Building a More Sociable Robot Can't find anyone who wants to hang out this weekend? Help may be on the way. Inventors are working to develop robots that can interact with people on a deeper level: communicating, responding to emotion and operating under specific rules of social behavior.
NPR logo

Building a More Sociable Robot

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
Building a More Sociable Robot

Building a More Sociable Robot

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript


This is Talk of the Nation: Science Friday. I'm Ira Flatow. What would it be like if you were to walk into your home and see a robot cooking dinner? Would it be strange if it asked you about your day or reminded you to call your mother or nagged you about watching too many football games? Many researchers envision a future where robots like this are part of our daily lives, helping us complete tasks like washing dishes, or playing the role of companions. And to help robots interact with people, researchers want to create robots with personalities. These social robots would understand social skills and express emotions. Special robots have been the stuff of science fiction from the "Twilight Zone" robotic companions - there are a whole lot of "Twilight Zone" episodes with robots as companions. Perhaps you remember Data on "Star Trek" who tried as he might to adopt human emotions. What are the technical challenges to creating these types of robots, and how lifelike do we want these machines to be? Maybe they'll be too lifelike, and we will get annoyed with them? What do you think?

Here to talk to us about the commercial robots available today, what is being planned for the future, here's Helen Greiner. She is co-founder and chairman of the board of iRobot. That is the company that brings us the Roomba Vacuum Cleaner. She joins us from Massachusetts. Welcome back to Science Friday.

Ms. HELEN GREINER (Co-founder and Chairman, iRobot): Oh, thanks for the invite again.

FLATOW: You're welcome. Also joining us is computer scientist Peter McOwen. Dr. McOwen is a professor at the University of London. He is project coordinator of the group Living with Robots and Interactive Companions. And last week, this group of European researchers launched a year-long project aiming to develop robots with personalities. Dr. McOwen joins us from the UK. Welcome to Science Friday.

Dr. PETER MCOWEN (Project Coordinator, Living with Robots and Interactive Companion): Good afternoon, Ira. It's a pleasure to be with you.

FLATOW: You're welcome. Doctor, why do we want robots with personalities?

Dr. MCOWEN: What we believe is that by being able to develop robots that have personality is that we're going to be able to make the sorts of technologies that we want to use in our life easier to use. So as human beings, we tend to interact very well with other human beings and the idea is to try and give robots and other sorts of technologies at least the beginnings of some type of personality, some type of quality of interaction there, where you feel that there's a companionship being built between you. But I think it is worth saying at this stage that we are not actually building robots that are going to be able to help you do the ironing and so on just yet. But we are beginning to try to develop some ideas about how we can make technologies, robots and also kind of graphical characters interesting enough in the interactions with them that you believe that they have some form of companionship with you.

FLATOW: But might we have these robots that don't have the personalities we like? I mean, don't we run that risk?

Dr. MCOWEN: No, because what we would ensure, obviously, in doing this is that the companions are the sorts of companions that you would want to have. There would be no point in developing a technology that made life more difficult. So clearly, we would want to be producing personalities that made life altogether more enjoyable.

FLATOW: 1-800-989-8255 is our number. Talking about robots with personalities. Also, if you're in Second Life, you can find us in Second Life Island. Also starting today, teenagers can find us on Science Friday in Teen Grid in Second Life; that's at the Eye4You Alliance Island. Helen, tell us about your robots. Do you find any need to make a vacuum cleaner have a personality?

Ms. GREINER: I think some of our users believe they already do. You know customers have told me for many years about naming their robots. I think they buy them as appliances. You know, they want to get the vacuuming done. But they get them home and you know they negotiate around the furniture, they back up from the stairs. They communicate with bleeps and bloops. They go back to the charging station and power themselves or eat. You know, I wish my cell phone could do that 'cause I keep forgetting to charge it. The people have had - have this type of experience before with pets in their home so they start giving them names. In fact, Georgia Tech just did a study, and they found that two-thirds of the people who are Roomba users actually give them names.

FLATOW: Well, wouldn't it be nice to be able to say to your Roomba, you know, hey, you missed a spot over there, and it goes over there and says, thank you very much. I appreciate that or something like that.

Ms. GREINER: You would think so. Roombas actually know when they are picking up dirt and they stay in the same place using intelligent algorithms. They can do that spot for longer. We do focus groups on our customers, and people say they want that kind of feature, but when you ask them how much are you going to pay for it, it's really not as much yet as the cost that it would take to add that kind of feature to their robots.

FLATOW: Peter, are people looking for just a substitute for their pet or something like that?

Dr. MCOWEN: It is interesting hearing what was being said there about the robotic cleaner and the fact that it emits kind of beeps and blops as it kind of wanders round, and clearly one of the reasons that it is so successful is that people do anthropomorphize; they give kind of human-like properties to something which is actually a very good piece of technology but it's doing very simple things. And so the question - the point there is that seems to make that technology more enjoyable to work with, to be around. One of the things we are doing in this project is we have a group of people who study the behavior of dogs and how humans and dogs interact with one another.

So we are looking at the sorts of things that make a relationship with a pet dog, for example, a fulfilling relationship and trying to see if we can take some of those ideas and build them into the way that the robots will interact with humans so as to kind of build that link. We are not building a robotic dog, but we are going to try to take some of the lessons that we can learn about how you build a relationship with a pet and build those into the robots.

FLATOW: And how do you give a robot a personality or emotion? How do you build it in?

Dr. MCOWEN: The research is at a very early stage. One of the issues, clearly, though, is for an entity - be it a robot or another human being, for that matter - for them to manifest any kind of personality, things that they are reacting to you in an individualized way, so you've got to be able to recognize, for example, the expression on the person's face or the body posture that they have, how they are positioning themselves with respect to you in the kind of social spaces. And so we can start to build computer vision systems that recognize, for example, whether somebody is smiling or frowning and whether or not there - the distance that they are and so on.

So you can begin to have some kind of meaningful input from the kind of vision point of it. Also looking at the tone of people's voices, as they are speaking, whether or not they are sounding happy or sad. So you got those inputs coming into it. Then the question is how do you relate those inputs to the actual behavior outputs, and that really is mediated by the personality that you're building into it. So the robot is aware of some element of your emotional state, and is then able to act accordingly, so it may be that it's when you are sounding annoyed and looking unhappy, then the robot will then go off into a corner and quietly wait until you are in a slightly better mood to interact with them, which is what you would expect maybe, a human companion who realized that you are having a bad day, is just going to kind of not bother you, get in your way and under your feet. And so that is kind of what we do. We are looking for something called semantic cues from the environment. Something that tells us a little bit about the human being's emotional state.

FLATOW: Do we want the robot to look like us, or a pet, or does it have to have a face, or...

Dr. MCOWEN: This is a very interesting question, and it is one that a great deal of research is actually being done on. There is classic research that was done in Japan, and it produced some interesting results, and something called the "uncanny valley." So it turns out that if you have robots that look like robots, they seem to be more acceptable to more people than robots that look too humanoid. If you move up to the kind of Mr. Data-type androids, and those aren't functioning as perfect human beings, then people find them actually quite difficult to be around, because the people report things like feeling that they are in the presence of a zombie. So it looks like a human being, but it is not doing exactly what a human being would want. So our robots aren't going to look anything like human beings. We are building robots that look like robots, but giving them the ability, for example, to have something which you can identify as a face, from which you can then take your cues, from the robot's face, so that there will be some way of being able to indicate the linkage with the robot through it having a kind of face and possibly, like the bleeping noises that the robot cleaner uses.

Ms. GREINER: I think Peter's right, because before we put the Roomba on the market, we ran focus groups to see if people wanted robot vacuum cleaners, and people would imagine, you know, like a humanoid robot pushing a vacuum. Most people said, "No, no! Not in my house!" That is kind of creepy, and also, "where would I store it?"

Dr. MCOWEN: That is a very good point, yes.

Ms. GREINER: So we have a little Roomba, and it goes under the couches, and under the bed, so it gets the job done more effectively, because it is small, and you know, people can actually have it in their house, and feel comfortable with it.

FLATOW: So, you don't want your customers to confuse it with being a human, or having a personality.

Ms. GREINER: It is not having a personality, because, you know, I was inspired by R2D2 myself, and he has a lot of personality, but it doesn't have to be a human personality. I will give you another story. Sometimes we get customer support calls, and you know, they say they have a problem, and the robot is still under warranty. We will say, send it back, and they will say, "No! No! Replace it? Not Rosie, you know, she is my robot," So you know, I swear they expect us to come to the door in an ambulance, perform open-chip surgery on their robots. This really is a whole new class of machine, and I am glad that universities, people like Peter's team are jumping in and studying really what is going on here, and you know, I am looking forward to seeing the results.

Dr. MCOWEN: And I think it's fair also to point out that part of this research is not only about building robots, but is actually trying to understand what makes a relationship, what gives something, a piece of technology a feeling of something called agency, that it seems to have a kind of an internal view of the world, of the way it interacts with you. There's a beautiful experiment that I remember reading about it that was done about 10 or 15 years ago, where they had a computer screen that had a series of squares and circles on it, and the squares and circles just moved around randomly, and...

FLATOW: Peter, I am going to have to interrupt and tell you to hold that thought, or else we will never hear the end of it.


FLATOW: Because we have to go to a break, but sounds like a great story, and I do not want to have to interrupt you right at the punchline, so stay with us. We will be right back talking more about robots with Peter McOwen and Helen Greiner. Our number, 1-800-989-8255. Stay with us. We'll be right back after this break.

(Soundbite of music)

FLATOW: This is Talk of the Nation: Science Friday. I am Ira Flatow. We are talking about social robots, robots with personalities. Maybe you would like to have one working for you with Peter McOwen and Helen Greiner. Our number, 1-800-989-8255, and when I rudely interrupted Peter, he was telling us a story about circles and squares, was it?

Dr. MCOWEN: Circles and squares, yes. So, the screen showed these shapes just moving round randomly, and people were asked to tell the scientists what they actually saw, and the way they reported that was, well, the squares were being chased by the circles, and then the circles went and kind of hid in the top corner, and the squares kind of chased them up and around. And what that shows is that even something very, very simple like a kind of random set of moving patterns, we tend to see some kind of pattern in there, in part because we like stories. As human beings, one of the things we have evolved over a period of time to do is to pass information on through a kind of narrative or a story. So we tend to anthropomorphize technologies and things a great deal, and even if you've got something very simple, like these kinds of moving shapes, you still see them as having some form of character - that the squares were the good guys, and the circles were the bad guys, for example. And I just think that is a very interesting experiment, because it is such a simple thing to do, and yet people see this as being, there is a real battle being fought out there between the squares and circles.

FLATOW: Pam in Aberdeen, Idaho. Hi, welcome to Science Friday, Pam.

PAM (Caller): Hi, Ira, my question is, will the people that are working on this be designing their robot keeping the geriatric population in mind, since all of us baby boomers are continuing to reach retirement age, so in other words, like, would it be more hi-fi, clap on, clap off, or "help, I've fallen," you know, that kind of thing?

Dr. MCOWEN: That is actually one of the application areas we are looking at, is assistance for people who want to remain in their home, and have some kind of companion/helper in there, so the nice thing about a robot is that if we design it such that you feel comfortable with it, is that that robot can not only do physical tasks, but it also can interface to the electronics in your home, and also through the Internet to other places, so it can provide information to you, provide a kind of channel by which you get information that's important to you coming through that way, and provide you with, kind of, assistance and the ability to be there and check if, you know, the lights are switched off in the back porch, and that you've taken your tablets when you were supposed to.

FLATOW: Thanks, Pam.

PAM: All right, great, thanks.

FLATOW: You are welcome. Why not the whole house as a robot? You know, I am thinking of Hal in "2001: A Space Odyssey," which is 40 years old this month. Can you believe that?

Dr. MCOWEN: It's amazing. That is - actually, yes, we are also looking at developing exactly that, something that we are calling at the moment The Spirit of the House. So the idea is that there is, your wired house has lots of technology in it, and as you are moving - so for example, your companion entity may download itself into the robot for a period of time, and then, later on when you go upstairs because robots tend not to be very good going upstairs, the same companion will load itself into, for example, a computer in another room of your house, but you will be able to identify that even though it is in a different body.

It is not in a robot body. It is in a kind of computer screen in front of you as like a computer graphics character, but you will be able to identify something about it being the same entity, the same agent that you've built this relationship with, and these are really fascinating questions. Will it work? We are not quite sure, because you've taken something from a situation where it is embodied - the technical term, it's embodied as a robot with the ability to move around in three-dimensional space, as opposed to being on a kind of two-dimensional screen, so these are some of the questions we're looking at.

Ms. GREINER: And that Smart Homes haven't entirely caught on, but, as Peter said, a robot's embodied, you know, you know if the robot's in the room with you. You know what it's doing there. And we're actually just going into beta testing on a virtual visiting robot. That's an Internet-connected robot as Peter was saying. We'll be in beta testing this year. And you'll be able to log on to it anywhere in the world, and you can see what the robot sees, you'll hear what it hears, and you can drive it wherever you want to go. And you know, you can use it to check on your pets, see what your kids are doing, see that your home is secure, and you know, make sure your elderly grandmother is taking her meds, as you said - the elderly demographic is a market that we are targeting.

FLATOW: Does it look like a person?

Ms. GREINER: No. No, it doesn't look like a person. We don't believe it has to, to either get the job done or to have a personality and to have quality interactions with people.

FLATOW: Are you building a personality into this?

Ms. GREINER: This one will take on - it's being controlled by somebody on the outside, so it will become them when they're logged onto it. But no reason the same micro-processor can't start to have a personality, and we look to that for future generations.

FLATOW: So, you could actually speak through it from a distance?

Ms. GREINER: You can - it's like you're there, except it's when you have to be away from home.

FLATOW: And I guess it could have a GPS in it to let you know exactly where it is in the house or something like that.

Ms. GREINER: Kind of a location system.

FLATOW: Yeah, location system.

Ms. GREINER: But, you know, even today, you know, we're speaking about the elderly. People tell me that they buy Roomba for their elderly relatives because they find it a challenge to lug out a vacuum and push it around and that, you know, helps people remain more independent, and that's one of the things that, you know, Peter and his group are looking at. How many more ways can robots help people remain more independent?

FLATOW: Well, you know, we're going to have Dean Kamen coming on a few minutes. Maybe he'll help you make it walk upstairs.

Dr. MCOWEN: That would be neat.

FLATOW: Well, he's got that wheelchair that does that. 1-800-989-8255 is our number. Talking about robots. Let's see if we can get a call or two in. Nancy in Louisville. Hi, Nancy. Welcome to Science Friday.

NANCY (Caller): Hi. Thanks. Love this topic, since I read "I Robot" years and years ago. I work with high school students and it's senioritis time, and I'm wondering if perhaps there's value in robots that were not as pleasant, not in a - that they might, for example, teach behavior, sort of do behavior modification. In other words, they would be as recalcitrant perhaps as some of those students are?

Ms. GREINER: I believe positive feedback from the robots would be the way to go. Maybe you'd be wanting to help - I mean, you want help quitting smoking, the robot could remind you, you know, the positive health benefits of that or losing weight or anything. I think positive reinforcement is the way to go, from robots.

NANCY: Well, I thought, and I was thinking also that, you know, positive when you are positive and negative when you are negative. In other words, it would, you know, it could teach, like it could teach good behavior.

FLATOW: There's also - remember there's a robot that if you want to know what it's like to have a baby, it acts as a baby?

NANCY: Oh, we have those. We have those, actually, at my school and the kids - it's really good for them. It's pretty creepy actually, but - and I do have a bigger question about educational uses of it. You've talked about the elderly. What about with kids? And I'll hang up. Thanks.

FLATOW: OK, thanks for calling, Nancy.

Ms. GREINER: Well, we find that when kids learn with robots, they get excited about studying engineering and sciences.

Dr. MCOWEN: Absolutely, yes.

Ms. GREINER: And what we've done is we made a developer's version of our rugged little reliable Roomba base, and it's being used today by researchers at CMU, MIT, USC and Brown. But it's also being adopted by Botball, which is a K-12 program, and they teach kids computer skills and engineering through robot design and programming. And, you know, it reaches over 5,000 kids today. And there's also, in the Boston Museum of Science, a Ph.D from Northeastern, they've set up a tangible computing interface where kids as young as four, they can go up to a robot and teach it to program - they program it by putting blocks in certain order, the way they want their software flow to operate. And we believe that that's going to encourage the next generation of technologists, including future iRobot employees.

FLATOW: Why, you could not have made my introduction any easier for me because we're going to spend the few minutes talking about teenagers and robots and robotic contests and having kids interested in it because there is a great robotic contest that just finished.

Unidentified Man #1: With five seconds, Thunder Chickens going off with three, two, one. He lights like a bonus at the buzzer.

FLATOW: Picture an arena, 60,000 screaming fans.

Unidentified Man #1: Unbelievable.

FLATOW: Blaring rock music, flashing multi-colored lights, and going head-to-head in the middle of the ring, are robots.

Unidentified Man #1: We'll get the official score with any penalties in just a moment, as we try to crown the champions of the first robotics competition, First Overdrive 2008.

FLATOW: This was the atmosphere at the First Robotics World Championship, held last weekend at the Georgia Dome in Atlanta. Thousands of kids from around the world competing to see whose "bot" was best. And that competition is in its 16th year. It was founded by inventor Dean Kamen.

Dean is also founder of DEKA Research and Development Corporation, which invents Quality of Life Technologies, and if you'd like to see a video of the competition, you are welcome to surf over to our website at, and you can watch the action as you listen to us talk about it in here.

A little bit later in the hour, we are going to be joined by a member of one of the winning first teams, so stick around for that. Dean Kamen joins us from Manchester, New Hampshire to talk about first, which former President Bush described as the WWF for smart people. Welcome back to Science Friday, Dean.

Mr. DEAN KAMEN (Inventor, Founder of Robotics World Championship): It's great to be here. I'm glad you have that quote. We are trying to make it our tag line.

FLATOW: Is that right? It sounded just like that from that little bit of audio we just listened to.

Mr. KAMEN: It was that line, really brought the house down, especially as it just came blurting out of the former president of the United States.

FLATOW: And it sound - and is that part of the intrigue of this to teenagers? It sounds like something that they would normally love to be involved in?

Mr. KAMEN: It's astounding. You get there, and it is just by any metric, whether it's the Super Bowl, or the World Series, or the Final Four, it is the most exciting sporting event you can attend, and you don't need to put any - accept that it is science, or accept that it is technology, or accept that it is a science fair. In the unlimited class of exciting competitions, we win.

FLATOW: You've been running this for 16 years. You have now gotten 31,000 kids, 23 schools involved. Did you ever envision it growing this big?

Mr. KAMEN: Actually, I envisioned it, and still envision it growing much bigger. I think it's critical that first, be something that is available to every kid in this country. They all need the option to see while they're still young enough to make career decisions, that science, and technology, and engineering, and problem solving is every bit as exciting, rewarding, fun as bouncing a ball, or anything else they put their passion to. Every kid deserves the opportunity to participate in FIRST.

FLATOW: Let me ask my other guests, Helen Greiner and Peter McOwen. Would you - is this something you would love to have had when you were a teen?

Mr. MCOWEN: Absolutely. Yes. I'm always impressed that the enthusiasm that young kids show, particularly with robots, because it really is kind of seen as a cool and exciting thing to do. And the more that we can get young people interested in science and the engineering parts of things, then the better.

That's something I spent the rest of my time doing is going out to schools, and running events to try and make people excited about science, because it really is absolutely critical, so kudos to you for doing it. It's wonderful stuff.

Ms. GREINER: I was inspired myself to go to MIT, because I saw the precursor to the first contest, which was a contest between MIT students, so it's made a huge difference in my life. I love going to the FIRST, and just feeling the excitement and the atmosphere. Kids excited about engineering and science, and kids excited about robots, I've got to love that.

FLATOW: Yeah. Dean, wasn't there a study done that looked at how kids participating in FIRST, were affected in the long run?

Mr. KAMEN: Yeah, it's so staggering that if I quote from it you won't believe me anyway. I'd encourage you to go to the First web site, and look at it, but basically, in every category of things like, will they stay in school relative to peers, and they - in quivalent school situation. Will they go on to college? Will they choose engineering careers?

In every category that your equivalent high school kids were matched between kids that participated in First, and kids that didn't, the difference wasn't the factor of three percent, four percent, five percent. It was a factor of three times more likely, four times more likely, six times more likely, to do the things that we all think they ought to be doing. The data is just so compelling, you'd have to be numb not to want to put First in every school you can reach.

FLATOW: We're talking about robotics this hour. Talk of the Nation Science Friday from NPR News, talking about the future of robots. Let me ask you did - Dean, to chime in on something we've been discussing, and that is personalities of robots. Do someone who invents robots or - your Segway Personal Transporter is very popular.

Do you think that you - as part of your competition, you might change it from sort of a who can knock the other competitor out, to who can make a more human-like robot?

Mr. KAMEN: Well, I was listening to some of that. I think what most people - and it amazes me that they don't understand, is just how good humans are...

Mr. MCOWEN: Absolutely.

Mr. KAMEN: At projecting their own likes and dislikes into almost anything, and making it anthropomorphic. That's why you don't give away your old shoes. They become part of you. So, it doesn't take much to make something look, quote, "real and animated and human." You know, kids start with teddy bears, and all you have to do is put a couple of squiggles of just the right shape on a white piece of paper, just to shape of that curve here and there, and oh, that's Donald Duck, or that's Mickey Mouse.

The fact that we can take very, very, very vague abstract references to something and, and make it anthropomorphic, should make it pretty easy to believe that if engineers actually intend to make stuff that really looks and acts human, do so, it will be very easy, in fact almost unavoidable, even if it's unintentional, to make people start to think that human - that machines have feelings, and are getting more and more life-like all the time.

FLATOW: So, it's - like you think it's unavoidable, so it's what you're saying?

Mr. KAMEN: I think it's unavoidable. I think it's, again, it doesn't matter whether you intentionally do it or not. It's just the ability humans have to project what they want. I think not to get into the religious side, but I think almost in that same game a great quote by Samuel Clemens, Mark Twain was, God created man in his own image, and man being a gentleman returned the favor.

I think we can project almost anything to be human-like, you know, a couple of pieces of soft material suddenly become a cuddly little living thing, and we even project what God looks like as a human with a beard and a staff.

FLATOW: Uh-huh. 1-800-989-8255. And of course, you decided to do a robotic competition, rather than a Math competition because you like to tinker with things?

Mr. KAMEN: No. I thought we would use robots because it is so easy to make them photogenic, and to put them in an environment that kids understand. A competition whether it's football, basketball or soccer. A robot running around, accomplishing a goal by scoring points, it just seem to us that in a culture that's obsessed with competitive sports, and visually exciting acts that happen quickly, why not get kids to do science and technology, by turning the science and technology into something that looks remarkably like other sports.

And as I said before, it doesn't take much to make a human see something that's not quite human, as getting close and being animated. So, we decided we could make the robot take the place of humans, and it would attract a lot of kids that might not otherwise think about science, and technology, and problem solving as a fun way to spend their time.

Ms. GREINER: Dean, you've been adding more autonomy to these robot contests?

Mr. KAMEN: Yea, for a lot of reasons. Number one, that's to a lot of people where robotics is going. It's all about the sensors, whether it's distance sensors, or object sensing, or differentiating things, or space, or people. And the sensors are getting better and cheaper. And the software is getting easier to, you know, drag and drop a whole program that can sense things, and distance, and speed. But we add that each year because we only want...

FLATOW: Hang on, Dean. We have to take a break.

Mr. KAMEN: Yeah.

FLATOW: So, at - as I most rudely do to everyone. So, stay with us. We'll be right back after this short break.

FLATOW: You're listening to Talk of the Nation Science Friday. I'm Ira Flatow. And we're talking this hour about robots, and robots that may have personalities, robots that are more autonomous with my guests Dean Kamen, inventor and founder of First Robotics Competition, Helen Greiner who is a cofounder and chairman of the board of iRobot, and Peter McOwen, who is professor at University of London, and Project Coordinator of the group Living with Robots and Interactive Companions. Our number, 1-800-989-8255.

And now when I rudely interrupted Dean, he was talking to us about a comment that Helen had made about the - his robots becoming more autonomous.

Mr. KAMEN: Well, I think, Helen brings up a great point which is, you know, what really will make robots seem to people to be totally animated, real living thinking things is the fact that they can respond to their environment, whether it's light, or sound, or distance, or smell, or temperature. And the sensors that are making that possible are getting simpler, and easier, and cheaper to make and sell. And the software that interface with them to have them do things, is getting simpler to implement.

And we, as she points out, each year try to put more, and more of that into our robot kits, because we want the kits to always be fresh and new and on the edge, and we want returning veterans to our program to always see new exciting stuff that raise the bar on what can be done, while we continue to keep the baseline of the kit accessible and fun for all the rookies that come in each year, that are playing the game for the first time. So, you need a very, very large dynamic range of capabilities in the kit, and sensors are where we've been focusing for the last few years, and I think we'll continue to focus for a while.

FLATOW: Uh-huh. Great. We're not going to bring in - we heard a little bit before about the contest, we heard a little bit of audio from the First Robotics Contest. And now, joining us is 18-year-old Grant Cox. Grant is a member of the Thunder Chickens. You heard that before. That was one of the winning FIRST teams. He's a senior at Eisenhower High School in Michigan where he's also - he also goes to a magnet school for Math and Science. He's on the phone with us today from Shelby Township, Michigan. Welcome to Science Friday, Grant.

Mr. GRANT COX, (Senior at Eisenhower High School and Member of Thunder Chickens): Hi. Nice to be here.

FLATOW: What made you join a robotics competition?

Mr. COX: Well, at the magnet school you mentioned, we don't really have, you know, football teams and such. But we are very involved in our robotics team. And it's just always been something, you know, I've always been interested in technology, and I'm a very competitive, outgoing person. And it just seemed like the perfect fit, so I gave it a shot and ended up loving it.

FLATOW: Tell us about your robot design. What did you have to do, or have it do?

Mr. COX: Well, each year they put out a new, GDC, puts out a new game. This year's game was called First Overdrive. It involves picking up a 40-inch exercise ball, and hurling it over a six-foot overpass, while driving around a track not unlike a NASCAR track. We came up with a pretty unique design for grabbing the ball with just these huge arms, and it can actually - at fully extended, it raises up to about 14 feet from a stow position of about five feet. And it was a big design challenge. They have different regulations that they use, and one of the main ones was that no horizontal dimension on the robot could be more than 80 inches. And the ball is 40 inches, so it's a big design challenge there.

FLATOW: That's interesting. In fact, we have a video of the competition on our site. It's and of your machine doing that sort of stuff. Say hello to Dean Kamen. You ever met Dean before?

Mr. COX: I haven't. Nice to chat on the phone with you, sir.

Mr. KAMEN: And that's a - congratulations for winning, and thank you for participating and, helping us raise the bar every year to make this more competitive and more exciting.

Mr. COX: Well, thank you for that.

FLATOW: How did you come up with Thunder Chickens?

Mr. COX: That's a common question. We started in 1999 and back then we were team Macomb - which is we are just a conglomerate of a bunch of school in Macomb County. And then we started to branched out form other teams and our team, the 217, became more localized to just the Utica community school district and the MSC you mentioned. So, we wanted to come up with the more personalized name so one of - we had several engineers from Ford Motor company, one of our major sponsors, and they were working on the Thunderbird engine and I guess the engineer joke was since it was slightly underpowered they called it the thunder chicken. And we just thought it as a fun name.

FLATOW: And where do you expect to go after you are done here?

Mr. COX: Me, personally?

FLATOW: Yes, you're a senior, you're heading to some place for robotics work?

Mr. COX: Yeah, I am trying to enroll in Michigan Tech in the fall for mechanical engineering. I had the opportunity to have a summer internship at Fanuc robotics and work with top mechanical engineers there. And it was really fascinating stuff. So I'm really looking forward to pursuing a career on that.

FLATOW: You know it is interesting that there is - it's almost, I think it's a WWF is an apt description, it's a sort of combination of a rock concert...

Mr. COX: Quote, unquote "homework" that we have to do and this year's Dean's homework was to get first more into media, and we've been really working hard to do that, First it's a great opportunity for students from all fields, from all, you know, social groups to get exposed to this technology that's engineering. And like I said, we are working to get more coverage, get it more into the public eye, and yes, it's becoming pretty mainstream, I think at least.

FLATOW: We have been talking about this hour before you joined us about making robots appear more human-like, give them feelings, socializing, would agree with that? Would like them to feel more like people?

Mr. COX: I think, I was listening in, I think Dean was right. That is a clear direction that things are going. I mean everybody seen the movies, you know, where robots takeover the world. But I really think - I've seen some of the stuff that Dean's group comes up with, I've seen some of the stuff, you know, groups like Darba come up with, and robots are becoming more and more intelligent, more and more life-like each day. And I definitely think that it's a positive direction, having a robotic companion, you know, I think in X number of years it's going to be just as common as having a cat or dog.

FLATOW: Helen, I know you do defense work with I am sure it's super-secret robotic stuff.

Ms. GREINER: Some of it is.

FLATOW: I am sure that if a robot is sitting next to you in a foxhole, you wanted it to be dependable trustworthy one.

Ms. GREINER: We delivered over 1500 robots to the military now, and they've been used on tens of thousands of missions, and one of the things I'm most proud is that they be credited with saving the lives of dozens and dozens of soldiers. They go out and they dispose of IEDs or improvised explosive devices instead of sending one of our troops up to these dangerous roadside bombs. So it's a great field for robots. That and the home robot field and the surgical robot field are the ones as a market has really been significant in the past few years.

FLATOW: Cameron in Tulsa welcome to Science Friday.

CAMERON (Caller): Hey thank you for taking my call.

FLATOW: Go ahead.

CAMERON: I am a member of a group called Humans United Against Robots and what we do is we make people aware of incidents like recently in South Africa, there is a military training accident where six soldiers were killed because the gun turned on them. And also I read this week about a military program in Iraq that was scrapped because the robot kept pointing at the soldiers. And I like to know what's going to be done to keep things like this from happening when these robots are in our homes.

FLATOW: That's breaking Asimov's first law of robotics, isn't it? You don't kill the person who invented you.

Ms. GREINER: Well, I think if the Roomba gets out of line, you pick it up and you smack it. Seriously, you know, you - they're less than five pounds, you turn them off, we don't envision any issues. They're built to be inherently safe. They won't fall down the stairs ever...

FLATOW: But they are not carrying a rifle either?

Ms. GREINER: We don't see a market for that in the homes.

FLATOW: But he had a serious question?


FLATOW: How do you...

Mr. MCOWEN: It's like any technology, if I can pipe in here. We are not planning to produce armed robots to live in the home, but if you look at any kind of technology, there can be accidents that happen with it, and technologies do go wrong, but we're not setting out to produce something that's lethal and dangerous. And I think it's, again, looking at the risk benefits analysis of that if we can produce something that is going to make people's lives somewhat better then its worth doing, but we obviously do it in the safest possible way.

FLATOW: Dean, I know that you spent your life helping people, you know, medically and physically to improve their lives as our baby boomers, all of us among them, get older and older, do you see more robotic products coming out?

Mr. KAMEN: Yeah, there is no question there will be more robotics projects and products and the caller is right that any product can malfunction, and depending on the size and scale and power of the product. Any malfunction of any product has a potential to do something surprising and sometimes that surprise can be harmful, but again as I said before, this idea that we imbue this things almost because of our own creativity with doing things for reasons or because of thought or soul or feeling, or happiness or anger, it's just something we have to get over. I mean, if you stood next to a snowblower that's got to have 10 horsepower to do its job and big blades to pickup stuff and crush ice and you accidentally get in front of it or if slips around on the ice and comes back and chops up your leg that's a terrible accident. But nobody would say there was malice involved that somehow this machine went through a thought process, turned around and ate your leg.

You would just say, the motor is connected to wheel. It was in drive. It went there, there was an accident and by definition, the machine that was capable enough to do what you wanted them to do, A, could end up doing B. And you have to be very careful to put in safety systems and interlocks and all sorts of things to prevent accidents. And a robot that can point the gun or do anything that the caller was talking about and needs to have the same interlock and he is completely right.

If something goes wrong with any machine it can do something bad, but you don't have to invent that it did it because somehow, unlike all the other machines we make, it was malicious or intentional. But machines who have lots of power can do lots of harm. You can - you know, a car gets out of control and kills about forty thousand people a year. It's not an evil car. It was something that we try to design better and better to eliminate accidents. He is exactly right, we need to do that, but to try to focus on the non-real emotional or intentional acts of machine is probably an inappropriate thing to focus on and unnecessary a complexity to add to the debate. It doesn't exist.

FLATOW: We are talking about robotics this hour, Talk of the Nations, Science Friday from NPR News. Dale in Saginaw, Michigan. Hi, welcome to Science Friday.

DALE (Caller): Hi. How are you?

FLATOW: Hi there. Go ahead.

DALE: Ira, I just wanted to say thanks for the show, and I listen to you every Friday on my way home from work. I am a teacher.

FLATOW: Thank you.

DALE: I teach sixth grade. And we have a robotics team, Team 49, in Saginaw and it's our 11th year.

Mr. KAMEN: Team 49, you're a real veteran.

DALE: Yeah, well not 16 yet. Yeah, Mr. Kamen, but we are getting there. We - it's been an amazing program for our school which is predominantly a minority school. We've had over a hundred graduates and out of that, 80 percent of them have gone on to study math, science, and engineering. In our regular program, fewer than three or four percent had gone on to study in math, science or engineering.

FLATOW: Do the other students look at your students as the geeks, you know, the weirdos, instead of joining the football team or the basketball or the soccer team or something like that?

DALE: Well, we have members of our team that on all of those, but we actually have more people try out - our first team that we have to try-out for all of our sports put together. Almost a third of our student body tries to get on our team every year.

FLATOW: And why is that?

DALE: I think the excitement. The fact that some of them see through our mentors from Dow Chemical Company that it's possible for them to be engineers or scientists and make a good living and be excited about what's going on.

FLATOW: Did they draw in any of the borderline students who might be interested but are afraid to think they're not smart enough to be involved in such...

DALE: We have. We do in fact, we have no grade points criteria for determining who can get on the team. What we look for is the excitement and the interest, and what's happened is that the kids that very many who are marginal have come in and been extremely successful.

FLATOW: Thanks for calling and good luck to you, Dale.

DALE: All right. Thanks.

FLATOW: 1-800-989-8255. Interesting story. I'm sure it's not unusual, Dean is it?

Mr. KAMEN: No, but I think the kind of data he just gave you, some of your listeners might dismiss as odd or a fluke, but you heard him, 80 percent of his students, over a hundred students have gone on from his school to careers in science and technology where we desperately need more people in this country and only three or four percent of the school kids in the same schools that didn't participate in first went on, and that is the kind of data we are getting from all the schools. And this year if you include (unintelligible) we have 13,911 schools involved. So, we are starting as - as a student pointed out to you, we are starting to become mainstream. And that's what I was intending all along and that's very exciting. And he is right, I asked, to see his homework assignment to get to media, that would be you. And help get our story out so that we can get more parents, more teachers, more companies, more sponsors, more foundations to take a look at the earth and get involved.

FLATOW: Grant, what kind of robotics would you like to be involved in? Any specific?

Mr. COX: No, nothing really at the moment, when I was at Fanuc, I was working on production line robots, stuff like for car assembly lines, industrial-type things, and I found that pretty interesting. They were, you know, fully autonomous paint boost robots and et cetera. But I just, I enjoy the engineering aspect of things, you know, I, wherever, my future employer takes me.

FLATOW: Well we got to have send you one of our sci-fri pocket protectors, when we get them handed out. I thank all of you for taking time to be with us. Dean Kamen, founder of FIRST Robotics Competition for young inventors, Grant Cox, a member of the team Thunder Chickens recently crowned world champions of the FIRST Robotics Competition. Also Helen Greiner, co-founder and chairman of the board of iRobot, Peter McOwen, is professor at the University of London and a project coordinator for the group Living with Robotics and Interactive Companions. Thanks for staying up late for us, Peter.

Mr. MCOWEN: Thank you.

FLATOW: All of you have a great weekend, thanks for taking time to be with us. Greg Smith, composer our theme music, and we had help today from NPR librarian Kee Malesky. Surf up into our web site if you want to see that robotics competition. We have a terrific video up there. You can watch how the contestants won that contest. Also, you can email us at Science Friday and also at We have blogs and podcast and looking for your input and ideas about how to make Science Friday a better way. What kind of things would you like to do. Have a great weekend. We will see you next week. I'm Ira Flatow in New York.

Copyright © 2008 NPR. All rights reserved. Visit our website terms of use and permissions pages at for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.