To understand current trends in end-of-arm tooling (EOAT), The Robot Report recently spoke with Dan O’Brien, president of Gibson Engineering Co., as well as Jean-Philippe Jobin, chief technology officer at Robotiq Inc. and Sam Bouchard, co-founder and president of Robotiq.
1. End-of-arm tooling proliferates
Although the collaborative robotics market slowed last year because of a dip in automotive manufacturing, most industry analysts expect it to continue to grow. The proliferation of predesigned end-of-arm tooling such as parallel and soft grippers or vacuum cups has accelerated adoption, noted O’Brien.
“Kristian Hulgard, [general manager, Americas, at OnRobot], recently told me of their plans to introduce between 20 and 30 separate new products this year,” he said. “That, along with their quick-change technology, should open up the number and types of applications that can be accomplished, along with speeding the time it takes to go from one application to the next. They really want to move the discussion from collaborative robots to collaborative applications.”
Robotiq is also well-known as an end-of-arm tooling provider, and it isn’t sitting on its laurels, either. Jobin said he expects the Quebec-based company to continue developing tools that help companies complete manufacturing tasks with cobots.
“There’s no reason why Robotiq would stop developing new tools,” he said. “It’s been one good aspect of the business that we’ve been recognized for grippers. Now we don’t want to be recognized just for that — we want to be recognized for everything we can bring around the application.”
“By taking the application first and bundling them together, I think we can simplify things at a certain level,” said Jobin. “The final goal is if we’re able to put things together and have them talk together, we would like the complexity to stay beneath, and only show the customer a ‘lever’ to make it work. The end user doesn’t care about the camera; they don’t care about the robot; they just want to do a task.”
2. End-of-arm tooling becomes vendor-agnostic
Many end-of-arm tooling providers focused on being compatible with only one brand of robot, observed O’Brien. As the early market leader in collaborative robot arms, Universal Robots was an obvious choice.
However, that has changed. “Based on the speed with which new players are entering the cobot market, it’s now standard practice to support multiple robot brands with their new designs,” O’Brien said. “OnRobot currently supports eight brands and plans to continue adding to that as new cobots enter the market.”
“In order to make the production, the end users don’t have enough people in order to do the manual task, so they want to automate,” said Jobin. “But on the other side, it’s even more difficult to find skilled people in order to put in production robots. So we are trying to solve those two aspects at the same time in that phase, creating products that could automate the task, and then on the other side, trying to automate the automation.”
3. Safety focus at the end of the arm
Cobot arms may be rated for hand guiding, power and force limiting, safety-rated monitored stopping, and speed and separation monitoring. But most vendors will remind you that a safety assessment is still necessary to evaluate the end effector, payload, and operating environment.
Improvements in sensors, computer vision, and connectivity are helping cobots be safer, which is especially important for emerging mobile manipulation. With sensing technology such as that from Veo Robotics, “collaborative” could become more of a property of industrial robots than a standalone category.
“After a recent meeting with Johannes Marktl, [head of sales at] Blue Danube Robotics, I was impressed to see that they’d extended their AIRSKIN technology to include the end-of-arm tooling,” said O’Brien. “AIRSKIN mounts like an outer layer to robots and/or tooling, and it senses a collision in real time and sends a signal to the robot safety circuit.”
“This allows for more traditional robots like the Mitsubishi RV series to work safely around people, and it allows the end-of-arm tooling to be part of a safe collaborative application,” he continued. “Blue Danube has also introduced AIRSKIN Module Pads, which is a cost-effective way to quickly and easily add safety to moving machinery, cartesian systems and custom end-of-arm tools.”
More and better sensors, in combination with the right end effector, can also improve productivity.
“We have a team continuing to look at vision, to see where it’s going,” said Jobin. “We believe in force-torque sensing and tactile sensing. We did develop technologies related to tactile sensing. Four years ago, we had some samples and tactile sensors in order to recognize a part, but it was not good enough in order to make a product, so we continue developing that.”
“Port placement is an important issue right now for our partners, so how can we solve that? There are many companies doing things such as intelligent trays and feeders, so the question is how would such a product be if a robot was doing that?” he asked. “Feeding the parts through a robot is sometimes in some application as important as picking the part and placing it in the machine because of the complexity of the input and the output.”
4. New use cases a ‘piece of cake’
New end effector options, wider interoperability, and improved safety all enable cobot users to apply automation to even more tasks, across industries.
“I remember the first robots we sold back in the mid-1990s,” recalled O’Brien. “IAI had introduced some low-cost SCARA robots, and we had visions of high-tech applications in factories.”
“It was eye-opening that the first robots we sold were installed in a bakery to grease cake pans,” he said. “It turned out that the operators who were doing that job didn’t love it, and they’d occasionally let a pan get through that wasn’t completely greased. When that happened, the bakery would bake the cake, but then it would be destroyed while they were getting it out of the pan.”
“It turned out that robots could do the task repeatably, ending the problem of the wasted cakes,” O’Brien said. “Back in those days, the robots had to be in cages, and typically you needed engineering to get a system up and running. With today’s robot and end-of-arm tooling options, the number and type of applications we can solve has dramatically increased, while the engineering and design overhead has been greatly reduced, along with the time to deployment.”
A related challenge facing manufacturers looking to automate is determining whether an application is good or bad for robots to handle.
“The trap there is that things that are so easy for humans are extremely difficult for robots, and vice versa,” said Bouchard. “That’s why sometimes if you’re new to robotics, it’s really hard to evaluate the complexity [of an application]. That’s where the knowledge of our team and partner network is very important.”
Bouchard said he advises partners to go to factories to see all of the applications and find out what the end user wants to automate. “If you see that it’s too complex because you know the complexity of robotics, make them understand that they should be targeting a simpler application,” he said.
“We have some guidelines and documents explaining the process, the parts, the parts presentation on the cycle time, etc.,” Bouchard added. “These are rough guidelines, but one of the challenges in this industry is that it’s hard to systemize all the learning, experience and tacit knowledge that adds a lot of value. That’s why once you’ve seen a lot of different projects, you can explain and really guide the partners in that direction.”