pic: Cyborg Swerve w/4 775 Pro Drive Motors



This isn’t quite finished because the bottom “caster box” still has to be modified and changed to a belted turning stage, but its close. I made this for a year-end project, and I wanted a sub-30 lb swerve with no water jetted parts except the frame. I got it down to about 6 lb a module and 7 lb for the frame. It’s lightened 4 x 1, I may have to beef up the truss-like pattern. It is geared for around 13 ft/s with almost no pushing power, similar to Bomb Squad’s style of Swerve. If you are trying to push everyone around with Swerve, you are kinda missing the point. That said, I am working on a 2 speed version involving 8 motors as well as both a 12 motor one-speed and a two speed. We went with a 12 motor one speed this year on our Steamworks robot.

Did you see match 120 Newton from 2016? I thought swerve was supposed the best of both worlds when it came to speed, push power, and maneuverability?

16 got pinned into a corner with no pin count and still got out of it by flipping the defender over. Seemed like swerve won there.

Nice design Sam! Keep up all the great work you and 4256 are doing!

I’m not sure how your post has anything to do with this design.

It is the best of both worlds. Few robots have enough power to push a tank 6 wheel pneumatic-tire robot sideways, because pneumatic tires have an extremely high amount of friction with the carpet.

Are you suggesting something else? How does your comment apply to the original post?

@Cothron

I’m just trying to explain that swerve should have more pushing power, and that it shouldn’t just be a fancy mecanum or holonomic drive.

OP’s quote:
“If you are trying to push everyone around with Swerve, you are kinda missing the point.”

Just that 16’s drive train is a total tank… that swerves :stuck_out_tongue:

So to clarify as a member of 16, our swerve and many others do not have great pushing power. It’s one of the trade offs to having excellent maneuverability. It’s not the best of both worlds it’s more one or the other. With high acceleration and being able to out maneuver another robot, you never need pushing power. Not even when playing defense as we showed in Midwest elims. That being said I don’t feel like any drivetrain could get out of the pin we faced in match 120 as we had a penalty zone, a large tower, and a sideways tank robot and a wall on all sides. We got out because they happened to drive on top of us. Just to clarify some misconceptions of swerve… it’s not about pushing power it’s about maneuverability…

On the main topic of this thread, great design! Always love to see swerve drives

I think a swerve can still have more traction than a mecanum or omni drive train, it just doesn’t win a lot of pushing matches.

With that in mind, what advantages would you say this has over a mecanum or omni drive? Also, are you geared for 13fps free speed or adjusted?

I see a lot of comments on ChiefDelphi discussing the pushing power of swerve versus other drive trains. Let me try to clarify it:

Given the same number of motors, type of motors, gear ratio, size of wheel, and coefficient of friction, a swerve drive will have more pushing power than a mecanum drive. This is a bit of a simplification, but with those wheels, the force vector doesn’t necessarily point in the direction of travel of the robot.

Let’s say you want to drive straight ahead. The wheels are all going to give a force vector off at a 45 degree angle to that. When you add up the four force vectors of the wheels, you get to travel straight ahead, but the magnitude of the force vector is a lot less than if the force vectors were all pointing in the same direction and added up. Consider if one wheel in a normal drive train gives off force X, the total force vector is 4x. With a mechanum, you end up with a right triangle, with each leg being 2x, so the hypotenuse (the direction you travel) is sqrt((2x)^2 + (2x)^2), or about 2.8*x. So you get about 30% less pushing power.

With a swerve drive, no mater which direction you want to travel, you point all the wheels in that direction and go, so you get 4x.

That being said, the “all things being equal” caveat from above is rarely true. For example, with a west coast drive,a lot of teams use 6 CIM motors, but a swerve rarely uses more than 4. So you need to be careful about generalizations about which drive train has more pushing power.

Of course, this is missing the point a bit with a swerve. Typically teams build swerve with a game plan of being fast and maneuverable and avoiding pushing matches all together. That has been my team’s strategy with swerve and we’ve been pretty successful with it, although we play in an admittedly defense-light area of the country (NYC).

On the other hand, great maneuverability without a crazy amount of pushing power can be used to deploy solid defense in a certain games. Take, for example, our matchup against the 1114-27-20 in 2016 Carver Semifinals. We (3419) used our maneuverability to make it extremely hard for them to score boulders: https://www.thebluealliance.com/match/2016carv_qf4m1

As for the original poster’s design: I like the simplicity and lightness of it. One word of caution, though: With a swerve drive, you can make very sharp turns. If your center of gravity is high, this can lead to a robot tipping over. With both the turn motor and drive motor pointing up like that, you might have a high center of gravity, of course depending on the rest of your robot. You might want to try turning those motors upside down and getting them closer to the ground. That introduces a lot of design challenges of their own, but I can speak from experience that it is a good direction to try to go in. For reference, our 2017 swerve had a COG about 4 inches off the ground (once the battery and bumpers were on) and never left the ground, even with some crazy driving. Our 2016 swerve, which had big pnuematic wheels and had motors up higher for ground clearance, had a COG about 9" off the ground, and would occasionally get up on 2 wheels (but fortunately never flipped). In a game with a more open field where we could get up to speed and turn quickly, I think that 2016 bot would have had flipping problems.

A swerve may not be able to push but it holds the advantage of being smoother motions and it’s not pushed as easily. It is very hard to push a swerve actually because you can lock the wheels (turn the towards the center of the robot) or simply spin off of a push. Yes you can do this with mecanum or omni but you get pushed more. In semifinals at midwest we used swerve to play defense on mecanum and pushed them around the airship.

I’ll call your match 120 Newton and raise ya almost any match involving 2767 on Daly and Einstein. This may be pretty arrogant, but our driver used the heck out of swerve. Very fun to watch.

OP’s quote is spot on:
“If you are trying to push everyone around with Swerve, you are kinda missing the point.”

That being said, we should remember there are other threads discussing the pros and cons of swerve.

Back to Cyborg Swerve, please check out the Stryke Force swerve white paper and GrabCAD has the CAD. You might shave off some more weight. It’s looking good. Keep up the good work!

@Mark

Thanks! Your robot has inspired the version of swerve I am working on presently, as the 3d printing and use of plastics is something I am looking into since our team might be purchasing a new 3D printer to help with it. I also love how you guys use the belt on a 1:1 pulley ratio for steering.