I was wondering of a 3 motor swerve design would be possible? I remember seeing a concept drawing some ware but the only 3 motor holonomic drive I can find is the three wheel omni drive.
3 motor? That would be tricky. You could do 1 modules, and 1 driven Omni wheel, but that’s not recommended. Now 3 wheels? Yes. A two wheel swerve can be done too, if you can figure out a way to support the 3rd side.
Here is the classic example of a 3 wheel swerve.
There is a such thing as a 3 motor mecanum drive using a complex differential setup, however with swerve, the modules need to independently rotate and spin the wheels which I do not know of a true swerve setup that can do this. A 3 motor crab drive without independently rotating modules is possible but wouldn’t be any good.
Three wheel? Yes. The latest one I’ve been involved with was 323’s 2017 off season bot, Ankle Biter.
Three motors total? Maybe. Should be doable but may be complicated to get all the controllable degrees of freedom.
found it,
here
I’m going to play a cheat card here.
One of our FTC teams built a swerve drive that used SERVOS for the steering instead of motors, so it was technically a 4 wheel, 4 motor (+4 servo) swerve drive. These modules could easily be used in a 3 wheel configuration resulting in a 3 motor swerve drive.
Hello! Yes, I drew this. I don’t think it’s completely workable, but I am interested in thinking about it.
Here is some noodling I did in Jan 2021 on the whole class of “underactuated” differential swerves. At the time I had a fairly naiive understanding of swerve kinematics so would definitely take a different approach now.
Then here’s what I said in Mar about it. Again, I’m not sure this still holds, but that’s what I thought at the time.

Here is the classic example of a 3 wheel swerve.
Tumbleweed is also an example of a 2DOF (two-degree of freedom) drive base. In theory it could have operated with just two motors, one to drive all the wheels and one to steer them all. In practice (where we play), it used one steering motor (visible in the 1:33 still frame you included) and SIX motors to drive the wheels. The drive motors are all combined into one massive gearbox: four CIMS and two FPs with planetary reducers that made them behave like CIMS.
Two degrees of freedom, three wheels, seven motors. One quick little demon.
@anon91315253 if you’re listening, please correct me if I got that wrong. Also please comment on how awesome Tumbleweed 2.0 could be using the FRC-legal motors, power transmission components, and control system components we have today.
Team 179 did one for their 2019 offseason robot. It was pretty interesting to see.

Also please comment on how awesome Tumbleweed 2.0 could be using the FRC-legal motors, power transmission components, and control system components we have today.
It’s missing 5 sides (and has 4 modules) but I think that’s what they were going for 2021. Looked pretty cool.
That cute little thing that appears between 2:17 and 2:28 …?
That too. Looked unstable (which I imagine is a big problem with 3 wheel “swerves” in general).
I’m curious what was learned with that prototype. I assume too unstable to go fast?
We build a small 24"x24" (4 module) swerve last year as an off season project. With the grippy tread on it, i could flip it while driving. I can only imagine how unstable a 3 wheel swerve would be using modern motors.
Can confirm, 3 wheel swerve is a bit unstable while driving
Pure joy!
That’s a one drive swerve.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.