Quote:
Originally Posted by niklas674
While our best year involved us using a mecanum drive (mistake), we practically never used it in matches, the positioning was close enough that we didn't have to make large changes. Bottom line is that with plenty of drivers practice, you won't need to be able to strafe because your drivers will get to the exact spot almost every time.
|
__________________________________________________ ___________
TL;DR: Even in a game free of opponent pushing where mecanum was a viable option for strafing motion, it still was not our first choice in the design of the drive system. In the end it failed to work as well as we wanted.
__________________________________________________ ___________
I recall a few factors in 2015 in deciding to use mecanum. One was the fact that we were trying to be a landfill clearing robot, which required fine maneuverability due to the perpendicular orientation of the totes. We also constructed our robot to ride along the alliance wall behind the yellow totes as part of our autonomous, which required the strafing motion. But even with that in mind, we didn't even opt to use mecanum first.
Our first drive system iteration was an H drive with 4 omnis on the outside and one perpendicular omni in the middle. We wanted H drive over mecanum first, because we didn't want to bother with programming mecanum when H drive was a simpler option, and furthermore, H drive strafing is not as dependent on mass distribution as mecanum, if at all (as far as I know, correct me if I am wrong please). We knew that grabbing totes and making stacks at the forward area of the robot would drastically change the distribution, so H drive seemed to be the smarter option.
Unfortunately, the center wheel of our H drive would cause the robot to lose contact with some of the floor when driving up the scoring platform. At this point of failure in our prototyping, we settled for mecanum, and I say "settled" for a reason. Through all our days of practice and competition, the strafing abilities were hampered due to poor distribution of mass, especially with the addition of totes in a held stack. Because of this, our intial idea of a strafing wall-riding autonomous was given up for simpler options.
As Niklas mentioned, we barely strafed at all and drove the robot more like a tank, and I feel like we should have just designed for tank. After all,
the only teams who were major scorers that beat us in all three of our regionals in finals did not have mecanum (118: 2 tread 2 omni tank, 254: 6 tread tank, 624: 5 omni H, 1678: 6 tread 2 omni tank) and two of them were the world champs, 1678 and 118, both of which we competed against at two separate competitions in the finals. I think I can chalk up our losses to these teams partly to a lack of proper control due to an improperly implemented mecanum drive.
The lesson I got from this was that even in a game where mecanum may seem like a good option due to the game design (no opponent pushing and lots of opportunity for fine control), it still might not be that great of a choice, so be careful with that decision. Some of the best teams use some form of tank drivein every game without fail, so I like to go their directions.