In our case, our fundamental strategy hasn't really changed at all -- just a lot of significant details in its implementation. As we worked through the design over the last several weeks, we made a lot of mistakes and changes and discoveries, all of which came with some number of band-aids and unplanned additions.
By starting over, we were able to take account of everything we'd learn and include all of the subsequents requirements in our design from the start. So, with some concessions to the design of the existing drivetrain, we've ended up with something that's more robust, lighter and more successful in accomplishing what we intend to do. We put in two very long days over the weekend and are caught up again. With luck, we'll be able to get the whole machine together by Sunday.
The robot is pretty simple, all things considered. The drive is 4WD with a single CIM on each side running through a custom gearbox. The overall ratio is 8.75:1. The intake roller has pneumatic tubing bristles that and a quarter round backing that bring the balls up to the conveyor system. The conveyor system consists of a pair of individually driven belts -- each powered by a FP motor, AM Planetary and the same gearbox we're using on our drive. Not shown is a flap at the top of the conveyor that will push the balls back into the hopper as the come out of the belting. That flap is "normally-closed," such that it covers the belting when we dump balls out.
The hopper can store ~15 balls and measures 14" x 35" x 24" or so. We'll get balls out by raising the height of the floor, essentially, and pushing the balls upward. They then roll over the flap mentioned above and into the opponents goal. The plan, of course, is for this to happen as quickly as possible and so that mechanism is driven by the remaining two CIMs, again through the same gearboxes as those on the drive.
Also, for the record, I don't advocate shooting the engineers.
