FoamBot builds a quadruped robot

FoamBot built of CKBot robot modules builds a quadruped robot which crawls away. Details of quadruped robot synthesis task from IROS 2011 presentation of Revzen, Bhoite, Macasieb and Yim.

• Category

  Science & Technology

[timerider4]

Couldn't you use Mecanum wheels instead of tilting each wheel?

[shrevz]

We could. Except than Mecanum wheels are expensive, and sensitive to particles getting jammed in them. The wheels we used are made of foam (different from the robot; it is the foam we'll use for the next generation of robots).

By using two CKBots for each wheel, we made a platform which -- in principle at least -- could be made out of the same materials as those we use for synthesis.

[randeDANDE]

Skynet.... WE'RE ALL FUCKED

[coldwindinmyhair]

1:58... FUCK!!! Lost my sh*t...

[theamazingsomethin]

It looks like a beautiful sea turtle c:

[nachoijp]

it made me remember of the mutant dog in the movie the fly... gross and creepy and scary D:

[EDbringer]

is this battle zone

[aznbleachmaster]

the moment the quadruped robot started moving i was creeped out, it looked like a freaky spider dying or something

[shrevz]

the print volumes of 3D printers are rather small, and the kinect isn't very accurate.

As in all things technical, the devil is in the details. But I like the way you think :-)

There isn't anything wrong with adding the kitchen sink, provided it is useful; in research our limitation is time and skilled labour -- not trying to make the cheapest, most streamlined product.

[aphetica]

Also I imagine a proprietary design would be a lot less bloated than a 3D printer on a mobile platform with a Kinect strapped to it. Eventually I'll be suggesting the addition of a kitchen sink as well. Carry on. :D

[aphetica]

Perhaps if it were combined with a Kinect for object detection? I'm not a software engineer though, I imagine there would be a lot of coding involved in getting all these things to work with eachother.

[shrevz]

@MarlosZappa, thank you -- you got the point of what we are doing exactly right. We're trying to create a new concept for "Unanticipated Challenge" mission profiles: missions where you have send your robot out there, but you don't know (yet) exactly what you need it to do. Planetary exploration is one domain where this could really shine -- if NASA has the money for it.