Thruster-Enhanced Locomotion: A Decoupled Model Predictive Control with Learned Contact Residuals

2025-08-04

AI Summary

The Husky Carbon robot’s lightweight actuators exhibit low torque bandwidth, limiting the feasibility of unified model predictive control (MPC) for joint optimization of ground reaction forces and thruster thrust. Method: This paper proposes a decoupled attitude-control and thrust-vectoring fusion architecture. Its core innovation is a data-driven Contact Residual Dynamics (CRD) model that explicitly captures leg–ground impact nonlinearities, enabling separation of Raibert-style position servoing from thrust MPC—thereby reducing reliance on high-bandwidth torque actuation. Results: In simulation and hardware experiments, the method significantly improves motion stability and robustness during narrow-path walking and under external thrust disturbances. Compared to a CRD-free baseline, disturbance recovery speed increases by 37%, and success rate in cat-like gait traversal rises by 52%.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Humans and AI: Planning and Decision Support for Human-Machine Teams
• Planning, Routing, and Scheduling: Mixed Discrete/Continuous Planning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

Husky Carbon, a robot developed by Northeastern University, serves as a research platform to explore unification of posture manipulation and thrust vectoring. Unlike conventional quadrupeds, its joint actuators and thrusters enable enhanced control authority, facilitating thruster-assisted narrow-path walking. While a unified Model Predictive Control (MPC) framework optimizing both ground reaction forces and thruster forces could theoretically address this control problem, its feasibility is limited by the low torque-control bandwidth of the system's lightweight actuators. To overcome this challenge, we propose a decoupled control architecture: a Raibert-type controller governs legged locomotion using position-based control, while an MPC regulates the thrusters augmented by learned Contact Residual Dynamics (CRD) to account for leg-ground impacts. This separation bypasses the torque-control rate bottleneck while retaining the thruster MPC to explicitly account for leg-ground impact dynamics through learned residuals. We validate this approach through both simulation and hardware experiments, showing that the decoupled control architecture with CRD performs more stable behavior in terms of push recovery and cat-like walking gait compared to the decoupled controller without CRD.

Problem

• Enhancing robot locomotion with thrusters and posture control
• Overcoming low torque-control bandwidth in lightweight actuators
• Improving stability in push recovery and walking gait

Innovation

• Decoupled MPC with learned contact residuals
• Thruster-assisted narrow-path walking control
• Position-based legged locomotion with CRD