Dual wheel per axle set crab-steer concept

Dual wheel wouldn't give you too much advantage for the added weight and complexity. However, if you did want to go with it, you don't need a differential. All you would need is a miter gear at a 90 degree angle from the axel of the dual wheels. This would effectively give you perfect transmission of power (with only a slight loss of efficiency).

[Veselin Kolev]

I have pondered the though for quite a while as well. I see the following pros with the design i came up with: you can use small wheels (about 2" diameter) with a differential in between. Also, you can have a CIM mounted vertically, feeding power straight into the differential. This lets you have a VERY small module turning radius. If you think about it, having bevel gears on a single wheel swerve to have the motor vertical is pretty close in size to having two smaller wheels and a differential, with a vertical motor. Sure, a few more gears, but worth a serious thought.

[Andrew Schuetze]

Here is an image found in a FIRST mechanism library posted in another thread on off-season crab-steer projects.

[Ben Piecuch]

Dating back to 2001, Penn Robotics (Team 135) used a dual-wheel per axle, crab-style drive system. It was extremely fast, and very maneuverable. Those kids obviously had some stick time behind it, as it was a blast to watch.

Here's a pic I dug up from the 2001 KSC regional. Maybe someone from the team can tell us a little more about it?

FRC135, 2001