Universal Robotic Gripper [Archive]

JohnBoucher

27-10-2010, 12:01

Coffee grounds in a balloon drawing a vacuum

University of Chicago, Cornell researchers develop universal robotic gripper (http://www.engadget.com/2010/10/27/university-of-chicago-cornell-researchers-develop-universal-rob/)

JesseK

27-10-2010, 12:37

:ahh:

That just may trump the roller claw.

Jon236

27-10-2010, 13:29

very nice.......probably coarse grind......expresso grind tends to stick

Chris is me

27-10-2010, 13:34

EricVanWyk

27-10-2010, 14:15

Sweet!

Reminds me of the jammer: http://www.youtube.com/watch?v=SbqHERKdlK8

jspatz1

27-10-2010, 21:13

Seems like an adaptation of the suction cup to better distribute vacuum force over objects with uneven surfaces. I want to play with it, and see how much better it holds various objects compared to suction cups. I wonder how feasible this is in FRC.

I don't believe it is applying suction to the object in the traditional sense as with a suction cup. That would imply that the balloon material is porous, and that would leak vacuum all over. If I correctly interpret what is happening, the vacuum inside the balloon is causing the balloon material to constrict around the object, with the resulting friction causing it to "grip." Notice all objects are 3D, not flat, and that they force the balloon around the object before gripping it. Great example of an totally new idea.

Andrew Bates

27-10-2010, 22:00

Now here's what I'm wondering, could we legally use it on a FIRST robot? If FIRST ever made a game with small objects this could be useful...

gblake

27-10-2010, 22:06

I don't believe it is applying suction to the object in the traditional sense as with a suction cup. That would imply that the balloon material is porous, and that would leak vacuum all over. If I correctly interpret what is happening, the vacuum inside the balloon is causing the balloon material to constrict around the object, with the resulting friction causing it to "grip." Notice all objects are 3D, not flat, and that they force the balloon around the object before gripping it. Great example of an totally new idea.

I think the balloon is porous because of the sounds in the video.

It might not be porous, but if it isn't I don't think some of those objects would stay against it. The egg and the small items in particular.

Jon236

27-10-2010, 22:17

Now here's what I'm wondering, could we legally use it on a FIRST robot? If FIRST ever made a game with small objects this could be useful...

We were only required to use those delrin wheels in 2009........let's see what rules await us in January!

Alan Anderson

28-10-2010, 08:30

I think the balloon is porous because of the sounds in the video.

It might not be porous, but if it isn't I don't think some of those objects would stay against it. The egg and the small items in particular.

Nope, the surface is not porous. It is grippy. It picks up the egg in much the same way you would, by squeezing it.

My son had a "Vac Man" toy (made by the same company that sold "Stretch Armstrong") that works along the same principles. Based on what it could do, I can imagine a much fancier universal gripper having "fingers" that might be capable of handling smaller or more delicate objects, or even curling into hooks to lift things with handles.

Greg Young

28-10-2010, 12:39

The press release says it's a latex party balloon, so definitely not porous.

And yes, probably a coarse grind since it depends on the material locking together when compressed.

I'm awestruck. Simple. Elegant. I would never have thought of this.

Chris is me

28-10-2010, 12:50

I'm at a loss as to how one should calculate grip force with that. Anyone have any ideas?

JesseK

28-10-2010, 13:19

I'm at a loss as to how one should calculate grip force with that. Anyone have any ideas?

The good ol' fashioned method of picking up heavier weights until the weights started falling would probably be the least-painful way for most of us to figure it out. It all depends on the friction of the balloon, tensile strength of the balloon, and I'm sure the vacuum plays a large part in it too. Then we'd have to consider the angle of grip as well.

Brandon Holley

28-10-2010, 13:41

The good ol' fashioned method of picking up heavier weights until the weights started falling would probably be the least-painful way for most of us to figure it out. It all depends on the friction of the balloon, tensile strength of the balloon, and I'm sure the vacuum plays a large part in it too. Then we'd have to consider the angle of grip as well.

In a similar fashion...pick up an object, place a load cell or some kind of spring gauge on the object in the gripper and pull til it pops out. That will give you an idea for that particular object. The gripping force would obviously vary from object to object.

-Brando

Alan Anderson

28-10-2010, 13:42

I'm at a loss as to how one should calculate grip force with that. Anyone have any ideas?

...It all depends on the friction of the balloon, tensile strength of the balloon, and I'm sure the vacuum plays a large part in it too. Then we'd have to consider the angle of grip as well.

It seems to me that grip force would be essentially independent of the balloon. The only relevant factors I can think of immediately are the amount of "shrinkage" of the filler material when vacuum is applied, and the stability (probably shear strength, mostly) of the vacuum-compacted material.

(The friction of the balloon definitely matters when you try to figure how much weight you can pick up with a given grip force, of course.)

lynca

28-10-2010, 13:59

We could use this for VEX Round-up right now !

jamie_1930

28-10-2010, 17:01

I'm at a loss as to how one should calculate grip force with that. Anyone have any ideas?

My thoughts would be that you need to find the surface area of contact with the balloon, the coefficient of friction between the balloon and object, and finally the force with which the vacuum is pulling the balloon in. Then from there you would be able to calculate the force with which the object will be held, however those are my initial thoughts which I feel are far off from what you would actually need in order to calculate the force.

JVN

28-10-2010, 18:10

In a similar fashion...pick up an object, place a load cell or some kind of spring gauge on the object in the gripper and pull til it pops out. That will give you an idea for that particular object. The gripping force would obviously vary from object to object.

-Brando

I would think it would vary GREATLY from grip to grip as well. I wish our team had build a suction cup at some point so I could go play with this concept and learn more.

Radical Pi

28-10-2010, 18:38

If you're serious about using this, I'd think that the coffee bean usage is a clear violation of the no food in the pits policy :P

Actually, how well would this work with styrofoam pellets? Does this require that the material inside the balloon not compress?

Jared Russell

28-10-2010, 20:10

The granules must not compress in the vacuum for this type of manipulator to work (well).

Jon236

28-10-2010, 21:55

If you're serious about using this, I'd think that the coffee bean usage is a clear violation of the no food in the pits policy :P

I think the rule deals with toxic substances....not ground coffee beans

Brandon Holley

29-10-2010, 08:57

I would think it would vary GREATLY from grip to grip as well. I wish our team had build a suction cup at some point so I could go play with this concept and learn more.

Yes I completely agree with that.

I could envision there being some objects that are very tough to pull out of the grip, while others are only just barely being held in.

RoboMaster

30-10-2010, 23:05

I wonder how you could make this with FIRST resources. The compressor could certainly be used for suction. You'd have to have some sort of filter system/mechanism to keep the coffee grains from sucking into the compressor. You might just be able to put a screen inside the balloon. It might be tricky to use this gripper and air tanks for pneumatic pistons, since both require different uses of the compressor and at different times. Then you'd have to figure out how to make/machine an attachment from the balloon to the air hose. This attachment would also have to have mounting holes on it for mounting it onto an arm or the robot.

apalrd

31-10-2010, 09:29

The compressor can't suck. It only compresses air, it doesn't create vacuums.

A few years ago (2007), a device was included in the KOP to create suction from air pressure, as well as two suction cups. (I forgot what it is called, but I know we still have ours in our shop).

I like this gripper. I still like roller claws better.

Edit: I looked it up. The device in the 2007 KOP was a Venturi valve.

RoboMaster

31-10-2010, 20:21

I thought the compressor is what made that suction kit work. But still, you could hook up a tube to the "in" side of the compressor and it would suction, right? I even remember that there was a rule stating that we couldn't do this (implying that it could be done).

Mark McLeod

31-10-2010, 20:41

The "in" for the KOP compressor is a felt patch on the side. There's no place to attach any fitting, and the draw there isn't all that great anyway.
It's not one of the two ports opposite each other on the top of the compressor. Those are both output ports.

A Venturi works because a flow of air is constricted by the device to speed it up. An increase in air velocity is matched by a corresponding decrease in pressure. The faster moving air draws a vacuum. Like blowing across the end of a soda straw.

So continuously forcing compressed air through the Venturi device results in a vacuum. It doesn't take a great deal of air to make it work.

RoboMaster

31-10-2010, 20:44

Oh, I see, I thought one of those output connections you talked about was an input.

But of course, like this Venturi device, there are other options.

jamie_1930

31-10-2010, 23:23

There should be fitting you can purchase that will take the air going in to create a vacuum effect, similar to how a carborator works. This was mentioned to me by a mentor last year on how we could utilize suction cup. I'll try and find an example of this fitting asap.

apalrd

01-11-2010, 07:46

There should be fitting you can purchase that will take the air going in to create a vacuum effect, similar to how a carborator works.

That would be a Venturi device.

marisaDKNY

01-11-2010, 14:27

Are there any teams out there thinking of researching this idea to implement it possibly to the FIRST Game?

jamie_1930

01-11-2010, 17:57

Are there any teams out there thinking of researching this idea to implement it possibly to the FIRST Game?

I myself think it is a possibility and am planning to introduce this, as well as the jammer, to my team on Wednesday. However actually implementing, this scaled device, would depend greatly on the objects that we will be manipulating this year, but in addition a larger model would be great if it can be optimized during the season. I think this would be an interesting challenge to do for the season, but right now I'm sitting here wondering if we would legally be allowed to use this for in technology many things become patented, and I'm not 100% sure if University of Chicago would allow us to, however am hopeful they would. What do you guys think?

dlavery

01-11-2010, 21:01

I wonder how well it could pick up one of these:

http://ecx.images-amazon.com/images/I/51mCdS4AIyL._SS500_.jpg

-dave

.

JesseK

01-11-2010, 21:38

[Insert random relation to water game here]

Mark Holschuh

01-11-2010, 22:51

Tennis anyone?

Jared Russell

01-11-2010, 23:06

Clearly we are going to be picking up pens this year!

RoboMaster

02-11-2010, 00:04

AARRAREARGAARGHHHHH!!!!

DAVE YOU QUIT MESSING WITH US!!!!
YOU'LL PAY FOR THAT! :mad::mad::mad:

[/stew]

;)

RMiller

02-11-2010, 13:52

Clearly we are going to be picking up pens this year!

Or maybe penguins. Or maybe Nittany Lions.

Daniel_LaFleur

02-11-2010, 17:54

I wonder how well it could pick up one of these:

http://ecx.images-amazon.com/images/I/51mCdS4AIyL._SS500_.jpg

-dave

.

... and so the red herring season begins :D

.

BJC

02-11-2010, 18:06

Tennis anyone?

On the contrary, I believe we can now safely eliminate tennis balls from the long list of possible game objects for next year.
Thanks Dave!
Unless... no no no, he wouldn't give the game piece away that easily, would he? Of course, we would never know until the actual game. Better just to pretend that Dave simply randomly posted a tennis ball with no inherent meaning at all and forget the whole thing.

rsisk

03-11-2010, 02:47

Of course he did end his post with a morse code "e"

Siri

03-11-2010, 09:50

Of course he did end his post with a morse code "e"Yay, pasta!

jamie_1930

03-11-2010, 13:07

Of course he did end his post with a morse code "e"

I believe I saw an "s" in there too

Bjenks548

03-11-2010, 18:25

I believe I saw an "s" in there too

Penn + e +s? Pennies? aka either the hardest surface to play on or 2 the hardest game piece to manipulate/ not let destroy your robot ever?

RoboDesigners

03-11-2010, 19:59

We could use this for VEX Round-up right now !

Totally agree! (although there are probably easier ways to pick up those tubes...)

lbarger

03-11-2010, 22:12

I wonder how well it could pick up one of these:

.

This thread was started to discuss the merit of a 'universal' gripper. It is shown gripping the fine rim of glassware which historically has been a challenge for classic and anthropomorphic grippers. If it is truly universal, it would also be able to grab a relatively large object as well. I too wonder how it would handle a tennis ball. How far around the side will it have to 'flow' before it would be able to pick up the tennis ball?

Karibou

03-11-2010, 22:22

Of course he did end his post with a morse code "e"

Yay, pasta!

I don't envy field reset. They're going to have a lot of cooking to do...

ChrisH

03-11-2010, 22:39

I wonder how well it could pick up one of these:

http://ecx.images-amazon.com/images/I/51mCdS4AIyL._SS500_.jpg

-dave
.

One of the first objects the gripper picked up in the video was a hemisphere. So the shape shouldn't be a problem. You might have to have a much larger gripper to have a workable size ratio with respect to the ball.

For our application a much more important question would be "Why use use this 'universal gripper' to pick up a single ball?" Usually, if we have small objects we need to handle many of them. I don't see any team picking up more than 8 or 9 in a match with this method, unless they do some pretty fancy automation. uh oh

Grim Tuesday

14-11-2010, 22:45

Back on the Gripper, I think my team needs to get in contact with Cornell! We live just down the hill from them, that gripper could come in quite handy :P

JesseK

16-02-2011, 16:34

so....

Who built one of these? Or, if you do not want to reveal yourselves yet -- will you at least take some video at your Regional competition?