Quadrocopter Ball Juggling, ETH Zurich

Ball juggling experiments with quadrotors in the ETH Flying Machine Arena

By Mark Müller, Sergei Lupashin and Raffaello D'Andrea

http://www.flyingmachinearena.org
IDSC, ETH Zürich, Switzerland

Common question: This is not human-piloted (please see the overview vid). The vehicles/ball are tracked by an overhead motion capture system and controlled by a pair of computers.

• Category

  Science & Technology

[Skullzero1]

i like how mad it gets when it misses

[max carter]

Skynet

[DevilNymph]

It's not. I saw them.

[Donghyung Kim]

It's really nice!

[MsYaDong]

Fake vdo 

[Julio Wong]

MQ-27 Dragonfire Inbound

[SamuelLPrime]

keep telling that to yourself, but don't come back to us crying when it travels back in time trying to kill you because you're lying about John Connor's location

[samlovemetal]

Soon these will have spinning razors attached to them and will ambush us when we enter the sewers.

[Dian Istining]

Hello! Thank you for this useful video. By the way, I hear a lot of people keep on talking about Tennisolanox System (search on google), but I'm not sure if it's good. Have you tried using this tennis program called Tennisolanox System? I've heard several unbelivable things about it.

[Alias Cummins]

But can it do a Mills Mess?

[Mark Richardson]

Ummm that doesn't make any sense. If each robot had the ability to throw without variation they would a) >>never<< miss or b) at least go for significantly longer than what I see here. Since they are reacting to flight path path based on algorithms that approximate velocity, arc to get them...close enough, they are prone to errors. Two humans would be able to play catch a lot longer than 10-15 passes. Human and quad would play longer as our catch algorithm is superior e.g. baseball players