Robots, from mechanical dogs that can learn new tricks to automated vacuum cleaners that avoid furniture, are steadily becoming a part of everyday life. But the real robot boom lies just ahead, experts say.
In the future, robots could help determine the outcome of wars and identify problems in data centres. Office buildings may come to life as they use Wi-Fi to dispatch robots to control human access, test heating and cooling systems, and fetch tools for workers.
IDG’s Lucas Mearian recently spoke about the future of robots with three experts: Chuck Thorpe, director of Carnegie Mellon University’s Robotics Institute; Jeanne Dietsch, CEO of MobileRobots.com; and Vijay Kumar, a professor in the Department of Mechanical Engineering and Applied Mechanics at the University of Pennsylvania. Here’s what they said:
Where will robots serve in the future?
Kumar: They’ll fight our wars. My robots will fight your robots. It’s like playing chess. It’ll be an expensive game, but I think it will minimise loss of life. I don’t know it’s something our leaders want to hear about, though.
What’s the future of robots in the services sector?
Thorpe: That’s the real frontier for robots — getting them out of the manufacturing world and interacting much more with people in our day-to-day chores. That’s a little tougher. If you think about an assembly line, it’s set up so that everything is predictable. If you think about the job of creating a robot butler to operate in your house, the difference between pile of clothes which should be picked up and a sleeping cat which shouldn’t be picked up is a little bit tricky.
The manipulation skills required to make a bed — flipping up the corner of the mattress and tucking in a sheet so it looks straight — is complicated. Some of those jobs will take a long time before robots can really do them as well as people do them.
To really have a humanoid robot walking around like C-3PO or like Rosie on The Jetsons is still a long ways off.
Can robots help information technology?
Dietsch: HP uses our PatrolBot in their data centres. It’s got temperature sensors. It drives around several times a day, and the data from these patrols is used to create 3-D models of the heat in the facility. If there’s a problem in the facility, the computer or a human can send the robot out to check on it.
Our robots are computers on wheels. When programmers start realising, “Oh, I can drive my computer around and make it do things,” they’ll come up with all sorts of interesting things. For example, they could keep track of any asset in the facility based on RFID tags.
What’s hard about making robots act like humans?
Thorpe: I think you could turn that question around and say, “Wow, humans are a lot more complicated than we thought.”
We’re striving for a couple different things with robots. Sometimes we’re trying to make them very much like humans, and sometimes we’re trying to make them be a lot better than humans. Where we’ve had our greatest success is when we let the robots be robots and do what they’re best at and let humans be humans and do what they’re best at.
A couple of guys from my lab have spun off a company that makes a little TV camera that watches the road as you’re driving. It doesn’t drive your car, but it’s there to watch, and if you start to get sleepy and start to drive off the road, it will beep at you and wake you up before you become an accident statistic.
What else are you working on?
Thorpe: We’ve got 61 faculty members with 120 projects. We’ve got robots doing everything from cleaning paint off ships to search and rescue in hazardous environments. We have snake robots that can crawl through skinny little holes, and great big military scout robots. We have teams of robots playing soccer.
That’s a great way to study teamwork: How do robots coordinate to tackle a project? In the latest version of that, we have a team that consists of robots and people; how do robots and people collaborate to solve a difficult task?
Can you speculate on some of the more far-out applications for robots?
Kumar: My colleagues would argue the biggest opportunity for robots is in entertainment.
There are many, many prototypes of robots that are intended to be domestic companions. They walk like humans and talk like humans, but unlike pets, they don’t dirty up your apartment.
Dietsch: Our robots are made to work indoors. We’re working next toward working outdoors. Everyone thinks GPS has solved the problem for robots, but it hasn’t at all. If you’re under tall buildings or trees, GPS is not reliable.
The problem with working outdoors is that you’re working on three-dimensional terrain. Going from building to building, depending on terrain, is not so bad. I’m hoping we can do that in one to two years. Following streets and sidewalks — if we can do that reliably in three or four years, that will be good. But we’ll still not be able to avoid cars going 40 mph.
Thorpe: Certainly, within 25 years, we’ll have robots assisting surgeons to be more precise, robots enabling elderly people to live in their houses longer and not have to go into nursing homes, (and) robots doing a lot of the dull jobs we don’t like to do, so we can have more fun and do the more interesting jobs.