Universal Gripper - Syringe Powered

40,963

166

52

Published

Introduction: Universal Gripper - Syringe Powered

About: I am a theatre designer and technical director. I work at Merely Players, a small non-profit theatre group in Southwest Colorado. Check us out on Facebook: Merely Players-Durango. You can also see some o...


The “universal gripper” developed by researchers from Cornell University, the University of Chicago, and iRobot inspired me to create my own version. The YouTube video is quite impressive. The gripper can form around very asymmetrical and smooth shapes and still pick up the object.

I gathered the materials below and decided to create a gripper of my own that could be incorporated into one of my robot designs. Adding a vacuum pump or a powerful air mover seemed difficult and probably unnecessary. So I started to think about vacuums.

I remembered that when they were filming “An American Werewolf in London” that Rick Baker used hypodermic syringes hooked up in a line. When a really big syringe was compressed at the end of the line, the little ones popped up creating the appearance of a spinal column erupting below the skin of the creature’s back.

I decided that the same idea might work with the “universal gripper.”


In addition, I thought that using a funnel could work nicely for balloon support. Playing around with it, I thought that adding a pressure sensor inside of the funnel would allow the Arduino to apply suction when the balloon came into contact with the object to pick up. This lead me to Plusea and her instructables using velostat and flexible materials.

Amazingly, it all seem to work. Follow these steps to see how I did it. You can probably see alterations that would improve the design.

Step 1: Materials

Materials:
  • Good quality latex balloon
  • Aquarium plastic tubing
  • 60 ml veterinarian syringe
  • 11/64” brass tubing
  • zip ties
  • Bond 527 cement
  • Conductive thread
  • Needle
  • Craft foam
  • some #6 bolts and nuts
  • 5 cm of U channel aluminum
  • TowerPro servo MG995
  • flat stock aluminum
  • an Arduino or other microprocessor

Step 2: Prepare the Balloon

I was worried that the latex balloon might have small pin-holes that would affect the rig’s ability to create a good vacuum. I decided to make a test. I inflated the two latex balloons I had purchased and put a clothespin on the bottom. The next morning I found that one balloon had deflated by half. The second balloon was still fully inflated, so I used it.

I filled the balloon with coffee. I chose the cheapest and coarsest coffee in the store. I think that a coarser grind would provide more jagged edges for the coffee granules to lock onto each other when the vacuum is applied. I didn’t test it, so it remains s theory.

I decided to only fill the balloon about ¾ full of coffee. When the vacuum is removed the coffee needs some room to become a flowing powder again.

Next I used a craft pipe cutter to cut off about 5 centimeters of the brass pipe. You can get pipe cutters from hardware and craft stores. You could also use a Dremel tool or even a hacksaw.

The aquarium tubing was pushed onto the brass tubing, and then the tubing was inserted into the balloon neck and locked on with zip ties.

Step 3: Test One

The first test used lung power. I found that I didn’t have to inhale all that much to create a good vacuum. The coffee became rigid, and I was successful in picking up a variety of objects.

However, I found that the coffee grounds also came up through the tube. This did not taste very good!

Step 4: Test Two

I attached the aquarium air hose to the 60 ml syringe. I added a zip tie to keep it pretty airtight.

Watch the video to see how this rig successfully picked up a variety of objects. I found that you did need to put a fair amount of pressure on the balloon so that it would form around the object and not just sit on top of it.



Step 5: Test Three

The next step automates it.

I added a funnel to give the balloon and tubing some structure. I took the small funnel I had and used a hacksaw to cut it off shorter. That way the balloon could hang freely, but have support when it was pressed against the object to be picked up.

I decided to put a fabric pressure sensor in the funnel. That way the Arduino controlling the servo could sense when enough pressure was applied to the balloon. Plusea has a great Instructable describing how to create pressure and bend sensors. These sensors can be very sensitive and also very accurate in reading pressure or bending.


I created mine out of craft foam from a local craft store. It is flexible, but has enough structure that it doesn’t collapse. You need some material that can be compressed, but won’t stay that way after the pressure is removed.

Instead of sewing the edges together, I used Bond 527 multi-purpose cement. It is flexible and also adheres well to the foam. The conductive thread was sewn in in a criss-cross pattern as Plusea describes. The Velostat middle layer was cut smaller than the outer pieces so that the glue did not adhere to it and cause compression. Velostat is available from The EMF Safety Site.

Conductive thread was left sticking out about an inch on both sides to allow connection to a wire running to the Arduino.



Step 6: Put It All Together

Build the Syringe Mechanism:

I tested the syringe and found that I only needed to pull the plunger back to 30ml in order to get a good vacuum. So I used a hacksaw to cut off the barrel of the syringe.

I also cut the plunger off about 3cm longer than the barrel.

I drilled a hold in the servo arm and also through the plunger. A #6 bolt with washers and nuts holds the plunger pretty stable as the servo pulls it back and pushes it forward.

I had some aluminum U channel and cut a piece about 20 cm long. I don’t know the technical name, but I had a plastic piece used to mount servos that I got in a box of servos from eBay. I attached the servo to the plastic mount and then bolted the mount to the U channel aluminum.



A piece of flat bar aluminum made the syringe support. The support has to rotate when the plunger is moved, so I put one #6 bolt to hold the barrel, and another bolt into the U channel. The U channel bolt was not tightened so that the whole assembly can rotate.

Wires from the pressure sensor go to the Arduino. I used the same setup as for a photoresistor on the Arduino site.

Some code reads the pressure sensor, and when enough pressure is applied the servo is activated. The Arduino waits 2 seconds and then pushes the plunger back into the barrel removing the vacuum.

I don’t have a robot arm (yet), but this could be really fun to put on one. Its strength is that it can pick up very irregular or smooth objects easily. Using the syringe, there is no need for vacuum pumps or vacuum reservoirs. The system seems pretty robust and reliable.

Let me know if you experiment and come up with improvements!

Step 7: Programming


The attached file is my Arduino code.

Share

    Recommendations

    • Planter Challenge

      Planter Challenge
    • Casting Contest

      Casting Contest
    • Oil Contest

      Oil Contest
    user

    We have a be nice policy.
    Please be positive and constructive.

    Tips

    Questions

    52 Comments

    a very nice instructable sir, indeed it is a very easy DIY version of cornell university gripper. also i like the idea of using the pressure sensor to deter the pressure created to pic up object, are there any ways other than the depicted one to make a cheap pressure sensor???

    1 reply

    Yikes - I have been traveling for a while and didn't see your question. I have had good luck with Plusea's pressure sensors. She has a couple of different designs. Here is one:
    https://www.instructables.com/id/Conductive-Thread-Pressure-Sensor/

    The nice thing is that you can have a foam or fabric structure, so it could work with the gripper's surface and flex.

    Brilliant!!! I have been wanting to mock up something like this but didn't really want to fool around with pumps etc...as you state power consumption is always an issue!!!

    Hi. I had a doubt. I am trying to make a similar prototype, but my idea is to automate the whole process by constructing the robotic arm. So I was wondering if there is a way in which we could confirm if the gripper has actually held on to the object or not? This is because as soon as the balloon presses either on the surface or around the object, it is bound to diverge and so a signal would be sent, so what if the gripper didnt make contact at all with the object? or maybe the gripper did make a contact but while lifting the object fell, the how do we check?

    user

    can it pick a glass of water like on youtube ?

    The more jagged edges you have, the more finely the grit is actually able to compact on itself, at least by one theory. Finer grit would be better, I think. The only thing it really has to have is sheer thickening (think cornstarch and water).

    2 replies

    If you read the article, the researchers discuss the optimum material that is fine but not smooth so that the particles grip onto each other. Their conclusion was that coffee grounds make an ideal material, and they are very light. I think your cement idea is interesting, but the material needs to be easily deformed so that when the balloon is pressed onto an object the material inside the balloon can "flow" around the object.

    Portland cement powder actually flows extremely well; when completely dry, it behaves very much like a liquid, actually. It will even splash up if you drop an item into it (it's far cheaper than coffee grounds, too: $5 gives you around 50lbs of it, which will fill 1-2 cubic feet of space). Plus, you can mix it with gravel, sand, and water to make concrete (if you want to be really fireproof, mix powdered fireclay with it).

    I'd probably use something like Portland cement for the fill material, because its much finer than coffee grounds, sand, and most other types of grit materials.

    Very cool, but is it my imagination that the gripper currently lets go as soon as the pressure is off?

    6 replies

    I programmed it to only keep the plunger pulled back for 2 seconds. I didn't test it yet, but if there are no leaks in the balloon or hose, you should be able to hold the object indefinitely (as long as the plunger is pulled back). When I was first getting the servo to work it definitely held the bolt for a minute or two while I was trying to get all of the connections right.

    Cool - so when it's on the arm, you'll need to programme separate "pick up" and "put down" commands.

    Yes, that would be necessary. What I like is that an arm could be programmed to descend until the pressure of the balloon against the funnel is at a certain level. That way the robot would know that it was in contact with the object even if the object was on an irregular surface.

    Or be fancy and do capacitive profiling instead.

    sir,im trying to build a wall climbing robot, i got problem about the suction to be held between the robot and the wall.in my search i found that gecko adhesive is the efficient one,but im using air pushres to held the vacuume pressure ,but im unable to design the body,ofcourse im trying but i need ur help

    That's probably better as a forum topic, then you get more help.

    Hi - nice instructable. I made something similar awhile back, but used a different approach to the suction device; should be perfect for your gripper and will let you get rid of the servo AND the syringe...  a miniature pump

    2 replies

    The pump looks cool. And so cheap! Depending upon how the pump is constructed, it might have to remain on in order to keep a vacuum. What I liked about the syringe is that once the plunger is pulled back, it takes no power to keep the grip on the object. And for most robots, power consumption is a major concern.

    Thanks for the pump source.

    Put a normally-shut solenoid valve between the pump and the gripper.

    This would be great for a quad rotor to pick up items and move them around!!!