Recently, Chris Budnick, the CEO of Vanguard Plastics in Southington, Connecticut, needed a new employee for his 30-person factory to help with the 24-hour schedule. He could have done what managers and CEOs have done for decades: post a job ad, sort through hundreds of resumes, interview dozens of potential candidates, conduct second or third interviews and pay for background checks, then offer a wage of $9-$15 an hour and plus health benefits, a profit-sharing program, and more.
Instead, Budnick hired someone else – or something else. His name is Baxter, a humanoid robot from Rethink Robotics.
Baxter is about the same height as a job candidate but wouldn’t hold up so well in a job interview. For one, he doesn’t talk and he doesn’t have legs. But these shortcomings don’t really matter to employers, because Baxter’s a hard worker – he can work 24-hour shifts, seven days a week, and never get tired. He also doesn’t need to eat, sleep or take smoke breaks. When he gets “sick,” he can usually be fixed with a reboot or a screwdriver.
Baxter also seems to, well, like his job. He looks at you with big oval eyes and raised eyebrows as if to say, “Give me work! I’m ready!” and a cylinder above his eyes flashes green to confirm it. When you show him what to do, he moves his arms steadily and precisely to complete the job, never taking his eyes off the task.
Watching him is almost meditative – he seems so content with what he’s doing. The only time he doesn’t seem happy is when something goes awry – the object he was supposed to pick up wasn’t there, or the place you told him to put it is out of range – and then one of his eyebrows droops and the other rises for a look of utter confusion. No one likes to see Baxter this way.
Baxter is, in a sense, the perfect employee. The best part, for Chris Budnick, at least, is that even if Baxter doesn’t meet job expectations, he’d be a very inexpensive mistake. He costs only $22,000 – or $27,000 with a three-year warranty – essentially working out to just $9,000 a year, or $1.50 an hour if Baxter works the 120-hour weeks Budnick has assigned him to. Compare that to $46,000 a year for a regular employee who works 40 hours a week. The math is simple. Why wouldn’t Budnick hire Baxter?
MANUFACTURING’S LAST MILE
Hundreds of small to mid-size manufacturers around the country are asking the same question, and many put in their order for Baxter as soon as he came onto the market last fall. “We’re sold out of our production capacity through April,” says Scott Eckert, CEO of Rethink Robotics.
Experts believe Baxter and others like him will completely change manufacturing as we know it.
“It’s getting to the last part of manufacturing, which is the final assembly – the part that takes a lot of dexterity and hand-eye coordination, which has so far have been the more protected job that couldn’t be done by heavy, industrial strength robots,” says Martin Ford, robotics expert and author of The Lights In the Tunnel: Automation, Accelerating Technology and the Economy of the Future. “Now robots are moving into that last mile and doing assembly-level work.” And Baxter 2.0 or 3.0 could completely upend other industries as well – a fast food Baxter could flip burgers, a Gap Baxter could fold t-shirts, or a Walmart Baxter could greet customers – all service jobs that employ 20 percent of American workers.
What makes Baxter different from every other robot before him? Three things:
(1) Baxter can work side by side with humans. Other factory machines tend to be behind cages. They would seriously injure human workers if they got too close. But Baxter can sense when a human is near. Stand in Baxter’s way and his arm might gently bump you – but then he’ll stop.
(2) It takes 10 to 15 minutes to train Baxter to do a new task. “Anyone with a high school degree can do it,” says Eckert. A trainer simply takes Baxter’s arm and moves it to where he should pick up and put down an object. If there are multiple objects, Baxter can take a 3D image of each with the cameras in his wrists to distinguish them. Want to repurpose Baxter for another area of the factory? There’s no high-level engineer needed to reprogram him. A few minutes – and Baxter’s at work.
(3) He’s cheap – he’s half the price of the cheapest industrial robots on the market, which typically range from $50,000 to $200,000.
A PERFECT FIT
There are about 300,000 small-to-medium manufacturers in the U.S. that Baxter would be perfect for – plastics factories like Vanguard, metal parts manufacturing, and consumer goods packaging, Eckert says. A lot of these manufacturers tend to make multiple items, or fill custom-made orders, and need a machine that can easily be reprogrammed to do different tasks.
Eckert believes technology like Baxter will help manufacturers start businesses in the U.S. instead of overseas, help other factories return to the U.S., and keep current ones here. “By getting more productivity out of people through technology and robots like Baxter, U.S. businesses are competing successfully with China and low-labor cost regions,” Eckert says. “The benefit for the economy is you keep companies like Vanguard Plastics and their employees here. That’s the approach we’ve been pursuing.”
Budnick has seen factory after factory leave Connecticut due to the tough business climate: The state is known for its high corporate taxes, strict regulations, and pricey electrical rates. Budnick believes Baxter will help the company survive. “The cost of doing business in Connecticut is nuts,” says Budnick. “We’re a family-run business and we want to stay here, but there have been no new plastic injection molding companies in the state for at least the last five years, and companies like ours moved long ago. Baxter will help us exist in this environment.”
Martin Ford argues that the team at Rethink is being “careful on how they spin it” and doesn’t want people to think of Baxter as a human replacement. “They want to position [Baxter] as a non-threatening robot that works next to people. But you can’t get away from the fact that if you’re working next to a robot [today], before that you would’ve been working next to another person,” says Ford.
Ford agrees with Eckert that Baxter will help manufacturers come back to the U.S., but says “the jobs aren’t coming with it.”
A JOBLESS BOOM
Apple set off a media frenzy late last year when it announced it was moving part of its manufacturing operations back to the U.S. from China. Many applauded the move, but Apple also announced they’d only be bringing 200 jobs with it – a miniscule drop in the bucket compared to the 618,365 employees GM once employed in 1979. Since 2000, the U.S. has shed nearly 6 million factory jobs, almost a third of the entire manufacturing industry.
Manufacturing profits, on the other hand, are booming. The industry has not only recovered from the recession, but profits are up 35 percent since before the recession in 2006. The trend is also happening globally. According to a 2012 McKinsey report, manufacturing contributed 20 percent of the growth in global economic output between 2000 and 2010, but advanced nations have lost 24 percent of their manufacturing workforce.
The few manufacturing jobs that are coming back tend to be higher-skilled jobs in engineering, design and sales. In fact, despite the loss of jobs and high unemployment, many employers have complained that they can’t fill their job openings. The National Association of Manufacturers estimates there are 600,000 jobs available that employers are struggling to fill.
While many credit the shortage to a skills mismatch, others say manufacturing has an image problem. Why invest in training and education for a job that may soon be outsourced or replaced by a machine? “Despite our high unemployment, people don’t go into manufacturing,” says Eckert. “The average manufacturing worker is 50 today. Manufacturers are trying to hold on to their workers and get as much productivity out of them as they can.”
Wages for higher-skilled jobs also aren’t competitive with those in other industries. Though entry-level manufacturing jobs now require more education and training, the pay has remained stagnant. At a metal manufacturer outside Milwaukee, a worker with an associate degree starts at about $15 an hour, which is a far cry from union wages offered before the recession. According to the BLS, the average hourly earnings for manufacturing workers are back to where they were in 2000.
THE ROLE OF HUMANS
When 52-year-old Rafaela Ventura, a veteran factory worker at Vanguard Plastics, first met Baxter, she found him to be “interesting.” Rethink brought him to the plant to see how he could handle some of the tasks, and while it was at first strange to see a robot do the same work she and her co-workers did, in the end she says she’s happy to have him on the team. “I’m glad because we can work on something else. It’ll give people a break from the [assembly line]. I’d rather have Baxter do it.”
Eckert says Ventura’s attitude is not uncommon. Many current factory workers look forward to learning more skills and moving up the ranks, leaving tedious assembly-line work to a machine. Chris Harbert, a sales manager at Rethink Robotics, sees Baxter as empowering workers. “Humans aren’t built to do repetitive motions over and over again,” he says. “They’re built to think creatively.” But when Ventura was asked if she’s worried about Baxter replacing her, her tone changed. She hesitated for a moment. “Yes, if they bring in more Baxters, maybe some people will lose their jobs, even me. But right now I don’t know. I try not to think about it.”
Budnick says he does hope to purchase more Baxters for the factory if all goes well with Baxter #1. For Ventura and others to keep their jobs, they’re going to have to learn how to manage a row of Baxters on an assembly line, or maybe even repair Baxter if anything goes wrong – something that may require engineering or computer science training.
“As a society, we have an obligation to educate the kids who are in school now and retrain people already in the workforce, so they can contribute effectively to the new economy,” says Seth Teller, a robotics researcher at MIT’s Computer Science and Artificial Intelligence Lab. “To just do technical innovation and not have massive resources going to improving skills is a mistake, and it’s a mistake that we seem to be making now.”
Who might the teachers of the future be? Ask Baxter.