Humanoid robot learns badminton skills from limited human motion data
Humanoid Badminton: Learning Dynamic Racket Skills from Limited Human Motion Data
Robotics
Summary
Playing badminton with a humanoid robot is very challenging because it needs to react quickly and hit the shuttle accurately. The authors created a new system that helps a robot learn how to play badminton by using a small set of human motion examples and smart learning techniques. Their approach breaks down the learning into stages, letting the robot practice a variety of strokes and decide how to play based on the shuttle’s position. When tested, the robot could play rallies with people, including tricky moves like jumping smashes.
What this means in practice
- •For robot developers: Create humanoid robots capable of playing fast-paced racket sports by learning diverse stroke skills from limited motion data.
- •For robot sports trainers: Train and test humanoid robots to perform complex, dynamic multi-skill rallies including jump returns against humans in real-world settings.
Authors
Jingzhi Cui, Zhexiong Wang, Bangjie Xu, Pengyu Zhao, Youyuan Li, Zhi Su, Peng Ren, Mengdi Xu, Chao Yu, Yi Wu, Luyang Wang, Zhongyu Li
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.