autonomous.....
 

hey guys... there was a lot of teams who had problems with autonmous last year... do you think that we will have autonomous for this coming season too... if we do how is that going to effect the whole competition...?

Re: autonomous.....
 

I definetly think that there will be autonomous for next year, but I don't think it will be implemented in the same way. I'de look for it closer to the end of a match.

Re: autonomous.....
 

i didnt think of that... but it sounds cool... and exciting... probably will be more competitive this time since many of the teams realized that they didnt had a head start for not having an auton...

Re: autonomous.....
 

I stil think it'd be fun if they randomly turned on autonomous mode.

Fun because I don't have to program it. :)

Re: autonomous.....
 

I'm pretty sure auto modes going to stick around. Especially since the new controller still has an autonomous user mode in the default code. I am leaning more toward what's been said before, that it will be closer to the end. Or, if it's at the beginning, the time may be shortened. most robots completed theirs or messed up within the first 10 seconds. Yes, having it closer to the end of the match would be very challenging, but it's not impossible. I think if the game is set up in a way that you could play it well autonomously at the end then it would be fine. Also, FIRST may require that robots return to their starting positions to run the automode, just like the pressure mats...
If so, you may see robots that attempt to hold off other robots from competing in their automode. That would be a bit annoying if everyone just got in a big pushing match to keep each other away from starting, and nothing happened for the last 15 seconds :D
Or, maybe they'll be 2 "hotspots" on the field where robots must compete to get to them so their auto is activated. That would also cause a few problems. I think the biggest issue with an automode at the end is positioning. Deadreckoning robots would be useless unless they could be lined up in the proper position. Sensory-driven automodes will more than likely be too slow to accomplish anything in 15 seconds. That's the biggest block. Come up with a system in the game where robots will want to let the other team get into automode, and be lined up.
Or, it may just trigger and you can start anywhere in the field.
But the more I think about it, more things just keep "adding" up negatively in my mind. They may just keep it at the beggining like last year. I mean, that was the point of automode. Just about every "strategist" knows that gaining the upper hand first in a match can win it, and to do that you had to function autonomously. They'll probably keep it at the beginning. Human-Player mode, however, may shift back to the in-game style. That would give a little chunk back to user drive time.

20 second auto mode, at the beging.

Then a 5 second inflatable clown period.

Wetzel

please! no more mention of clowns! *shudders*

i do believe that the auton mode is going to stick around, where it is in the match however is kind of hard to determine. it will most likely not be in the middle though. the problem with running auton at the end is the idea that things are so much harder to plan for your advantage. and how upsetting would it be if you're autonomous ruined your score and ended up benefitting the other team?

Re: autonomous.....
 

We could go into "pong" mode and just start bouncing randomly off the walls, other robots, or obstacles.
It may not help but it sure could make it interesting to watch.

Re: autonomous.....
 

I'm thinking for auto this year they might do something along the lines of, if your robot can put big ball on top of tower in auto you get 20 points or if your robot can put said ball on top of the power in driver mode you get 10 points. Something like that would help encourage more teams to look into auto programming sooner in the season instead of making it at competition. It wasn't that teams had a huge problem with it, just most teams left it till the very end when it could have been worked on a bit sooner. *remembers waiting in line at nats to use the test field for auto code testing*

Re: autonomous.....
 

I think most teams were trying to just have robots that would work under human control, they just didn't have the time to code for auto because they didn't have a robot to test with.

That said, the field section that is designed to be reacted to/with durring auto mode should be cheaper to build then last years ramp. Mayhaps something similar to the troughs from 2000 and having to remove the inflatable clowns from them?


Wetzel
~~~~~~~~~~~~~~
Perhaps cherry inflatable clowns...

Re: autonomous.....
 

I would like to see an "optional" autonomous mode that you could switch to human controlled mode when you want to, and the longer you stayed in autonomous, the more points you get. This could possibly be done as a multiplier of either the total points scored or the points scored while in autonomous mode.

Re: autonomous.....
 

i think that since this is only the second year of auton mode, it will most probably be at the start because some teams may still be adjusting to it (like the teams whose robot did not operate for those 15 seconds last year). however i think that maybe 4 or 5 years down the road when auton has settled in with everyone, then it might be at the end of the match to make it more challenging ...

Re: autonomous.....
 

Johnson, I love it! :D

Optional would give every team a chance, even struggling first year teams. If they had some way to let the crowd know it was switched, that would look good from the audience, too.

Re: autonomous.....
 

Re: autonomous.....
 

If autonomous is moved to the end of a match (or any time where the robot's start position is unknown) I hope we get some sort of fixed objects to locate a position off of.

I think it would be interesting to have the first 5 or 10 seconds and the last 20 or 30 second be autonomous. The start of the game would be similar to this year. The robot knows where it is and how to get to some predetermined places on the field. Things can complicate this, but its generally not too difficult to dead reckon your way around (yes, even with bins stacked in your way :p ). The end of the game would be more interesting. Some teams would attempt to move their robot into a known location (corner of the field maybe) so they could dead reckon the end-game- others would try and find their location and/or navigate the field. Maybe it would be possible to carry variables over into this end of game autonomous mode. A selector switch set mid-game could determine the robot's behavior in the last few seconds.

I liked autonomous mode last year, looks even more fun this year with the new controller. I just hope FIRST presents a wider range of options in autonomous.

Last year: "go forward, turn (left or right), go forward, crash into things".
This year: "<do stuff>, crash into things".

Greg

Re: autonomous.....
 

i think i will just have to wait anther 5 weeks to get to know about the auton... :confused:

Re: autonomous.....
 

I would REALLY love to see optional auton mode. A few of us on our team have discussed it off and on, and it gives the rookies a chance to function without having really experienced programmers, and veteran teams a chance to go above and beyond and develop some really cool systems to help them work in autonomous mode for longer periods of time. The only problem I see is showing the audience which robots are in autonomous modes when, and some way of keeping track. Other than that, I see that as a definate possibility. :D

A set autonomous mode at the beginning of the match is also perfectly viable too, as this is only the second year since autonomous mode has been introduced. At the end would definately be a big challenge for teams, since you wouldnt know exactly where and in what orientation your robot is in. Personally, I dont think that they would jump to autonomous mode at the end, but the beginnning is a definate possibility. Oh, wait, know what would be neat? a 10-15 sec mandatory autonomous mode at the beginning of the match, and then you can stay in autonomous mode for longer for extra points, or a multiplier of the points scored during the autonomous mode. The audience could tell who was in autonomous mode by some sort of highlighting of the team's number on the scoreboard, and by who was still a few feet from the conrtols, just have both an emergency stop and an autonomous end button on the controls, instead of just an E-STOP button. I think this is definately feasable and would be REALLY cool to see implemented. ;)

only 4 weeks until we find out!

Re: autonomous.....
 

i think that is a really good idea... a lot of teams can go beyond... and it will be easier for the rookies too... :)

Re: autonomous.....
 

Brandon, you're evil.

As a first-year C programmer, heading a team that's just as inexperienced as I am, I'm hoping very much that the auton will be like it was last year - at the beginning of the match, and relatively easy to plan. Anything that makes things easier for us is a huge benefit. :)

Re: autonomous.....
 

I personally would be very surprised to see it at any time other than the begining of the match. I think that trying to make autonomous mode at any other time, compounded with the new programming language this year, would make it more a level of difficulty above that of simply "challenging" for any but the most experienced teams.

While you need to have the game be challenging, you need to make some of the aspects be attainable by all teams. For instance, an autonomous program is challenging, but not impossible. New teams can do it, and veteran teams can do an EXCELLENT job of it.


On a side note, one way you could get around some sort of local positioning system would be to just end your robot in a specific position and orientation. However, I see a lot of room for foul play that could end up occuring by shoving teams robots around right before the end. The main problem with this would be that teams could potentially damaging their machines with much greater ease.

Just my two cents... thoughts?

Re: autonomous.....
 

Autonomous will probably be around again next year, hopefully at the beginning of the match. I like the idea of keeping autonomous on for a longer time to get more points for your team/alliance. FIRST would probably set a time limit for how long autonomous mode could be run if teams could decide when to take over to start driving. My guess for autonomous mode: MANDATORY 15 seconds, auto mode turned off after 45 seconds or 1 minute.

Re: autonomous.....
 

Yeah, I hope it's back at the beginning. My only concern is that if they expand the required auto mode time, they need to expand the total match time. I felt like time was cut a little short after last year for the drivers, and to cut it even shorter would be a strain.

Re: autonomous.....
 

What most of us did last year wasn't really autonomy. It was pre-programmed mode.

If they wanted to emphasize autonomy, the would give us a blind maze with a prize in it. You could maneuver your robot to the entrance to the maze and set it into autonomous mode. It would have to find its way through, get the prize, and find its way back out without driver intervention.

Now, most people wouldn't even attempt this if it were optional, since you might lose your robot in the maze with no hope of retrieving it before the end of the match.

im blind
 

I strongly believe that auton will stay in the beginning. Last year, the robots started in a measured predetermined place, and people still had trouble. The only way that there could ever be autonomous any other time would be to have some sort of absolute positioning system, which would allow a robot to know exactly where it is at all times. I highly doubt that this will ever happen. I agree with the previous post that says, which stated that what most teams did last year was preprogrammed mode rather than autonomous. Most of them were running open loop relying merely on timing. If FIRST wants real autonomy we are going to need more sophisticated sensors.

Re: autonomous.....
 

i think we should go into auto mode sometime in the middle. If it is in the begining it is like pre-program and it does not need much stragey. Putting it at the end will make everyone do the same thing and that is too go after the king of the hill points or whatever it will be called.

It most likely will be put at the begining just like it was last year. We will have to wait to find out.

Cyber blue team 234

Yep
 

And I think that hits it on the head, boys and girls.

We're gonna be doing pre-programmed movement until we get better sensors, case closed. You can't expect immensely intelligent autonomous movement with the low-end sensors provided and a $200 budget.

Matt

Re: Yep
 

What sensors do you not have that you wish you did? If the kids in Lego League can complete all of the missions with the limited set of sensors they have then you all should be able to do the same. This is supposed to be hard! You need to push the capabilities of what you're given.

Re: Yep
 

But most of the lego kids used simple dead-reconing combined with rotation & touch sensors. Is that all you think FIRST teams are capable of? They didn't make decisions on the fly. They didn't do any type of positioning system. They didn't have to react to a changing playing field because they were the only ones affecting it. The challenge in FIRST is much greater and even a simple robot in FIRST is more complex than the Lego league robots. They're just asking for more sensors to take autonomous to the next level. That's not an unreasonable request.

Re: Yep
 

I didn't imply that they were capable of no more than the FLL kids, but there seems to be a lot of people who don't think they're able to do anything meaningful with the sensor set provided in the kit of parts. What did you all add to create your impressive control system last year? If I remember correctly all you used was the gyro, a pot, and an encoder. Anybody could have bought that part from digi-key but you were the only team that had such a robust positioning system. You have more software expertise than most teams, but the new controller should be able to handle most of processing you implemented externally.

Short of providing a plug-and-play positioning system I don't see what they could provide that you can't already buy or build. I don't think its unreasonable to expand the sensor suite in the kit, I'm just wondering what people would like to see added.

Re: Yep
 

That's correct, we used the standard gyro, a 100k pot, and two of the Banner light sensors from the kit (not an off the shelf encoder).

They were all fed into an external box that calculated the coordinates of the robot. The RC would ask that box for the three pieces of data (x,y,theta/rotation).

I think that the key to our success was the clear division of tasks. The circuit kept track of the position, while the RC figured out how to get from where it was to where it needed to go. In doing this, each was able to focus on one task.

This allowed the processing power of each unit to be used most efficiently. As mentioned in other threads, we had so much code in the RC that we were actually missing packets from the OI. If we tried to cram in the positioning information, we would have missed so many packets that we would be in a constant state of reset...that is, if there was enough memory space for the code.

It also allowed us to be able to disconnect the box and completely disable autonomous movement. This was done by doing a handshake when the robot was turned on.

Another plus is that if for whatever reason the RC reset, we would be able to pick up where we left off because we didn't lose the information.

While you are correct in saying that the new OI would probably be able to handle the calculations that we did, I think (without knowing the specs) that if we had to redo the system, we would still keep the division of labor between an external circuit and the RC.

Designing an external circuit is a very difficult task, and there are teams that may not have the resources to do so. In these cases, the RC should have enough power to handle basic sensor-based autonomy.

Re: Yep
 

That sounds like the perfect topic for a white paper for the teams. Glad you volunteered!! :p (and, yes, I still need to poke at Sean as well to write one on the co-processor board he built last year)

-dave


-------------------

15 days to go!!!!