Last month, President Barack Obama announced the National Robotics Initiative, a major program to develop next-generation robots for manufacturing, healthcare, and other areas. The robotics community received the new initiative with enthusiasm, but some observers expressed concern about an expansion in automation, raising a perennial question in robotics: Do robots take people’s jobs?
“The real purpose of automating manufacturing is to eliminate skilled workers and replace them with low paid button pushers—preferably offshore,” commented one IEEE Spectrum reader who’s worked as a control engineer for 25 years. Said another: "As jobs at all levels, from McDonald's to college-educated knowledge-workers, are increasingly automated, there will be more unemployment." Other readers voiced similar concerns.
To hear what the pro-robots camp has to say, I spoke to John Dulchinos [photo, right], president and CEO of Adept Technology, the largest U.S. industrial robotics company. Adept, based in Pleasanton, Calif., offers a variety of robotics products, including SCARA, parallel, linear, and mobile robots. Dulchinos, a mechanical engineer by training, says he became interested in robots during college. “A publication by IEEE got me into robotics,” he says. “It talked about the personal robotics revolution, how it was going to be bigger than the computer industry, and I said, I want to go into robots.”
Dulchinos says that automation, though it might take some people’s jobs in the short term, is essential for keeping companies competitive, and thus able to expand and hire more workers. That's why more and more companies in industries as varied as food packaging and electronics manufacturing are embracing robots.
“If you look out far enough, machines are going to win," he says. "The human body is not a machine. It wears out. It was not designed to be a factory machine. It was designed to be a thinking machine."
Dulchinos believes that "robotics is going to be one of the transformative technologies of the 21st century." The entire robotics industry is only a 5 billion dollar market today, he says, but according to some estimates it will grow to 100 billion by 2020. He envisions future domestic robots helping people at home and factory robots that are not job takers but rather robotic assistants that work alongside human workers. (Just don't let a robot borrow your iPhone.)
Read below my full interview with Dulchinos, in which he discusses how countries like Germany and China are using robots to improve manufacturing, and how a new generation of smart factory robots could do the same in the United States.
Erico Guizzo: Let’s start with a question that comes up again and again: Do robots take people’s jobs?
John Dulchinos: Robots are not the enemy. It’s low-cost labor that’s the enemy. If you want to look at where jobs are going, it’s not robots taking people’s jobs; it’s entire companies and industries moving overseas. Robots elicit an emotional response from people because they are personified as people, but really what robots do is they enhance productivity and they free people up to do other tasks. In a global economy where cost rules, the only way for Western countries to be able to compete effectively against low-cost labor markets is through productivity gains, and robots are one way to achieve that.
Let me give you some background: In the last 15 years the United States has lost somewhere between 2 and 3 million jobs in manufacturing. And in that period of time, China has grown to surpass every other country now except the United States in total manufacturing capacity; in fact, in the next year or two, China is expected to surpass even the United States. Germany has actually grown its manufacturing population. Germany and Japan have the highest density of robots [number of industrial robots per 10,000 manufacturing workers]. And Germany has used robots to grow their manufacturing employee base, because they’ve been able to be competitive and bring manufacturing plants back to Germany. With that comes not only the manufacturing jobs but all the other indirect jobs as well.
So do robots take jobs away? I’m asked that question many times. I would draw some comparisons to other industries. In 1900, about 50 percent of the U.S. population was in farming. Today it’s less than 5 percent. Yet our output is far greater than it was in 1900. And while you’d argue that tractors and cotton mills and other machines eliminated a number of farming jobs, the reality is that the mechanization enhanced the productivity of farming to a point where people went into manufacturing, into engineering, into a variety of industries that spun the industrial age and then the information age in the United States. Had that mechanization not occurred in farming, we wouldn’t have much of the advances we had in the 20th century, because we’d all be on the farms producing food.
EG: Still, some argue that companies can use robots to become more productive and even expand, but in the end they’ll need less workers…
JD: Let me ask you something: Do computers take jobs?
EG: That’s a good question. There was a big debate in the 1980s and ’90s about computers replacing office workers and whether or not productivity was increasing.
JD: What happened to all the secretaries in the 1970s and 1980s? Where are they today? If you follow this logic that if a specific task is automated that means a person now goes to the unemployment line, then we would have tens or hundreds of thousands of secretaries in the unemployment line. And they are not. They've upgraded their skills, are using computers, and learned to do more complex tasks as opposed to being relegated to menial tasks. Computers have been quite an enabler. And robots are just a computer with an arm connected to it.
I can put it in numbers for you. Over the last 15 years, there’s probably been about 10,000 to 15,000 industrial robots deployed a year in the United States. So that's around 100,000 robots deployed from 1995 to 2010. If each robot replaced two people, which I think would be a stretch, but let’s say it did, that would be 200,000 people, or maybe if you want to be really aggressive, 400,000 or 500,000 people. We’ve actually lost millions of jobs in manufacturing over the last 15 years. Really what happened is, manufacturing is going away because the United State isn’t competitive in the global economy anymore.
And I’ll give you a scary thought. China last year grew to become the No. 4 market in industrial robots. In the next two years it will likely pass the United States in number of robots installed every year. And so if you want to extrapolate that, you'll realize that China, which has the lowest labor costs in the industrialized world, is putting in robots at a pace that is on par with the United States and soon will be faster than the United States. To me, that’s putting the Chinese manufacturing economy on steroids.
"Move over, Dulchinos, I want to be CEO."
EG: Government agencies and companies have invested in robotics R&D for decades, but things seem to move so slowly. Why will this new National Robotics Initiative be able to help now? What’s different?
JD: Robots were pioneered in the United States 50 years ago. What happened is the United States didn’t embrace them. Japan did. Look at the significant gains that Japan made in the 1970s and ’80s in manufacturing. Those were all fueled by the creation of a very strong and vibrant domestic robotics industry.
What’s different about now is that robotic technology has gotten to a point in the research lab where there’s the potential for a new generation of smarter, more flexible, safer robots. This new generation of machines promises to expand the applications of robotics not only in manufacturing but also in healthcare, military, and domestic applications.
That has the potential to shake up the industry and create an inflection point where the United States can compete with Japan, Korea, Europe, and China—all of which are all spending much more money on robotics R&D than the United States—and gain a leadership position in next-generation robotics to enhance global competitiveness across a host of industries.
So if you look at who has the potential to win in the 21st century robotics race, government funding will be a factor in that race. It’s not the only one, for sure; in fact, I think the United States has better foundational technology to win that race, but it can’t without some help to try to level the playing field. The Obama initiative, which is a beginning, is a way to bring some focus and value that can push the U.S. robotics industry forward, and give companies a chance to compete fairly in the global economy.
EG: What are some technology barriers that you think should be attacked?
JD: It starts at the sensing level. A robot is a programmable machine, but it’s a dumb machine if it can’t be aware of its environment and adapt to its environment. Sensing technologies like vision, which gives the robot the ability to perceive and see things around it, need to improve a lot. So does tactile sensing, so robots can feel objects in the same way or better than a person. And we need not only better sensors, we need to make them cheaper, so we can use them in applications that aren’t cost effective today.
Another area for improvement is collaborative robots. Robots that can work safely alongside people. The robot of the future is not a job taker; it’s a job assistant. So think about what a robot is going to do in the future. You’ll have, in a production facility, people and robots intermixed together doing tasks. Robots are going to allow people to be more effective. You see this already in surgical robots. These robots are not taking doctors' jobs away—they’re making doctors more efficient, enabling them to perform a surgery with a higher level of precision and consistency.
Lastly, we also need new materials, new motor technologies, and safety strategies. I think that once robots reach a certain cost point, you'll have a robot in your home, doing laundry for you, helping with vacuuming and cleaning, or imagine robots at hospitals delivering supplies to the OR. These applications require robots that are smarter, more adaptable to their environment, more able to work alongside people and assist people. I think those are the technologies that could come out of a more focused effort in terms of R&D in robotics.
China is putting in robots at a pace that is on par and soon will be faster than the United States. To me, that’s putting the Chinese manufacturing economy on steroids.
EG: Companies like yours do invest in R&D, I’m sure, and here comes the government funding universities and other businesses that may become your competitor. Aren’t you upset?
JD: Yes, but I’m more upset that Japan has been subsidizing my Japanese competitors and Korea is subsidizing my Korean competitors. I will give you an example. This is how it plays out for a company like mine. We manufacture robots all around the world. We produce some of them here, some in Japan, some in Europe. And the motors that I use, the best motors for my SCARA robots are made in Japan. I am at a competitive disadvantage in the purchase of those motors because I am a U.S. company; all of my Japanese competitors get much better pricing on those motors because they’re in the Japanese community. So I’m competing on unlevel playing field, and the reason that there’s a lot of competition coming from places like Japan is for two reasons. One is Japan was a big market for robots many years ago, and second there was government investment.
So what I think the government can do—and I would say that the last thing that this industry needs is a handout from the government—what I’m looking for the Obama administration and the government to help with is to create a robotics domestic market here. We’re looking to build a coordinated activity that allows us to take some of the barriers down. Another barrier is regulation. To allow people to create a new generation robotics company and don’t have to worry about fighting a lot of regulatory tape or being sued for early applications, we need to take down some of the regulatory barriers. One of the big areas is to define the environment for how robots and people will work together. Robots get held to a completely different standard than other machinery in terms of safety. The United States could take a leadership role in defining how robots and people can coexist in a safe manner and enable this assistive robotics environment to expand.
It goes all the way to cars. Google demonstrated earlier this year that they drove a robotic car over 100,000 miles and that allowed them to prove that an autonomous vehicle has potential to be in highways. There’s 40,000 people a year who die in car accidents, and that could be significantly reduced if we automated the highways and have robotic cars driving. Paving the way for that kind of technology is, again, one example of how the government could help.
EG: What’s the most exciting emerging market for Adept? And how do you see manufacturing robots evolving?
JD: I think packaging is a very exciting area. The jobs are challenging, margins are very thin, throughputs are very high. A typical poultry plant might have 50 percent turnover on an annual basis. You think people like those jobs? You think people are lining up for those jobs? We see jobs where a person is picking up 10-pound cylinders of meat and putting into a box, 30 to 40 times a minute, eight hours a day. You go pick up a 10-pound weight and put it into a box and see how long you last. The other thing that’s driving that market is the crackdown on immigration. Those jobs are all being filled by illegal aliens. And why do you think that is? Because you can’t get people to do those jobs anymore; they hate them.
Another exciting area is electronics. The robots we sell go into disk drive factories, for example. When you’re putting together a disk drive, one of the biggest enemies is contamination. The biggest source of contamination is people. People are dirty. When a disk driver manufacturer decides to automate a factory, it’s not because they want to get rid of labor per se, it’s because people can’t do the job at the level of cleanliness and yield that is required to be competitive in that space.
EG: What about more complex electronics? We hear about problems involving workers assembling iPhones in high volume. Will robots be able to fully assemble a smartphone?
JD: There are certain jobs that require dexterity at a level that robots don’t have yet. But they’re getting better and they’ll get there. The question is: What do we need to do to have high-tech electronics manufactured here in the United States? Let me tell you a story about cellphones. In the mid-1990s, one of Adept’s largest markets was actually selling robots to cellphone manufacturers. That was when the market was a couple hundred million cellphones sold a year. Fast forward 15 years later. There isn’t s a single cellphone built in the United States anymore. And I can point to those old factories—all those people are gone. All the production is sitting in China right now, done largely by a mix of Chinese labor and automation.
EG: All the robots you sold to cellphone manufacturers in the 1990s are not being used?
JD: No. The problem was that robots weren’t flexible enough to be able to automate those factories. The product lifecycles were so short that with the level of technology in the mid-1990s, it was difficult to automate those lines with robotics at a level that would be cost justifiable. So that’s a great example of where, had robots been capable at the time, we would have a cellphone manufacturing industry today in the Untied States. But because robots weren’t capable enough, the market went to low-cost labor in China. And now China is starting to use robots to automate their factories. China owns the industry, and no one is going to be able to take it back.
So how can robots help moving forward? What robots enable, and this is what’s exciting about them, is that if you can build a domestic market, robots can help equalize the manufacturing costs anywhere in the world. And then what really matters is infrastructure costs and what’s the regulatory, tax, and business environment and where’s the market. And that’s what determines where you’re going to build a factory; not just because the labor costs are cheaper.
So I think we’re at a very exciting point. This is a chance over the next decade for the United States to regain leadership in robotics. And with a little bit of focus from the government to help push some technologies forward and push the regulatory environment in the right direction, I think that the United States can win this race. Success 10 years from now will be that there’s a number of U.S. robotics companies and they’re leaders in creating next-generation robots that permeate our everyday lives.
This interview has been edited and condensed.
Erico Guizzo is the digital product manager at IEEE Spectrum. He oversees the operation, integration, and new feature development for all digital properties and platforms, including the Spectrum website, newsletters, CMS, editorial workflow systems, and analytics and AI tools. He’s the cofounder of the IEEE Robots Guide, an award-winning interactive site about robotics. An IEEE Member, he is an electrical engineer by training and has a master’s degree in science writing from MIT.