[dtengineering]

For Rack'n'Roll we built a mecanum drive... for many of the reasons that have been advocated here.

It was great... I'm glad we did it... and when we eventually got everything working and our driver learned how to maximize the freedom it gave him, it worked great.

But it came at a cost... mecanum requires much more software development than tank drive, particularly if you want closed-loop feedback for motor speed controls. Or at least it DID... the Jaguars are well set up to offload speed control processing from the main controller, and even then, running PID speed control in labview is next to a "no-brainer" compared to what it was on the PICs. (I actually liked the fact that you had to write efficient code to accomplish complex operations on the PIC...)

Mecanum also required a lot more time for the driver to learn how to maximize it's abilities. A good tank-drive push-bot is easy to learn... but mecanum gives up raw pushing ability for elegant maneuverability. It takes longer to learn to be elegant than it does to be a brute.

In the end, I think from a competitive point of view we would have been better of to go with a 6wd system... but people still love to check out the mecanum wheels and see the robot strafe. If you want a robot that is fun after April, you might consider taking a risk on mecanum drive. Just keep in mind that you're likely building a mecanum because it is cool, not because it is necessarily the best drive train for playing this particular game.

Finally we built a 4 CIM, 8wd for last year's game. We put 14:1 gearing on it, direct drive to 6" VEX Pro traction wheels. That robot could accellerate to full-speed, come to a full stop, turn 90 degrees and get back up to full speed basically as quickly as our less-torquey mecanum drive could. We had basically achieved maneuverability close to that of the mecanum but in a robot with around 200 pounds of pushing force and the ability to "get air" over the bumps.

I'd suggest that high rates of accelleration, rather than top speed will be critical for the game. Consider how much time you will spend getting up to speed, and slowing down relative to the amount of time you will spend at top speed. If you want the best of both worlds, a shifting gear box might be in your design, but that is only useful if you actually shift gears frequently during a match.

Jason