Ask West Michigan manufacturers how they’re addressing the challenges of finding qualified workers, and many of them likely will cite automation equipment as a key part of their solution.
Industry statistics would seem to bear that out. Capital investment in robotic equipment has been skyrocketing at a double-digit pace, with the automation supply chain struggling to keep up with increasing order volumes. Additionally, the cost of automation continues to decline, meaning that more companies should be able to afford the technology.
But economist Paul Isely said there’s just one problem with the commonly repeated narrative proclaiming the rise of the robot: Labor productivity statistics for Michigan suggest that it’s not true.
“We’re really in a world where the output per worker is static or even dropping,” said Isely, the associate dean and professor of economics at the Grand Valley State University Seidman College of Business.
If manufacturers were turning to automation in droves, worker output should be increasing, not moving in the opposite direction, he said.
“What that tells me is that we’re not replacing workers with automation. If we were, you’d need fewer workers to produce the same amount of stuff, and we’re not seeing that. We’re actually seeing that workers are becoming less efficient (and) that those workers are producing less stuff each,” Isely said.
While manufacturing labor productivity data — defined as output per hour worked — increased sharply following the recession as companies leaned their operations, it’s largely flatlined during the recovery. Labor productivity grew approximately 11 percent in 2010, but has hovered around 1 percent since then, according to data collected by the Bureau of Labor Statistics (BLS).
During past recessionary periods, labor productivity has grown at nearly twice that rate, Isely said.
“As we look at the things that really drive us here in West Michigan — those manufacturing components — there’s just not evidence that people are running to have machines build everything,” he said.
For the economist, the disparity suggests perceptions of the industrial market may not match reality.
“We certainly have a lot of anecdotal evidence that the companies selling automation to manufacturing are doing a very brisk business right now, but it’s not breaking through to the end result where we are seeing output go up,” he said.
Unsurprisingly, executives in the automation sector take issue with Isely’s analysis, citing evidence that companies are increasingly adopting automation, even if that hasn’t translated into productivity statistics.
“I don’t think everyone would agree with that economist,” said Jeff Burnstein, president of the Ann Arbor-based Robotic Industries Association.
While Burnstein agreed productivity was not growing quickly nationwide, he countered that automation equipment was only one factor that contributes to increasing the labor productivity rate.
“We’re one element of productivity improvement technologies and we’re not growing fast enough to increase the productivity rate in the U.S. just on robot sales,” Burnstein said.
Although sales of robotic and automation equipment have surged in recent years, Burstein believes the technology has not fully penetrated into all segments of the manufacturing industry.
“For small and medium-size companies, many are just now getting around to the idea that this is something they need to do to stay globally competitive,” Burstein said. “For a long time, companies thought the best way to compete was to outsource and that’s proven to not be the best way to do things. As the cost of labor is going up in China and elsewhere, a lot of companies are now saying, ‘Maybe we could do this by automating.’”
Shipments of robotic equipment increased 10 percent to 30,875 units in 2016, according to data from the Robotic Industries Association. Meanwhile, capital investment in robotic equipment reached $1.81 billion last year, an increase of approximately 13 percent compared to $1.6 million in 2015.
The disparity between productivity data and robotic equipment sales could stem from the type of technology that companies are adopting, said Mark Ermatinger, vice president of sales for Zeeland-based Industrial Control Service Inc.
In a positive economy, companies put more emphasis on quality control, which leads to more investment in automation technologies like machine vision, according to Ermatinger. But automated quality equipment often requires a human operator to oversee it, he said. That means a company’s headcount won’t necessarily decrease when they add the automation equipment.
“Gauging and inspecting with machine vision, these are all things where if I add that automation, I still have that operator,” Ermatinger said. “If production goes up, I have to hire more operators. I would say that 50 percent of what I sell isn’t going to change one operator to another.”
Industrial Control Service has benefited from manufacturers’ adoption of automated quality systems. The company, which consults with manufacturers in building automation lines, is on pace to generate $10.5 million in annual sales this year, up from $7 million in 2016.
For his part, Isely notes that manufacturers could be maintaining their human workers in parallel with the automated systems as they integrate the new technology into their operations.
“I’ve talked to several employers in West Michigan that are doing that — they’re keeping parallel systems,” Isely said. “When they’re sure the machine is doing a good enough job, then they can replace them, and then you’re going to see a precipitous drop in the total number of workers and you’ll see a big jump in the total amount of productivity. But I don’t know when that would be.”
Isely also posits that manufacturers may put off investing in automation equipment simply because they can’t justify the return on investment, despite the cost-savings narrative.
“Those are profit-making firms. They’re going to be making the decisions that they think benefit them the most,” Isely said. “If they’re not putting in automation or at least not doing it at a high enough rate that we’re seeing it in the aggregate data, they’re not doing it because it’s not cost effective yet.”
But robotics industry executives dispute that notion, citing industry statistics that suggest otherwise. According to the Robotics Industries Association, the average cost of a single piece of robotic equipment in 2006 stood at around $75,586. A decade later, the cost per robot fell roughly 22 percent to $58,714.
“Automation has never been cheaper,” Ermatinger said. “It’s crazy how far down the cost has gone. It’s never been more affordable.”
As he sees it, the affordability of robotics and the current economic conditions that have led to yearlong order backlogs could be causing a delay in the capital investment in automation showing up in the productivity data.
“I don’t think (companies) can buy automation because they can’t get anyone to build it for them,” Ermatinger said. “People are just way too busy. Machine builders are all strapped, they’re all stressed. Everyone I talk to is so stressed.”
A LACK OF SPACE
Isely acknowledges that the constrained industrial real estate market in West Michigan and other cities nationwide could play a factor in some manufacturers delaying the investment in automation simply because they lack space.
Manufacturers moving to West Michigan or expanding in the area have struggled in recent months to find available modern buildings that suit their needs, as MiBiz previously reported. The available buildings are often so outdated that it is more cost effective to construct a new facility, sources said.
Isely believes that dynamic could affect how much companies are investing in automation technology.
“You can’t bring in the machines at the same time you have your regular workers doing the transition if there’s not available space to transition into,” he said. “It might be that we have that type of thing going on.”
Ermatinger agrees with Isely on that point, noting that several manufacturers he’s spoken with are “dying for space” and are waiting before they add automation equipment for that reason.
Despite Isely’s skepticism on manufacturers’ adoption of automation equipment, he thinks the data will show an uptick in productivity related to the technology in the years ahead.
“To me, it’s just odd because the data isn’t showing what we’d expect,” Isely said. “I think we just ran out of labor and (prior to that) it was cheaper to keep increasing labor than to move to automation. It very well might be as we get to the end of 2017 and beginning of 2018, the data will start to show it.”