Mechanum Wheels?

Originally Posted by theprgramerdude

Controls aren't part of the drive code; they're the parameters that are passed into the drive code. What your programmers don't seem to understand is it's their code, not the WPILib code, that can be improved.

It's like saying even though the motor controller is already set to 100%, the code can be improved to set it to 110% via "better controls". That doesn't happen.

We don't use mecanums. Never have. The point they made was in order to have the best quality controls for the drivers and have the driving experience be the most controllable and intuitive that our drivers can have, more complex programming would be required for mecanums compared to the traditional tank drive.

That's that, let's not derail this thread further.

[Joe Schornak]

My team has done quite a bit of experimentation with mecanum wheels this season, since we were initially doubtful of their effectiveness in certain critical areas, like traversing the barrier and climbing the bridge. Unfortunately, until we get some videos and pictures up (hopefully this weekend!) you'll have to trust me that our inadequately-described mechanisms actually work.

Surprisingly, the mecanum roller material grips better on plastic than on plywood. I suspect that fine sawdust is the culprit here, as our test of taping a sheet of plastic to one half of the bridge and then driving up resulted in the robot turning away from the side with the plastic, demonstrating higher traction on that side. We would be able to ascend while strafing, but our bridge has a significant lip that makes it very difficult to go from the ground to the bridge while sideways. No problems with turning sideways partway up and balancing that way, though!

We devised a system of raised powered traction wheels to pull the robot onto the barrier and provide a better point of contact to the mecanum wheels. Our prototype long chassis has no problems climbing over the barrier, even from a stop very close to the barrier without a high-speed run-up. The mecanum rollers function exactly as they should, in that the force vectors cancel due to the wheel orientation and no disadvantageous rolling occurs.

We use mecanum wheels because the extra range of motion is highly useful in offensive play. The ability to strafe has been particularly helpful when lining up on the bridge and will likely be very useful when the robot is sideways while trying to balance three robots. From a control perspective, we use flight simulator joysticks with a twist axis, a setup which has proved highly intuitive and easy to explain to 6th-graders during our school's open house.

In short, while mecanum wheels are certainly not perfect for everything, they have none of the negative traits frequently attributed to them, namely low traction, and we expect to use them to great advantage this season.

[372 lives on]

Originally Posted by SuperNerd256

We don't use mecanums. Never have. The point they made was in order to have the best quality controls for the drivers and have the driving experience be the most controllable and intuitive that our drivers can have, more complex programming would be required for mecanums compared to the traditional tank drive.

That's that, let's not derail this thread further.

just stop...

they aren't that hard to use and if you gave them a chance you would see that. now can you stop posting all of your opinions on something you clearly don't understand because it could mess up peoples decisions on what wheels to use.

[jvriezen]

Originally Posted by theprgramerdude

Controls aren't part of the drive code; they're the parameters that are passed into the drive code. What your programmers don't seem to understand is it's their code, not the WPILib code, that can be improved.

It's like saying even though the motor controller is already set to 100%, the code can be improved to set it to 110% via "better controls". That doesn't happen.

One relatively simple improvement you can make for mecanum drive is to add the option for 'field oriented' drive. Using a gyro, you can arrange the code to so that pushing the joystick forward *always* makes the bot move away from you, regardless of which way it is pointing. Moving stick right, the bot always moves directly to the right side wall of the field. This can make driving easier. Our bot last year allowed you to choose via a joystick button which type of drive mode you wanted.

[nlknauss]

Originally Posted by BJC

What I believe he is saying is that the midpoint of the wheel is level with the top of the barrier. Because of this the forces exerted by the robot must be entirely up until the wheel gains some purchase on the barrier and will continue to be able to pull the robot forward (read: over) the barrier.

If when your robot is climbing the barrier where the initial wheel contact with the barrier is lower then the middle of the wheel more forces remain pushing in the forward direction. In the same way, when you are on flat ground the wheels are pushing entirely forward. As a good rule of thumb if you can get the initial contact point inside the bottom 1/3 of the wheel that wheel will fairly easily surmount the point of contact while maintaining its forward force.

This is why it was suggested to put little ramps or something in front of your wheels. So the initial contact point of the barrier to the wheel is as low to the wheel as possible and forward momentum can be maintained.

Regards, Bryan

We did some tests with the bump and the AM kit chassis with 8" wheels and found out exactly what you are describing Bryan. Here is a tweet with a picture of the test to put a visual to this part of the discussion. The center of the 8" wheel hit the middle of the bump directly. We ended up putting a slope on the front edge of the chassis to help with horizontal forces.

http://t.co/4GxQFVqU

We used mecanum for the first time last year and had some success with them and the rest of our robot design. I always tell my students that the chassis decision should be based on our robot/game strategy. We built an offensive robot and of course tried to stay in front of our own rack, worked great. We decided not to use mecanum this year for a few reason with the main having to do with high level game strategy.

[JaredSanMiguel]

Originally Posted by SuperNerd256

-You won't be driving circles around other's bots unless you have a lot of practice.

I don't think this is the case at all. I was my teams driver two years ago when we had mecanums in break away. After a couple days of driving the maneuverability of our robot was evident.

Not to discredit what you are saying, if you don't have a couple days to spend on driving then learning at competition may be difficult.

Just don't want to discourage teams, the driving can be mastered in a relatively short amount of time.

[Koko Ed]

As of last year I still think a robot with Mechanum has never made it to Einstein.

[Taylor]

Originally Posted by Koko Ed

As of last year I still think a robot with Mechanum has never made it to Einstein.

Straw man.

That's like saying women are poor drivers because no woman has ever won the NASCAR championship.

As of last year, no robot with a Kinect has ever made it to Einstein either.

[thefro526]

Originally Posted by Koko Ed

As of last year I still think a robot with Mechanum has never made it to Einstein.

This is correct.

Though, 2826 made it a stone's throw from Einstein in 2011 with their 'Octocanum' drive. Had 71 not had issues in the Curie finals, mecanums may have made it. We're still quite some time off from seeing a pure mecanum make it to Einstein though.

Originally Posted by Taylor

Straw man.

That's like saying women are poor drivers because no woman has ever won the NASCAR championship.

As of last year, no robot with a Kinect has ever made it to Einstein either.

Is this really a valid comparison?

Look at the number of women in Nascar vs the Number of teams who use Mecanum wheels in FRC.

And the Kinect point is completely irrelevant here, Mecanum wheels have been in FRC since 2005, at my count that's 7 seasons that they could've and or should've made it to Einstein if they were superior (Though, Mecanum usage really exploded in 2007 and they were illegal in 2009) - the Kinect is new this season, maybe you can make the Kinect comment again in a few years.

The reason a Mecanum robot hasn't made it to Einstein is a really deep topic. I believe, and may be wrong here, but 51 is the only team in the Modern Era to make it to Einstein with any sort of non-swerve omni drive. Maybe Omni Drives that rely on Omni Wheels and Mecanum Wheels just aren't a good fit for FRC, who knows?

[Chris is me]

I'm just going to chime in and add that while mecanums aren't ALWAYS bad... 2012 really isn't the year for them.

[lcoreyl]

Originally Posted by Chris is me

I'm just going to chime in and add that while mecanums aren't ALWAYS bad... 2012 really isn't the year for them.

for whom? If you are a rookie or just inexperienced team, and you want to be able to shoot, mecanum is pretty simple to implement (if you buy AM wheels anyway) and takes care of your drive and gives you an easy way to line up your shot and your approach to the bridge.

[EricH]

Originally Posted by lcoreyl

for whom? If you are a rookie or just inexperienced team, and you want to be able to shoot, mecanum is pretty simple to implement (if you buy AM wheels anyway) and takes care of your drive and gives you an easy way to line up your shot and your approach to the bridge.

Mechanically, yes.

Programming, not really. If you haven't done mecanum in the offseason, it's generally a bad idea to do it in the normal season. The programming can take a while.

[lcoreyl]

Originally Posted by EricH

Mechanically, yes.

Programming, not really. If you haven't done mecanum in the offseason, it's generally a bad idea to do it in the normal season. The programming can take a while.

We have no experienced programmers (what is C++?) and received them during week 2 last year, and got them running in one day. We got gyro and encoder feedback working within a week.

[Chris is me]

Originally Posted by lcoreyl

We have no experienced programmers (what is C++?) and received them during week 2 last year, and got them running in one day. We got gyro and encoder feedback working within a week.

Oh, I didn't know it was THAT easy. I just figured 6WD takes basically zero setup time.

[EricH]

Originally Posted by lcoreyl

We have no experienced programmers (what is C++?) and received them during week 2 last year, and got them running in one day. We got gyro and encoder feedback working within a week.

Back in my day, we didn't get any of this pre-canned code or pre-made wheels. You young whippersnappers don't know how good you have it. It took us several weeks or more to get anything mecanum together!</geezer>

And yes, that is actually a true statement. It took 330 two years of off-and-on work to go from seeing a set for the first time, not in an FRC application, to a full-size, fully-functional set (2003 after Championship to 2005 build season). And they still haven't used mecanums in competition.