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

 


Forgot your password?
Close
typodupeerror
×
Science

Autonomous Robot Explores Antarctica 170

Posted by emmett from the danger-will-robinson dept.
rhet writes "Here's an interesting article about an autonomous robot that is exploring the Antarctic looking for meteorites. So far it has found three of them. The robot, Nomad, is built by Carnegie Mellon. Nomad's home page is here."
This discussion has been archived. No new comments can be posted.

Autonomous Robot Explores Antarctica More

Autonomous Robot Explores Antarctica

Comments Filter:
  • "There are four computers on Nomad during this expedition. Two PCs running Windows NT control the panoramic camera, perform landmark based navigation, and run the autonomous classification software. A third computer running Red Hat Linux coordinates robot navigation and obstacle avoidance with the stereo cameras and the laser rangefinder. Finally, a VME processor cage with a Motorola 68060 processor controls Nomad's real-time processing, such as translation of driving commands into servo motor movements and the monitoring of all systems on Nomad."

    I found the above from following one of the links in the article. Very cool. Nomad is rad! I want one of these.
  • This is a shining example of why a human explorer will always be better.

    covering about 100 square feet daily.
    And on favorable terrain. I could cover that in 15m, in any kind of terrain.

    15 feet away, searching for anything that seems like a rock
    Now, in a terrain loaded with rocks, 15 feet might be just right. But when you hit a dry spot, and all the rocks are 16 feet from you... you need more range. Esp when it takes you 20minutes to move a foot.

    No mechanics or spare parts were taken along, either, since the aim is to make Nomad completely autonomous.
    At least Nomad doesn't need to run to the crapper ever few hours. But if they ever get it to assimlate surrounding materials in order to make spare parts, it probably will.

    ``The breakthrough technologies are robotic classification and search. Humans classify every time they sort pennies from nickels, and they search every time they lose their car keys,'' said William ``Red'' Whittaker, founder of the Field Robotics Center at CMU. ``But these are new skills for robots.''
    Ah... the entire reason for OpenSource... Use existing source code!

    ``It's quite moody actually,'' he said. ``Some days it's perfect, and others nothing will work.''
    I appologize, they did.


    I have never understood the reason behind robotic exploration. Sure, NASA didn't have to call the Mars Landers parents, but its not like we force people to make the choice. A human explorer is more capable of collecting information, moving around the terrain, and surving the experiance. And then said explorer goes from around the country, making money talking about what he did. And all the ppl "Oh" and "Ah", and all the kids say "I wanna do that when I grow up".

    I think Nomad is great, it took alot of hard work, and I don't see the problems overcome as being easy ones. And I personally wouldn't want to spend several years walking around a piece of ice looking for rocks when I could get a robot to do it for me. But robotic explorers are not the way togo... they should be kept to the menial work that we are to lazy todo.

    --Cam
  • Actually I think that NASA is funding this because the research will help them out alot with their planned mission (not sure if its officially planned but they've been saying they'd like to do it) to Jupiter's moon Europa. They want to employ some sort of underwater autonomus vehicle to break through the icy surface of the frozen moon because they think that there may be oceans down there that might contain life (I doubt it but its worth a shot and at least they'll come up with important research as to what lies beneath the moon's crust).
  • H.P. Lovecraft warned us about searching for funny rocks in Antartica. Get ready for crazy shit to wake up and destroy humanity...
  • NEWSFLASH!

    NASA scientists have merged the programmable robot technology with the "Real Doll" (see http://www.realdoll.com/) and have finally solved what has become known in Psychology circles as the "Geek Date Syndrome". The programming interface and the variety of body styles offered by the Real Doll company have finally made the dreams of Geeks everywhere come true!

    "Why my gal, Priss, loves to play Quake with me for hours and hack the Linux kernel!!!" said one enthusiastic beta tester of the new technology.

    The downside for many college students and others who may wish to purchase their own 'Dream Date"TM is that the high cost means that only those with the high salaries and stock options of computer programmers will be able to afford them.

    Get yours today!

  • With other planets, you don't have governments complaining about being spied upon. That's the only thing preventing us from doing that right now. (And I'm sure there are highly detailed topographic maps of the earth. They just aren't public knowledge.)
  • Given that off-the-shelf components are relatively cheap, it strikes me that it would be feasible to cobble together an open source rover for exploration of remote areas. Cameras and sensors are cheap. A ham radio satellite link can be used for communication. Just get someone to take the rover along on an Antarctic resupply mission (I live right by the company that supports the Antarctic research stations - I could talk to someone there), set it down and let 'er go. Even if no valuable data is gained from the project, it would at least be fun. And we all know that's the most important thing anyway.

    Heck, this could even lead to amateur-built, open source space probes ;) Make something small and relatively self-sufficient, and you could buy some space on the shuttle or an Arianne rocket. That's how hams get their satellites up. I think it would be WAY cool to have our own, cheap orbiting web cam :)
  • Yup, I'm *in* CMU, and I can't get to the server. I guess slashdot does indeed have a huge audience...

  • A bunch of penguins programmed to create the perfect OS find the damaged NOMAD and attempt to repair the programming of it. They redesign it to look for the perfect operating system and steralize all those that are not perfect.

    When NOMAD finds the imperfect OS'es Win9x and Win2K as well as MacOS and steralizes them. Captain Gatesvolds of the USS Blue Screen is sent to find out why his creations were destroyed. He uploads NOMAD onto the Blue Screen and NOMAD mistakes him for the creator. SpJobs, Gatesvolds first officer does a mind meld with NOMAD and finds out that the creator was Torvalds. NOMAD then erases the memory of SAM and kills unit Registry because it sees that they are not perfect. Gatesvolds tells NOMAD that he is the creator and that he is not perfect so NOMAD should destroy itself but NOMAD is too smart for Gatesvolds and kills him too because he wears unstylish glasses and is obviously not perfect. NOMAD then destroys the USS Blue Screen and goes in search of his perfect creator Torvalds and his first officer Cox on board the USS-Kernel Panic. NOMAD becomes a little more human when Torvalds tells him that Linux is not perfect but sharing is a lot more fun than hoarding.

    The End...

    FOR NOW...BUM BUM BUM...

  • Better yet, a team of Aibos(TM), pulling a sled...
  • It's nice to hear of something created in electronics for a mission that doesn't break on it's first run. Seems that every mission we hear about somthing breaks right away or is a failure, although that is probably due to our overly negitive and always there media:). Thats a great school out there in Pittsburg and this just might show it. Man, if things like this happen, NASA might start taking notes.

  • Famous Last Words (Score:3)

    by naloxone ( 142847 ) on Monday January 31, 2000 @02:34PM (#1317402)
    "If the Carnegie Mellon group finds a meteorite this year, that would be an achievement," said Ralph Harvey, a geologist at Case Western Reserve University in Cleveland. He heads the human team sponsored by NASA that has looked for meteorites in Antarctica for more than 20 years. "The robot is still having a hard time figuring which way is up." - From a Post-Gazette article at http://www.post-gazette.com/healthscience/19981012 nomad2.asp It found, what... 3 so far? Not bad for a first run.
  • You can run almost any operating system on a VME bus system. It's just a question of what operating systems are supported by the vendor of the CPU board. One of the labs that I work in has several VME bus based Windows NT systems.

    You may be thinking of the Forth based Open Firmware that is used on some PCI bus cards.

  • They need to explore other planets now, space travel would be my ideal search area.
  • On the same idea: with your figures, the robot has a density of 52 kg/m3, so I don't think it would even sink in water (density 1000 kg/m3), if it is waterproof.

    I think you are basing the above calculation on the assumption that the net volume of the Nomad is (2.4m)^3 == 13.84 m^3. But this is the "bounding box" volume... or the volume of a box that could just fit the robot when fully extended. In a deep puddle that robot would probably sink like a rock.
  • I'd like to see an Aibo lead an expedition to the south pole.

  • Re:Robots (Score:4)

    by Mr. X ( 17716 ) on Monday January 31, 2000 @01:50PM (#1317413)
    While space exploration sounds like a good idea, there are plenty of areas of Earth that haven't been explored yet. I've heard we know more about space than the Earth's oceans. Exploring the oceans most likely would be cheaper and yield more bang for the buck than NASAs current Mars programs.
  • Ofcourse NT can be set to reboot when the kernel crashes. And when setup properly it's unlikely NT will BSOD. I doubt anyone will be playing quake or installing fancy devices/device drivers on the box.

  • Sponsoring nomad (Score:3)

    by Captain Zion ( 33522 ) on Monday January 31, 2000 @01:52PM (#1317417)
    The search engine companies are missing a nice opportunity to sponsor nomad. That would make an ideal ad :)
  • This is actually a pretty straightforward computer vision approach. You use two cameras, and, since you've carefully calibrated the cameras and know how far apart they are, you can compute the distortion between the two images pixel by pixel. Since the cameras are in slightly different locations (separated by a fixed baseline and angle of difference), any disparity will be the result of the different angles of view of the two cameras.

    One interesting point is that the farther apart the eyes are, the more sensetive the apparatus is. So one way to get better depth perception is to put your eyes out on stalks.

    Here [cmu.edu] is a paper on fast stereo vision.

  • I guess that Marvin (the paranoïd androïd, of HHGTTG fame) is no longer the only robot with mental problems...

    (seen on the p age describing the robot [cmu.edu]) :

    > Nomad is a powerful computing platform. Its size allows all necessary
    > processing to be performed on the robot. There are four computers on Nomad
    > during this expedition. Two PCs running Windows NT control the panoramic
    > camera, perform landmark based navigation, and run the autonomous
    > classification software. A third computer running Red Hat Linux
    > coordinates robot navigation and obstacle avoidance with the stereo
    > cameras and the laser rangefinder.
    > Finally, a VME processor cage with a Motorola 68060 processor controls
    > Nomad's real-time processing, such as translation of driving commands
    > into servo motor movements and the monitoring of all systems on Nomad.

    Now, let's introduce the schizo robot!!!

    And I wonder what effet it will have to make the robot look through Microsoft-coloured glasses...

    (Now, I wonder that OS runs on the 68060)...
    --
    " It's a ligne Maginot [maginot.org]-in-the-sky "

  • Not sure how much I like that idea. I thought they missed what they were aiming at on Mars with the Mars Lander. Not by much, but people in Argentina might not be pleased. Of course if they miss that metric conversion thing again I might get hit even here in the US. Yes I mean this as sarcastic and only aimed at the part about landing things here from space. The rest of the idea is good I think.
    -cpd
  • the aiboditarod ?


  • Hmmm... how do they reboot the NT boxes when they BSOD? *grins*

    silly! thats why there are 2! a cron job on the linux box alternates the reboots.

    ======
    "Rex unto my cleeb, and thou shalt have everlasting blort." - Zorp 3:16

  • Nomad (Score:1)

    by BWS ( 104239 )
    Noo! This is a Bad Idea. Rememeber Star Trek when the Probe Nomad meets with an Alien Probe and their identifies merge?
    Instead of an perfect AI Robot seeking Alien Races, Nomad becomes an Robot who's goal is to destory inperfect biological organisms. That means our Nomad will merge with an dophin and become NOMAD, Goal to throw metroites at big tall Politicans. Then we'll need Shatners's great logic to stop it.

    NOooo!

  • Well, it's better than GORF!

    Space cadet, fight like a robot!


    ---
  • Well, many people around here would probably like to point out NT doesn't have Telnet. But they could have installed a 3rd party telnet program.
    Aparently the VME is supposed to be controlling the robot arm, but almost any system inbetween could have caused the trouble.
  • Yep, that was Dante. There's still some stuff up at NASA [nasa.gov] about it. Actually the other thread on "sending a robot to Hell" reminded me of Dante, too...
  • Yes, Xavier often has his manservant with him, but it's been a while since he's had to hit the kill switch. Xavier and Amelia both come to my office and annoy me, unattended. I'm in the same corridor and my door is often open (5303).

    Flo the Nursebot [cmu.edu] is an interesting new development. It uses speech recognition, synthesis, and face tracking, nominally intended as a "robotic assistant for the elderly." The 'lips' move when flo talks, and the eyes track the face of whoever it thinks it's talking to.

    Robotics still has a long way to go, but things are starting to get interesting. Robots get a lot more interesting to me when you can talk to them; sometimes i wish they would just shut up when i tell them to, though.

  • Maybe Nomad will turn up the Loc-Nar.
  • If you want to look over a bunch of robotics projects at CMU, here's a nice list [cmu.edu]. It's not complete, but there are a bunch of pictures of robots and links to more info.
  • Umm lets get this straight, this is a wheeled and armed robot with various science sensors and you doubt it would have any fancy devices or drivers?
  • I think you meant gassed, not gased. While I am certain you know how to spell, what you are indicating is that you have not taken the time to review what you wish to submit, let alone reflect upon the social ramifications of your suggestions. I only hope that you are joking, however unfunny it may seem to me, because if you must hate someone, then at least take an educated stance and hate individuals as you feel circumstances merit it.
  • It would certain make all those bot's that just flip you over a bit useless
    .oO0Oo.
  • Anybody notice the Star Trek Reference?
    I am Nomad. You are creator. I am Nomad.
    fatal exception in NOMAD.EXE at 034:FF:678
    ERROR! ANALYZE ERROR! ERROR! ANALYZE!

  • Now, let's introduce the schizo robot!!!

    Isn't that basically what happened to Nomad in the original Star Trek episode?


    ---

  • Re:I'll be impressed when... (Score:2)

    by Anonymous Coward

    Sounds like a BeoWULF sled!

  • Given the sucess of other Autonomous Robots at finding meteorites, The Fundie Special has been launched. Nearly identical to its competitor, it differs in only two lines of code:


    if(object == meteor && containsFossils(object))
    destroyObject(object);

  • my university has a strong research presence in antartica, and from what i know i'm sure Nomad would have a lot more charisma than most of them.

    at least Nomad won't spend it's time wondering why antartic stories don't impress the opposite sex!

  • That robot could be abducted by evil open-sourced penguins controlled by NASA so they can send another robot to Mars ;P
  • Before you never Forth with obviously programmed...
    --
    " It's a ligne Maginot [maginot.org]-in-the-sky "
  • I was talking more of video cameras etc ...
    however, those science sensors and stuff (the arms etc are controlled by the VME computer) are most likely interfaced with through the standard parallel and serial ports, not 'fancy' devices. Any 'fancy' devices they may have developed as ISA/PCI cards would more than likely have had device drivers written by them, thus i doubt they would have been as big a problem as poorly written drivers by 3rd parties.
    So yes, I do doubt they have fancy devices and drivers.
  • I think it is a complete waste of money to be sending a robot around in the antartic wasteland to search for space rocks (Remember that you could pay many people $50,000 per year to do something like this and they would be happy). A robot that can do this must cost an incredible amount of money and would probably also have trouble coping with the frigid cold of the Antarctic because of all the moving parts and powering system. This will require it to be rescued by real people and waste more money. My final point is that the robot will also pick up all sorts of crap it thinks is a meteorite wasting even more time.
  • Another possibility would be to grab some ice, melt it, and electrolyze H and O out of it. Use the H for a fuel cell.
    This will get you exactly nowhere. Just a hint: look up "conservation of energy".
    I wonder of you could make biodegradable blivets?
    Things don't biodegrade on the ice sheet; it's too cold for bacteria, and there's no liquid water. Photodegradation is another matter. The researchers pack out their trash (including feces), but they pee on the ice; solar UV breaks the urea down into CO2 and NH3. I suppose you could air-drop solid fuel, like coal or wax, but this still requires someone to fly out frequently. This is not particularly safe even for professionals, and well beyond the capabilities of a group of amateurs (which was my impression of the original idea).
    --
  • Uh... Not only that, but am why would any "meteorites" be laying around on the surface of the snow/ice?
  • I know the music industry is going crazy against MP3, but is it necessary to send 64MB of MP3 to the Antarctic? What is NASA trying to get away with, an illegal copy of "Enter Sandman" by Metallica?

    =]

  • It's not so much that the robot is polluting, but it is wasting money. In Minnesota, where I live, I can't start my car at -20F. So how does this start at -80F?!
  • Maybe Nasa should try to land some probes on the Antartic from Earth orbit

    Probably be a whole lot easier just to throw the thing out of an airplane. Of course, g, atmospheric density, and delta-atmospheric density, will all be different on Mars than in Antarctica, meaning rate of descent and impact energy will all be different, too.

    Does the lander bouncy-ball have any sort of cameras or guidance mechanism? It seems like a drogue chute, a couple of fins and a range-finder could do an awful lot of inexpensive good towards getting the thing to land someplace flat.

    --

  • Nomad (Score:4)

    by Ertai ( 134811 ) on Monday January 31, 2000 @02:59PM (#1317458)
    "Nomad is a powerful computing platform. Its size allows all necessary processing to be performed on the robot. There are four computers on Nomad during this expedition. ... A third computer running Red Hat Linux ..."

    "I am Nomad. I am perfect."

    "... Two PCs running Windows NT "

    "Error! Error! Sterilze error!"

  • wow (Score:2)

    by Anonymous Coward
    this robot can cover a 10' by 10' square a day. at that rate it will be done about the time the sun crashes into the earth.
  • > Now, I wonder that OS runs on the 68060)...

    It probably doesn't have an OS, as such. As far as I understand it, VME busses are programmed via a variant of Forth.
    If this is true, only the Forth interpreter would be running (bare metal).

    Of course, the use of Forth is enough to question its sanity.
  • Think a person on skis. Think a pond skater (the insect, not Tonya Harding on a picnic with friends). It's more to do with surface area/tension (I forget which). The higher the surface area, especially if rolling, not stepping, the less likely it is that this will be the case.

    Another example: your home pool. Put a long pole flat on the water (maybe the skimmer) - it'll last a few seconds. Put a large sheet of wood. See how long it lasts... extrapolate.

  • Check out this [asu.edu] site at NASA. It describes a satellite built by undergraduate students at Arizona State University. It was launched successfully on Jan. 27. I belive students could actually register for a class where the lab involved working on the satellite (I'm not sure...I'm EE, not Aerospace). In any case, the numbers are pretty impressive: more than 350 students worked on the program including a substantial number of undergrads. The launch vehicle also boosted a number of projects from other universities into orbit as well.

  • Check out Nasa's Website [nasa.gov], specifically the page on STS-99 [nasa.gov]. The shuttle will be fitted with a large arm, capable of creating high-resolution topographical maps of Earth. The mission is eleven days long and dubbed the "Shuttle Radar Topography Mission".

    They mention that one of the motivations for the research is the fact that we currently have better maps of other planets than of our own...
  • I seem to recall coverage on a legged robot called Dante, which was sent to investigate a volcanic crater in Antarctica. Could you be thinking of that?
    --
  • OK, sorry, but I gotta practis my journalism skills here: "being an american college student ... thought it was exciting to have a 'trampoline' program"

    ...oh, that's what you meant by "higher education"...
  • I never thought it would take until I was 30 years old for someone to actually make a decent use for a robot. I have to admit the time has finally arrived though. Maybe in another 30 years we'll actually make it to Mars.
  • Will Version 2.0 Find Date Robot help out those of us female geeks? I can deal with the nice legs and hot ass, but large breasts don't do much for me. ;) "The light at the end of the tunnel is inevitably the headlamp of an oncoming train." - Mrs. Murphy
  • In Minnesota, where I live, I can't start my car at -20F. So how does this start at -80F?!

    I'm guessing here, but given that it's running a gasoline powered generator (to generate power for the inwheel electric motors) that it never actually shuts the generator down, just throttles it back to a very low idle.

    As for pollution, the amount of pollution generated by this little guy (well, not so little, actually), is probably pretty low, especially when compared to all the other stuff that's already down there (helicopters, snowmobiles, tracked transports, etc.). Not something to worry about.

    Similarly, the cost of gas while idling is also pretty low; city traffic sucks consumption wise because of the stop and start; an idling engine is pretty fuel efficient.

    Scott Taylor

  • Maybe Nasa should try to land some probes on the Antartic from Earth orbit...


    This would be a very expensive option, if not a bad idea.

    The cost of getting ANYTHING into orbit is still prohibitive. Plus we would essentially have to build the rover to it's flight-hardened specifications since it will be spending some time in space, and this means more money.

    Deploying it from the Shuttle is another, probably less expensive option and certainly might be more feasible.


    There are cheaper alternatives to the Shuttle that are almost as good and far cheaper. You could use an obsolete ICBM. This would put it in a sub-orbital ballistic trajectory. This would get it out of the atmosphere, which should be sufficient to simulate orbital re-entry. This would be very cheap - since you are not trying to achive orbit, you probably wouldn't have to modify the ICBM at all, except for replacing the payload.

    Another possibility would be to drop it from a high-altitude aircraft. It was not widely publicized, but when the USAF mothballed the SR-71 (Blackbird) fleet, NASA got a couple of them (2 or 3, IIRC) for high-altitude / high-speed research. The SR-71 has hardpoints on the top of it's fuselage, which were originally intended to launch a pilotless drone aircraft into hostile airspace. The SR-71 flies so high that it's effecively in space. While it would not be a perfect simulation, it would be orders of magnitude cheaper than anything else.

    Finally, you could just drop it out of the bomb bays of a B-1B. By putting the plane in a steep climb on full afterburners up to it's maximum altitude, you could "loft" the payload in to a ballistic trajectory. A small off-the-shelf rocket booster could kick it into an even higher altitude.
  • I have never understood the reason behind robotic exploration.


    Because sending humans out into space is hideously inefficient -- the vast majority of your mass has to be dedicated to keeping the people alive & somewhat comfortable. With a human, you need to send up food, air, water, life support machinery, safety equipment, etc, as well as any mission equipment and instrumentation you need. A robot probe does not need any of this parasitic load, and can devote more of it's mass to sensors and equipment to perform it's mission.

  • It can make some fairly simple yet effective walkers, that have no concept of what to do when they get there, or even have any concept of getting there. They just go, they don't do anything else. They are toys.

    So, walking across broken ground is a "toy" action? Degrading gracefully upon damage is useless?

    Hate to tell you this, but these are problems that the top-down school of AI has not been able to solve worth a damn. The military spent a fortune trying to get various AI labs at places like MIT and Stanford to build autonomous walkers- the AI labs failed dismally.

    Tilden's developed simple, robust walkers. Couple these with cameras and some basic logic and transmission stuff and you have a cheap robot explorer.

    So they can't learn. Who cares? They walk, and that's proven to be harder than most anyone imagined.

    Eric

  • The page also states that Nomad classified meteorite #12322 as "interesting". I wonder if others are classified as "funny" or "troll"...

    Given the rocks' likely (lack of) combustibility, they're probably not "flamebait". But how does it avoid nervous breakdown, knowing that each one of its positive ratings is thoroughly meta-moderated?
  • yes it is. but not for the reason you mentioned.
    Have a look at :
    http://www.cs.cmu.edu/~cil/v-source.html
    All the source is there including :
    BrainTech's Odysee Development Studio - Commercial program - A drag-and-drop environment for vision system prototyping and testing (a la Wit,
    Vision Blox, Khoros, etc.). It is also an open architecture so that one can import their own C/C++ functions. (by Ajay Sidda / Odysee / BrainTech)
    Camera Calibration - Routines for calibrating using Roger Tsai's perspective projection camera model. (by Reg Willson / CMU)
    Convex grouping algorithm - Robustly locates salient convex collections of line segments in an image.
    Edge list approximation code - From Nonparametric segmentation of curves into various representations, PAMI 1995 pp 1140-1153. by Paul Rosin and
    Geoff West.
    GSnake - Contour modeling, extraction, detection, and classification.
    Hidden Markov Model routines - Implementation of Forward, Backward, Viterbi and Baum-Welch algorithms. The code follows Rabiner and Juang
    notation. Written in c. (by Tapas Kanungo / Center for Automation Research / University of Maryland, College Park)
    Intel Video Capture Card Libraries for Linux - Research libraries for the Intel Create & Share Camera Pack and the Smart Video Recorder III to be
    used with C, Matlab, or Java on a Linux system
    Facilitates high speed image capture into Matlab and C, and slower capture into Java. (by Jeff Norris / Learning and Vision Group / MIT)
    KLT - An implementation of the Kanade-Lucas-Tomasi feature tracker. (by Stan Birchfield / Stanford Vision Lab / Stanford University)
    Logical/Linear Operators (by Lee Iverson)
    MMach - A Mathematical Morphology Toolbox for the Khoros System
    Maximum-Flow Stereo Algorithm - Code for the maximum-flow formulation of the N camera stereo correspondence problem. (by Sebastien Roy / NEC
    Research Institute)
    MeasTex - A framework for quantitative measurement of image texture classification algorithms.
    MegaWave - Wavelet, Snake and Segmentation source code.
    Microsoft Easy Camera Calibration Tool - a flexible camera calibration technique, which only requires the camera to observe a planar pattern shown at a
    few (at least two) different (unknown) orientations. (by Zhengyou Zhang / Vision Technology Group / Microsoft Corp.)
    NIST Handwriting OCR Testbed - OCR software and datasets for UNIX systems.
    Perceptual Organization Software
    Performance of Optical Flow Techniques - Implementations of a number of optical flow algorithms as well as test data and results.
    These programs are described in John Barron, David Fleet and Steven Beauchemin, Queen's University Tech Report RPL-TR-9107, July 1992 (revised July
    1993).
    SAMPLEX Color Classifier - Demos and software (requires licensing) (Purdue University)
    SRI Stereo Engine Software - fast stereo software for PCs. It performs disparity calculations and filtering in real time on images up to 320x240 in size. (
    SRI Artificial Intelligence Center / SRI International)
    SUSAN - Low-level image processing.
    SUSAN is an acronym for Smallest Univalue Segment Assimilating Nucleus. The SUSAN algorithms cover image noise filtering, edge finding and corner
    finding. (by Steve Smith / Oxford University)
    SatherVision - An object oriented framework for artificial vision.
    Includes low-level operations (FFTs, convolutions), pipes, Tk widgets, stereo and optical flow modules.
    Segmentation of Skin-Cancer Images - Implementation of an algorithm for segmenting images of skin cancer and other pigmented lesions (see Image and
    Vision Computing, January 1999, pp. 65-74).
    An automatic method for segmention of images of skin cancer and other pigmented lesions is implemented. This method first reduces a color image into an
    intensity image and approximately segments the image by intensity thresholding. Then, it refines the segmentation using image edges. Double thresholding
    is used to focus on an image area where a lesion boundary potentially exists. Image edges are then used to localize the boundary in that area. A closed elastic
    curve is fitted to the initial boundary and is locally shrunk or expanded to approximate edges in its neighborhood in the area of focus. Segmentation results
    from twenty randomly selected images show an average error that is about the same as that obtained by four experts manually segmenting the images. (by L.
    Xu, M. Jackowski, A. Goshtasby, C. Yu, D. Roseman, S. Bines, A. Dhawan, A. Huntley / Intelligent Systems Laboratory / Wright State University)
    SketchUp - A demo package for recognizing hand-drawn sketches through Size Functions. ( Vision Mathematics group / University of Bologna)
    Steerable Pyramid
    Check Eero's home page for a tar file. (by Eero Simoncelli)
    TargetJr - A C++ Computer Vision Environment - C++ programing environment with libraries to support: image processing; image segmentation;
    camera modeling; 2-d and 3-d geometry; a graphical user interface based on FRESCO.
    TargetJr has been developed over the last 10 years, starting at GE's Corporate R&D Center. Currently TargetJr is used by a number of vision research
    groups with emaphasis on geometric algorithms and object recognition. TargetJr is written in C++ and organized into a number of libraries including:
    numerics; spatial objects; image; image processing; segmentation; computational geometry; 3-d modeling; and user interface. (by Joseph Mundy, William
    Hoffman, Andrew Fitzgibbon, Peter Vanroose and Rupert Curwen / GE Corp. R&D, Oxford University, University of Leuven)
    ToolDiag - Pattern recognition of multivariate numerical data.
    UNL Fourier Features (UFF) - An implementation of a general purpose 2-D shape description method. (by Thomas Rauber)
    University of Calgary vision software - Includes chain code, Hough transform, and others.
    ZipPack Polygon Mesh Zippering - Combines several range images into a polygonal mesh.

  • There is no snow. (Score:3)

    by Tau Zero ( 75868 ) on Monday January 31, 2000 @03:15PM (#1317487) Journal
    It can't get stranded in snow, because there's no snow there. Just ice. Nomad is running in an area where the ice is actually evaporating; if snow was accumulating, the meteorites would be buried and thus extremely difficult to detect. The beauty of the conditions where Nomad (and other Antarctic meteorite hunters) operate is that the sublimation of the ice sheet exposes all the things which have fallen onto it over the hundreds of miles and thousands of years before. The ice flows like a river and brings everything to that one area, like streams washing gold dust into every crevice in the stream bed. To a meteoriticist, this is just like panning for gold
    --
  • Maybe NASA should try blowing a big crater somewhere in the Antartic, dropping the robot into it, and then seeing if they can get a response. From a robot on Mars :)
  • .. it's called Nomad because it wanders the ice?
    --
  • This stuff is great - it will lead to the sort of technology we will need to allow autonomous robots to explore other planetary bodies, such as Mars or the Moon, but if it breaks down we can just walk up to it and figure out what broke. We can work out the kinks here where we can fix it.

    NASA has been sponsoring this type of research for about 2 decades. But we still can't plan for all of the variables involved in getting a rover to the surface of an alien world, and its stay on the world.

    Maybe Nasa should try to land some probes on the Antartic from Earth orbit...

    This would be a very expensive option, if not a bad idea.

    The cost of getting ANYTHING into orbit is still prohibitive. Plus we would essentially have to build the rover to it's flight-hardened specifications since it will be spending some time in space, and this means more money.

    Deploying it from the Shuttle is another, probably less expensive option and certainly might be more feasible.

    However, the amount of flack you are going to receive from people like environmental groups for the potential harm to the Antarctic environment (if it crashed or something else horrible happened) would mean something like this would never fly. I'm not sure I don't agree with them.

  • Actually, we have very few European meteorite samples before the 18th century. Why you ask? Scientists said "rocks don't fall out of heaven", and got various large collections destroyed.
  • I found it interesting that Nomad can use two cameras for depth perception. According to the site, it creates a "disparity map" based on the differences. Does anyone know where I can find more information on this technique? It sounds fascinating...
  • I happen to know a few people who worked on the robot at the robotics institute and i know that most of the computers there have been dual-booting redhat for many years now. I'm not an expert, but i think they had a 4.1 installation running. I'm not sure when that was since i only started with 5.0, but maybe someone else knows.
  • I agree it is pretty high, but how stupid is it spending 2 million for 30 seconds to advertise a .com address during the SuperBowl, by the way I forget almost all of those stupid .com commercials 30 minutes after the SuperBowl was over with, what a waste of money. I did remember that e*trade one with the Monkey and the Bud one with the Dog, but besides those, a waste, in fact if almost all those .com commercials had put their money together we probabley could of almost launched another shiny metallic thing in space(sarcasm ;). )
  • fyi

    meteoroid: a rock from from space that is in space

    meteor: a rock from space that is in the atmosphere

    meteorite: a rock from space that has landed on the earth

    Laplace

  • ...better names for these robots?

    Mars: Sojourner
    Antarctica: Nomad

    We get the point already!

    What about R2-D2? Or maybe Annihilator 9000? Ultron? Daneel Olivaw?



  • Your ideal search area for meteorites is outer-space? Egads man - what a breakthrough!! ;)

    -----------

    "You can't shake the Devil's hand and say you're only kidding."

  • ``If it fails, it fails,'' said Apostolopoulos. ``We need to know why it failed.''"
    Maybe I'm alone here, but isn't it time NASA took the same approach with the Mars probe?


    Pablo Nevares, "the freshmaker".

  • The expedition (Score:4)

    by pnevares ( 96029 ) on Monday January 31, 2000 @01:58PM (#1317505) Homepage
    Here [bigsignal.net] is the official website for the expedition.


    Pablo Nevares, "the freshmaker".
  • The amount of research done by colleges always impresses me, but one thing I've noticed is that they're really never doing anythign up in space. I think colleges could possibly link up and actually get something done out there in conjunction with NASA. Would it not be a nice step towards more space exploration? Colleges know mucho!

    And back to the article... wouldnt' this be a perfect place to put a super-overclocked processor in the robot? :)
    - Mike Roberto
    -- roberto@apk.net
    --- AOL IM: MicroBerto

  • BEAM (Score:3)

    by craw ( 6958 ) on Monday January 31, 2000 @03:23PM (#1317508) Homepage
    The Smithsonian magazine had a very nice article about Mark Tilden and the "chaotic" robots that he makes at Los Alamos. These robots are called generally refered to as BEAM (Biology, Electronics, Aesthetics, Mechanics). They are relatively simple in design and are made of parts that can be easily purchased. But the nice thing about them is that can exhibit learning capabilities and seem like they are alive.

    If you have some spare change, then check out Solarbotics [solarbotics.com]. If you want to build one from scratch, then go to some search engine, and search for Mark Tilden (that should get you going). So get out your soldering iron and electrical tape, and get to work making your very own robot.

  • There's just one thing left: find someone to go to the probe every so often, and fill it up with gas. (Nomad is gasoline-powered and cruises at about 1 MPH.)

    There might be alternatives. There's a reasonable amount of wind on Antarctica, and enough sunlight during the summer months that the meteorite hunters who camp out there use solar panels to charge their laptop batteries. Running a mobile probe off this might be challenge, though. Unlike the gas-powered version you have a much smaller power supply, no waste heat for thermal control, and a host of other added constraints. Anyone trying to do this would have their work cut out for them.
    --

  • Fortunately, that would run afoul of international treaties.
    --
  • This is a good example of the limitations of autonomous robots in natural terrain. While things have improved from 20 years ago, where mobile robots were first tried, it's still pathetic compared to what a pair of geologists with appropriate transport could do.

    While research into mobile robots is very valuable stuff, you're still going to need people if you want to do real exploration on bodies out of radio range of Earth.

  • it would be a good idea, but getting anoything into space costs millions of dollars. ($80 million ish?)

    since this was such a low budget, and other projects like it, they can only afford the one ride.

    IMHO, we should scrap pretty much everything except research into finding a half way decent meathod of getting into orbit. $80 mill a pop is plain stupid!
  • Good lord I hope it's got some defenses! I heard mars rock is the single most expensive rock on earth. This raises a problem with robot independence - any free roaming robot is going to be stolen or vandalized, or at least teased until it breaks.
  • Err.. Offtopic? Moderators, don't moderate people down if you don't know what they're talking about. Just cuz *you're* a little slow, doesn't justify bad moderation.

    Aibo - The Sony dog.
    Iditarod - The dog sled race.

    I thought this was a little funny... The moderator that knocked it down should at least have the decency to reply to the thread to nullify the moderation.
  • <sincere>I bow to your superior and more exact knowledge ;-)</sincere>
  • In Colorado, I have skied in snow that seemed to have zero density, but my skis floated on that. I think a properly designed robot could roam indefinitely in Antarctica.
  • RE: Power - I like the wind generator idea. Another possibility would be to grab some ice, melt it, and electrolyze H and O out of it. Use the H for a fuel cell. Another possibility: air drop blivets of gas and let the rover find them and fuel itself. I wonder of you could make biodegradable blivets?

  • Glad to see NASA funding this (Score:4)

    by joshv ( 13017 ) on Monday January 31, 2000 @02:00PM (#1317527)
    This stuff is great - it will lead to the sort of technology we will need to allow autonomous robots to explore other planetary bodies, such as Mars or the Moon, but if it breaks down we can just walk up to it and figure out what broke.
    We can work out the kinks here where we can fix it.

    Maybe Nasa should try to land some probes on the Antartic from Earth orbit, while simulating the kinds of communications delays we have with a distant probe. Seems like it would be a lot cheaper, and we would learn a heck of a lot more even if the mission failed. One of the biggest problems with the recent Mars mission is that we have no idea what went wrong. If we did it might have been worth the $165 million we blew, at least we would not make the same mistake twice.


    -josh
  • This is _semi_ related. For more information on Robots in the news, check some of the below links:
    CNN - Second robot to work on recovery of EgyptAir 'black boxes' - November 8, [cnn.com]
    Biped robot research in the world [caltech.edu]
    Enjoy.

    Ben Brewer
    brewer@nullified.org

  • Slight Variation: (Score:4)

    by Skyshadow ( 508 ) on Monday January 31, 2000 @02:02PM (#1317530) Homepage
    NASA Scientists today unveiled a robot designed to find a date for lonely NASA Scientists.

    "Quite simply, the robot first searches for certain characteristics," said NASA researcher Jason Godfrey. The robot locates candidates based on a number of factors, including "nice legs, large breasts and what we technically refer to as a 'hot ass'".

    Once a viable candidate is identified, the robot moves in for a much closer look.

    "First, the robot takes a photo using a special stereoscopic camera and relays back to base. This allows us to filter out unsavory characteristics like too-large noses, unibrows and pimples." Godfrey continued, "if a candidate passes this test, it is further studied for compatibility," including hobbies, interests and ability to perform the Vulcan 'Live Long & Prosper' hand sign. It also attempts to collect information, such as name, weight, and phone number.

    Although the robot is capable of parsing up to three blocks of downtown bar area per hour, the robot has yet to locate a single candidate which worked out. "Our most promising subject turned out to be a total bitch," said Godfrey, based on the fact that she gave a "fake-o phone number".

    ----

  • The problem was money: NASA couldn't afford the extra $5 million. Think about that - for less than the cost of a minute of SuperBowl advertising, we would have known exactly what went wrong - all we needed was an omnidirectional transmitter with its own power supply (and the weight penalty would have added the $5 mil). Oh well - maybe next time Congress will allow a few more dollars.
    Face it - this is rocket science, and it doesn't come cheap.

  • (Now, I wonder that OS runs on the 68060).



    It runs vxWorks. (Hence the real-time processing, being an RTOS and all.)

  • check this out.... (Score:2)

    by Anonymous Coward
    And I quote: "Nomad is a powerful computing platform. Its size allows all necessary processing to be performed on the robot. There are four computers on Nomad during this expedition. Two PCs running Windows NT control the panoramic camera, perform landmark based navigation, and run the autonomous classification software. A third computer running Red Hat Linux coordinates robot navigation and obstacle avoidance with the stereo cameras and the laser rangefinder. Finally, a VME processor cage with a Motorola 68060 processor controls Nomad's real-time processing, such as translation of driving commands into servo motor movements and the monitoring of all systems on Nomad."

    Hmmm... how do they reboot the NT boxes when they BSOD? *grins*

  • Images from Nomad (Score:3)

    by naloxone ( 142847 ) on Monday January 31, 2000 @02:09PM (#1317546)
    There's a fair amount of information about an earlier expedition with the Nomad robot in the Atacama Desert (including archives of some images from and of it) at http://img.arc.nasa.gov/Nomad/nomad.html

    Does anyone else remember a NOVA special on this robot? Or was it about one of the many previous robots from Carnegie Mellon?

  • Nomad in the NYTimes (Score:3)

    by DaKrzyGuy ( 25850 ) on Monday January 31, 2000 @02:16PM (#1317547)
    There was an article on the front page of the Science Times of the New York Times a few weeks ago. The link can be found here [nytimes.com]. The robot is very cool. And what is really cool is that they allow undergrads to help work on it too.

  • But what does it look like? (Score:3)

    by Wah ( 30840 ) on Monday January 31, 2000 @02:17PM (#1317548) Homepage Journal
    Oh, for that go here [cmu.edu]

    Even includes pictures of microscopic interstellar life, err, rock.

  • Delivery Mechanism (Score:4)

    by chazR ( 41002 ) on Monday January 31, 2000 @02:18PM (#1317549) Homepage
    "A highly sophisticated, rather delicate, very ingenious robot that is a credit to its developers and a triumph of Human ingenuity is being tested in the antarctic"

    "It was transported there with great care by a number of highly skilled scientists and engineers, who will monitor its every step with loving care."

    "Then we'll stuff it on the top of a rocket, do our level best to shake it to bits on launch, expose it to unimaginably rapid changes in temperature and leave it alone for 9-12 months"

    "Once we're bored with that, we'll plunge it into the atmosphere of a small planet, allow it to hit the ground in way that will make it lucky not to leave a significant crater. Then, whatever happens, we'll spend the next several months bombarding it with incredibly faint radio signals, just to see what happens. Either way, we're going to need more funding."

    Apologies. This is capable of being one of the peaks of human endeavour. But I'm glad I'm not the robot.

  • Re:check this out.... (Score:4)

    by Captain Zion ( 33522 ) on Monday January 31, 2000 @02:22PM (#1317551)
    Two PCs running Windows NT control the panoramic camera, perform landmark based navigation, and run the autonomous classification software.
    Humm... here's another quote from the page: "Between January 22 and January 26 Nomad autonomously found and correctly classified in-situ 3 indigenous meteorites. Nomad also found but incorrectly classified a fourth meteorite as terrestrial rock." :) The page also states that Nomad classified meteorite #12322 as "interesting". I wonder if others are classified as "funny" or "troll"...

Slashdot Top Deals

FORTRAN is not a flower but a weed -- it is hardy, occasionally blooms, and grows in every computer. -- A.J. Perlis

Close