Future of Autonomous Mode

[/Pyro]

Originally Posted by 10intheCrunch

I know FIRST isn't supposed to be fair, but autonomous mode as it is out of reach for many teams.

FIRST IS meant to be fair. Team 1241 showed us at GLR that a Rookie can do just as well as veteran teams. They had an amazing design and idea. Obviously veteran teams are doing well with awesome designs as well. Autonomous mode is simply something else that teams can use to show off their designs and strategies. We can't look away from the fact that some teams have more engineers than others. This will make their robot do more things in the game, but a specialized robot can do just as well as a robot that can do everything.
In respects to the goal of FIRST and the competitions, it is meant to be very fair. I hate it when people say FIRST isn't fair. We have been a team without any engineers for all 6 of our years as a team. Our students have always designed and built the robots by themselves. We even managed to win Midwest and Newton last year. It all comes down to strategy and how hard the students want to work.

[roberthan]

I too, think programer should have a bigger part in FRIST, I think they start by a better RC

[theory6]

Autonomous mode is probably the most important parts of the whole 2 minutes probably because not all teams perfect it. For example our team NEEDED an alliance who could knock off the bonus ball during autonomous. If they succeeded we could guarantee a win 90% of the time. If they failed to knock the ball we would be stuck there without any strategy for 45 seconds making us incredibly vulnerable. My point is that teams might as well take advantage of this given 15 seconds during the game since It could make a huge difference in the end.

[FizMan]

I think the autonomous probably won't become much more complicated than what it already is now... for the very simple reason that as the complexity increases so does the complexity of the testing requirements.

Not all teams are capable of building a full mockup field, for example. Nor do all teams manage to finish their robot with weeks to spare in order to test these autonomous programs.

I think the key in developing the autonomous area of the game is to keep the procedure relatively simple (not much more complicated than what it was this year) and increase it's strategic significance in the game. In last year's stack attack, had the potential to make the autonomous really worth while... but then the king of the hill came in and seemed to pretty much negate that. This year, again, you'd think knocking off (or grabbing) that 10 point ball would be worth more, but herding those balls (and managing to throw them in successfully) turned out sometimes to be less successful than just hanging from the bar.

(of course, there are always exceptions with those amazing robots)

I remember my adult mentor discussing this with someone else... (for the life of me I can't remember... it was probably with me ) if they wanted to make the autonomous more of a priority for teams to make it worth a set number of points. For example, rather than just having your score based on the final state of the field, if you had your robot knock off that 10 point ball, you'd automatically get 20 points for example.

Unfortunately, doing so would probably further unbalance the game. Where teams with the programming and testing resources would be able to develop a flawless system and teams where they can't afford to build testing fields or have time to test and debug the program... or even the experience and education to write the programs and work the miscellaneous sensors would be left in the dust.


So where does the future of autonomous lie?

Probably right where it is now; purely for a strategic importance. Getting your robot in place to climbe the stairs. Or knock off the 10 point ball. Something that would help you (or if you have bad luck, break you) but it isn't required by any means to have a successful match.

[Ryan M.]

Originally Posted by FizMan

Not all teams are capable of building a full mockup field, for example. Nor do all teams manage to finish their robot with weeks to spare in order to test these autonomous programs.

Amen.

I totally agree with that. However, I feel that autonomous mode should still become a larger part of the game.

Plus, I want a better microprocessor.

[Joel J]

Autonomous should definitely have a future in FIRST. As many others have said, it need not become more complicated; however, I don't believe it should remain the type of simple it is now. We should be able to choose among different objective tasks in autonomous mode. For example, we should be able to decide whether or not we want to knock over a stack of boxes (shudders), or remove the 10-point ball from its holder, or move a mobile goal into position, or etc. Essentially, the range of options should definitely be increased, but the complexity need not become greater as a result.

[JoshJ]

I like what Fizman said, having auto mode get u points directly if you could do a certain task.

As far as complexity, I think we are good for what we need to do. For now, it is only a 15 second run. Teams struggle getting that to work right. For FIRST, all you really need is consistency in what you do, no complexity. If you think about the DARPA Grand Challenge, those teams had everything they could want at their disposal, yet no one completed the race.

The FIRST matches are hard enough for people to make decisions, I think it would take some of the fun (and defintitely most of the pre-match strategy) out of it if the computer made all the decisios(if it even could). Thats my two cents.
-Josh

[ejthe4th]

Not many decisions for autonomous mode? There are many different things that can be accomplished during autonomous mode...a robot can attempt to knock off the 10pt ball, block another robot from getting to the ball, push a goal into the ball corral, interfere with a robot, set yourself up for the bar, strategically sit in place for 15 seconds, wildly drive around hoping to do something, ect... There are many things that can be done during that 15 seconds for autonomous mode, sure not that many teams are actually able to do them consistently, but the option to attempt different strategies is defiantly there.

[Marc P.]

Originally Posted by roberthan

I too, think programer should have a bigger part in FRIST, I think they start by a better RC

This year FIRST did provide a better RC. The PICs (while not perfect) are a huge improvement over the pbasic stamps we've used for the past 4-5 years. If not in programming simplicity (I'm fluent in basic... know peanuts about C), in capability. I'm not a C programmer, but I can certainly appreciate the power of such a language, and the hardware it's now running on for competition. It's the sort of thing that may take a year or so to pick up before one can start to truely take advantage of it, and I feel this is that year. There are teams with a number of members who went into this season with a working ability to read and write in C, and plenty of other teams with not one member who's ever seen C before.

That said, I agree with Joel. The autonomous is just right as it is now for a high school level competition. Basic functionality will get things done in the match (just moving forward while aimed at the 10 point balls), while more advanced programming can accomplish more (knock off 10 point ball, then collect balls, move goal, etc). The more advanced the objective would get, the more it would fall out of reach of more teams. However, I don't see any teams at a distinct disadvantage due to lack of programming ability: the objective was simple enough where a basic program would suffice. I'd fear that should the complexity increase.

[10intheCrunch]

Originally Posted by ejthe4th

Not many decisions for autonomous mode? There are many different things that can be accomplished during autonomous mode...a robot can attempt to knock off the 10pt ball, block another robot from getting to the ball, push a goal into the ball corral, interfere with a robot, set yourself up for the bar, strategically sit in place for 15 seconds, wildly drive around hoping to do something, ect... There are many things that can be done during that 15 seconds for autonomous mode, sure not that many teams are actually able to do them consistently, but the option to attempt different strategies is defiantly there.

But the robot itself is never forced to make these decisions during a match. It knows what it is going to do, for the most part, before it starts. If a robot had to play an entire match autonomously, it would have to track tons of balls, herd, dodge other robots, cap by itself with balls flying everywhere, hang by itself. That's beyond a high school level, at the edge of what we can do right now. Some day for First? Maybe, but not now, next year, or for a long while.

[Ryan M.]

Originally Posted by 10intheCrunch

But the robot itself is never forced to make these decisions during a match. It knows what it is going to do, for the most part, before it starts. If a robot had to play an entire match autonomously, it would have to track tons of balls, herd, dodge other robots, cap by itself with balls flying everywhere, hang by itself. That's beyond a high school level, at the edge of what we can do right now. Some day for First? Maybe, but not now, next year, or for a long while.

You'd have to serious processing power, a great programmer (like the entire Microslop () corporation. Maybe Linus Torvaldes.), and video feedback or something equally hard to be able to do this. Nothing is stopping you if you really want to go out and do that.

It would be exremely hard, although if anyone could do it successfully, it would be awesome.

[Adam Y.]

You'd have to serious processing power, a great programmer (like the entire Microslop () corporation. Maybe Linus Torvaldes.), and video feedback or something equally hard to be able to do this. Nothing is stopping you if you really want to go out and do that.

Im not really sure about that. If you were to chop up the problem into easier to solve problems it really wouldn't be that hard. I myself would go for the master/slave configuration where a bunch of little slave PIC's would be feeding information into the master microcontroller. Line following in itself is something so simple you really do not even need a microprocessor to build a robot.

[10intheCrunch]

Originally Posted by Texan

You'd have to serious processing power, a great programmer (like the entire Microslop () corporation. Maybe Linus Torvaldes.), and video feedback or something equally hard to be able to do this. Nothing is stopping you if you really want to go out and do that.

It would be exremely hard, although if anyone could do it successfully, it would be awesome.

Not with a $300 or so electronics budget.

[deltacoder1020]

hmm, i wonder if you could get them to make a special allowance for the budget if you compromised that there was no human control at all...

[Ryan M.]

Originally Posted by Adam Y.

Im not really sure about that. If you were to chop up the problem into easier to solve problems it really wouldn't be that hard. I myself would go for the master/slave configuration where a bunch of little slave PIC's would be feeding information into the master microcontroller. Line following in itself is something so simple you really do not even need a microprocessor to build a robot.

True, doing those things in themselves might not be that bad. But, it would be diffucult to have the robot decided, "Hey, it's time to go chinup!" or, the same robot saying, "There's an opponent just sitting there waiting to decap this goal. I can get more points by defending it than going and doing the chinup." or "I'm on the platform and can hook myself, but my opponents are coming up. They'll win if I let them hook, so I have to hold them off till the very last possible second." That's more along the lines of what I was thinking.