Zero-Shot Retargeting of Learned Quadruped Locomotion Policies Using Hybrid Kinodynamic Model Predictive Control

TL;DR

This paper introduces a hybrid kinodynamic MPC framework that enables zero-shot transfer of learned quadruped locomotion policies across different robot platforms, reducing the need for retraining.

Contribution

The work presents a novel planning-and-control pipeline combining RL and MPC with a hybrid kinodynamic model for effective policy retargeting across diverse quadruped robots.

Findings

Successful transfer of policies from A1 and Laikago to MIT Mini Cheetah

No policy re-tuning required during transfer

Robust and stable locomotion demonstrated on hardware

Abstract

Reinforcement Learning (RL) has witnessed great strides for quadruped locomotion, with continued progress in the reliable sim-to-real transfer of policies. However, it remains a challenge to reuse a policy on another robot, which could save time for retraining. In this work, we present a framework for zero-shot policy retargeting wherein diverse motor skills can be transferred between robots of different shapes and sizes. The new framework centers on a planning-and-control pipeline that systematically integrates RL and Model Predictive Control (MPC). The planning stage employs RL to generate a dynamically plausible trajectory as well as the contact schedule, avoiding the combinatorial complexity of contact sequence optimization. This information is then used to seed the MPC to stabilize and robustify the policy roll-out via a new Hybrid Kinodynamic (HKD) model that implicitly optimizes the foothold locations. Hardware results show an ability to transfer policies from both the A1 and Laikago robots to the MIT Mini Cheetah robot without requiring any policy re-tuning.