http://www.nandotimes.com/technology...-6056222c.html You probably wanna check that out. It's really interesting, it's autonomously run robots playing soccer (or futbol :)). They have no human interaction once they go.... compared to what that league does, we barely compare in the sense of autonomous... but the thing is... if we all worked a bit more and taught ourselves and had to literally have programming teams work to learn (which seems to be what is happening from autonomous, and that's an excellent thing I think) and make robots that can work with their environment, not just timers... and react to other team's actions, it would really give us a real-world experience and we'd learn lots more about what happens in engineering.

They also have resuce robots which are pretty cool.
Here is another news link on it:
http://http://www.usatoday.com/tech/...o-soccer_x.htm
Here is the website for the college people:
www.robocup.org
Here it is for the junior devision:
http://satchmo.cs.columbia.edu/rcj/rcj2003/

If you want to see what I thought was the best autonomous (I may sound biased, but it was 68 that did a majority of the work ;))mode run by KoH bots, here is match 63 at the championship between 45 and 68 against 118 and 541. 118 got tipped early, but 541 played tough and only lost 63 to 61. Excellent match.

Hey the best part is that the human player mode features me, although that was a more routine stack round for me....not my best round ;)

http://www.soap108.com/2003/movies/gal/gal_063.asf

Sure, post the round we got tipped! Er... one of. My favorite autonomous round (no video, sorry) was the first one we were in on Friday at Lone Star - the robot made it to the top of the ramp and locked down (or at least it does in my memory) after a complete transmission rebuild the day before. A few of us weren't sure the thing would even roll.

Another evil idea
 

Hmmmm reading through these posts I have just two comments..

If we have Middle School Kids doing full autonomous in Lego League competition...why is it that our high school teams couldn't do full autonomous..??? I am not advocating that position just asking the question....I do believe with the right tools and the right game...full autonomous might be very interesting...


(2) What about an event in which the human controllers can only "see" through the eyes of the robot? The robot would need a variety of sensors...including a camera perhaps...but human control would only be through the senses of the robot for positioning, etc. This would require some very interesting options for AI robots in which gross errors could be compensated for by feed back to HUMAN controllers....might be very interesting too...

Just ideas...
not rumours...

Bob

Re: Another evil idea
 

Yes, but as has been said before, Lego League teams don't have to worry about having 3 other robots on the field during their matches. It's a lot harder to do a fully autonomous game when you have 3 unpredictable machines out there with you.

Another neat approach might be to make the player stations opaque and allow driving/manipulating -only- from a robot mounted camera.

This would be a more real-world tele-operated example. As with full autonomy, it would require a complete change from the existing controller.

On another topic...if we had better sensors and a better control unit, more teams would integrate controls, signal processing, and autonomy into their existing robot.

Autonomous code encouraged teams to work together with their programmers. I think actually...from a team standpoint...it brought the programmers even more into the picture. Teams were required to figure out how they were going to do the autonomous mode. In 2002 a group of programmers got together and started talking about their positions on their team. Although not everybody felt this way...there were a good number of them who felt like they were a bit separated from the rest of the team. With the advent of autonomous...the programmer suddenly became a bigger consideration and programming became more important if you wanted to move during autonomous mode. There were other teams where the programmers were really incorporated well into their teams, and they worked well together. The programmers I met from those teams tended to have actually done some bit of tricky code to make their bot work in a special way, or they had incorporated something neat that their program processed the input and output of.

Although this years competition was structured such that there was an emphasis on cooperation, I think they should take that one step further for next year. For example, some task that required cooperation to complete for a bonus.

Wireless communication between robots on the same team would also be neat. So they could talk to each other durring autonomous mode.

I like the opaque shield and having all control is through sensors/ a camera on board the robot.
I think the fully autonomous would be not interesting to spectators.

I would hate to burst your buble but there is a competition like lego league where there is six robots on the field at any given time.

I completly agree, an opaque shield with the only interaction being through sensors would be cool, but... the problem remains that the teams could look up to the big screen TV to get visual aid and removing the TV's cuts way back on the spectator quality of the competition.

We should definitly keep autonomous either way though.

Just my two hundreths of a fourty three Haitian gourds (2 cents)

Programming problems
 

Sorry, this is a bit off the subject, but how many teams actually were able to test their autonomous program before the practice day at regionals? Since we dont have a lot of money or any space, we couldnt build an adequate set-up in order to test our program. Since so many things effect the program (like carpet and so forth), without that set-up, it was hard to do any testing. So when we got to our first regional on the practice day, we had to just put our robot on the field, turn it on, and hope it didnt kill someone (which it nearly did the first practice round we used our arm). Since you can't keep tweaking your program and downloading it to try it out until you get it right, like you can if you are at the workshop practicing on your own set-up, we found it very difficult to get our autonomous mode to do what we wanted it to do. We initially wanted to have our arm sweep across the ramp and then have our robot drive forward, turn and start heading up the ramp, but we could not for the life of us get our bot up the ramp. Every time we had it written for the robot to move after we swung the arm, it was a disaster. So we did the only thing we could, and just cut that part out, preventing any further threats to innocent by-standers' safety. But i was just wondering if most teams ran into the same problem that we did? Thats basically the only problem i have with keeping the autonomous mode.

We were one of the teams that got a great deal of time to work with our programming. I noticed that many teams at least at the regionals I attended, did not have the time or resources to do any testing. I think that this puts teams at a disadvantage. What I would recommend, to buy software that can test "artificially" on the computer you're programming. I would contact university's that may already have the software, and may be able to lend you a copy. That way you can walk in with an idea of where to begin. With practice it is hard, and the more testing you have the better. If you guys ever need help with you're program give me an e-mail and I would be more then happy to help in any way I can.

Like many other Teams we were only able to run autonomous once on the robot prior to shipping. While auto worked out of the box at the regional, at the first practice the robot went through the stacks like a bullet with the arms going the opposite direction and not opening and smashed into the driver wall at full speed (we thought there'd be other robots to push against). We spent all the practice sessions and the Qualifying Matches fixing, modifying, fine-tuning and playing with different autonomous programs. In our video tapes we can really see the programs changing drastically at first then gradually demonstrating smaller and smaller corrections. We only stopped changing the code when Finals began.

We were lucky enough to have a dedicated core programming team of two students and two mentors, so we were able to spend a lot of time on development, but it was still the first year for all of us. We foresaw not being able to test on the real robot and countered the problem by designing autonomous to run as a script. That allowed us to safely make big behavior changes at the first regional.

Full autonomous would be quite a challenge over and above the navigational issues. New teams would not have the background to know how to integrate the sensors through a custom circuit or even that they must to get the accuracy required by match long autonomous. FIRST would have to formalize a standard solution to this that all the teams have. Then the more experienced Teams could add additional capabilities, but every Team would start with a basic level of competitiveness.