Counting Cars to Find the Fight: High Over Hostile Territory, Squadron Uses New Tools to Make Sense of Aerial Images
Analytics tames avalanche of drone video, other data
Streaking over Libya in a Sentinel surveillance jet, squadron leader Chris “Moose” Melville leaned over a computer screen and studied the image of a town far below.
Commanders wanted to know who was winning a battle between forces loyal to Libyan leader Moammar Ghadafi and anti-Ghadafi fighters in the town east of Brega. The front line was hard to see amid sporadic, house-to-house fighting – but the Sentinel’s crew had a solution.
Townspeople would avoid dangerous areas, they reasoned. So using the plane’s Raytheon radar and processing equipment, his crew quickly mapped the flow of traffic in the town. Soon they had a map of yellow arrows showing where cars were backing up or turning around. In the middle was an empty area outlined in purple – the front line.
“We’re giving information, not just data,” Melville, of the Royal Air Force, told reporters at the Farnborough International Airshow this week. “We analyze it and turn it into this: information that a ground commander can use to make an informed decision.”
Picking out useful information is becoming increasingly important as drones and surveillance planes beam down a growing torrent of video and data, Raytheon experts say. Last year drones collected so much full-motion video that it would take a person 24 years to watch it all, Raytheon research shows – three times as much as in 2007.
“Data is what we see as the raw stuff coming out of the sensor,” said Jim Hvizd, Raytheon’s vice president of International Strategy and Business Development for its Space and Airborne Systems business. “How do we turn data into information, and then information into action? That’s really the key element of our technology solutions.”
This week Raytheon announced new surveillance developments, including a $191 million deal to provide its Multispectral Targeting System and a separate contract to develop a surveillance pod, both for the Air Force’s MQ-9 Reaper drone. It also announced the fall release of a Google Android-based system that delivers surveillance video, chat and “buddy tracking” to soldiers.
To make surveillance data more usable, Raytheon is developing sensors that can do some basic analysis by themselves, said Brooke Griffith, director of business development for Raytheon’s Intelligence and Information Systems business. That’s especially important as enemies become more aggressive about jamming transmissions, he said.
“Getting that stuff closer to the sensor makes it a lot simpler,” Griffth said.
Raytheon is also developing automated analytics tools to tame the avalanche of video and other data created by the information revolution, Griffith said.
In 2011 alone, the Internet generated 1.8 zettabytes of data - so much that it would take a human 42 million years, reading 24 hours a day, to consume it all. A zettabyte is 1 trillion gigabytes.
To hone its skill at searching data, this year Raytheon launched a series of internal “war games” and innovation efforts called the Grand Challenge 2012. Ideas from the exercises will make their way into the company’s products, Griffith said.
New Raytheon tools should be able to reduce analyst costs by more than 50 percent by doubling their output, he said.
“There’s a landslide of data out there, and in this time of tight budgets, automated analysis is the only solution,” Griffith said.