2011 GEOINT Symposium Speech

By General C. Robert Kehler | Omaha, Neb. | Oct. 19, 2011

Good morning everybody it's a pleasure to be here

I have to say that when I was standing off stage here watching the CIOs from the intelligence community walk off it reminded me that of course the CIO of Stratcom is also our J6, and I typically find our J6 in one of two mental conditions. I either find her in the corner facing the corner deeply depressed, or she's standing in front of my desk blowing into a paper bag.


They looked surprisingly good I thought. It was a lot of fun to listen to them in the back too because I'm going to talk about some things that they teed up I believe.

But before I do have to tell you it reminds me of another quick story. A man is standing in his living room when he hears a car approaching rapidly down the street. It turns out this car is careening back and forth and the car screeches to a halt in his driveway and his wife jumps out. She pounds up the stairs, flings the front door open and yells "I've won the lottery pack up your bags!" and he jumps up and says "Honey that's great should I pack for the mountains or the beach?" and she replies "I don't care just pack your bags and get out."


I felt a little bit like that after I talked at the GEOINT symposium in 2008, I felt a little bit like the guy. "We're excited, but pack your bags."

That's because the last time I was hear I gave a talk entitled "one size does not fit all.", it was about collections satellites, specifically about GEOINT collection satellites, and the assertion that I made was that one size does not fit all and that I thought that we were making a mistake and I thought we were making a mistake in trying to have one size fit all.

That actually got some playback from some of you in the audience, in fact I still occasionally get emails and comments about that. It was about 50/50, about 50% of the people said "Yup that was a great idea", about 50% of the people thought I was crazy I think.

So what I'd like to do today is set that one aside, of me being the guy that I am. But I want to raise another one with you today and hopefully I can raise this in a way that is going to continue to generate the kind of debate that the last talk that I gave here generated.

Because I want to have this dialog about how we are managing the intelligence enterprise in support of national security today. When I think as the title of the symposium suggests we can all agree that forging integrated intelligence is both important and appropriate.

But I would argue that how we process, exploit, and disseminate the massive amounts of data that we generate, and the resulting information is where I think at least the real challenge lies.

We need to think hard about how we surf through the growing volume of raw data that we collect to find the critical information. Not just the information we seek, but the information that indicates an unanticipated event, or some kind of a strategic shift that could take us by surprise.

In essence, how do we turn collection into focused knowledge and eventually into action? So in my remarks I'd like to talk about today's operational environment, because it's an environment in which our adversaries leverage uncertainty, and they do that to mitigate our bypass, our advantages, and to impose surprise.

By the way in my humble opinion, surprise is our deadliest foe.

I'd also like to talk a little bit about the gap that exists between this huge and growing capacity to collect data, and our limited capacity to process that data, and I'd like to offer some ideas on how we might bridge the gap improving our capability to fully exploit and disseminate the data we collect. So I've got a T-shirt around somewhere that I got years ago in London that says "Mind the gap", and those of you that have been there know why they have those, it has to do with the subway and there's a gap between the train and the platform, and so the saying there is mind the gap. So if I had an informal moniker to lay on this talk today let's just call it mine the gap, because there is a gap between what we collect and what we can do with it.

So this is going to be no surprise to any of you in this audience, but today's operating environment is unlike any operating environment we have ever seen before. It's characterized by protracted conflict, constant change, enormous complexity and I would argue profound uncertainty. One of the biggest strategic mistakes we can make in my view is to assume we know the future, because we will get it wrong. Typically we have gotten it wrong.

I asked my staff what were we collectively thinking on the 6th of December of 1941, and what were we thinking on the 8th of December of 1941. What were we thinking on the 10th of September of 2001, and what were we thinking on the 12th of September 2011, and did we get wrong in terms of what we thought could or would happen?

So I think profound uncertainty it goes without saying is a feature today's global environment, and as you look to the future I think what we're going to see is increasing competition for resources, and we're going to see demographic change, and we're going to see continued violent extremism, and unless it's carefully managed we're going to see the continued proliferation of weapons of mass destruction.

Now recognizing that the U.S. has conventional dominance competitors can front our power asymmetrically in all domains. This is not a new concept in conflict; adversaries always seek to exploit your vulnerabilities, challenge your strategy, make your investments irrelevant, and pursue surprises and force multipliers.

We're going to continue to see higher up combinations of strategies tactics and capabilities, and those are going to remain the norm. The battle space is growing beyond traditional geographic boundaries. Certainly its competition and I think eventually conflict includes space and cyberspace, and transregional actors acquire transregional capabilities like ballistic missiles, or surrogates to act on their behalf.

All of this is compressing our decision space, stressing our operational concepts, and stressing our command relations. Now underlying all of this future is the uncertainty of the nations debt crisis and the budgetary pressures that come with it not that will come with it but that are coming with it today, and that are impacting and influencing every single thing that we do.

So this is today's reality at least in my view, our capabilities and our advantages are under constant threat, our resources are increasingly constrained and the danger of surprise looms large. Now I am not a glass half empty kind of a person, because I think if managed correctly these challenges are not insurmountable.

In fact, I think they present us with an opportunity to do some things, refine our processes for example, hone our own strategies and improve our effectiveness. Our own parallel GEOINT capabilities present us with tremendous opportunities to mitigate risks, to response to emerging threats, to prepare for uncertainty, to anticipate surprise, and in return to be surprising ourselves at times and in the places of our choosing.

But before we can really maximize that well spring of GEOINT capability we have to learn to better manage and make sense of the growing volume of data we collect and the intelligence we produce.

So now first the good news: confronting Al-Qaeda, fighting insurgencies in Iraq and Afghanistan and dealing with a variety of global threats has led to a series of leaps forward in our ability to gather intelligence data that continues today.

So the good news is, we are collecting more than 1,500% percent more data than we did just five years ago.

Here's the bad news, we're collecting more than 1,500% percent more data than we did five years ago, and the worst news is during the same time our traditional PED capability has grown by about 30%.

Now conservative estimates predict that the exponential growth of collection growth capacity requires about 100% increase in analysts. A solution that I would argue is not realistic in any environment, but alone today's fiscal environment.

So, how do we mind the gap? How do we process, how do we exploit, how do we disseminate? What do we do with all of this excess data that resides in the gap; do we let it fall to the virtual cutting room floor? Do we collect it and store it for some future use? Do we collect to our PED capability and no more?

Do we take a strict approach to requirements, and collect only what commanders ask for, based on what we think they need? By the way I think there is enormous danger in that kind of approach. Although these approach us as appropriate as each of them restricts our collection capability vital to understanding the battle space. Commanders need more than what they ask for; they need more than what they think they need. Simply put they also need to know what they don't anticipate.

For that's the only way to avoid surprise. Now, latest case in point the Arab Spring, we're we caught by surprise? Certainly, we reacted to the realities but I would ask you did we miss the buzz on social media? Did we only ramp up our ISR collection after the events were well underway in Egypt and Tunisia?

While we welcome the unexpected spread of democracy and hopefully that's the way all of that unfolds, I would argue that at least at some level we failed to see it coming. It may turn out to have been a pleasant surprise although I think that remains to be seen although it was still a surprise at some level.

So we can't let our adversaries take advantage of our inattentiveness or our distraction because of a singular focus on what we want to know, or where we want to know it, rather than what we need to know.

So when we throw more people at the problem, I would argue that that's unrealistic in today's fiscal environment, but even in the best economic times I think we couldn't expect to grow our force structure at the rate needed to keep up with the massive growth of data. The information age, and its very nature is going to outpace the ability of us to throw people at the problem.

Do we throw computers at the problem? Absolutely.

While much work is underway in this field you heard some of it from the last conversation, we need to pursue this option more aggressively. We need to find new ways to sift through the data to find patterns and detect important activity. Automating capabilities might be able to reduce the data to a manageable level for exploitation.

But we need to remember it will always require human beings to make sense of the data and convey its meaning and relevance. Well how about we can manage the PED more efficiently. Of course we can. There are improvements to be gained through integration, and I would offer through bringing the unity of effort to PED management.

Can we short the distance between the commander and the collector, or the user and the collector, whether it's a military commander or not? Certainly we can.

With our advantages in cyberspace we can do this virtually through a federated approach and as you heard, cloud computing. So I what would argue is that the way ahead to mind the gap is to first establish that we don't fear the gap. Those are some of the options that are available to us but I think the first step in framing the issue is to recognize that this gap is not a problem but it's an opportunity. The gap between what we can collect which is huge and growing, and what we can process which is currently finite and constrained, is a reflection of strength not weakness.

We've created the ability to collect the mass amount of data that was never before available to us. Think of the untapped knowledge that resides in that gap. If we can effectively access it we will gain unprecedented levels of flexibility, agility and foresight. It represents opportunities we have never had, and can reveal unknown threats that we might be able to anticipate and forestall.

We should strive to fully exploit this enormous collection capacity that we have so carefully constructed. Now the worst thing that we can do in my view is to view the gap as a problem and try to minimize it by stopping or reducing our collection. Instead of shrinking our collections we need to grow our PED capacity, to close the information gap.

So then I propose a two part effort. First in the long term we have to pursue innovative methods of automation that will generate an unprecedented level of effectiveness. In the short term on the other hand we should take immediate steps to reduce the gap through globally connected, globally focused integration, and by managing the PED the way we manage the collection platforms. In short, in a unified structure.

Now the elegant solution is automation, and it's always easy to say well we'll just go automate this, but I still think our first priority here needs to be to let machines handle the data, sort it, relate it, fuse it, so that humans can handle the decisions, make sense of it, make it relevant, and make it relevant intelligence for users and commanders.

We need to get away from humans doing the sorting because there aren't enough people to do it and it's not efficient, and rather we should leverage innovation, technology, the strength of industry to build automated processes that sort, relate, and fuse the data in ways that facilitate interpretation and decision making.

We've seen this by the way as a current feature of NGA's current mission, but even more emphasis I believe should be placed in this area and it will be difficult to do in this budget environment.

So we wouldn't reach the appropriate level of automation tomorrow, but I think it remains a worthy goal, and it will reward success with a massive increase in head effectiveness.

Now in combination with our automation efforts now and in the future there remains a need to mass our available PED today in an agile way against both our current and our future needs. And of course, massive unity of effort and our time honored military principles I think can be brought to bare here with great effectiveness.

Now the PED architecture that we use today I would argue also is a product of the industrial age. It arose prior to the internet basically when technological limitation meant that analysts had to be located in the region they supported to do their jobs. Collection analysis and decision all occurred locally when not tied to a geographic location like the national stuff was not specifically tied to a geographical location; I would argue PED was often tied to platforms.

So today, I would say geographic location or platform focus should not be the determining factors for our PED, today we are nearly all wired together with nearly instantaneous communication capabilities. Integration of current capabilities and connectedness will be the key to increasing the efficiency of today's PED. While and industrial age PED had to be geographically or platform focused, an efficient information age PED should be globally connected with the ability to be locally focused.

In this way an analyst in Germany for example could focus on EUCOM priorities and rapidly shift focus to a PACOM issue within seconds should the need arise. Bring the information to the analyst, not the analyst to the information.

Now a key question in this regard is who would have the authority to quickly direct that shift in the best interests of the nation? Well there's that pesky authority term, and dang I hate it when that one comes up.

What command or agency exists to manage PED, the way we manage platforms? I think the deafening answer is none. At least not one I know about; that unifying authority does not exist, but I submit to you that it should, and it must, and we should be able to synchronize under such an authority and do so in the best interests of all of us.

Now the PED needs a single focal point, dare I say a single commander that's certainly the way the military would look at it, but let's just say for a moment a single focal point who has the authority to allocate resources and drive the effort from a global perspective, not ignoring the local needs but massing all the capability to do what's needed and by the way I would argue what we eventually do in a crisis after we decide that there is a crisis and we go pull people from lots of different places and we swarm on a problem.

My view is that you can get that same effect on a day to day basis if we manage it differently, and if we have some kind of authority that can do so. And we've done this, we've done this as recently as last spring, where we rapidly shifted assets to sustain support commanders involved in combat operations in CENTCOM, providing humanitarian support in PACOM, and supporting agency operations in Libya. But, we had to go through I would argue too painstaking a process to balance regional priorities, global situational awareness, and rapidly evolving contingencies. What we could not do was allocate PED accordingly.

So now eventually it was worked out and the results were outstanding as they always are. The people in this room made it work, and you always make it work, but I think it could have been more efficient and done with a greater sense of global priority under some kind of a single authority.

So as with our reconnaissance assets our PED synchronization must include both service and agency assets, in addition where such assets exist they should include PED capabilities and the rest of government as well. Those exist and every place I travel I find more of them all the time. What do you do here in agency X? I do essentially GEOINT analysis. Gosh, how do we get you in the fight? I'm in the fight, I'm in this part over here, etc, etc.

So there are probably more people out there that we can get our arms around if we go and seek. So, in these fiscally lean times in particular work in this area I think is vitally important and worthy of our efforts.

You don't want to diminish anybodies responsibilities; you don't want to diminish any of our capabilities it isn't about taking away from anyone. It isn't about taking authority away, it isn't about taking capability away. It is coming up with the way to leverage the capabilities we have, to exploit this gap that is real and is wrong.

So, maybe we could locate this global PED synchronization within Stratcom's Joint Functional Component command for intelligence, surveillance and reconnaissance. Doing so would extend the work being down today with reconnaissance systems. It's a purpose built capability that's inside the beltway with easy access to command services and agencies and it's a common ground for both the military and the intelligence community.

But I only offer that because it's something that could be offered. What I would say more broadly is, there may be other and better solutions, I don't know and I'm not looking for more work for certain. But I would certainly both encourage and welcome such a dialog, as we consider these issues more broadly.

So let me just conclude by saying that the demand for intelligence to reduce surprise in an uncertain world led us down a path that emphasized investment in collection capability. That capacity has exponentially grown, and we now have more data than we can analyze, and we have more data that can actually help us understand and anticipate in ways that we can't even conceive today.

Resource constraints prevent us from throwing more money and people at the gap, problem. Sifting through the clutter to find the nuggets of information, and those nuggets in particular that we really need that are most important is our challenge.

So to do that we need tools that automate processing, that make the data accessible to analysts across a global federated enterprise. I've mentioned some of the options available to us as we seek to effectively manage what is rapidly becoming an unmanageable volume of information. Keep in mind too that all effective solutions will provide balance, and help us get better, smarter, and faster at processing information.

So I would say we will get better by improving our processes, and creating more holistic management of PED in a global federated enterprise. We will get smarter by creating a PED enterprise that is designed for an uncertain future, and this will require us to leverage technology and innovation to rapidly and accurately analyze the vast amounts of data we're collecting. And we will become faster through a federated, fused, and refined collection effort that quickly reaches the priorities of commanders and other users.

Shortening the distance between the PED and commanders, and the PED and users, by arming analysts with direct access to the data they need to be responsive regardless of their location is a goal. Take the information to the analysts, federate the analysts in a global way, don't take the analyst to the information.

I'm confident that we can make the enormous volume of data we collect work in our favor, leveraging the power of automation, flexibility, with federate enterprise. With the clear prioritization of PED and capacity, and the synchronization of collectors and PED, we can chaff, reveal the relevant, and provide a depth of understanding to decision makers.

Well I hope that gives you something to chew on for the remaining hours in the conference, and I certainly hope that it gives you something to think about and send me emails on, for the next couple of months as we go forward. It has been a real pleasure to be with you today and I look forward to your questions.

Thank you very much.