Speaking from the programming end, this year has been leaps and bounds more difficult to program for auton. Last year we really didn't worry too much about other robots, and we were only trying to hit a single target.
This year, avoiding other robots while still getting to the scoring locations and lapping as quickly as possible has already resulted in 3 or 4 total rewrites of the autonomous code as we try to improve our methods.
Plus we don't really have a drivable chassis yet because we decided to do a far more complex drive train than we've ever done before. The whole team is nervous.
