One program!

Team 2710 ran the board at the Flint MI tournament on one program. When they were at my table, they scored a 390 - the power lines were 1/4 inch off so they only connected 2 properties. They had other problems on other rounds during the day. An amazing bit of programming. They would come back to base and wait a few seconds, as the team gave the robot new pieces, without ever touching the robot. Sometimes the team pre-positioned pieces in base and the robot snagged them as it passed by. They even released elements on the field and then regained control of them for better positioning.

Great work, Brother’s Keepers!

Cool! Anybody have any video of it? NXT or RCX?

NXT. I was a bit busy with reffing to do any video taping. If the coach responds on the FLL forum, I’ll suggest they post one.

Wow, that’s great to hear! Congrats to them!
I have seen a lot of really interesting programming techniques from FLL teams this year, but one of my favorite by far is the use of the touch sensor. A few teams used the sensor to allow them to quickly switch between programs, saving them precious seconds. But I’m very impressed by this team’s LACK of need to switch at all!

I second that video request!

We heard there was some interest in watching the video.

When we watch it all we can think of is all the hair we pulled out but we would do it again in a second.

http://video.google.com/videoplay?docid=-4765357511095837007

P.S.

Gary Voshol is the one who caused this in the frist place.
He made a post that said it could not be done.
There is nothing like a challenge.

You’re going to make me go back and look for my exact words, aren’t you? I know I said that last year there was a New Hampshire team that did it all on one program because Skye Sweeney said it couldn’t be done. I don’t recall if I repeated that doubt for this year or not.

Great job, 2710!

No problem we will quote you.

"A team in New Hampshire accomplished all of last year’s missions in one run. That is a very rare accomplishment. If you search through the archives from last year, you may find a link to a video of it.

A problem with running only one program is, what happens if something goes wrong and you have to restart? The robot will have to waste time retracing its steps on the missions or parts of missions that have been completed before it can try the one where it got off track. The reason the team in NH even attempted it was because one of the Forum’s premier posters, Skye himself, had the opinion that it couldn’t be done and they wanted to prove him wrong.

A team from Wisconsin completed all the missions without having to make any modifications to their robot (no changing arms, attachments, etc). They didn’t use just one program without touching the robot in between. If I recall correctly, that was during the Ocean Odyessey challenge 2 years ago.

This year it would be even harder to do everything on one run, because so many of the missions are linked together with multiple parts.

You certainly can write a program that will accomplish several missions at the same time. In fact, this would likely be the only way to complete a significant number of the missions. There’s just not enough time in 2.5 mintutes to waste it coming and going from Base for every mission individually. Pick 2 or 3 missions in the same vacinity of the table, and see if you can accomplish more than one of them using the same program.

  • Gary"

Beautiful!

Did you guys enjoy the rest of the competition as well?

Yes, we placed second overall.

The team worked real hard and knew they had done well in all four parts but when they called our number for second we all just looked at each other and wondered if they got the numbers mixed up somehow.

We have been doing this long enough to know there is always somebody who has come up with something so fantastic that there is no way to match them.

But no matter how you place, the thing to watch for is the character change in the team members.
The trophies are only important for a few hours.
The character that is built will last a lifetime.

Duane

I think you are right, Duane.
Congratulations again on what sounds like an outstanding competition.
Jane

Thats amazing.

Can somebody from that team post their code so future teams can see what they can do?

Does NXT have a text based language yet? Or are you stuck using the buggy drag and drop one that comes with the NXT kits?

Needless to say, I’m very, very impressed. I did FLL for 5 years, and am well aware of the challenges that you face. I like how you bounce of walls to find your position, we always found it to be faster, easier and more reliable than tracking lines. And perhaps most impressive is that you managed to do this reliably. We always split our missions up into multiple programs not only for attachment changing, but also so if one turn messed up our whole program wasn’t shot.

The only legal languages are NXT-G, what comes with the kit, and Robolab, version 2.9 now supports the NXT.

I would also love a look at the program, as I am interested in looking at if the robot compensated other than touching the walls.

Our team also used one program. At the warm up, we had the one program, and also all of the single missions loaded as separate files as well so if something went wrong, they could also use the individual programs. It did go wrong, and the time did not allow for manually changing between programs.

The kids came up with a one program menu system that allowed for you to go back ad run a mission if it failed. Watch the state round 2 video, Wes ran the truck grab 3 times before going to the next task.

Our programs and robot pics are on our site, http://www.oxfordenergy.us
I also put up a link to the brotherkeepers video, as some other FLL coaches were looking for it. Let me know if there are any issues with this, and I will pull it from my server.

Scott (programming coach)
OxfordEnergy #0234
(4th Overall Novi Mi States)

I’m pretty sure you can only use four sensor and on the side description thing it said 2 touch, 2 light, and the ultrasonic sensor and if I’m wrong someone please correct me.

The robot in question does indeed have more than four sensors. To accomplish this they used the old style RCX sensors and the NXT-RCX adapter cables. With the adapter cables you can use the RCX style piggyback trick to place a touch and a light sensor on the same input port.

-Skye

Is it legal?

With the NXT you’re allowed 2 light and 2 touch sensors, the ultrasonic sensor. Plus 3 rotation sensors, including the ones built into the motors.