The Signal in the Noise

Oct. 27, 2015

The Air Force wants to make the most of its hard-built intelligence, surveillance, reconnaissance, and command and control capabilities. To do so, it has to analyze, exploit, and disseminate the information more effectively and faster, senior leaders noted at the Air Force Association’s Air & Space Conference in September.

USAF’s ability to project both power and global reach around the globe has grown more potent due to advances made to its networks and ISR nodes, in particular the Distributed Common Ground System (DCGS) and the service’s combined air and space operations centers (CAOCs). Now composed of 27 geographically separated and networked sites, the DCGS collects and processes information from a vast array of sensors on aircraft such as the U-2, the RQ-4 Global Hawk, and the MQ-9 Reaper.

But as USAF adapts its ISR architecture from one that supported low-intensity fights to one better equipped for contested and anti-access challenges, senior leaders are emphasizing that the service must get a handle on its information management and exploitation practices. Modernizing the DCGS and the CAOC architecture are critical to the future success of the combat air force, Air Combat Command’s Gen. Herbert J. “Hawk” Carlisle noted.

Airmen “are drowning in data,” he said during his speech on fifth generation warfare. “We have to help them figure out how to sift through that raw data and get better at it. We need to change data into knowledge to give us the decision advantage against our adversaries.”

As sensors and networks get more capable, the problems are mounting. The Air Force needs to reinvest in trade craft, and build more “all-source” analysts who can solve increasingly difficult, time-sensitive problems. They will turn raw data from satellites, sensors, and even open sources into targetable information—rather than just stare at screens for hours at a time tracking 24/7 full-motion video feeds, several leaders noted.

Lt. Gen. Robert P. “Bob” Otto, deputy chief of staff for ISR on the Air Staff, told attendees that USAF has to figure out how it will leverage the “big data” revolution—as its implications affect not only capability but also retention. “One of the struggles we have today is we are taking very bright airmen, and … having them do some pretty mundane stuff,” Otto said. These geospatial analysts who are doing shift work trying to keep up with FMV requirements will get bored, and eventually leave, he added.

Instead, USAF could train intelligence airmen to go “solve problems,” use databases and insights to a key question in real time. “We are much more likely to harness their interests and therefore retain them over time,” he added. Analytics in the Air Force of the future will be very different than they are today, he emphasized. “The ability to rapidly retrieve and assess and act, kind of indiscriminate of the sensor or the domain, is the goal,” he said.

Refining information quickly is a critical challenge in today’s combat operations—which is why investing in intelligence skills such as targeting is so vital. Air strikes on ISIS forces are overwhelmingly successful because the process of turning intelligence into “targetable data” over Iraq and Syria has improved vastly since the opening days of Operation Inherent Resolve, said Lt. Gen. John W. Hesterman III, USAF’s assistant vice chief of staff and former Air Forces Central Command boss.

“In my humble opinion, … our ability to do [rapid targeting] had atrophied a little bit before we started this campaign,” Hesterman added.

Closing The Tech Gap

Otto said one reason targeting has atrophied is because the services and combatant commands cut the number of targeteers at the same time precision guided munitions were proliferating and becoming more capable. Now a single aircraft can hit 10 different targets on a mission. Targeting is a “higher level analytic skill” that the Air Force needs to invest in, he added, especially as it confronts challenges in contested environments against highly capable adversaries with advanced weapons.

One of the issues exacerbating the stress on analysts is the rapid growth in network capacity. Otto pointed out when he commanded the 9th Reconnaissance Wing at Beale AFB, Calif. (from 2008 to 2010), data network transfer rates were around five megabits per second (Mbps), while today those rates can hover around 3,000 Mbps. From full-motion video to light detection and ranging information (LIDAR) to classified and commercial space imagery, there is an increasing flood of data “through the pipes,” Otto said.

In response, USAF must expand its partnership with commercial industry—where the state of the art is being constantly refined in fields such as data analytics, tagging, speech-to-text technology, machine learning, and other information exploitation advances. “The commercial sector is going to have tools, some of which are phenomenal,” Otto added. But the question is whether the Air Force can vet these technologies and improve them on a timetable.

Automation and machine learning are two areas that could have great impact on how analysts sift through seas of data, and a field where commercial firms are making great progress.

Speaking on the panel with Otto were James Crawford, CEO of the geo-mapping firm Orbital Insight, and Samuel Druker, director of data science at Microsoft. Crawford noted that commercial satellite imagery is more accessible than ever and his company can now track data related to industries ranging from oil and gas to construction and agriculture yields. Druker called his company’s process of finding insight for customers “pixels to numbers,” that helps humans look at smaller more relevant data sets. “The numbers are just a signal, and … that signal is hidden within a huge amount of noise,” he said. While technology and data science will help assemble this information, the human will remain incredibly important. “In some sense, what you are looking for is a needle in a haystack. And what we are trying to do with the automation is to sift out the haystack into a very, very small pile of hay, so that the humans don’t have very much to look through,” said Druker.

Beyond the technological gap, what will make the difference long term is investment in intelligence tradecraft, said Steven Rogers, USAF’s senior scientist for automatic target recognition and sensor fusion. “How we interface humans and computers together to do what we do … is unique and that’s where our advantage will lie,” said Rogers. “It won’t be in the technology, … it will be in how we do that tradecraft.”