Colin Angle, CEO of iRobot, co-founded the firm straight after graduating from MIT. It’s arguably the world’s most successful robot company, creating solutions that range from the robot vacuum Roomba to resilient military solutions like Packbot. Wired.co.uk met up with Angle in New York City to talk about the future of robots, why Google’s Larry Page is a robot geek and just when we’ll be able to swap our eyeballs for robot replacements …
Wired.co.uk: What did you make of Google’s autonomous car project? Is Google going to become a competitor?
Colin Angle: Larry Page is passionate about robots. I know Larry. He’s a robot geek. He has a Roomba. I went to his wedding and the first DARPA Grand Challenge with him. He’s an excited, visionary thinker. Put an excited visionary thinker with arbitrary wealth and access to great other minds and exciting interesting things happen. Did Google need to make robot cars in order to make Streetview work? Absolutely not. It’s the equivalent of saying you need an walking robot in order to push an upright vacuum cleaner. It’s gratuitous robotics!
People often speculate about robot intelligence as a frightening thing. How do you see that?
People ask me if I’m worried about robots taking over, I say: absolutely not. People taking robot and sensing technology and incorporating them into their own bodies is much more imminent. We already do cochlea implants for the hard of hearing and neural implants for eyes [that] allow people to regain some perception of light. What happens when you can elective surgery to replace your eyeballs to give your AR-vision? Would you take out your perfectly good eye to do it? Not today. But maybe in 20 years. That creates real questions about morality and the expansion of the gulf between haves and have nots. But all of those upgrades are easier than trying to make a robot with a human level intelligence.
Will we develop robots that can compete with humans in terms of intellect? Is Ray Kurzweil right when he talks up the concept of the singularity?
Ray Kurzweil is wrong when it comes to timeframe. Nothing has moved fast in the robot industry. We’ll see an acceleration as new players enter it but if you look at the level of complexity required, it’s incredibly daunting. 10 years is a heartbeat. iRobot has been pushing the envelope for 20 years. If you’d asked me when I’d started how long it would take to make Roomba, my answer would have been next year, not 12 years from now.
__What about the sci-fi vision of android servants? Is C3PO waiting for us in the future? ____ __
We will not have humanoid androids. It’s interesting: when you start trying to make robots look more human, you end up making them look more grotesque. It takes very little to go from super-attractive robot to hideous robot. When we built Roomba, we explicitly designed it to not have a face. We didn’t want to think it was cute, we wanted people to take it seriously so we gave it more of an industrial look. People personified their Roomba anyway. Over 80 percent of people name their robot. We did nothing to encourage people to do that but they do it anyway. You don’t have to try, you just have to make the robot useful and mobile. There is a moment when Roomba stops being a robot and starts being part of the family.
We’re going to have robots in the home but they’re not going to be walking. Legs are complicated, unreliable and costly. Robots are going to look and be designed to meet the function they’re supposed to perform. People will still name them and connect with them.
*Continue reading ...
*iRobot recently announced a new healthcare division. What’s your vision for medical robots?
My goal is quite simple: to extend the amount of time you can live independently at home. You need to be able to maintain your home. That completely fits with what we’re already doing. Roomba is the most successful healthcare robot ever created. Older people are among the most rabid purchasers of Roomba. We get stories about people’s lives being changed by their Roomba. The floor washing robot Scooba is the same -- that’s a manually intensive thing that older people struggle to do.
There are a lot of interactions with people themselves that need to happen. We found with Packbot [iRobot’s military robot] that some prototypes have shown it’s possible to get a remote person to drive a robot around. There are already telepresence robots out on the market. That’s going to be part of the solution, all of a sudden doctors and nurses, depending on the robot and the accessories on it, can start to make house calls again. Back in the 1800s doctors made house calls all the time but they had to be generalists and the end of the house call was the arrival of equipment that was too heavy to lug around and a sudden need for specialists. Healthcare got more and more centralised. With telepresence and miniaturisation of sensors we can turn that inside out and bring most health care to your own home.”
There’s a huge community of Roomba hackers online. What’s your take on that?
We’re a company that was founded by engineers and really loves the fact that [a Roomba hacking] community exists. Why do you think we made it so you can take the top off and find a beautiful serial port? Why do you think we published the API for the serial port? It’s an expense but it was the right thing to do. People should hack robots. There is a cottage industry of people making money from Roomba accessories out there today. I hope that it will grow over time. People are getting excited and learning about robotics. Roomba is a durable, dependable and affordable platform for doing robotic research.
Are you planning on developing apps for robots like Roomba and Scooba?
The robot operating system architecture will divide in half. The mobile industry is moving far faster and is far larger than the robot industry. You’ve got a couple of wonderful front runners, Google and Apple, which have developed software platforms that are optimised around communication, voice recognition, graphics and touch screen interfaces. That’s enabling for the robot industry but it’s not sufficient. If your phone dies nothing that serious happens. But if you’re robot dies and it’s bigger than a Roomba, you don’t want it to topple down the stairs. There’s a need for reliable, safe, secure software at the core of the robot.
But there will be a division between the core robot OS which is carefully designed and has fail safes and the cool, sexy UI for the consumer. Things like iPhone control first evolve in the informal hacking communities but over time the robots will have much more sophisticated operating systems and be able to link in to other systems. Ultimately though if the robot’s function is to be vacuum cleaner, it needs to do that well first.
One day I see lots of robots managed by a butler robot. I talk to it and it talks to the other robots. At that point you’ll see a lot of human interaction features on the main robot. You could have Android OS running on part of that robot alongside a safe and secure robot OS. There’s a place for co-operation.
You’ve been pretty scathing about the idea of robot games. What puts you off the idea?
I’ve always loved gadgets and experimenting with what seems like fad and what seems genuinely valuable. I’ve had smart home technologies in my home for decades. One of the things I’m deeply passionate about is that robots need to be practical and useful. Software doesn’t. You can make a beer-drinking iPhone app and make some money. A robot is a serious design and capital investment. The robot needs to serve a purpose and meet a physical need. Robot video games wouldn’t do that: it’s 15 seconds of novel fun to be eclipsed in minutes by a 3D graphics video game. What does a flying robot give you if you’re just watching it that a 3D projection doesn’t? Your actual robot will fly for 10 minutes before it needs new batteries and you can’t actually fire lasers from your robot and blow things up. It’s so handicapped relative to video games. Robotics is a technology we should use to solve major societal problems -- let’s solve bomb disposal or the energy crisis -- not create more toys.
