Big Data Analytics for Locating MH 370

By CIOReview | Wednesday, March 19, 2014

FREMONT, CA: An unprecedented event of a jumbo carrier flying with 239 people suddenly disappearing from the view of the men and the machine is quite astonishing. While several contrasting theories have emerged trying to elucidate what might have happened, there is no concrete evidence to backup speculations. To facilitate the search, DigitalGlobe has opened up its repository of the satellite captured images to the people to analyze and tag the pictures they find of interest in locating the aircraft.

DigitalGlobe is a commercial vendor of space imagery and geospatial content and has five private satellites providing images taken from around 400 miles of distance above earth covering over 4.5Bn square kilometers of global coverage as reported by Chris Preimesberger of eWeek. All those images will be available to view in the, a crowdsourcing website.  Tagging the pictures will allow the Big Data engine to filter those regions that has greater amount of tags and provide clue to the authorities to fine grain their search for the aircraft. Clues could be oil slick in the waters, wreckage of the aircraft or raft. People can observe those images and tag suspicious regions for any debris.

The Tomnod platform has earlier been used for finding a missing plane in Idaho, U.S. as well as mapping the damages caused due to the natural calamities that struck Philippines and U.S. in the form of typhoon Haiyan and Oklahoma Tornado respectively.

"Anyone can click on the link and begin searching the images, tagging anything that looks suspicious. Each pixel on a computer screen represents 0.5m on the ocean's surface. We'll say, 'here are our 10 top suspicious or interesting locations'. Is it really an aircraft wing that's been chopped in half or is this some other debris floating on the ocean?" says Luke Barrington, Senior Manager of Geospatial Big Data for DigitalGlobe while speaking to ABC News as reported by Vasudevan Sridharan of International Business Times.