Did Anyone Accept the Challenge?

[Geek 2.0]

If my memory serves me correctly, there was a thread last year where a challenge was put forth for teams to make their robot entirely autonomous, with no operator control. A few teams pledged to do it.

I in no way criticize anyone who said they would and then didn't, because it's something that is really difficult to do, and not just anyone can do it. But I would like to know if you've seen anyone (or you are someone) who actually did this, because if they did, they deserve the utmost respect.

[Alan Anderson]

If a team fielded a robot that used only its onboard sensors and programming and took no input from the drivers, I'd wonder why they chose to show off their programming skills instead of playing the game to win. I would feel sorry for any of their alliance partners.

Unless, of course, the fully autonomous robot could score more points than the same robot could under driver control. That achievement would indeed be inspirational.

[mwtidd]

Originally Posted by Geek 2.0

If my memory serves me correctly, there was a thread last year where a challenge was put forth for teams to make their robot entirely autonomous, with no operator control. A few teams pledged to do it.

I in no way criticize anyone who said they would and then didn't, because it's something that is really difficult to do, and not just anyone can do it. But I would like to know if you've seen anyone (or you are someone) who actually did this, because if they did, they deserve the utmost respect.

This goal is still very alive in my eyes. Where I don't expect to see any in the regular season, I think come IRI we may see one that can pull it off.

Actually my defensive bot would be fully autonomous. It would start off right next to one of the towers and create a wall within the 84" limit stopping teams from putting up one of the mini bots. Also it would block a significant portion of the lane....

I am going to be working on a fully autonomous robot once I raise the money for it. (4 rev swerves, a killer bees 4 speed shifter, 4 range finders, and a second camera)

I think once I have the platform built, it will take about a 6 week period to program it. Hopefully come this years IRI, I will be able to bring it down and showcase it.

[davidthefat]

Even I, the one who proposed the challenge, did not do it for several reasons. I choose not to express why, but I can say it was from personal issues. Even if the robot was fully autonomous, and it dominated, we would not have implemented it. I will say that people have a fear of failure; that includes the fear of the unknown.

Mentors were very optimistic about it. I will attempt to implement it next year when it would be my last year.

[SudoSammich]

If anyone managed it I'll be rather surprised. Hell, over here we didn't even put out a robot with full code on it since the builders were inside the thing up until it was put in the bag. We'd be thrilled to have enough time to run A test on the bot before it's shipped...a team that can get full autonomous robot testing in has to be moving pretty close to the speed of light...

[mwtidd]

Originally Posted by SudoSammich

If anyone managed it I'll be rather surprised. Hell, over here we didn't even put out a robot with full code on it since the builders were inside the thing up until it was put in the bag. We'd be thrilled to have enough time to run A test on the bot before it's shipped...a team that can get full autonomous robot testing in has to be moving pretty close to the speed of light...

Or have their robot already finished at kickoff... cough PINK cough

but in seriousness I think it's an almost impossible task to accomplish in 6 weeks. If you had a fully built robot it may be possible, but you would have to have exclusive rights to the robot (ie no drive team practicing with it).

I wish FIRST would allow the zigbee module so we could begin programming coordinated autonomous programs. Three autonomous robot working together would be pretty awesome

[davidthefat]

I personally think it is more than doable; the teams just need to have a very expandable and scalable engine. That will take a long time. But once you have the engine, I believe it is doable in 6 weeks. My worry is using too much funds for such goals. I don't want to put a cap on how much the other departments can use for the robot. The sensors do not pay for themselves.

My other worry is: "Woop dee doo, you have a fully autonomous robot; what is the big deal?" Programmers would be impressed; no one else would be. It mostly is a public statement to even have autonomy. The drive team was mostly seniors, and a sophomore and I will be the only ones left next year. I don't think drive team would mind.

Third, it seems like a very selfish goal. I am purposely choosing NOT to be the team captain next year for the fear that I would turn out to be a dictator and drag the team down with me. I am nominating someone else on purpose.

Oh and BTW, I love your signature Mr. Mike

[Jetweb]

Originally Posted by lineskier

Or have their robot already finished at kickoff... cough PINK cough

HAHA anyone who knows pink knows they have to work like crazy just to finish there robot by Saturday afternoon...

On the topic of the thread i think this year has to may things involved to really pull this off. Making logos, finding the right peaces, navigating the field the list goes on and on. In order for this to happen we need another game like lunacy there the field is open, there are very few goals, very simple strategy, and very few (preferably one) type of game piece.

I for one hope to see this happen one day (or have the chance to try it myself) but i don't think its gonna happen successfully this year. Even if it was pulled off i think i would leave it in the shop and leave the real work to our drivers, autonomous is enough of a burden.

[mwtidd]

Originally Posted by davidthefat

I personally think it is more than doable; the teams just need to have a very expandable and scalable engine. That will take a long time. But once you have the engine, I believe it is doable in 6 weeks. My worry is using too much funds for such goals. I don't want to put a cap on how much the other departments can use for the robot. The sensors do not pay for themselves.

My other worry is: "Woop dee doo, you have a fully autonomous robot; what is the big deal?" Programmers would be impressed; no one else would be. It mostly is a public statement to even have autonomy. The drive team was mostly seniors, and a sophomore and I will be the only ones left next year. I don't think drive team would mind.

Third, it seems like a very selfish goal. I am purposely choosing NOT to be the team captain next year for the fear that I would turn out to be a dictator and drag the team down with me. I am nominating someone else on purpose.

Oh and BTW, I love your signature Mr. Mike

Haha thanks! It was my quote for my slot in the senior yearbook.

I definitely think it is possible in 6 weeks also, but there would be some requirements. First using almost an identical chassis. Second, a swerve or a mecanum would make the task a lot easier. Third, working libraries for each sensor (many could be a layer on top of wpiLib) and a working camera library on the driver station (I just got Open CV working in netbeans on my mac).

My list of absolute requirements for autonomous logomotion:
2 cameras
2 rangefinders
strafe (mecanum, swerve, drop strafe)
encoder on arm/lift
5 ft / s @ 90% on at least one speed (hopefully not the only speed)

ideal additions
a third camera for the robot (on board processing)
1 compass
1 gyro
encoders on everything
multi-speed transmission

[EricLeifermann]

Aren't the mini bots fully autonomous?... then any team with a mini bot has a fully autonomous robot this year

[mwtidd]

Originally Posted by Jetweb

HAHA anyone who knows pink knows they have to work like crazy just to finish there robot by Saturday afternoon...

On the topic of the thread i think this year has to may things involved to really pull this off. Making logos, finding the right peaces, navigating the field the list goes on and on. In order for this to happen we need another game like lunacy there the field is open, there are very few goals, very simple strategy, and very few (preferably one) type of game piece.

I for one hope to see this happen one day (or have the chance to try it myself) but i don't think its gonna happen successfully this year. Even if it was pulled off i think i would leave it in the shop and leave the real work to our drivers, autonomous is enough of a burden.

it was a joke referring to the picture they posted
The concept was give me a robot close to what this years will be, and I could play with it all season.

I agree this year is a bit ambitious to accomplish. I think the most difficult part would be avoiding penalties. finding tubes, picking them up, and knowing where to cap them can be done reasonably with off-robot processing using something like open CV

[JamesBrown]

David and lineskier ot of curiosity how much experience do you guys have with programming autonomous robots? I think you guys are drastically under estimating the difficulty of this task.

[davidthefat]

Originally Posted by JamesBrown

David and lineskier ot of curiosity how much experience do you guys have with programming autonomous robots? I think you guys are drastically under estimating the difficulty of this task.

I really have no experience, I believe that it can't be that hard in a such confined environment such as a FIRST game compared to the world. I was only planning on being defensive anyway. A fully autonomous robot could have been just a wall like pong. Just strafing with the bot in front

[Vikesrock]

Originally Posted by davidthefat

I really have no experience, I believe that it can't be that hard in a such confined environment such as a FIRST game compared to the world. I was only planning on being defensive anyway.

It may be even harder. In "the world" it is often not a requirement to navigate around objects moving around at 10-15FPS.

[davidthefat]

Originally Posted by Vikesrock

It may be even harder. In "the world" it is often not a requirement to navigate around objects moving around at 10-15FPS.

Eh, true, but we won't need to be calculating realtime to compensate for uneven terrain or some other stuff like that. I see no reason why this will not be possible in 1 year.