Mecanum or Swerve?

[Andrew Schreiber]

Originally Posted by ajlapp

This is the reason we wanted to introduce commercial swerve components to the market. There is always lots of discussion amongst the community about how to keep FIRST competitive....

I disagree with making the game easier or artificially leveling the playing field through other means.

I'd much rather see a large portion of teams building high quality robots by using off-the-shelf items when necessary to bolster an area of their team where they may not have the expertise.

That said, swerve is tough to pull off even if you start with pre-made transmissions. Cyber Blue is hosting a swerve discussion this week in Atlanta about the months they spent leading up to competition developing the technology.

The hard part is not the hardware. The devil is in the software. I admit, after 7 years of FRC and countless off season projects I am still intimidated by the thought of having to code a functional swerve drive during the build season.

Originally Posted by joek

but, most tank drives use only 2 drive motors, whereas omni and mechanum ALWAYS use 4, so there is pore power being delivered, despite the 30% decrease, we have mechanum, and were able to push around tank drives like it was notheing

What Tank Drive bots are you looking at? Ive seen some using 6 drive motors and most using 4.

[kevin.li.rit]

Originally Posted by joek

but, most tank drives use only 2 drive motors, whereas omni and mechanum ALWAYS use 4, so there is pore power being delivered, despite the 30% decrease, we have mechanum, and were able to push around tank drives like it was notheing

While I haven't seen any first omni robots using less than 4 wheels. There are omni-wheeled robots with 3 wheels. In the past we've considered using a 3 wheel setup.

[Tom Line]

Originally Posted by thefro526

I thought a bit more on this Subject last night, and came to the realization that Swerve Drives are going to become more and more common in FRC.

I would tend to disagree. Full mechanical designs for swerves (or the pictures to allow you to easily reverse engineer them) have been around for years. I believe 217 posted a cutaway of their serve from 5 or more years ago. Those teams are more than willing to lend help to anyone who needs it as well.

However, you still see a lot of teams who have already DONE swerve not using them going forward. Nearly every big name team has done swerve several times, yet most do not pick it as their drivetrain of choice.

Why?

1. Complexity. Both mechanical and programmatically.

2. Advantage gained. I would put forward that last year and this year are two of the biggest swerve-advantaged games so far. Yet if you look at the top tier of teams in OPR, the proportion of swerve teams comes no where near 50%. Swerve has a huge number of trade offs, and the advantages are actually questionable nearly every year. You'd have a hard time telling me that 67, 469, and 1114's lack of swerve this year is hurting them. I bet the choice not to do swerve HELPED them in a big way during the build season - it was that much more time to work on their ball handling systems.

3. Cost. Whether you purchase a turnkey system like the 221 one, or whether you build your own, there is no small cost in engineering, materials, and machining.

I keep pointing out and I will continue to point out that Swerve is cool, but most teams that build it realize they'd rather be spending their time solving the game rather than solving a drivetrain, and a 6-wheeled or 8-wheeled drivetrain will perform nearly as well in most applications, and better in many.

[Chris is me]

Originally Posted by joek

but, most tank drives use only 2 drive motors, whereas omni and mechanum ALWAYS use 4, so there is pore power being delivered, despite the 30% decrease, we have mechanum, and were able to push around tank drives like it was notheing

Very few respectable tank drives use only 2 motors, and very few use relatively low traction wheels like KoP wheels that your mecanum drive could push. I guess if they used the wheels in the KoP you could push them around, but that's hardly the level of competition you have to play defense against.

A mecanum drive is limited by traction faster than a tank drive, thus, it will have less "pushing power" in most situations.

[Andrew Schreiber]

Originally Posted by Tom Line

I would tend to disagree. Full mechanical designs for swerves (or the pictures to allow you to easily reverse engineer them) have been around for years. I believe 217 posted a cutaway of their serve from 5 or more years ago. Those teams are more than willing to lend help to anyone who needs it as well.

However, you still see a lot of teams who have already DONE swerve not using them going forward. Nearly every big name team has done swerve several times, yet most do not pick it as their drivetrain of choice.

Why?

1. Complexity. Both mechanical and programmatically.

2. Advantage gained. I would put forward that last year and this year are two of the biggest swerve-advantaged games so far. Yet if you look at the top tier of teams in OPR, the proportion of swerve teams comes no where near 50%. Swerve has a huge number of trade offs, and the advantages are actually questionable nearly every year. You'd have a hard time telling me that 67, 469, and 1114's lack of swerve this year is hurting them. I bet the choice not to do swerve HELPED them in a big way during the build season - it was that much more time to work on their ball handling systems.

3. Cost. Whether you purchase a turnkey system like the 221 one, or whether you build your own, there is no small cost in engineering, materials, and machining.

I keep pointing out and I will continue to point out that Swerve is cool, but most teams that build it realize they'd rather be spending their time solving the game rather than solving a drivetrain, and a 6-wheeled or 8-wheeled drivetrain will perform nearly as well in most applications, and better in many.

What is of most interest to me:

217: Swerved in 2003. Hasn't done it again. (Nonadrive is the closest they've come)
67: Swerved in 2005. Hasn't gone back. (This was a cool flop bot swerve though)
1114: Swerved in 2004. Hasn't gone back.
33: Swerved in 2005, switched to different drive train halfway through the season. Again in 2009: Maybe?
71: has been using swerve since 2005. (Of particular interest: Hasn't won a Championship since 2004)
111: even I don't know how long.
68: 2008,2009,2010. This year their swerve drive was too much of a technical undertaking for them.
148: Swerved in 2008. Hasn't gone back since.

So many veteran teams have tried swerve and gone back to traditional 6wd/8wd machines. 2009 was not a FULL swerve, only 2 wheels actuated as far as I can recall.

67, 6wd the last couple years. 8wd this year.
1114, 6wd. 8wd this year
33, 6wd.
217, 6wd (excepting this year)

I know correlation does not imply causation but I have a hunch that there is a reason why none of these teams have gone swerving again. It may be a cool thing, and definitely a design every team should have in their arsenal, but in most cases it is not the most efficient design.

[thefro526]

Originally Posted by Tom Line

*snip* I would tend to disagree. Full mechanical designs for swerves (or the pictures to allow you to easily reverse engineer them) have been around for years. I believe 217 posted a cutaway of their serve from 5 or more years ago. Those teams are more than willing to lend help to anyone who needs it as well.

However, you still see a lot of teams who have already DONE swerve not using them going forward. Nearly every big name team has done swerve several times, yet most do not pick it as their drivetrain of choice. *snip*

Even though full designs for swerves had been posted for years, most teams lacked the resources to build one. The same thing was true for Mecanum wheels until 2007. There were a few teams who made their own in 2005 and 2006 and posted designs and research on them, but most teams lacked the ability to make the wheels and chassis. Then, in 2007 AM came out with their own line of Mecanum wheels and a huge number of teams began using them. I think the same thing will happen with Swerve Drives in the next few years with the introduction of the Team 221 Swerve Modules.

[Ether]

Originally Posted by Chris is me

A mecanum drive is limited by traction faster than a tank drive, thus, it will have less "pushing power" in most situations.

Yes! In the pure forward and reverse directions, a mecanum wheel's pushing force is limited by traction, not by a vector reduction in the amount of available force.

Even if a mecanum wheel's rollers were made of the exact same material as a standard wheel of the same wheel diameter, the mecanum would lose traction before the standard wheel would.

This is because the reaction force of the floor (carpet) on the mecanum wheel's roller is larger than the reaction force on a standard wheel, given the same driving torque on the wheel.


~

[Tom Line]

Originally Posted by Andrew Schreiber

What is of most interest to me:

217: Swerved in 2003. Hasn't done it again. (Nonadrive is the closest they've come)
67: Swerved in 2005. Hasn't gone back. (This was a cool flop bot swerve though)
1114: Swerved in 2004. Hasn't gone back.
33: Swerved in 2005, switched to different drive train halfway through the season. Again in 2009: Maybe?
71: has been using swerve since 2005. (Of particular interest: Hasn't won a Championship since 2004)
111: even I don't know how long.
68: 2008,2009,2010. This year their swerve drive was too much of a technical undertaking for them.
148: Swerved in 2008. Hasn't gone back since.

So many veteran teams have tried swerve and gone back to traditional 6wd/8wd machines. 2009 was not a FULL swerve, only 2 wheels actuated as far as I can recall.

67, 6wd the last couple years. 8wd this year.
1114, 6wd. 8wd this year
33, 6wd.
217, 6wd (excepting this year)

I know correlation does not imply causation but I have a hunch that there is a reason why none of these teams have gone swerving again. It may be a cool thing, and definitely a design every team should have in their arsenal, but in most cases it is not the most efficient design.

One very slight correction. 67's 2008 overdrive bot actually had rear-wheel-only pseudo swerve. If you watch their matches that year, it will suddenly dawn on your how they can turn their corners so quickly.

[Chris is me]

Originally Posted by Tom Line

One very slight correction. 67's 2008 overdrive bot actually had rear-wheel-only pseudo swerve. If you watch their matches that year, it will suddenly dawn on your how they can turn their corners so quickly.

IIRC from reading an old thread, the back wheels were unpowered, just turreted.

[Andrew Schreiber]

Originally Posted by Tom Line

One very slight correction. 67's 2008 overdrive bot actually had rear-wheel-only pseudo swerve. If you watch their matches that year, it will suddenly dawn on your how they can turn their corners so quickly.

Ah yes. I forgot that. This is outlined Here

[joek]

Originally Posted by Coffeeism

While I haven't seen any first omni robots using less than 4 wheels. There are omni-wheeled robots with 3 wheels. In the past we've considered using a 3 wheel setup.

i was talking about true omni drive, the holonomic kind

[joek]

Originally Posted by Andrew Schreiber

The hard part is not the hardware. The devil is in the software. I admit, after 7 years of FRC and countless off season projects I am still intimidated by the thought of having to code a functional swerve drive during the build season.



What Tank Drive bots are you looking at? Ive seen some using 6 drive motors and most using 4.

then they must be using Fisher Price motors to drive, because we're only allowed 5 cims

[=Martin=Taylor=]

Originally Posted by Andrew Schreiber

....
...I know correlation does not imply causation but I have a hunch that there is a reason why none of these teams have gone swerving again. It may be a cool thing, and definitely a design every team should have in their arsenal, but in most cases it is not the most efficient design.

Last time 111 didn't build a swerve they won both their regionals.

The same is true for 1625.

hmmmmm...

[sdcantrell56]

Originally Posted by joek

then they must be using Fisher Price motors to drive, because we're only allowed 5 cims

We're using 8wd tank drive with 6 motors this year (4cims and 2 FP's)


Any team only using 2 cims in the drive this year is seriously hurting themselves

[dtengineering]

Wow! Did you ever get a lot of responses, quickly, on this one thread. You'd think you had posted a game hint for next year, or something!

But you asked two questions... one was:

"What should we build?"

and the other was "What does your team prefer?"

Those are two entirely different questions.

Frankly, I really don't think that the "What does your team prefer?" question is relevant to the choice that you have to make. First of all, very, VERY few teams or people posting here will have built and worked with BOTH mecanum and swerve drive systems. Many people will have SEEN both systems in action... and a few will have built and worked with both... but very few will be able to give you a first hand opinion on what they prefer.

What you need to consider is WHY you want to build a robot in the off season. Do you want to build it to work on design and machining skills? Do you want to improve programming skills? Do you just want to get first hand experience with at least one form of omni-directional drive? Do you want to have a cool demo robot? All of the above?

We have built a mecanum... but not a swerve. It was pretty easy to build... but took a bit of work to program (at least if you want 4 wheel PID speed control on an IFI control system... the cRio should make it a bit easier.) Unless you choose to build your own wheels, or develop a fancy suspension system, a mecanum drive is a very simple build challenge... particularly if you use a direct drive from either a Banebots or AM gearbox.

Outside of the discussions surrounding FRC competition robots... which are really kind of irrelevant to an off-season build project, the #1 advantage of a mecanum drive is that 99% of the people on this planet have no idea what a mecanum wheel is... and aside from a brief shot of a forklift on the recent Star Trek movie... have never seen one, either. Think about that... these wheels fit in a STAR TREK movie! Honestly, there is not much that is cooler from a teacher's viewpoint than watching a grade 10 explain to a P.Eng how your wheels work.

If you're looking for a bigger machining challenge, however, a swerve has all sorts of intricate parts that need to fit together just so. Sure, you can buy some COTS parts now to make that easier... but you are still working on a more mechanically complex system.

That is the reason we have avoided swerve up until now... we just don't have the manufacturing resources (mostly human resources... we've got the machines...) to confidently put together a good working swerve during build season. It would certainly be less daunting a task if we had built one as an off-season project.

But your team needs to think about why you want to build this thing... what you want it to be able to do... how much you want it to cost (in terms of money AND time invested in it) and then go with the machine that will make your team a better team.

Who knows... we might be back on regolith next year... maybe we'll have to climb stairs, or maybe the field will be made of corrugated iron. Or maybe wheels will be outlawed entirely.

Focus on the team, not the machine, and you can't go wrong.

Jason