View Full Version : Week 1 lessons learned: balancing strategy
I've been watching the KC and Alamo regionals streaming today and have seen many attempts to balance a bridge with two robots, and some successes. Many of those attempts end up with a flipped robot. I hope other teams are watching as well, and will think hard about their balancing strategy.
What works: Drive onto the bridge. Tip it towards the other robot. Allow the other robot to push you onto the bridge and do the backwards scoot to do the final balance.
How to flip over the other robot: Drive onto the bridge. Tip it towards the other robot. Allow the other robot to drive partially onto the bridge, and back away from it. Attempt to balance the bridge yourself.
This second method nearly always results in both robots fighting each other for control of the balancing. Sooner or later one of them ends up partially off the bridge when the other is on the far end, flipping the one on the downhill side.
DON'T be the team that flips anyone who tries to balance with them. Please.
wilhitern1
02-03-2012, 15:25
Our problem at the moment is that we have questionable traction (we're working to improve it). Pushing another robot up the bridge seems unlikely for the time being... We expect to be able to get on the bridge without help however.
GCentola
02-03-2012, 15:26
Another thing I've noticed is that teams arent allowing enough time. The bridges seem to be last-second thoughts, and often dont turn out successfully.
On the cooportition bridge, it doesn't appear that there is any logical plan for approaching it. Are alliances discussing it pre-match? It would seem as if some sort of pre-planned process would help balance (as you said, two robots trying to adjust for each other just makes it wobble back and forth)
pandamonium
02-03-2012, 15:33
Why are some robots spending the entire match trying to pick up balls or shoot and miss a hand full of times but then never have time for the bridge? They could spend the majority of the match focusing on the bridge. I think it is important to be realistic with the strengths and weaknesses of your robot. 20 points seems to be enough to win the majority of qualification matches...
P.S. I would love to see 1126 and 1511 on a cooportition bridge together next weekend.
GCentola
02-03-2012, 15:49
P.S. I would love to see 1126 and 1511 on a cooportition bridge together next weekend.
Same! We are looking forward to FLR! You guys definitely have a cool robot! If this happens, i'm sure we will know what to do after reading this thread.
Grim Tuesday
02-03-2012, 15:55
Right now I'm feeling really good about making our robot have a very high traction drive system. There seem to be a lot of teams who are slipping all around the bridge and are completely unable to pull off a balance. I would be surprised to see many of these teams in eliminations.
GCentola
02-03-2012, 16:05
Speaking of low traction, that seems to be the death of 1 Co-op point. If both alliances are on the C-Bridge, but it isnt balanced, you still get one CP each, but it looks like right after the match ends, the bridge tips to one side and the one on top pushes the other one off accidentally when it slides down.
DominickC
02-03-2012, 16:07
How to flip over the other robot: Drive onto the bridge. Tip it towards the other robot. Allow the other robot to drive partially onto the bridge, and back away from it. Attempt to balance the bridge yourself.
I'd be weary of doing this. It might get you flagged for unnecessary aggression. It's also not in the nature of GP...do you want your team to be known as "that team who flips over other bots"?
AdamHeard
02-03-2012, 16:13
I'd be weary of doing this. It might get you flagged for unnecessary aggression. It's also not in the nature of GP...do you want your team to be known as "that team who flips over other bots"?
I don't think his statement was in regards to how to intentionally tip the opponent, but rather pointing out how many teams have repeated this same operation to tip their opponents today in the process of trying to legitimately cooperate.
DominickC
02-03-2012, 16:17
Oh, I see. Silly me :p
I expect this will be like mini bots last year. By week three there where 2 or 3 methods that made it look easy.
I was a little surprised how easy it is to flip a bot. Seeing it makes it kinda a DUH. Agreed that it does not seem to be intentional. But if you see the other robot backing away from you and you are more than half off it, you need to be on the gas following it.
Saw a few teams with 4wd that struggled getting the second pair of wheels on the bridge once it started to tip
On the cooportition bridge, it doesn't appear that there is any logical plan for approaching it. Are alliances discussing it pre-match? It would seem as if some sort of pre-planned process would help balance (as you said, two robots trying to adjust for each other just makes it wobble back and forth)
At Kettering there seems to be a fairly consistent planned attempt for the cooperation bridge. (Whether is works is a different story however...)
And yes, the second robot pushing the first robot on the bridge works better than each team trying to balance individually. Warning: Less is more. Small baby steps will get you further than plowing straight through.
Tetraman
02-03-2012, 23:15
I was the driving mentor/instructor for 174 this year, and I can say we have practiced the art of balancing every chance we could. Mastering such a thing is always difficult for anyone as a slight movement alters everything. But the number one thing we discussed was allowing time to score on the bridge and being ready for a team that makes a last second mad dash. If we get on the bridge with 30 seconds, or even a whole minute left on the clock, as long as it will end up with +10, +20 or +40 to our score (or +2 qualifying points) it will be worth it.
Another thing I've noticed is that teams arent allowing enough time. The bridges seem to be last-second thoughts, and often dont turn out successfully.
On the cooportition bridge, it doesn't appear that there is any logical plan for approaching it. Are alliances discussing it pre-match? It would seem as if some sort of pre-planned process would help balance (as you said, two robots trying to adjust for each other just makes it wobble back and forth)
Definitely discuss pre-match. I usually go out and find our alliance members 3 matches in advance to discuss these things.
I was the driving mentor/instructor for 174 this year, and I can say we have practiced the art of balancing every chance we could. Mastering such a thing is always difficult for anyone as a slight movement alters everything. But the number one thing we discussed was allowing time to score on the bridge and being ready for a team that makes a last second mad dash. If we get on the bridge with 30 seconds, or even a whole minute left on the clock, as long as it will end up with +10, +20 or +40 to our score (or +2 qualifying points) it will be worth it.
My thoughts exactly. Our team was struggling but once I realized this, it was a good thing.
Andrew Lawrence
03-03-2012, 13:17
Which is better: Auto-balancing, or manual balancing? Also, for those in eliminations, does it matter if your partner's robots can actuate the bridge, or can one or two of them be bridgeless, and have you actuate it for them.
The choo-choo train methodology seems to work well. The 2936-148-922 alliance at Alamo attempted the three balance and made it look easy the first time. Choo-choo up, easy over. Put the high traction high torque bot as the last one up. They made it seem as if 3-bot balances are as easy (perhaps easier) than 2-bot balances since there's so much mass at the edges (the same effect that helps tight-rope walkers).
RufflesRidge
03-03-2012, 20:21
They made it seem as if 3-bot balances are as easy (perhaps easier) than 2-bot balances since there's so much mass at the edges (the same effect that helps tight-rope walkers).
I think 148's bridge leveler also played a factor in that.
I think 148's bridge leveler also played a factor in that.
Yea, premature post. I saw the balance in The Red Alliance's 6-frame format so I didn't see the details. My bad.
Mentor Mac here back from Philadelphia, William Grove. Team 341, Miss Daisey
kick but. 11-0-1 first place. Wins tournment. Scores 5 baskets in hybrid mode. Second game of finals. Blue Allinance shoots five balls at start. They all get stuck in the basket. Game is replayed; they loose.
Be Safe Thomas McCubbin
GaryVoshol
04-03-2012, 07:03
How not to balance:
Balance one robot with about 40 seconds to go
Decide that a second robot will get on the bridge
Unbalance the bridge for the second robot to drive up
Dither around for 30 seconds trying to get the second robot on the bridge, nearly tipping both robots over
Fail to get the bridge balanced
Lose the match 3-0
It's not getting the amount of hype as the 148 alliance triple balance but 1477, 231 and 457 had a good strategy of having 457 go over to the other side foo the field and wait for 1477 and 231 to go up on the ramp and use their bumpers to quickly stabilize the ramp. Ii expect many teams to adopt this strategy.
Alpha Beta
04-03-2012, 11:33
It's not getting the amount of hype as the 148 alliance triple balance but 1477, 231 and 457 had a good strategy of having 457 go over to the other side foo the field and wait for 1477 and 231 to go up on the ramp and use their bumpers to quickly stabilize the ramp. Ii expect many teams to adopt this strategy.
Brilliant!
On a not so brilliant note, I saw a match where a robot balanced, and then in the last few seconds a partner taps the bridge (seemingly intentionally) to screw it up. Alliance still wins 4 to 2.
Match 40: http://www.more.net/content/2012-qualification-matches
DampRobot
04-03-2012, 12:35
It seems like there are a fair number of teams who go for two (or three in elims) robots balancing when just one (or two) would win the match, and neither teams end up balancing. Maybe people need to corrordinte within their own alliance more. Keep in mind, just win/loss counts for seeding (as well as coopertition bridge), not match score.
It's not getting the amount of hype as the 148 alliance triple balance but 1477, 231 and 457 had a good strategy of having 457 go over to the other side foo the field and wait for 1477 and 231 to go up on the ramp and use their bumpers to quickly stabilize the ramp. Ii expect many teams to adopt this strategy.
We have a purpose built mechanism just for this purpose; it may not be as impressive as a triple bot balance, but a garunteed double balance in thirty seconds or less allowed us to play some strategic defense and the rest of the alliance to score until the last part of the match.
I'm surprised so many teams wrote off balancing as easy or simple, and there weren't more dedicated robots or mechanisms.
RRLedford
04-03-2012, 15:44
We have a purpose built mechanism just for this purpose; it may not be as impressive as a triple bot balance, but a garunteed double balance in thirty seconds or less allowed us to play some strategic defense and the rest of the alliance to score until the last part of the match.
I'm surprised so many teams wrote off balancing as easy or simple, and there weren't more dedicated robots or mechanisms.
Our whole design is based on the assumption that the winning alliance will be capable of getting consistent and fast 3-bot balancing in elims. We designed our ~33" wide bot to to be able to roll up the side of the bridge with ~20" of our width overhanging into the air, out beyond the side, and only ~13" above the the bridge.
We wanted be able play defense and then use end with our super strong plow arm to get up onto the bridge with ~25 seconds remaining to pre-tilt it toward our partners on the offensive end. We can reach balance in 3 seconds after our first wheel goes up. We then lean the bridge toward them while they continue shooting as long a possible. We can ride along the side rail of the bridge (without holding or gripping it) from end to end, while most of our width is cantilevered out over the edge guide. Our engagement with it guides us accurately (even when wheels slip) so we can stay laterally balanced and avoid tipping off the side.
Once we tilt bridge toward our offensive , our two non-wide (max ~34") partners are able to quickly and smoothly drive right up the bridge, bumper-to-bumper fashion, stopping on either side of the center, as we simultaneously roll up alongside them toward to center and finish the balance. Our bumper can also help align & guide them, if they are having traction issues, since we cannot be moved laterally much (< 1") or be easily pushed or tipped off the side from the way we overlap the aluminum edge guide and from our lowest possible height bumper positions.
We are a day or two from finishing an identical practice bot to give us two full weeks of driving practice, a first for our small team.
-Dick Ledford
We balanced 9 times in matches (two CP bridges) and 4 times on the practice field.
he best strategy seemed to have the first team lay off the controls completely and the other team push them up the bridge. What seemed to work best was releasing the controls the second the bridge began to tip and it would center itself.
Balancing while both teams are moving is next to impossible.
Video Example: Balance (http://www.youtube.com/watch?v=RT96V-VvYBk) starts at 1:20 and ends at 1:35.
annie1939
04-03-2012, 16:01
At Kansas City, our team of rookies (FIRST Team 3931) in match 91 was honored to be part of a group that had all 6 robots balanced on the bridges.
It sounds like that was one of the first times at any of the regionals. It was AWESOME and we all got a standing ovation from the crowd.
Andrew Lawrence
04-03-2012, 16:03
Now that the "secret" of successful balancing is out, how do you think balancing will change in week two?
Now that the "secret" of successful balancing is out, how do you think balancing will change in week two?
It'll get defensed.
Teams were getting attacked at the bump to keep them from setting up the ramp climb.
D.Allred
04-03-2012, 16:19
Saw a few teams with 4wd that struggled getting the second pair of wheels on the bridge once it started to tip
Just to be clear about this, 4 wheel robots getting on the bridge second is very dangerous.
Speaking of low traction, that seems to be the death of 1 Co-op point. If both alliances are on the C-Bridge, but it isnt balanced, you still get one CP each, but it looks like right after the match ends, the bridge tips to one side and the one on top pushes the other one off accidentally when it slides down.
Yes we thought of this during build session and ended up using a two speed transmission this helps a ton on the bridge because it allows us to make very slow movements while on the bridge and it allows us to stay on the bridge if its tilted or not. the only bad thing is getting it of the bridge after we balanced as its hard to roll.
Alpha Beta
04-03-2012, 19:00
Yes we thought of this during build session and ended up using a two speed transmission this helps a ton on the bridge because it allows us to make very slow movements while on the bridge and it allows us to stay on the bridge if its tilted or not. the only bad thing is getting it of the bridge after we balanced as its hard to roll.
Your post reminded me to go back and watch our 2nd match. It looks like our low gear keeps everyone on the co-op bridge for 5 seconds after the match ends, even though we are not at rest. According to G37, and G41 this should have been worth 1 co-op point to both alliances. It was not awarded, we did not even think to challenge for it at the time, and it does not appear as if the refs are even looking for the 5 second count.
Video (Match 17): http://www.more.net/content/2012-qualification-matches
Match ends at 3:43 on the video, and wheels don't touch carpet until 3:52. At that point the ref gives the all clear signal. Head ref is clearly watching the robot roll off the bridge, but has no reference to the 5 second stipulation from G37.
Lesson Learned: If you have a low enough gear to put yourself in this situation you may want to ask the Head ref how they will determine when 5 seconds has elapsed during the drivers meeting on the Thursday practice day.
PS. What difference would this have made? In hind sight, none. At the time it could have been huge.
GilaMonsterAlex
04-03-2012, 23:36
I was really surprised to see that so many teams didn't get the strategy of balancing the bridges (not to say it's easy). I was expecting more team work.
Since the kick off, I was harping to my students about how we should "choo-choo" onto the bridge with another team. It was a major reason why we withheld our bridge manipulator. We weren't sure we needed it.
Now I'm going to make sure to harp about communication with the alliances about our balancing strategy. It will make for a much better competition.
The one thing I'm going to predict is that more and more teams will abandon ball scoring and line up on the bridge balancing. This to me makes the game a bit more boring and takes away the challenge. I wouldn't be surprised to see a rule change where you have to wait til the 30 second mark before you make an attempt to balance the bride (not crossing).
Also, just so I understand this correctly. 148 uses a piston or shaft down to the ground to level out the bridge correct? Do they lift the piston right before the end?
wireties
05-03-2012, 00:40
Also, just so I understand this correctly. 148 uses a piston or shaft down to the ground to level out the bridge correct? Do they lift the piston right before the end?
Looked like that to me...
dj bauch
06-03-2012, 22:48
Also, just so I understand this correctly. 148 uses a piston or shaft down to the ground to level out the bridge correct? Do they lift the piston right before the end?
Yes, they had a little "foot" they could use to stabilize the bridge. Very clever design. Once the bridge was stable, they would lift the foot back into the robot.
vBulletin® v3.6.4, Copyright ©2000-2017, Jelsoft Enterprises Ltd.