Re: Future of Autonomous Mode
 

Neural nets? Whoa, calm down. =)

Isn't sending input to the robot also part of the challenge? I mean, I guess you could argue that your input would be coming from the outputs of your program, but still...

15 seconds is a pretty good length of time for autonomous mode. It can either be a really long amount of time or a really short amount of time, depending on what your robot can (or can't) do. It's a good amount of time to let the teams with the resources and abilities to show off some cool stuff, while not leaving some of the less experienced teams in the dust.

I'm sure plenty of people would love to work with things that "sophisticated," but sometimes it's good to remember that simplicity is the greatest form of sophistication. The definition of true innovation is something that accomplishes a task or solves a problem in the simplest way possible.

And that's what we're all going for, right?

I mean, isn't efficiency the main goal of everyone's robot? :p

Re: Future of Autonomous Mode
 

I don't think that the hardware this year is a problem at all. Even if you do outgrow it ,there is nothing preventing you from adding somethign more sophisticad to augment it. The real challenge(at least for out team) is the six weeks. If this were extended ,then i could see alot more happening. For now i think we continue to see improvement, but only as much as the time limit allows. If Autonomous doesn't change all that much, then this is one area that teams that work during the off-season will have a signifigant advantage.

There also comes a point where you have to consider what is really needed to complete the task at hand. I think we would all love to see more sophistication, but FIRST really isn't a very sophisticated competition.
Its sad but true that elaborate autonomous systems are for the most part unnecessary. I saw lots of robots whose autonomous modes did little more than "go forward for x seconds" "turn for x seconds" etc. These worked fine and got the job done. If we want to see more sophistication then i think two things need to change. first, i think that the build time would need to be extended. Second i think that the complexity and scoring benifits of autonomous would need to be increased.

I could also see some room for upgraded hardware, but nothign more sophisticated than a real DSP with floating point math and maybie a little more memory. Even that is for the most part unnecesary. If you know how to use it, the current hardware is perfectly fine so long as the provision for custom circuits stays in the rules.

Re: Future of Autonomous Mode
 

Which is why it is there. So that old teams don't leave rookies in the dust.

Re: Future of Autonomous Mode
 

lol, I just see all my friends doing all the mechanical stuff so eloquently and yet I can't visualize any mechanical system for beans. I want to be more useful because I love the concept of this program so much!!! Ohh well, we had so many mechanical problems as rookie team as is. Part of that came from serious underfunding, 0 machining and tools in the school, and minimal mentor/adult support. It was too bad, there is just so much potential in our school that was just sitting there, waiting to be unlocked. There's one thing for sure, I sure ain't letting FIRST slip by me next year!!! J-WALKER II (this year's bot was called J-WALKER) will be back for vengance!!! :) :D :D

Re: Future of Autonomous Mode
 

It really is not a matter of programing in my eyes. It is a lack of sensors. You really can't know where you are/going unless you can see. The sensors FIRST gives us are a good start but more can be done. I have schematics ranging from ultrasonic sensors to metal dectors suitable for detecting goals. I also do not see much problem working on autonomous off-season. Just as long as you:
A) Offer whatever you build to all teams.
B) Rebuild whatever you develop during the six weeks.

Re: Future of Autonomous Mode
 

Not even a wood shop? We used fiberglass because we don't have much as far as aluminum goes. I think many teams use a sponsor's shop.

Re: Future of Autonomous Mode
 

Man, all autonomous would be scary..... Though I would like to see more from autonomous, because I really like a really hard challenge that keeps me up late at night, beating my head on walls, and screaming at the computer. I think that there is alot of potential for what you can do during autonomous, but I didn't see anything spectacular on the mathces I got to see at either of the 2 regionals we went to :(.

Re: Future of Autonomous Mode
 

I prefer challenges that I feel like I'm that close to getting it perfect, but I'm missing something.
Me too.

And longer means we can line follow. :D

Re: Future of Autonomous Mode
 

I havent seen one autonomous that took advantage of the capabilities of the upgraded chips. Until the programmers in this competition start learning this stuff, I think FIRST should just give use a handlfull of logic ICs for autonomous. And secondly, You dont need Pentium 4s and webcams for image processing or AI. :ahh:

_________________
TheD00d
Keep on chooglin
_________________

Re: Future of Autonomous Mode
 

to be perfectly honest , I think it would be difficult to run a whole auto match , strictly because some teams dont even have a programmer. this is obvious because many teams did not have an autonomys mode at all. its hard to use the upgraded chips because their were only a few things that would be favorable in auto mode.... like stuffing the ball chute with the mobile goal , hanging , or going for the 10 point ball. all of these had an extreamly simple way of getting it done, and dead reconing never failed on my robot.
a pentium 4 type processor would be pretty impossible to use, no teams would ever use it to its full capacity ... I think that the chips we have are efficiant enough for the game we had last season. :yikes:

Re: Future of Autonomous Mode
 

Who needs webcams? To be perfectly honest, a pair of contacts on the front for continuity are great to find out if you are against a bot or the plexiglass wall, in addition to bumpers and what-not. Big companies are doing image recognition, I know I'm not ready to try it.

The only major improvement I can think of for the controller is a floating-point unit so we can do trig. Other than that, it's cool.

Re: Future of Autonomous Mode
 

True, the sky would be the limit for autonomous mode, but the sky would also have to be the limit for weight, and build time :-D.