Big data: Border protection's first line of defense
- By Frank Konkel
- Oct 30, 2013
The Department of Homeland Security's biggest weapon in securing American borders is not a fence or a fingerprint, but data -- and lots of it.
DHS and its component agencies clear some 1 million people across U.S. borders and process some 60,000 maritime and air-freight containers daily, just a small portion of the $2 trillion worth of imported goods that enter the country every year.
The incoming flow of goods and people also comprises massive datasets DHS continuously runs against other federal databases – no-fly and watch lists, for example – allowing the agency to differentiate among risks specific to a particular passenger or cargo.
DHS has been collecting and logging these datasets for upwards of 10 years, building petabyte-sized "haystacks of data" in which hide elusive needles of useful information. The department's evolutionary successes in handling the information has turned big data into a big part of its strategy in protecting borders, according to Alan Bersin, assistant secretary of international affairs and chief diplomatic officer for DHS.
Big data, he said, has changed the way DHS looks at borders.
"We typically think of borders as the first line of defense, with hard lines, but in the global context, our job is to keep dangerous people and dangerous things away from the homeland," said Bersin, speaking Oct. 30 at the Bipartisan Policy Center. "We don't do that by looking at borders as the first line of defense, but as the last line."
By Bersin's count, there are three ways to approach the problem. The government's post-Sept. 11 attempt at one of them – looking at every piece of data – was a failure, Bersin said, noting that border personnel can't possibly open every trunk of every vehicle that comes by.
The second, using specific intelligence such as a suspected terrorist's phone number to find the person works only if that specific intelligence is available.
Where DHS has been successful in recent years is with the third approach, making the proverbial haystack smaller, Bersin said. The ability to differentiate among risks "is where big data comes into play," he said.
The case of Umar Farouk Abdulmutallab, the Nigerian man jailed for attempting to bring down Northwest flight 253 over Detroit in 2009 with a bomb hidden in his underwear, is "the seminal case that changed the way we looked at using data," Bersin said.
Abdulmutallab purchased a one-way ticket to the U.S. in Africa and carried no luggage. Today those unusual data points would be scanned against previous flight records and a series of other databases, leading authorities to a rapid secondary examination of such a passenger.
Were Abdulmutallab's scenario to unfold again, DHS's big data efforts would likely have prevented him from getting aboard his target flight. As it was, he was foiled by a quick thinking flight attendant.
Frank Konkel is a staff writer covering big data, mobile, open government and a range of science/technology issues. Connect with him on Twitter at @Frank_Konkel.