I hate to break it to everyone, but the skilled labor shortage isn’t going away anytime soon. Nor are import tariffs and reshoring initiatives going to make overseas suppliers any less competitive, despite the claims of those responsible for such measures. And as the past year has shown, random bits of genetic material called viruses tend to pop up from time to time, making workers sick or worse.
The writing is on the wall: Automation is the clear path forward for shops that wish to stay relevant in 2021 and beyond. But questions abound: What robot to buy, how to implement it, who can program it and where on the shop floor to put it. When Luke Skywalker chose R2-D2 from a long row of similar-looking droids, he had The Force guiding him. For everyone else, though, a little assistance in the decision-making process could come in handy.
Stefan Drakensjo, vice president of marketing and sales for general industry at ABB Robotics and Discrete Automation, has some good advice on taking the right approach. He suggests that, while return on investment (ROI) is important, those new to automation should generally start with whatever’s easiest to implement.
“If you’re a small to medium enterprise at the beginning of an automation journey, begin with a project that will secure a smooth integration and get your staff up to speed quickly,” he says. “That’s not necessarily low-hanging fruit for every company – it could be the biggest bottleneck or the highest volume part – but at least try to avoid very complex applications and use a standard solution wherever possible.”
It’s obvious that ABB Robotics and the other companies that contributed to this article offer such solutions, in some cases dozens of them. All are high quality and represent a significant step forward for any shop new to automation. What’s more important than any specific brand, model or feature, however, is that the managers and owners of these shops recognize the fundamental problems outlined in the opening paragraph and remain committed to addressing them.
Drakensjo notes that automation helps ease the skilled labor dilemma in several ways, although not in the ways one might expect. A robot eliminates the need for a human, yes, but what’s often more relevant is the company growth that typically occurs after automating. This in turn creates more jobs, and better yet, the kinds of jobs that are easier to fill.
When a manufacturer installs robots, he explains, it creates new kinds of employment opportunities. The people who were doing the manual labor previously often end up managing, operating and programming the robots. They also have more time available to help with continuous improvement and process optimization activities. It raises the level of employment for these people, which usually pays more and is generally more attractive to the younger generations who grew up with technology.
At the same time, part quality improves and the company becomes increasingly competitive, able to take on different work and additional orders. Simply put, it’s a winning situation for all involved.
“I work for a robot company, so I’m admittedly a bit biased, but it’s a known fact that the countries with the highest robot density have higher GDP rates,” Drakensjo says. “Similarly, companies that implement robots are known to have higher performance levels. And because they’re able to offload the dull, dirty and dangerous work to a robot, they can offer better, safer jobs with higher retention levels.”
So, what kinds of tasks can robots perform? In a word, all of them. Machine tending, assembly and packing, painting, deburring and especially welding – these are just a few of the roles that robots assume every day in large and small shops alike, even on lower quantity work that has traditionally been off limits for automation.
Mark Scherler, general manager of the materials joining group at Fanuc America Corp., can speak at length about welding, although he’s quick to point out robots can do much more than welding in a typical shop. He and others attribute at least some of this to the increased use of vision with robotics, which together with force sensing and advanced software systems give these little droids almost human-like capabilities.
“Consider a parts-picking application at the end of a laser cutter or turret punch press,” he says. “It’s easy to have a vision system-equipped robot identify, pick and stack the different parts cut from a nested sheet. The challenge comes when the sheet is a little warped or the parts are sitting differently than the program expected. This is why we and others have developed intelligent vision systems able to detect and then resolve these problems without human intervention. We have also added artificial intelligence to make these systems easier for users to set up and operate.”
This level of intelligence has also opened the door to automated parts sorting and bin picking applications and has reduced the need for dedicated fixturing and conveyors on machine tending operations. In many instances, it’s possible to just drop a crate filled with raw material blanks in front of the machine and the robot takes care of the rest.
“The beauty of today’s robots is that they’ve become very flexible,” Scherler says. “If you need to adjust from one part to the next, simply reprogram the robot. That’s it. The only limitation might be the tool that’s on the end of the robot and what it can accommodate for picking up a part or whatever the other application is. That’s why I usually recommend that shops buy a robot that might seem a little oversized at the time. This gives them the ability to take on larger work as the need arises, further increasing their flexibility.”
Scherler also points to the growing use of collaborative robots, or cobots, throughout manufacturing. Though initially a self-professed doubter, he now says that these friendlier versions of industrial robots are suitable for
all of the tasks mentioned earlier, including welding. “Maybe they’re not as fast or strong as a traditional robot, but unless you need that capability for your production, a cobot’s simplicity and ease of use should be considered,” he says.
Joe Campbell was never a cobot doubter. The senior manager of applications development at Universal Robots (UR) notes that UR pioneered the use of cobots and continues to build on that success. “We have a number of partners in the arc welding space that have rolled out cost-effective solutions. It’s easy to find one complete with robot, torch, wire feed, power supply, mounting base and worktable for around $80,000, making them ideal for smaller shops engaged in low-volume, high-mix work.”
Such investment takes the pressure off a shop’s skilled arc welders by picking up the simpler work – the brackets and small assemblies – freeing the welders to do more interesting parts. “That’s the sweet spot for a cobot, particularly with the low cost and ease of implementation,” Campbell says. “It’s common to hear of cobots welding parts within a few hours of their arrival.”
He agrees with Scherler’s assessment concerning the broad range of tasks that cobots can tackle. The only exception, he warns, is painting. Due to the need for special sealing, UR does not support Class 1, Division 1 solids-based painting operations, and has no intention of changing that. “We’re quite happy with the product line,” Campbell says. “It fits where it fits, and we know it exactly.”
That fit took on an all-too-common twist with the Covid-19 pandemic as shops struggled to remain operational in the face of social distancing measures. Suddenly, the ability to have robots working among humans became more than a convenience – it became a potentially life-saving necessity.
For instance, assemblers or shipping clerks can now have a robotic companion, one that sits between them and shares in the workload while keeping them safely apart. It’s this scenario that has sparked tremendous interest in cobots of late, and Campbell doesn’t see that changing.
But even without the pandemic, manufacturers have begun embracing automation like never before. “In a typical job shop, you have a skilled machinist who sets up a CNC lathe or mill and then spends the next four or five hours loading and unloading parts,” he says. “Many companies now realize that it’s far more cost-effective – not to mention a lot more interesting for the machinist – to let a cobot tend the machine. And while it’s running, the high-paid human can go off and repeat the process, setting up the next machine and the one after that while robots do the boring work.”
Michael Castor, product manager at Yaskawa America Inc., Motoman Robotics division, sees much the same thing. “It’s very hard for shops to find workers,” he says. “And when they do find someone, it’s not uncommon for that person to walk out the door a couple of months later when another shop offers them more money. Robots don’t do that. They don’t complain, they don’t take breaks or go on vacation, and you never have to worry about missing a delivery date because one didn’t show up for work.”
Motoman’s product manager for welding, Joshua Leath, says that welding has long been the manufacturing industry’s most commonly automated process, although that’s beginning to change as robot use becomes more widespread. Aside from those already named, he lists several additional tasks such as fastener insertion, glue dispensing, parts inspection and palletization.
“We have a variety of software tools that make these and many other common operations quite easy to automate,” Castor adds. “In this last example, you might have boxes coming down a conveyor belt and you need to get them on a pallet. It’s tedious, back-breaking work, but a robot doesn’t care. Provided it’s a fairly common box size, you just upload a program or teach it what to do. Problem solved.”
Note the term “back-breaking.” Leath and Castor each suggest that robot shoppers frequently overlook this consideration when calculating their automation ROI, in that a single worker’s compensation claim for repetitive strain or similar work-related injury might cost more than a robot. But even without that unfortunate event, a one-year ROI is not unusual, and is often much shorter.
“As a rule of thumb, I tell shops that one robot can do the work of three and a half welders,” Leath says. “Similar benchmarks exist for other applications, but whatever the activity, robots can generally do it more cost-effectively. And when you add to that the broad array of programming and simulation packages that are available, the end-of-arm tooling selection and the ease with which robots can be implemented today, it’s really become a no-brainer for most manufacturers.”