Forgot your password?
typodupeerror
United States News

Help Find Steve Fossett 439

Posted by kdawson
from the spare-wetware-cycles dept.
An anonymous reader invites us to join in the hunt for the missing Steve Fossett using Amazon's Mechanical Turk. DigitalGlobe, one of Google's imaging partners, has acquired new high-resolution satellite imagery of the area where Fossett disappeared on Monday. The public can now go through this imagery and quickly flag any images that might contain Fossett's plane. Flagged images will receive further review by search and rescue experts.
This discussion has been archived. No new comments can be posted.

Help Find Steve Fossett

Comments Filter:
  • what's he wearing? (Score:3, Informative)

    by datapharmer (1099455) on Saturday September 08, 2007 @05:12PM (#20523273) Homepage
    What's he wearing red and white stripes? Seriously though, this is a pretty cool tool even if it is a bit ridiculous considering all the missing persons there are out there who get no attention...
    • Re: (Score:2, Insightful)

      by BungaDunga (801391)
      Well, most missing persons aren't going to be visible from the air, are they?
    • by searchr (564109) <searchr@gmail. c o m> on Saturday September 08, 2007 @05:22PM (#20523343)
      To be fair, most missing persons are hiding in bus terminals and seedy motels. Even if it sadly takes someone of celebrity, even someone whose personal hobby is to put themselves into ridiculous danger, to develop a new form of distributed wetware computing, it's still for the better.

      Maybe if someone had thought of this earlier, that unlucky family in Oregon wouldn't have been stranded in their car for a week. Or maybe, now there's a new option for the next time that does happen.

      Forget SETI-at-Home. I'd much rather play "FindTheLostPeople-at-Home".
    • Like who? (Score:5, Insightful)

      by Quadraginta (902985) on Saturday September 08, 2007 @07:26PM (#20524139)
      it is a bit ridiculous considering all the missing persons there are out there who get no attention...

      Did you just make this up because it makes you seem like a Sensitive and Thoughtful Person? Or can you actually name someone who went missing in the wilderness and "got no attention"?

      FYI, rangers and such take their jobs very seriously. So far as I know, everyone reported missing in the wilderness gets a full spare-no-expense search and rescue effort. They look for "nobodies" just as hard as they're looking for Fossett, and the dedicated folks who do those tough jobs would take great offense at your ignorant suggestion otherwise.
      • Re:Like who? (Score:5, Interesting)

        by ejito (700826) on Saturday September 08, 2007 @11:26PM (#20525609)
        They found SIX [google.com] separate undocumented plane crash sites while searching for Fosset... why is it they can find these plane crashes while searching for Fosset, but couldn't find any of those before hand?

        It's also quite illogical to ask "who" didn't get attention -- if they got attention, then we'd know who they were.
        • Re:Like who? (Score:4, Interesting)

          by zippthorne (748122) on Sunday September 09, 2007 @12:05AM (#20525809) Journal
          You can't look for people if you don't even know they're missing. If they were flying below the radar (not hard to do in the mountains), filed no flight plans, and left behind no one who cared about them enough to notice they're missing, then how exactly were the rescue agencies supposed to find them?

          Further, if they were doing all those things, what're the odds the undocumented wreckage contains remains of undocumented would-be workers or non-medicinal pharmaceuticals?
      • Re: (Score:3, Insightful)

        by h2g2bob (948006)

        it is a bit ridiculous considering all the missing persons there are out there who get no attention...
        Did you just make this up [...] Or can you actually name someone who went missing in the wilderness and "got no attention"?
        I can't name anyone who went missing and got no media attention. And I should know: I've been studying the newspapers for it every day.
  • by 10e6Steve (545457) on Saturday September 08, 2007 @05:14PM (#20523283)
    but I found Waldo!
  • Google Earth (Score:2, Interesting)

    by Rebelgecko (893016)
    You can also look at the imagery in Google Earth

    Viewing in Google Earth:
    If you wish to view images in additional detail, you can pull them up in Google Earth. To do that you must: Download and Install Google Earth. Open the following KML file: http://s3.amazonaws.com/fossett/geo-eye.kml [amazonaws.com] Cut and paste the co-ordinates found next to the image tile below into the "Fly To" box in the top left corner of Google Earth. For the best experience, you will likely want to turn OFF terrain by unchecking the "Terrai

    • by solevita (967690) on Saturday September 08, 2007 @05:19PM (#20523319)

      into the "Fly To" box in the top left corner of Google Earth.
      It sure is tempting to burn some karma now; could even mention the new flightsim in Google Earth...
    • Re: (Score:2, Redundant)

      by John3 (85454)
      Google Earth images appear to be seriously outdated. I did a search for my business and the image shows our building as it was in 2005 prior to a second floor addition. Did Google Earth update the images for the region where Steve's plane may have crashed? If not, there isn't much point to viewing that area using Google Earth.
      • KML files can be used to overlay external data. There are KML files for "live" cloud cover images, for example.
      • Re: (Score:3, Informative)

        by G Fab (1142219)
        Yes, Google Earth has updated images of the region. You can tell because they are obviously satellite photos and not overflight. Notice that everything is shot from straight up instead of the normal angle and also note the lack of color.

        Read the article, and you'd see that they explain how authorities helped facilitate new images.
        • by John3 (85454)
          I didn't RTFA so I'm at fault for that. However, the first section I looked at via Google Earth was in color...at least the trees were green.

          In my defense, I was so anxious to start looking for Steve that I immediately jumped to Amazon to start looking rather than stopping to read the article. Who has time to read articles when there is a man out in the wilderness waiting for us to Amazon or Google him?
      • I tried looking to see if the "Series of Tubes" Senator's house in Girdwood, AK had a small shadow or a long one, because it was jacked up two floors in 2000 [adn.com] as a favor from a local corrupt oil company. Unfortunately this [google.com] is the best you can get from Google Maps- a fuzzy satellite view.

        The map is different in Google Earth- there, you can see that each one of those short stubby little roads ends in a nice stately circle.
  • by jesco (598308) on Saturday September 08, 2007 @05:22PM (#20523341)
    I suppose there are already trained people looking at the images. From the Police, Fire-Department, or whatever organization handles these kinds of emergencies in Nevada. I stress the word trained because the satellite data definetely needs experienced eyes to look out for the right stuff.

    The article starts by explaining what to look for on these images. This is good, but to substitute for experience in looking at such images.
    • Well, then, any novices who join in here may gain some experience!
    • One day we'll be telling our children, "When I was your age, we actually had people comparing satellite imagery to find lost people!"

      Seriously, though, can't computers do this sort of thing more efficiently? I'm no expert on the state of image recognition research, but you think it would be good enough that a computer could pick out potential "hits" for further review by trained professionals, perhaps by searching for what looks like man-made objects in remote areas or comparing old imagery with the curren

      • Agree on two points,
        The One-click Mechanical Turk is highly inefficient as it requires loads of clicks and scrolling for each of the impossible small search areas provided.

        While something is better than nothing - what's the point of a large community effort if not to advance the technology, and maintain the technology so that in future cases, it can be deployed more effectively.

        If this is the best google can do - i'd sell their stock.

        AIK
        • Re: (Score:3, Informative)

          by Dan East (318230)
          I press the "End" key to scroll all the way down. Look at the image. Click Yes or No. Click Submit Hit. Three interactions per image.

          Dan East
          • by ozric99 (162412)
            I hit "End" and scroll all the way down. Look at the image. Attempt to click yes or no. It's grayed out. Hit Accept HIT. Hit "End" to scroll down. Now the radio buttons are working. I hit Yes or No and click Submit. I then scroll all the way down, look at the picture, don't see anything so go to click "No" - it's grayed out again. Close Browser.
      • most OCRs for images other than text need to be trained on a sample image. This type of rough, broken terrain makes for really bad sample images- every photo is different. You'd spend more time teaching the computer what wasn't a plane than you'd save using the computer. We use an image recognition automated inspection tool at my workplace to inspect chips for defects and as regular as most of our chip features are, we still haven't gotten it down to a really usable level of false positives.
      • can't computers do this sort of thing more efficiently? you think a computer could pick out potential "hits" for further review by trained professionals, perhaps by searching for what looks like man-made objects in remote areas or comparing old imagery with the current, updated samples.

        But what exactly are you looking for?

        Remember the "Face on Mars?" It is very easy to find significance in patterns that are pure chance. The computer can be no more rational and objective than its programmer.

        In World War T

      • by timeOday (582209) on Saturday September 08, 2007 @06:38PM (#20523823)

        Seriously, though, can't computers do this sort of thing more efficiently?
        Nope. The govt. has spent millions or billions on this problem over the years, but they still employ analysts to do it manually for the most part.

        What, you thought there was no interesting CS research left to do?

      • by pongo000 (97357)
        Seriously, though, can't computers do this sort of thing more efficiently?

        I'll have to disagree with the poster that claims this can't be done.

        Yes, it can. It's known as "pattern recognition." The problem is that for the most effective recognition, you need either before/after images (which would be normalized and then overlaid at appropriate registration points), or a series of pics fed to the algorithm, one of which contains the artifact you're interested in (supervised learning). The former is often us
    • by Hatta (162192) on Saturday September 08, 2007 @05:49PM (#20523511) Journal
      If amateurs can find new meteor craters [astroseti.org] with google earth, why not airplanes? How trained do you really have to be to spot an oddly shaped bright feature in otherwise mundane terrain?
    • It probably can't hurt, but you're right in that it can be difficult for even a trained observer to spot the wreckage of a small plane at that resolution - or even from 5,000 feet with your own eyeballs.

      I spent a few years on the local search and rescue team and fortunately only got to see one serious crash up close. From the air, it looked more or less like a bunch of trash strewn across a 100-foot stretch of hillside. Nothing you'd identify immediately as an aircraft, though in this case the huge burn m
  • by nurb432 (527695) on Saturday September 08, 2007 @05:24PM (#20523353) Homepage Journal
    Cool, can i get some of my neighborhood? The stuff on google is a good 5 years old, if not older. The resolution is pretty poor too.
    • Cool, can i get some of my neighborhood?

      Try Microsoft (booo! hisss!) Live Search. Google had only the usual low-res satellite image of my neighborhood, but MSFT had high res airplane photos... and from multiple perspectives!
      • by nurb432 (527695)
        Tried that a while ago, and It looked like the same exact image for my area.

        Reason i know its at least 5 years old is due to the lack of any car in the driveway. No way for me to know how much older tho.
  • Nevada (Score:4, Informative)

    by mysterious_mark (577643) on Saturday September 08, 2007 @05:32PM (#20523395)
    The area of Nevada where he is missing is actually rugged and mountainous ( I have some proerties in those parts myself ). Look on Google earth if you don't believe me, the name 'Nevada' means ',mountains. Also area 51 is now where nearby. There's a lot of rugged and inaccessible terrain he could've gone down, unfortunatley, and 5 days is a long time without water, its dry and hot out this time of year. I'd say the situation doesn't look good at this point, but we can always hope for a miracle, best of luck to the SAR and CAP people.
    • Re:Nevada (Score:4, Informative)

      by Tofof (199751) on Saturday September 08, 2007 @05:46PM (#20523477)
      the name 'Nevada' means ',mountains. No, I think you've got your states confused. The word 'nevada' means 'snow-covered.' The word 'montana' means 'mountain.'
      • Re: (Score:3, Informative)

        by Garabito (720521)
        Actually the name 'Nevada' came from 'Sierra Nevada', which means 'snow-covered mountain range', so the OP wasn't so wrong after all.
  • by Dakkus (567781)
    Hmm. I found something that was interesting that is of correct sizr and somewhat airplane shaped. Probably nothing, but there's still the possibility. However, the frigging site doesn't accept my clicks on either of the radio buttons under the sample image. The browsers I'm using are Safari, Camino and Firefox
    Am I doing it wrong or is the page really picky when it comes to peoples' browser choises?
    "No Windows, no helping"?

    Anyone got it working?
  • Found a plane... (Score:5, Informative)

    by Anonymous Coward on Saturday September 08, 2007 @05:49PM (#20523507)
    38 29' 03.51" N
    119 24' 21.64" W
    • Re:Found a plane... (Score:5, Interesting)

      by Fullerene (1151313) on Saturday September 08, 2007 @06:03PM (#20523593)
      The scale make it a little bit small, but I think it is exactly the sort of thing that they are telling us to report to them. Contact the HIT requestor via Amazon perhaps?

      Looks like a plane to me too.

      • Re: (Score:3, Informative)

        by Fullerene (1151313)
        Follow-up. I have contacted the person who put the satellite picture on Amazon and sent them AC's coordinates. No need for everyone to do it, I'm sure they're busy right now.
        • Ok, so what do we do if we find a plane just browsing in Google Earth? (way more efficient than refreshing the webpage).

          38 7'34.00"N, 11929'4.81"W

          Much more fuzzy than the AC plane, so this is probably nothing, but the size and shape is about right (a bit shorter, but of the plane is angled, it could easily show up shorter).
        • What do you think about

            38 27'2.88"N 119 25'25.17"W

          as possible wreckage of another plane, from some time ago?

          I think this is a great chance to go back and bring closure to those families of people missing in unrecovered crashes in the past, in the area. I hope they go ahead and let the survey complete, even if they find Fossett soon.
    • Mod Parent Up (Score:5, Interesting)

      by Nasarius (593729) on Saturday September 08, 2007 @06:05PM (#20523615)
      I'd say that's a hit. The object matches the dimensions of a Super Decathlon, according to Google Earth.
  • by pongo000 (97357) on Saturday September 08, 2007 @05:55PM (#20523555)
    Most of us aren't SAR experts, and wouldn't know a burn mark from a ridge shadow. The SARs that will be sifting through the public's mostly incorrect identification of accident artifacts would be better utilized in direct search efforts (either in the air or using imagery), rather than being distracted by what could best be considered a somewhat morbid game of "Where's Steve".

    The time to test this type of technology out isn't during a live SAR mission. Leave the search and rescue to the experts, and please don't tie up their time with your well-meaning, but ultimately time-wasting, suppositions.

    • Re: (Score:3, Insightful)

      by Scrameustache (459504)

      Most of us aren't SAR experts, and wouldn't know a burn mark from a ridge shadow. The SARs that will be sifting through the public's mostly incorrect identification of accident artifacts would be better utilized in direct search efforts

      They don't re-send those images to other random users to filter out the results before passing them along to a pro?
      'you sure?

      Because, if they did that, then "x people think there's something here" might make a nice priority queue for those pros, rather than trying blind.

  • where is Bin Laden, or WMDs, or Waldo...
  • by AmericanInKiev (453362) on Saturday September 08, 2007 @06:04PM (#20523611) Homepage
    These pictures are lousy - to really get useable images would require a fly-over.
    Manned flyovers are expensive, slow, and often dangerous if a person is lost due to inclimate weather;
    However Unmanned flyovers can be conducted in poor weather, at very low cost, and without pilot fatigue or airspace crowding concerns.

    It is ironic that private pilots have been objecting to uav, and now their hero doesn't have the benefit of private UAV flights for search and rescue in his time of need.

    Not to gloat, but this would be a fitting time for the private pilots associations to change course on elbowing out UAV's and giving another nascent industry to europe.

    AIK

  • by sdo1 (213835) on Saturday September 08, 2007 @06:06PM (#20523637) Journal
    This is a very good use of the technology. I hope this works if for no other reason than to bring closure to his family if he hasn't survived.

    My problem is the way they've got the web page set up. Every time I submit a new "HIT", I have to scroll all the way down the page again to see the next image. It's great that they have a "primer" a the top, but I've done a couple hundred now... I don't need to keep seeing that over and over again. Just cut to the chase and show me the next picture to examine.

    Also, looking at the Google Earth swath that this is covering, I can't help but think that he might be outside of that. Can anyone comment? Or do they know "if he's anywhere, he's in that area."?

    -S
    • I think part of the problem is the simple fact that Amazon had to shoehorn this project to fit the Mechanical Turk model (no time to develop a custom solution).

      What I really feel is missing: why aren't they putting up an older set of pictures alongside the new pictures? If there's a borderline case, knowing that "this white shiny thing wasn't there nine months ago" could be very important. It would also provide important context.

      Ah well, back to searching.
  • Perhaps.. (Score:2, Funny)

    by Ginger_Chris (1068390)
    I recently lost my mobile phone somewhere in Wadham College gardens, Oxford, and I was wondering if I could get this kind of help to find it. Any takers?
  • by Joao (155665) on Saturday September 08, 2007 @06:56PM (#20523937) Homepage
    I was about to post how distorted the image is, when on a hunch I decided to unclick the "terrain" box on GE. The image becomes a LOT clearer, but I still don't think it is nearly clear enough to identify something as small as a 22" plane.
  • by Dan East (318230) on Saturday September 08, 2007 @07:05PM (#20524003) Homepage Journal
    I keep hearing people whine about the Mechanical Turk interface. Each image only requires one mouse click and two keyboard presses (no mouse movement at all required). Here's how:

    1 - When you manually accept the first hit, make sure you check "Automatically accept the next HIT".
    2 - Press the END key to scroll all the way down to see the image.
    3 - Click the mouse on Yes or No.
    4 - Press the ENTER key to accept the HIT.
    5 - Goto 2

    I've found two images that are really good candidates for a crash. One was at 38.020248,-119.368515. It looks like a line of tree damage, with a bright object at the edge of the tree line.

    Next, I keep hearing people saying that laypeople aren't useful for something like this. This is simply to flag interesting images so experts can spend their time looking only at the most likely candidates. Also, this is free for them. So they could use an algorithm something like this:
    Show each image to at least 5 people.
    Each time someone says "Yes" to a specific image, show it to two additional people, up to a max of 20 reviews.
    Sort the images by descending Yes vote count and show them to the experts in that order.

    Dan East
  • If he crashed in Nevada, why are all the pictures I'm getting located in Yosemite National Park?
  • ...others. Anyway that's what I surmise after hearing an official state that searchers have found several other crash sites and she said that's good news of a sort, because it means that several other families will at least now have closure.

    What it meant to ME is that some people apparently warrant searches that other people don't warrant. Government officials apparently chose to search for ONE jackass who didn't file a flight plan, but I guess the others weren't worth it. We don't know whether THEY filed f

    • Steve is ridiculously wealthy, and I don't mean this in a bad way. All of the expenses of searching for him can be recovered from him (or his estate), so there is basically a blank check on all of these costs. And yes, he's famous, so people will care and do more about finding him compared to some other schmuck - this is human nature and isn't going to be changing any time soon.
  • 38&#176;08'01.34"N 119&#176;25'20.19"W

    There is a strange rectangular object to the left of this coordinate. It doesn't seem like it would be a rock. I imagine if a plane crashed the wings would be gone and such. I dunno, just wondering because I thought it seemed odd.
    • by Em Ellel (523581)
      Way to large - looks like its about 40' across and few hundred feet long. Submit it to be reviewed though - never know.
  • Too bad it's Amazon (Score:2, Informative)

    by DogDude (805747)
    Too bad it's done by Amazon. It's an absolutely fantastic idea, and a really new technology application (getting the public to scan pictures). Unfortunately, I don't have, and won't have an Amazon ID due to their continued promotion of dog fighting [amazon.com]

    Fuck you, Amazon.
  • Jim Gray (Score:3, Interesting)

    by tji (74570) on Saturday September 08, 2007 @07:40PM (#20524247)
    The last time an effort like this was undertaken, it was for Jim Gray (Database researcher, Microsoft Fellow), who had disappeared sailing from San Francisco. I checked on that for a while, but never saw any more information.

    Was anything ever found in the search for Jim Gray? No remnants of his boat, or other signs of what happened?

    • Re: (Score:3, Informative)

      by aug24 (38229)
      From Wikipedia, because I can be arsed to hit Google for 5 seconds...

      "On February 1, 2007, the DigitalGlobe satellite did a scan of the area, generating thousands of images.[9] The images were posted to Amazon Mechanical Turk in order to distribute the work of searching through them, in hopes of spotting his boat.

      "On February 16, 2007, the Friends of Jim Gray Group suspended their search,[10] but continue to follow any important leads. The family ended its search May 31, 2007. The massive high-tech effort d
  • Quite ironic.... (Score:4, Interesting)

    by thrill12 (711899) on Saturday September 08, 2007 @08:12PM (#20524449) Journal
    ...as Steve Fossett originally set out to take this journey to find a flat and long enough place to do his world land-speed record. Now Google has high-resolution imagery of the whole place, which makes the whole undertaking a bit obsolete in retrospect ?
  • by Backyard Billy (1153843) on Saturday September 08, 2007 @11:15PM (#20525555)
    38 3'24.02"N 11914'56.55"W It looks like a plane, it's about 20 ft in length and has a 23-24 ft wingspan. I don't know how to contact them but if anyone here can communicate with them, please do. Don't forget to credit me if it turns out to be his plane.
  • Question (Score:3, Interesting)

    by truesaer (135079) on Sunday September 09, 2007 @01:03AM (#20526125) Homepage
    I've seen on the news that they've checked out like 6 or 7 plane crash sites that turned out not to be Fossett. What that left me wondering was, are those sites where they just left wreckage because it was remote or are they previously unknown crashes where a plane went missing and was never found?


    How often does that happen with light aircraft? Do they vanish entirely very often?

  • by Tribbin (565963) on Sunday September 09, 2007 @01:44AM (#20526315) Homepage
    This might have been said already but I'm too drunk and tired to check the posts.

    Take several detailed pictures and look for the diffs; there he is; if he's alive.
  • by ckedge (192996) on Sunday September 09, 2007 @03:12PM (#20530479) Journal
    When I started to help this morning there were 32,000 work units (called hits, images to be reviewed) available. They were disappearing at a rate of 5-10,000 per hour, meaning that all things being equal there were 50-100 people looking at them.

    However over the past half hour the work units available have been *increasing*. Currently 12,000 and increasing. Clearly they are adding more to be done faster than we're doing them. So anyone who helped out at the beginning - don't assume the hits are "all done". There could be more at any time.

    In my old version of IE I couldn't see the scale bars or the example image, looking at the same coords of a unique scene in google maps I estimated the image was 125m x 125m - which would be half meter resolution. Now I see they claim the images are actually 85x85m, which would be 1.08ft resolution.

    Based on that and that I've done 400 units, that mean's I've searched one full square mile.

    It also means the 32,000 units I saw when I started is only 10 miles x 10 miles, 100 square miles. I heard someone else say that they only have 500 square miles of imagery. Looking at Google Earth, assuming the new imagery is the kinda-rectangular patch that is all the same color/brightness - they have approx 1700 square miles. That means there is approximately 600,000 work units in total that need done. If everone does a square mile (shouldn't take more than an hour) then we need 1700 people helping.

    But as someone else noted - they're really artificially limiting the search area, considering the range on his plane. Assuming he went certain places or crashed on his way back to the ranch. That doesn't bode well.

    PS: It'd be way way more effective if they showed a "image before crash" so that people could self-discover their false positives, without forcing people to download google earth and figure it's before/after out, and/or be smart enough to copy/paste the coords into google-maps satellite view.

    PPS: If they were really smart, they'd have a second private pool of the public's false positives being reviewed by amateurs or employees whom they know have much much smaller false positive rates, whom they know are comparing the two available before images (google maps and google earth) against the current images.

    BTW: Here are images of the actual specific plane he was flying. http://www.airport-data.com/aircraft/N240R.html [airport-data.com] (Aviation buffs take pictures and index online everything that flies, apparently :) I'm guessing that although from the side it's mostly blue, that the top of the wings are white.

There is no royal road to geometry. -- Euclid

Working...