Autonomous Longer????

[Pavan Dave]

Would you support a five to ten second longer autonomous period if FIRST decided to implement it. I feel that many programmers do not feel it is worth it to spend their time for fifteen seconds and that if there was a little more time than more people would find it more important and thus making more people work on autonomous.

Please give your input on whether you think it should be made longer or not.

Pavan.

[Joe Matt]

Same said about build time, game durration, etc.

You can always have more time, but the point of a challenge is not having everything you need and still trying to finish it.

I don't support making auton longer and I think it's fine enough time, and those that won't score a keeper now won't score a keeper if there was an additional 5 or 10 seconds.

[Dan Petrovic]

How often do you see autonomous modes get cut off due to the time limitation?

Even last year, with the shorter autonomous mode, there were robots sitting around for a couple seconds.

I think it's fine the way it is.

[JaneYoung]

Just from trying to explain the autonomous period to folks in the stands that come for a day or two (spectators) - it is difficult to explain. And it can be pretty boring. 5 to 10 more seconds would not help make it any more exciting at this point.

[Not2B]

Sure - I think it would be a nice change. Not too long, mind you.

I'm still (waiting for the day) (wondering when will be the day) (fearing the day) FIRST puts automous at the end of the game.

Big *stinking green lights and cameras makes that very doable. Double points for anything scored in the 15 seconds AFTER the human quit?

15 before, and 15 after?

*The lights don't stink. I like the lights. They are like robot mood lighting.

[thefro526]

I think a longer autonomous mode would be great. Autonomous this year could be anywhere from 5 to 15 seconds longer and still be good. As long as the autonomous mode is not like FVC last year (45 seconds was a bit excessive) autonomous mode would be great.

[Bongle]

No, I like it how it is. Many teams have large difficulties simply doing autonomous, but that isn't because of the length of time given, it's because they find it simply impossible to do whatever action it is in any amount of time.

I think the best way to improve autonomous participation rates would be less fragile or easier to use autonomous-focused equipment. On both teams I've been on in 2006 and 2007, the camera broke or would inexplicably stop working pretty frequently. It takes a long time to set up, a long time to test, and requires a large area to practice in properly. Laptops with serial ports are growing fewer and fewer, and the ones that my teams have been able to afford typically have nonfunctional batteries, so on-field debugging is impossible.

I would LOVE an end-of-game autonomous mode. You could have the end-of-game bonus be being some specific point on the field at the end of the teleoperated period, and at another point at the end of the ending autonomous period.

Problem with an end of game one would be that you'd have issues with game damage possibly getting exacerbated by the robot doing things autonomously. For example: in 2004, our robot had two high-torque motors driving seperate worm gears that moved the arm. Once, one of the wires came undone but the team didn't notice, and the unsynchronized worm gear action actually torqued the entire arm. With start-of-game autonomous, you can disable the autonomous mode to not damage things you've found are damaged. But if match damage breaks a chain or loosens a wire, trouble could ensue.

[Brad Voracek]

Originally Posted by Bongle

No, I like it how it is. Many teams have large difficulties simply doing autonomous, but that isn't because of the length of time given, it's because they find it simply impossible to do whatever action it is in any amount of time.

I think the best way to improve autonomous participation rates would be less fragile or easier to use autonomous-focused equipment. On both teams I've been on in 2006 and 2007, the camera broke or would inexplicably stop working pretty frequently. It takes a long time to set up, a long time to test, and requires a large area to practice in properly. Laptops with serial ports are growing fewer and fewer, and the ones that my teams have been able to afford typically have nonfunctional batteries, so on-field debugging is impossible.

I would LOVE an end-of-game autonomous mode. You could have the end-of-game bonus be being some specific point on the field at the end of the teleoperated period, and at another point at the end of the ending autonomous period.

Problem with an end of game one would be that you'd have issues with game damage possibly getting exacerbated by the robot doing things autonomously. For example: in 2004, our robot had two high-torque motors driving seperate worm gears that moved the arm. Once, one of the wires came undone but the team didn't notice, and the unsynchronized worm gear action actually torqued the entire arm. With start-of-game autonomous, you can disable the autonomous mode to not damage things you've found are damaged. But if match damage breaks a chain or loosens a wire, trouble could ensue.

I would love an end game auto too, it would be interesting. To stop the problems you said, just e-stop yourself before auto starts...

[Pavan Dave]

Originally Posted by Bongle

No, I like it how it is. Many teams have large difficulties simply doing autonomous, but that isn't because of the length of time given, it's because they find it simply impossible to do whatever action it is in any amount of time.

Honestly, to combat that there are products out like the one Eldarion made available at http://falconir.com/ and there is a CODE database provided by a team (can't remember the name or site) which is accepting past year code donations and I think that if people donated previous years' autonomous codes, new teams could "reverse engineer" them and learn by subdividing (wrong word?) what is already there.


Pavan.

[Gdeaver]

Last year was good. At the beginning of auto, robots moved - did something. Our team even scored 4 to 6 balls in the center goal with only timers. This year is bad. Robots don't do anything until human control. This year the incentive to put resources into auto is all most nil. Does it really matter to the game if you score a keeper? The length of auto time is not the key it's the value of possible points and game wining advantage that push teams to focus on auto. As far as auto value, the game designers blew it this year.

I find the concept of an end of match auto ---- interesting.

With Easy-C even teams like ours can do something useful in auto but it has to have a high value to the game. This year we will sit and wait for human play.

Isn't robotics about taking human interaction out of the loop?

[mluckham]

Isn't robotics about taking human interaction out of the loop?

I agree this year's Autonomous was of dubious value - although in high-scoring matches like the Waterloo Regional just concluded, the 'keeper' did have some importance.

Making Autonomous work though, requires a good machine, and excellent/innovative programming - it stretches the minds and creativity of the entire team. It engages the minds, not just the arm controlling the joystick!

While Easy/C makes it possible to have a one-liner Operator Control program - TankDrive(..,..,..,..) - Autonomous demands the programmers and the team add hardware (encoders and do a lot more work).

I hope, next year, the "rewards" for doing Autonomous well are once again "worth it" - as they were in the 2006 competition. Even this year, had the keeper ring been worth extra (20 points? 10 points?) it would have been more worthwhile.

[Qbranch]

I can say that in this year's game especially the 15seconds is a really short time to do anything seriously cool...

for instance, I'm doing an ultrasonic radar system only autonomous (for style points ) and it takes 14.78 seconds with NO SLIP UPS. Granted, the world doesn't work this way, so usually it runs overtime even in simulation and can't complete in the time allotted... even 5 extra seconds would be great!

-q

[cziggy343]

i wouldn't b in support of it. i say that because from what i have seen, only a small percentage of people take advantage of the autonomous mode. adding a couple of seconds would only add time to the "staring competition" that takes place at the begining of most of the matches.

[Not2B]

I guess I don't understand why more teams don't do... SOMETHING in auto mode.

No one said you had to try and score.

If you have a arm all folded up, why don't more people just unfold the arm and get ready for human play. Or move forward onto the field. Maybe you don't have the time to perfect a camera based auto mode - but it's not as hard to "get ready" for game play.

Maybe alot of teams never really thought about it...

hmmm....... I never really thought about it before now.......

[Bongle]

Originally Posted by Not2B

If you have a arm all folded up, why don't more people just unfold the arm and get ready for human play. Or move forward onto the field. Maybe you don't have the time to perfect a camera based auto mode - but it's not as hard to "get ready" for game play.

Why not arm unfolding? Because you can do that in the time it takes your human player to run forward and toss the first ringer

Why not move forward? Risk of ringer-ing yourself or an alliance member ringer-ing you on the first toss. You're safer in the 'shadow' of the driver wall.

Why not move across the field and prepare to defend? I saw a few teams get tangled on the rack trying this. In order to avoid entangling yourself, you need the gyros and accelerometers as well as the programming skill to use them, which is nearly as hard as the camera.

I agree it is the skill of the programmer, but also it is the skill of hardware team. We have a very successful auto because we got the robot over a week before shipping date. They set goals to finish early and kept to it. so we had time to work on this hard autonomous game.

Totally agree with this. You CANNOT make an effective autonomous mode from scratch at the competition, unless you have a godlike hardware team. At the competition, there will always be repairs to be done, bolts to be tightened, kajiggers to be lubed and polished. The programming team only gets robot access when all those things are done, which is never. You need substantial alone time with the robot during the build period in order to develop and test autonomous mode, which requires a fast hardware team.