[Discuss] Cool 5 minute talk about a Ball bot.

Wed Nov 23 11:03:48 PST 2011

Hey Robert!,

I just sent a link for that to the Maker Space discuss list (CCed 
above). ;-)

My video player happened to freeze at just the right frame for me to see 
that there are three largish ball bearings on the bottom pressed in with 
spring steel. At the top of the ball it only contacts the three driving 
wheels. The three axis are obviously at 120 degrees, I wonder if the 
down tilt is also 120? ;-)

Yes it does look like a fun project doesn’t it? Reminds me of the home 
built seguways that this guy from hackers was making on the cheep. He 
would have parts for sure. Hmmmm....

So you need at least a good six axis accelerometer, Three motors and 
drivers, Controller and batteries! Oh, and those fancy sliding wheels 
would be handy too. :-D


Robert Kelley wrote:
> Check out this amazing TED Talk:
> Péter Fankhauser: Meet Rezero, the dancing ballbot
> http://www.ted.com/talks/peter_fankhauser_meet_rezero_the_dancing_ballbot.html
> I like the human scale of it, though smaller ones have been built.
> Larger ones could be ridden.
> The devil is in the details, I suppose.  Do you suppose the ball cage
> is lined with rollers, or Teflon?
> The ball is rigid, rubber coated, like the ball of a mechanical mouse.
> Do you suppose the three axles are orthogonal?
> http://www.mathworks.com/matlabcentral/fileexchange/23931
> http://www.msl.ri.cmu.edu/publications/pdfs/isrr05.pdf
> http://www.msl.ri.cmu.edu/publications/pdfs/ballbot_ICRA06_web.pdf
> http://www.msl.ri.cmu.edu/publications/pdfs/Ballbot_TrajectoryPlanning_ICRA_2009.pdf
> http://www.msl.ri.cmu.edu/publications/pdfs/BallbotTransition_ICRA._2009.pdf
> http://www.msl.ri.cmu.edu/publications/pdfs/BallbotPhysicalInteraction_HRI_2009.pdf
> http://www.msl.ri.cmu.edu/publications/pdfs/Ballbot_ShapePlanner_RSS.pdf
> Got any pieces?

Mr. Clif wrote:
> Here is a quick demo about a bot that balances on a ball.
> http://www.ted.com/talks/peter_fankhauser_meet_rezero_the_dancing_ballbot.html 
> Clif

More information about the Discuss mailing list