USAF primed to launch new phase of data strategy

The Air Force has full-on embraced data and its possibilities. Listen to acquisition chief, Will Roper, for more than five minutes and you'll quickly understand why the Air and Space Force want to bring the internet of things for the military to reality.

But there is much work to be done to make that happen. Defense Systems talked with Eileen Vidrine, the Air Force's chief data officer, to understand her office's priorities, how they've changed this year, and how the Defense Department's data strategy ties it all together.

The following interview has been edited for clarity and length.

I'd really like to get to know you and explore your priorities as the chief data officer for the Air Force. What have you been focusing on this year? I know that a lot has been going on this year, so I'm interested in understanding what your priorities have been and how they have changed given the events of 2020.

VIDRINE: It's been a really exciting year. Just recently, the DOD data strategy was published, and that data strategy was really a collaborative effort where the DOD CDO, David Spirk [worked] with the military department CDOs -- Department of Navy, Tom Sasala, David Markowitz, who is the new Army chief data officer, and myself. It was really a collaborative effort to really focus on data as a strategic asset. And so that publication has guiding principles for the entire Department of Defense and it uses the core principles which we call VAULTIS -- visible, accessible, understandable, linked, and trustworthy, and the "I" stands for interoperable and "S" is for secure. Those are really the construct [for] all of the work that we've been focusing on in our portfolio; it's really just laser-focused.

The last six months or so have just given us great velocity in terms of drive behind the efforts that were actually starting to mature. So in the fiscal year 2020 we built, in the Department of the Air Force, which now includes two military services, air and space, we built the VAULT data platform, which stands for visible, accessible, understandable, linked, and trustworthy. And really the vision was to have a self-service capability where airmen and space professionals at all levels could access data and use state-of-the-art, open-source and commercial products to really drive insights from data. And so we started with an IL 4 platform, today we're up to an Impact Level 6 [classified] platform with cross-domain multi-tenancy. It allows airmen and space professionals anywhere in the Air Force to log on and do that unclassified work regardless of their physical location.

Using Impact Level 6, our secret-classified solution, it has also actually allowed us to optimize scheduling when we do have to bring people into physical locations. So it's been a really exciting time because we had this foundational capability that we were maturing and we used an agile development model, so every two to three weeks, we're actually building out new capabilities...And a lot of the asks are actually coming from our airmen and our space professionals out in the field. So they're identifying some of the capabilities that they need to really work efficiently.

The other piece that I think is really important is that we built great trusted partnerships, not just across the department, but with our other military services so that we can leverage each other's work and really use it as springboards to the next capability possible. So Tom Sasala and I, we're constantly collaborating.

The strategy has been in existence to the public for a little over a month now, and you mentioned the collaborative effort there. Can you talk to me a little bit about what you're doing to hit the ground running or even what you were doing behind the scenes? I imagine that this has been in process for many months. What is it that you're looking to implement first to make sure that strategy takes hold for the Air Force?

VIDRINE: Now that the DOD data strategy has been officially published, each of the military departments, including the Department of the Air Force, will be publishing an implementation plan. Our implementation plan has been drafted, and it's currently being staffed. And that will give us a road map of actionable items to move forward. One is the data capability of the platform to give our air and space professionals the tools they need to do their work at the physical location that they happen to be working at where we're going to be agile moving forward. But we're also building out the policy piece, we're working on our first Air Force Instruction for Data Management.

We have built a data lab, which is based out of Andrews Air Force Base physically, but geographically people work at various locations right now. And we're actually working on a couple dozen enterprise-focused use cases...from condition-based maintenance to training. Most of our cases are readiness focused. And then the other critical piece to that is our workforce...making sure that we have a workforce that is ready to do this type of work.

In fiscal year 2018, we developed the framework for a data science intern in our department, and we actually brought in our first intern. We continued that last summer, even in the unique work environment that we had. In addition to that, our academy out in Colorado Springs this fall rolled out to find an undergraduate degree program with a major and minor in data science. That's really exciting for us because when our cadets graduate and actually come on active duty, they're going to be in every job field in the department. And so that is going to be an accelerator to really begin a new inflection of great talent, and specifically focused on data science as well as other components. So that blender, that fusion between functional expertise and data science expertise, we're building that capability right now.

We're also investing in the workforce that we have today, running a couple of pathfinders. One is a partnership with the Air Force Institute of Technology based out in Dayton, Ohio at Wright Patterson Air Force Base. They are piloting an online graduate certificate program in data science to upskill and build acumen across our workforce. We also ran a pilot to upskill our data officers in data governance... a short boot camp to get our first 70-plus data professionals in the Air Force at a common baseline to build from when just last quarter. Also last quarter, we ran one of our first datathon partnering with AFWERX. We virtually brought together groups of air and space professionals and worked on a single problem set, specifically, on the C-17 scheduling. That datathon offered a 92% scheduling accuracy rate using automated capabilities, using data sets.

But I think what was the most interesting about that was that there were also some really other great ideas that came from some of the other groups. And we were able to take all of those ideas and capabilities and send them back to our partners at the [U.S. Air Force's Artificial Intelligence Accelerator program with the Massachusetts Institute of Technology] to have some of the best people in the United States that are super smart in artificial intelligence to really continue to perfect and accelerate that type of change. We're now planning for our next datathon, which will happen right after the first of the year, focusing more on data prep.

How has the Air Force been working with other services to meld these different capabilities like in the Army's Project Convergence 2021 experiment, which intends to invite other military services?

VIDRINE: We partner with both the Department of the Navy and the Department of the Army on many of our efforts. And so right now, under this convergence strategy, we have a different name for it. The VAULT Data Platform and the Army Vantage Platform, we can push and pull data from each of those platforms together. So that interoperability was something that we rolled out in one of our most recent sprints because that ability, the technical capability to make sure that we're being able to appropriately share to leverage each other's work is critical.

Fiscal year 2021 has to be about what I call velocity. How do we take what we're doing and really use it to start moving that intentionally and deliver faster. Because we have built a lot of the building blocks to really start moving out smartly. So rather than starting over, then we're actually using the work that's been done. Innovation Connect has been one of the opportunities for us.

How do you plan to keep that velocity going?

VIDRINE: Well, that's where I go back to -- We have a data strategy that's codified. It is the scaffolding, the guidance to help us. It clearly and specifically aligns to the NDS. The concept around our visible, accessible, understandable, linked, trustworthy, interoperable and secure [VAULTIS] -- interoperable is critical as we work across the Department of Defense. And when we start having great high-end quality data, that's consistent across our department that we have confidence in, it's going to build our trustworthiness. And under the federal data strategy action plan, each of the military departments has been working very deliberately on our data maturity models. That baseline work has been done in this calendar year, we are now ready to launch into our next phase of that.

There are a lot of folks that talk about AI and machine learning, in order to do artificial intelligence and machine learning, we need really great volumes of high-quality data and the strategy actually specifically addresses a path forward in that arena. Having that critical document to really be the baseline, the North Star that we work towards, I think is really going to help us stay laser-focused on driving mission results.

I'm a career fed, it's about the mission...and driving change and being good stewards of our resources. Whether it's our workforce is time or whether it's a dollar in our budget, we have to really maximize return on investment.

This article first appeared on FCW, a Defense Systems partner site.