in crime

Distributed search of a different sort

I drove to the Oregon Coast today and the whole time I kept thinking about James Kim and his family. They have been missing for a week now and since there have been no phonecalls or credit card use, it’s not merely getting stuck in some snow (which we had last sunday or monday). It’s likely something worse, which caused me to scan the forests and embankments all the way to the coast.

I read about how they concentrated the search on the 38 highway, but if you check it out on Google Maps, there are 5 or 6 major roads (all about 60 miles long) that link the main cities along the 5 freeway with the coast. I know on my first trip to Oregon, I just randomly picked one and drove along the coast instead of the freeway. They could have certainly done the same.

I was thinking about how helpless it feels to sit at home and worry about this family, and how you could harness the power and goodwill of everyone. I’ve seen some pretty amazing stuff come out of MetaFilter, when people collaborate on a real world problem. Then it hit me. There are only 5 or 6 major roads to the coast, and they’re not that long. Why not run a surveillance plane 500′-1000′ feet above each of the roads, going slow enough that it takes maybe 20-30 minutes to follow the roads to the ocean. If the camera view could capture 100-200′ north and south of the road, you could probably film all 5 or 6 major roads in a single clear day like today.

If each recording is say 30 minutes long for a road, split it into 10 equal parts, 3 minutes long, and upload all of them to youtube. Ask viewers to leave comments pointing out when they see anything strange. The Kims were in a silver Saab wagon, so it’s probably something that can be seen from above. In total, there’d be 50 or 60 short clips and in a matter of hours you could have millions of people closely scan then and start pointing out the things worth looking into on the ground. If everyone says there’s a silver glint in the trees on video #6 from the highway 18 group, at 1:55 in, you could send a police unit out to investigate.

Hopefully an approach like the one I described is fairly normal in the future.


  1. Ideas like yours make the future of technology and social online volunteerism seem exciting and tangible – I pray from the Kims.

  2. This sounds like it’d be a great use case for Mechanical Turk, assuming you can keep people from trying to game it like they try to game everything else on there.

  3. Might be worthwhile contacting someone about this. IIRC, they already have helicopters searching the area. Should be possible to stick a camera onto one of them and upload footage onto the net.

  4. Here in Australia tend to lose people out bushwalking all the time, or out to sea.
    If only there was an application like Google Earth that had real time high resolution satellite data and some kind of collaboration facility made available for this purpose.
    If everyone was set to search at a specific resolution, and areas divided up into units that had markers indicating how often they had been searched, it could actually work.

  5. It works a lot better if you remove the time aspect — you can take HD video, combine overlapping frames to increase resolution further, and then mosaic into one long static image, maybe with links to motion video of that section… that way you can randomly assign people sections to look at, rather than having everyone start at the beginning of the video (the least likely place to find someone.) Like flickr, they can note locations in that 2D static image easily, as well.

  6. One pass with one high resolution military satellite.
    Get the whole area at one time.
    I realize there’s a war on, but his tax dollars helped pay for it too.

  7. I would suggest doing some image processing before letting people look at it.
    For e.g., some basic color filtering could be done to locate silver hues in image sequences, infact if infrared photography was possible then searches on both sides of the road to look for heat objects might be useful…basically run the video through a whole range of filters that would take very little time and if nothing works let millions look at it.

  8. This is good, but it also would rely on the person(s) likely being near the vehicle in question and/or not underneath any other form of covering.
    I hope they’re found soon.

  9. Ram is correct. Computers are limited in analyzing context but great at colors and patterns. Cross-reference visible, infrared and that crazy thing the army uses to look through walls and they will appear.

  10. It’s moments like this that make me excited to have a career that touches the web, and that lead me to people’s blogs like Matt’s. These are the ideas that need to be heard, and because of technology, can be accomplished. What sounds like a great idea today has a pretty good chance of being a reality tomorrow. Until then, hopefully the analog world will find them soon.

  11. Well, they’re already distributing bits of radio telescope signals to distributed clients to search for aliens, so it shouldn’t be much harder to send different pictures to a distributed group.
    It would at least be a much better use of people’s time than Digg…

  12. If something like this because the norm, lost people would know to make large landmarks, arrows or “I M HERE” made from branches or stones or the like, documenting their path in the knowledge that someone will see them.

  13. Satellite is nowhere as useful as low-level reconnaissance video — first of all, you have to wait for a clear day (and a lot of these situations happen in inclement weather, and are most urgent during those periods) and the resolution just isn’t there. With video data from a low-flying plane, not only can you see in cloudy weather, but if done correctly you can look at a particular position from multiple angles based on the shots taken as a plane flies over.
    An unmanned drone sounds neat, but getting one that’s stable enough/large enough to hold a high-resolution camera is probably a lot harder than simply having a local news helicopter with stabilized HD cameras fly over — most large media markets have one these days — or a fixed-wing aircraft with a camera attached in a housing.

  14. Quote James: “If only there was an application like Google Earth that had real time high resolution satellite data and some kind of collaboration facility made available for this purpose.”
    I’m sure it can already be done by the powers that be… some folks probably get a good look at Iraq like that every day… but even scarier than government use is when its publicly available. Getting lost isn’t always a bad thing.
    As for the Kims – if they are hurt or in trouble, I wish them well – and if this is something more than an accident, all the techno gizmo’s in the world can’t really stop bad people from doing bad things.

  15. This is a silly idea. The chances of finding them out in the open, easily seen from above are minuscule.

  16. The distributed search idea is interesting, but you might be able to solve 80% of the problem simply by setting up a community that matches up missing person reports with private pilots and survivalists.
    I don’t know about the web 2.0 world, but in the software industry lots of people have their pilot’s license, and a few even own their own plane. I bet a good number of them would love to use their love of flying to help rescue people. They should bring along a spotter or two, binoculars, and a GPS system. Video would be niec, but it might take some time to code up the software that is being proposed here. Until then, spotters in light planes might save a few lives.

  17. b1tr0t, you might as well go the next step and have some sort of organized search-and-rescue organization for which private pilots can volunteer their time … oh wait! we already do! Most regions seem to have volunteer air search and rescue organizations.

  18. The state of Texas just concluded a project that applies this concept to a different problem: border patrol. They put up a bunch of webcams along the border, and people could sign up to “patrol” the border from the comfort of their own homes.
    Apparently they had 200,000 people sign up, and 25 million viewings. I don’t know how many crossings were reported.

    “A commercial satellite-imagery company said Tuesday it is rerouting one of its satellites to fly over the Oregon wilderness where rescue crews search for CNET editor James Kim.

    Another concern, said Bender, is where to send the pictures. The company only flies the satellite and doesn’t employ analysts to comb the photos. Bender said his company was notified by a concerned citizen interested in helping to find Kim.”
    May be too late now, but hopefully something useful will come out of it.

  20. I was so sorry they didn’t find James Kim alive, but so relieved the rest of the family came through okay.
    Matt, this idea – when you Twittered it at the time – reminded me of the air-dropped webcams idea in Patrick Farley’s extended web comic The Spiders
    from 2001-2003

Comments are closed.