RMA: Rapid Motor Adaptation for Legged Robots | HTML5

Successful real-world deployment of legged robots would require them to adapt in real-time to unseen scenarios like changing terrains, changing payloads, wear and tear. This paper presents Rapid Motor Adaptation (RMA) algorithm to solve this problem of real-time online adaptation in quadruped robots. RMA consists of two components: a base policy and an adaptation module. The combination of these components enables the robot to adapt to novel situations in fractions of a second. RMA is trained completely in simulation without using any domain knowledge like reference trajectories or predefined foot trajectory generators and is deployed on the A1 robot without any fine-tuning. We train RMA on a varied terrain generator using bioenergetics-inspired rewards and deploy it on a variety of difficult terrains including rocky, slippery, deformable surfaces in environments with grass, long vegetation, concrete, pebbles, stairs, sand, etc. RMA shows state-of-the-art performance across diverse rea

RMA: Rapid Motor Adaptation for Legged Robots Ashish Kumar UC Berkeley Zipeng Fu Carnegie Mellon University Deepak Pathak Carnegie Mellon University Jitendra Malik UC Berkeley, Facebook Abstract Successful real-world deployment of legged robots would require them to adapt in real-time to unseen scenarios like changing terrains, changing payloads, wear and tear. This paper presents Rapid Motor Adaptation (RMA) algorithm to solve this problem of real-time online adaptation in quadruped robots. RMA consists of two components: a base policy and an adaptation module. The combination of these compon

related reading

• A Unified Model for Motion-Conditioned Robot Co-designtransformer-transformer.github.io
• Precise Manipulation with Efficient Online RLpi.website
• Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learningarxiv.org
• [2509.22407] EMMA: Generalizing Real-World Robot Manipulation via Generative Visual Transferarxiv.org
• Versatile Loco-Manipulation through Flexible Interlimb Coordinationarxiv.org
• [2401.14403] Adaptive Mobile Manipulation for Articulated Objects In the Open Worldar5iv.labs.arxiv.org
• Emergent Dexterity via Diverse Resets and Large-Scale Reinforcement Learningarxiv.org
• Learning Sim-to-Real Humanoid Locomotion in 15 Minutesyounggyo.me
• [2603.08546] Interactive World Simulator for Robot Policy Training and Evaluationarxiv.org
• [1910.07113] Solving Rubik's Cube with a Robot Handarxiv.org
• RI Seminar: Jitendra Malik : Robot Learning, With Inspiration From Child Developmentyoutube.com
• [1709.06560] Deep Reinforcement Learning that Mattersarxiv.org