Robot end-of-arm tooling (EOAT) is a hotbed for development right now. It’s also one of the most critical parts of any robotic application. Industry experts cannot overstate the importance of end effectors, the direct interface between the robot and the part that it’s handling or processing.
The variety and intricacy of objects robots are being asked to manipulate and the need for more user-friendly solutions are driving radical change, giving birth to new technologies and suppliers. Plus increased demand for higher speeds and reduced energy consumption are putting pressure on leading EOAT manufacturers to raise their game.
The 5-finger robotic hand is a glimpse into our future. A feat of engineering, this anthropomorphic end effector boasts 9 servo-electric motors controlling 20 axes, or joints.
“It mimics a human hand in not only size and proportion, but in all the motions and dexterity of the human hand,” says RJ Ruberti, Team Leader of Linear Systems at SCHUNK in Morrisville, North Carolina.
This video courtesy of SCHUNK shows the high-fiver in action. With the help of a 6 degrees-of-freedom arm (DOF), the hand gestures and points, even counts to three.
While undeniably cool, the 5-finger manipulator is still a long way from the factory floor according to Ruberti. It’s largely geared towards the fledgling service roboticsindustry and R&D efforts in that arena.
“Most of these kinds of robotic hands go to universities,” says Ruberti. “They’re not going to factories to be sold for automation applications. If it ever got to the point where these were mass-produced and more affordable, it’s possible that it could be in an automation application. That’s why we develop these grippers. If we’re going to be the world’s largest gripper company, we definitely want to stay ahead of the game.”
“We also have a 3-finger hand with tactile sensor pads on each finger,” says Ruberti. “So when you grip something, you can actually see where the fingers are gripping the part.” Sporting a more robust design, the 3-finger hand is primed for industrial applications as well as service robotics.
You may recall seeing a SCHUNK 3-finger hand coupled with a Universal Robots UR5 arm triggering the closing bell on the NASDAQ stage last November. This historic event marked the launch of the first globally traded stock index fund for robotics.
“It’s easy to pick out a gripper to handle only one part,” says Ruberti. “What’s difficult is when you have a lot of different parts, or parts that may be similar but all different sizes. What’s difficult is figuring out how many different end effectors you need to handle this range of parts.”
“We have a lot of solutions,” says Ruberti. “We have ways that a robot can quickly switch out tools. We have quick-change gripper fingers that allow you to go from one set of fingers to another in a matter of seconds. It acknowledges that you need specialized solutions, but we’re trying to make it as easy as possible.”
While leading EOAT manufacturers are striving to stay ahead of the game, others are changing the game altogether.
A start-up in 2012, Empire Robotics Inc., lit up the high-tech news channels with an end effector innovation based on jamming transition technology. What evolved in Cornell University’s Creative Machines Lab is now a full-blown product kit released in January and being marketed to the industrial automation sector.
The VERSABALL® busts through the traditional end effector categories, namely mechanical, vacuum and magnetic, to create its own archetype. Its co-inventor and company cofounder describes how the unique end effector came to fruition.
“When I arrived at graduate school (Cornell), two simultaneous things were occurring and I found myself working in the place where they intersected,” says John Amend, Ph.D., Chief Technology Officer at Boston-based Empire Robotics. “There was lots of new interest in jamming transition technology in the physics community, and on the robotics side there was an increased push towards soft robotics in research circles.”
He explains how the Defense Advanced Research Projects Agency (DARPA) was sponsoring several different research programs, all with the aim of developing robots that are soft and flexible.
“My professor at the time, Hod Lipson, was collaborating with Heinrich Jaeger, who is a physics professor at the University of Chicago,” says Amend. “They were using granular materials to do soft robotics development for the DARPA Programmable Matter project. I started using granular materials as a way to stiffen soft robots. It wasn’t too long before we had our eureka idea for this gripper.”
Amend says later they discovered that other researchers have had similar ideas in the past. The advantage was that his team had experts in both granular materials and robotics. This allowed them to delve deeper into the research and get better results. The initial patent filed at Cornell has since spawned additional patents at Empire Robotics. He explains how the unique end effector combines vacuum and jamming transition technologies for a universal gripping effect.
“The way it works is that you have this granular material inside a balloon membrane. You blow in air to fluidize it and make it soft, and then you push it onto an object, where it conforms passively to take that object’s shape.”
Amend then describes how vacuum is used to contract the granular material (think vacuum-sealed coffee grounds), which quickly hardens to grasp the object firmly. Subsequently, positive pressure is used again to reverse the transition, effectively returning the balloon to a deformable state and releasing the object.
“The jamming transition is essentially a glass transition,” explains Amend. “It’s a transition from a liquid-like state to a disordered solid state. The interesting thing about that from the perspective of making a gripper like this is that you need a very small volume change and no change in temperature to get a really large change in the hardness of the bulk material. So when it’s soft, our gripper is just one percent larger than when it’s hard.”
The lime-green ball has been used to pick and place everything from plastic bottles and cups, to smartphones and broken pieces of glass, even a raw egg.
This video courtesy of Empire Robotics shows the VERSABALL kitting toy parts.
Designed for Industrial Use
Back in the lab at Cornell, the research team was building prototypes with common party balloons filled with coffee grounds. The current design has come a long way. Now Empire has a polymer scientist on staff who works full-time improving the performance of the balloon membrane materials for industrial use.
“We’ll work up a formula in the lab, test it, and then send it out for manufacturing in higher quantity,” says Amend. “Basically, we’re blending your elastomeric materials, anything that you might find in a surgical glove. Durability is the main issue.”
Empire is also experimenting with different balloon textures and may introduce specialty balloons for different applications. The granular materials have also evolved.
“We generally want to use smaller granular materials, typically sub-millimeter size,” says Amend. “It’s a trade-off though. You want them to be very small because it helps conform to the shape of the object. But if they get too small, you have air permeability issues, so it’s harder to draw a vacuum through the material. Smaller grains also have a tendency to clog our filters, so it’s always a balancing act.”
Amend says the amount of energy consumption is on par with suction cups. To save energy, the gripper only uses air at two points, to create the initial grip and to release the object. During transfer of the object, the gripper is sealed so it doesn’t use any air.
He says their unique end effector is primarily addressing applications for which there is currently no solution.
“It takes a lot of engineering effort to deal with end-of-arm tooling, especially for robot integrators trying to design industrial automated manufacturing lines. People will say things to us like, ‘I know your VERSABALL is not the optimal gripper for these objects, but if it will do the job and it will do it well enough, then I can have my engineers work on other things.’ They don’t have to spend time and money trying to design or track down grippers that are optimal.”
The end effector does have its limitations, according to Amend. “If you try to pick up an object from the side, the object is usually not heavy enough to resist that force, and so the object will slide and you won’t get anywhere. You have to pin it against something.”
He says conveyance shouldn’t be an issue. “If you match the speed of the conveyor to the robot arm, then the gripper and the object aren’t moving relative to each other. Whatever object is coming down the line, if you try to grip it at center, you will do pretty well.”
The end effector has mostly been tested on six-axis articulated robots, but Amend says it will work with all different kinds of robots. He does note, however, that because it can take more than a tenth of a second to draw the vacuum, they can’t quite take full advantage of delta and SCARA robot speeds, yet.
Amend says they get a lot of requests to create jaws that combine two or more VERSABALL units, or to flip them upside down and use them to fixture objects. The commercially available kit currently contains a 3.5-inch ball and a 6.5-inch ball. They are considering other sizes and investigating specific applications, such as workholding, food, prosthetics, and even underwater applications. All of these are in various stages of development in the lab.
Jamming transition is not the only new material handling technology vying to be part of the conversation. Electroadhesion technology is adding to the recent excitement.
This video shows electroadhesion technology being used for grippers, fixtures and conveyors.
Traditional vacuum gripper manufacturers are also fighting to stay ahead of the game. Faster speeds, higher performance, and lower energy consumption are the big trends.
“We see more products being packaged in stand-up pouches and poly bags,” says Ed McGovern, Vice President of Sales & Business Development for North America at Piab USA Inc. in Hingham, Massachusetts. “Everything from tuna fish to pet food, even cereal you’re starting to see in these stand-up packages. The challenge from a robotic point of view is how to grab those things. They’re not very kind to traditional grippers or suction cups, because the bags crease and the surface area is different every time you pick it up.”
McGovern says there used to be a trade-off between the grip and the machine handling speed for these types of applications. You either had to slow the speed down or oversize your vacuum pump to compensate for leakage. The piGRIP® suction cup, shown gripping a chip bag in the photo, solved that problem.
“It was difficult to get a seal, so we developed this bag cup which has an extremely soft, pliable lip with very firm bellows,” explains McGovern. “The lip is made of liquid silicone to create a real good seal between the stand-up pouch, or bag, and the suction cup.”
This video courtesy of Piab shows the technology in action with a dual-arm collaborative robot picking and placing relatively heavy bags filled with liquid.
“If you picked up that bag with any standard suction cup, it wouldn’t work well,” says McGovern. “The bag would crease, then there would be leakage, then it would drop. That bag lip is so good that they can pick it up with one suction cup.”
Suction Cup with Fingers
How about a suction cup with a set of compliant fingers?! Check out the second part of the previous video.
“The advantage compared to regular clamping end effectors is that it’s gentler on the product,” says McGovern. “The piGRAB gripper is an inexpensive compliant gripper compared to a much more complex solution.”
McGovern says the compliant vacuum gripper was developed in collaboration with one of the foremost experts in vision guided robotics and bin picking, the idea being that it could someday be used in 3D random bin picking applications.
Piab’s vacuum manipulator is not the only compliant gripper handling fragile cargo. This video shows a dual-arm robot equipped with adaptive grippers on a mission to crack the case.
The Critical Interface
What many robotics users don’t realize, stresses McGovern, is the importance suction cups play in the overall success of the robotic work cell. The interface or suction cups are a critical part of the tooling.
“It’s amazing to me how little time engineers will spend on making sure they have the right suction cup to handle their part,” says McGovern. “A suction cup is really the main interface between the robot, the end-of-arm tooling and the product you’re handling. You can buy an $80,000 robot and if you have the wrong cup on there, it’s going to slow things down. And it doesn’t matter what size your robot is or how good it is, the wrong cup will kill performance.”
He says the selection of the suction cups needs to occur before the end-of-arm tooling is designed. That’s when the critical questions need to be asked.
“What type of material am I handling? Is it porous, is it nonporous? Is it textured, is it curved, are there oils? What’s the nature of the product, how quickly do I want to move, and what are the major issues with the product actually being handled? What types of technology in suction cups, for example friction cups or bag cups, are available to be able to provide the best seal?”
“If you have a poor suction cup design, it will significantly impact your robotic performance,” adds McGovern. “It’s a big issue and it’s an issue that can be easily addressed, and it can be addressed relatively inexpensively.”
Lowering the energy consumption of vacuum grippers is a big trend. Piab developed the first multistage ejector vacuum generator.
“Piab is different from other vacuum manufacturers in that we have this patented multistage technology called COAX®,” says McGovern. “Unlike a single-stage venturi with one nozzle, we typically have several nozzles to extract more performance from the compressed air. We take the exhaust air that would come out of a single-stage venturi and we use that again in the second nozzle. Then we use that exhaust again in the third nozzle. So by the time the air is exhausted, it’s at a very low pressure and you’ve generated a lot of vacuum flow, which is equivalent to speed.”
“This is critical in end-of-arm tooling,” adds McGovern. “The result is fast vacuum with low energy consumption.”
Another energy saver is a decentralized vacuum system. To that end, Piab is working to reduce the weight of generator components so they can be moved closer to the end-of-arm tool.
“There’s a big energy benefit when you can decentralize a vacuum system,” says McGovern. “Think of a long straw that you have to suck air through. The longer the straw, the more difficult it is to evacuate that suction cup on the other end. That’s why we suggest considering a decentralized vacuum system to get the vacuum close to the point of suction.”
“With our piCOMPACT® vacuum ejectorwe can separate the controls from the generator. We can put the actual vacuum engine on the end-of-arm tooling, but all the controls including the valves, sensors and switches can be separated and mounted on the robot so that the weight isn’t added to the end effector. That’s a unique design to Piab.”
Standard vs. Custom End-of-Arm Tooling
In the world of robot end effectors – practically an industry onto itself – there’s an ongoing push-pull between standard and custom end-of-arm tooling.
“There’s a general trend towards a universal end effector, one tool to pick up multiple products, which leans more towards standardization,” says McGovern. “But overall the trend is more towards specialization, only because a lot of robot companies have driven the design of the end effector out to their integrator network. The integrators to a large extent are designing the end effectors.”
Weldon Solutions, a robot integrator in York, Pennsylvania, can attest to that. “The end effector manufacturers are generally providing all the different components for the end-of-arm tooling, but they’re not pulling it altogether,” says Charles Gales, Manager of Automation Sales. “For the vast majority of projects that we do, we use our engineering staff to design and develop our own custom end-of-arm tooling utilizing standard components wherever possible.”
Weldon in fact designed and built a custom end-of-arm tool using a unique magnetic end effector for a boring mill machine tending application.
“The large castings we were handling had an unusual geometry, so we had to work very closely with the magnet supplier,” says Gales. “It’s a unique tool with a permanent magnet in it, but it’s not energized until you rotate the poles. That allows us to create a magnetic force only when we wanted it, and do it pneumatically versus using an electromagnet. In the past, we’ve used different means to strip off the part. But this rotary device that misaligns the poles it’s definitely a better means of disengaging the part.”
“Then we also recognized the need to pick up a vacuum array, so that we could move the plywood sheets that separate the layers of parts,” he explains. “I think it was a pretty good use of the magnet to be able to pick up that plane that held the vacuum apparatus. That idea came from our engineering staff.”
He notes that the vacuum EOAT was mounted out of the way along the side of the work cell, a good use of floor space.
“We could have had some retractable vacuum cups that jack-knife along the robot arm, but in this case we didn’t have the space,” says Gales. “Or we could have used a tool changer, but this method provided better cycle time.” He notes that it’s important in the design phase to take into account the space that will be required for all the pneumatic fittings, electrical couplings and dress-out.
Gales says a retractable coil hose supplied compressed air to the generators on the vacuum gripper. The hose was fed from overhead so it did not have to be run along the robot arm.
This video courtesy of FANUC America Corporation shows the end-of-arm tooling playing double-duty on several fronts.
Magnetic Grippers Reimagined
The two magnetic grippers, one to unload the boring mill and the other to load it, were supplied by Magswitch Technology. This company’s patented switchable magnetic technology first came across our radar last summer when we profiled it in this EOAT trends article.
Magswitch AR-Series magnetic grippers are designed to handle irregular shaped parts, such as those in Weldon’s machine tending application.
“The two grippers in this application are both AR110 HDC and they are capable of a maximum breakaway of 3,300 lbf with appropriate pole shoes,” says Michael Blanchard, Engineering Manager at Magswitch Technology Inc. in Westminster, Colorado. “The pole shoes and armor on these units were also designed by Magswitch.”
“We have expanded our application assistance program to include electromagnetic field simulation software,” says Blanchard. “Using this technology we can accurately predict holding forces on custom part geometries and materials given the 3D CAD file and appropriate material properties.”
“What is shown in the video was very difficult prior to Magswitch,” Blanchard explains. “The parts do not have a high surface area or material mass near the magnet face, which would prevent smaller magnets and suction cups from gripping the part. We used a larger magnet with a deep field to grip the part and hold in shear successfully.”
Evident in this machine tending application is the pendulum swinging between standard and custom end-of-arm tooling.
“As there is a wider proliferation of robotics, what happens is that a lot of the different styles and types of end-of-arm tools have already been done to some extent over the last 10 or 15 years,” says Weldon’s Gales. “There is some reuse of concepts. But the other side of that wider proliferation coin is that the robots are starting to get into new applications where they weren’t before, so there are new things being developed.”
Whether they’re existing designs, custom creations or disruptive technologies, robot end effectors and end-of-arm tooling will continue to play a critical role. Not only in the success of specific applications, but also in the proliferation of robotics in our work and home lives.