Hybrid Challenge-No Robot Left Behind

[JesseK]

Originally Posted by Travis Hoffman

I think there is a compromise that must be struck between "teaching a rookie everything there is to know about MPLAB and autonomous programming from scratch" vs. "here's a complete MPLAB project - it'll work - download it and go". This is the build season, not the summer and fall where longer term classes and educational sessions can be held. It's crunch time, ladies and gentlemen. If veterans are to assist rookies and the inexperienced NOW, during the build and at the competitions, ANNNNND we hope to preserve our sanity because, ya know, we still have to design and build our robots too, then we're going to have to meet in the middle on this.

I completely agree with this. Learning curves are exponential and unpredictable: if you start with nothing it takes an excessive amount of time to get anywhere. You can give some knowledge and time to ensure understanding to a rookie team and who knows how fast they'll run with it. Chaos theory at work indeed.

[iwin2000]

This forum is for teaching new people. I am learning a lot by reading the posts that people leave on here, and I really hope people don't stop.

I think it is fine to help the beginners with learning to program. However, if you are holding back from the little secrets that you put in your code, that is fine as well.

[Steve_Alaniz]

I ...agree in THEORY .... with leaving no team behind, but realistically, isn't this, in effect reducing the autonomous period to another operator controlled period? The documentation for the IR board is a bit shaky and I think there is a learning curve here but by Regional time we'll have everything worked out and cataloged so that it will be a plug and play situation.
FIRST gives a default program so you don't HAVE to program if your team is weak in that area and we'll soon have (in fact... already have) the code for simple forward, left, right and reverse(or stop). In fact, virtually a driver controlled "drive" mode
That being said I also agree that a true autonomous robot can benefit by having a creative autonomous program supplemented by the Hybrid control.
but the point is, including everybody.
( I personally think the autonomous should be a separate task that happens concurrent with the round but scores differently. Then, at the end of the autonomous, they join in regular play... but thats a pipe dream.)
In the end, the autonomous periods have never had a scoring structure that makes an insurmountable difference in the game play and they have always been relatively short for the amount of work that goes into them.
Bottom line :
1) Sure, give code if asked.
2) Involve other teams if they want.
3) but keep the autonomous period a challenge to some degree.

Heck I wouldn't mind doing away with autonomous if we're really excluding people, but I kind of like the challenge. ( By the way... We've never had a truly successful autonomous code... but I still like it!)

Steve

[jgannon]

Originally Posted by Steve_Alaniz

I ...agree in THEORY .... with leaving no team behind, but realistically, isn't this, in effect reducing the autonomous period to another operator controlled period?

I think you'll find that when there are six different IR transmitters going at once, it will be virtually impossible to successfully turn hybrid mode into another driver control period. All teams will need to exercise some amount of autonomy to perform reliably, and that the knowledge of how to do that should be passed on to those who don't have it. Believe me when I say that this will still undoubtedly be a challenge for all involved.

[Steve_Alaniz]

Originally Posted by jgannon

I think you'll find that when there are six different IR transmitters going at once, it will be virtually impossible to successfully turn hybrid mode into another driver control period. All teams will need to exercise some amount of autonomy to perform reliably, and that the knowledge of how to do that should be passed on to those who don't have it. Believe me when I say that this will still undoubtedly be a challenge for all involved.

You make good points but what you are saying is that hybrid is so unreliable it should not be tried...only pure autonomous has a chance. If that is the case, many autonomous actions will probably be threatened by run away hybrids. You may be right but we won't truly know until the first regional happens.

Steve

[jgannon]

Originally Posted by Steve_Alaniz

You make good points but what you are saying is that hybrid is so unreliable it should not be tried...only pure autonomous has a chance. If that is the case, many autonomous actions will probably be threatened by run away hybrids. You may be right but we won't truly know until the first regional happens.

Steve

That wasn't quite what I was going for. As waialua359 has pointed out in other threads, hybrid modes that are most likely to be successful are ones that take a minimal amount of input from the robocoach, perhaps solely at the start. As you point out, complete autonomy is at the mercy of the other robots. The limitations of the IR system reported by other teams indicate that full teleoperation will be problematic, too. Successful hybrid modes will be truly hybrid, and this is shaping up to be a pretty intense challenge. It will take a great deal of learning and planning, both now and at the regionals, to ensure that no robot is left behind.

[Tottanka]

see this:
http://www.chiefdelphi.com/media/papers/2070
inspired by this thread.

It is kind of a beta version, so i will consider adding there anything you want to =]

[Qbranch]

Hey... a note... I'm guessing I'm not the only one who's noticed that the IR remote reciever doesn't work at all if there's more than one remote in the general vicinity.

Is anyone else leaving the IR remote control as a last resort and going without?

Or has anyone figured out a way of making it function dependably?

For now, 1024 is improving it's SDAR (which was first debuted at our pit in Atlanta last year) or Sonar Direction and Ranging. Right now we have it to a point where we can take up to 3 simultaneous datastreams with a little over 1 update/sec, scan range 7', angle accuracy +/-2deg at Fast scan speed. Up to 9' range with +/-0.35deg accuracy.

We're going for a PID-based electronically geared path following routine with SDAR-based obstacle avoidance.

Okay, so i digressed a little. Videos to be posted (hopefully) in a couple weeks.

-q

[Travis Hoffman]

Please see this thread for free and easy software tools that facilitate connections and partnerships between teams over distances via the Internet.

Using these tools in tandem lets any veteran team reach out and provide programming or any other kind of educational assistance to rookie and inexperienced teams without ever leaving the comfort of their shops.

Realtime voice/videoconferencing combined with a powerful remote desktop/file transfer application - I invite you to check these tools out and see what kinds of creative applications you can discover.

Even if you're two veteran teams from the same region who aren't collaborating on a design, using these tools allows you to more easily ping ideas and prototypes off of mentors and students from other local teams - if people can get past the us vs. them competitive mentality and learn to spend more time "talking shop" with other local teams in person and via this software, I believe all the individual robotics programs involved would stand to benefit a great deal.

[Doug Leppard]

1902 team accepts the challenge and will support teams to have a simple hybrid mode.

PLEASE BUILD INTO THE HYBRID MODE TIME DELAYS.

One of my fears this year is not my opinents knocking me out but crashing into my partners coming out of the gate. We have built in switches that allow 0 0.5 1 and 2 second delays so we can give others time to get out of the way.

You don't need switches but you can have jumpers on the robot controller to change the delay time in the software.

[lkdjm]

As the senior programmer on Longwood's Digital Impact Team 564, I accept the challenge. Gracious professionalism is more important than competition.
One of our main goals at competition this year will be to help teams with their code, as so many helped me when I was learning.

[lukevanoort]

A goal of mine this year is to gradually make myself redundant, since I am graduating. So, hopefully I will succeed at that and thus have some time at the competitions to go out and help others with their programming. Actually, assuming they all make it to competition and all learn C (some are just starting), we should have six programmers available to help other teams on their hybrid modes.

If anyone is at the VCU regional (or the championship, but since we don't know divisions yet, it is kind of hard to make that offer) and lacks a hybrid mode (or is having any other robot-related issues) come on by the team 587 pit and ask for "Luke." If nobody in the pit is wearing a blue shirt embroidered with the FIRST logo (ie judges) I'll happily come by and help out. A word of warning: I don't know how to use EasyC and neither do most of our other coders, so if your robot code is in EasyC, we probably won't be very helpful.

[SU 39]

I think that this is a great idea for the learning experience. However, I personally won't be participating(my programming skills...suck). I just wanted to say that the complication of just saying "go straight" is that not all robots go in a perfectly straight line. Having 3 robots move in half the space as usual will make this even more difficult, with the risk of having one robot cut off the rest of the alliance and effectively impeding all possible scoring chances.
Also, for people who are participating in the challenge, it's worthy to note that if a team isn't coming with a preprogrammed hybrid mode, they probably don't have the IR board attached. So unless you have and are willing to lend extra IR boards + remotes, the program will have to be just a simple "go straight".

[Doug Leppard]

Originally Posted by SU 39

I think that this is a great idea for the learning experience. However, I personally won't be participating(my programming skills...suck). I just wanted to say that the complication of just saying "go straight" is that not all robots go in a perfectly straight line. Having 3 robots move in half the space as usual will make this even more difficult, with the risk of having one robot cut off the rest of the alliance and effectively impeding all possible scoring chances.
Also, for people who are participating in the challenge, it's worthy to note that if a team isn't coming with a preprogrammed hybrid mode, they probably don't have the IR board attached. So unless you have and are willing to lend extra IR boards + remotes, the program will have to be just a simple "go straight".

I agree about the collision. I fear this year more from my teammates than opponents. I would suggest that everyone should at least try to go forward and get the four points.

So I suggest that the go forward for four points have a delay of at least one second. We will hopefully have a significant hybrid mode and would appreciate not being hit. But at the same time we would appreciate having the extra points from our partners without hitting our robot.

So help evaryone get the four points and not be moving, but include a delay.

[Tottanka]

Originally Posted by Doug Leppard

I agree about the collision. I fear this year more from my teammates than opponents. I would suggest that everyone should at least try to go forward and get the four points.

So I suggest that the go forward for four points have a delay of at least one second. We will hopefully have a significant hybrid mode and would appreciate not being hit. But at the same time we would appreciate having the extra points from our partners without hitting our robot.

So help evaryone get the four points and not be moving, but include a delay.

I would even include there a 5 second delay, and not jsut go straight, but also go left, for 8 points.

The 5 second delay is good for 2 reasons:
1)it gives your alliance members enough time to get out of your way when you will be blindely going forward/left.
2)that is justa bout the tame it will take to a robot from the oppoisng alliance to get to your location, if they go straight forward frm the start of the match. Therefore you might have a good chance of blocking them from recieving more points, and even if you don't - it's still good.

As a team that is gonna be having (hopefully) a significant hybrid mode, we will make sure that those in our aliance would not ram us down on the way, and i suggest the same for all of you guys.