Quadruped robot plays badminton with you using AI

https://static.foxnews.com/foxnews.com/content/uploads/2025/06/5-quadruped-robot-plays-badminton-with-you-using-ai.jpg

At ETH Zurich’s Robotic Systems Lab, engineers have created ANYmal-D, a four-legged robot that can play badminton with people. 

This project brings together robotics, artificial intelligence and sports, showing how advanced robots can take on dynamic, fast-paced games. 

ANYmal-D’s design and abilities are opening up new possibilities for human-robot collaboration in sports and beyond.

Sign up for my FREE CyberGuy Report
Get my best tech tips, urgent security alerts, and exclusive deals delivered straight to your inbox. Plus, you’ll get instant access to my Ultimate Scam Survival Guide — free when you join.

How does ANYmal-D play badminton with humans?

Badminton is a game that requires quick footwork, fast reactions, and precise hand-eye coordination. To give a robot a chance on the court, the ETH Zurich team equipped ANYmal-D with four legs for stability and agility, a dynamic arm to swing the racket, and a stereo camera to track the shuttlecock. The robot uses a reinforcement learning-based controller, which allows it to predict and react to the shuttlecock’s movement in real-time. ANYmal-D can move around the court, adjust its posture, and time its swings, keeping rallies going with human players for up to 10 shots.

NO TENNIS PARTNER? NO WORRIES WITH THIS AI ROBOT

The technology behind ANYmal-D’s badminton skills

ANYmal-D’s stereo camera serves as its eyes, constantly monitoring the shuttlecock. The robot uses a “perception noise model” to compare what it sees with data from its training, helping it track the shuttlecock even when it moves unpredictably. The robot can pitch its body to keep the shuttlecock in view, mimicking how a human player might lean in for a tricky shot.

HUMANPLUS ROBOT CAN GO FROM PLAYING THE PIANO TO PING-PONG TO BOXING

Unified reinforcement learning for whole-body control

Coordinating legs and an arm is tough for any robot. The ETH Zurich team developed a unified control policy using reinforcement learning, allowing ANYmal-D to move and swing as a coordinated whole. This system was trained in simulation, so the robot learned how to handle a wide range of shots and situations before stepping onto a real court.

AI HUMANOID ROBOT LEARNS TO MIMIC HUMAN EMOTIONS AND BEHAVIOR

Hardware integration: What’s inside the robot?

ANYmal-D combines a sturdy quadrupedal base with the DynaArm, and its racket is set at a 45-degree angle for effective striking. The robot’s state estimation runs at 400 Hz, the control policy updates at 100 Hz, and the perception system operates at 60 Hz. All of this runs on a Jetson AGX Orin module, making the robot responsive and ready for action.

GET A FREE SCAN TO FIND OUT IF YOUR PERSONAL INFORMATION IS ALREADY OUT ON THE WEB

Challenges of playing badminton with a robot

Getting the robot’s legs and arm to work together smoothly is a major challenge. Most robots handle these tasks separately, but this limits agility. By combining locomotion and arm control into a single system, ANYmal-D can adjust its posture and gait based on the shuttlecock’s path, moving more like a human player.

AI TENNIS ROBOT COACH BRINGS PROFESSIONAL TRAINING TO PLAYERS

Active perception: How ANYmal-D sees the game

Robots don’t have human eyes, so their cameras can struggle with frame rates and field of view. ANYmal-D’s perception-aware controller keeps its camera moving smoothly, always tracking the shuttlecock. The perception noise model helps bridge the gap between simulation and real matches, making the robot more reliable during games.

WHAT IS ARTIFICIAL INTELLIGENCE (AI)?

Real-world deployment: Bringing the robot to the court

Bringing ANYmal-D from the lab to the badminton court meant dealing with practical issues like power limits and communication delays. Despite these challenges, the robot managed to keep up with human players, responding to different shot speeds and landing positions, and maintaining rallies that showcased its adaptability and skill.

ANYmal-D’s badminton performance: What did the tests show?

In collaborative games with amateur players, ANYmal-D tracked, intercepted, and returned shuttlecocks with impressive consistency. On average, it took about 0.357 seconds to process the shuttlecock’s trajectory after a human hit, leaving just over half a second to get into position and make the shot. While it didn’t return every shot, the robot’s ability to maintain rallies and adjust to the pace of the game highlights how far robotics has come in dynamic sports scenarios.

SUBSCRIBE TO KURT’S YOUTUBE CHANNEL FOR QUICK VIDEO TIPS ON HOW TO WORK ALL OF YOUR TECH DEVICES

Kurt’s key takeaways

ANYmal-D really shows how far robotics has come, especially when it comes to working alongside people in fast-paced activities like badminton. It’s interesting to see a robot not just keeping up on the court, but actually rallying with human players and adapting to the game as it unfolds. As these technologies keep improving, it’s easy to picture more robots joining us in all sorts of sports and activities, making play and teamwork even more fun for everyone.

CLICK HERE TO GET THE FOX NEWS APP

Would you be curious to play a match against a robot like ANYmal-D, or do you think nothing can replace the experience of playing badminton with another human? Let us know by writing us at Cyberguy.com/Contact.

For more of my tech tips and security alerts, subscribe to my free CyberGuy Report Newsletter by heading to Cyberguy.com/Newsletter.

Ask Kurt a question or let us know what stories you’d like us to cover

Follow Kurt on his social channels

Answers to the most asked CyberGuy questions:

New from Kurt:

Copyright 2025 CyberGuy.com.  All rights reserved.