It's no surprise that uber-geek and iRobot CEO Colin Angle is a Star Wars fan.
But even though the heroic R2-D2 and humanoid C-3PO get all the glory, Angle admires the lowly MSE-6, a small, box-shaped droid that scurried around the floor to make repairs and lead the thousands of storm troopers around the Death Star.
Why? "We could build that robot!" says Angle. Instead of idolising R2-D2 and C-3PO, the world would be in much better shape if the robotics industry focused on solving real problems, even if they're mundane, ugly or dangerous.
iRobot CEO and co-founder
Angle gave the opening talk at the RoboBusiness conference in Boston this week where he urged entrepreneurs and engineers to be practical and business-savvy.
There is no shortage of examples of how that approach has worked for iRobot. In addition to selling millions of Roomba vacuum cleaners, the company has deployed thousands of its PackBot robots to defuse roadside bombs in Iraq and search caves in Afghanistan.
Company engineers also jumped in with customised robots for searching underwater sea oil plumes after last year's Gulf of Mexico oil spill and to do inspections after the partial meltdown at the Fukushima nuclear power plant.
That doesn't mean consumer robots are passé. On the contrary, iRobot is actively imagining what the robot-enabled home of the future will look like. The product to watch here is iRobot's AVA, which is a robotic pedestal with an iPad or Android tablet as the "head" and man-machine interface.
After his talk, Angle spoke with me about how the revolutions in mobile computing and gaming are putting robotics development in the fast lane — as long as people stop chasing C-3PO and Rosie the maid from the Jetsons.
CNET: Why are you so hard on C-3PO and R2-D2?
Angle: Robotics has been around forever and it's been the next big thing forever, and it is so exciting and compelling that it's easy to get carried away. People almost always do and that's one of the things that has held back the industry. It is both true and important that when you see the [Star Wars] movie, everyone is inspired to pursue that unachievable reality [of] building C-3PO or at least R2-D2. In Japan it almost has become an art form of how elaborate an engineering demo can be. But it has nothing to do with building a practical robot industry. It's a remarkable era to be building companies and big money can be spent on grand demos. Less money is spent on [finding] where the need is, how do I create an approach to that need that is not doomed from the start.
Go look at all the companies making healthcare robots and imagine that you commercialise that [US]$250,000 home healthcare nurse. Does anyone care? It's this maddeningly exciting but super constrained set of engineering challenges that keep [robotics] the next great thing since 1962 when Rosie the robot from The Jetsons came out and we all knew that robots should be cleaning our floors.
You took a different approach to cleaning the floor. Rather than humanising it, you automated it.
That's right. It's not about building a humanoid robot pushing a manual vacuum cleaner. It's about creating the simplest way to automatically clean the floor that we can come up with. Even today there are vacuums that are more about the technology than cleaning the floor and that's not what we want.
Do you think people will still be inspired by robotics when and if industry produces more workday machines doing grunt work?
That's a great question. On the surface, no. At a deeper level, yes. When we first launched Roomba, we didn't even call it a robot — it was the media that called a robot. If you ask the typical two- or three-year-old or a teenager what a robot is, they will think about a humanoid that does my homework for me or walks the dog. When I go and talk to kids and pull out the Roomba, it's not this big "wow!" moment. Yet 85 per cent of Roomba owners name the Roomba. People get so attached to these devices because they make their lives easier so that you get this deep emotional commitment.
We have a robot in our office called Scooby Doo (used for spotting roadside bombs in Iraq) and when it was finally blown up, the operator carried it back with tears like it was a fallen comrade. Why? Because if it weren't for that robot, he'd be dead. Once again, it's sex appeal versus importance.
There are so many opportunities to make a bad decision in building a robot company on top of all the normal ways that entrepreneurs screw up that it is incredibly difficult to truly create value because it is so cost-sensitive. In the end, robots do things that people can do. So there is a cost above which you can hire somebody to do it and that bounds the opportunity.
On that point, there's more concern about jobs in general and the worry about robots taking jobs. What are your thoughts on taking away people's jobs?
Not something that I believe is a legitimate worry. Why? Because the types of jobs that robots are taking on or could take on are the jobs where there are currently jobs shortages. If you look at janitorial jobs, there is a shortage of people willing to do that work today. If you look at security jobs, there is a shortage of people willing to do that work today and it's getting worse.
I was doing a radio show and one person called to complain about robots taking jobs and the next caller called in, and said, "I lost my job to a robot filling pharmaceuticals. Now I have much better job fixing the robot."
Innovation technology disrupts the world in different ways. The washing machine: how many jobs did that eliminate? Certainly it eliminated countless hours of manual labour but created opportunities for women to become enfranchised, go to school, get jobs and contribute to our GDP. The internet, which arguably single-handedly eliminated more manual-entry routine secretarial work than anything else, has inspired and engendered whole industries. I believe the opportunity robots represent will be a substantial net positive to the job market and the GDP of the world.
One way to look at robots is that they are basically moving computers. So how does Moore's Law and other technology trends affect your development cycle?
There are three tech trends that are impacting the industry positively and certainly Moore's Law is one of them. What Moore's Law means to the robot industry is that things like machine vision become more affordable for robots. So robots will have a better ability to understand their environments and take on more complex tasks and higher-value missions.
The second is the video gaming industry. The Xbox game sensor made by PrimeSense is an incredibly disruptive sensor because now for a very low cost — under [US]$100 — your robot can have a gestural interface. The multibillion-dollar game industry is investing in gestural interfaces at a scale of investment inaccessible to the robot industry at this point.
The last and certainly the biggest trend is the mobile computing revolution where the mobile industry is now solving problems that the robot industry thought it had to solve. So voice and video over IP, machine vision, object recognition, human-independent voice recognition, very sophisticated man-machine interfaces.
Will the AVA robot that you introduced at CES be leading the charge on this?
I think it's going to be one of the most important robots we've ever made. It is a robot where there is explicitly "the robot part", which is the neck down, and the mobile computing part, which is the head. And so it's a robot where there is mobility, creating maps, navigating, avoiding obstacles and moving around physically in human environments, which can be very crowded human environments. But then on top of that, just by putting an Android tablet or an iOS tablet on its head, I get voice recognition, the ability to display video, touchscreen interfaces, facial recognition by using technologies made for the iPad. I now have an incredibly useful robot and because development tools are so sophisticated for these tablets, I also empower the world to become roboticists. Those three trends really marked the beginning of a period of great innovation in this space.
What will be the apps for AVA?
Things like security, things like home healthcare, being able to visit my kids when I'm on the road and interact with them more richly than if I just call them on their cell phone — I could play games of hide and seek or something. Some things like that will ultimately become core robot apps and people will buy a robot because they will be able to do one or two of those and then more apps will come in. Think of how all the current iPad apps could be made more cool by becoming physically proactive.
How could an iPad app become more physically proactive?
It could go find you. The idea is that the information is finding us as opposed to us finding information. Would we like to be less bound to our cell phones and instead have some kind of mitigation in between? Mobile information, making information proactive — that's kind of phase one. Then you get into where robots will start to truly differentiate themselves and that's moving stuff, having more physical functions.
You start off with moving sensors around her house. I want home security [but] I don't particularly like the idea of putting cameras all over my home because I don't know who's watching them. I much prefer having the metaphor of a security guard where it's in the room with me, I can see it. When it's not, I don't. Plus I don't need to rewire my home. At what price do I prefer buying an AVA at my house versus putting cameras up? There's a very clear answer for that.
On the medical side, making compliance with complicated medical regime more effective — that's a big one. Carrying stuff around. Do you want a stereo that follows you around? Do you want a robot that goes around the house that finds when the husband leaves a toilet seat up? Someone will write an app for that. Who knows where it will go? The difference is the proactivity, the mobility, the physicality. It's not super important for all apps but for a certain class of apps, it is going to drive things.
Does the industry need more people who think like entrepreneurs?
Absolutely. The robot industry has done a poor job of attracting enough business-oriented entrepreneurs and the ones that we do attract sometimes fall into the trap of imagining robots that can do more than they can actually do. You can probably find plenty of entrepreneurs who believe that all we need to do is build C-3PO and the world will be ours. What's rare is the entrepreneurs who think if we build MSE-6 and the world will be ours. That's the missing ingredient in having a true entrepreneurial explosion in the space. The technology is truly not our limit — the Roomba could have been built in 1990. It's the good business models and the intersection of those good business models with what can technically be created and realistically manufactured.
Back in the 1970s, Bill Gates and Microsoft had the vision of a personal computer in every home which, at least in rich countries, has largely panned out. Will he have a robot in every home too?
I think we're well on our way to a robot in every home. There have been 6 million Roombas sold and in Spain one out of four vacuums sold this year was robotic so there is a mainstreaming of the industry. So the question is how many robots will be in every home and what will be the next Roomba-scale success.
People don't like to vacuum. They hate to mop. Robots that can disinfect and scrub floors clean are a legitimate contender [to succeed like] the Roomba. [iRobot's Scooba floor cleaner] has the potential to be the next robot in your home. If you get many beyond two robots, you are going to need robots to manage your robots so that will also be part of the future.