UPDATE: The Czech roboticists updated their video with labels and a slow motion sequence, making it easier to see how the robot juggles three, and then four, and finally five balls, and also how it drops one ball at the end.
Robots are especially good at juggling. This is not to say that juggling is a particularly easy problem to tackle, because it's not, but it's a fun excuse to design a robot to demonstrate precision control and high-speed object tracking. The robot in the video above, for example, was built by three masters students from the Department of Control Engineering at the Czech Technical University in Prague. It uses three linear motors, including one for each arm and a third for a central ball deployment system, along with two pivoting "hands" to catch and toss up to five balls at once. The feedback loop is closed using data from encoders built in the motors, and a high-speed camera helps to fine-tune the trajectories of the balls.
You don't actually need a high-speed vision system to juggle, though. We first met the "blind juggler" with it's fascinating passively adaptive single ball juggling capability back in 2009. Since then, it's gotten a fairly significant upgrade over the past year or so, which its designer, Philip Reist from ETH Zurich, presented last month at ICRA:
Remember, there's no sensing going on here. No cameras, no force sensors, no microphones, nothing at all. The robot is able to juggle without having any idea what the ball is doing, simply by virtue of the level of feedback control inherent in its design. It's really quite beautiful... You know, from a mechanical perspective. You can read about how this is possible here.
The Pendulum Juggler robot was presented in an ICRA paper unsurprisingly entitled "Design of the Pendulum Juggler," by Philipp Reist and Raffaello D’Andrea from the Institute for Dynamic Systems and Control, ETH Zurich, Switzerland.
[ Five Ball Juggler ] via [ Hack a Day ]
[ Blind Juggler ]
Evan Ackerman is a senior editor at IEEE Spectrum. Since 2007, he has written over 6,000 articles on robotics and technology. He has a degree in Martian geology and is excellent at playing bagpipes.