Poll: Turrets for Rebound Rumble
 

What is your opinion on using a rotating turret for this year's game? How far should it turn? How would you use it?

*This is for my own curiosity, and will not effect 256's decision making process*

Re: Poll: Turrets for Rebound Rumble
 

We felt that the ability to score with any robot orientation is important especially since the number of balls is fewer than we'd like. It is easier to pick up a ball and go for another one and still be able to shoot. We think the benefits outweighs the complexity.

Re: Poll: Turrets for Rebound Rumble
 

Wow, 100% no (at this time), that's pretty convincing...:rolleyes:

We considered this, and actually did it for Aim High, but it adds weight, complexity, and doesn't add any features we can't do with drivetrain. Yeah, I guess we can't drive sideways and shoot, but if we're touching the key, defense will be a challenge. So we stop, turn and aim.

Re: Poll: Turrets for Rebound Rumble
 

Well, we decided that it is too risky to shoot form afar, so we are already lining up with the fender, and it should not be a problem.

Re: Poll: Turrets for Rebound Rumble
 

Lining up with the fender is pretty much the only place on the field when defense can be played on you.

Back on topic, the turret adds a huge amount of complexity and complication in design where it would not have been before. 90% of that functionality can be done by moving the drive train.

Re: Poll: Turrets for Rebound Rumble
 

Here's something one of our members put together showing the relationship between distance and accuracy.



I will leave it to others to determine if a turret is necessary, but the precision required to sink shots from long distances is substantial.

It may be possible that hitting the backboard and bouncing balls in will be more forgiving. We don't yet have good data on that.

Re: Poll: Turrets for Rebound Rumble
 

The key doesnt start at 12 feet it starts at 101 inches, which is less than 9ft.

Re: Poll: Turrets for Rebound Rumble
 

We still thought that defense wouldn't be played that much, and that there would not be a downside. Of course, that why we went with 2 speed for the first time this year, just in case.

Re: Poll: Turrets for Rebound Rumble
 

We got our turret made tonight, although the robot to put it on, and the shooter to put on it, are not really even started. We are looking for about 45 degrees maximum travel, but that can be changed pretty easily, it's capable of close to 180, I think. We want to just have enough travel in the turret to do what's needed, which is to get the shooter aimed at the basket when we park on the key.

The complexity of our turret is not very high, but then we don't have it working yet, so we are not sure if it has sufficient complexity :)

The turret we made in 2009 worked well and was also relatively simple.

Re: Poll: Turrets for Rebound Rumble
 

I would say go for the turret.

that way, if you get pushed, so what. esp. if you have camera code ala Aim high that can track the target... you get pushed, the turret auto-aims and you can still (hopefully) make the shot.

not to mention all the legal motors... the AM gearmotor looks to be an excellent turret motor; it reminds me of the globe motors, which my team's bot in 2009 used for a turret control... it worked well, dare I say perfect.

Re: Poll: Turrets for Rebound Rumble
 

True enough; though the rest of the information is correct. Teams will need to aim with high precision for long shots. I think folks are grossly underestimating the difficulty involved here.

Re: Poll: Turrets for Rebound Rumble
 

We are adding a turret for two reasons:

1. We've done it before, easily.
2. It involves software to aim well, but you can also manually set our turret and shooter assembly for autonomous mode this year.

If either of those things weren't true, I would support a non-turreted shooter for my team. We just did well last time with a COTS lazy susan type bearing setup that we're refining for 2012.

There's rarely a question of "best" in FRC. There is only "best for your team". That said, a turret for many ball shooters is a decision with a much lower opportunity cost than, say, a swerve drive. Thus you will likely see turrets at the top of play no matter what.

Re: Poll: Turrets for Rebound Rumble
 

There are many factors that go into whether or not building a turret is the "best" strategy for this game.
1. Prior experience with turrets (off-season or previous years).
2. Whether you can get the desired precision with the drivetrain.
3. Motor allocation.
4. Time taken for iteration and development.

We ultimately decided against the turret due to several of these factors (one of which has yet to be determined). It comes down to realistically looking at what you can achieve, and what resources you are willing to allocate to achieve it. For us a turret may have been achievable, but it would require allocating more of our resources than we were willing to use in it's development. Whether or not this puts us at a competitive disadvantage is yet to be seen, but it is less likely to put us at a huge disadvantage compared to attempting to do more than we realistically can.

Re: Poll: Turrets for Rebound Rumble
 

I think you're mixing up the width with the position. The key is 144 in., or 12ft, from the Alliance Station wall, and 48 in, or 4 ft deep. So 12 to 16ft.

Re: Poll: Turrets for Rebound Rumble
 

We talked about a turret for many of the reasons already mentioned, and we have done it before and it would be fairly trivial to implement. However, we find it critical to keep the CG as low as possible, and adding a turreting mechanism adds weight up high and potentially moves shooter weight higher. Given how easily one can turn the robot to aim when protected in the key, and the relative scarcity of balls, we don't think it's worth it.

Re: Poll: Turrets for Rebound Rumble
 

This is a great table. That said, achieving sub-1 degree azimuth accuracy does not in itself necessitate a turret.

Re: Poll: Turrets for Rebound Rumble
 

Right-o. I think that the difference between passable teams, great teams, and awesome teams will be accuracy at range...

...and accuracy at close range is better than innaccuracy at long range (except for certain feederbot applications).

Re: Poll: Turrets for Rebound Rumble
 

In my experience... nothing is trivial....

Re: Poll: Turrets for Rebound Rumble
 

If you don't have your design down at this point and you read one sentence on Chief Delphi, read this one.

Re: Poll: Turrets for Rebound Rumble
 

Our turret so far. The plywood disc sits on a 6" lazy susan bearing. The ball is fed from the side. The shooter sits on top of this. The turret only moves a little bit, like 45 degrees. The belt is screwed to the wood disc. The pulley on the gearmotor had it's hole opened up to 10mm to fit the gearmotor shaft. Design calls for about one second to move 45 degrees, so we have about an 8:1 ratio, with the roughly 1 rev/second gearmotor.

Re: Poll: Turrets for Rebound Rumble
 

Looks nice! Are you just using friction for the drive belt, or are you going to put teeth on the plywood disc as well?

Re: Poll: Turrets for Rebound Rumble
 

Personally, i don't think a turret is needed for this game, it adds a variable that drivers have to more actively control. that is not to say it can't be a useful asset, but if you don't have the driver practice time to dial them in, than i believe it can be more of a hinderance to team performance, than a help.

a driver that simply drives to a certain spot on the field and points the chasis in the right direction will have a lot more success than a driver going to random locations, dialing in the shooter speed and pointing a turret in the right direction. i think many turret teams will find that they do the same thing as the non-turret teams, because their drivers can't get all the variables down consistently.

that being said, one area where turrets could be useful is on defense... a turret team driving in the backfield with turret pointed toward your end. the chasis is driving, turning and picking up balls while the turret just keeps shooting them to your end.

Re: Poll: Turrets for Rebound Rumble
 

Jim -- you could use Ethernet cable for your sensor wire there. Might tidy things up a bit and be easier to manage. Big spools are much cheaper (per foot) than what you'd get at a electronics store. The school's IT department may also have some spare you can use (10' should be enough). No need for connectors, just splice the twisted pairs like any normal wire. We did that for our 2 limit switches + 1 encoder last year on our wrist joint and it worked great.

Re: Poll: Turrets for Rebound Rumble
 

On our turret, the belt was cut to the right length, and the ends screwed into the plywood disc. There's no need for much rotation, so we didn't make it so it could rotate much.

We are planning to use the turret for computer controlled aiming with the camera, so the drive just parks the robot so it's facing the goal, and the robot does the turret turning to "fine tune" the aim right at the basket. The idea is to make it easier on the driver, not more difficult.

I'm not in charge of wiring, but I'll suggest that to the electronics team. For now they just grabbed some old limit switches and wire to get it working.

Re: Poll: Turrets for Rebound Rumble
 

Gotcha! I was wondering because the addition of teeth would have added accuracy over just tension on the belt -- looks like you've got that taken care of with the drive sprocket.

Re: Poll: Turrets for Rebound Rumble
 

I'm sure that's where they got the number, but if you put your back wheels on the front of the key, and aim for the center of the basket, you're looking at about a 9 foot shot anyway.

Re: Poll: Turrets for Rebound Rumble
 

Video of a turret in action, but not connected to a shooter yet.

http://www.youtube.com/watch?v=Hu2llOOlR18

Re: Poll: Turrets for Rebound Rumble
 

Looks great! The AndyMark gearmotor will probably be our choice of motor for our turret, too.

Re: Poll: Turrets for Rebound Rumble
 

if you're going to have a camera system auto-aim your turret, you can have a camera system auto-aim your robot.

Re: Poll: Turrets for Rebound Rumble
 

Yeah....maybe if you have a drivetrain that works a bit better than ours will.

Re: Poll: Turrets for Rebound Rumble
 

Why not both? we were thinking of making a program so our camera sees how far off we are, and makes large changes via drivetrain moving and smaller, precision changes via turret.

Re: Poll: Turrets for Rebound Rumble
 

We're trying to develop a system-un turret, so we could just hit a button for which hoop to aim at, and let the 'bot do the rest. Plus, we think it'd be cool to have TARGET LOCK flash up on the driver's station.:p

Re: Poll: Turrets for Rebound Rumble
 

I think it would be neat to have some lights on the robot that indicate things like "acquiring target" and "target lock"

Just to give the fans something fun to watch.

Re: Poll: Turrets for Rebound Rumble
 

We say "there will be no defense this year" every year. Then there will be the 1729's and the 3467's that will squash your chances of scoring every match.

Re: Poll: Turrets for Rebound Rumble
 

On the other hand, you only have to be touching the key, so the safe zone for a 37" robot with a 14" appendage is approx 8 to 20 feet. Yes, I expect to see at least a few rotating hands and shooters at the end of the appendage. If you think a standard turret is complex, then stay away from this option but kudos to those who try.

Re: Poll: Turrets for Rebound Rumble
 

Well, we have red and blue LED assemblies.
BTW, I'd still like to borrow your o'scope to demonstrate its power to the students.

Re: Poll: Turrets for Rebound Rumble
 

We decided to go with a turret precisely because of the precision required. Without careful design, we've found that most drive trains just have too much lash to be used for precision aiming. When you take the mass of the robot and the floor scrubbing inherent to turning in place into account, you end up having to apply fairly high forces over very small distances. This is not easy with gearmotors and likely to lead to oscillation without careful mechanical and algorithm design.

Its certainly not impossible, but its much easier to use a small, precise gearmotor to turn a lighter turret than it is to try and adjust the whole robot. In general, through Aim High and Lunacy, we found that turrets really aren't all that complicated with a little cleverness. Certainly easier than trying to engineer lash out of a drive train or oscillation out of a PID loop with four independent outputs and inputs.

Re: Poll: Turrets for Rebound Rumble
 

I'd like to hear of any team that did Aim High in '06, making a shooter, tell me that a turret is not necessary.

If teams are going to rely on the driver to aim the robot, I think they'll be in for a rude awakening.

Re: Poll: Turrets for Rebound Rumble
 

We had a static shooter in '06.

We're building a turret this year!

Re: Poll: Turrets for Rebound Rumble
 

My thought is that aiming with the drive base can work, but I think the big advantage of turrets comes into play when you're being actively harassed by a defender. If you can push your way to the key, you're good to go, but if not, or if you want to shoot close-range, a turret will be invaluable against aggressive defense.

Re: Poll: Turrets for Rebound Rumble
 

In Aim High there wasn't a big spot in the middle of the floor that gave you 3 points any time somebody so much as touched you :)

Re: Poll: Turrets for Rebound Rumble
 

Our team started in 2007 so i wasn't around for aim high, but for 2009 we went with a turret because we had moving targets, and so we thought that we needed that ability to be able to track and score on enemy trailers. we soon realised that long-range shooting = very innaccurate and it was easy to simply drive up to the trailers turret facing forward, and score (even with the constant collisions).

This year, we have stationary targets and protected areas on the field. I think turret teams will soon find that the variablity in ball size, shape, mass, density... variability in shooter speed, compression, feed orientation, feed position... will still be difficult to overcome and compensate for even with a slow moving turret.

that is not to say that a turret couldn't help the accuracy of a robot. I simply don't think the benifit is worth the complexity in control, design and programming or the allocation of weight and resources. a turret might add 5% more accuracy to a robot shooting from the key (probably less from closer, more from farther), but in a game where a robot shoots 6 balls a match, going from 65% to 70% accuracy... you still make 4 out of 6 balls.

(this all having been said my team is building a turret, despite my objections.:rolleyes: )

Re: Poll: Turrets for Rebound Rumble
 

I hope you don't think turrets and manual aim are mutually exclusive. We plan on being able to rotate about the center of the robot and use vision tracking to acquire targets. Sure, you can add on a turret system to aid in auto-aim, but you can also build a turret for a manual shooter.

Re: Poll: Turrets for Rebound Rumble
 

Has anyone devoted any thought to having a limited-traverse mechanism capable of aiming maybe 20 or 30 degrees or so to get precise aim, instead of a fixed shooter or a fully rotating turret? You could still make accurate shots but it would be a simpler design.

We had a turret in 06 and decided that it was better to do a fixed shooter this year.

Re: Poll: Turrets for Rebound Rumble
 

Ours is designed for about 45 degrees rotation, as we figured that's about all it needs.

Re: Poll: Turrets for Rebound Rumble
 

968 and 254 probably would say that, with the other attributes of my teams robot this year we determined it wasn't necessary.

Re: Poll: Turrets for Rebound Rumble
 

We made a fixed shooter in Aim High.

We're making a fixed shooter this year.


Of course, we're also making another unicorn drive.

Re: Poll: Turrets for Rebound Rumble
 

I think that whether or not a turret is a good idea depends entirely on what kind of drive train you have. There are enough protected zones in this game that drivetrain aiming (under computer control/assist) will be every bit as reliable as turret aiming (under computer control/assist) IF your drivetrain is (or at least can be) as maneuverable as a turret.

Anyone relying solely on human aim and/or dead reckoning will not be that effective, methinks.

Re: Poll: Turrets for Rebound Rumble
 

There's not all that much about this game that requires a turret. We are going for two very specific shots this year.

Keep in mind, in Aim High, there was a large obstacle in front of the goal, and the goal was in a completely different orientation. I mean, the last time I used Aim High as rationale to make design decisions, we ended up with a turreted shooter in 2009, a clearly suboptimal design.

Re: Poll: Turrets for Rebound Rumble
 

We're going for one very specific shot. And we're taking it from a spot where we can't be touched. Granted we'll have the flexibility to launch 'em from anywhere (and make 'em... totally), but we're focusing on those 3 pts everytime, and if no one can touch you, why turret?

Re: Poll: Turrets for Rebound Rumble
 

That sounds familiar...

Baba Ghanoush was an awesome robot, but wow was she awful at Lunacy!

We're hoping for that one specific shot (I assume that you guys, being as smart as we guys are, are thinking of the same shot) -- but we're also designing for the possibility that what we hope to accomplish and what we can't accomplish vis-a-vis aimed shots at various ranges might well be different things!

Re: Poll: Turrets for Rebound Rumble
 

Aren,
When I re-read what I wrote, I didnt add the part of an '06 team who had a turret explain why they would NOT do it this year.
There are several responses of those who said they would do that, but not really explain why, other than the key.

I knew that someone would bring up 254/968. I'll see what Cory folks think.
They had a fixed shooter for which I am very well aware of playing with both of them on Newton at CMP.
I never thought their auto mode shooting Poof Balls were anything spectacular and sometimes they missed quite a bit of shots since it was done purely on aiming the robot.
However, their robot was advantageous due to the fact that they could plant their robot right beneath the goal and fire straight above them. Shots that were too hard, bounced right out, back into their hopper, and straight into the goal again. Yet, they could be pushed from the side, causing them to miss again.
Look at teams 25, 111, 217, 296 and the many others that had turrets. From anywhere near the vicinity, they could lock and load, eliminating the need for the driver to do any aiming at all. 25's ability to BOTH lock and load, and not be pushed around by locking their drivetrain made them highly successful, including overcoming the fact that they could ONLY human load to fill their hopper.
I wanted to reiterate what Karthik has always encouraged teams to do......design and build within your capabilities and expertise. Back in 2006, we had a fixed shooter for that very reason.
If teams are doing that for that specific reason, props to them as well.

Re: Poll: Turrets for Rebound Rumble
 

Oh we had one in 2006, and i quite enjoyed it, i just pointed them out to be annoying :p

My thoughts on this game mainly differ on the basis of having some protected places to shoot from harassment free, l think this makes the lack of a turret a much more non limiting decision.

Re: Poll: Turrets for Rebound Rumble
 

The thing to remember is that while we have protected areas this year, the shots are MUCH harder to make than in 2006 since the target is both smaller and mounted horizontally. From a guidance standpoint, the fact that the target isn't on the same plane as the goal (like 2006) makes the vision problem much harder because you have considerable parallax, depending on where you are shooting from.

I think they added the key precisely because scoring would turn into a game of chance for most teams without both a precision shooter and a precision guidance system. Having a straight shot, in line with the vision targets, makes it easier.

Knowing that, there's plenty of opportunity for defense bots to try and keep you out of the key. Yes, its risky, but if you're a consistent scorer, most teams will assume its better to risk a handful of penalties than let a good scorer nail unobstructed 3-pointers for 2 minutes!

Re: Poll: Turrets for Rebound Rumble
 

We're trying for a turret, but the design is not yet complete. Our drive train, a basic 6-wheel with gearing for about 13 fps max, will probably not allow for precision aiming of the whole robot, even under computer control. But we attempt the turret with the fallback plan that if it doesn't work well, as long as we can lock it down and still shoot, then we aim with the drive wheels.

We spend most of our lives in Plan B... 8-)

Re: Poll: Turrets for Rebound Rumble
 

:cool: Haha love you guys! :)

While this is true, this is the first shooting year with a safe zone aka the key. Shooting without an plan for aiming is extremely risky. You only have 3 shots and you can't waste 1 or 2 until your driver has the shot even if he is that good. Our plan is to make all of our shots from the key and use our robot base to auto aim at a really slow speed. We'll see how this works but a turret adds more complexity to our design when we already can aim our robot.

Good luck!

Re: Poll: Turrets for Rebound Rumble
 

That is a risky strategy.

Every incidental contact that occurs when the scoring robot is touching their key is worth exactly the same in foul points to the scoring robot as actually sinking that shot in the high goal.

Even worse: try to play defense on said robot while they manage to just get their corner wheel over the key and bam!, technical foul and 9 pts to scoring robot's alliance. That little accident just gave the offense's alliance just as much points as if they actually shot three balls with 100% accuracy into the high goal.

For these reasons, I can easily see a continued scrub with a defense bot near the key racking up 2-3x more points for the offense's alliance in fouls than they probably could have scored on their own.

Re: Poll: Turrets for Rebound Rumble
 

100% agree, this year the defense is WAY too risky. What is even worse is that most efforts to stop a robot from scoring ends in a penalty and them getting there anyway.

Re: Poll: Turrets for Rebound Rumble
 

Using a turret seems pretty mandatory for determined shooters, however I agree it all depends on how efficiently your team can develop the shooter and not spend the whole season trying to build it.

Re: Poll: Turrets for Rebound Rumble
 

We're doing it, 'cause why not?

I can see how many top teams will be fine in their choice not to however.

The key is a nice place to shoot from in that regard.