Collaborating with Robots

Between “Blade Runner,” “Terminator,” “The Matrix” and other blockbuster movies, Hollywood has painted a frightening picture—one in which intelligent machines attempt to destroy humanity. And while artificial intelligence (AI) promises to change much about our future, it’s unlikely that neither it nor advanced automatons will do more than make our lives easier and more productive over the coming years. Let’s hope.

In the meantime, robots are bringing significant and quite necessary benefits to the production floor. As has been discussed at great length, the skilled labor shortage remains a chronic problem, and the most expedient solution is to automate wherever possible. While AI will help expedite these initiatives, its impact has not yet been felt to any meaningful degree. Rest assured, however, it’s coming—and sooner than any of us expect.

Joe Campbell is the senior manager for strategic marketing and applications development at Universal Robots USA Inc. of Ann Arbor, Mich., a manufacturer of collaborative robots, better known as “cobots.” A self-professed “automation old guy,” Campbell has watched as robots evolved from a high-volume, automotive-only technology to today’s landscape of “where can we put one to work?”

To anyone wringing their hands right now over potential AI-related job losses, he suggests there’s little need to worry. “I heard someone talking the other day about how Tesla’s autonomous semitrucks are going to put all the long-haul drivers out of work,” Campbell says. “But you know what? Even if the technology were available today—which it’s not—implementing it would be a major capital investment. It’s going to take some time.”

Granted, we’re seeing robots take on plenty of other jobs, everything from drywall finishing to hamburger flipping, but those positions are increasingly difficult to fill anyway, he adds. “And, at the end of the day, society has weathered these transitions before, and we’ll get through this one just like all the others.”

He’s right. Consider agriculture. According to a 2005 report titled “The 20th Century Transformation of U.S. Agriculture and Farm Policy,” 41% of the nation’s workforce was employed on small farms during the early 1900s. One century later, this rate fell to just 1.9% thanks to the development of “highly productive and mechanized farms” that rely on millions of tractors. Going forward, agricultural employment will drop even further as robots assume tasks such as planting, weeding and sorting, many of them equipped with AI-enabled cameras.

This article is not about AI, even though the technology will play a huge role in future automation. It’s about the relationship between robots and humans in the workplace—their collaboration—a topic about which FANUC America Corp. of Rochester Hills, Mich., has plenty to say.

District Sales Manager Adam Willea offers several examples of such collaboration. It could be where the robot inspects a human’s work, during an equipment installation or maintenance, for instance. There’s the collaboration that’s growing quite common in machine shops and sheet-metal fabricating, where a robot loads and unloads parts, possibly deburring and washing them afterward. Assembly lines are another opportunity for collaborative work, as is robotic welding, trimming and deflashing of molded plastic parts, packaging, palletizing, sorting and more.

In each case, humans work nearby or periodically check on the robot’s progress. Is this collaborative? “Absolutely,” Willea maintains but notes the robot itself might not be “collaborative” by definition. “Number one, it’s fenceless. There’s no barrier between robot and human. Secondly, it’s interactive, meaning the human can touch the robot and easily tell it when to start and stop. Most importantly: It’s safe. There’s no chance of injury to the human. All of these conditions must be met in a collaborative environment.”

The curveball many people fail to catch is that the robot in this environment could be a cobot, but given the proper programming and safety equipment— for example, an area scanner that makes the robot slow down and stop when a human approaches—even a so-called industrial robot can be made collaborative. “We sell both styles, and I can tell you that the term ‘collaborative’ has evolved into something it’s not,” Willea explains.

We won’t get into the relative merits of each robot type here, nor compare their installed costs. That’s an argument for people like Willea and Universal Robots’ Campbell to arm wrestle over. We will note, however, that cobots are generally far slower than robots, less accurate and much easier to implement and program (although each of these attributes is improving daily as cobot manufacturers continue to make advances).

That said, cobots are suitable for a huge number of industrial tasks, and because it’s quite easy to “roll one out,” they often serve as a gateway to shops new to robotics.

“The majority of them are going into small shops as well as some mid-size manufacturing facilities,” notes Corey Ryan of KUKA Robotics Corp., Shelby Township, Mich.

Ryan is KUKA’s director of medical robotics, a market segment that consumes large numbers of cobots. Yet his purview also extends to industrial uses. He notes that the focus over the past few years has shifted from a robot’s collaborative properties to its ease of programming. “Whether they’re looking at cobots or robots, people want simplicity,” Ryan explains. “So, if they can get it up and running quickly with their current employees—many of whom might be technical but have no experience with robotics—they’re willing to settle for a cobot’s lower speed.”

Aside from coping with the general inability to find new employees, there’s the need to retain existing staff. Doing so is much easier if you can automate the dull, dirty and dangerous tasks and redeploy workers to more interesting activities—setting up and programming robots is one. The result? Happier employees, greater throughput without increased headcount, and more predictable part accuracy and product flow.

Like FANUC, and most other “industrial robot” providers, KUKA offers a line of cobots. Ryan sells loads of them, which he asserts are every bit as good as his competitors’ products. But if a shop is looking for maximum throughput and round-the-clock operation, he’ll point them to one of the company’s industrial offerings, despite their greater complexity.

“Maybe I’ll get in trouble for saying this, but cobots are quite simply less robust,” he says. “If you run one a few hours a day, it might last 15 years, but industrial robots are designed to deal with high-volume, non-stop work. So even though they’re a bit more challenging to implement and program, they’ll run circles around a cobot.”

But, hold on; check the literature for any industrial robot. Surf the manufacturer’s website, talk to a salesperson—all will boast that their wares are more user-friendly than ever. FANUC has its ROBOGUIDE Simulation Software, for instance. There is also KUKA.AppTech, ABB Automation’s Wizard, Yaskawa’s MotoSim and Comau’s Smart packages, to name a few. Each provides an intuitive software interface, advanced simulation and application-specific tools that aim to level the cobot-robot programming field.

Usability and performance comparisons aside, the industry is never going back to the way it once was. So says Patrick Varley, product marketing manager for robots at Mitsubishi Electric Automation Inc. (Vernon Hills, Illinois). Varley has been in the industry for more than three decades and, like Universal Robots’ Joe Campbell, watched as robots went from niche to mainstream.

Varley agrees with his counterparts on the growing need for automation to combat the labor shortage, but as Ryan of KUKA pointed out, the problem is more than finding employees—it’s keeping them around once you do. “If you can’t get them to stay, you’re always in retrain mode,” Varley explains. “Your quality goes down, production suffers and when you finally get them up to speed, they go work somewhere else for another 50 cents an hour. It’s very discouraging.”

Automation brings consistency—to part quality, production costs, job scheduling and machine uptime—that manufacturers can’t otherwise achieve. “Even so, I don’t see a never-ending increase in robot density, because there will always be some jobs that robots can’t do or that people do better,” Varley adds.

Having just returned from grocery shopping, Varley offered an example that everyone can relate to: picking the best bag of cherries from the display case. Even the smartest, most capable robots lack human judgment and intuition, he notes. And while AI, vision systems and tactile feedback will bring droids closer to our innate capabilities, they’re unlikely to surpass us anytime soon.

“Even relatively simple tasks like putting two halves of a phone together or inserting a small screw into a tapped hole can be fairly difficult for a robot, and teaching one to perform such jobs can be cost prohibitive, at least where smaller quantities and rapid changeovers are required,” Varley says.

Here again, ease of use and programming will be crucial to robotics’ cause as we advance into this brave new droid world. So will innovations that have yet to be developed, which is where companies such as MassRobotics of Boston come in. A non-profit supported by various industry partners (including Mitsubishi Electric), the organization’s stated mission “is to help create and scale the next generation of successful global robotics and connected device companies by providing entrepreneurs and innovative robotics and automation startups with the workspace and resources they need to develop, prototype, test and commercialize their products and solutions.”

“I’m personally involved with them and see where these entrepreneurs are going,” Varley says. “Although much of it is not yet ready for production, some of their developments are amazing. The next few years will be quite interesting.”

While the trucking industry has little to fear from automation (at least for the next few decades), that might not be the case for those who drive forklifts for a living. Denmark-based Mobile Industrial Robots (MiR) A/S intends to automate warehouse floors everywhere with fleets of autonomous mobile robots (AMRs) that are said to maximize the efficiency of logistics operations.

As Brian Betts, director of applications projects for MiR Americas points out, the company is well on its way; MiR has installed more than 8,000 robots globally and has a large presence in the United States.

Betts ticked off several of the company’s AMR models, from its diminutive MiR100 to the MiR1350, with the digits representing carrying capacity in kilograms. Each is a battery-powered, wheeled platform designed to move boxes, pallets and containers, and when coupled with MiR Fleet—a supervisory software system that acts as a traffic cop—can navigate the warehouse and production floor autonomously, as well as manage robot missions and send the device to its charging station when needed.

“We also have a cloud-based tool called MiR Insights that connects to the fleet and allows users to visualize its operations,” says Betts. “By understanding traffic patterns and gathering various productivity metrics, customers can more easily optimize product flows and the like.”

What does it take to make material movement automated? Little more than a wireless connection, some static objects the robot can reference, and a few hours to map the floor and its many obstacles. “We have a great demonstration of this process on YouTube, but simply put, the robots use laser scanners and 3D cameras to drive around, avoid obstacles and determine the best way to navigate the facility,” adds Betts.

Whatever the future of automation, it will continue to require supporting tools and accessories. Cobots and robots use all manner of end effectors, suction cups, welding torches, drills and screwdrivers, laser heads and various other attachments to perform their duties, without which they’d be expensive paperweights.

So it is with AMRs. These self-driving droids are, in some cases, equipped with articulated arms to grab items from shelves or latch onto a bin, but as Carsten Sørensen points out, one of the most common accessories appears to be nothing more than a wheeled cart.

Sørensen is a managing partner at ROEQ ApS, a MiR neighbor and part of the Danish robotics cluster on the country’s island of Funen. In 2017, this startup company saw a need for products that would add capabilities to MiR’s growing AMR line, and has since developed carts, lifters and conveyors for it and several competing brands.

“An AMR is great at finding its way from point A to B, so to speak,” he says. “What we do is provide equipment that makes it into a more useful tool.”

Much of this equipment comes in the form of roller modules and lifter modules that attach to the AMR. Instead of goods being placed directly onto the robot’s top surface, these modules provide a smart interface that—as the names suggest—have rollers or lifters that not only secure and support the AMR’s load, but communicate with its onboard computer for increased safety and additional functionality. And, as with all of the company’s cart solutions, these increase the AMR’s carrying capacity—in some cases doubling it—while greatly increasing its usefulness as a delivery vehicle.

Accessories will be a crucial aspect of the automation industry as it moves forward, making robots, cobots and AMRs more capable, flexible and, above all, accepted by humans who might harbor fears about losing their jobs. As most who’ve taken part in the robot revolution will attest, this last concern is a nothing burger. Automation has repeatedly been shown to increase company growth and efficiency, resulting in greater security for its employees and, in most cases, leading to additional job opportunities.

Connect With Us