The Navy Reserve is finally moving out of “pilot purgatory.”
This is where new efforts, particularly around data analytics, get hamstrung between the test phase and the production phase.
Capt. Kathleen Powell, the force data officer for the Navy Reserve, said her organization no longer struggles to operationalize technology to provide repeatable value.
“The [Department of Navy’s] Jupiter platform has changed all of that. We now combined the power of cloud-based analytic capabilities of Jupiter with domain expertise. This allows us to rapidly ingest and curate and integrate the data for decision making purposes,” Powell said on Ask the CIO. “The platform gives analysts the ability to develop and share insights. They do that to a broad audience in real time, and this helps senior leaders to not only understand the current operating environment, but optimize it. That’s key, to optimize it through the use of predictive analytics.”
In addition to the analytics, Jupiter lets the Navy “democratize and operationalize” data science, which makes the tools and data available to everyone.
The Department of Navy launched the Naval Enterprise Data and Analytics Environment, called Jupiter, in 2020 in an effort to make it easier for commands and bureaus to take advantage of these tools.
The Navy Reserve began using Jupiter in 2021 to address issues like readiness and training. The reserve established the force data office also in 2021, and Powell came aboard to lead the effort in 2022.
“Our goal is to optimize and modernize how the Navy Reserve uses data to assess improving warfighting readiness of the force. We use data to increase operational advantage and efficiency, with the goal of encouraging data informed decisions at all echelons,” she said. “I’ve been on board for about [seven] months, the office was set up about a year ago, but we’ve been working the data analytics portion of the office for about four years. We believe strongly that we need to leverage our vast troves of data to improve force management, doing things like identifying gaps and deficiencies, using the data to streamline processes and optimizing limited training dollars. Most importantly, we need to use data to mitigate impacts to warfighting readiness.”
Powell works closely with Navy Reserve Chief Information Officer Capt. Chris Peppel, whose office supports the technology infrastructure and tools that the data office and others rely on.
End-to-end decision support services
Peppel said having the right tools to understand and use the data becomes more important as budget and resources become more constrained.
“Where we are right now is figuring out where do we get the data from, who must give us permission, what systems need to be engineered to facilitate that on the other side, just the physical computation side? Capt. Powell and her team have done a lot in terms of acquisition of specific technologies,” Peppel said. “We need something to crunch the numbers and to present the products that will be useful to leadership. They’re not necessarily going to want to see huge datasets with lots of data points in there. They’re going to want to see what’s the bottom line. So that involves creating the product, and once we get the data connections in place and get everything into a lake where we can manipulate it, then that’s where Capt. Powell’s team can take action. We’re certainly not there yet. But we’re working on it. As a previous mentor of mine used to say, in order to make a diamond, you need pressure and time. So we’re getting after that.”
One area where time and pressure has come together is in the Navy Reserve enclave on Jupiter. Starting in 2020, the reserve started to bring data together to provide end-to-end business applications and decision support services.
Powell said the reserve pulls information from multiple domains and uses the tools in Jupiter to “slice and dice” it in different ways.
“Today we’re able to look across those information domains and access authoritative data, collate it and answer senior leader questions, and as one of the first DoN organizations to develop and operationalize advanced analytic solutions on the platform,” she said. “One use case is to enable warfighting readiness through training and through operational support. To that end, the command manages a large training budget annually, and due to the high and annual variability in missions and also in participation rates among selected reservists, allocations of that budget are extremely complex. We want to make sure that we optimize it, use it to the best of our ability to drive that warfighting readiness and make sure that it’s being used for that purpose and for increasing our readiness writ large.”
“To answer these questions, we’ve created solutions in Jupiter that link multiple data sources and distill millions of data points and then provide intuitive visualizations,” Powell said. “This allows senior leaders to view the data from a number of perspectives and then they can evaluate participation metrics by rank, by rate, by command, by the mission that people are doing and then just see if they think we’re going down the wrong path or what the money is actually doing the most good for.”
Reserve leaders get the data in real time that leads to fact-based decisions instead of opinions or hearsay, which sometimes happened before the Jupiter implementation.
“Our team has delivered over 80 data analytic solutions along with our more recent effort to consolidate our reserves specific data into Jupiter, which was a nice big project of its own,” Powell said. “We measure impact in several ways. We do that through operational efficiencies such as cost avoidance and cost savings, and the time savings which drive those savings right, as well as intangible benefits — which are a little more difficult to measure — but intangible benefits such as improved quality of life, which is increasingly more important in a tight labor market, where individuals have many options on how they’re going to spend their employment hours.”
Creating the data lake
Powell said moving tools to the cloud and relying on application programming interfaces (APIs) have accelerated the time to decision for her staff and leadership.
“In the past, we tried to do this stuff on our desktops or on basic laptops, and we really were restricted in the modeling world. It’s fantastic today to be able to work in this environment and it’s also allowed the high skill data science talent that we have on board to actually exercise in the realm of the possible with the data that we have available to us,” she said. “On the software side, I mentioned access to powerful analytic data science and visualization tools certainly enable our users to integrate, share and exploit the data. More importantly, exploit it at speed and scale because we need to be able to do this stuff fast for operational advantage and also for increased efficiency. These are all drivers of warfighting readiness that we bring to bear, leveraging these more modern tools that we’ve been able to bring into the mix.”
A key piece of this data analytics puzzle is the creation of a data lake that Jupiter or any other tool can pull information from.
Powell said data comes into the lake on a daily basis through semi-automatic and, more and more, direct feeds and it is standardized before it reaches the platform.
“That’s been great because now our data is persistently accurate, persistently being updated so our interfaces and our visualizations can be immediately updated with current data so that someone could get a daily update, for example, on questions we get all the time, like what’s the average number of days people have done this year?” she said. “We used to have to have people go running around and asking each other and then pulling the data and everything together. The data is just going to be there persistently. Then as long as we have the metric we established, then you can get today’s answer at any point in time.”
Powell added the reserve also is applying artificial intelligence and machine learning tools to improve the data’s value and speed to decisions.