IRobot’s Colin Angle on the Home Run of Robotics (It’s Not A.I.)

Xconomy Boston — 

When it comes to consumer robotics, iRobot has to be the poster child. It was born 27 years ago out of MIT, founded by legendary roboticist Rodney Brooks, now professor emeritus, and two of his students, Helen Greiner and Colin Angle. Over the years, the company has made many forays into various products—including the My Real Baby doll on the consumer side, but also bomb-sniffing robots for the military. It found its groove, though, in the seemingly ubiquitous Roomba robot vacuum cleaner, its mainstay product.

Right now, the Bedford, MA-based company (NASDAQ: IRBT) is on a pretty nice roll. Revenue for the first three quarters of the year soared to $557 million, up 24 percent from $448 million in the same period a year earlier, with profits rising 64 percent to $46 million. IRobot’s stock price is around $68 a share, up from $52 a year ago—though well down from its peak of over $107 per share this summer. But the good times haven’t come easy. A lot of hard lessons have been learned, and the company has come a long way since its original business model—“private mission to the moon, sell the movie rights,” in the words of Angle.

In recent years, iRobot has shifted dramatically. It shed its defense and enterprise-remote presence businesses (the latter included its healthcare telepresence operations) last year to focus on the home market—led by the Roomba—and in 2015 launched a venture group to help robotics entrepreneurs grow the industry (it’s made about a dozen investments to date). More recently, it has set its sights on how robots can help enable and fulfill the vision of the smart home.

Success with this grander vision won’t come easy either. I recently sat down with Angle, who’s been CEO since 1997—and is the only one of the three founders still working at the company. (Greiner and Brooks both have their own startups, drone-maker CyPhy Works and industrial robot company Rethink Robotics, respectively.) Angle says innovation in robotics is like working in “anti-dog years,” meaning what you thought you could do in one year actually takes seven.

Among other things, we talked about lessons learned, his vision of the smart home and how to get there, and the ultimate home run of robotics—“helping people stay in their home as they age and maintain the lifestyle advantages of living at home.” And that means, he says, “We’re going to need a lot of robots.”

Here is an edited version of our conversation.

Xconomy: Why don’t we start with an overview of where you are with iRobot and where things are headed.

Colin Angle: IRobot has the last two years really focused on the consumer. We’ve divested our defense business, we’ve spun out commercial activities, and are laser focused on the home market. How are devices and robots going to evolve over time? What’s the interface and synergy between robotics and the smartphone? And we’ve pretty much obsoleted the upright vacuum cleaner. Twenty percent of all money going to vacuuming is going to robot vacuuming, and that’s been really exciting to see that first great mainstreaming.

X: Was the Roomba the origin story of the bigger smart home focus?

CA: Yes. The bigger smart home focus is enabled by the fact that the installed base of connected mapping robots is a reality. Millions of maps are being created every week of where the robot’s going, and that creates opportunities over time for homes to get smarter about what they’re supposed to do—and the Roomba is central to that future.

X: Can you elaborate on where you see the smart home going, and how that fits what you’re doing at iRobot?

CA: The challenge with the current smart home is that the home doesn’t understand enough about itself to actually be smart. We talk about the smart home, but really what we’re talking about is a cell phone-controlled piece of electronics–whether there be a light bulb you can turn on with your cell phone, or a toaster oven, or a thermostat. There’s some convenience there, but not a ton, because turning on your lights with the cell phone is not a huge step forward from pushing a light switch. In fact, many people would say it’s a step in the wrong direction because I have to spend 20 seconds to turn on a light bulb [with a phone] when I can do it in half a second with a switch. So what is needed is really a compelling idea about what the smart home is supposed to be—and then how do we get there.

So this idea that I should just live my life, and the home should do the right thing, that’s compelling. I’m not turning things on and off. I walk into a room, the lights go on, the heating adjusts appropriately. When I leave for more than a certain amount of time, maybe that room shuts down to save me energy. That’s a home operating in an intelligent fashion. And we haven’t even gotten into more exciting goods and services that your home can do for you.

But if that’s the vision, it begs some real questions. If I’m going to turn on the lights in a room, I need to know what a room is. It’s really that simple. Our homes don’t understand anything about even what’s in them. And so how are you going to control it? You could say well, we’ll just program it. And that sounds good, other than the fact that we couldn’t program the clock on a VCR. We haven’t suddenly got smarter. So either I hire someone to do it for me, which is expensive, or it has to happen automatically. And that’s where robots come in.

We now have robots that can drive around your house building an understanding of what they’ve done. That’s game-changing, because we can start talking about automatically generated data sets that include a concept of a room [and] what’s in that room, so that homes can program themselves. You couple that knowledge [with] where am I, and then [Angle snaps his fingers], your home becomes smart. Because you can have a set of preferences as to what you want to have happen, or your home can just say ‘Oh, whenever Colin is sitting in the living room the following things happen—so why don’t I just do it automatically, and he can tell me if I’m wrong and I’ll learn very quickly what I’m supposed to do.’ That, I think, is going to be the next step in building a smart home.

X: What’s been missed up until this point when people think about smart homes?

CA: They don’t approach the smart home by thinking of the home as a robot. They say, ‘I’m going to connect everything, I’m going to create a voice interface to the home’ without really thinking of what that means. Really cool products like Alexa and Google Home exist, but they weren’t originally conceived of as the enabler of the smart home. It was a voice interface to the Internet—and if I could go and talk to the Internet as opposed to browsing the Internet, that might lead to some interesting places.

Somewhere along the way people started thinking, ‘Hey maybe this is the answer to the smart home,’ when it’s really only part of the answer. As a robot guy who tries to figure out how to make things work automatically, context and understanding of the local environment goes hand-in-hand with artificial intelligence. We’ve watched A.I. rocket past robotics in very real terms by constraining the world that the A.I. acts on to the digital space. Whereas we poor robot guys are saying, ‘I’d really love to be working on how I make a robot go to the kitchen and bring me a beer, but I really haven’t figured out where the kitchen is.’

X:  Little things like that.

CA: Little things like that—and so I’m stymied. But the exciting thing in robotics is that we’re actually rolling out solutions to understanding the environment so that these higher-level tasks are suddenly more within our grasp. And that’s amazingly exciting. It’s only taken 27 years to go from where we started this at iRobot to where there is hope that within a very short amount of time we’ll be able to understand where the kitchen is.

X: Way to go.

CA: Yep, yep. It takes thick skin and patience and a fair amount of dedication to be in this industry, but there is definitely an acceleration that’s happening both because of the commercial success of home robotics and the advent of this technology to allow us to understand our environment enough to do more stuff.

X: So probably therein is the answer to how you differentiate iRobot from the things being done by big players like Google and Amazon. It seems like you see them more as potential partners than adversaries.

CA: Oh, absolutely. Those companies’ ability to have speaker-independent voice recognition, rich interfaces that allow you to ask their systems anything and get anywhere from a minimally satisfying to a completely satisfying result, requires tremendous investment and resources. Where iRobot comes in is we can give you the spatial information required for a system like that to do the right thing. And if you tell Alexa to turn on the lights in the kitchen and it doesn’t know what the kitchen is, nor does it know what lights are in the kitchen, it doesn’t know what to do. If we can help with that, that’s a great partnering opportunity.

X: That in some ways would mean you’re a data company, which clearly you are, but you’re also building robots.

CA: The data that is collected enables the owner of the Roomba to make their homes smarter by allowing some type of data sharing with an Alexa. In the future—it doesn’t exist today. You can turn on your Roomba with Alexa, but can’t do much more than that today. So there is a real data dimension to iRobot which is growing very, very rapidly. It’s something that differentiates our products in the marketplace. We think that our future is quite simple: there are multiple robots in everyone’s future. There’s this inherent promise in robotics to make our lives easier and so if we can build up this ecosystem of robots in the home—we’ve got vacuuming and mopping and are hard at work at other things—we will be very successful. With the information robots are collecting in order to work better—with permission of the homeowner—there’s opportunities to make other dimensions of the smart home work better as well.

X: So can you give us some quick stats on the Roomba and mopping—and then share a little bit about what you’re working on?

CA: Roomba is 85-90 percent of our revenue and is growing quite rapidly. In 2017, we expect sales of Roomba in North America to increase by over 40 percent. It is really quite wonderful. Last year, our entry level Roomba was the top-selling vacuum cleaner in the United States. So that is a rocket ship. The mopping category—our Braava line of robots—is starting from a much lower base, but it grew by even a larger percentage in North America [although] it’s not going to displace Roomba any time soon. But here is further example that if you get the function right, and at a price people can afford, there is an excitement to bring robots into the home.

As you think about the future, there’s actually two dimensions. There’s what are the next products? When am I going to get a robot that folds laundry? When are we going to have robots that mow lawns—things like that. We will see robots like that over time—there are different challenges, [like] having dexterous manipulation. But we’re past the point of, ‘Are home robots in the future, or is this something real gathering momentum?’ It absolutely is. The domestic appliance industry is going to go through a tremendous disruption as people start to expect more from their home appliances. I’m a little disappointed that we don’t [yet] see a greater diversity of products from the rest of the consumer electronics world. But so be it.

X: Is there anything out there that you wish you would have done?

CA: Not yet. If someone comes out with a laundry-folding robot, I’ll wish I had done that. There’s one company that has a big machine to do it. I’m not sure they’re at mass market price points. And Franklin Robotics has a neat gardening robot [Editor: It’s called Tertill] that’s a cool beginning of outdoor robotics. But home maintenance is full of these treadmill tasks where we just have to do them every day. Every one of them is an opportunity for robotics and automation.

X: So for iRobot, is the smart home right now the interior of the home as opposed to the grounds?

CA: We certainly look holistically at the home. Lawn-mowing is something we’ve talked about as being an area of interest. The challenge is what are the technological barriers that stand between idea and product. Robotics is tremendously exciting and tremendously frustrating at the same time, because the pace of new product introduction has never been satisfying. I joke that give me 10 minutes and a blank piece of paper, and I’ll write down a lifetime of projects for an aspiring roboticist. But at the same time it is moving, and every year we see more good robot companies emerging. And as we see successes and dollars entering the space, that’s a spinning up of the flywheel. But we need innovation on the ‘go do hard, physical work’ side in order for the industry to take its next steps.

X: You’ve been open about your failed business models and products—and you even put failed products in the museum near your lobby. Do you have any key insights from these to share?

CA: Certainly that innovation and failure come hand-in-hand. If you’re not failing and succeeding, you’re not trying. Being an entrepreneur in general, and probably a robotics entrepreneur in specific, requires tremendous optimism and patience. We’re sitting in my office and I’m surrounded by images of 50 robots that we’ve tried throughout the years, and the vast majority are not Roombas. So every time you build something, you learn something about this unforgiving field.

Our first business model—private mission to the moon, sell the movie rights—led to the growth and launching of micro-rovers like Spirit and Opportunity and Sojourner. Using robots to defuse bombs for the military was, ‘Hey, that’s a bad idea to have people doing this. Why can’t we build a robot to do it?’ So we did. The halls of iRobot, of our museum, are full of these things that started off as crazy ideas or Lego models, and this confidence that maybe there’s a better way being played through to practical reality.

X: Let me switch gears and get a couple quick takes. Is there any update on your venture fund you can share?

CA: IRobot continues to be very committed to investing in early-stage startups in the robotics and IoT areas. We actually have a pretty good track record of success and are quite pleased with the returns and the potential of the companies that we have put money into. It started off as an experiment—and at this point we view it as a valuable part of iRobot’s activities, particularly as it relates to both innovation and supporting this growing industry.

X: Are you doing it as a strategic investment where you think it will help your future business, or are you sometimes just helping a young company?

CA: If iRobot is to be involved in your company, it has to be in an area that we view as of interest to iRobot. So we wouldn’t invest in an industrial robotics company unless we thought there was a really cool manipulation technology that could one day be useful in the consumer space. But we’ll also invest in companies that might have a strategic technology that enables home understanding, even if its applications were broader than just enabling robotics. So we try to be thematically relevant to the company, and not just because we think if they succeed they become an acquisition target. It’s mostly because we understand the places that we play. And if we think you have a technology that’s going to be important, well then maybe we have some insight that suggests you’re on the right track. We could influence and help that company make good decisions.

X: How do you see the international market for home robotics? Especially China, which seems to be a challenge.

CA: Well, the international market for robotics is lagging, but chasing the U.S. market. The growth rates are up in all markets. We’re seeing good traction in Japan and Europe. We’re seeing great growth and traction in China. In the China market, we actually had a reduction in revenues but 20 percent growth in sell through. We got caught a little flat-footed at the end of last year, beginning of this year, on what the market wanted from iRobot and there’s some major competitors that have been really successful at the low to mid-range, which gave us some headaches on our inventory strategy. But the market itself is growing well, and our sell through is growing well.

X: So quickly, are there are a few takes on hype versus reality in this field?

CA: Oh my god. Hype [and] reality in robotics go hand in hand. The promise of robotics is so profound, and starts at such a young age, that the reality of robotics can never stay in step. Every two-year-old is passionate about whales, dinosaurs, and robots. And from that early age our mental imagining of what the robotics industry is supposed to be is so exciting, and the challenge of robotics is so hard, that they’re just nearly incompatible. I’m guilty of this, too. If you asked me in 1990 when iRobot was founded, ‘OK Colin. Twenty-seven years have passed, and you are a successful public company. What do you build?’ There’s no way I would have said 90 percent of my revenue is coming from the vacuuming industry. I think I would have said, ‘Well, we are a diversified robot company with androids.’ It would just be completely different than reality.

You know, I’m depressed and excited at the same time. There’s just this massive disconnect between our fears and worries and expectations of robots versus the reality of what’s happening.

X: Where do you come down on whether robotics takes away jobs, or adds jobs?

CA: Robotics changes the world, just like the dishwasher changed the world and autonomous cars could change the world. It means that some very, very hard problems now have solutions. We tend to jump to ‘Oh my, am I going to lose my job?’ without thinking of the fact that unless we can extend our ability to live independently at home, we are going to see a tangible reduction in the standard of living as our society gets older—because there just are not enough young people to take care of the old people.  We need to find solutions. The home run of robotics is not vacuuming. The home run of robotics in my mind is helping people stay in their home as they age and maintain the lifestyle advantages of living at home. And thus, we’re going to need a lot of robots.

It’s going to be an exciting industry, and it is going to employ a ton of people. ‘N’ years from now, I don’t know if it’s 20 or 50 or 100, you just won’t be able to imagine how life was possible without robots.

X: Well, let’s try again. Twenty-seven years from now, what’s iRobot going to be?

CA: It’s going to be less diverse and more important. Twenty-seven isn’t that long, first off. Not in robot years, which are worse than human years—anti-dog years. We’re going to have multiple robots in our homes that maintain the home and deliver services into the home such that our living spaces can be to a large extent self-managing.

X: You’re talking about physical robots?

CA: Physical robots. Different sizes. There’ll be a robot that you interact with, robots that go and do physical work, and your home will also be quite robotic. It will be a system that you buy in parts, that network together, and give you a self-maintaining experience. Homes will probably be much smaller than they are today. You could have rooms that reconfigure. There’s cool startups today that are working on that eventuality already.

When you talk about what is the robot future, you have to also have an idea of what the future is going to look like. The current generation lives a significant part of their existence in the virtual world. So the home 27 years from now is going to blur real and virtual quite seamlessly. The robot dimension of that needs to play into that future, but the robot’s job in this wild and crazy gray area is to keep the physical part of your life organized and managed. You’re still going to have to vacuum the floor. You’re still going to have to get your food from wherever it’s prepared to wherever you want to eat, even if you’re too frail to be able to move it there yourself. The physicality of robots is going to become more and more important as we get older. We’ll be increasingly dependent on robots to deliver the stuff that keeps a physical human body nourished and happy.