Leveraging Data Science, Delivering a Data Driven Decision Model Providing Facility Capabilities Meeting Mission Needs

The Air Force’s (AF) installations are its power projection platform, their safeguard is critical to meet national strategic objectives. Guarding AF bases is dependent on a robust and sustaining stream of weapons qualified personnel, making small arms ranges a mission-critical capability. Small arms qualification is enabled by small arms ranges, ensuring their sustainability and modernization involves consideration of many variables and management criteria.

Retired Senior Master Sgt. James Eselgroth (right), and Eric Kennedy, Small Arms Ranges action officer, identifying the magnitude of the situation of the small arms ranges in 2017.
Photo by: Vicki K. Stein | VIRIN: 170626-F-GX074-001.JPG

This video will detail how this D3 model came to be and what barriers had to be overcome. Data from three separate enterprise databases had to be combined, but this only provided ~40% of the answer. Additional information and analysis techniques had to be obtained by other means or developed to generate the information needed to perform the analysis not found in AF enterprise systems. The effort was worth it, the analysis generated was decisive in improving overall operational readiness coupled with the prudent use of taxpayer dollars. We did so by right-sizing the inventory by size and type; investing in those mission areas needing it most, without a loss in training requirements, while also identifying opportunities to optimize the utilization capacity, reducing waste by 19%, through small investments and leveraging community partnerships while simultaneously correctly configuring ranges. The total planned cost prior to this endeavor was $1.2B, with the use of the D3 model and new governance structure we saved $246M, precious dollars can now be spent in other mission critical areas.

Participants in the M4 rifle qualification course lay in the prone position in preparation to firing their weapons at the 99th Security Forces Squadron Combat Arms Training and Maintenance firing range at Nellis Air Force Base, Nevada, Aug. 18, 2015.
Photo by: Airman st Class Jake Carter | VIRIN: 150818-F-UN699-135.JPG

Delivering the D3 model, attendees will learn how we leveraged Data Science. Some of the techniques we used such as decision tree analysis, multicriteria analysis, predictive analytics, domain expertise, linear algebra, algorithm development, and storytelling with data were critical in delivering the D3 model. The use of all these Data Science techniques allowed for the development of the D3 model so stakeholders at all levels can see and understand the complex relationships associated with the 2d, 3d, and 4th order effects while making a decision of this magnitude.

Figure 1. Data Driven Decision | Model Process

Throughout this presentation, the D3 model will showcase the AF small arms range example, but fundamentally the making of this discovery process universal applicability will be highlighted. The outcome of the presentation will be a structured logic flow that can be used to apply the process to develop the D3 model to many problems involving facilities, funding, operational risk analysis towards developing long-term investment strategies.

1. How does the presentation relate to ISO 55000?

The approach used, and will be presented, follows the intent of ISO 55000 by leveraging the owner/operator of the asset capturing the purpose of the facility. Combining the operating context of the asset within the Planning, Programing, Budgeting, and Execution process assists civil engineers in their management of assets. Together the shared responsibility between the mission owner and operator of the asset, the civil engineers and their responsibility of caring for and managing the asset, coupled with the financial managers on pursing the funds needed for the delivery and sustainment of the aforementioned all support the intent of ISO 55000. All three of these key players assists the senior leaders to make a more informed decision on where the use of the taxpayer dollar should or should not be spent.

2. What asset management best practice(s) will be featured?

Best practices we will feature are the use of data science tools, techniques, and procedures delivering a data driven decision (D3) model. Some of the techniques we used such as decision tree analysis, multicriteria analysis, predictive analytics, domain expertise, linear algebra, algorithm development, and storytelling with data were critical in delivering the D3 model. The use of all these Data Science techniques allowed for the development of the D3 model so stakeholders at all levels can see and understand the complex relationships associated with the 2d, 3d, and 4th order effects while making a decision of this magnitude.

3. How does the presentation relate to the theme Smart, Secure, Sustainable?

Using a D3 model is a smart method towards Augmented Intelligence. Where Augmented Intelligence is the process of crashing the information gathering process by helping people make better decisions through data driven decision making, focusing on objectivity before subjectivity, transforming organizations from reactionary to proactive ones through predictive and ultimately prescriptive analytics. The use of D3 is a sustainable process as many of the tools, techniques, and procedures used by data scientists can be found throughout industry. Finally, tools like the D3 model are designed to be in the closed loop of an organizations data stack making it a secure approach to delivering information for D3.

Connect with one of our consultants to see if a data driven decision solution is right for your organization.