Lessons Learned - The Negative

[Koko Ed]

Originally Posted by Chris is me

I do have a negative!

Why did this game feature and emphasize camera tracking of non-lit targets when lighting conditions were inconsistent between regionals? Hell, in Atlanta the lighting conditions changed with the number of clouds in the sky and this screwed up robots. Vision Targets should be lit if they're going to be featured.

So next year FIRST puts a Weather control device in the KOP and everything will be fine.

[martin417]

Several teams on Curie were able to track consistently with the camera, some scored in autonomous, notably 27 and 1771. It was a challenge, but it was possible... at least until 5:08 PM (when the sun went behind part of the dome).. then it wouldn't work at all.

[Nate Smith]

Originally Posted by BurtGummer

Once again, I will state that FIRST needs to look into something here. We are not the only school who has started out in FRC and not FTC or FLL. I did not set our team up, our sponsor rep did. We didn't even know what FRC was. All I am trying to say is that there are two different types of rookies in FRC. Rookies with prior robot experience in FTC and FLL, and complete newbies to the entire field of robotics. I think that is a fair statement, and not one which criticizes any teams that do have prior experience. It isn't their problem, it's something FIRST should look into.

I apologize that my post sounded like I am criticizing these teams. I found the facts about their history, and I applaud them for it. It is a deal with how FIRST is handling it, not them.

Unfortunately, once you decide to call an FTC team who transitions to FRC for the first time anything other than a rookie, you get into a whole mess of other "what ifs." Just a few that came to mind include:

• A team who has competed in VRC, but then jumps straight to FRC
• A team who has previously competed in Botball
• A team who has previously competed in BEST
• A team who has previously competed in NRC

This only includes other high school level competitions, and doesn't take into account other factors, such as an existing robotics curriculum at the school, or previous individual experience of any member of the team.

To look at it from a different angle, look at the sports world that FIRST is trying to some degree to emulate. I don't believe (and I may be wrong at this, as I don't really follow sports to this degree) that you would find any pro sports program who would say that a "rookie" player would only be considered one who had never touched the ball before in their life.

[synth3tk]

That's true. A rookie to the NBA may have played ball since middle school, but they still call him a "rookie" because he never played for the NBA in his life, yet he still played the game sometime before.

[EricH]

Originally Posted by Nate Smith

To look at it from a different angle, look at the sports world that FIRST is trying to some degree to emulate. I don't believe (and I may be wrong at this, as I don't really follow sports to this degree) that you would find any pro sports program who would say that a "rookie" player would only be considered one who had never touched the ball before in their life.

If any pro sports program did that, they'd probably be violating child labor laws...

To even get into the pro sports leagues, you have to be at least decent. That means you have to handle the game object a bit...

[FRC4ME]

Originally Posted by EricH

If any pro sports program did that, they'd probably be violating child labor laws...

To even get into the pro sports leagues, you have to be at least decent. That means you have to handle the game object a bit...

True; I don't think anyone has ever picked up a basketball for the first time in his life and instantly been a star player ("hey, I think I have a knack for this!").

[Greg McKaskle]

I'm going to echo this.
If FIRST is going to provide teams with an expensive piece of technology and even more expensive controls designed to work with it, they should ensure that said technology should be not only possible, but plausible to work with at each and every event/venue.

The lighting was indeed challenging, but I don't think the clouds were that much of an issue. I was measuring the light level at various times during the day on Saturday. The light level measured at the front of the stands stayed within a factor of two of the spec level. I wasn't able to check on other days, where it may have had more impact. On the other hand, when the timeout was called on Einstein, the disco lights started going full tilt. Good thing teams weren't trying to calibrate their camera during their timeout.

I can't directly compare to previous years that used the camera, but levels were checked at the majority of the regional events, and only needed adjustments a few times to bring them into spec. That doesn't mean it was ideal, however. The dark curtain, the lack of conistency in the lighting in the background, and the spotlight positioning made it difficult for the camera to meter the light and get a good exposure. With the data and experience gained, I think the camera will be better utilized in coming years. Additionally, having a moving target being shot at from a moving platform with a slow projectile is more than just a vision problem. Personally I think the lead estimation and the lack of scoring benefit was the primary reason for the camera being underutilized.

While discussing the event lighting, I'll also mention that few teams took advantage of the lunch hour for measuring the field or calibrating the camera. At the three regionals where I assisted with this, three or four teams gave it a shot. In Atlanta, the red seats and pink target were close to each other in color, but they could be differentiated with a small change to the hue threshold. I don't know how many people asked the NI AIs, but only two teams asked me for help with vision. The vision sessions I presented were also not well attended. IMO, inconsistent lighting isn't the only issue.

Greg McKaskle

[MrForbes]

Greg--we gave up at our first regional....although Steve gave us a lot of help that Thursday at lunch, we didn't have time to get the camera working reliably, so we changed strategy and everything worked out ok.

[Alex Cormier]

The final 4 setup, show and all was LAME. Where did the 4 division champs robot parade go? Where did the big entrance for Dave go? Why were there so many speeches in between the matches? The big finish was merely a boring cable show.

[GoSparx]

Originally Posted by Alex Cormier

The final 4 setup, show and all was LAME. Where did the 4 division champs robot parade go? Where did the big entrance for Dave go? Why were there so many speeches in between the matches? The big finish was merely a boring cable show.

Agreed. My best memory of FIRST Championships is Dave V. riding out in a car to ZZ Top's Sharp Dressed Man. Best ever.

[Lil' Lavery]

Originally Posted by Greg McKaskle

The lighting was indeed challenging, but I don't think the clouds were that much of an issue. I was measuring the light level at various times during the day on Saturday. The light level measured at the front of the stands stayed within a factor of two of the spec level. I wasn't able to check on other days, where it may have had more impact. On the other hand, when the timeout was called on Einstein, the disco lights started going full tilt. Good thing teams weren't trying to calibrate their camera during their timeout.

I can't directly compare to previous years that used the camera, but levels were checked at the majority of the regional events, and only needed adjustments a few times to bring them into spec. That doesn't mean it was ideal, however. The dark curtain, the lack of conistency in the lighting in the background, and the spotlight positioning made it difficult for the camera to meter the light and get a good exposure. With the data and experience gained, I think the camera will be better utilized in coming years. Additionally, having a moving target being shot at from a moving platform with a slow projectile is more than just a vision problem. Personally I think the lead estimation and the lack of scoring benefit was the primary reason for the camera being underutilized.

While discussing the event lighting, I'll also mention that few teams took advantage of the lunch hour for measuring the field or calibrating the camera. At the three regionals where I assisted with this, three or four teams gave it a shot. In Atlanta, the red seats and pink target were close to each other in color, but they could be differentiated with a small change to the hue threshold. I don't know how many people asked the NI AIs, but only two teams asked me for help with vision. The vision sessions I presented were also not well attended. IMO, inconsistent lighting isn't the only issue.

Greg McKaskle

Greg,

I certainly don't think I had intended that as a shot against the camera or technology or support that NI supplied this year. It was NOT intended that way.

I agree that the game challenge was the biggest cause of the camera being underused by teams, but I feel that has as much to do with the specific choice of target teams were asked to follow. Similar to the last time we had an unlit target (2005), there was minimal success by teams. However, with a backlit target in 2006 ans 2007, success was much more widespread. This is especially true in 2006, when the value of tracking the target was readily apparent and incredibly high.

[Greg McKaskle]

I certainly don't think I had intended that as a shot against the camera or technology or support that NI supplied this year. It was NOT intended that way.

I didn't take it that way. Sorry if my response was a bit negative. I was delighted to see so many videos of people tracking and playing with the stuff in their shops. I met with a number of students that went pretty deep and wrote their own code, so all in all, I think that the learning happened even if the game wasn't a great showcase for it.

Greg McKaskle