Catapult vs Wheeled Shooter

[3072Cap]

I'm curious on what went through your heads when you were deciding between a catapult and a wheeled shooter when designing your robot. Team 3072 went with the wheeled shooter so that an opposing robot couldn't hit us hard enough, then dislodge our ball out of our shooter. I've noticed about an equal number of top level robots that have both designs.

[Jon Stratis]

Our decision was based on past experience with rebound rumble - that year, there was so much variability with the balls that wheeled shooters had a real tough time being consistent. We assumed it would be the same with these balls!

[dradel]

We went with a wheeled shooter this year, and it worked well, but as the balls got abused it lost some of its range. As most years it was a learning experience.

[s_forbes]

We had two main competing designs during the first week: a single wheel turreted shooter for batter/midrange shots, and a catapult for shots from the outerworks.

Initially the entire design team was leaning towards the turreted shooter, but we built prototypes of both concepts and weren't happy with the ball to ball variance of the wheeled shooter. We were worried about aiming accuracy for the outerworks concept as well, but the catapult prototype showed good repeatability and was able to hit the target at that distance, so we settled on the catapult.

Then we wasted half of our build season on a drivetrain.

[3072Cap]

I think 401's shooter was deadly, and pretty precise.

3072 had a couple positions of being able to shoot at different positions on the field, so we could make it from the outerworks, all the way to the batter. I'd be confident if we could shoot from the neutral zone on making it. Our prototypes for our shooter vs catapult were heavily on the shooter side. We just never got there with a catapult.

[PayneTrain]

Originally Posted by 3072Cap

I think 401's shooter was deadly, and pretty precise.

I feel like 401 can agree that if you wanted to add a fourth adjective to their shooter, it would be "temperamental". Wheeled shooters, unless nearly expertly honed in, can be temperamental. When they are on a roll though, watch out.

[happyWobot]

Originally Posted by 3072Cap

I'm curious on what went through your heads when you were deciding between a catapult and a wheeled shooter when designing your robot. Team 3072 went with the wheeled shooter so that an opposing robot couldn't hit us hard enough, then dislodge our ball out of our shooter. I've noticed about an equal number of top level robots that have both designs.

That's actually a more complex question than it might seem. Several things first:

Cons - catapult

Catapults take up a lot of space.
Energy transfer and release can be problematic without the right equipment
Dependent upon the stored energy you use, it can take time to replenish
Release angle can be limited without extra articulation or features
Need extra systems to load it

Pro-catapult
Energy transfer is more consistent because the game piece is in contact with the catapult longer.
Simple and reliable if built right

Cons -wheeled
The need to understand the relationship between wheels, game piece, compression, energy loss variability, damage to game piece, flywheel effect, flywheel size, flywheel weight, speed, and loading.
The need to understand the difference between speed, torque, motor kickback, and density of game piece
Outside the robot they are vulnerable to damage
Enormous amount of energy required to articulate if on an arm
Harder to set at precision angle
Harder to design especially if outside perimeter

Pros - wheeled
Instantly reset for use
Can perform both ingress and egress with same design
Variable control over power
Variable control over angle if on a pivot (can batter shoot or long shoot for example)



Someone mentioned Aerial Assist. Thats agreat example. That game piece was very elastic, it needed lots of time in contact with the energy being transferred so it could retain it. That's a really good example where a catapult is better

But you have to consider other things. In this game particular chassis comes first, not the shooter. We used large wheels with a high chassis floor. A narrow frame so it would not get trapped going through a defense at angle. After that it would not been difficult to make a catapult but would have been difficult to load it. Tall wheels take a lot of space. Point is the wheeled shooter was the best choice in this game for us based upon how we wanted to play the game and the kind of chassis we used. The bumper rules also drove that decision as well forcing the shooter outside the frame.

But as for how difficult it is to make a wheeled shooter I wouldn't say it's that hard. Several things though.

Small wheels are harder than large ones. Weight on outer perimeter of wheel is more important than its diameter. Rpms is not enough. You can pair a mini cim direct with a heavy 6-8 inch wheel or a high rpm 775pro at 4:1 with a light 4 inch wheel. Either will work. But a mini cim with a light 4 inch isn't going to shoot far. The heavy wheel reduces kickback. And the amount of compression on the ball and the contact material between wheel and ball matter significantly. A lot of energy loss can be taken there if it's the wrong contact material. And no matter how much math you throw at it, you can never be prepared for the variability in the balls surface, density and elasticity characteristics. You have to adjust everything to take that into account but it's more gut, intuition, and feel than science. In that regard I recommend you listen to what your bot is telling you. Literally. Listen for the kickback, the smoothness of egress, grinding, etc. Dust and marks on the ball tell you a lot too.

Finally some one mentioned 401. They were loading in at one event and their students brought the bot in to the pit. I asked one of the students about it and he looked at me with this serious look and said "this thing is scary". Was he not joking. Thats a wheeled shooter that doubles as a Howitzer. They overcome a lot of the game piece variability by shooting in a straight line but they need high precision control over angle to do that and hit the mark. Their design is really quite remarkable. Ballistic trajectory is what most people including us use, 401 is shooting laser beams.

The point? Can't shoot "laser beams" with a catapult. Only ballistic. If you store enough energy to catapult in a straight line its probably pushing the limit of legal. Besides even if you do it would be crazy difficult to control angle for threading the needle the way 401 does it.You use catapult or wheeled shooter depending on what you are trying to accomplish. Neither is better than the other until you define what it is you want to do. But that has to take into account all aspects of your design including chassis and navigation.

[hectorcastillo]

Originally we were thinking about making some kind of catapult of pneumatic puncher because we thought it would be more consistent than a wheeled shooter. However, we weren't really getting anywhere with our prototypes, and we had done a two-wheeled shooter in 2012, so that's what we went with.

I'm happy with the way our shooter turned out. We debuted in Alamo shooting an average of four high goals a game and only missing two during the regional, including one in a practice match. We've been able to tweak it a lot throughout the season (adding encoders, adjusting the ranges for our close and far shots, etc.) This was largely due to the fact that we designed in a lot of adjustability for the angle and speed of our shots.

[orangemoore]

Each type of shooter has its own set of Pros and Cons.

Whichever one you choose should be based on which fits best with your design and what else you are trying to do.

I would always suggest that anyone interested in learning how best to utilize each type of shooter is spend time with them in the off season.

[KaseyB]

Neither. 4469 decided to use the same type of pneumatic setup as they did in 2014.

[Drakxii]

Originally Posted by happyWobot

Someone mentioned Aerial Assist. Thats a great example. That game piece was very elastic, it needed lots of time in contact with the energy being transferred so it could retain it. That's a really good example where a catapult is better.

Actually it's a bad example as nearly no one did wheeled shooters with Aerial Assist because the ball were so large. A wheeled shooter needs go around at least 2 side of a ball, that's hard to do safely when the ball barely fit in the frame perimeter and anything outside of the frame perimeter is fair game to hit.

[backdrive]

Originally Posted by Drakxii

Actually it's a bad example as nearly no one did wheeled shooters with Aerial Assist because the ball were so large. A wheeled shooter needs go around at least 2 side of a ball, that's hard to do safely when the ball barely fit in the frame perimeter and anything outside of the frame perimeter is fair game to hit.

It's not true that nearly no one did flywheel shooters, though they were definitely not the most common design; in fact, the winning and finalist alliance on Einstein both had a flywheel shooter on them (254 and 1114).

[evoluti1]

For game pieces like this year's boulders, Aerial Assist balls, and even the balls from 2012, I think the notion that catapults inherently shoot more repeatably than wheeled shooters is a myth. If you're trying to launch a Lunacy moon rock or one of last year's recycle containers, then yes, a catapult is probably your best bet. But without more data directly comparing the catapults and wheeled shooters teams have actually built, I don't think anyone should be making confident claims to this effect.

Here's some slow-motion footage of one of the shots we took with our shooter prototype to test repeatability: https://youtu.be/lQUPcOZiL3c (one-wheeled shooter with a curved hood and 2" of compression. The hood on this prototype went through about 60 degrees of arc I think.)

The blue dots on the board represent where previous shots landed. As you can see, the consistency is as good or better than what we needed to make the shot from the outer works, and this was without even running closed-loop control on the wheel speed.

[Eric Scheuing]

We spent the better part of 3 weeks honing the design of our wheeled shooter to get it to perform consistently. Many tweaks were made in terms of wheel spacing, compression, and delivery mechanisms, but we manged to come up with something that seems fairly resilient to variations in boulders. Of course, no wheeled shooter is going to be 100% perfect, and occasionally we do get issues with a boulder pulling to the left or right a little, but we've got a pretty high accuracy all things considered.

If I had to do it all over again, I would definitely make a catapult though. It has benefits aside from accuracy. Most notably, they don't have to wait for their shooter wheels to spool up.

[Ari423]

I think this question is missing a category of shooter completely (or combining two where they shouldn't necessarily be combined). The category I am referring to is linear punch shooters. There is a big difference between a catapult (which throws the ball) and a linear punch (which punches the ball). I can't speak to the whole world, but I know a number of MAR teams successfully pulled off a linear punch (see 1089, 272). I'm not saying that that's necessarily the best shooter type, but I don't think it should be overlooked.