The Internet of (Robot) Things

Not many people understood the computer science and artificial intelligence research Tom Wagner ’00PhD did while studying at UMass Amherst. “When you’re doing doctorate research, you’re usually focusing on an unusual or novel thing,” Wagner points out. “But then the world evolves.” Since his time at UMass, Wagner has become a pioneer in a booming robotics and artificial intelligence industry. The projects he leads have brought robots into realms never before imagined.

A decade ago, Wagner was the chief technical officer of iRobot, the company that brought you the Roomba Robot Vacuum. The company also created innovative robots that save lives. iRobot’s PackBots were used by the U.S. military to hunt improvised explosive devices (a menace that has killed or maimed thousands of U.S. troops in Iraq and Afghanistan). In the aftermath of the Fukushima Daiichi nuclear plant disaster in Japan, the company’s robots were deployed to explore areas that were too dangerous for humans.

Toy robot standing

Wagner left iRobot in 2012 to become the founder and CEO of the Boston-based firm Berkshire Grey which automates menial work in industrial environments in the name of creating more opportunity for humans. Wagner has spent his career thinking about how technology, robots, and artificial intelligence can make our lives better—and in some ways they already have.

Wagner is driven to invent practical technology that delivers value for customers and he says that often times we don’t know we need a technology until it comes along. Consider something as commonplace as photo sharing apps. “In the past you’d have to print the photo, type the letter, address hundreds of envelopes, and send all those things out to share a photo with friends,” Wagner says. None of us could conceive of how easy sharing a photo would become. That’s Wagner’s job—to imagine what most of us can’t. We asked him to imagine what the near future with robots might look like.

First, robots need to smarten up

“When you see a robot painting a car, that robot isn’t thinking. Conceptually, you could pull the car out and it would paint the air.” In order to create what Wagner calls “an impactful way to use robotic systems,” he says robots themselves need to get smarter—and the robotic systems he’s creating for retailers and e-commerce companies are moving in that direction. “We put artificial intelligence into robots to enable them to do tasks that are too hard for conventional, unintelligent robotic automation,” he explains. Robots in shipping warehouses can identify, pick, and place the bottle of shampoo you ordered, select the right size box for it, even organize and pack items. This new technology makes use of computer vision, sensors, novel grippers, and machine learning to automate tasks that have never before been performed by robots in commercial settings. Berkshire Grey’s portfolio of robotic solutions pick objects up, pack them in boxes, and ship them out the door to both consumers and stores—all of which saves companies big bucks. “When someone rings your doorbell two days later,” Wagner says, “that package you ordered online may have been handled by one of our robotic systems.” It might simply be a cool concept for consumers, but for small and large e-commerce businesses, it’s a lifesaver—it allows them to compete with juggernauts like Amazon.

Robots have a lot more thinking to do

“What you will see over the next 10 years are more robotic systems designed to solve a very specific problem—mow your grass, for example, or make a cup of coffee exactly as you want it,” Wagner predicts. Some tasks will be too hard for full automation and you’ll see these specialized robots working with people, e.g., “A pizza delivery robot knows a little about roads and sidewalks and front doors, but they’ll be much more challenged by things like apartment buildings, steps, door knobs, and buzzers. You’ll see pizza delivery robots working aside people doing pizza delivery or we may get used to going to the street to pick up our pizza. The easier and more manageable things will come first.”

Robots: a limited (but useful) workforce

“Our customers have thousands of positions they can’t fill,” he says. When robotic systems step into those unfilled positions, companies can survive and thrive—and provide jobs to living, breathing humans. Certain things are better handled by people. If you expect a high-end consumer good to be beautifully presented in a box, for example, that goes to a person who does the delicate, artistic, thoughtful work. We don’t automate that.” At least, not yet.

We’re on the lookout

Share your most intriguing nooks, niches, coordinates, or curiosities on campus or anywhere in the region. Email magazine@umass.edu and we’ll investigate!