Could a Robot Pool arm/machine

C.Milian said:
And this is simple? I don't see how execution of moving a sphere from point A to B with a cue by use of a human body can be replicated with incorrect, insufficient data...by an inanimate object.
Click to expand...

I don't think an inanimate object would be the best bet. We probably need something that can move. Modeling after the human body would be pointless we only want something that can strike a cue ball using a single degree of freedom for the stroke.

Most of the data can be collected with high resolution cameras which would provide much more accurate and sufficient data than can be gleaned by the human eye. Hygrometers provide much more accurate humidity than you can feel, and all this is logged over time and adjusted for.

The difficulty IMO is a matter of how accurately the machine could position itself to strike the ball. The algorithms and formulas already exist.


THAT BEING SAID, pool is a game. It's not for computers; it is for human enjoyment.
 
angldemn said:
I don't think an inanimate object would be the best bet. We probably need something that can move. Modeling after the human body would be pointless we only want something that can strike a cue ball using a single degree of freedom for the stroke.

Most of the data can be collected with high resolution cameras which would provide much more accurate and sufficient data than can be gleaned by the human eye. Hygrometers provide much more accurate humidity than you can feel, and all this is logged over time and adjusted for.

The difficulty IMO is a matter of how accurately the machine could position itself to strike the ball. The algorithms and formulas already exist.


THAT BEING SAID, pool is a game. It's not for computers; it is for human enjoyment.
Click to expand...

Then why use a cue to move the balls. A cue was built for the human to use. Another tool should be conceived based on the ability of whatever thing is using the tool. At the moment, the only written information I have ever come across is finding a contact point and sending the cueball to it. This is just logistics. It dose not represent a robot that can replicate a pool arm. The pool arm is way beyond contact point equating final relocation of the object. To think of it this way is like putting a big block on your brain when trying to figure out the proper way to move things.


It's not checkers. You can computerize checkers not something like a pool player. It takes thinking beings many years to use the cue properly, how does it make sense that you can create a robot to do this?
 
For a well programmed robot stroke is not an issue.

broken said:
Does any one know the difference between a robot poking at a ball, and a robot following through with it's stroke?

It seems that with humans, focusing on the follow through can sometimes help your stroke stay true and straight. In comparison to folks who "poke and pray", they may not focus on staying true and straight.
Click to expand...

Follow through for a robot would not be an issue. Even a human can poke at the ball and be accurate. In fact you can be more accurate without a follow through.

One of the key things to a god stroke is letting the cue do the work. Follow through tends to ensure that we are stroking straight as we contact the CB but that's all that is necessary, to be stroking straight WHEN you make contact.

Robots cannot be easily programmed to runout a pool table.

They are just getting to the point in Germany where using stereoscopic cameras, they can program a computer to drive a car successfully. The tolerances there are much lower than playing pool.

The only way it would even be feasible right now to program a robot to run out on it's own is to make a robot whose axes is greater than the size of the entire table and to use an ultra fine grid pattern on the pool table that can be visually converted into ultra fine coordinates allowing the robotic arm to go to any part of the table without upsetting its' base.

Then it MIGHT be possible, but you probably aren't going to be able to get any higher than good short stop accuracy even then at about .877.

Jaden
 
Jaden said:
... Robots cannot be easily programmed to runout a pool table.
...
Click to expand...
Virtual Pool in man-against-computer mode does a good job of running out the table. Each shot is performed with reasonable physics and sources of error. What would probably interfere with getting similar results on a real table is the vision part (as noted above) and the accuracy and alignment of the mechanics.

As far as the vision stuff goes, it's currently possible but expensive to determine the location of each ball to better than a tenth of a millimeter. That is better than typical human vision for most shots.
 
Last edited:
Bob Jewett said:
Virtual Pool in man-against-computer mode does a good job of running out the table. Each shot is performed with reasonable physics and sources of error. What would probably interfere with getting similar results on a real table is the vision part (as noted above) and the accuracy and alignment of the mechanics.
Click to expand...

That is not all their limitations. As example, you can build a machine to beat my grandmother, but to have a robot play like a true pool player, I'm sure it can't be done. At best, it can SHOOT better than a hack but it will never play like a player. Players do things with the cue only few can even imagine doing.
 
Don't drop the soap PJ

Patrick Johnson said:
There's no practical difference. Followthrough is only necessary for humans - so they don't screw up the part before the followthrough. Robots don't have this problem.


Right - except it's more like always.

pj
chgo
Click to expand...

This is elementary, not even breaking the surface of the cause of follow through, let alone it's purpose...ponder this while you eat those jam up joint burgers, then you will know if it is necessary. I'd clue you in but you never ask for it.
 
C.Milian said:
This is elementary, not even breaking the surface of the cause of follow through, let alone it's purpose...ponder this while you eat those jam up joint burgers, then you will know if it is necessary. I'd clue you in but you never ask for it.
Click to expand...

I like to learn everything I can. I humbly ask you the cause and purpose of follow through. Would you please explain?
 
Saved by Dr. Dave

Dodged a bullet in the middle of a thesis....my poor little fingers needed rest. Thank you Dr. Dave. You shouldn't give pool secrets away so freely. The God's will set your butt on fire. In this case the guy was sincerely looking for legit info. I couldn't say no either...it's part of our creed. Thank you sir.
 
The reason I dont think a machine can replicate a human while playing pool is best described using the bank shot as an example.

A computer can figure the exact reflection of the angle, meaning the degree of angle going into the rail equals the degree of angle coming out. But can the robot contemplate how much the rail will deform when using "X" amount of power. How much will the tangent line change when using 1 1/2 tips of left hand english with that same amount of power. How much does the room humidity effect the rails.

There is a long list of things that the robot isn't doing that our mind does in a split second. Whether or not there is follow through is on the back burner.


And as far as follow through is concerned here is my take. Does it effect the shot after final contact is made with the cue-ball? NO

But, it makes a world of difference when trying to accurately apply good fundamentals to a shot. It help train your brain to not poke at the ball and to finish your stroke.

Show a person with no follow through and I will show you a person who pokes the ball the majority of the time.
 
Cdryden said:
A computer can figure the exact reflection of the angle, meaning the degree of angle going into the rail equals the degree of angle coming out. But can the robot contemplate how much the rail will deform when using "X" amount of power. How much will the tangent line change when using 1 1/2 tips of left hand english with that same amount of power. How much does the room humidity effect the rails.
Click to expand...

Actually that's the point of computer learning I mentioned in the first post. The formulas for calculation exist as a guide, but "experience" allows the computer to continuously fine tune factors in aim, speed, angle, English and as many other factors as one would care to program. There are many techniques for "learning," and a number of different ones could be applied.


C.Milian said:
Click Dr.Dave's link Alice
Click to expand...

I have read this and many other of Dr. Dave's articles; I love his work. It's undoubtedly the most reliable and accurate information I've seen in pool instruction, and has been very helpful to me.

And, it does not support your argument in any way. Please don't bother posting your thesis, I think we've all had enough trolling in this thread Susan.
 
angldemn said:
Actually that's the point of computer learning I mentioned in the first post. The formulas for calculation exist as a guide, but "experience" allows the computer to continuously fine tune factors in aim, speed, angle, English and as many other factors as one would care to program. There are many techniques for "learning," and a number of different ones could be applied.




I have read this and many other of Dr. Dave's articles; I love his work. It's undoubtedly the most reliable and accurate information I've seen in pool instruction, and has been very helpful to me.

And, it does not support your argument in any way. Please don't bother posting your thesis, I think we've all had enough trolling in this thread Susan.
Click to expand...

You asked and now that you got something else which you think is reliable and accurate information is enough to throw this bullshit ball at me? WTF happened to being humble?
 
angldemn said:
Actually that's the point of computer learning I mentioned in the first post. The formulas for calculation exist as a guide, but "experience" allows the computer to continuously fine tune factors in aim, speed, angle, English and as many other factors as one would care to program. There are many techniques for "learning," and a number of different ones could be applied.




I have read this and many other of Dr. Dave's articles; I love his work. It's undoubtedly the most reliable and accurate information I've seen in pool instruction, and has been very helpful to me.

And, it does not support your argument in any way. Please don't bother posting your thesis, I think we've all had enough trolling in this thread Susan.
Click to expand...

This all sounds too familiar....yet you accuse me of trolling.
 
this discussion makes me think about people at work who think they are irreplaceable and then get the sad wakeup call one day that everyone is replaceable.
 
Maybe, maybe not. But if it can program it to get me a beer from the fridge and give me a
hand job, I might consider shooting a game with it.
 
Bob Jewett said:
Some current pool robots have no problem generating power and doing it with great consistency.
Click to expand...

Is that force? How does it relate to power? I can imagine an engine generating enough force to move a 500lbs door. Or is it electrical power that moves the door? It would take me a lifetime to generate the type of power it takes to swing that door a few revolutions. Is this the power you're talking about that these robots have?

What experiment is this you are working on and how is it funded?.....who is getting paid for their work from who?
 
C.Milian said:
Is that force? How does it relate to power? I can imagine an engine generating enough force to move a 500lbs door. Or is it electrical power that moves the door? It would take me a lifetime to generate the type of power it takes to swing that door a few revolutions. Is this the power you're talking about that these robots have?

What experiment is this you are working on and how is it funded?.....who is getting paid for their work from who?
Click to expand...
I think Wikipedia would be a good place for you to start to fill in the holes in your physics education -- if you feels it's really necessary, which for most pool purposes it isn't.

The experiment I referred to was the so-called "Jacksonville Experiment" that I and several other fanatics worked on in 1998. We paid for it ourselves, although Clawson Cues shared the cost of the high-speed camera and provided the table and Iron Willie for us to play with for a week. Search for Jacksonville on this page: http://www.sfbilliards.com/articles/BD_articles.html
 
You must log in or register to reply here.
Back
Top