Driving Artificial Intelligence and Robotics on the Factory Floor
Rodney Brooks wasn’t willing in 2002 to hazard a guess about when the challenges of robotics would be overcome. “Not in two years or three years,” Brooks asserted during an IndustryWeek interview that year. “But is it going to be here in 30 years or 40 years? I’m not quite prepared [to make a prediction].”
Still in a book published then and the interview, he gave us a peek at what it would look like when it arrived: “robots that take instruction easily rather than ones that require hours of programming time; robots that can help in small-batch operations rather than ones that only make financial sense in continuous or nearly continuous fixed operation settings with long runs; robots with social interaction skills;” and robots that reduce costs on the factory floor.
When Brooks and his company—he’s founder and CTO of Rethink Robotics—introduced Baxter in 2012, he’d delivered on every specification. Last year, with the introduction of Sawyer, he delivered on still yet another important specification--the ability to handle tasks that require more precision, including machine tending and circuit-board testing.
In doing so, he has done more than any other individual to pioneer and popularize collaborative robots, or cobots, that can safely work alongside people on a factory assembly line. As well, he’s made possible the next generation of automation in factory operations and, with the low-cost of the robots, made practical the use of robots in small and midsize manufacturing plants.
That’s why IW editors have named Brooks the first recipient of the IW Manufacturing Technology Leader, one of a new series of IW Industry Excellence Awards. They think that Baxter and Sawyer are critical combinations of technologies uniquely suited to help rejuvenate the U.S. manufacturing sector.
As important as the functionality the robots make available to manufacturers, the arrival of robots, among other technologies, has business and public policy leaders taking a second look at the manufacturing sector, which had been all but written off as dying the past two decades. During that time, the manufacturing sector struggled for attention as information technology, software and social companies captured the spotlight.
For Brooks, however, the jury’s still out as to whether the robots will help spur a U.S. manufacturing revival. “I hope so. I hope so,” he says, when asked. “It’s still early.”
A Research Scientist Solves Practical Problems on the Factory Floor
Still he’s up to the task of touring manufacturing plants and bringing sci-fi artificial intelligence to life in the service of solving manufacturers’ problems. Indeed, it was his close involvement in manufacturing that led him to make Baxter and Sawyer what they are. While working with contract manufacturers at offshore locations and as a member of the Technology Advisory Board of John Deere, he learned of the impending manufacturing labor shortage; while touring automobile plants, he noted the divide between where robots and humans tended different tasks. Together, these facts “led me to think that we had to do something different in manufacturing, that there would be a demand for something different in manufacturing coming down the line.” That is, robots that work collaboratively with people.
Brooks’ research and role at the company perfectly suits his “boyhood fascination” with robots. Brooks recalls that he started building robots when he was six or seven and, because he imagined the robots needed to be smart, he decided to study artificial intelligence. Ultimately, however, he always wanted to build “things that are physical.”
The physical manifestation [of artificial intelligence] is actually important."— Rodney Brooks, founder and CTO, Rethink Robotics
“The physical manifestation [of artificial intelligence] is actually important,” he declares. “You know, I think we’ve become computational chauvinists, where we think computation can do everything,” Brooks muses during an interview. “But the physical stuff is really important. There are laws of physics you can’t get around.” For example,he notes that physical wings provide a lot of the things necessary for flight, “the particular shape of them and the air going over the wings provides upward lift.”
He continues: “You can’t imagine flying just through computation, so likewise with robots, you can’t actually do stuff in the world just through computation.”
It’s this technological mix, “the computational with the physical,” that guides Brooks’ work and that directs deep scientific thinking toward making useful things.“I’ve had aspirations to understand intelligence generally, as a researcher, aspirations to understand what it is that’ll make things alive verses things that are not alive, and those are very deep scientific questions,” Brooks says. “But what I’ve done along the way is take what little we do know and engineered something that’s useful in the world.”
Which is an understatement. Brooks’ contributions parallel the history of robotics, from early work developing robots that move, which was adopted by NASA and included in the Sojourner, the first robot on Mars; to robots that clean up mines, which evolved to become the Roomba, the self-navigating vacuum cleaner, and the PackBot, used in Iraq and Afghanistan to find IEDs; and then to Baxter and Sawyer, robots that do real work in factories.
Finding the Value Pathways through Incremental Adoption of Visionary Technology
Amidst his own success and the hype around artificial intelligence, virtual reality and the Internet of Things, Brooks maintains his practical outlook. Indeed, he thinks many in the press and business community may be a little overenthusiastic about how soon many of these new technologies will have an impact.
About the Internet of Things, for example, “I think that it’s great that we have things like industry 4.0, that we have China 2020. These are visions of the future,” Brooks says. “But just having that vision doesn’t mean that people are going to go out and adopt that whole thing all at once.”
There have to be pathways, where you do a little bit, you get some benefit; you do a little bit more, you get some value, and it doesn’t screw things up along the way."— Rodney Brooks
“There have to be pathways, where you do a little bit, you get some benefit; you do a little bit more, you get some value, and it doesn’t screw things up along the way. That’s a bit of what we’re missing now: How to get the incremental value through incremental adoption.”
He adds, however, “the people who figure out how to use those trends effectively, not foolishly, to increase their productivity, increase their capabilities and increase their quality are the ones who are going to win out in the long term.”
As for Brooks’ next step, he’s spending a lot of time with customers, listening and looking for ways to “reduce the friction of putting a robot in a place they’d like to put it but it’s too much work, too much third-party engineering.” He continues, “How do we make it so it’s more like a consumer device that you can just take out of the box and it just works?” Those ideas are already in the company’s roadmaps for new software and hardware releases.
Looking further ahead, “I was in a couple factories less than two weeks ago, which led me to a whole bunch of new thoughts that I’m writing up and doing some analysis,” says Brooks. “It’s the early stage, way before it gets into the product roadmaps.”
One thing seems certain. Manufacturers can count on Brooks to continue to blend the possible with the practical to bring the best of artificial intelligence and robotics to manufacturing.
