The Department of the Navy’s first Chief Data Officer, Thomas Sasala, was designated in his position in October 2019. He previously served as the director of the Army’s Architecture Integration Center and chief data officer for the Chief Information Officer/G-6 office.
In November, Sasala was a keynote speaker at the Naval Information Warfare Center Atlantic Data Science and Analytics Workshop where he described the drivers for a modern naval data environment and provided a high-level overview of the naval data management concept of operations and the proposed naval data architecture.
Naval data management is centered on 12 information domains, he explained at the workshop. The DON data governance board (DGB) governs all naval data policies, regulations and guidance. Each domain will be managed by a data steward through the DGB, and decisions and policy will be made through the governance structure.
Sasala joins Chief Technology Officer Jane Rathbun (and DASN for Information Warfare and Enterprise Services), Chief Digital Innovation Officer Michael Galbraith, and Chief Information Security Officer Christopher Cleary under the direction of DON Chief Information Officer Aaron Weis in a mandate based on the SECNAV Cybersecurity Readiness Review to modernize and secure the DON’s IT/cyber infrastructure and maximize its use of data in decision making.
CHIPS senior editor Sharon Anderson interviewed Mr. Sasala in mid-January.
CHIPS: Can you talk about how you are working within the new DON CIO directorates and other stakeholders to implement the changes necessary for a digital transformation in the department?
Sasala: Aaron Weis, the DON CIO, has established three big broad categories we are working on which are modernize, innovate and defend. So we are organizing under the modernize, innovate and defend themes to catalyze our different initiatives. I am particularly working closely with the three other division chiefs, but specifically with Mike Galbraith, who is the chief digital innovation officer, to integrate how we are going to govern and change data management across the Navy and Marine Corps.
I am working with Jane Rathbun to establish an enterprise data environment. What that means is realigning some of the programs that we already have inside the department, as well as creating new capabilities, in the form of new IT services, new business services, to allow for greater data integration and to manage data more effectively.
We started with a data architecture that defines a suite of capabilities. Jane is taking those capabilities as the CTO and identifying existing programs that provide that type of service and where the gaps are. We are seeking funding in what Mike is doing on the innovation side. Aaron is working on a realignment of the IT budget to fill those gaps.
CHIPS: Are you starting from scratch or was there already something in place?
Sasala: There was nothing in place from a data management perspective. I came into the job in April and the undersecretary, just a few weeks before that, asked us to set up a Data Governance Board to provide context to the data. We wrote a concept of operations and lined out just how the Data Governance Board fits in the larger data management structure and then we began to weave in things, like artificial intelligence and machine learning, into the data management plan. We are trying to leverage existing, pre-existing architecture artifacts, and the pre-existing IT solutions that were there as much as we can. But the concept of operations and the data management plan didn’t exist prior to this. We are trying to stitch them all together to provide those CIO themes (lines of effort) across the different buckets of activities.
CHIPS: Can you talk about the 12 information domains and your short-range objectives to increase the security and usability of DON data?
Sasala: In the concept of operations, we identified that rather than managing and governing information and data by organization or by programs or systems, we are going to start putting like information together and managing like information as a dataset. For example, one of the information domains is financial management, another one is medical information, and another is legal, law enforcement information. They are three of the high-level information domains. What we want to do is, from a policy and procedural perspective, manage all financial information similarly, all medical information similarly, and so on, no matter which organization holds the information and no matter what system the information is in.
We have taken control of these information domains and provided data stewards, who are responsible for identifying where the information needs to go, what systems need to go in those information domains and what policies and processes do we need to develop around the information domains.
The short-term objectives in that regard are really coming up with use cases and mission threads for an information domain and then developing questions people might want to ask about the data so that we can identify what systems hold that data and then prioritize some of our activities around what we are calling ‘high value targets’ for information.
For example, one of the questions is, How much are we spending on IT equipment in the Department of the Navy? It is a legitimate question and one that has been asked many times. We would use this question and direct it to the variety of different databases we have in the department. On that specific question, we have identified between five and eight different systems that we have to pull information from, integrate it, and then produce an answer. The goal is to put that in a data environment where the information is pre-integrated, so that every time someone asks the same question, we don’t have to go into eight different systems and pull a bunch of different spreadsheets, integrate the information and hope we have our analysis correct.
That kind of theme repeats across every one of the information domains, whether it is readiness for our different carriers and strike groups, financial management, or budgeting and programming. These are just examples of the type of things we are working on right now in the different domains.
CHIPS: I understand there will be a DON Implementation Plan based on the Department of Defense Data Strategy. The DoD Data Strategy, which sets the vision, guiding principles, goals and objectives to transform the DON into a data-driven organization, is expected to be signed any time now followed by the release of the DON I-Plan in early 2020. Can you discuss the lines of effort that will be part of the I-Plan?
Sasala: The DoD has not issued the data strategy yet. Based on a meeting I had last week [mid-January], the tentative schedule for release is in February when the strategy will be signed by the Deputy Secretary of Defense.
The Navy is not going to have any form of a data strategy nor will the Marine Corps. Instead, we are going to write an implementation plan based on the DoD strategy.
We have drafted an implementation plan. We have been socializing it, collaborating on it across the department, here at the Secretariat level as well as in the Navy and Marine Corps. We sent it to a number of different people, as well as the 12 data stewards, asking for input on some of the initiatives and tasks.
Within the implementation plan are five lines of effort. Line of effort 1 is building the foundation; LOE 2 is evolving the workforce; LOE 3 is positioning and protecting data; LOE 4 is building, optimizing, operationalizing and rationalizing the environment; and LOE 5 is managing and governing the environment.
A lot of the work we are doing within the information domains falls into a combination of the first line of effort, which is setting the foundation, as well as the fifth, which is managing and governing the environment. Then what I mentioned earlier, which is looking into the programs and leveraging what is already there, which falls into the positioning and protecting and building and optimizing.
The other component that we are working feverishly is evolving the workforce. This involves creating literacy across the department at the lowest level up to the highest level as we educate the workforce on basic data fundamentals, as well as subsets of the workforce, on what I’ll call higher order capabilities. We will be training data scientists, data engineers, data stewards, as well as functional domain managers, on the nuances of the algorithms — how to generate things like machine learning, deep learning algorithms — to a smaller subset of folks.
The idea is to get everyone in the department, regardless of your grade or position, an awareness of data accuracy, data timeliness and their importance to the department. We want them to know how they individually contribute to the DON’s broader initiatives by doing their jobs more efficiently, more effectively; making sure that whatever they enter into their systems is accurate. That is the I-Plan in a snapshot.
One other thing, the initiatives in the I-Plan will be either added to, or as subsets to the information systems that are already in the Business Operations Plan that the Acting Secretary is using to guide and drive the department. The Department of the Navy Business Operations Plan is one of the fundamental documents that we are deriving these initiatives from as well as putting initiatives into to help the department move forward.
There is going to be a lot more activity in a couple of months with the next release of the BOP. Aaron is going to add a number of initiatives from the DON CIO perspective into the BOP, in addition to the things that are there just for data management. Right now, we have eight initiatives in the BOP for data; he will be adding some for every one of his lines of effort.
CHIPS: Will the I-Plan define a desired end state for the DON’s digital transformation?
Sasala: We are going to establish metrics for each of our objectives and we will monitor our progress against the metrics and the objectives. We will be able to show either we are evolving or transforming over time. I don’t particularly care for the term ‘end state’ because it implies that you are finished when you get to a certain point. I would say that we are going to establish a target date and we are going to monitor and manage our progress toward the target date. I suspect that when certain aspects of the data environment hit their target metrics we will update the target metric, change it, and move it out a little further into the future.
Some things are hard to measure quantitatively. The completeness of data is subjective based on the mission and the questions that you are trying to ask. Timeliness is also subjective based on mission needs.
Many of the initiatives in LOE 1 of the implementation plan call for us to establish metrics and then populate the values of the metrics. We will be briefing the Acting Secretary quarterly on the values. The domains that are not hitting their targets or showing progress will get special attention.
CHIPS: Put another way, will we know we are in a digital transformation?
Sasala: I hope so. The idea is once we clean up the data fundamentals and make data more readily available and accessible, we are able to use that data to make data-driven decisions. For example, Mike Galbraith is working on things like artificial intelligence. He has objectives that he wants to achieve but without having the data accessible, clean and curated and kept up to date, any artificial intelligence or machine learning activity is not going to be as productive or impactful as it could be.
You will see more things moving from the evolutionary to the revolutionary and transformational stage in the next six months or so. The biggest impact will be in about 12 months and over the next two years on the data side.
CHIPS: Is there anything else you would like to discuss?
Sasala: The only thing I would add is we have one Chief Data Officer for the Department of the Navy, and we will have two deputies, one each for the Navy and Marine Corps. They are not going to be called chief data officers; they will be called ‘DON Deputy Data Officers’ to align with the DON Deputy CIOs. The consensus is we want to avoid having many different chief data officers and just have data officers who are responsible for their portion of the organization.