At the HOUSTEX show back in March, as the flood of attendees hit the expo floor, Dan Allford sat back in his booth to watch the traffic roll in. He had the perfect view.
Allford is president of ARC Specialties Inc., an automated welding systems producer, which meant he and his booth were squarely in the middle of robot territory.
"From my position, I could see the traffic going through all the robotics booths," he recalls. "As I watched, all these people kept going right around the traditional robot suppliers to investigate these odd little user-friendly, inexpensive robots instead—they walked right by the big ones to go see the Universal Robots booth."
Basically, he had a front row seat to a full market shift. The buyers and operators stalking the floor were no longer quite as interested in the big metal automatons anymore. Instead, they were clamoring to see these tiny little machines that swapped super power and lightning speeds for safety and simplicity.
That night Allford would have an epiphany—and traded places from spectator to major player on this new landscape.
"I went to dinner with the guys from Universal Robots that evening," he says. "ARC Specialties was an integrator the next day."
On the outset, this is a pretty surprising move. ARC Specialties falls much more comfortably into the traditional robot camp than with these new collaborative systems. Its core business is as an integrator for long-armed welding robots and a producer of large dedicated welding machines and gigantic Cartesian and cylindrical coordinate robots. That last business line in particular is a special source of pride for Allford.
"We're one of the last in America to build these big, heavy ones," he says. "We're welding things like blow-out preventers that can weigh up to 10,000 lb. I challenge you to find any other manufacturer that has a system that can tilt and rotate a part that big while welding it."
With all this heavy steel in its portfolio, "surprising" is probably too weak a word to describe the move to work with Universal Robots.
But to Allford, it made perfect sense.
"You have to understand our background to get it," he says. "We make welding approachable, that's what we all do for a living here. And that's why we've been so successful since 1983."
All of the company's traditional machines, he says, are purposefully built to be easy to program and operate, no matter how large the project they are designed to serve. So when he had the opportunity to create something on a smaller scale with a collaborative system, he jumped on it.
"I've never seen a robot with a stitch weld feature," he explains. "And I thought, 'Wouldn't it be nice to program two points and then build a robot to put 1-in. welds on 3-in. centers? Wouldn't it be nice to program the robot by physically manipulating it into position?' Well, now we're doing that."
Fellow Collaborator
On the other side of this arrangement sits Universal Robots—one of the founding companies of the entire collaborative robot market. In business just since 2008, the company's slick gray robotic arms have already become a common sight in factories and research facilities around the world. They perform an absurdly broad range of applications that stretch all the way from tending machines to landing planes.
UR's entrance into the robotics market was a rare kind of pure disruption that took everyone by surprise.
"We created a market that no one else could see," recalls Esben Østergaard, co-founder and CTO of Universal Robots, which is now owned by U.S.-based Teradyne.
"It's a new market for robots, a new wave of automation," he says. "It's a whole new way to look at manufacturing."
And that sounds pretty hyperbolic. But, looking back at the impact the company has had in its short history, it seems rather accurate. Prior to Universal Robots' launch—and that of competitor Rethink Robotics the same year—the automation market was ruled exclusively by gigantic, head-smashing, purely industrial robots. They were complex and cold machines; they were super-powerful and super dangerous, surrounded always by fences and gates and lockouts and light curtains to make sure they never made any kind of contact with their human operators. They were also some of the most expensive assets in most plants, particularly once you factored in the heavy programming and integration costs they required for every job.
But Universal Robots ignored basically that entire paradigm. By focusing on safety and simplicity over speed and power, it brought robots out from behind the guards and onto the line next to human colleagues. It also completely re-envisioned the programming system, creating a "teaching" system that allowed anyone on the floor to program movements in minutes. And, to top it off, compared to the traditional systems, these were downright cheap.
It also sought to automate an entirely new segment of the manufacturing workload. While the traditional machines took on the heavy-duty jobs, building cars and tossing tons of materials across factory floors, the cobot targeted what had always been considered human jobs—pick and place operations, machine tending; simple, repetitive tasks. And, because most setups don't require fencing, integrators, or the high investments of the traditional systems, it could offer this new automation to small and mid-sized shops that could never fit or afford them before.
It was a mix that, though few perceived ahead of time, fit a giant hole in the market. According to Østergaard, the company has already put well over 18,000 of these machines into the market, contributing to a bustling new collaborative robot market that is estimated to be worth nearly $4.3 billion in the next five years.
In that sense, Allford, watching the tide shift from his booth, read the waters correctly: this is where the market is going, and now is the time to get in.
It is, however an odd mix. A giant welding robot maker suddenly making tiny welding robots? A light-duty robot maker going heavy-duty? It was a stretch, but it worked in a way.
"Welding is a new twist here," Østergaard says. "But we see all kinds of applications for our robots appearing in so many industries all around the world, so I'm not really surprised."
And, he says, it adds a kind of symmetry to automation history that really makes a lot of sense.
"It's kind of making a full circle," he says. "Robots came from welding, that's where it all began. Now these collaborative robots are going into that again, but from a whole new way. There's a lot of value in that."
Welding Industries
The match between these two companies began at a tradeshow, so it only made sense to release the result of it at another. And so last month at FABTECH—after a shockingly quick development cycle—Universal Robots and ARC Specialties introduced SnapWeld, which merges the user-friendly versatility of cobots with the calculated precision of automated welding.
The system equally balances the assets of each company. From ARC Specialties, it brings a Profax wire feeder and water-cooled torch for welds up to 600 amps—with torch bracket, cables, and hose packages included. From Universal Robots, there is, of course, the robotic arm, which comes equipped with its simplified Universal Robots+ platform that allows users to set vital process aspects—wire feed speed, burn back time, gas flow time, crater fill time, volts, amps, etc—directly on the teach pendant.
This set-up does change the fence-free nature of the robots, though. We're talking arc welding, after all, not machine tending, so operators need to be protected from light and heat just like in any other automated welding system.
But, Allford says, the collaborative nature of the bots remains a key element to the system.
"When you're programming the machine, the operator is right in there with it, teaching it what to do without any safety concerns," he says. "That's when the collaborative aspect really shines."
The result of all this is an interactive, easy-to-use welding system that allows operators even in small fabrication shops can setup, manage, and deploy wherever they are needed—a system that would have been unimaginable just a few years ago.
"[SnapWeld] fills a niche that the industry needs right now," says Allford. "Everybody is trying to save money; everybody is trying to select technologies that are simple and fast. This system qualifies on all these fronts."
While it is far too early to tell if this will play out as well as Universal Robots' past successes, Allford suspects that it will make up a "significant portion" of his company's revenues in 2018. And, looking at the general state of the welding industry, he could be right.
According to the American Welding Society, the average age of a welder today is around 55 years old. As those experienced workers retire out, the industry could face a shortage of as many as 372,000 welders by 2026.
If done right, the SnapWeld system could be a tool to help fill that shortage with some real robotic expertise.
Robots vs Humans
It's impossible to conclude a discussion of any automated system—particularly not one dealing with collaboratives—without addressing the jobs issue.
Perhaps because cobots perform jobs at roughly human speeds, and perhaps because those jobs have been performed by humans for so long, these robots have added new fuel to the robot takeover fears that has been raging for the last 50+ years. Even with the dire future facing the welding industry, developing what might appear to be a robot replacement for experience workers threatens to fan fresh flames.
But that line of thinking, Østergaard argues, is missing the point of the technology.
"Our robots are not going for lights-out, fully-automated factories—they are a mix of man and machine," he explains. "They are creating a new kind of factory that mixes machines with the process experts and the product experts so they can get more out of their floor space.
"It's not replacing humans," he adds. "It's optimizing their efficiency."
SnapWeld System Includes:
• UR Robot
• Profax Wire Feeder
• Water-Cooled Torch
• Welds up to 600 Amps
• Torch Bracket
• Cable and Hose Packages
Universal Robots+ Platform to control wire feed speed, burn back time, gas flow time, crater fill time, volts, amps, and more directly on the teach pendant.
See more on the NED Directory
The system, he says, depends on humans, depends on those experts already in place who know how to create good welds. Without them, the robot is useless.
"People working on the shop floor know the processes, they know the products, they know what good quality looks like," he explains. "Replacing humans with robots is a bad idea—you are just throwing away the core knowledge of your product."
On the other side of this, Allford seems to take this same human-centric perspective.
"Even with automation, I still need somebody that understands what a weld puddle should look like," he says. "You give me that guy and I will teach him how to run the machine and get him out from underneath the welding wood. He'll still be a welder, of course, but now he's a programmer, also."
In that sense, they agree, a system like this is far from a threat to the industry. It's just, as Østergaard explains, another tool.
"By giving these people a robot to use as tool to do their work more efficiently, rather than replacing them with a robot, you are creating more value for your operation," he says. "I know that sounds abstract, but it's true."
Next Steps
Even with the SnapWeld system barely on the market for a month at this point, Allford is already dreaming up new ways to apply Universal Robots' collaborative technologies in the welding industry. For his next trick, he and his team are working on a new system called "CutBot."
Though it is under fairly tight wraps until its release in May 2018, Allford was able to provide a few on-the-record details for this new scheme.
"We're going to give the robot a plasma cutting torch and, with the current plan, a magnetic base so you can position the robot anywhere on your workpiece," he explains. "It's basically a robotic equivalent to a magnetic base drill."
The difference, though, is that, while drills limit you with round holes, CutBot will achieve far more complex results.
"Wouldn't it be nice to be able to do an octagon with a quarter inch radius corners and 12 degree bevels?" he suggests. "No human could do that. But CutBot will."
And that, one could say, completes the automation circle for collaboratives, taking us from far beyond the realm of light-duty automation to add new skills and capabilities to the market. Which is exactly the trajectory we are on.
Read more about SnapWeld on the NED Directory.