Gripper Box 3.0: FEEDBACK/TESTING REQUESTED

[joecross]

This is Jason Bryant's proposed replacement for the tensioner box on the Raptor Reloaded. It provides adaptive grip so that each finger in the hand will adjust to the shape of the object being held.

We want to get this integrated into the official Reloaded design, so please reply here to let Jason know you can help fabricate and/or test pilot it.



Adaptive grip gives the user a better hold on objects. On the regular Raptor design, picking up a cup might cause all the force to be on the thumb and the index finger. The other three fingers will hang in the air without contributing anything to grip power or stability. With the Gripper Box, the power is balanced between the fingers and they take on the shape of the object being held, making the grip more secure.

The box was based on the whipple tree designs by Skip Metze.

STL Files and Assembly Instructions: https://www.dropbox.com/sh/rpln62laoxan ... BctVa?dl=0

R&D thread for the Gripper Box: viewtopic.php?f=11&t=12

[Jason M Bryant]

Give me a day before anyone downloads this. I'm looking at some leverage issues.

[Jason M Bryant]

I've reworked a few of the parts now. If you've downloaded it within the last few days, get the whippletree file again. It should be ready for testing now.

[joecross]

I printed all of the parts but got a bit stuck on assembly. Could you post a few pics? Especially of how the pin and box pieces go together.

IMG_20150331_194420.jpg (198.41 KiB) Viewed 7797 times

[Jason M Bryant]

Hi Joe. Thanks for testing!

I'll try to get some pics posted. I'll need to restring my hand to take them. Unfortunately, this is my busy time of the week, so it will either happen in the next few hours, or two days from now. For right now, here's a picture of the fully assembled box so you can see how they fit.

DSCF3215.JPG (50.05 KiB) Viewed 7792 times

Did you get the instructions file?

[Jason M Bryant]

Okay, download the Gripper Box Instructions file from dropbox again. I've added a ton of pictures that should clarify things.

[joecross]

Got the new instructions - thanks! Of course I only need one pin and one box - duh...

[Jason M Bryant]

I've replaced the thumb_tensioner_pin. The previous version was too short.

[Jason M Bryant]

I've changed the orientation of the swivel pins. The geometry is the same, but I think they will be stronger when printed on their sides.

I also changed the names of the files to make things more clear.

[LesHall]

OK I"ve got my gripper box files and instructions. I tried to read the instructions but they kinda require that you have a hand and the gripper box parts in front of ya, so i will wait on grokking them. I will get my filament within a few days and then i can go to town and rock the box! This hand will also have the special picking things up friendly fingers and the Spidey Sense Touch and Temperature (TnT) electronics mounted on a Piccadilly rail made by Skip Meetze. It'll be a showcase of our combined talents!

Les

[hertfordkc]

Jason,
I finally got around to rigging the Raptor Classic and Extended Gauntlet with the files which you so promptly provided. Thanks again.
While there is some adaptive grip, it is not working as I feel it should. I printed at 133%, and everything looks great. However, when testing the grip, the whippletree appeared to be very limited in its pivot. I tied the whippletree about 3mm from the tensioner. Shouldn't that be enough? Any other thoughts?
Kent

[Jason M Bryant]

Hi Kent. Thanks for the feedback!

How much can you straighten one or two fingers while the others are still gripping? With my 100% hand I can do this:

DSCF3255.JPG (41.68 KiB) Viewed 7381 times

That doesn't cause the whippletree to move very much. Straightening two fingers makes it move a little more, but it isn't a huge amount.

Here's another test that I do to see if all the fingers are adapting to a shape. I put the hand on the edge of a table and grip the fingers on the edge. Then I start turning it to see how much angle the fingers can adapt to.

DSCF3256.JPG (40.88 KiB) Viewed 7381 times

Try those couple of things and see how much the fingers adapt. What you're looking for is that the fingers that have bent the most have some force on them. If they're just hanging there, they're not useful. Also, try picking up a few things around the house, like cups. That doesn't require as much adaptation as when we pull a couple of fingers completely straight.

Here's another thing that you could check for me, if you don' mind. On mine, the swivel pin sticks up a little, pushing the whippletree against the roof of the box. If that's happening to other people, then I might need to adjust the screw holes on the box a bit. Is that happening to you?

Thanks again!

[hertfordkc]

I found two or three things that were affecting the flexibility: One elastic over tightened, left two fingers tightened more than right two fingers, perhaps a sticky joint. I'm going to play with it some more. It does appear to meet one of the recipient's criteria: the hand can hold a trumpet.
In other news, I'm worried that your project list may be shrinking.
How about locking pins for the whippletree, so the grip can be either fully adaptive, partially adaptive, or fixed?

[Jason M Bryant]

Don't worry about my project list. I'm obsessed with fingers now!

I'm actually going to start experimenting with the dental elastic solution from the Falcon hand. I think a lot of the times that I think adaptive grip isn't working, it' just that the elastics aren't stretching easily enough.

Tell me more about the locking pins idea. What situations would be better with partially adaptive or fixed fingers?

[hertfordkc]

My suggestion for adaptive/non-adaptive fingers was based only on my observation of the hand working with non-adaptive fingers to carry a lunch pail or a grocery bag, picking up small pieces of candy, all of which work, and trying to hold a trumpet, which a recipient specifically requested, which doesn't work without adaptive grip. I hope to give the adaptive grip hand to him for testing in about 2 weeks. The hand is ready, we just haven't been able to get together.
Don't expect any recipient-based feedback from here is less than a month. The young man is receptive to different designs, but he is very deliberate in his judgments. I'm sure others will have feedback much sooner.
BTW, the hand has thinner fingers. I haven't measured to see exactly how much thinner. So much for doing incremental development and testing.

[joecross]

I printed out the latest Gripper Box at 100% to fit a demo hand I'm finishing up. Unfortunately, I couldn't fit anything into the whippletree or pin except a single loop of .5mm cord. The photos seem to show a metal wire. Is that what I should be using?

[Jason M Bryant]

I'm using fishing wire. At the back I'm using some strong double braided stuff. In the front I'm using some cheap wire that I only use for testing. It's not metal, it just looks that way because of how the light is catching it in the photos.

The good fishing wire that I use it pretty strong, but pretty thin. The size is #8, but I don't know what that means in real measurements.

The hole in the whippletree is a lot bigger than .5mm. It's actually 2.5mm in diameter. You should be able to fit a few loops of wire through it. Is it possible that the holes are not printing very clearly? Do you an external fan pointed at your extruder? If something is blocking it, you could try bending a paperclip and pushing it through to clear the tunnels.

I'd love to see a photo of the holes on your whippletree. If you can post one, that would help me see if it is printing out properly.

[Jason M Bryant]

Joe, from the photo you sent me, it looks like you fit two loops through the whippletree with room to spare. Is that right?

From the angle of the picture, I can't see the see the hole in the swivel pin.

[joecross]

Yeah, I was able to get two loops, but I was only using a single thread from a braided nylon cord. Maybe that's strong enough?

[Jason M Bryant]

It should be enough. Remember, 100% is the size for a three-year-old. With decent cord, I'd be much more worried about tying a good knot than two strands of cord breaking.

Well, that's what testing is for, at least.

I'd really like to find a solution that does away with tensioner pins entirely. I haven't found anything I'm happy with yet.

[Jason M Bryant]

I tried making the whippletree a little wider. I thought that might give a little more leverage and make it easier to pull up the left two fingers while pulling the right two fingers down. It wasn't a noticeable improvement.

I'm really not sure that it is necessary, anyway. The test we all go for is to hold two fingers up and try to completely close the other two. While this looks really cool, it's actually very difficult (and painful) to do this with a normal flesh hand. The range of adaptive motion with the gripper box is a little less than with a flesh hand, but it's still more than enough to pick up a cup.

[hertfordkc]

I fitted a Raptor Classic with a Gripper Box for Alex this weekend. He received a conventional Raptor Classic about Jan. 1, so he has it as a backup.
I thought that I took several pictures, but botched something. The pictures would have showed him picking up a CO2 cartridge, and holding a trumpet, which he could not do with the conventional hand. I was pleased to see that he had not lost dexterity vis a vis the conventional hand, and was elated at his ability to hold a trumpet which he plays. He couldn't hold the trumpet with the conventional hand.
Unfortunately, Alex is very reserved, and I don't expect any feedback for several weeks ... or longer.
I have two questions concerning the adaptable grip: 1) Is there a critical size for the whippletree? In other words, if it were too big, perhaps the fingers couldn't be closed. Conversely, if it were too small, it might be impossible to close just one finger. Is the whippletree (and box) really scaleable? 2) Would there be value in having a brake on the whippletree so that the hand could be operated in either mode? I hope Alex will answer #2
Kudos to you and Skip Meetze.

[Jason M Bryant]

Thanks for the feedback. I'm so glad he's getting something out of it!

I don't think there's a minimum size. I do all of my testing at 100%, and there seems to be decent adaptability in the fingers at that size. I find that there isn't too much trouble in getting the fingers around a cup or most other objects.

I don't think there's a problem with large scales, either. If anything, there is more leverage with a larger whippletree, and it is easier to pull the fingers around things. The one hand I've made for a recipient was at about 180%, and it seemed to work pretty well.

I don't think there's a need for a break. From my testing, it feels like anything that we can do without adaptive grip can also be done with it. My test was picking up a bucket with five pounds of weight in it. Since I was pulling straight up, the fingers didn't slide around as long as the grip was strong enough to hold it. Even when I angled the hand a bit, the fingers only moved a little, redistributing the load.

However, more testing would be good. Are there other situations besides holding a bucket? If you or your recipient come with with something where a fixed grip might be better, let me know. We can start testing and see if there's a noticeable difference.

Oh, and we *definitely* need some pictures of him playing the trumpet! I think I need that as the background on my desktop!

[droomurray]

Eventually got these printed out to try with my new crimping system

Drew.

[franciscofernandez]

Hi
I finished making yesterday a gripper box, it was a lot easier than it looked at first impression.
I made an alteration to my raptor with an screw on addon that allows to have an oposable thumb using the existing raptor design just by attaching it. The thumb and index worked great but without adaptive grip the other fingers would not bend all the way so i looked up for an adaptive system and found the gripper box. The 3.0 is definitely the best, fewer parts, good solid design and easy to build.

In this video, http://youtu.be/Ou715-H7NGY
you can see the raptor with gripper box and opposable thumb attachment.
I grab a paper roll, as you can see by the pressure on the paper and the tightness when i shake it, it has an adaptive grip without losing strength which i thought it was gonna lose some. In fact i think is stronger. I also grab a paper cutter and even though is quite a particurlarly heavy one, it holds it without problem.

[Jason M Bryant]

Hi Francisco. This is really neat. I like the opposable thumb system. Unfortunately, I can't see the youtube video that you linked to. It was set to private. Could you change it to public?

[franciscofernandez]

[GreggDennison]

I printed this out for Luke to try out along with the revised Raptor Reloaded(using the rubber bands). I followed the instructions but I have noticed that the index finger bends way too much and too fast, not allowing it to pinch with the thumb, which is adjusted accordingly. I'm pretty sure everything is adjusted right. I will try to get some video posted.

[Jason M Bryant]

Gregg, the first thing I would suggest is that you check the tension in the elastic cables. If one is not as tight as the others, it will bend sooner.

[GreggDennison]

Luke has been testing this out for a couple weeks now and I've not had any issues with it. But I haven't really noticed it make a difference either. Maybe I don't have it put together just right. I don't know. I followed the instructions.

One thing I did notice is the index finger curls a lot more than the other fingers. It's not a gradual close, the tip closes first. The tip actually flexes completely on the joint, then the middle of the finger closes whereas all the other fingers bend at the knuckle block first. Not sure if this is intended. I can send a pic if you need me too.

Also, since I tested out the whole revised Raptor Reloaded, I feel like the "posts" for the rubber bands need to be slightly taller. It would make replacing the bands extremely easier. I have had to resort to using needle nose pliers on a couple fingers. I'm also using "loom bands" that kids use for bracelets and whatnot. 600 in a pack for $3 and they've held up pretty well but I do have to double them up on all but the index finger first joint.

Edit: just saw I already posted about the index finger. Been so busy, I forgot. I will look into what you said tho.

Double edit: You're right, it was the band. For some reason, I thought something was weird about that joint when I first banded it so I left it loose on purpose. Oh well, it's working better now. Thanks.

[Jason M Bryant]

I'm glad you figured out the issue, Greg. Thanks for the feedback, it's very useful.

What size is the hand?

[GreggDennison]

It's a 130%er. If you watch any of the Maker Faire Bay Area footage or any of the pictures posted by anyone there, you will see him wearing the Raptor Reloaded Revision. Your fingers, Gauntlet with string bars, Gripper Box, Andreas' thermomesh palm. And easily said, by Luke himself, it's his most favorite hand so far. He likes that he can repair the bands if they pop off and the palm especially. He wants to try out a thermomesh Gauntlet next.

[john.diamond]

I've attached and archive containing the STL and OpenSCAD source files for the modified reverse dovetail gripper box and whippletree components that I discussed during the Monday 3rd August R&D hangout.

[Jason M Bryant]

Thank you. I look forward to playing with this.