Auto mode 2007

[nuggetsyl]

The more our team works on auto mode the more i am seeing the 2005 problem all over again. I just see teams having a tough time doing a good auto. I also see teams playing a ton of defence in auto this year.

[Andrew Blair]

Successful denial of autonomous is far easier than actual scoring, so you are right. I have a feeling that operator control will mean more than anything thos year.

[EricH]

I agree. The main difference is that there are more targets to place on and see and lighted targets to see.

[Andrew Blair]

Originally Posted by EricH

I agree. The main difference is that there are more targets to place on and see and lighted targets to see.

Very true- more rungs mean a better chance statistically- but the biggest difference is the "deniability" of the rack. Any team can go balls to the wall and run into the rack. Even teams with an extremely fast drivetrain probably cannot get over to the rack and score before a slow robot shoves it at all.

I expect the Great CarNack to have a post regarding this issue shortly.

EDIT: There already has been! 1/08/07

[Donut]

I see hanging a keeper to be quite rare this year, not as bad as 2005 (when 0% of teams at the AZ regional hung a tetra with the vision system), but no where near the success of last year.

Defense I think will take a hit also, as all the "suicidal ramming mode" auto will get you this year is stopping opponents who are positioned near the sides of the rack; ramming the rack may work, but it's not as good a defense as last year, and you always run the risk of getting penalized for that.

[Netherlander]

The 2005 tetra stacking camera's failed to work because the lighting at the events fluctuated constantly causing our robot to be looking for the wrong colour almost every match. We got it working at our playing field at our school where we could keep the lighting constant but it failed at competition. Therefore I don't see this year being nearly as bad as two years ago as far as autonomous is concerned.

[dtengineering]

There is no rule requiring anyone to use a camera for auto mode. Although I'm sure some teams will have a great deal of success with it there are many other inexpensive ways of determining the location of an object. Infra-red is one that we've experimented with and had poor success (the IR light reflects away from the sensor on the shiny diamond plate at anything other than almost exactly 90 degrees) so we've moved on.

It won't, however, be a simple task. I expect on Thursday morning in Portland (first weekend regional, practice day) hardly anyone will score in auto, and by Saturday afternoon in Toronto (last weekend, finals) at least 2/3 robots on each alliance will be scoring.

It is great fun to see the level of competition rise every week as teams gain and share experience.

Jason

[Al Skierkiewicz]

Originally Posted by dtengineering

expect on Thursday morning in Portland (first weekend regional, practice day) hardly anyone will score in auto, and by Saturday afternoon in Toronto (last weekend, finals) at least 2/3 robots on each alliance will be scoring.

Jason,
I would hope your prediction will be accurate but I think that it is wishful thinking at best. With the short software time table and the low incentive for scoring in auto, I think the best event might have 10% of the teams attempting and far less then that successful.

[Rich Ross]

I think Al hit it right on the head. The small timetable hurts your chances, but what it all comes down to is Incentive. Last year, 10 point bonus, '05 was extra tetras, '04 was an early ball release, giving teams more "5-point possibilities" to work with. Auto mode has usually contained a significant advantage. and even though keepers cant be spoiled, the ringers next to the keepers can be spoiled. Not that big of a bonus.

[Jared Russell]

I think that a lot of teams will attempt it, if nothing else.

It isn't that hard to do a dead reckoning program that will score if the rack happens to be oriented a certain way - depending on how forgiving a given team's scoring mechanism may be with regards to precision, I could see that strategy working up to 20% of the time when done well.

The camera also works pretty well, and I would up the chance of succeeding on average to about 2/3 IF the team can write quality camera code.

The reasons I'm optimistic are (a) I'm an eternal optimist, and (b) there really isn't much else to try to do during autonomous mode. You can position your robot, but the human load zone is right next to the robot start. You could try to navigate to the other side of the field to get ringers, but again, why bother if they're right next to you. Lastly, you could play defense, but honestly, I think I'd rather just sit there and let the other team take their shot before I'd risk damaging the robot and/or field by slamming into the rack. So I see a lot of teams at least trying - I guess we'll see how successful the teams are at scoring!

[Barry Bonzack]

You could try to navigate to the other side of the field to get ringers, but again, why bother if they're right next to you.

If you wanted your first ringer to on the other side, your autonomous would be already have you there for the beginning of the game.

90% is a steep number. about 1-2 robots I would think can do this per regional. I'm betting many more fall in the 50% range.