Tapping into the Maritime-Tracking Space Race

Author: 

http://www.institutionalinvestor.com - Fred R. Bleakley
Companies like exactEarth, Orbcomm and Spire Global have deployed satellite networks to collect data on shipping traffic for government and commercial users.
Show

The quest to turn big data on cargo shipping into winning trading strategies would not be possible if not for recent advances in satellites, microprocessing and analytics. One of the pioneers in the field is Peter Mabson, president of Cambridge, Ontario-based exactEarth, founded in 2009 by Canadian space hardware manufacturer COM DEV International.

In 2005, Mabson, then a COM DEV technology and business development executive who had built the payloads for hundreds of satellites over the previous two decades, "stumbled across this new thing, AIS," which he describes as "sort of a GPS for ships." Officially known as the Automatic Identification System, AIS is made up of radio transponders on large seafaring vessels that several times a minute automatically transmit data about a ship's identity, position, course and speed.

Mabson recalls, "It was fascinating that, coincidentally, new microprocessors could at the same time make satellites cheaper and smaller to put into space." The satellites allow commercial and government observers to pick up the signals of vessels far out at sea rather than relying only on land-based monitors that typically lose track of ships more than 50 miles from shore. The problem for exactEarth was how to identify individual ships from AIS signals of tens of thousands of vessels.

For Mabson the solution was a combination of launching small, microwave oven-size satellites 500 miles above Earth to record these AIS signals and then having the satellites "fire all those jumbled-up signals back to a large processing center in Toronto." He adds that the "secret sauce that makes it all work is a massive engine full of algorithms that detangle the signals." Over the past six years, exactEarth has collected 6.5 billion data sets that track the movements of the world's commercial shipping fleet. It sells the full set to government entities and on a case-by-case, sometimes limited basis to commercial buyers, hedge funds and commodities traders. The company currently has eight satellites in orbit and plans to add dozens more over the next two years. Mabson has competitors in the satellite AIS space race. Rochelle Park, New Jersey-based Orbcomm, which started about the same time as exactEarth, launched 11 new satellites in December aboard a SpaceX Falcon rocket. Spire Global, a San Francisco-based company started in 2012, in September launched four Lemur-2 satellites that can do both maritime monitoring and atmospheric measurements.

Picture-taking satellites are also proliferating. One such provider is San Francisco's Planet Labs, founded by three former officials of the National Aeronautics and Space Administration. The company already has 40 small satellites in orbit, which it is using to build historical patterns on everything from agriculture yield predictions to the lengths of shadows of liquid energy containers.

The satellite companies' clients include Louisville, Kentucky-based Genscape and London's IHS Maritime & Trade, analytics firms that combine other data sources with AIS signals. At Genscape's Vesseltracker headquarters in Hamburg, computers show maps of large and small regions of the globe with hundreds or thousands of dots, each representing a vessel. As an analyst hovers a cursor over a dot, a photo of the vessel pops up with dozens of lines of identification, including location, size, name, owner, speed and destination, plus the estimated time of arrival. Genscape, whose clients include Scott Borgerson's CargoMetrics Technologies, also uses historical commodities prices, port logs of cargo and 2,000 land-based AIS sensors.

"If we see a ship that ten of the last 12 times went to the same port carrying the same cargo, we can make assumptions validated by history," says Genscape CEO Matthew Burkley.

Level: 

Category: 

field_vote: