Swerve Concept using Kitbot

Hello CD, I wanted to share an idea that I thought would be interesting. Feel free to let me know if this idea won’t work, I’m not sure whether this idea has any potential but I thought it was interesting so I felt I should share.


Kitbot Swerve.pdf (169.4 KB)


I’ve always thought that an “active omni wheel” would be cool, but I think you’d have a number of difficulties fielding it.

  • The crossed helical gears are inefficient, so you’ll have to deal with a lot of friction, heating and wear.
  • Since every roller needs a geartrain, the wheels will end up having a lot of moving pieces inside.
  • Even if you get the mechanics to work, the little ~0.75" rollers won’t have much grip on the carpet pile, making them drive a little unpredictably I think.

James Bruton on Youtube attempted to build a couple of powered-omnis recently. You might be interested in the challenges he faced:


Thanks so much!!! I am also somewhat worried about the grip on the carpet. I was thinking of putting some kind of grippy material on the outer wheels to try to negate some of that slip. Do you think that could be a potential solution?

1 Like

I have zero experience, so don’t trust my word for it. But it seems to me that the rollers are small enough that they might stir up the individual carpet piles more than they will want to drive across them. You might need something with bigger diameter rollers


I’ll have to test it. I’m making a spreadsheet with testing configurations so I’ll add wheel diameter to the list. I appreciate the feedback!!!

1 Like

Update 1 - July 28th

Added bigger diameter wheels.

Currently working on gearbox to mount to AM-14U4 kitbot

1 Like

The idea is cool, but pretty impractical as far as FRC goes. In exchange for not having to design a custom chassis (aka a few tubes and gussets), you now have to field an entirely custom drivetrain with hundreds of delicate moving parts. Really, moving sideways isn’t for every team. I wouldn’t recommend any team go from KoP chassis directly to traditional swerve without experimenting with a custom WCD in between. And swerve at this point has a lot of tried and tested implementations to copy with example code readily available. If you can’t build a custom chassis, you probably shouldn’t be building a swerve at all, and you certainly shouldn’t be designing a sensitive, untested custom drivetrain.


That makes sense. I’m mainly focused on the development of an idea that I thought was interesting. Sort of to just take an idea and run with it. I thought it might have some kind of niche for FRC if it ever got reliable enough to be worth doing an offseason project on (because right now I’m working on this as a home project rather than a traditional offseason project.) at this point I’m reconsidering whether or not it’s worth pursuing further, any thoughts?

1 Like