Over the Horizon (OTH): Big Data analytic constructs figure significantly into operating in a more fluid and integrated multi-domain future. What are your thoughts on how that future system would help identify cross-domain threats as well as opportunities for cross-domain pivots and action?
Dr. Jon Kimminau (JK): I hadn’t tried to bridge these thought streams before and it’s kind of a challenging question. I think there’s no doubt that if you bring all the data together you ought to have no domain boundaries, right, and that certainly seems to be—at least in one sense—part of what I was talking about in breaking down all those stovepipes on the data entry side of things anyway. So if we have all that data, we should be able to do cross-domain thinking, shouldn’t we?
Well, maybe not, though. There are three challenges for us in data analytics, even in our approaches, to make sure to pay attention to if we don’t want to become domain focused in what we’re doing and I think the first is stovepiping the data. So one of the challenges that’s out there at this early stage, you know I mentioned all these different projects and campaigns that are going on as bubbles. Well, each one of them is also, although it’s Big Data they’re only going after particular types of big data, particular sources. They aren’t actually going after all data yet. So as soon as you say I’m only going after some data, you have immediately put some boundaries up. You’re limiting your domain somehow.
So hypothetically, we could be applying an activity based intelligence approach, but the only data we’re looking at is data for, let’s say, the Middle East. Well, we can certainly get a lot of value out of that, but we are probably not going to see things crossing, let’s say hypothetically, to Russian operations in the Ukraine or even up in the Baltics. We aren’t going to be able to see any connections there if we are only looking at that area in terms of data. So that would be the number one challenge: if we stovepipe the data, we aren’t going to be able to do cross-domains.
The second challenge would be in, “How we might stovepipe our questions?” If we, let’s say in our program of analysis or in our national intelligence priorities framework, if we’re asking the questions and saying pay attention to country X, we may have created some boundaries there to seeing cross-state or cross-functional or cross-domain activities just because the questions we’re asking people to pay attention to immediately sets some boundaries.
And the third challenge, which may sound related to that, is the organizing people issue. Let’s say I have a shop that is Africa focused and we may do great at seeing things across African nations, but we aren’t necessarily going to see things, or we may have an impediment to seeing things, crossing from other nations into Africa or other continents.
So that was my first reaction to your question, I certainly think the foundation is there, I think it’s obvious the foundation is there for us to do cross-domain analytics, but the dangers are let’s say domain boundaries on either the data or the questions we ask or how we organize our people to do the business.
OTH: Thinking about these three challenges you mention to using Big Data analytics to enable fluid multi-domain operations, does this have implications for military Services that are organized primarily by domain (Air and Space, Land, and Sea)? An Air Force leader is going to ask questions that are primarily air related and organize their intelligence force around answering air-related questions.
JK: Definitely. To me, the things that people aren’t openly talking about or paying attention to are issues like this. Related to this is the history of the Intelligence Community and the organizations we have. They are fundamentally about intelligence sources and so we have people in each of these big agencies who work on that particular source, have specialized in that source. If we truly reach a point where we are sharing all the data, have we not just removed one of the fundamental purposes of the separate agencies? Do we need to rethink how we’re organized, perhaps not organized by type of intelligence, but organized by something else? I don’t think anyone is looking at that, but I think it’s a brick wall we’re going to hit in the near future if we realize what we’re trying to do.
Related to that is a question I got asked by a Senate staffer when I was briefing our vision of where we want to go with data analytics. He asked me: “Well, okay, I get it and yeah, I can see how the analysts would be better equipped to answer more and do it better, but what drives what they look at to do this? If they can access all data, then why wouldn’t Sue and Jim and Bob, even if they’re at three different places, all be chasing that really exciting activity over in Country X? Or why wouldn’t all of them be looking at what’s happening in the Ukraine? What drives them to look at different places?”
I call this the children’s soccer game. We don’t have everything in place, once we provide everybody all the information to go look at, we don’t have necessarily anything in place to stop it from becoming a children’s soccer game where everybody gathers around the current most exciting question and tries to answer it in all their 10 or 12 different ways. So, that’s another over the horizon thing I’m not sure we’re grappling with. If we break down the boundaries in the data, and you said it before, how do we now organize our people? And what missions do we give them?
OTH: Related to this, there is a sense that advanced analytics will compress the decision-making cycle. How do you see future analytics impacting decision-making and the time component?
JK: This is another interesting one and I hadn’t thought about that in depth either, but it kind of leads me to philosophy again. Does data analytics compress decision cycles? Well, my first thing would be what causes us to wait on decisions, and do they wait on getting more information or do decisions get paced by events? Philosophically, I lean toward the latter, that decisions are paced more by events than they are by the amount of information we have. Is data analytics going to change the pace of events or is it going to change the amount or quality of information we have? I think data analytics changes the latter which would lead me to, kind of my first blush answer to the question: “Does it compress decision cycles?” I don’t think it does. That may be a radical answer, but I don’t think it compresses decision cycles, because I think decision cycles are driven by things outside what we put in our Observe-Orient-Decide-Act cycle. The things that influence us to complete a cycle are external to those four steps. I think they’re events that drive us to say “I have to have a decision now.”
It also relates to this old adage I used with people as a commander and supervisor, you know sometimes people do sit around saying, “I have to have more information, I have to have a study before I decide that.” And my old adage is that if you have full information you no longer have a decision. That’s not what decision making is about. Decision making is always about making a choice when you don’t have full information. So I guess it relates to this in the sense of, again, I don’t think better information-generating tools affect our decision cycles.
OTH: A final question we ask all Over the Horizon guests: what is something just over the horizon that the international security community should be paying close attention to or trying to figure out?
JK: Well, this has nothing to do with analytics, but I personally am still surprised that we haven’t seen a terrorist set off some kind of nuclear or radiological weapon, because there is no concern for life there, there’s no concern for many of these guys with what they do. I’m just personally surprised we still haven’t seen that happen and I just keep expecting to see that around the corner or over the horizon. It doesn’t relate to data analytics, but that’s one that I still think of.
Jon “Doc” Kimminau is the Air Force Analysis Mission Technical Advisor for the Deputy Chief of Staff, Intelligence, Surveillance and Reconnaissance. He is a Defense Intelligence Senior Leader (DISL) serving as the principal advisor on analytic tradecraft, substantive intelligence capabilities, acquisition of analysis technology, human capital, and standards. Previously, he served nearly 30 years on active duty as an Air Force intelligence officer. Dr. Kimminau holds a Master’s in Public Policy from the Kennedy School of Government, Harvard University, a Master’s in Airpower Art and Science from the School of Advanced Airpower Studies (SAAS), and a PhD in Political Science from the Ohio State University.
Disclaimer: The views expressed are those of the author and do not reflect the official policy or position of the Department of Defense or the U.S. Government.