The Department of Defense has laid the foundation for better data management and governance through its data strategy and subsequent data decrees, but there remains room for improvement.
The DoD’s Data Strategy focuses on operationalizing data to improve decision-making across the DoD enterprise, which will support mission operations and maintain military advantage.
As part of its strategy, the DoD established a chief data officer (CDO) to oversee the data governance process, data standards and policies, and improved data acumen across the DoD workforce. The DoD’s Data Strategy also created a CDO Council and federated data catalog for more unified data management. These steps are crucial to determine what data is sensitive and needs protection, as well as who can access it and under what circumstances.
In DataOps and DevOps spaces, a key focus has been on balancing security and user access requirements while removing operational bottlenecks. This is accomplished through automation workflows, which simplify data access on cloud architectures. Consequently, the cost and elasticity benefits of the available storage and compute will accelerate analytics modernization. To accomplish this, data teams should use the DevSecOps approach, which ensures a proper balance between data security and utility.
To accelerate data modernization, the DoD workforce needs the right tools, environment, resources and access to establish data-driven programs. Successful data-driven agencies are creating enterprise services for data discovery and establishing guidelines and guardrails to ensure that costs, the protection of sensitive information and legal compliance are effectively managed.
Opportunities exist for the DoD to learn these best practices and apply them within agencies. Here are five areas DoD decision-makers should consider when implementing a data management strategy.
Bring in highly qualified experts from outside the DoD who can introduce new ideas and lead innovative efforts.
The DoD is assessing the talent pool to identify civilian and military data professionals, data fluency across the DoD and critical skill gaps within its workforce. The goal here is to establish additional public and private partnerships — particularly those outside of the typical contractor community — to quickly overcome the perceived deficit. This presents opportunities for DoD leaders to gain new expertise.
Provide additional innovation funds or special projects to DoD components to develop prototypes and test innovative approaches.
The DoD and military services already use Other Transaction Agreement (OTA) contracts to build prototypes. OTAs cut through bureaucratic red tape and speed up prototyping and the delivery of new military capabilities. Agencies should evaluate OTAs and other acquisition and funding programs to build innovative prototypes.
Encourage participation in conferences and events that highlight successful applications.
Conferences that provide the latest insights into data readiness and analytics, data governance and compliance, data tools, and transitioning to a data-driven government are growing. The DoD workforce should attend private and public sector events to stay informed about the latest trends and technologies that improve data-driven decision-making.
Invest in workforce training for the cloud.
To operationalize data, agencies need a modern cloud infrastructure and native services to catalog and automate the data discovery process while harnessing the power of artificial intelligence (AI) and machine learning technologies. The DoD is transitioning from a single cloud provider to a multi-cloud provider with the Joint Warfighter Cloud Capability (JWCC) and the CIA’s Commercial Cloud Enterprise (C2E) contract. This shift will provide more services and increase the digital footprint of defense agencies. While the DoD will see benefits during this transition, it will also increase the burden on IT administrators and security teams, who will require more training. Making workforce development a priority is essential.
Automate data access for analytics at the data layer focusing on zero trust.
Historically, zero trust has focused on automation and continuous network and application monitoring. This works for transactional system architectures but leaves a security gap in analytic environments, as data from source systems is co-located. The only way to ensure proper access is to deploy granular, attribute-based access control (ABAC) security models at the data layer. This design should take into account not only user attributes but also data attributes, while using tools to automate relevant data access, governance and privacy policies at scale. Role-based access control lacks the granularity needed and requires significant effort to establish and maintain. By automating access, agencies can reduce current DataOps bottlenecks that affect users.
AI’s growing role and the future of DoD data
Defense leaders realize the promise of applying AI in their agencies, as priorities have shifted to near peer adversaries. Successful application of AI has identified cancer cells in medical imagery, and the DoD has had success with predictive maintenance for aircraft and trucks. However, images from remote sensors on satellites or unmanned aerial vehicles have their own challenges, including weather or camouflage. DoD officials must be realistic about what AI can do. To that end, the DoD’s Joint Artificial Intelligence Center (JAIC) office is launching the Artificial Intelligence and Data Accelerator (AIDA) to help DoD commanders better understand their data to improve operational decision-making and identify use cases for combatant commands. The goal here is to advance what is possible with AI while applying the technology to solve operational challenges within the DoD.
DoD agencies are currently deploying a zero trust architecture across government and partner networks. If the DoD does not find a way to balance security with utility, its data challenges will only increase. Automation, cloud services, an authoritative source for data access management and a highly skilled workforce are key ingredients to supply the tools that analysts need to optimally manage agency data.
Danny Holloway is the public sector field chief technology officer at Immuta.