Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Technology

Social Robot? 192

smashr writes "Researchers are currently putting the finishing touches on robots that will be attending the AAAI (American Association for Artificial Intelligence) conference this year as part of the AAAI robot challenge. In addition to robots wearing tuxedos and serving drinks, several robots designed to actually register themselves will be participating in the conference. One such robot is GRACE, being built by Carnegie Mellon University and the Naval Research Lab (among others). GRACE features a digital face and speech recognition to interact with people attending the conference. (She even runs Linux!) Her goal is to register for the conference, give a speech and answer questions. Stories at: Pittsburgh Post-Gazette, CNN.com, and USA Today."
This discussion has been archived. No new comments can be posted.

Social Robot?

Comments Filter:
  • *blush* (Score:3, Funny)

    by Zaphod B ( 94313 ) on Friday July 26, 2002 @03:27PM (#3961098) Journal
    Gosh, I wonder what else lonely geeks will be developing robots for!
    • Well, they did design it to be female, and they admitted the team is almost exclusively males, so you can't be too far off. However, if you looked at the picture, GRACE is anything but what you would want to be doing that kind of activity with. She looks more like an oil drum with a monitor stuck on top of it than any female I know. Unless oil drums have become suddenly attractive to us males, I don't think what you're suggesting is going to happen.
      • Actually, that oil drum looks more like an industrial strength vacuum cleaner (wet-vacs). You know, the kind that could..., ummm, I better stop here.
    • Re:*blush* (Score:3, Funny)

      by fobbman ( 131816 )
      Betcha next years /. "meatings" will be better attended by robots than this years organic model attendance.

    • http://www.sexuality.org/l/fetish/robofaq.html
    • I don't think www.fu-fme.com will be a joke for very much longer.
  • Ro-bot (Score:5, Funny)

    by Skyshadow ( 508 ) on Friday July 26, 2002 @03:27PM (#3961105) Homepage
    Robots with social skills? Dear God, they've surpassed 80% of the population of /.!
    • Re:Ro-bot (Score:2, Funny)

      by Wirr ( 157970 )
      It depends.

      I mean she runs Linux. Can't you just imagine her telling everybody that she told you a thousand times that you shouldn't use Windows and that the RIAA is evil ?

  • it's too perfect! (Score:3, Interesting)

    by krog ( 25663 ) on Friday July 26, 2002 @03:28PM (#3961111) Homepage
    maybe one of these robots could give Bruce Perens' illegal speech! what a setup!!
    • sure! why couldn't a robot like this be setup with an anonymous upload that could allow a tech show participant (like bruce) to load presentations. ok, so sometimes it might get loaded with a 'sex-ed' program. ummm.... so either way it's educational.... riiiight.

      but really, who are 'they' going to prosecute?
  • Now my plan to take over the world can be completed! Excellent...
  • Is she available for dating? I have a Apple IIgs I've been trying to get hooked up. It's been rather depressed lately.
  • Since this thing is running GNU/Linux... I suppose someone could
    1337@h4x0r>$finger GRACE
  • (She even runs Linux!)
    A girl that runs linux? Rriigghhtt ;)

    everyone I know that runs linux is male (unfortunately).
  • Robots that will attend the AAA [aaa.com]
    (American Automobile Association)...
    If robots will really do that - they'll be really social robots!
    • [quote]
      Robots that will attend the AAA [aaa.com] (American Automobile Association)... If robots will really do that - they'll be really social robots!
      [/quote]

      Of course, you'll know the robots have gotten too social when they start attending AA.
  • woudl I like to finger that.... aw, nevermind.
  • Now even ROBOTS will get more dates than I do! Sigh... where's my copy of Akira...
  • "Her goal is to register for the conference, give a speech and answer questions."

    Given the audience, I wonder how many questions will be answered with, "No, I won't meet you in your hotel room."
  • A paradox (Score:4, Funny)

    by Tall Rob Mc ( 579885 ) on Friday July 26, 2002 @03:36PM (#3961196)
    Its an interesting paradox that the nerdiest of computer geeks are programming robots to interact socially...

    These people are the most qualified and least qualified at the same time!
    • Well at least we know the robots will never have sex and be able to reproduce to take over the world.
    • by fobbman ( 131816 ) on Friday July 26, 2002 @04:06PM (#3961445) Homepage
      People have been getting robots to do what is too unsafe for humans to do for a long time. Studying the insides of volcanos, defusing bombs, roaming the surface of Mars, and interacting with others in public all hold the same level of danger to geeks.

    • Well I hate to be anal about a joke, but the CNN article actually mentioned that the researchers hired Drama Students to help teach the 'social skills'..

      That is interesting in itself. I guess these geeks see their weaknesses quite well?
  • Now all those lonely computer geeks out there can finally get a date, even if she is just a robot. Oh wait I am one of those computer geeks... Doh! Minus the fact that she looks like something I hit with my car last night, she is still better looking then my last girlfriend which coincidentaly is CmdTaco's last girlfriend/sister.
  • When I saw the article title all I could think of was Bill Gates sipping soda at a cocktail party...

  • by Autonomous Crowhard ( 205058 ) on Friday July 26, 2002 @03:37PM (#3961215)
    Until it can do games and/or porn.

    (Did you ever notice that no one wnats to admit this. They always try for the educational uses for a while first)

  • AI (Score:4, Insightful)

    by Uruk ( 4907 ) on Friday July 26, 2002 @03:38PM (#3961223)
    While I'm sure this robot will be impressive in some respects, I really wish people would go ahead and segment the AI field out into two separate fields - "Cognitive Modeling" and "Cognitive Imitation". These types of robots which simulate high-level human behaviour that we realistically just don't understand how it works are essentially parlor tricks. They're entertaining, but do they really tell us more about how the brain works? I don't think so - they mostly push the envelope in the field of pattern recognition and imitation.

    In other words, when it comes to "AI", these things tend to be really heavy on the "A" and really light on the "I".

    Course that's just my opinion, I could be wrong.
    • by El Camino SS ( 264212 ) on Friday July 26, 2002 @04:09PM (#3961470)

      do they really tell us more about how the brain works? I don't think so - they mostly push the envelope in the field of pattern recognition and imitation.

      So pattern recognition and imitation are not considered a part of intelligence now?

      Wow. I better call my friends at every elementary school on the planet and tell them to radically change their teaching style.
    • Re:AI (Score:3, Interesting)

      by Pemdas ( 33265 )
      I disagree. AI may have this making-artificial-people mythos about it, but if you look at the direction of AI research in the past 20 years, that's not really what the field is about.

      There are exceptions; people are trying to make magical leaps over some percieved barrier, usually self-awareness. The cyc [cyc.com] project works along those lines.

      By and large, though, AI research occurs in little steps, and most of those steps are like this one. There is a need for a good understanding of what works and what doesn't work for smaller tasks before we're going to solve the bigger problems. What you seem to call "Cognitive Imitation" I would often call "Trying to understand cognition".

      In other words, when it comes to "AI", these things tend to be really heavy on the "A" and really light on the "I".

      I'd like to hear of an approach that you think is light on the "A" and heavy on the "I".

      • I'd like to hear of an approach that you think is light on the "A" and heavy on the "I".

        Adding variable amounts human genes to other mammal's chromosomes.

        (Although, I will admit that it's not all that light on the "A", and the "I" of the experimenter would be decidedly questionable.)
      • I disagree. AI may have this making-artificial-people mythos about it, but if you look at the direction of AI research in the past 20 years, that's not really what the field is about.

        There are exceptions; people are trying to make magical leaps over some percieved barrier, usually self-awareness. The cyc [cyc.com] project works along those lines.


        I also disagree with grandparent, but for other reasons. Cyc is an interesting expert system, but it hasn't made any magical leaps. Wallace's disparagements towards the project are plenty accurate.

        By and large, though, AI research occurs in little steps, and most of those steps are like this one. There is a need for a good understanding of what works and what doesn't work for smaller tasks before we're going to solve the bigger problems. What you seem to call "Cognitive Imitation" I would often call "Trying to understand cognition".

        Yeah. Exactly. Cyc is a little step too.
  • by drox ( 18559 ) on Friday July 26, 2002 @03:39PM (#3961226)
    ...they sent Vikia out into hallways to see if she could get passersby to stop and talk with her and if she could learn to predict the behavior of people.

    So it's a neat exercise. Is this really what we want robots to do? Recognizing human emotional states and predicting their responses from facial expressions and actions is one of the things humans do best. Why work at making a robot do it? It would seem to make more sense to design robots to do things that humans are BAD at, rather than having them try to do things we're GOOD at.

    • So it's a neat exercise. Is this really what we want robots to do? Recognizing human emotional states and predicting their responses from facial expressions and actions is one of the things humans do best. Why work at making a robot do it? It would seem to make more sense to design robots to do things that humans are BAD at, rather than having them try to do things we're GOOD at.

      Because people don't communicate to each other with a keyboard or mouse. If we want robots/computers/silicon-whatever to be able to perform tasks of any sort for non-experts, we need to be able to communicate with them more naturaly, and that means includes gestures and expressions.

      -Andrew
      • You could take that a step further, and say integrate an AI "friend" with your applications. It could say stuff like "Hey, It looks like you are trying to print a document, but you keep screwing up, let me help". You could even make the AI non-threatning looking, something like a little paper clip. Revolutionary!
      • Because people don't communicate to each other with a keyboard or mouse

        It's what I'm doing now.

        If we want robots/computers/silicon-whatever to be able to perform tasks of any sort for non-experts...

        I'm not saying that the short-term goal isn't a laudable one - or at least a stimulating intellectual exercise. I'm more concerned about the long term. Do we WANT robots performing tasks for non-experts? There are more than enough human non-experts around, and we keep making more. I guess it just seems more sensible to me to work on making robots that ARE experts. Particularly if they're made to be expert at things human have difficulty doing.

        ... we need to be able to communicate with them more naturaly, and that means includes gestures and expressions.

        Gestures and expressions are just another form of communication, and if you want to use them to communicate with your robot, be my guest. But gestures and expressions are extremely subtle (fractions of a centimeter in eyelid position can communicate volumes) and alarmingly variable, at least as humans use them. They also very subject to contextual cues. I expect that getting a robot to reliably recognize and interpret human gestures and expression will be far more difficult that getting it to recognize and interpret human speech. With speech we at least know most of the rules. We're still learning about nonverbal human communications ourselves.

    • So it's a neat exercise. Is this really what we want robots to do? Recognizing human emotional states and predicting their responses from facial expressions and actions is one of the things humans do best. Why work at making a robot do it? It would seem to make more sense to design robots to do things that humans are BAD at, rather than having them try to do things we're GOOD at.

      Hwah?

      Humans, *good* at social interaction?

      Sure, we're leading the pack species-wise, but I wouldn't go so far as to call humanity as a whole "good" at social graces.

      As proof, I submit any number of individuals employed in the "service" industry. Maybe one out of every ten has -good- social skills, and the primary role of a service person is to interact with the general public!

      I much prefer interacting with a cheerful, courteous automaton than with a surly, mumbling automaton, human or robotic...

    • by solarrhino ( 581267 ) on Friday July 26, 2002 @04:09PM (#3961469) Homepage Journal
      I happened to cross paths with someone who worked on Grace, and he mentioned that he taught it "how to ride an elevator".

      That's all he said, but I've been thinking about it ever since. Would I want to get in an elevator with a huge, industrial looking faceless machine? Okay, yes I would, but would my mom? No. What about a robot that has a face, but that stares grimly straight ahead? Or one that continually tries to make eye contact? Wouldn't those all seem creepy inside an elevator?

      The fact is, human behavior is highly context specific, and we feel uncomfortable and /or threatened when "people" around us behave improperly. Don't believe me? Try peeking over a bathroom stall sometime. That's why we need a socially aware robot.

    • The real question is why not do it?
      You can not just work on one part of science at a time, you must progress the whole pack slowly. Because a robot that can interact with us isn't just about a robot maid, its about the future, and how computers will soon be able to know what we want when we say something. Computers will not be so anal, when you give a command, it might not take it literally as proper english suggests, but how we actually talk. I might sense our emotions and understand someone is mad and doesn't mean what they say. There are so many applications for the future.

      Think of the star trek doors, how do they know when to open? It isn't motion sensored primitive stuff, since they will walk right by it, or close to it and it won't open, but it is based on a AI that monitors your motion, and walking patterns, maybe your conversation too. I know its just a show with guys pulling the doors, but thats what it is meant to be anyways.
    • by thrillbert ( 146343 ) on Friday July 26, 2002 @04:32PM (#3961624) Homepage
      Recognizing human emotional states and predicting their responses from facial expressions and actions is one of the things humans do best ...<snip>... rather than having them try to do things we're GOOD at.

      Bwuahahahahahahahahaa!!!!!

      Sorry.. but if we're so good at recognizing emotions and predicting responses from facial expressions, then why are there so many divorces? That alone is proof that we SUCK at recognizing each other's needs..

      On that note, maybe I'll start work on AICounselor_v1.0Beta.tar.gz now.. I'm sure GRACE would enjoy that module.

      ---
      With all things being equal.. well, no, that's boring..
      • On that note, maybe I'll start work on AICounselor_v1.0Beta.tar.gz now.. I'm sure GRACE would enjoy that module.

        Better yet, why don't you start work on AICounselor_Troi_v1.0Beta.tar.gz now! I'm sure WE would enjoy that module.

        GMD

      • Rhetorical question, I'm sure, but I felt the urge to respond anyway.

        Sorry.. but if we're so good at recognizing emotions and predicting responses from facial expressions, then why are there so many divorces?

        If we're so bad at it, why are there so many marriages? People are at least good enough at it to convince themselves and others that they'd be compatible as lifelong partners. Try getting a robot to do that.

        That alone is proof that we SUCK at recognizing each other's needs..

        Not necessarily. Humans are pretty good at recognizing each others needs, by both speech and non-verbal cues. They may not always be so good at meeting those needs though.
      • Correction, that should read:
        AICounselor_v1.0Beta.zoid.tar.gz

        -
      • ..ah you better rethink that statement because of the following: At Siggraph this year there was a neat paper about eye-movement simulation, and allthough the results weren't particularly impressive like in the rest of the papers at the (btw very great) conference, it showed one important message: context is everything! The speaker demonstrated that indeed, she had managed to make the eyes more or less move in a natural way, but despite her efforts, I think most of the people in that room would agree that it still looked bad. She kind of appologized for it and said that you realy ought to track the eye lids, the frowns, the cheaks and all that... to get a realy convincing AI eye simulation. On top of that, the eyes have specific behaviour that diverge wildly when the person is engaged in a social pattern, or when they are saying something. For instance: Studies have shown that one tends to look more to the left while thinking about social stuff, while looking the other way when thinking about some technical details. Things can of course be reversed, but these tendencies to stare into infitity in a particular direction exist. They come natural to us, but if you think about it it's rather odd.. Even males and females don't do the same things with their eyes.

        So, without wanntig to diss on your remark, because you are right in many ways: while we may seem to be doing rather average reading the face of someone, it is much more likely that it is due to the fact that we don't know the person that well, don't know how to give an appropriate response, instead of being unable to read their face, that causes all these human issues. Maybe one day we will be having stupid arguments with robots, who knows..

        Hozever, imho we're still a very long way from truely intelligent behavior, because a) you need lots of context and b) you need lots of noise. Both of them are impractical for current silicon. The best quantum computer can deal with just 8 bits.. not terribly impressive..

        cheers,
  • I would like to see a demonstration of this technology in person. Or if that isn't feasible, maybe they could demo it on Leno or something?
    • I attend Carnegie Mellon. Grace is often in the atrium entrance of one of the computer science buildings here. She tells jokes and asks questions. It's quite entertaining, and disturbing.

      Robots wander the halls around often. I'm oddly used to it.
  • by Dthoma ( 593797 ) on Friday July 26, 2002 @03:40PM (#3961242) Journal
    This has been done before [newscientist.com]. Back in 1999, five robots had to make their way around the American Assocation for Artifical Intelligence annual meeting in Orlando.
  • by Rupert ( 28001 )
    Air-conditioned, psychoanalysed
    You're very nearly human you're so well disguised.
  • Hmmm, unsocial geeks programming social robots. Now tell me if I am wrong but, dont you think if an unsocial geek and program a social robot, the geek himself has the capabilities of being social himself?
    • Grammar correction, sorry:
      Hmmm, unsocial geeks programming social robots. Now tell me if I am wrong but, dont you think if an unsocial geek can program a social robot, the geek himself has the capabilities of being social himself?
      • Re:Topic goes here (Score:2, Interesting)

        by Maniakes ( 216039 )
        Now tell me if I am wrong but, dont you think if an unsocial geek can program a social robot, the geek himself has the capabilities of being social himself?

        Being able to program a computer to do something well doesn't mean you can do it yourself. I've written programs that play chess much better than I can.

        Another factor here is the "dancing bear" effect. A dancing bear isn't impressive because it dances well. It is impressive because it dances at all. Behaviour that is amazingly social for a robot may be pretty dismal for a human.
    • Hmmm, unsocial geeks programming social robots. Now tell me if I am wrong but, dont you think if an unsocial geek and program a social robot, the geek himself has the capabilities of being social himself? Not without root access.
  • by Anonymous Coward on Friday July 26, 2002 @03:42PM (#3961254)
    How can they imply a machine as female when it does not even reproduce? Not only that, they have no sexual organs, nor sex chromosome! Most ironic part is, all the photo has shown MEN constructing the machines!!! If they really want to respect the other gender, they should have put female engineers at work instead of making a poor excuse of 'female robot'.
    • by Anonymous Coward
      Sorry, girls can't make robots. Just look here www.girlscantmakerobots.net [girlscantmakerobots.net]

      Hell, it 's down. I guess girls can't make web pages either.

    • by Anonymous Coward
      Bahahaha. You had me up to "female engineers". Yeah, honey, "engineer" me a beer out of the fridge. Thanks.
    • How can they imply a machine as female when it does not even reproduce?

      If a woman has cervical cancer and can no longer reproduce, is she no longer a female? Or if she loses her breasts to cancer, how about then?

      You are female if you act like a female.

      My dictionary lists this as one definition of female: "Characteristic of or appropriate to this sex; feminine."

      And why shouldn't males be able to develop a female robot? Guys stare at females all the time, they ought to know what they look like enough to make a female robot.

      By your logic, we should not have male gynecologists either.

      Yeah....
    • Isn't that a matter of gender versus sex? Do you assign gender to things in your life that can't reproduce?
    • by sielwolf ( 246764 ) on Friday July 26, 2002 @04:27PM (#3961590) Homepage Journal
      The Register had a good article on pronoun usage.

      In English the masculine is used in gender-neutral cases where the gender cannot be specified before hand ("The officer should have his standard equipment present").

      The feminine is used in cases of abstract personification ("The United States has her hands full") of a concept or an object (Freedom or automobiles).

      Much like a car, to give it "personality" would move it from the first case to the second. Thus when you give a generic machine a specific voice, it is (usually) female).

      But, as with all usage, it is up to the user. There is a strong push for gender neutral language in most things. The problem is that in cases that are purely up to personal taste, these rules apply.

      So you can't tell someone to not make their robot feminine. But you can tell someone to make their manual not masculine.

      And you seem to think their purpose was purely political (ie that there was a certain quota that needed to be maintained so they added a female robot to offset the male engineers). I really think it is much simplier than that.
  • Does it have the Genuine People Personalities feature?
    Share and enjoy!
  • Nothing will go wrong...
    go wrong
    go wrong
  • on my quest to find a replacement body, I eventually found true love with the tracker I hired to take me into "The Zone", E. (Edith) Johnson [imdb.com].
  • Did anyone else laugh out loud when then pictured a robot wearing a tuxedo? That's about the funniest thing I heard all week.

    On a side note, with the robot conference and the Milky Way mapping thing [slashdot.org], does anyone else get the feeling we're living in an Isaac Asimov book? (Go here [amazon.com] and put author: asimov AND title: (robots OR foundation) in the Power Search box and buy and read all matching items if you don't get the reference)
  • touch, finger, uzip, configure(you know what i mean *wink*), mount, force, make, umount, sleep
  • The REAL question all the guys wanna know down at the trailer park is can she REALLY suck the chrome off a trailer hitch.

    That is all

    --toq
  • Am I the only one who already sees this in reflecteted back to humans in say, 10-15 years? "In other news, Summer 2018 Natural Intelligence will have Dolphins Delegation with 10 minutes screech. There are some unconfirmed rumors that humans could be physically attending this conference too..."

  • How Long... (Score:4, Funny)

    by John Hasler ( 414242 ) on Friday July 26, 2002 @03:59PM (#3961389) Homepage
    ...before the organization is renamed "The American Association of Artificial Intelligences"?
  • I thought this would be a story about Slashdot readers...
  • She'd get along great with Gaak, the robot that escaped [slashdot.org] from its masters and made it to the parking lot. I see a wonderful romance developing, or at least a great Thelma-and-Louise chase scene.

    On a related note, wouldn't a battle bot capable of social engineering the judges be a great thing? "I protest! My opponent is cheating!" "Never mind that broken wheel, t'is but a flesh wound!" Of course, she'd have be toughened up and given a weapon of some kind; sarcastic banter vs. a wedge is hardly a fair fight.

    • wouldn't a battle bot capable of social engineering the judges be a great thing?

      That reminds me of a beer commercial where the bot opens a fridge full of ice-cold booze, the opposition manuevers their bot into position to grab the beer and bring it back, and suddenly the beer-bearing fridge pulls out a hammer from behind and whacks the other bot to smitherings.
  • AAAI

    Does this mean in 20 year we will be reading how the AAAI is sueing people for circumnavigating there robots copyprotection scheme?

  • Well, I have finally gotten my Social Robot web/ drink server up! It is running in high positronic memory with about a 4k footprint. This server can do about 2-3 dps (drinks per second), and handle approx. 4 conversations at a time.

    I also have plans to do a Quake server that will actually come to your LAN party and kick your ass in person!

    Please check it out! http://www.socialrobotservesu.org [socialrobotservesu.org] It might be unavailable right now, I'm having a barbeque!

  • To make it more autonomous, grace should be able to connect internet (especially google). So, when user asks a question, all it has to do to ask the same question to google. There will be some reasonable answer in the search results.
    • by shadowduck ( 203571 ) on Friday July 26, 2002 @04:35PM (#3961647)
      Actually, that is planned for the future. We currently have the ability to locate the AAAI badges that are given to all conference attendees and use OCR to extract the name of the person we're talking to (Swarthmore did that portion of the work). Next year, we'd like to try to schmooze with conference attendees about their current research by pulling down their current papers, c.v.'s, etc. off the web. In time, in time. :)

  • Wow. I mean, she may be nice, but put a bag over that head! In fact, put one over yours in case hers falls off!
  • Have a look at this! [cmu.edu]
  • "Lowtax - I misprogrammed it. I tried to do good, but the robot jumped up and pushed grandma's head and she started spitting and her teeth flew out and the robot shot sparks and grandma fell down the stairs onto my uncle.

    It was the worst Christmas ever.

    Corn_Boy - oh no! sparks, did anything catch on fire

    Lowtax - Grandma did, but I got a Pusher robot to shove her outside into the snow."

    Credit where credit is due [somethingawful.com]

  • several robots designed to actually register themselves will be participating in the conference

    Wonderful. First we have them registering themselves at conferences, the next thing you know they will register their own software installed on themselves.

    Does this mean in a few years I'll have to get my robot her own e-mail address for all the spam she gets?
  • I good indication of how far AI has advanced is the fact that GRACE is a brunette and not a blonde. In the next major advance, GRACE will don glasses and be renamed VELMA.
  • GRACE runs Linux or she runs on Linux?
  • by guttentag ( 313541 ) on Friday July 26, 2002 @06:05PM (#3962058) Journal
    Her goal is to register for the conference, give a speech and answer questions.
    I realize these are not easy tasks, but I have to wonder what challenging goals will be like in the future. Will we be reading an article on Slashdot.NET in 2050 that talks about a robot's goals at the conference including:
    • Driving itself to the conference in an inconspicuous, beat-up old 2040-model car?
    • Spilling its drink on an attractive woman's shirt and taking first-person-perspective photos of the cleanup efforts?
    • Fooling the attendees into believing it is a real person?
    • Fooling the other robots posing as attendees into believing it is a real person?
    • Fooling the attendees into believing they are real people?
    By those standards, registering oneself for the conference, giving a speech and answering questions seems easy.
  • Laser rangefinders? (Score:2, Informative)

    by iamroot ( 319400 )
    "Looking over photos of the convention center in Edmonton, one of Simmons' fears is the center's glass elevators.

    Grace uses floor-level laser range-finders to find her way around; the laser beams go straight through glass, rather than being reflected as they are with other obstacles."


    I can see the laser rangefinders at the bottom, but it also has a bunch of ultrasound transducers in the middle, and some other lens(laser scanner?) above those in front. The glass shouldn't be too much of a problem provided that the ultrasonics are working.

    Sonar does have some problems, but it works well enough on the robots I've seen/used/built. The ultrasound clicks they make can be annoying, but it works. At a university I toured once, they were demonstrating some robots, a heathkit one, and two custom ones which looked exactly like GRACE, execpt they had thermal, visible, and UV(I think) cameras, 24 ultrasonics, laser rangefinders, a 3D laser scanner on top, and no screen. The robots were roaming around, and greeting people. They didn't seem to have any problem avoiding obstacles, it was actually kind of funny when the heathkit said "Pardon me." to a supply cart:)
  • (She even runs Linux!)

    Yeah, blue screens of death are very socially degrading, aren't they?

    Still, what I wouldn't give to hear the robot screaming "Kernel Panic", "Kernel Panic", lift up a knife and start tearing up the place.

    On the other side of the room, a robot running FreeBSD lets out a war cry...

"The great question... which I have not been able to answer... is, `What does woman want?'" -- Sigmund Freud

Working...