They scraped the camera :-(

[Dr Nick]

For all the hype they put on the camera the past two years, I would have sworn they would have done more with it this year somehow. I was really looking forward to working with it again (though would liked to see something other than the green light again)

[bear24rw]

I completly agree with you, im not too happy about them not having the cmucam again..

[Kyle Love]

From my past experience with the CMU cam with my team, I found it to be something to be put on the put-a-side list and done as an afterthought more less. I will say its a let done to work for 3 years now and not have it around anymore.

[Mark Pendergast]

Sorry to spoil the party, but I am glad the camera is gone (at least for a while.) We have never been able to get the camera to work consistently. We have a small team and have never had the time or resources to debug it. We had autonomous software that would have worked if the camera was not checking out the overhead lights or other points of interest, but we rarely got a consistent lock on anything. I believe that has been the experience of the majority of teams.

<RANT >
From what I have seen, the camera has been a failure. It is balky and unreliable. In 2006 and 2007 very few teams at the Boilermaker were able to make it work consistently. (Even with the 'best' teams at IRI in 2007, autonomous ringers were few and far between.)

Also, they have not made a big difference in the game. In 2006, most teams used dead reckoning - ignoring the camera. In 2007, very few teams even tried to score, let alone made a ringer - and the advantage for doing so was minimal.
</RANT>

I am excited about the new IR challenge. I hope that it will be reliable than the camera.

[bear24rw]

idk, for all the teams who say its was tracking all the overhead lights and stuff you just have to spend a couple hours and play around with sataturation, brightness and other values in labview, when we got ours calibrated it was the key to our success in autonomous mode in 06.

[Roger]

In spite of my programmers, I am a little sorry it isn't in this year. But of course it took us 2006 and 2007 to understand it. And yes, the field lights are not forgiving to find that ol' green light. I think, in terms of programming, the hardware has to catch up with the software on this part.

[Tottanka]

Originally Posted by Mark Pendergast

Sorry to spoil the party, but I am glad the camera is gone (at least for a while.) We have never been able to get the camera to work consistently. We have a small team and have never had the time or resources to debug it. We had autonomous software that would have worked if the camera was not checking out the overhead lights or other points of interest, but we rarely got a consistent lock on anything. I believe that has been the experience of the majority of teams.

<RANT >
From what I have seen, the camera has been a failure. It is balky and unreliable. In 2006 and 2007 very few teams at the Boilermaker were able to make it work consistently. (Even with the 'best' teams at IRI in 2007, autonomous ringers were few and far between.)

Also, they have not made a big difference in the game. In 2006, most teams used dead reckoning - ignoring the camera. In 2007, very few teams even tried to score, let alone made a ringer - and the advantage for doing so was minimal.
</RANT>

I am excited about the new IR challenge. I hope that it will be reliable than the camera.

I have to dissagree.
I think that the reason that CMUcam2 has been removed, is because it was mastered.

[mcmouse]

We had the camera working pretty good - TX to the programming team.
an i would have like to see something like it again, sience we always had trouble other sensors used for navigation.

I had the idea to use the camera with the preprogrammed tracking code to as a carrier forthe IR sensor. then give the ROBOCoach a green light at his location. the camera logs on to the light, no matter were the robot goes. and the IR comunication can be used.
what are your thoughts, esp. about a big green light for the Robocoach?
and if the 2007 camera would be allowed in 2008?
or any other ideas to use the IR sensor while the robot is driving and being ushed arround?
The CatAttack 451 - Jens

[pheadxdll]

I don't know about you guys, but I'm happy I'll never have to mess with the CMUCam again. It was really unreliable and flaky. It was the cause to many late nights fighting with it follow a green light. Eventually I got a routine down that found the light and hung a tube most of the time but once we got onto the competition field, the robot had a hard time finding the green light amongst all the ambient lighting in the station.

Good-bye CMUCam (for this year), you won't be missed.

[Koko Ed]

Originally Posted by Tottanka

I have to dissagree.
I think that the reason that CMUcam2 has been removed, is because it was mastered.

By what? 2% of the competitors in FIRST?

[Zyik]

We had enough programming problems that for the most part we didn't even bother with the camera. I'm glad we can finally move onto a new challenge.

[King Duke]

I don't think they've scrapped it at all. the CMUcam is still a totally viable system, we just have to buy it ourselves now and include it in expenses. I think it could be used for places other than tracking green lights. Maybe with the correct filtering and what not you could track an alliance ball. Also 1 of the 2 balls has white spots on it, maybe the camera could lock onto that.

of course this is pure conjecture and has not yet been tested, nor have we had much success with the camera in the past.

I think FIRST is trying to get more teams involved in making their vehicles more robotic and so they met us half way with this hybrid mode.

anyways, good luck guys.

[Chris_Elston]

Originally Posted by Dr Nick

For all the hype they put on the camera the past two years, I would have sworn they would have done more with it this year somehow. I was really looking forward to working with it again (though would liked to see something other than the green light again)

You can still use it. We are doing some testing, but the Camera is a great sensor. We've had no problems using it.

In 2005, was tough, though we did manage to find the Vision Tetra on the field, lighting was the main problem.

In 2006, we nailed the auto mode every time and if someone hit us, we still tried to correct and get a second aim. We also used the camera in tele-mode to "auto aim". Driver pulled the trigger and let the robot take aim, rather than human take the aim, worked pretty good for us.

The only troubles we noticed was the backup battery. We changed the backup battery after every match in 2005, then we we got the rechargeable circuit in 2006 for the battery that helped out ALOT...Camera was rock solid both years for us. In 2007 we was a lifter only robot, so no camera for us. we had no grippers to handle rings.

[David Fort]

Originally Posted by mcmouse

...
I had the idea to use the camera with the preprogrammed tracking code to as a carrier forthe IR sensor. then give the ROBOCoach a green light at his location. the camera [locks] on to the light, no matter were the robot goes....
what are your thoughts, esp. about a big green light for the Robocoach?
and if the 2007 camera would be allowed in 2008?
or any other ideas to use the IR sensor while the robot is driving and being [pushed] arround?
The CatAttack 451 - Jens

We were kicking this same idea around.

I worried that this might be considered sending more than 4 discrete pieces of information from the RoboCoach to the Robot, but frankly, I think the idea is so cool that it would be worth getting DSQd just to see it work! I promise that if you call me as a witness at your trial, I will testify that wearing hats with big green lights is all the rage among teenagers these days, and certainly can't be construed as trying to influence a robot.

I am not an authority on it, but it seems to me the Robot rules preclude use of the CMU camera as shipped in previous KOP because it is a custom made part for the FIRST competition.

For example, here is a piece of <R36>
The item must not be a part custom made for the FIRST competition and provided in the
Kit Of Parts for a previous FIRST Robotics Competition (e.g. 2006 FRC transmissions,
custom-made motor couplers, custom sensor strips, 2006 IFI CMUcam II modules, etc.
are not permitted)

And I agree with both sides of the comments here: it was really hard to get the CMUCam to stop looking at the MountainDew signs and stage lights, but now that we've figured that out, we wish it was integral to the competition again this year. And the other side I agree with is that now that we've figured it out, it is time to remove it and put some other stress on our brains. (but I really hope it comes back again in the future).

I think the right answer is to have something you could do if you can get the camera working, but also have some useful things you can do even if you can't.

For this year's game I think a good twist would have been to allow the camera, but not have the green light. That will present a whole new set of calibration issues (similar to the Tetra challenge) if you want to use it to figure out which ball to knock off, or whatever.

I don't see any preclusion against using cameras this year, it seems you just aren't allowed to use the one all the veteran teams already have. I can understand that as a way to prevent Rookie disadvantage, but hey, we Vet teams are broke too. :-(

[EDIT: I read on some other threads that they think just the 2006 camera is explicitly disallowed, and the 2007 camera is OK because you can buy it COTS: http://www.ifirobotics.com/camera.shtml - - again: only imho]

[baop858]

I am quite disappointed as well about the disappearance of the camera this year. There was quite a bit of a learning curve but teams that have the resource to do an autonomous mode should have been able to pick it up by at least the 3rd year. And last year letting us go on the field and calibrate the camera did make tracking a lot easier during competition.