Searching the Vasty Deep

The Internet is now being used as a tool for seeking lost individuals and ships lost at sea. Satellite and aerial mapping data, in the case below, is being shared out to numerous volunteers. Page codes are returned to a central web site with recommendations – "nothing found", or "please review".

James Gray, 63, a Microsoft researcher and winner of the prestigious Turing Award, failed to return home from a sailing trip on Sunday, January 28.

He left from San Francisco Bay aboard a 40-foot sailing boat, and had intended to scatter his mother's ashes at the nearby Farallon Islands.

Several days of intensive searching by the US Coast Guard and private planes revealed no sign of Dr Gray or his boat, so desperate friends and colleagues turned to the internet for help.

On Friday, engineers from NASA, online retailer Amazon and technology companies such as Google and Microsoft organised a satellite and high-altitude aircraft to photograph the area where he was believed to be located.

The photographs were then split into smaller tiles and uploaded to Amazon's Mechanical Turk website, allowing virtually anyone to take part in the search effort.

If you want to help, click here. This methodology is essentially distributed computing, similar to that used in the SETI project, except that the parallel component consists of human visual analysis units. I wonder whether this open community approach might have helped with the Beagle 2, an ESA lander lost on Mars, Christmas Day, 2003.

Remote sensing and satellite imagery are finding more uses every year. It’s really remarkable. Search theory and methodology have also matured dramatically within my lifetime. I remember Grace Hopper mentioning the search problem in a speech at my school. As a naval officer, she couldn’t help but be exposed to the problem. A recent episode of Numb3rs went into some detail on how to find a sunken ship.

We may think the world seems small today – with international communications and online mapping, but things can still be lost on a single plot of land, and the Pacific is a whole other thing. I’m sure Hopper would be gratified, but not surprised, at how computers have facilitated this work.

It is ironic that once imaging becomes sufficiently detailed and available to do this kind of work online, countervailing forces act to politically impose reduced availability and resolution. If India can change an international Google standard, I can imagine what China has requested.

2/6/2007 1:56 AM

Tags: search theory, James Gray, distributed computing