Mecanum Wheel Considerations

[abross]

Hello. I have read a couple threads that talk about mecanum wheels, but I have a more specific question. My team is trying to decide whether we want tank drive, or to use mecanum wheels. This is our second season, so as of now, we only really have experience with tank drive, but we're willing to try out something new.

I was wondering if anyone knows how well mecanum wheels can stand up to another robot pushing or 'ramming'. Will they slide more? Will they have a tendency to quickly lose control?

Any and all opinions/advice would be greatly appreciated.

Thank You.

[wilsonmw04]

greetings,
There are a few things that you give up to gain the maneuverability that comes with Mechanum drive.

1. lack of friction. With rollers as you contact point with the floor, you will find it hard to push other robots or climb the bumps.

2. lack of "power". With the way mechanum, works you lose half your force by canceling vectors perpendicular to forward motion.

advice: pick the drive train you think will work best for the game or what you want to accomplish in the game. Stay with what you know and what you are comfortable right now. IMO build season is not the time to go out on a limb and try a new drive train.

[EricH]

Originally Posted by abross

I was wondering if anyone knows how well mecanum wheels can stand up to another robot pushing or 'ramming'. Will they slide more? Will they have a tendency to quickly lose control?

Any and all opinions/advice would be greatly appreciated.

Thank You.

They won't stand up to pushing. You'll have to avoid pushing, which that extra degree of freedom you get with a mecanum drive will hopefully allow you to do.

[Kevin Sevcik]

With the lack of pushing power in mind, watch some of the videos of mecanum bots in action, and realize two things:
1. You're going to need to develop some way for your drivers to control a robot that can move in two directions and rotate all at the same time.
2. Your drivers are going to have to figure out how to drive the darned thing well enough to avoid getting cornered.

This means you're going to need to finish a driveable mecanum base early enough that your programmers can get it moving AND that your drivers can get comfortable with it.

[abross]

I have already considered the programming involved with programming a practical driving setup, but the fact that they can't stand up well to pushing is what is concerning me at the moment. Thanks for the help.

[thefro526]

Realistically, if you're building a Mecanum Drive you shouldn't really ever need to push someone because you can drive around them. In the situation that you were being pushed, then you would drive around them. Your increased mobility is what you would use instead of the ability to push.

As far as controlling it goes, (I believe someone mentioned the difficulties controlling a mecanum drive above) You can use a combination of 2 joysticks, one for X/Y travel and the other for rotation, or a 3 axis joystick where you have the conventional X and Y and then also have a twisting Z axis.

Btw, you guys are pretty close to 816, if you ever need anything let us know.

[dtengineering]

Originally Posted by thefro526

Realistically, if you're building a Mecanum Drive you shouldn't really ever need to push someone because you can drive around them. In the situation that you were being pushed, then you would drive around them. Your increased mobility is what you would use instead of the ability to push.

I agree with that, but I'd say "theoretically" instead of "realistically".

Our experience with mecanums is that you do give up brute force, and in return gain a fair bit of finesse.

Without well-tuned software and a talented driver, you can't take advantage of the finesse. Any idiot can take advantage of brute force. (Admittedly, there aren't a lot of idiots playing FRC... but you get my point.)

So in Rack'n'Roll we built a mecanum for many of the same reasons you're likely considering it for this game. After two regionals we had everything tuned in just perfectly, and our driver executed some amazing plays. In one match, he faked right, went left, and executed a 180 degree spin while strafing sideways past one of the best defensive bots at the event, finishing with a "slam dunk" of the tube onto the rack. It was probably the nicest maneuver I've seen our team do... ever. Unfortunately it was our last match as it had taken us two regionals to figure out how to really take advantage of the mecanums.

BUT... if you want to build a mecanum... do it. They are really, really cool, and whether you win matches or not, you'll have a bit of robot "bling" in your collection for years to come... one of my favorite things is watching grade 10 students explain mecanum wheels to engineers who've never seen them before!

Go for it... but be aware.

Jason

[MrForbes]

We built a mecanum drive T shirt cannon robot a few years ago for learning/fun/promotions/display, so we know how mecanum drive works. But as yet we haven't been brave enough to use this drive system on a competition robot.

[idahorobot]

Good advice! Do omni wheels get the same reviews as Mecanum?

[KHall]

Hello All,

Two years ago Team 2171 won the West Michigan Regional on mecanum wheels playing FIRST Overdrive. So naturally when we got the new controller last year, we experimented with the Mecanum wheels using LabView.

We used the LabView software driver and it worked just fine. Converting the joystick's output to the format the mecanum driver needs proved to be an interesting problem with a surprisingly simple solution, BTW.

I must add that I have not yet looked at the LabView software for the 2010 season to see if anything has changed on the mecanum driver's inputs. I'm guessing (and hoping) that it didn't change so we can just re-write the code we had working last year.

The key to success with these wheels is practice. The robot's motion is not at all intuitive, so your drivers will need plenty of practice to get a feel for its abilities.

Good luck playing Breakaway everybody!

Keith Hall
Mentor
Team 2171, Crown Point Indiana RoboDogs.

[ebarker]

In the Labview vi is the driver control one or two joysticks ?

Do you have both options ? I suppose you could write code to do it either way depending on driver preference.

[KHall]

Hi Ed,

If memory serves, LabView expects the mecanum wheels will be driven from a single joystick. (Has anyone looked at the 2010 LabView code who can confirm this?)

The Easy-C driver for the older PIC-based controllers used a modified tank-drive which required two joysticks.

Another difference is how you enable spin mode on the robot. With Easy-C, I seem to recall, you would pull one stick down and push the other one up to enable spin mode. With LabView you'll probably have to do that some other way.

Of course, you need to make sure that the robot's abilities are aligned with your game strategy. For example, ask yourself under what circumstances would spinning be an advantage? What are the potential risks and pitfalls from spinning? Just because a robot can do something doesn't necessarily mean that doing it will help your alliance win the match.

Regards,
Keith Hall
2171

[abross]

Well thanks everyone for the help. I finally have more knowledge to be able to decide whether or not we should attept this type of wheel. I'll update you when I find out more. And I'll probably be back to ask more questions.

[Lil' Lavery]

Originally Posted by idahorobot

Good advice! Do omni wheels get the same reviews as Mecanum?

In general, the physics between the two are essentially identical (assuming it's a 4-wheeled holonomic drive with wheels at 90deg angles). The coefficient of friction may vary slightly, depending on what brand omni-wheel you buy, but they are usually in the same ballpark. Omnis are usually much cheaper and lighter than mecanum wheels. On the other hand, mecanums are (typically) better at handling inclines (such as the bump) than most omni systems, though both systems may require a suspension to cross the bump.


An important note about both. The vector-based physics of the systems will limit you to 50-71% of your total power of your drivetrain, and the angles of the rollers also limit you to 50-71% of your total contact area resisting outside motion (your total traction). That % varies based on what direction your driving/being pushed. So the system will both suffer while trying to push another robot, and while being pushed.

[pfreivald]

On control, keep in mind that if you make one 'strafe' and one 'turn' joystick, it will work very much like an Xbox controller...

...which begs the question: have you figured out how to get an Xbox controller interfaced and working with the new control system? (We have not, as of yet, but we certainly want to!)