Ball Pitcher Hood Design

[dtengineering]

Originally Posted by JellyMan

We would like to build a turret for ease of aiming.
Is this a correct assumption? Thanks!

I can address the turret part based on 2006 experience... we built a full pan/tilt turrent... it is awesome to have a turrent... it makes the robot look great, and is fabulous for showing off.

But it is exceptionally difficult for humans to aim. The driver would be turning one way and the "gunner" would have to compensate. We didn't get our auto-aim working quickly enough or reliably enough... if we had, it would have made a huge difference. Although we occasionally made a beautiful long-distance, over the back shot, we were thumped by machines that could get close and score fast.

This year the hardware is faster and the software better developed... a turret might be a good idea, but don't count on humans to be fast enough or reliable enough to aim from any kind of distance.

And certainly don't count on your opponents letting you have time to line up a nice clear outside shot without a well-timed bumper-based greeting.

If you go with a turret, and plan to shoot from outside, computer-based aiming will be critical...

Jason

[MrForbes]

We're thinking that computer based aiming will be essential, no matter where we are...and we plan to be over the colored half circle if there are any opposing robots on our side.

[JellyMan]

"Video, or it did not happen"? Awe come on Mark, smile when you say that so I know that you are joking.
... and a fellow Falcon no less! Hey, its all fun.

Jason,
Our team is investigating methods of range-finding. We would like to use this input to automaticallly set the speed of the pitching wheel to hit the target basket. I agree that auto-aiming would nice as well, but the range finding will need to come first.
- Jeff

[gpetilli]

Originally Posted by squirrel

Here's ours, which sounds like it's working about the same as theirs.

http://www.youtube.com/watch?v=jeLe2WWemWA

We never got around to trimming the top of the shooter, so the students are holding it at an angle. It's not quite 45 degrees. There is a Lexan flap behind the shooter, that is being held down over the top of the ball before inserting the ball into the shooter.

http://www.youtube.com/watch?v=jeLe2WWemWA

It seems that the amount you compress the ball is a critical parameter. Too little and you get slip, too much and you waste energy. We are looking at about 1.5inch compression of the wheel on the ball. Can you share what you are using and how you calibrated it?

[MrForbes]

In the video, the ball is being compressed to about 6". We're working on modifying it so we can do more testing.

[waialua359]

Originally Posted by squirrel

In the video, the ball is being compressed to about 6". We're working on modifying it so we can do more testing.

Jim,
could you have gotten away with less compression? Have you tried 1" and 2"?

[MrForbes]

We did get a chance to play with it, and it really does like the 2" compression. Also we noticed that having the ball escape while it is still compressed, results in quite a bit of variation in height. But using a flat plate that lets the ball uncompress gradually, allows very precise height control. So make sure your shooter has some "follow through" with the hood extending gradually away from the wheel.

[Gdeaver]

Squirrel, your plywood box looks very much like ours. We are using 2 wheels together but, about the same effect. Now how to get the balls off the floor, and cross the bump and get them up 2 the shooter and control feeding and aim and calculate shoot speed and and and.

[JesseK]

Jim, it looks like you have 2 wheels sandwiched together with a gap in between them. Is your 6" compression from the shroud to the edge of the separated wheels?

We got incredibly precise shots from the top of the key with a single pneumatic wheel and much less compression (< 1").

[MrForbes]

We have a single wheel. I expect there are quite a few different ways to tackle this problem....

[Mk.32]

Originally Posted by JesseK

Jim, it looks like you have 2 wheels sandwiched together with a gap in between them. Is your 6" compression from the shroud to the edge of the separated wheels?

We got incredibly precise shots from the top of the key with a single pneumatic wheel and much less compression (< 1").

What was the range?

[JesseK]

Originally Posted by Mk.32

What was the range?

We haven't tested the full range yet because it was an electric drill running it and we were testing things like energy absorption of the backboard, consistency of shot, best angle to approach the basket from the key, etc. Our results are similar to Jim's video, but I suspect that the reason for our convergence to similar results from different setups is the fact that we use a heavier/bulkier wheel with grippier tread versus his lighter/slippier -- both are tradeoffs between the designs and there doesn't appear to be a 'better' one so long as the implementation makes up for any potential shortcomings. It's pretty interesting.

As-is I suspect our shooter gets the ball far enough to cover half the field before it hits the ground again, which is good enough for a ball-delivery bot should the situation arise. Once we get a starting point for a more realistic motor/gearbox we'll adjust speed with some extra room for maximum "oomph" (and also some breathing room for drained batteries). It won't do full-field shots, but we're not trying to do those in the first place. At best we'll try to hit the mid or high goal from ~24 feet away.

[Brandon Holley]

Jim-

How are you guys doing on consistency? Looks like I saw 2 balls in the video being cycled through. Do you have more? Have you tried more? How about highly damaged balls versus new balls?

Nice work so far!

-Brando

[MrForbes]

With no feedback on the motor, we have to let the motor get back to speed for about two seconds before making the second shot, or it will be short.

The balls have lots of faint lines marked on them, and the labels are getting faded, but they don't appear to have any damage from the shooter.

[waialua359]

Originally Posted by squirrel

We did get a chance to play with it, and it really does like the 2" compression. Also we noticed that having the ball escape while it is still compressed, results in quite a bit of variation in height. But using a flat plate that lets the ball uncompress gradually, allows very precise height control. So make sure your shooter has some "follow through" with the hood extending gradually away from the wheel.

Jim,
We tried your suggestion with an extended top piece. It was a little more consistent shot after shot. We did not see a difference between a 1 1/2 and 2 in compression, for our design.
The AM wheel from 2008 is definitely much better than the grippy ones this year. We shredded 3 balls in 5 minutes with them. The smooth older ones definitely grip better with less slippage and doesn't damage the ball other than small scuffs.