by Paul J. Heney, Editorial Director
Robotics is a fast changing field, and it’s hard even for the most tuned-in engineer to keep up to date with the latest technology, trends, design ideas and industry pressures. Earlier this year, the International Federation of Robotics held a high-level panel on how robots and humans will interact together in the future. Members of the panel included:
• Phil Crowther, Global Product Manager, Small Robots, ABB, China
• Professor Henrik Christensen, Executive Director, Institute for Robotics and Intelligent Machines, Georgia Tech
• Rodney Brooks, Chairman and CTO, Rethink Robotics
• Enrico Krog Iversen, CEO, Universal Robots, Denmark
• Masahiro Ogawa, Chairman and CEO, Yaskawa America
• Stuart Shepherd, Vice President Sales, Americas, KUKA US Holding
Following are areas discussed that will affect the nature of robots, from specialized industrial models to consumer-oriented devices. Watch for these trends in the coming months and years.
1. More collaboration with humans
Brooks discussed how much more consumer robots have proliferated than industrial models have.
“A previous company, iRobot, builds military robots and robots for the home, sells 2 million per year, which is 10 times as many as industrial robots; they’re much cheaper, but 2 million per year. People are seeing robots in their lives. They’re cleaning … they’re also seeing that their cars are becoming robots. They get inside a robot and it has autonomous systems.”
Brooks said that in the industrial sphere, historically, the robots were separated, dangerous, didn’t have much processing, and didn’t have enough sensors.
“Now the processing senses are available. Why can’t those robots, the industrial robots, be more like the vast majority of robots in the world, which are things that people can interact with?”
“I think there’s another thing that’s starting to happen. To use an analogy, you buy this industrial robot, and you can get somebody to program it to make phone calls for you, if you want; but instead, we’re going to the model that is here, that is a software stack with a sensor, with sensor algorithms, with planning algorithms, with all those algorithms on there. Once you have that built in, then the robot becomes something that you can safely interact with from day one, from when you get it out of the box. That’s a change of thinking.”
“There are a lot of benefits [to a software stack],” Iversen agreed.
“If you see it from the customers’ point of view, which is really important, they have a much faster deployment of the robot. They have a much smaller investment; there’s a lot of equipment they don’t need to buy; safety equipment, they don’t need to put around the robot. They need a lot less space. Basically, they are out there, achieving their savings a lot faster than they would with traditional robots.”
“By allowing us to have collaborative robots, a lot of the auxiliary hardware goes away, and at the same time, the programming goes down tremendously,” Christen said. “It implies that we deploy products that are much cheaper, we can do it faster, and we really get out there. At the end of the day, if it didn’t make economic sense, people wouldn’t use it.”
Brooks said that collaborative robots can be a lot more than the way many people see them now.
“I think we’ve seen a little of it—here’s an industrial robot, we’ll call it collaborative because we’ll get rid of the cage, maybe … collaborative robots, to me, are getting away from the idea that you build the environment for the robot, and for the people who are sort of peripheral around it.”
“Imagine if you want to put robots in McDonald’s, and you said first you completely rebuild the way McDonald’s is set up, then you put the robot in … that’s going to be a long production. I see it as more robots and people being close together, and if there are people around, it means it’s an unstructured environment; these things move around. … Now, with the processing and sensors, robots can sense how environments change. It doesn’t need to be down to some millimeter of accuracy because robots can adapt; and that, to me, is a big part of collaborative robots, that it’s not the clean, pristine workspace, but it’s the workspace where people are, and once you have people and robots in the same workspace, the robot doesn’t have to be 100% perfect because people can do some of the harder tasks.”
Shepherd said he sees the drive toward collaboration as being the best of both worlds.
“You get the best of what human beings do, and that’s think and act and adapt to the marketplace, and adapt to the application. A robot can do all the other things, be precise, be repeatable, be active, which is incredibly difficult for a human being to do.”
2. Robots will replace people in some manufacturing
With a new breed of interactive robots, certain manufacturing tasks will benefit the most, and that will mean a lot of changes for manufacturers and how they’ll use robots in the future.
Shepherd said that when you have a person and a machine together, you focus on the adaptability of the human being and the precision of the robot.
“Sometimes just in sheer speed, any place where you can teach something once, or do some adaptive training, the robot can perform faster than a human being can, and also be able to record that it has, in fact, done the task correctly. This is extremely important because one of the challenges for the human being is—how do you know if they did the job right? Whereas if we’ve trained a robot and validate the process, then you know the robot will do it right, time and time again.”
Ogawa agreed, and mentioned that when 100% precision is required all of the time, such as in some automotive applications, that’s an impossible situation for a human being over longer time periods.
Iversen wondered which manufacturing tasks would not benefit from collaborative robots.
“Even if you say that [a task] could be done also with traditional robots or with people … from a financial point of view, this is really where our customers are focused. It makes sense to do the automation with collaborative robots, and the more we move collaborative robots away from being dangerous machines and move them over to becoming safe tools—I think that is part of the transition that we are seeing today—the more applications it will make sense to automate. The way I look at it is that anything I can get my hands on, that will make sense to automate with a collaborative robot,” he said.
Christen gave the example of an automotive manufacturer, where robots have taken over for people in the plate shop, the welding shop, the paint shop, and they haven’t really penetrated the remaining part of the automotive line.
“It’s primarily because we are difficult customers,” he said. “We all want custom products. If you take a simple car, like a Jaguar, it’s available in one million different configurations. So as soon as you get away from the standard model of product and into the latter, [we can] have humans and robots work together on this. There are things like putting in wire harnesses and things like that, it’s difficult to do for a robot. That’s where humans add a lot of value.”
“We’re going away from the robots in the cages, so even in an area like automotive, we can see massive penetration,” Christen said. “We can see it in electronics, we can see it in aerospace … it’s going to be harder to see the places where we’ll not come this far.”
And Brooks said that in the 90% of manufacturing places where there are no robots, that’s where he sees putting in collaborative robots. Why there?
“Because that’s where there’s lots of people, and they’re doing some task which is repetitive, and it puts stress on them, or it’s just unpleasant,” he said. “But you’re not going to convince anyone of return on investment or just throwing out all the people, bringing in all the robots, set the thing up, and then have it happen. It’s got to be robots, incrementally, a little bit at a time.”
3. There’s potential for small- to medium-sized manufacturers
Iversen explained that smaller manufacturers should look at robotic integration from a strategic point of view. They should ask,“What can I do to make my company stay competitive?” and then go through the manufacturing to see where do they have people sitting today, and where it would make sense to put in robots.
“I think this is something that you would do like any other kind of automation task, or anything else that you would do to optimize your company,” he said. “You are being responsible for optimizing your company, and that means you are looking at new technology, and collaborative robots is one of the new types of technology that you need to look at. It’s no different than getting a new IT system or anything else that will basically make you more competitive, and more profitable. I don’t think that there’s a specific recipe here, but I think what we do see in the SMEs is that some … management is extremely focused on the daily operations. Probably the first step is to sort of step back, and then get away from daily operations and actually have a view on a more strategic development of the company.”
Shepherd thinks there are two pieces to integrating robots.
“One is the technical knowledge required to go through … A lot of small- to medium-sized companies don’t have the people internally who have been exposed to how to make those kinds of decisions. The first thing I would recommend is ask for help because the help is out there, and what may seem like a difficult application may in fact not be that bad,” he said.
“The other thing, frankly, is the financing. I think the thing that slows down the vast majority of small to midsize companies is they don’t know how to approach their lending institution to talk about robotic automation. They don’t understand, that much, how to get the value out of it themselves, so it’s tough for them to write that into the business plan and present that to the banker. Robots actually have a much longer life than what other capital equipment has because they’re flexible and can be reused. Many can be rebuilt or even have life extensions added to them, and they’ll go through three or four applications and last much longer than ELCs or PCs … as the small to midsize company’s business changes, the robot can be still used and still adapt to those changes.”
Christen thinks a major factor is how manufacturers get educated.
“A lot of them are still thinking about old-style robotics. It’s difficult, it takes a long time before anything starts moving. As an industry, we need to be much better at telling people, ‘Those are the old days, we can get you much faster.’ … It’s not such a big investment, but right now their expectation is this is going to be a big investment … we can’t afford to do this.”
Iversen thinks SMEs are open to the new robotic technology—and they’re buying.
“We probably have 3,000-plus collaborative robots installed in SMEs today, on a global scale, and this is just scratching the surface of what is about to happen in this industry; so it’s definitely out there,” he said.
Crowther agreed. “The smaller companies are the fastest sales cycle because the product review board is the owner of the company. With the larger companies, we’ve gone to a safety meeting, and 20 safety people are on the other side of the table … smaller companies can make the decision much more quickly,” he said.
Summing it all up
Christen sees a lot of areas where we see a mix of what is traditional manufacturing and service, such as in the supply chain area, where robotics are used for picking up materials and handling totes.
“I think we’re seeing that the new collaborative robots in many cases have a relatively small footprint, in comparison, and that implies that they’re relatively easy to put on a mobile platform. You see a mobile network’s market growing for a lot of new applications, both in terms of some doing office collaborations to the traditional material handling, to healthcare-type of applications,” he said.
Iversen described an application where industrial robots are used in a large bar in Holland.
“There are three stories, and they put in rails on the walls. They basically have the robot [run around] whenever you order a drink. It can carry three glasses at a time; it will go and get the drink, and serve it, and basically then the bartender is really down to just doing all the chit chatting and taking the payment,” he said.
“I was in San Francisco a few months ago, and a guy came up to me and said, ‘I’m opening a new restaurant, and I need to [incorporate robots],’” said Crowther. He was serious about that, so he came up with the total design of how he wanted them to handle the eggs and meat.”
Shepherd noted that more and more robots are programmed offline—especially large robots, because you can’t physically reach them or it’s not safe to go in and do the programming.
“The more that our simulation program can be what [people are] used to in the gaming world, the easier our life will become; so from a training point of view, yes, it’s the language of the user, whichever they prefer, but it’s also the accuracy and precision of the controls that we’re providing today that again brings a mid-grade-user ability into the environment.”