Juggling Robot Uses No Sensors

[filipemtx]

can it mills mess?

[MustyOldBooks]

Most of these clips make me go Wow!

This isn't one of them.

The "robot" could just have easily been made 100 years ago,

Compare 1939's Elektro with today's ASIMO, now that's progress.

[versiontwenty]

hahaha. It's official. A racism/political discussion pops up in every youtube video at one point in time.

Cool video!

[somedudefromnorway]

this is so unessesary... but wicked cool xD

[pandamanman89]

u cant call obama a liar u racist! im going to write a referral of disapproval unless u apologize.

[SDTricker]

What about my comment was considered making me a racist. Just because I don't like someones politics and because I think they are a liar, doesn't mean I am after their skin color. Get a brain.

[jeffsadowski]

Isn't mechanical feedback considered a sensor? Very interesting video. The feedback is measured correct?

[becomepostal]

This is almost a shift of paradigm to me. I've always believed that the efficiency of a robot was linked to the "quality" of the feedback loop: the more precise and fast is this loop, the more precise and fast is the movement of the robot. Now this experiment proves that with the worst quality possible of a loop (this being just no feedback at all) some stability can be achieved. I hope future systems will include this "mechanical" or "natural" stability and add it to usual feedback mechanisms.

[screenflicker1]

how did you do the last part without using sensors?

[becomepostal]

The web site explains it all. The surface is not plane, the surface of the paddle is incurved. The motion and the shape of the paddle have "just" been computed in order to obtain this stability. The paper is quite short, and the idea is strikingly so simple. Impressive.

[screenflicker1]

Thank you!

[CraigViewer]

amazing