Autonomous Mode

This is just personal opinion, but every year that I’ve been involved (since 2000), FIRST comes out with some new twist or gimmick to the match, to make it difficult in a slightly new fashion.

We’ve seen 4 vs. 0, which was sometimes quite the strategy headache, and the advent of Autonomous Mode has pushed programmers.

This year, I believe that we’ll see Autonomous Mode moved to the end of the match, with the scoring left at the end of the match as well.

With this feature, it will be easy for FIRST to design a game in which Veterans can achieve great points with a highly effective autonomous program, and Rookies are able to maintain a good score easily with decent performance during the manual control period and some stable defense for atonomous mode.
$0.02
TW

I can also see a big difference in the point value of different items, or the same point value for a very large number of items (soccer balls and goals from '02, and the boxes and ramp from last year) I dont think they would design a game that favors verterans too much, they need to give the rookies a good shot at winning or they’d just see the same old faces at nationals. With the varrying point values, it gives teams, new and old, a large variation in the strategies. Also, I forsee a very easy game to visualize, score-wize. I remember hearing somewhere that they were going to try to keep it spectator-friendly and with any game too complicated, it gets the crowd lost too easily. Anywhoo. Thats my $0.02, bringing the total value of this post up to a whopping $0.04. :smiley:

I think that would be cool, but heres why i don’t think it will happen. In order to do anything in autonomous mode you must have some idea of where you are. Developing a system to track position through an entire match is probably out of the reach of most teams. I guess you could sortof drive your robot in some fixed spot. If there were lines on the feild nearby you could align yourself to those and then start your auton.

Perhaps the competition control will give you the robot’s exact field coordinates.

Maybe there will be stationary objects which are easy to get your location off of? (two fixed points covered in retro tape for instance).

Maybe autonomous mode won’t require driving.

There are lots of things that would allow first to move the autonomous mode to the end of the match. I don’t think its very likely though.
Greg

Don’t be so sure that lack of knowledge of field position will keep them from moving auton. In the real world, autonomous robots have to operate based on one thing: sensor input. My guess is that if they do indeed change the timing of user/auton modes, the objects on the field will have some way to give good sensor input. However, seeing that we have these brand new microprocessors to learn, I don’t think they’ll throw another big programming trick our way this season…

Auto mode was not bad… I think it would be cool to see it moved to the end of the match… We didn’t too bad with it last year considering we got a design award for that and the “big swing” which still wasn’t powerful enough to knock all the boxes off…

I doubt we’ll see auton at the end of the match. the only real purpose for auton is to be able to perform some fixed goal. But if the autonomous is at the end, your goals will differ hugely depending on how the match has gone.

I would like to see a longer auton, though.

intx13- true but you can also have multiple programs… Not just one big one… I believe a couple teams last year had like 4 or 5 different modes…

We (1097) had 2…4…6…8…10…11 programs to choose from last year. Gave me a headache though, because i had to change the program everytime because we would change the robots configuration, change starting position, or our allies would decide that since we were rookies, our autonomous would be less successful than theirs (the irony was that thier auton always screwed up when this happened, even GUNN and Monta Vista). Anyways, Autonomous wasn;t much of a problem lasy year. The programming was simple. I believe this year autonomous will be much longer, maybe even half the match. Having it at the end of the match would be fun, very fun and quite a challenge, which i always love.

It really isnt that hard. Gyroscope and accelerometer will do it all with some integration.

But there is another option. If you looked last year, there was one little antenna on each of the four corner posts. They looked like the robot antenna. Anyways, i have good information that those antennas are used for FPS (Field positioning signals), basically GPS in a small scale. FIRST just hasn’t open it up to us yet, but if they decide to move autonomous this year to the end, or even if they make it much longer, they might give us the hardware and information we need to use them. Makes programming that much more fun.

-Kesich

A simple way of getting around gyros and reflective tape to find your position, if indeed auton was moved to the end of the game, would be to position yoour robot in a specifc spot and then run a certain auto program. For example, if this were applied to last year’s game, then you could program the robot to drive ontop of the ramp and run, say program number 5, to have the robot drive 5 feet past the end of the ramp, turn left and into a stack (the stack would be there, other wise you would have run a different program).

There are many problems with this, but atleast its a simple solution. The first problem being that your referance point could be blocked by a robot, or you may not want to give up your KOTH position. 2nd, there most likely would be a robot gaurding the stack.

I’m curious to see how big tracking/triangluation routines would be. Lest be forget, once you find out where you are, you have to figure out your oriantation and move somewhere. While moving, sweep for any robots that could be in the way. I think that you would be very much straped for program space. Think about how you would do it if you where in an place in the scoring zone and wanted to climb up the rap to get an additional 25 points.

Thats my $0.02 PLN ($0.005439 USD)

range finders with the dimensions of the field programmed in solve it all. If the bot sees anything, it would then confer with the equations for the walls and see if they confer. If they do, its a wall, otherwise its a stack (or what have you) or another bot.

And I hope you take that cuz im too broke to afford $0.02, or is it $0.04, or $0.06 now… aww I don’t care, im not pating anyways.

-Kesich

Well, since this is the rumor mill, I’ll tell you what I’ve been hearing.

FIRST is attempting to make autonomous a bigger part of the game (you all obviously know this), but from my understanding the want it in the actual game- as in autonmous functions (such as picking something up without using the controls, arms that are self-guided, etc) that take place during play, not before (or after).

That would be one of the biggest reasons for switching to C. Being able to define functions externally from the main program and such is a rather nice feature.

meh, thats just my $.01 (I just paid college applications, i couldn’t afford two cents).