mlsanityData to 13 Oct 2026

RoboNaldo: Accurate, Stable and Powerful Humanoid Soccer Shooting via Motion-Guided Curriculum Reinforcement Learning

arXiv 2606.11092 ↗First posted 9 Jun 2026cs.RO cs.AIcited by 0

On the map: Deep reinforcement learning methods → RL for robot locomotion and manipulation

Abstract (arXiv)

Elite humanoid soccer shooting requires whole-body stability, high-impulse whole-body interactions, and accuracy to targets. Motion tracking-driven reinforcement learning (RL) provides stability in whole-body movement coordination, but a fixed reference makes it hard to adapt to varied ball positions and strike timings; in contrast, task reward-driven RL struggles to explore and discover valid kicks from scratch. We therefore introduce RoboNaldo, a three-stage motion-guided curriculum RL framework for high-impulse humanoid interaction. A single human-kick reference is used as a scaffold and progressively shifts optimization towards shooting performance. The curriculum first learns a stable whole-body kicking prior, then adapts the kick to free-kick settings where the ball is stationary at random positions, and finally extends it to moving-ball shooting through a locomotion-command and kick-trigger interface. A high-level heuristic planner controls this interface during training, while alternative high-level controllers can drive the same low-level policy at inference. In simulation, RoboNaldo demonstrates free-kick shot error 48.6% lower and shoot velocity 2.96x than prior work baselines. In real world on a Unitree G1 with onboard perception, RoboNaldo attains 0.73 m and 0.86 m average target shooting error from 3 m away in free-kick and moving-ball cases, accordingly. And the post-contact ball velocity reaches 13.10 m/s, which is 59-71% of reported professional open-play shot speed. Project page: https://opendrivelab.com/RoboNaldo.

Key facts. No citing papers in the map yet (data through 13 Oct 2026). Placed in RL for robot locomotion and manipulation, one of 15 directions in Deep reinforcement learning methods. Builds on 24 of its 31 references that are in the map. First posted on arXiv 9 Jun 2026.

Builds on

24 of 31 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025Humanoid Goalkeeper: Learning from Position Conditioned Task-Motion Constraints20 Oct
  2. 2025Learning coordinated badminton skills for legged manipulators29 May
  3. 2025Toward Real-World Cooperative and Competitive Soccer with Quadrupedal Robot Teams20 May

Other references · 21

  1. 2017Proximal Policy Optimization Algorithmscited by 14,592
  2. 2017Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real Worldcited by 1,411
  3. 2017Sim-to-Real Transfer of Robotic Control with Dynamics Randomizationcited by 622
  4. 2021Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learningcited by 399
  5. 2020Curriculum Learning for Reinforcement Learning Domains: A Framework and Surveycited by 351
  6. 2023Orbit: A Unified Simulation Framework for Interactive Robot Learning Environmentscited by 240
  7. 2025BeyondMimic: From Motion Tracking to Versatile Humanoid Control via Guided Diffusioncited by 132
  8. 2024World-Grounded Human Motion Recovery via Gravity-View Coordinatescited by 129
  9. 2023Learning Agile Soccer Skills for a Bipedal Robot with Deep Reinforcement Learningcited by 111
  10. 2025InterMimic: Towards Universal Whole-Body Control for Physics-Based Human-Object Interactionscited by 68
  11. 2025Retargeting Matters: General Motion Retargeting for Humanoid Motion Trackingcited by 64
  12. 2025ResMimic: From General Motion Tracking to Humanoid Whole-body Loco-Manipulation via Residual Learningcited by 36
9 more
  1. 2025HDMI: Learning Interactive Humanoid Whole-Body Control from Human Videoscited by 34
  2. 2025FALCON: Learning Force-Adaptive Humanoid Loco-Manipulationcited by 29
  3. 2025HITTER: A HumanoId Table TEnnis Robot via Hierarchical Planning and Learningcited by 29
  4. 2025DreamControl: Human-Inspired Whole-Body Humanoid Control for Scene Interaction via Guided Diffusioncited by 14
  5. 2026HumanX: Toward Agile and Generalizable Humanoid Interaction Skills from Human Videoscited by 11
  6. 2025Learning Vision-Driven Reactive Soccer Skills for Humanoid Robotscited by 4
  7. 2025Learning Agile Striker Skills for Humanoid Soccer Robots from Noisy Sensory Inputcited by 3
  8. 2026Learning Soccer Skills for Humanoid Robots: A Progressive Perception-Action Frameworkcited by 3
  9. 2025Like Playing a Video Game: Spatial-Temporal Optimization of Foot Trajectories for Controlled Football Kicking in Bipedal Robotscited by 1

Built on by

Not cited yet.