And you thought you had cool omni-wheels

All I can say is go to this website and watch the video, seems like this design doesn’t have many of the drawbacks of the perpendicular rollers and allows you to strafe sideways without rotating the drive units


Airtrax wheel.jpg

Airtrax wheel.jpg

All I have to say is :eek:. Those look awesome.


Those are amazing, i love how smoothly it seems they drive sideways and also jut the size is overwhelming… Imagine a robot next year with 3’ diameter air trax omnis, they might run over the allotted budget but it would be a site to see…

/me wants to own that vehical in the demo…

If I remember correctly, these sorts of wheels are used inside the hold of Boeing cargo aircraft. They make for sliding the storage cans about very easy. I first saw them on a documentary two years ago or so and have since been trying to find a source.

Thanks, Matt.

ahhhh noooo! you found out about those! we have a few people working and making those for next years robot as we speak but i didnt say anything, nice catch :wink:

These omni’s are pretty awesome. That fork lift in the video might as well be hovering. :slight_smile:

Thanks for bringing this link back to my attention because I thought I had seen it before on these forums.


Yeah, your manuverable, but they aren’t really anything new. They’re called Mechanum wheels. And you can only exert 50% of your force in any one direction…

not true, you can utalise 100% in forward, reverse, strafeing left and right and rotating. only when it moves exactly in the 45% directions is it at 50% meaning its running on only 2 motors

this is true to the extent that tytus said but if you look at this compaired to a regular tanny where you can;t exert any force perpendicular to your wheels 50% is an increase. you also have probably less then 50% with a standard tank drive on any turn because the oposite directions of the wheels is canceling each other out while in this case the opposite direction of the wheels is working together. i will be interested to see if any team attempts to use these because they are MUCH more sophisticated then regular omni wheels.

also for you programers out there how would you deal with these?

forward = all 4 forward
back = all 4 back
left= front 2 forward, back 2 back
right front 2 back, back 2 forward

is this right?


Left = left-rear and right-front forward, left-front and right-rear backward
Right = left-rear and right-front backward, left-front and right-rear forward

Actually, the wheels exert 70% of the force in the forward/reverse direction. It’s the cosine of 45 degrees (1/sqrt(2)).

Look for several robots to use a similar idea next year. If the game requires it.

So do omnis require that each wheel be individually powered?

i will be interested to see if any team attempts to use these because they are MUCH more sophisticated then regular omni wheels.

well, some teams use 4 “standard” omniwheels at each corner at a 45 degree angle. this is exactly the same thing…basically…isnt it? each ‘small’ roller is contacting the ground at a 45 degree angle from the “forward”. and the controls would be the same too. theoretically if there were two robots. exactly the same in every aspect except robot A had “normal” omnis at a 45 degree angle at each corner, and B had these “new?” omnis, and both omnis had the same diameter, and same friction with ground, robots A and B will accelerate exactly the same, push exactly the same, and handle exactly the same.
the reason forklifts dont use normal omnis at a 45 is probably because its just harder to make a forklift that way. well, not harder but more like impractical… 45 degrees will take up more space than a straight wheel… but for our robot applications, unless you buy these, machining normal omnis will be much easier than these ones… and mounting wheels at 45 degrees on a robot is not that hard to do…
hope someone understands me…

yes. unless you’re using them as a caster/ ball transfer replacement

not quite…
when they are going “forward” the rollers on the wheels don’t roll, and the mechanum/ilon wheels act as a normal wheel would. this allows it to put 100% of its power forward, where a “traditional” holonomic platform (with four wheels) would only be able to put out approx. 71% of its full motor power.

ahhh, yes, you’re right… so where a *“traditional” holonomic platform * goes twice(well, for simplicity’s sake, allow me to say twice) the speed going at a 45 degree angle than going straight forward or sideways, this “new” one will go twice as fast forwards and daigonally than sideways… am i right? or am i just getting even more confused?

I’m not sure but I do know you got me more confused. I think I’ll just stick with plain 'ol wheels.

actually with the “new” one, it’ll be twice (again, for simplicity) as powerful going forward, and i believe sideways (<–subject to some debate), than it would going a a 45 degree angle.

yeah. me, i like round ones. how about you?


Let’s try some inductive reasoning:

Assume the mecanum wheels do put 100% of the power forward. We know the gearboxes can’t output more than 100%, so there can’t be any force sideways. Therefore, the robot could not strafe sideways.

However, we know that the robot can strafe sideways. Therefore our assumption is false.

The mecanum wheels do not put 100% of the power forward. They behave just like regular omnis at 45* angles. You get a maximum of 71% [1/sqrt(2) as Warren put it] power by moving forward, reverse, left, or right.