Humanoid Badminton: Learning Dynamic Racket Skills from Limited Human Motion Data

2026-09-25

AI Summary

This study addresses the challenge of acquiring high-speed dynamic badminton skills in humanoid robots due to the scarcity of human motion data. To this end, we propose a three-stage hierarchical reinforcement learning framework that pioneers the integration of task-randomized motion augmentation with context-conditioned adversarial regularization. By modeling a latent skill space, our approach generates diverse hitting motions from limited data and composes them online during execution, effectively balancing movement naturalness with task performance. This work demonstrates, for the first time on a physical humanoid robot, multi-skill human-robot rallies encompassing forehand strokes, backhand strokes, and highly dynamic jumping returns.

Abstract

High-speed racket sports provide a demanding testbed for humanoid robots, requiring time-critical decisions, precise striking, and dynamic whole-body coordination. In badminton, fast-changing shuttle trajectories require timely contact decisions, while successful returns demand precise racket pose and velocity within a brief contact window and across a broad three-dimensional striking workspace. Human motion data provide valuable priors for such athletic skills, but usable badminton references are limited and imperfect. Direct tracking provides insufficient executable variation for diverse shuttle conditions, while purely task-driven optimization may produce unnatural motion. To address these challenges, we present a three-stage hierarchical reinforcement learning framework for dynamic humanoid badminton. First, task-randomized motion augmentation expands sparse annotated hitting events into executable target-conditioned stroke variations, forming a continuous latent skill space. Second, a high-level planner outputs continuous latent skill codes to compose these skills online according to the observed shuttle state. Third, a context-conditioned adversarial regularizer encourages more natural planner-level skill usage while preserving return performance. When deployed on a real humanoid robot, our system achieves sustained multi-skill rallies with human players, including forehand, backhand, and highly dynamic jump returns. This is the first real-world humanoid racket-sport system to demonstrate multi-skill human--robot rallies including highly dynamic jump returns.

Problem

• humanoid robot
• badminton
• dynamic racket skills
• limited human motion data
• whole-body coordination

Innovation

• Hierarchical Reinforcement Learning
• Humanoid Robot
• Motion Augmentation
• Latent Skill Space
• Adversarial Regularizer