Olympus: A Jumping Quadruped for Planetary Exploration Utilizing Reinforcement Learning for In-Flight Attitude Control

2025-03-05

AI Summary

To address the low mobility efficiency of conventional rovers in low-gravity environments (e.g., Moon and Mars), this work proposes Olympus, a hopping quadruped robot designed to tackle autonomous in-flight attitude regulation. Methodologically, we integrate physics-based multi-objective structural optimization, Proximal Policy Optimization (PPO)-based reinforcement learning (RL) for flight control policy design, and a real-time embedded attitude control framework. Notably, this is the first successful deployment of RL for in-flight attitude control on a planetary hopping robot, coupled with high-fidelity sim-to-real transfer. Experimental validation on physical hardware demonstrates robust multimodal aerial tumbling and stable landing: vertical hops reach 1.8 m, forward hops exceed 3.2 m, and attitude tracking error remains below 3°. These results significantly enhance maneuverability and mission adaptability under low-gravity conditions.

Technology Category

• Intelligent Robots: Learning & Optimization for ROB
• Search and Optimization: Learning to Search
• Planning, Routing, and Scheduling: Replanning and Plan Repair

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
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

Exploring planetary bodies with lower gravity, such as the moon and Mars, allows legged robots to utilize jumping as an efficient form of locomotion thus giving them a valuable advantage over traditional rovers for exploration. Motivated by this fact, this paper presents the design, simulation, and learning-based"in-flight"attitude control of Olympus, a jumping legged robot tailored to the gravity of Mars. First, the design requirements are outlined followed by detailing how simulation enabled optimizing the robot's design - from its legs to the overall configuration - towards high vertical jumping, forward jumping distance, and in-flight attitude reorientation. Subsequently, the reinforcement learning policy used to track desired in-flight attitude maneuvers is presented. Successfully crossing the sim2real gap, extensive experimental studies of attitude reorientation tests are demonstrated.

Problem

• Designing a jumping robot for planetary exploration
• Optimizing robot design for high jumping performance
• Implementing reinforcement learning for in-flight attitude control

Innovation

• Reinforcement learning for in-flight attitude control
• Simulation-optimized design for high vertical jumping
• Sim2real gap crossing in experimental studies