Experimental Jamming 'Beanbag' Robot Gripper
ฝัง
- เผยแพร่เมื่อ 18 เม.ย. 2021
- Get your first 10 PCBs for free at www.pcbway.com/
I've been thinking about building a jamming or 'beanbag' gripper for some time. This gripper uses a flexible membrane full of granular material which can form around the item to be picked up. The air is then sucked out of the membrane which cases the grains to be pulled together so that is becomes rigid. This is like a brick of vacuum packed coffee, but in this case I'm using BB pellets. My membrane is made from a squeezy toy and the rest of the assembly is 3D printed. I'm using a Dynamixel servo to pull a 200CC syringe which pulls the air out of the membrane.
CAD is at: github.com/XRobots/JammingGri...
Thanks to ROBOTIS for the Dynamixel servo: Check out: www.dynamixel.com/
This servo is: www.robotis.us/dynamixel-xm54...
You can support me on Patreon or buy my Merchandise:
***************************
Patreon: / xrobots
Merchandise: teespring.com/stores/james-br...
***************************
Affiliate links - I will get some money of you use them to sign up or buy something:
***************************
Matterhackers 3D printing supplies: www.matterhackers.com?aff=7500
Music for your TH-cam videos: share.epidemicsound.com/xrobots
***************************
Other socials:
***************************
Instagram: / xrobotsuk
Facebook: / xrobotsuk
Twitter: / xrobotsuk
***************************
CAD and Code for my projects: github.com/XRobots
Huge thanks to my Patrons, without whom my standard of living would drastically decline. Like, inside out-Farm Foods bag decline. Plus a very special shoutout to Lulzbot, Inc who keep me in LulzBot 3D printers and support me via Patreon.
HARDWARE/SOFTWARE
Below you can also find a lot of the typical tools, equipment and supplies used in my projects:
Filament from: www.3dfuel.com/
Lulzbot 3D Printers: bit.ly/2Sj6nil
Lincoln Electric Welder: bit.ly/2Rqhqos
CNC Router: bit.ly/2QdsNjt
Ryobi Tools: bit.ly/2RhArcD
Axminster Micro Lathe: bit.ly/2Sj6eeN
3D Printer Filament: bit.ly/2PdcdUu
Soldering Iron: bit.ly/2DrNWDR
Vectric CNC Software: bit.ly/2zxpZqv
Why not join my community, who are mostly made up of actual geniuses. There’s a Facebook group and everything: / 287089964833488
XROBOTS
Former toy designer, current TH-cam maker and general robotics, electrical and mechanical engineer, I’m a fan of doing it yourself and innovation by trial and error. My channel is where I share some of my useful and not-so-useful inventions, designs and maker advice. Iron Man is my go-to cosplay, and 3D printing can solve most issues - broken bolts, missing parts, world hunger, you name it.
XRobots is the community around my content where you can get in touch, share tips and advice, and more build FAQs, schematics and designs are also available. - วิทยาศาสตร์และเทคโนโลยี
![เก่งแต่ลับหลัง (FINSTA) - ส้ม มารี (Zom Marie) Ft. Punch 4EVE [Official MV]](http://i.ytimg.com/vi/OOolBU2aC70/mqdefault.jpg)
"I 3D printed a funnel" when you are on another plane of extistence and you don't need to buy common household items anymore
Came down to say this
six hours later
It's one of the nice things about having a printer, you can just make custom common items.
@@DanielBeaver this. It's probably for precision if anything. If he can get bb pellets he can get a funnel
Why not? Is cheaper and faster than ordering one
I think you simply just need to use smaller bb's. The bigger bb's allow a much bigger gap in between the bb's and the objects with a more sophisticated shape.
Yeah, with smaller balls it would increase the contact area and that should help.
Maybe softer balls (the original ones) may work, too?
I once saw a video where they used grinded coffee (the powder) in place of BBs.
I think that would work much better.
I've seen a similar gripper that used coffee grounds; that's probably the same idea!
How about replacing BBs with a less soft shape? They could interlock in order to retain its vacuum shape better. I'm thinking of the tetrapods that are used to break waves on shores.
We use a similar technique for pre-hospital splints for fractures, and even a whole body mattress for possible spinal injuries. I’m sure the beads are expanded polystyrene, as they’re solid, but with a little bit of give which helps them mould together when a vacuum is applied, plus very light.
came here to comment the same. It's probably the same stuff those huge bean bags are filled with. I bet a simple balloon of any size filled with those would work better because of the size and give of the beads.
What are pre-hospital splints and mattresses? Something to use in ambulances on the way to a hospital? Sounds pretty cool if it's that.
@@olekaarvaag9405 it is. it forms to the individual's contours, without pressure and hardens with vacuum to stabilize. great, if there might be spine injuries or to keep fractures in place. any movement would cause more damage and, frankly, hurts like a bitch.
@@therealpanse had to ride 45 minutes to a hospital holding my left arm in my left hand because of a break. can confirm, every little bump in the road hurts
@@kingmasterlord had the same experience with a broken shoulder, the weight of the whole arm tugging on it even with the smallest vibration. Not fun. they make those splints like arm slings too, and it gets strapped to your body, so it supports the weight.
Great job on your video! Another source for the loss of vacuum that you are experiencing is with the vinyl tubing that you are using to connect the syringe to the funnel. If you were to use mesh reinforced tubing, and an overall shorter length of tubing, you would see less of a loss of force. But really, great job!
4 screws,,, screw it on with 4 screwws
Yo the king of grippers ian davis is here i love your vids
There was a Stretch Armstrong villain called Vac Man that used much finger grains to achieve a similar effect; you stretch him into a position then vacuum out the air to hold the pose. I think you'd have more success on less detailed objects with finer grains. It looks like the original does this.
I agree smaller balls will enable finer griping resolution, and potentially better mechanical gripping
I knew Id seen this somewhere
As I recall the inside of the vac man was something granular but irregular, something like dried chopped millet seeds, not sand because that would be too heavy but definitely something with a non spherical shape, irregular grains are going to jam together more effectively than spheres.
Am I the only one that realized that you have made a functional one of the toilet plunger hands that Daleks have?
I had the exact thought.
Davros, the early years
@ wins the internet
Nope, through it I thought this would great for a Dalek build.
To be fair, daleks are pretty useful for their occupants...
i'm still attached to the fact that u 3d print a regular funnel.
...I've done that too. Sometimes you just gotta 3d print for the fun of it
I somehow doubt he didn't have an appropriate funnel at his home ready.
Also the fact he didn't recycle the funnel as the cup.
yes that's kinda overkill.
I chuckled a bit when he did that. But hey, if you have a 3d printer, a lot of plastic and don't have a funnel, why not? I assume he used vase mode and it probably didn't even take that long.
@@olekaarvaag9405 Which is the easiest print setting to get nice and still it looks wacky, he needs to tune it just a bit.
It wouldn’t last too long but I think I’d try literally just using coffee grounds first. With a conical burr mill you can make the grounds any size you want and I think the rough angular shapes cause the grounds to lock up nicely when under pressure.
in fact I've literally seen such grippers using ordinary coffee grounds, and it works great! The one in this article uses coffee, for instance: spectrum.ieee.org/automaton/robotics/industrial-robots/universal-jamming-gripper
i imagine it'd be easier to use fine sand
"alien frog spawn" killed me lmao i was not expecting that
watching you screw around with the first stretchy toy felt invasive lmaoo
I goatse that coming
How It's Made: Caviar
😏
Hey James fill it with coarse ground coffee to the brim, you're welcome.
And your funnel diameter should be max 1/3 of your squeeze ball.
biggest flex of all time xD "i 3d printed a funnel"
3D printed funnels are great if you want a very small one or a thread at the end
In addition to the size of the particles inside, the surface texture of the membrane probably has a big effect on grip strength
Stretch Armstrong had a villain toy that used this. It had much smaller pieces inside compared to those bearings. I would say a smaller "grain size" and a rounder shape as you suggest would be great.
A student of mine built a universal jamming gripper like this as part of his final-year mechanical engineering project where he was automating a fruit packing line (he also did pick and place operations using machine vision). He found that using much smaller granular material was the trick required to pick up the apples. He used a pneumatic cylinder driven by a servo instead of a syringe, which is much heavier and more expensive, but worked very well.
The front of the membrane that you have secured with that ring: it's being stretched out and away from center, while the rear of the membrane still has lots of free play. All the weight of the bbs are pushing forward against the object you are lifting, while the front of the membrane which needs the most 'play' is restricted by the ring. Try securing the rubber to the rear of the fixture directly and leaving the front half completely free to contort around the objects to be lifted without being pulled away from center. Great video!
I love that you found a way to legitimately compare apples and oranges :)
reminds me of the mail sorting machine they built for the train on manlab
Is this an inside joke? Because I swear that was a tomato. I have never seen an orange that looks like that.
i think you should use smaller bbs or something smaller so that it can grip things better
I laughed so hard when I realized he was sucking it with his lungs this entire time
Might be beneficial to have separate bag regions, like four surrounding a central one. Outer regions vacuumed for shape then inner region vacuumed for grip. Separate regions would also reduce the need for the massively large bags, and plumbing could be reduced by connecting all the outer regions together as the concept has them functioning at the same time.
I feel like the spherical form of the gripper is important. The spherical shape works together with the inner pressure to "hug" whatever you want to pick up whenever you press it against it, having it "flat" makes the surface tension work against you, giving the BBs a hard time huging the object. I imagine it like being stomped by an enormous gym ball vs a trampoline, the gym ball is going to hug and form my shape better (weird example, I know)
Yep - I'll probably make a custom one next time.
Yeah, exactly that.
The original one was a large sphere that, when the air was pumped out, would contract - which applied a bit of inward-pressure on the object it enevloped.
With the membrane being fixated at the perimeter sucking the air out will pull the membrane away from the object making it a far looser grip.
Two improvements come to mind: smaller grains and perhaps some method of blowing air around the inside of the membrane to liquefy the grains so they can mould around objects better. I'm thinking of Veritasium's latest video on burrowing soft robots.
Dang, beat me to it! I was thinking the same; coffee grounds or course sand or something, with an agitator to get them to really flow around the part being picked up.
Maybe an in and outlet inside the gripper, so you could flow air through, but on griping use both for suction?
the orange, is this some kind of inside joke? new to this channel :-D
He didn't launch a tomato with it
Why'd you 3d print a funnel? Is this a moment of a man with a hammer see's everything as a nail? Wonderful video and pretty neat.
Hmm, I see a lot of advantages and little disadvantages. While industrial manufacturing should be a lot more efficient, I believe distribution for a single funnel wouldn't be if he orders it online. He should be able to get it in a store nearby, depending on where you live. But that might involve moving a car using fuel.
Also the cost probably is a lot less because of the markup on single items.
And you can get the exact size you want.
Some disadvantages would be that it is probably not as sturdy and you shouldn't use it for food contact.
@@neur303 also take into account energy - the delivery of the 3D fuel is a trip anyway (and probably heavier given offcuts etc)and the energy cost of running the printer will far exceed the cost of an injection moulded funnel. I assume it’s a symptom of needing to finish before a deadline, however it was a weird detail to point out.
@@InvalidEntry That's what I meant to include when saying to be efficient. But I believe you would lose that efficiency very fast once a truck or car is involved for a light item like a funnel.
Funnels are suprisingly hard to get nowadays, it's probably mildly more convenient to print one
Lets say he didn't have a funnel. He could go out unnecessarily during covid, for 45 minutes to find one, or he could print one at home in 30 minutes with the fast printers he has and the extremely simply vase print a funnel is.
In this case, the hammer was the reasonable choice.
So this is how power puff girls hands worked.😁
That's pretty
C
U
R
S
E
D
Ponys too
Somewhere I saw one that used rice instead of bbs and that worked well. It would increase the surface contact with the object you're trying to pick up.
Just as a FYI, in germany we have mattraces and tubes that work by the same principle and are used in medical emergencys. If you need to fixate something quick, you just put the limb inside the corresponding "bag/tube" pull out the air and the limb is immobillized. Or when you have someone with a back injury, you put em on a air mattrace type of thing, fold it gently around the person and pull out the air.
Naomi Wu could probably put you in touch with a manufacturer of larger silicone components. 😉 From her video on the subject they're specified in cc's and available in many incremental sizes. The manufacturer doesn't necessarily describe them as squeezy toys, but the name still fits.
You could probably improve this significantly without going to a release valve:
1. Use lighter plastic BB's
2. Enlarge the hole. Mount the bag using two flanges that go inside the hole to hold the balloon in place. This would allow it to envelop an object better.
3. Tube/flanges in the center so it is more symmetrical and can envelop objects better
4. Improve the code so that it fully expands the balloon to expel an object, then retracts slightly to get rid of excess pressure.
You would need to play around with number of pellets and pressures in order to optimize grip. Basically the main problem IMO is the way you had four contact points to pull the surface area tight like a drum. You want maximum flop.
I suggest using non-round objects with high friction coefficients. This is why coffee worked in their case. In my experiments, I used laser-cut wooden cubes, which were quite lightweight but because of shape and high friction held objects quite well. Also, layer-jamming in which paper is used is also pretty cool.
The really original was done by researchers at Cornell university. They tested different materials for their gripper, thickness of the membrane, filling and so on.
Using a latex balloon and grounded coffee was the best they figured out.
At my university I tried that combination also, but using normal balloons, not latex. It works very well, but with time the coffee damages the membrane... Probably chemistry doing its thing. It is quite some work to tidy up all that spilled coffee from all over the work bench.
how do you manage so many project and work so fast!!!
maybe he is not the only working and editing the videos ?)
@@juanchirino7135 true that would take a long time out of his projects
He must have managed to clone himself somehow.
Heck, just print time alone would be massive.
@@jacobhargiss3839 the good thing with lulzbots is you don't have to watch them all the time and you can work while they are printing.
this video is gripping!!
It's got a hold on me.
James, you're kicking goals every video. I love it.
Wow, great project 👍
I like your approach to the project, "what can you get hold of".
Thanks for sharing your experience with all of us 👍😀
For casting a custom membrane you could use a resin printer to make a mold for some polymer.
Would smaller bb's help it contour around things better?
Yeah, I feel like filling with something smaller would really help with the gripper "resolution". Ideally, something light weight too. Metal BBs would be pretty heavy, maybe something more like beanbag pellets.
@@Timestamp_Guy honestly probably something even like rice could work pretty well. Obviously best case scenario would be that weird ball filling they put in beanbags and stuffed animals though
@@Timestamp_Guy I wonder if rice might work.
@@leogallis8959 Rice would probably be an improvement. Rice tends to be really asymmetric, rather than approximately spherical. Though, I don't actually know if that would hurt or improve gripper performance.
@@Timestamp_Guy yeah plus since he probably already has some in the pantry, a quick test with it would definetly let you know if higher resolution would be helpful
Jamming phase transition, like Vac Man, Stretch Armstrong’s arch nemesis! Thank you, Action Lab.
Oh hey I remember bringing this up some time ago for the exo project, cool that you're trying it out!
I just really like your voice!
Try smaller BBs - even a powder. This will increase the 3d "resolution" of the gripper.
I built one as a kid - used a household vacuum cleaner, coffee filter, rubber band, balloon, and I think ground coffee and it worked like a charm on pretty much any item
Oddly enough, granular beanbag grippers like this are somewhat famous within the My Little Pony fandom as a popular theory to explain how they can pick things up without fingers! Awesome creation btw, very interesting to see it in action :)
Use smaller bids, i think is too little contact area between the membrane and the object with the way you use it, also maybe will be better if the membrane is more full, a good idea to replace the bids is using rice, if is enough you can also try crushing it.
hope this helps, i like your projects very much, sorry if my english is no clear enough
If you bid too small you will be outbid so I don't think that will work
Also your English is perfect, just one misspelling
I wonder how this would react to compressible beads (like styrofoam). Could the grabber shrink around the object to grab it harder?
I think that wouldn't work; when the air is sucked out, the compressible beads would compressed away from the object. Maybe a compressible membrane would be better; it would get squished between the beads and the part; it would be pulled thinner when the air is removed, but if it's thick enough, it may still be squished against the part enough to generate grip.
@@SpencerPaire I thought about it as an analogue to a metal expanding and shrinking due to temperature. A common way to fit metal parts is to heat the part with a hole so that the hole (unintuitively) expands, allowing for another part to fit inside. Then, when the metal cools, it shrinks around the part for a tight fit.
My favorite video of your catalog! Probably because this would just be within my “grasp” if I wanted to make it:)
A great attempt, James! Certainly a good stepping off point for anyone who wants to develop an open-source version of the original gripper.
I'll probably be doing a V2 at some point
@@jamesbruton 👍🏻
I wonder how small you could make this with like sand could be really good for a pick and place robot
Nice. This is the kind of trying out of methods and tech I like.
That's a cool concept! You could incorporate a check valve in it that exhausts air to let it squish easier onto objects
I think I found other possible use for jamming phase transition stuff...
A driver's seat for a race car. When combined with a six-point harness (with inflatable straps) and a quick-release mechanism (allows all of the air back into the material in an emergency), the safety levels should go through the roof.
I've used those grippers before they use a sand like material inside and use a slight positive pressure when forming around the object. They actually grip quite well.
I second the idea of smaller bbs or powder. Great project!
Great low-cost implementation of the idea!
Sensing and varying the volume sounds like a good idea. Also probably just a second syringe might work better than a bigger one. You surely still can drive them from one servo or just use a second one.
To solve the issue with the positive pressure in the "open" state you could integrate a valve into the plunger. drill a small hole in the bottom of the cylinder and add a spring loaded valve that is opened by the plunger when it is at the bottom of its stroke, but as soon as the plunger lifts the valve closes and it allows the plunger to draw a vacuum. Once the plunger returns to the bottom it pushes on a pin attached to the valve and opens it.
This is really cool James, thank you for sharing it!!
I remember that article! I made a little one with a baloon, a syringe, and coffee grounds as a kid. It looks like you're stretching the ball a bit in order to secure it to the holding. I don't really see a way around that, but that probably makes the gripper shrink into the back of the grabber more, and less around the object, making it harder to hold. Swapping out the BB pellets for something more compressable would probably help too. Not too compressable, just compressable enough that there's something springy pushing against the object from all sides. I think the article specifically recommended coffee grounds for this reason.
I love these robotics maker bucket list experiments. Keep up the good work. Next one, yoga ball!
I had no idea this was even possible! Awesome video as always!
So cool! I'd have never thought of this!
This test setup works great at showing off the principle. A few suggestions on improving functionality:
1. Smaller grain size
2. Overfill the membrane with grains
3. Slightly stiffer membrane material
As well, industrial robotics use venturi effect vacuum nozzles that only require compressed air to function. They're not too expensive if you look around (or maybe even design and 3D print one?)
Awesome video, like always! Thanks for sharing your knowledge 😁
I remember seeing similar projects on Hackaday a long time ago (I just checked, they were from 2010).
They seem to use coffee grounds for the gripper as is also in some of the other comments here.
Inventers see coffee beans in a vacuum pack, Inventers be like "YES".
My thoughts on seeing this are, needs more BBS and smaller ones like ball bearings or the glass beads from sandblasting. And two instead of using air try using a fluid.
Your have an amazing mind. Thank you for the work you put into this channel.
I really recommend replacing bbs with ground coffe. Preferably French press ground. Improves the weight of the system and makes shaping around objects better! Good luck. Great job
I tried making one of these with coffee grounds a little while back. The problem with granular jamming is that, unlike regular robot grippers, there's no positive gripping force at all. Once you've sucked the air out, you're reliant on the new geometry (and friction) to lock the part in place. If the granules get disturbed in any meaningful way, the grip is broken, unlike with a conventional gripper that can use force feedback to further close the fingers if the object moves.
Cool gripper design!
If you tried sealing the 3d printed funnel or printed in resin. You could then clamp a circle cutout of a silicone sheet or those large thick balloons over the opening. If you can get a reliable seal, it would open up a lot of options for finding the best material and thickness to get the best grip.
I read the paper on the concept when the gripper came out. I believe it stated that a finer medium and irregular shaped medium was critical to grip. Ground coffee was mentioned as great starting point. It is also lighter than metal BBs.
Try a Ballon filled with flour mixed with rice. This makes for a good stress relive ball, which could work great for this project.
Fantastic project.
Makes a awesome and smart griper, procedes to confuse a orange and a tangerine.
Congrats...you have reinvented the Dalek’s grippy thing. ;-)
Hi I am 12 years old and I’m not expert but I think you should use sand or smaller bbs because there would be less gaps in between and the smaller would have more details which might help grip. Also you are a huge inspiration.
I agree that the partical size has a definite impact. Airport bb would work better, but the shape is also a factor. The one shown at the start has much more exposed material allowing for the gripping section to flow around the object and get a better hold your version does not allow for enough expansion since most of your gripper is encased in your print and can not form around the picked up object.
I built one of those like 6 or 7+ years ago. My recommendation from my experience would be smaller grain size. So long as you have a filter that is good enough for the grain size you choose the smaller the size the better it can conform to the shape(within reason there still has to be sufficient airflow around the particles to allow a decent vacuum. Also a cheaper alternative to the squishy toy is to just use a rubber balloon that you can pick up anywhere for a few cents. The syringe with a servo is a good idea though. Personally in an industrial setting at least I would just use a vacuum transducer (can be easily 3d printed so that might be an idea for a future project) and compressed air to generate the vacuum or a vacuum pump.
i truly admire your knowledge
I was quite scared that you would inhale a bb. A peice of fine mesh at the end would work perfectly. Amazing idea and amazing outcome. I can’t wait till yku use this in a future project
A few years ago popular mechanics made this with a balloon and ground coffee. I was able to make it using the same materials and a small vacuum pump
You simply need a smoother gripping surface which means smaller beads in the gripper. polystyrene can work but you have to pack it full or pull out a lot of air since they compress easily otherwise a more solid option could be bead/foam putty but one thats not too sticky...
Excellent! It looks lovely in those colors, just like a toy. Maybe a deeper cup would allow to encompass objects more wholly ?
Three way valve will help. One position to let air out of the squashy bit only, leaving air in syringe, 2nd position to pass through squashy to syringe and 3rd to allow air into syringe. That way you can do all the things you need.
3D printing a funnel is such a 1st world solution to a simple problem
That is a very cool concept! The armchair expert in me are thinking of variables that might make it better;
• The membrane having a soft texture and being stickier. Just some small textures that would increase surface area, like "hairs" or something as simple as a diamond pattern or anything in that ballpark. (Maybe glue a couple of gecko feet on it? /s)
• Much smaller BBs to have the membrane conform better to objects, especially small objects with contours. Something like
The MIT project used coffee grounds in a balloon, which worked well given their stiff yet light structure and good packing ratio.
Instead bb's you can just use any spherical stock to reduce the weight. Also (as mentioned), you need smaller filler-balls like poppy seed, or might even use sand. You can combine poppy seed with larger polystyrene packaging balls.
I just have came across the video of @Practical Engineering about Suez Canal, where he explains, that fillers containing non-spherical pieces (like sand) grabs better under pressure, because the pieces will likely to rotate causing the whole volume to expand.
This also works with latex balloons, much easier to find in any size you want. Look for "punching balloons", they're a thick latex that's really strong. Also, consider using smaller bbs, maybe something like rice would work, the uneven shape should add more texture for gripping.
I think the issue is the retaining ring holding the membrane taut.
It's preventing the vacuum from closing the means around the lower end of the object and the only thing holding parts in place is friction against the sides.
If you allow the membrane to close around the part is say your gripper will work better.
That has to be the slime sound effect when he squeezes it, it's to perfect
I think the beads used for making jewelry the plastic, wood, or glass would do great for this. They come in all sizes
Couple ideas. 1. Smaller beads 2. Some sort of filter to keep The beads out of your line. 3. A softer bead, like foam that each bead molds to the small contours. 4. Thinner membrane. Obviously the issue with that is thinner generally means weaker
It's a very interesting idea!! Some future ideas for optimization would be:
more pressure/suction
bigger membrane
semispherical membrane instead of spherical (cut down on weight)
some texture on the membrane for grip
smaller bbs or even small ball bearing or birds shot. Or maybe even a combination of granular material sizes to balance out weight/ air volume/ and grip
Very Interesting concept! Quite curious to see another video of this maybe using smaller pallets. (For example: raw plastic pallets used for injection moulding)
Another suggestion would be to initialise the syringe at 25%. This will reduce the suction capability but it would also allow the objects to pop out better :)
This dude must be a millionaire after getting like 5 sponsors in a single video
For weight concerns a replacement of almost similar grain size would be airsoft pellets but i would recommend a finer material, sand or dry packet silica with a cloth filter as a barrier works nicely flour and a balloon with a a few cloth /silk/ fine air pernissive membrane is optimal if thats to fine oats or granola works well and maintains low weight parameters
Havent tried Styrofoam yet that could be need due to potential squash factors
Great project, seems like It would greatly benefit from a larger surface area and finer detail; I think it can be done with smaller bb's and a smaller housing of the ball without any side stretching.
It's like a stress ball, except instead of foam or starch it's, something called orbeez here in America. I'm sure, this will be interesting for a project later. Like the ball moving machine.
I think you need a more convex surface for the gripper, having a hemisphere seemed to work much better than a stretched-out diagram. It also prevents the membrane from wrapping around the object as it can only ever grab the top. Less bb's as well might work better.
use a solenoid to equalise the pressure before each stroke... to make it softer when picking up items and let more air out before pulling the syringe and also the opposite, before releasing, open the solenoid to let air back in before pushing the syringe to inflate. It essentially doubles the useable range of the syringe