|
Hmm... we tried practicing our autonomous mode at Philly too. We borrowed a piece of carpet from another team (I wish I could remember the number), and went outside in a clear area to test it (we were just going to be turning 45 degrees in place, so not a lot of room was needed). We were told by a Drexel official, however, that robot teams, contrary to years past, were not allowed to practice their robots on the ground (i.e., it had to be up on blocks, or some other form where it wouldn't actually move on its own accord). This was due to liability concerns (isn't everything?).
Anyway, I believe most autonomous testing can be done just by observing the wheels -- obviously you can't get the "exact" values, but in the least a general feeling about "safe" values ... and anyway, this is what the practice rounds are for. And if it isn't done by then, what's to stop teams from figuring it all out during the actual match? Sure, there's a need for FIRST to plan to have safe areas to practice out autonomous/ general performance ... but isn't this the practice matches on Thursday? At Philly there were 4 practice matches, and with 2 "matches" each time that means 8 trials of the auton program. That provides a lot of data to get it right, in my opinion.
And to the main point, I think its probably better to try to do 'something' than nothing ... a simple dead reckoning system can be quickly developed for your robot, and it helps a lot to get a good position in the first 15 seconds of the match. Though, for various reasons, it might be advantageous to have a mode where the autonomous mode does nothing -- this, however, is trivial, and there isn't any real excuse to have to rely on the human player to disable the auton mode. If it's what you have though, use it to full advantage.
|