View Single Post
  #70   Spotlight this post!  
Unread 16-01-2007, 09:09
Assassin Shadow's Avatar
Assassin Shadow Assassin Shadow is offline
Inventor of the Shadow-gram
AKA: James Bilger
FRC #0342 (The Burning Magnetos)
Team Role: Programmer
 
Join Date: Oct 2005
Rookie Year: 2005
Location: South Carolina
Posts: 16
Assassin Shadow is a jewel in the roughAssassin Shadow is a jewel in the roughAssassin Shadow is a jewel in the rough
Re: Advantage given by succesful autonomous mode

Ok. Let me take these arguments one at a time...

First...
Quote:
Originally Posted by Salik Syed View Post
You are overlooking a few things, it is true the camera can track two targets, but how do you know WHICH of the two targets your ally will go for???
Ask them of course! Which of the two? There may well be three visible to the team...

Secondly...
Quote:
Originally Posted by Salik Syed View Post
The rack is rotated randomly and there is no reliance on what the other teams program might do,
and while I agree, it is certainly possible and not even ridiculously hard to figure out a way to work this out. I just don't see many teams doing it. You talk about implementing vision like it is a very simple task, but realistically it is not.
Nothing is necessarily easy, but everything is possible...Like my dad and grandpa always tell me, "There's no such thing as impossible, just harder."

The rack is rotated randomly, however, why does the orientation matter? If you know that the lights are 90 degrees apart and if you find out at what distance you can only see one, then you can find out exactly where you are in relation to the spider legs. You could even set it up to make itself drive until it is equidistant between two lights, then you're aimed at a leg in between two lights. Remember that scoring is not dependant on the rack's orientation on the field, it's dependant on the rack existing at all!

Next...
Quote:
Originally Posted by Salik Syed View Post
Last year I saw about 50-60% of teams do something in autonomous mode (at our regionals) most teams were unable to accomplish anything other than defense. If only 15-20% of teams can program a robot to go in a straight line and dump balls in a goal efficiently what do you think that says about implementing a succesful working vision system?
I think it says that teams should overlap their jobs. On my team, we have the programmers playing with the vision system attached to an old robot, even though this years robot is still being built and the design isn't anything near being finalized yet. Therefore the program team can have several weeks to practice while waiting on a robot to program.

And on to the math...

Quote:
Originally Posted by Salik Syed View Post
Lastly you said that if you "do the math" the ringers have only 6 inches of play in each direction.
Do you know how incredibly much that is? you are attempting to put a ring with a hole of 13-14" (?) diameter onto a rack whose cap is 10" in diameter.
I don't think 6" is "very little".
Robots that cannot score in autonomous will move the rack and make it incredibly difficult if not impossible to score.
Ok, here is how I reached the conclusion of about 6 inches. Try and follow me here... If the inside diameter of the tube(the hole) is 13 inches, and it is being put on a target 10 inches in diameter, then there is 3 inches(1 1/2 in any direction) right there. Now, take into acount that the tube is beveled or curved. If you attempt to push it onto the target, even if it is not exactly on target, the spider leg(which hangs from chains) will move so that it fits into the hole. Soooooo...if the diameter of the tube(cross section) is 9 inches, then there is about 4 1/2 inches where it will be pushed onto the leg, now if we tone it down a little so that you are more likely to be on target, then we'll call it 3 1/2 inches added to the 1 1/2 originally is equal to 5 inches of play in any direction. Understand? If not, say so and I'll try to explain it better...

And Lastly...

Quote:
Originally Posted by Salik Syed View Post
Lastly, experience does have something to do with it. I'm not saying a rookie team can't pull it off because you're right skill and knowledge will result in sucess. However the vision system is one of those things where it is of great benefit to teams who have the resources to have a practice robot to test code on and prototype. Most rookies scramble to get the bots together, and don't have time. I'm not saying vision is a bad thing but I think more options for play during autonomous would be better (other than just defense) ... I think 2006 was a great example: You could score in side goals, block or score in the center goal
Please don't compare the game from this year to the one last year. They are very differant. The cams are, however, extremely similar. So yes, experiance will help veteran teams. But, like I said before, if the programmers/electrical team put together the camera and two motors with the electronics, then they can play with the code until they have an actual robot to program. Always remember that you don't have to wait to test something like that! As a matter of fact, if teams havn't started putting together the camera and playing with the code by now, then I don't know what to think. Remeber there is a time limit, so get what you can done while you can!

Ok, sorry it was a little long winded...
__________________
Team 342

"If it doesn't move enought - WD-40, if it moves to much - Duct Tape."
-page 1, Life's Troubleshooting Guide