Quote:
Originally Posted by platypus
6. Autonomous is a bust, which is kind of sad but completely unsurprising. One keeper was scored today. Just one. As a software guy, I know that it's not an easy autonomous to write this year. Last year, you could use the light to align a shooter and fire from the starting position. This year, you have to move while tracking the light, and move a manipulator to the proper height. This is not easy, and there really isn't much incentive to make it work. Since spoilers aren't used, keepers are basically ringers (and even if spoilers were used, it wouldn't matter very often as spoilers would break the row on the ringer next to the keeper). That's 2 whole points for your trouble, which is why teams just aren't bothering. (My team for one instead focused on an advanced teleoperated control system with preset postitions recorded into EEPROM, and relegated autonomous to that bottom part of the software todo list that never sees a share of the limited integration time.)
|
I have to agree that autonomous isn't worth as much as it could be. At St. Louis, maybe one or two spoilers were placed ever. I thought we would be spending most of our time working on autonomous, when in fact we spent most of our time working on perfecting our teleoperated code. We do plan to write an autonomous mode, but we think it will mostly be for show.
This year, more than ever, a sophisticated teleoperated mode is incredibly useful, particularly on 868's robot. Our tube handling system has 3 separate controllable parts, and a simple switch-on-the-OI system would make the job of the operator incredibly complex. Fortunately for me, depending on the job at hand, there are only a few distinct setups that the system should have. (The grabber should never be open when the arm is up in the air unless you're trying to take off a spoiler, or you've hit the rack)