Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Technical Discussion (http://www.chiefdelphi.com/forums/forumdisplay.php?f=22)
-   -   Mecanum vs Treads (http://www.chiefdelphi.com/forums/showthread.php?t=110793)

Kevin Sevcik 09-01-2013 14:49

Re: Mecanum vs Treads
 
If I might summarize the Mecanum side of things, for easy reference...

Pros: Highly maneuverable, Fast (if you gear it to be), Cheapest Holonomic drive available.

Cons: Less traction, More driver training to use effectively, more expensive than a 6-wheel drive, and prone to wonky driving on uneven surfaces unless you have a suspension, use encoder and gyro closed loop control, and check and maintain the rollers on a regular basis.

So, unless maneuverability is really that incredibly important to you, you probably want to use something other than Mecanums.

On the treads side of things, as far as I'm concerned, a tank tread drivetrain is almost identical to a good 6 wheel drive, only it's more expensive and complicated and less reliable.

Bryscus 09-01-2013 14:51

Re: Mecanum vs Treads
 
Quote:

Originally Posted by Kevin Sevcik (Post 1211575)
On the treads side of things, as far as I'm concerned, a tank tread drivetrain is almost identical to a good 6 wheel drive, only it's more expensive and complicated and less reliable.

It's even more similar to a treaded 6-wheel robot on level ground...

- Bryce

CalTran 09-01-2013 16:07

Re: Mecanum vs Treads
 
Quote:

Originally Posted by Peter Matteson (Post 1211381)
I believe it's 100% of 9 sided robots, it was 3 wheeled.

No 9-wheeled robot has made it.

Also I believe 3 wheeled robots are 2 for 3.
67 -2005 Champion
16 - 2008 Finalist
148 -2008 Champion

Quote:

Originally Posted by Bryscus (Post 1211577)
It's even more similar to a treaded 6-wheel robot on level ground...

- Bryce

Well, I'll argue that every single one of six wheel swerve robots have made it to Einstein. Therefore, you should look into six wheeled swerve. All the advantages of traction with the omnidirectional motion of mecanums. :rolleyes:

I have not had experience with Treads (of the tank variety) but one of our current mentors came over from Team 1987, Broncobots, and has informed me many many times that treads are simply not worth their trouble. Without a lot of background and prototyping on them, they are difficult to make, repair, and you're probably going to be traversing the field dead slow. Also, for this year, keep in mind that tank treads will be a heavy system.

I have had experience with mecanum wheels and, while not endorsing them over the other, will share what I know about them. They are somewhat difficult to service between matches, unless you bring a full spare wheel or two to just replace the questionable wheel and fix it at your leisure. While it is true that it is an omnidirectional system, it does require a fairly significant amount of driver practice (as compared to a standard drive) because you have to train yourself to utilize the onmidirectional motion availiable.

I, like those before, do submit that you look into a solid six or eight wheel drive system, as there are many many threads discussing the merits of mecanum vs (6 or 8) wheel drive. It is vastly simpler, lighter than both of the aforementioned drives, and it will be more intuitive for your driver.

sybert1ger 09-01-2013 16:19

Re: Mecanum vs Treads
 
You may want to step back and define what you're trying to accomplish with your drive train. You can't decide which is "better" unless you've defined some metrics you can use for comparison.

As explained in detail by many of these posts, your metrics should include more than just on field performance.

It's really easy to jump right into design comparison but you'll end up with better robot performance if you slow down a few minutes and define what you want your drive train to accomplish independent of implementation.

karomata 09-01-2013 16:56

Re: Mecanum vs Treads
 
If you think that you could pull it off, a teammate and I developed a drop center tread/mechanum drivetrain. If you want to know more and maybe see some image designs, let me know.

efoote868 09-01-2013 17:31

Re: Mecanum vs Treads
 
While AM mecanum wheels may not have as much grip as high traction wheels, that doesn't matter too much. You might not be able to push your opponent across the field, but defense with mecanum wheels in a shooting game is as simple as colliding at a medium speed with your opponents.

pfreivald 09-01-2013 17:55

Re: Mecanum vs Treads
 
The woes of mecanum are not nearly so great as many would like you to believe, but:

1551 has used mecanum for five years (not counting The Game That Shall Not Be Named), and there's a reason last year we switched to octocanum. (Well, several, not counting wanting to challenge ourselves with something new.) Given the experience we've had, I would not choose mecanum-only for this game (and it has nothing to do with wonky driving over the tiny bit of uneven carpet -- that wouldn't stop me for a second!)

CalTran 09-01-2013 18:51

Re: Mecanum vs Treads
 
Quote:

Originally Posted by karomata (Post 1211664)
If you think that you could pull it off, a teammate and I developed a drop center tread/mechanum drivetrain. If you want to know more and maybe see some image designs, let me know.

So like Octocanum with treads then.

Dumper FTW 435 09-01-2013 19:30

Re: Mecanum vs Treads
 
Quote:

Originally Posted by commodoredl (Post 1211562)
Mecanums?
Treads?
Why not both? :p

(Not actually endorsing mecanums, treads, or mecanum-treads)

Please tell me you have more pictures of this or video of it driving. This is the greatest thing ever :ahh:

cmrnpizzo14 09-01-2013 19:55

Re: Mecanum vs Treads
 
Quote:

Originally Posted by StevenB (Post 1211216)
I suggest you count the number of Einstein robots which used treads and mecanums.

I don't think that this is ever a good strategy......

341, 67, and 1717 did not make it to Einstein last year, ergo shooting 5+ balls in hybrid, utility arms, and unique end game strategies are a bad choice.

If this is how we think, there would be no new ideas ever.

Out team personally is using mechanums this year because we believe that being able to maneuver around the defense is more effective than withstanding the hits or pushing bots. I don't think that a bump less than .5" tall will be any sort of noticeable problem for the mechanums. I might be wrong, but I wouldn't be too worried about it.

Also, nice job, Eric, on those treads they are sweet. (and dreadfully heavy)

karomata 09-01-2013 20:24

Re: Mecanum vs Treads
 
Quote:

Originally Posted by CalTran (Post 1211739)
So like Octocanum with treads then.

Not quite, I will be sure to post it though.

dtengineering 09-01-2013 23:30

Re: Mecanum vs Treads
 
This is all very practical advice... and I agree that for winning this years' game, likely neither tank treads nor mecanums are the optimal choice.

Not to understimate the significance of competition, but allow me to point out that your robot will retire from competition just four months from now. Okay... maybe the odd off-season event, but really, once the championship is over, every design choice you made for "winning the game" suddenly becomes redundant.

It is totally okay to use treads or mecanums or swerve or hovercraft technology just beacause you think it is awesome.

We used mecanum wheels in 2007 (it didn't work nearly as well as we'd hoped) but those wheels have kept on being awesome. People still look at the machine and ask about the wheels... allowing me to give a brief lecture on vector mechanics. Our design choice might not have accomplished its mission on the field.... but as far as inspiring science and technology discussions it is still hard at work six years later.

Sure, if you have the time and resources you could build an off-season bot (especially if you are going for the hovercraft drive), or just demonstrate mecanums and treads using VEX... but that isn't quite as cool as having it on a "real" machine.

So kudos for asking advice... you've got plenty of good advice so far in this column... but honestly, if Orville and Wilbur Wright had taken good advice they would have packed up and gone back to their bicycle shop.

Of course, Orville did darn near kill himself flying one of those things. Ignore the advice at your own peril... but feel free to ignore it!

Jason

lcoreyl 20-01-2013 05:44

Re: Mecanum vs Treads
 
I had to revive this thread due to some great discussion found here. Our team has posted a whitepaper addressing a few mecanum misconceptions, and I think all of them are here in this thread: mecanum sacrifices torque and/or speed, is inherently inefficient, et al. The link is in my signature.

*while i'm about to address some points in this thread, i suggest that your responses specific to mecanum go in this thread

Some points more geared to a teams experience with mecanum:

*NOTE: I'm not advocating for mecanum this year (although we are using it... most likely) I've just decided for some reason that now after most teams have decided was a good time to start weeding out bad information. :rolleyes:

-We don't do a good job with weight distribution or frame flex and have NEVER even considered a suspension, and our bot operates just fine (i.e. very little deviation from expectation). We even balanced on the bridge last year. (would regular wheels have been easier? sure)

-our inexperienced programmers get gyro/encoder feedback working every year without much trouble and then it reallyoptimizes the driving.

-our drivers have found robot-centric drive vs. regular about the same, and strafing when needed was very easy and intuitive. Field-centric drive is REALLY intuitive and a major upgrade. (I guess you could do a 6wd field-centric program... just never heard of it)

-While we did get 2 sets of andymark 8" (because everyone scared us about them), we never had to do any maintenance during competition (on the wheels that is)

-I've seen over and over the "straw man" argument that if you want to be a block/push defense bot, mecanum is a terrible choice. I've never seen anyone advocate that, and can't recall a team asking input for that idea. Last year driving towards while also orienting the intake towards a ball was actually a defensive technique for which mecanum could be a good choice. The "keep away" defense. Playing "get in the way" style can also have success. Our alliance member last year actually did some heavyweight blocking/pushing (although not as their main strategy) with good success. would 6wd push better? sure. Does mecanum suck at it? no.

-Cool factor IS a valid consideration since we are looking to attract mentors and sponsors

PVCpirate 20-01-2013 17:12

Re: Mecanum vs Treads
 
Quote:

Originally Posted by Kevin Sevcik (Post 1211575)
Cons: Less traction, More driver training to use effectively, more expensive than a 6-wheel drive, and prone to wonky driving on uneven surfaces unless you have a suspension, use encoder and gyro closed loop control, and check and maintain the rollers on a regular basis.

I'm gonna have to disagree with you on one point here. If you use field-oriented driving controls, it probably makes it easier to learn to drive the bot, since it will always move in the direction the joystick is pointed. Combined with a spin control the driver is comfortable with, you've got an intuitively controlled omnidirectional drivetrain, and its driver doesn't need to think about which way it's facing to move somewhere.

Robofreak3130 20-01-2013 17:25

Re: Mecanum vs Treads
 
I would say neither. Both options are way toheavy if you are planning to climb higher than level one.


All times are GMT -5. The time now is 18:10.

Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi