-sim and sim-to-real transfer challenges in robotics
• Enables reproducible whole-body control for agile manipulation tasks

Innovation

• Open pipeline for RL-based quadruped loco-manipulation control
• Unified sim-to-sim and sim-to-real transfer via ROS
• Hardware abstraction layer enabling cross-simulation and real deployment