Rethink Robotics GmbH
Sawyer
Sawyer
Sawyer
Sawyer
Sawyer

The Great Robot Takeover: Fact or Fiction?

June 10, 2019
It's inevitable. The robot invasion is coming. But that's not necessarily a bad thing.

"If you were 22 and had a job where you were treated like a machine and knew you had about 30 years to go, how would you feel?”

This line is so timeless and universal that it could be from any era in any country, but it's a quote from a UAW official made in 1972 in Ohio's Mahoning Valley about the infamous unrest at General Motors' Lordstown plant.

In 1970, the plant had the most advanced manufacturing automation in the world. Its 26 robots performed 520 welds per car, churning out 100 Chevrolet Vegas per hour. At the time, the plant was the fastest in the world and represented the impact automation could have on production.

Automation cut the standard 60-second takt time down to just 36 seconds. The problem was, the humans on the assembly line could not reasonably match the robots' speed.

Quality often suffered as a result—kind of like that "I Love Lucy" bit with the chocolates and conveyor belt, only hilarity did not ensue. A control box was set on fire, seats and wiring were slashed, and management cried sabotage. The overall discontent led to a 22-day worker strike in 1972 that cost GM $150 million.

It's a valuable lesson on why humans don’t like being treated like machines.

Nearly 50 years later, the manufacturing industry is facing a similar battle. With a new breed of robots on the market and more sophisticated automation solutions arriving every day, many workers once again feel they're on the front lines of a full-scale robot takeover.

The fact is, robots will take some jobs and create others. And the tech is not that advanced…yet. Amazon, at the forefront of automation, says it's at least a decade from fully automating a single order.

But the takeover isn’t all fiction. Or all bad. But it is becoming clearer, and how the tech will affect everything from plant culture to quality to safety and viability. The goal is to help you sort out what's real and what isn’t so you can make the right decisions in what we can all agree is an unprecedented era of change.

The Automation Army

For Mark Jagiela, CEO of automation and test equipment manufacturer Teradyne, the robot/human clash goes back a full generation. As a teen in the 1970s, his father worked for an industrial robot manufacturer serving the automotive industry. At the time, the industry had a complicated love-hate relationship with automation—management loved the productivity, while workers hated the idea of being replaced.

"People blamed automation for the displacement of jobs," says Jagiela. "Factory workers conspired to thwart automation by gumming up the works.”

"I understand the emotion—people who don’t have a job and seeing a machine doing something they can do," he adds. "It's not new; it's been around a hundred years since the industrial revolution began."

For many, the Lordstown antics of the '70s conjured images of the famous Scottish Luddites, who rebelled against some of the first industrial machines of the previous century.

Meanwhile, in Japan, businesses (and consumers) embraced robots of every kind, which Jagiela saw firsthand as general manager of Teradyne’s Japan division.

"In the 1980s, Japan was viewed as the biggest threat to U.S. industry because of the leadership they had in automating auto-motive manufacturing and semiconductors," he notes.

Now the typical U.S. factory is fighting a war on two new fronts: one with time, specifically the aging workforce running out of it; the other with interest, which is totally lacking from the potential reinforcements due to the repetitive, boring nature of the jobs.

Jagiela, who became president of Teradyne in 2013 and CEO the following year, is seeking to broker peace between Americans and automatons by marshaling a new legion of automated workers called collaborative robots, or cobots. Defined by safety features such as padded surfaces, limited speed, and force-torque sensors to prevent pinching and crushing, cobots are finding employment in factories and job shops of every size, as well as hospitals, homes, and everywhere in between.

The first recruit was Universal Robots, which introduced the first collaborative robot back in 2008 and is the current market leader in the space. The Danish company has grown around 500% since being acquired in 2015 by Teradyne and sold its 30,000th unit last year. (To put things in perspective, it only sold around 6,300 total from 2012 to 2015.)

According to the Robotic Industries Association, cobots, which accounted for an estimated 3% of all robot sales—or 11,416 in 2017—are expected to capture 34% of that market in the next seven years.

Next was Mobile Industrial Robots (MiR), the market leader in autonomous mobile robots, bought in 2018. The rapidly growing company, also headquartered in Denmark, has reported two consecutive years of 300% growth and hired 100 more employees.

Finally, last year Teradyne also snapped up Cambridge, Massachusetts-based Energid, which makes a robotic control framework that makes it easier to train the machines to move more precisely.

With these critical divisions at his disposal, you could call Jagiela a key leader of the great robot takeover.

But first, you have to buy in that a takeover is even happening. And if by that you mean that these new cobots are taking over for their larger industrial brethren locked in cages and the workers who would prefer a more cognitively challenging, less mundane workday, it’s a clear fact.

But if you are thinking more in terms of robots taking all the U.S. jobs—or at least the 47% predicted by University of Oxford researchers in 2013—Jagiela says that is a work of true fiction.

"It's pretty clear that prediction is wrong," he says of the damaging 6-year-old study that the robot industry can’t seem to shake. "Nobody is confronted with any meaningful sense of robots stealing jobs. Look at employment."

In May U.S. unemployment hit 3.6%, the lowest since 1969. Japan has a 2.5% unemployment rate and its robot takeover happened three decades ago.

"Despite all this fearmongering and sensationalism, it's just not happening," he says. "And I don't think it will be. All the evidence suggests that historically it's been a tremendous benefit to job growth."

So nothing to fear but fear itself and all that, right? Not quite. Like everything with robotics, it's complicated.

"It's more of a societal benefit," Jagiela continues, "But that doesn’t mean along the way the nature of work doesn’t change and some people are going to be affected along the way in the short term."

Auto Mates

The first industrial robot, Unimate, started in a GM plant, and automotive productivity wouldn’t be anything close to where it is without the automation innovations that followed. But until now the factory has split people and bots, due to obvious safety concerns.

But we're learning how much more efficient workers can be with a little help from technology, such as augmented reality. It's a lesson Ford seems to be quickly understanding as it rolls out more cobots onto the factory floor. Ford has about 100 cobots spread across 24 plants, using their new Advanced Manufacturing Center in Redford, Michigan, as a testing ground for new applications and where human work can be­come superhuman with a robot sidekick.

They have a Universal Robots silver, grey, and powder-blue arm mounted with cameras to inspect vehicle chassis and another methodically sorting screws with a pair of parallel grippers. Next to those a more robust Fanuc CR-35ia cobot simulates running down fasteners in electric car battery packs.

"It's growing exponentially," says Harry Kekedjian, Ford advanced controls and digital factory manager, about the au­tomaker's use of cobots.

"The increase in efficiency they bring over traditional robots is opening up our eyes on different ways to process the work content that normally we wouldn’t have thought of before."

The most successful implementation to date involves a KUKA LBR iiwa cobot, which performs vital but monoto­nous engine inspections. The machine's visioning system can be programmed to scan for discrepancies and then alert the nearby worker to fix electrical connections.

This application is used on 16 lines across Ford engine plants. It's a perfect pair­ing because the robot excels at the task that humans generally find extraordinarily boring—basically staring down engine blocks all day like they’re the back of a Highlights magazine.

One extra benefit is all the data are being collected, which allows designers to improve engineering and reduce faults in the future.

"Automated inspection allows us to get a better sense of how well we're performing a lot of those manual tasks and be able to provide that feedback instantly to the production team," Kekedjian says.

He adds that the automated quality scan backed by a hu­man creates "a tremendous reduction in those early concerns at the dealership" while also improving productivity. He couldn’t provide hard data but says the proof is in the repli­cation rate of the application, which continues to grow.

Most importantly, due to the safer nature of these cobots, which will shut off if they so much as brush up against you, and their use as aug­menting tools and not replacements, the workforce is much more accept­ing of these bots.

"We've had a really good response with the collaborative robots as op­posed to other types of automation," Kekedjian says. "They are usually be­ing introduced to enhance the work, so it's there to help them do their job, which would be much more mundane and easy to miss."

Enhancing SMEs

Autonomous Mobile Robots were more conspicuous at Ford. Some KUKA mobile platforms are parked in a corner, one with the 7-axis robot on top. During a holiday party at the center in Decem­ber, Ford employed this machine to roll around and hand out drinks.

At IMTS in 2018, KUKA had a similar setup for in­specting a car's paint job.

At MiR's sprawling display, a MiR200 had a UR-5 on top to simulate picking up PCBs. These are certainly effective at showing what robots can do for large companies, but Jagiela has his sights set on dis­rupting industrial work on a much bigger battlefield: small and medium enterprises.

SMEs comprise nearly 99% of the total number of busi­nesses in America, and they need the most help. When asked to think of the most representative candidate that Teradyne's tech can help, Jagiela imagines a small manufacturer in mid­dle America, Iowa or perhaps Minnesota, where a worker is pushing a cart of material from the warehouse to "islands of people " at disparate cells, where a dozen more workers busily assemble product.

He sees a solution where a low-profile mobile robot, fitted with sensors to safely navigate and able to carry more than 1,000 pounds, delivers new material to a mix of maybe seven workers and three cobots, then takes the assembled product back to the warehouse.

He says this incremental path is a real strategy for SMEs to become globally competitive.

"The small company no longer has this massive barrier or disadvantage," says Juergen von Hollen, Universal Robot's president. "They have a tool similar to the large companies that generate efficiencies and scale."

Von Hollen points to the cobot’s flexibility. While a typical industrial robot is fixed in place and needs a huge footprint and safety gating, cobots can be placed virtually anywhere, programmed to do many jobs. And they also fit into SMEs' more modest investment timelines.

"With the megatrend towards mass customization, it be­comes more and more difficult to create a business case of five, 10 or 20 years for a very static automation process or line," von Hollen says. "If it's more than three years, most companies are not willing to take the risk. The world has changed."

That's perhaps why cobots could have a third of the market by 2025: they are affordable, automated Swiss army knives.

"We're bringing back the concept of the robots as a tool," von Hollen says. "That's all it is."

It's actually that and more. Recent data from small business mentoring company SCORE says that 89% of small man­ufacturers cannot fill all their job openings. And in 2018, it took more than three months to fill a production job, Deloitte found, an increase of 33% from 2015. The future looks even bleaker. The Manufacturing Institute and Deloitte's 2018 Skills Gap study reports less than half of the 4.6 open manu­facturing jobs over the next decade will be filled.

So it's not as much SMEs need cobots to remain globally competitive. They will need them to exist at all.

Cohabitating with Cobots

About 30 miles north of Ford's innovation center, a 67-year-old contract manufacturer called Fitzgerald Manufacturing Co. has found how valuable cobots can be to solve the local skills gap. In 2018, I spoke with their president Kevin La­Comb about their recent implemen­tation of Rethink Robotic's Sawyer, a red cobot with a touchscreen that can become an emoting face.

Sawyer takes metal motion control cylinders, short or long, from a pin board to a honing machine. When the parts are done, it washes, dries, and packages them. Previously, a human would have to sit there and do this fairly simple job, instead of running one of the shop's 100+ other more complicated machine tools.

"Humans would be bored to tears," says LaComb, adding that boredom eventually affects quality. "No one really understands how that creeps into the finished product."

Sawyer doesn't get bored or distracted. It also rarely stops and is able to run up to an entire day sometimes with­out an adjustment. "Sawyer runs seamlessly, and there's not a lot of hand-holding and babysitting," LaComb says.

This cobot currently costs under $40,000 and at the time LaComb estimated the ROI would come at seven or eight months. The added productivity allows Fitzgerald to accept more jobs and keep employees. In 2018, LaComb said the company had 86 employees but was short about a dozen.

"We have to remain relevant and this is a path forward in many directions," LaComb says. "It's harder and harder to find people, and when we do find good people, we're driving to put them on more advanced processes."

Rethink closed down in October 2018, after a decade of operation. The HAHN Group, which installs about 2,000 robots a year, acquired its IP, so Sawyer (but not two-armed sib­ling Baxter) lives on. HAHN plans to combine the proprietary software called Intera, which "allows programming in a mat­ter of minutes," with their German engineering knowhow to innovate more customer-oriented cobots, says Philipp Un­terhalt, CEO of Rethink Robotics and HAHN's managing director.

Takeover Non-Fiction

Cobots do allow companies to accept more business and spackle the gaps left by today's monotony-averse younger workers. But no one really knows what these cobots will evolve into as artificial intelligence takes hold and the grip­pers become more dexterous. Predictions range from creat­ing one billion jobs to stealing 2 billion. These saviors could become an invasive species soon enough.

There's a greater problem American manufacturers have right now when it comes to robots.

"We don’t have enough plant tech­nicians to take care of automation being put into manufacturing in the United States," Universal Robots' von Hollen says. "These are high pay­ing jobs. You create another indus­try that requires people to do these things."

Overall, the level of robotics knowledge the next generation has versus what it needs is concerning. At our Manufacturing & Technology event in Pittsburgh earlier this year, I asked three high school females from a successful robot club if they are con­sidering robotics as a career. All three said "no."

There's clearly still a disconnect, and the real takeover needs to be part of the whole robot narrative.

Robots aren’t just going to be in factories in the next decade, but also on hospital floors delivering drugs or moving patients, in homes assisting the elderly or disabled, and hopefully doing my yard work. They are going to be ev­erywhere, and we need to prepare.

So, the real story from here needs to be how to create the skilled workers to install, run and maintain these robots, and how to responsibly transition the people replaced by automation to new jobs. Boston Dynamics' Atlas regularly goes viral for new feats of agility, while Japan's National Institute of Advanced Industrial Science and Technology is using the HRP-5P humanoid robot to (slowly) perform heavy labor such as drywalling.

In ten years, it's conceivable that robots will assume more trades jobs. We have a new opportunity to redefine labor and restore humanity and dignity to the workers who break their bodies to boost productivity. It’s a debate that’s gone on for long enough and we have the technology to end it.

The question is, do we have leaders able to will it into reality?