Mecanum or Swerve?

During the offseason this year, my team has discussed developing swerve drive, but after seeing a number of teams use mecanum with great sucess, I am torn between the two. Mecanum seems simpler and lighter, while swerve seems to give better control and more pushing power/resistance to being pushed. What are the experiences of the rest of the FRC world?

I think the choice between mecanum and swerve lies in where your robot is best suited to play (offense/midfield/defense).

Offense: Swerve
From watching quite a few matches, swerve seems alot better suited to collect balls and maneuver them into the goal.

Midfielf: Swerve
When your playing the midfield balls constantly keep entering, so you need to be able to quickly “pick up” balls from your zone. And if your in the midfield, theres a higher chance of you needing to go over the bump. So if you use mecanum it seems you need to drive straight to go over on mecanum, but you can go also sideways with swerve over the bump

Defense: Mecanum
Mecanum seems better here because its alot easier to get from goal to goal and just block the goal from other robots scoring.

I say go for the 4 or 6 legged robot. First because with 6 legs, 3 legs are ALWAYS touching the ground (Of course unless its falling off a cliff) minimum of 3 legs needs to be on the ground to be stable. Yes I do mean LEGS… Its more adaptable then wheeled vehicles. I mean look at the Protoss Colossus

http://starcraft.incgamers.com/gallery/data/517/Colossus.jpg

In My personal opinion, Swerve > Mecanum > Holonomic, and this goes for just about any game.

If you’ve got the resources, both Manufacturing and Programming, go with a swerve. Swerve drives have the traction advantage over all of the other omni-drives out there. Swerve drives can be complicated to build and drive though, so be careful (Though, you could look into the team 221 swerve modules). Personally, if I ever built and omni-drive for any game I would build a swerve.

Mecanum Drives are probably the most accessible form of omni-drive out there. You can use the kit frame, AM Mecanums, and some off the shelf transmission and have a reliable Mecanum drive. Mecanums lack the traction of a swerve or traditional drive so watch out. Mecanums can climb decently though, so this is an advantage over a Holonomic.

Holonomic Drives are interesting. They’re somewhere between a swerve and a mecanum in terms of build difficulty, due to the fact that the wheels must be mounted 90* from one another. (or 120*) They also can be a bit of a handful to drive, and they REALLY don’t like un-level playing fields. Holonomic drives also don’t push well.

We had looked into the 221 modules, but had decided that with our small team (12 members) it would be better to have an operable robot with a 6wd than a swerve that did nothing. We also only have 1 programmer and build in a basement, so our manufacturing abilities are very limited.

This was my primary reason for wanting to explore Mecanum…but it seems as though swerve offers better performance if it works right

If you want an omnidirectional drivetrain, it appears that swerve would be the best because it gives omnidirectional motion with better traction and without the power loss of mecanum.
However, our meccanum drivetrain has served us very well this year. When we decided we wanted to have omni directional capabilites they were our best option because our team is not capable of building a successful swerve in 2 weeks.
The mecanum does have (atleast) one advantage, swerve has a lag while the pods turn. One other might be weight. If you direct drive 6" mecanums then it should surely be lighter than 4 (or 6) swerve pods, chain, and sterring assemblies. This did not prove to be the case for us this year because we are chain driving 4 8" wheels with 4 toughboxes.

For a game like breakaway where pushing isn’t a factor unless your playing a defensive strategy, you really can’t go wrong with either.

No - at least not by our scouting methods. Mecanum robots are immediately removed from our defensive pic list, as are robots with slick wheels / omni wheels. A decently geared robot with traction wheels will have no problem moving a robot with slick/omni or mechanum out of their way and scoring.

Mechanum are a neat idea (as are omni wheels), but once you bring robot to robot contact into the equation, I’d much rather have robots with traction that won’t get pushed out of the way easily.

Regarding any lag time with swerve modules: If you have it programmed correctly, your swerve modules should never have to turn more than 90 degrees from any given point. When you keep in mind that a tank drive has to “turn” before it drives forward, there really is no lag in a well-done swerve drive when compared to a tank drive. When you take into account acceleration time, there really isn’t any lag in a swerve compared to a mechanum either.

You may want to note, however, that except for a few teams, most teams do NOT do swerve every year. Even teams that have done swerve, generally don’t repeat it much. That’s because it takes so much time, machining, programming etc to make it work well.

Swerve is that thing that every team has to try at least once. They try it, they may win a couple engineering awards with it: then most teams rarely do it again.

Here’s my thoughts on the matter. Full disclosure being that I have yet to assemble a swerve drive in anything other than concept sketches, and my mecanum experience is also somewhat limited.

Mecanum drives are overrated, heavily. Often, people decide that they need the ability to strafe, or they decide “maneuverability” is important, so they jump to the conclusion that they should build a mecanum drive. Mecanum drive is a very specific tradeoff. You exchange drive efficiency, resistance to defense, and a bit of speed for strafing. If teams spent more time prototyping bases to determine how well a 6WD does what they aim for, and how much better a mecanum does the same job, I imagine there would be a few less mecanums around. Well driven mecanum robots have seen success in FRC, though, especially this year. 2008 had several, this year there’s teams like 190, 230, 188 rocking the mecanum.

Swerve takes away some of the disadvantages of mecanum, while adding extreme complexity in design, build, and driving. Extensive preseason testing should be done with a swerve base before most teams consider the option of building one. From what little experience I have, it’s a completely different ball game.

So basically, The question shouldn’t be “mecanum or swerve”, it should be “what traits of a drivetrain are most important this year?”, and you should pick based on what is most important.

Extensive prototyping will always help you make more informed decisions on any part of the robot, including drivetrains. I would encourage any team considering anything new to do it!

A question: Do you remove half and half teams from the list (2 traction 2 omni)? Personally I rate them lower but not completely off.

(I also generally avoid mecanums with any pick, personal preference, but there are exceptions to every rule)

In my six years of competing in FIRST I don’t remember a single mecanum robot ever dominating a competition.

The powerhouse teams have built swerves, tank-drives, and more unusual designs like 71’s shuffler and the 148/217 nonadrives. But I can’t remember any of the top teams EVER using mecanum.

I can only think of a few cases in which a mecanum robot even won a competition.

Why is this?

Is it simply two difficult to write controllable code?
Are they just too inefficient?
Not enough lateral traction?
Or has the mecanum drive simply never been perfected?

I’ve seen those mecanum forklifts, I know what the technology is capable of…

I have to disagree. This is the third year my team has done mecanum drive, and we were able to play defense well. Defense doesn’t necessarily have anything to do with pushing if you’re maneuverable enough.

If you want to find an holonomic drive that also has pushing power, you might try over the summer looking into mecanums that have a mechanism which locks all the mech rollers.

What about a 2 wheel drive? It can turn very tight and stuff

Another drive system you could research is a **kicker drive **system, also known as a slide drive system. It’s basically a 4WD setup with omni wheels, with a fifth powered omni wheel mounted in the center of the chassis, that allows the robot to strafe.

We’ve never built one for FRC, but many for VRC. I’d love to prototype one that would be FRC-ready though.

-Nick

But you can strife with regular omni.

While true, a robot that can push is almost always more defensively useful than a robot that basically can’t. I’m a pretty strong advocate of “defense != pushing”, but I’d still rather do it with an XWD.

Except you lose significant efficiency in most directions, when a slide drive lets you make the primary direction of travel more efficient than the others. For example, if you have 5 CIMs, you can have a 100% efficient 4 cim forward drivetrain, while only using CIM 5 to go sideways. With a holonomic drive, you are stuck with 4 cims and the inherent efficiency losses that come with them.

Mecanum and Holonomic involve great loses of power that drives like Swerve tend to provide for you. But swerve drives require weight and a certain skill level. A lot of teams this year saw that they had the need for the maneuverability, but not necessarily the experience/resources for a swerve, and mecanums offered an easy solution.

I would disagree with that. A great offensive robot can be easily shut down if it can’t push back.

Darn, I don’t think I can give you anymore rep for a while.

This is basically what the 148/217 drive is based on. I’d love to play with one at the FRC scale aswell.

While this is true, a “Slide Drive” has a few advantages over a regular omni drive. Theoretically, it’s easier to drive, and because you aren’t always relying on the omni wheels slipping it’ll have more pushing power.

I completely agree. Along those lines, one thing I simply don’t understand is why so many teams build a mecanum drive but never or rarely strafe. I just don’t get it. Why would a team spend the time, money and weight on a mecanum drive and basically have a 4 wheel drive that can’t push and can be pushed around? It is one very quick way to get on my ‘no pick list’. Not only is it not effective at all but it raises a lot of questions of a team’s competency.

If you decide you need mecanum for the game (2008 is the only game I could consider using one), but sure take full advantage to all of its capibilities.

1 Like

weight of an 8" AM mecanum = weight of one of our swerve modules, without making them anorexic :stuck_out_tongue:

The weight of a swerve system will come down iteration by iteration, but the same goes for a solid tank system which can go much lower.

Chris - that’s exactly what we did. With the proliferation of robots using omni this year we didn’t eliminate the traction / omni combos. They simply got lowered in the ratings.