Sponsored

DOD’s Data-Driven Future: Shared Knowledge, Near Real-Time Answers

The impact of data-driven decision making reaches far beyond strategic planning. While data is essential to creating and maintaining a competitive advantage over adversaries, data-driven decision making applies to everything from warfighter responses under battlefield conditions to logistics and maintenance.

Complete, accurate data is crucial to supporting the OODA Loop (Observe – Orient – Decide – Act), which was developed by USAF Colonel John Boyd. However, only timely data is useful — information that arrives too late to support a decision is as good as no information at all. The speed of access to shared resources is the gating factor. To be effective, answers must be generated in seconds, not hours or days. And those answers must be forward deployable to support troops and commanders as events unfold, so security is imperative.

The first two parts of this series discussed how meeting the goals of shared data mandates, including ABMS and JADC2, will require users across the DoD to be able to access authoritative, up-to-the-second data from across all sources. Those articles also covered how data can be secured and compartmentalized, and can also amplify cybersecurity, helping to mitigate threats before they impact the mission. In this final article, Elastic looks at what near real-time answers from comprehensive data can mean for real-world situations.

Start at the Source

Developing an all-encompassing view requires diverse sources, including data from computing systems as well as the information captured from operational technology (OT). The sheer breadth of OT systems providing data — everything from flight line diagnostics to security endpoints, and even including seemingly unrelated systems such as HVAC and badging — can provide a wealth of information relevant to readiness, long-term planning and day-to-day operations.

  • The Internet of Battlefield Things (IoBT) has a measurable impact on both the quantity and quality of information that can support better, faster decision making; sensors (both stationary and mobile), on-vehicle telemetry, and weapons systems can feed continuous data streams to command and control systems.
  • Wearables can also provide essential inputs: Location data and personnel health monitoring are two obvious uses, but they can also enable essential tasks such as contact tracing for disease and physical threat prevention.
  • To ensure overall readiness, the Secretary of Defense has mandated an 85 percent operational level for airframes and vehicles. Reaching that goal requires predictive maintenance, which is data-driven by definition. But identical parts can perform differently in real-world applications due to environment and mission requirements. Today, many predictive maintenance efforts use standard data science practices, which can take weeks or months of analysis. AI-based solutions can deliver accurate results in a fraction of the time.

Comprehensive OT data can also drive inventive new capabilities. A prime example is the Air Force’ Base of the Future initiative. Intended to protect infrastructure and enable resilience, the program encompasses data-driven processes such as improving physical and cyber defense; using emerging technologies, including AI, to promote innovation; monitoring personnel well-being; and ensuring legacy systems can be supported while building a more agile direction for the long term. All of this is dependent on fast, secure data sharing to enable more effective automation and better-informed decisions.

The practical applications of this interconnected web of data are limitless. Transportation commands, for example, can employ machine learning to view an entire fleet’s operations simultaneously, allowing leaders to identify recurring issues while factoring in terrain, vehicle health, and personnel performance. Geofencing can be applied to highlight problems affecting a specific area. Capturing data from as many relevant sources as possible can even lead to the next stage of AI-powered capabilities: prescriptive analytics — in which systems can respond automatically even when dealing with changing conditions.

With Elastic’s machine learning and natural language processing querying capabilities, users can cross-correlate multiple data sets that normally would not interact to determine what’s outside of the realm of normal — mitigating analysis paralysis and driving effective, budget-conscious decisions.

Seamlessly Blend Legacy and Modernized Data Sets

Hybrid environments, composed of both legacy and cloud systems, will be with us for the foreseeable future. There are simply some proprietary systems that can’t be replaced by cloud solutions and some analog capabilities that cannot be superseded by digital ones. The MIL-STD-1553 aircraft communications bus comes to mind. This standard dates from the 1960s, yet nearly every helicopter’s communications system relies on it, along with airframes such as the F-15 and F-16, among others. Even the emerging standard, MIL-STD-1760, will remain analog for operational effectiveness under combat conditions.

Elastic can manage data captured by these essential systems, then correlate it with information from modernized systems. The solution — standardizing and normalizing data formats using an open schema — permits easy access across resources. This means there’s no need to rip-and replace valued systems, but the data can still be used wherever needed.

Empower Users with Near Real-Time Search

The majority of data queries have been limited by two factors: access to shared data and time. Most of us are familiar with the overnight question—it takes all day to figure out what to ask, and the system takes all night to come back with an answer. The “coffee question” can be answered more quickly but leaves plenty of time to say, “I’ll go get a cup of coffee while I wait.”

By presenting answers in just seconds, Elastic empowers users to ask “What if?” questions, veering away from the status quo in search of better solutions. For example, if a cyberattack is detected against an Air Force base, can that same attack be identified elsewhere? Or if an artillery transport truck deviates from its schedule, can it quickly be determined if the issue is an engine failure or a possible threat?

Grant Access … Securely

Maintaining security and privacy is crucial to data-sharing strategies. Not all data needs to be shared, and the effort involved with copying and moving data to centralize it is simply unnecessary. Elastic operationalizes the data lake concept by allowing information to be kept in place and accessed by whomever needs it — but only per the data owner’s policies. This keeps local data local, shares what’s important to the group, and ensures that private or secure information remains so.

Start Today: It’s About Time

Because timeliness is critical, it’s crucial that the most up-to-date data is included in queries. Enterprise search solutions typically index data as it’s requested, which could mean the latest information fails to be considered. That’s why Elastic indexes data as it is ingested, so it is immediately available to be analyzed and queried.

Systems powered by this comprehensive search capability also empower junior operators to understand and identify issues and take appropriate action faster, while freeing up higher-level operators to focus on more complex challenges.

With both personnel and automated systems hungry for information, access to data across the enterprise frees people up to focus on the decisions that only humans should make, especially those involving extreme consequences. Elastic’s search and analysis tools highlight details that could otherwise be obscured, while answers can be as granular as necessary. With answers available in seconds from all relevant sources, faster, smarter decision-making is only a question away.