Charlie Kemp builds robots. His office and an adjoining lab, on the second floor of a Technology Square building, are filled with functioning robots and the assorted computers and mechanical parts that accompany the task. Kemp, an assistant professor in biomedical engineering at Tech, founded the Healthcare Robotics Lab in 2007. His goal is to design robots that can be used in health care, particularly to help the millions of people with physical impairments who require assistance in their daily lives. Before coming to Tech, Kemp earned three degrees from MIT. While in Boston, he was a member of the electro-pop band Electric Laser People. The group is known for its album Straight Talk on Raising Kids, which has been a mainstay on college radio stations.
Robotic inspiration: I got into this because I was interested in artificial intelligence. My adviser at MIT thought robotics was a good way to understand intelligence. If it’s grounded in this body and the world, it can be valuable.
A booming field: Robotics finally has begun to deliver on the grand dreams of science fiction, dreams that pop culture has had for years. Computing power is increasing, and we have better sensors. It’s an excellent time to be in robotics.
Purpose for robots: Coming to Georgia Tech, I wanted to take seriously the notion that robots are machines that can help us in daily life, particularly in health care. For those who need assistance, robots offer privacy, independence, round-the-clock assistance and consistent performance.
Robots in the home: Robots have been very successful in factories, which are controlled environments, with no people around them. Now they’re entering our home environment, our office space. How are we going to interact with robots? And how are they going to interact with the world? Those are our challenges.
Robot toy: This one my wife’s parents got me. It just looks cool. I like how it’s made out of random objects, but it all comes together. It’s reminiscent of how we pull parts together to build robots.
Science fiction: I love sci-fi. For current authors, I like Vernor Vinge. He’s really talented and captures these worlds where everything fits together. For classic authors, I like Asimov. He’s great with robots.
Film robots: There’s a new movie, Moon, I was really impressed with that. It was just this guy with this robot, and it only had a smiley face. But it could communicate so much. And Blade Runner is an obvious one. Are there any robots I haven’t liked? I’m a sucker for it. If you put a robot in a movie, I’m going to watch it.
Robot arm: It’s simple, and it works. It’s fun to play around and see how much you can do with it. Robot arms are really expensive. New ones cost about $100,000. So there’s a need for cheap arms.
Feeding robot: This is a robot we’re really excited about. Its name is Gatsbii. We’re collaborating with Psychology, Computing and the Aware Home, looking at how to use robots to help older adults live at home independently. The maker gave out the robots to 11 teams based on their proposals. The code is open source, which is great. We can all develop something that the future of robotics can be built on.
Robotic firsts: This is something I love to see. [Gatsbii has just successfully scooped and delivered M&Ms.] I come out of my office, and robots are doing something new. This is a time of firsts. Once, I had a robot prepare pancakes. It’s important to develop things to allow robots to do something new.
Using sensors: We added an RFID antenna ourselves. We’re looking at the idea of adding labels to little things around the room. For a light switch, it would tell the robot that it can turn it on or off.
Harsh retirement: This is an older robot, EL-E, and it’s kind of going into robot retirement. It’s dangerous being an old robot in a robotics lab. You tend to be cannibalized for parts.
Robot mobility: Cody has two arms and wheels that allow it to go anywhere, which is important. They can move in any direction, just like people can. The arms are flexible, just like our arms. We’re making this open doors and drawers right now. That’s something people take for granted. There’s a lot of the world hidden behind doors and drawers.
Adding AI: Another nice thing is the robot doesn’t have to have an exact model of whatever it opens. We just tell it where the handle is. It doesn’t need to know if it’s a door or a drawer, it just figures it out. Robots, so far, have not had any common sense. Generally, people know how much force it will take to open something. We’re trying to develop that for robots, which relates to the AI side.
Robot sponge bath: Cody was also cleaning people recently, which got some attention. Hygiene is a big issue, and the sponge bath is our first effort to have robots help with that. The broader notion is that the robot needs to reach out and make contact with people. We’re looking at how they can do that effectively.
Importance of touch: We did a study that showed that leading a robot through an obstacle course by hand is more effective than using a gaming controller. One of the subjects was a gamer, and he said he liked the controller, but he still hit fewer obstacles leading the robot by hand.
Robot retriever: This is Dusty, which can pick up objects. We found that dropped objects are a big issue for people. This is able to pick them up really easily and consistently. We did a study, and people had success using it. They really light up to have that independence.
Future of robotics: Robots need to be designed so they can meet preferences. It’s probably not going to be one robot fits all. Just like with cars, there are so many different types to meet what people want. But we still need the Model T of robots.
Twitter