Please help find Steve Fossett! Steve's disappearance has been all over the news, but so far no trace of him has been found. You can help search, by using Amazon Mechanical Turk. Several organizations have pulled together to use recent satellite imagery for this effort. The image on the right-hand side of this page is one of the actual images that I just looked at as part of my contribution to the search. Have to say that some of this terrain is incredibly rugged.
Here are some highlights of the imagery, copied from one of the "HITs" in Amazon Mechanical Turk:
- The images shown below are recent and of the relevant search area.
- Each tile is roughly 256x256 pixels, representing an area of 85m or 278ft square.
- Steve's airplane will appear in the images as a 30pixel wingspan and 21 pixels in length.
- We're asking between 5 and 10 people to review the same images so we can be certain nothing was missed.
A cool feature of this particular set of hits is that you can view the satellite imagery in Google Earth. But a note about that... I am using a new computer, and Google Earth was not installed. I had to restart Internet Explorer before the links functioned properly for Google Earth integration. Think it's a mime type association issue.
This search effort may sound familiar--a few months ago the same sort of effort was launched to look for missing Microsoft Researcher Jim Gray. Unfortunately no trace of Jim was ever found.
The basic idea of Amazon Mechanical Turk is to deliver a highly scalable workforce for tasks. In this case, a large number of satellite images were injected into the system for people to review. Because each image is made available to multiple people for review, there is a de-facto quality check built into the workflow. While the example at hand is unfortunate in its circumstances, its existence does showcase the innovation of Amazon Mechanical Turk. Take a look at the other work in the system to see how others leverage this workforce.
I'm a pilot, and within the General Aviation community there is a long-standing debate around the subject of ELTs, which is short for Emergency Locater Transmitters. These beacons were originally required by law after a couple on members of Congress went missing in Alaska. (They never were found.)
The original beacons were simple analog radio signals that transmit on 121.5 MHz, the universal emergency frequency for aircraft. Satellites in orbit also monitor for signals on this frequency. A number of pilots maintain that any beacon at all is a waste of money; if you're old enough, this is reminiscent of the airbag and seatbelt rhetoric of years ago. The problem with the old technology is false alarms: over 100,000 PER YEAR.
Modern digital technology is now used, often coupled with a GPS receiver to make it much easier to pinpoint a downed aircraft. Problem is that these new beacons are extremely expensive. Anything certified for aircraft is insanely priced--and currently 406 MHz beacons cost over $1,000 per plane. That's a lot in aircraft that are flown occasionally for recreational use.
If you want more info on the two technologies, I've posted an article about the two ELT versions on an aviation website that I run called PopularAviation.com.
Finally, as a pilot who has participated in a number of Search and Rescue missions, the probability of the wreckage looking like an airplane is low. This image shows what the plane would look like if it is resting horizontally after a normal landing. In fact wreckage usually looks more like random junk. Anything that looks out of place should be considered "suspect" in these images.