Solving the “Find, Fix, Finish” Equation
Interview with Lt. Gen. David Deptula, USAF (Ret.), CEO, Mav6
In July 2006, Lt. Gen. Dave Deptula was appointed as the U.S. Air Force’s first deputy chief of staff for intelligence, surveillance, and reconnaissance (ISR), responsible for policy formulation, planning, evaluation, oversight, and leadership of ISR capabilities. He also served as the Air Force’s senior intelligence officer.
A veteran F-15 Eagle pilot, Deptula was the principal coalition air campaign attack planner during Operation Desert Storm and later served as joint task force commander for Operation Northern Watch, during which he flew 82 combat missions enforcing the “no-fly” zone over northern Iraq in 1998-1999. He returned to Southwest Asia in 2001 as director of the Combined Air Operations Center for Operation Enduring Freedom, orchestrating air operations over Afghanistan at the start of the war ignited by 9/11.
Following his retirement in 2010, Deptula was named CEO of Mav6, a newly created defense contractor focused on bringing lighter-than-air (LTA) airships, modernized with cutting-edge technology, back into military service, primarily as ISR platforms.
Deptula recently spoke with Defense senior writer J.R. Wilson about the status and future of ISR.
J.R. Wilson: What is the current U.S. state of the art for ISR?
Lt. Gen. David Deptula: I think it’s very healthy, from a macro-level perspective, in terms of collection capability. There are a variety of leading-edge technologies that have been applied to sensors in all the domains – airborne, space-based, at sea, on the ground, and throughout the different disciplines of intelligence – signals, human, imagery.
There are two areas that require work to increase the collective effectiveness of those technology capabilities:
First, integration of the information that comes from these sensors – the ability to cross-cue sensors and information automatically.
Second, the organizational integration of not just the collection sensors but the entire ISR enterprise, as opposed to the segregated collection, analysis, and dissemination we grew up with in the last century.
Without adding any additional technology, platforms, or people, you probably could achieve a 30 to 40 percent increase in effectiveness in Afghanistan today, with all the ISR assets deployed there, if you just had a well-integrated ISR management concept of operations.
Because we have capitalized on technology to develop ISR capabilities – and done so well at that – we are now increasing data collection at extraordinary rates, well beyond our ability to analyze it all in the fashion we used to – individual sets of eyes looking to everything collected to pull out points of interest.
From the beginning of time until the year 2003, there were about five exobytes of information created. We now create five exobytes of data every two days – and that is accelerating.
How do you deal with that?
Completely different concepts than the way we have designed equipment today. We now collect imagery by UAV [unmanned aerial vehicle], for example, stream it to the ground and then analyze it after it gets offboard the system. However, an airborne surveillance system collects so much data there is not a big enough pipe to offboard it all. So you have to either offboard select elements or wait until you land, offload the information, and then analyze it.
The problem is you are now well behind the time line in terms of being able to act on that information in real- or near-real time. So the answer is to move to onboard processing of these vast amounts of information.
And what would that require?
It will take some very serious application of effort to change the organizations, processes, and procedures now in place. And one of the biggest difficulties we have is institutional change, because the organizations responsible for different modalities of information all are very well established bureaucratic institutions that have evolved over time. They are magnificent at doing what they do, but integration of different modalities is not one of those.