Main Subcommittee Category
Policy, Standards, Management, and AcquisitionHow would you label your submission?
GovernmentAbstract
The digital world is grappling with the transition in two vector spaces: 1) centralized to distributed architecture and 2) the continuum between human and machine actors. This evolution requires deliberate action and contextual awareness of mission space and local restrictions such as access-denied environments, semantic conflicts, Intellectual Property (IP) sensitivities, coalition partnerships, and machine interoperability. Each of these is a different aspect of the uphill battle to establish a data mesh responsible for negotiating across physical, regulatory, ontological, and digital borders.
Extending the data mesh (“mesh”) beyond its heretofore idealistic yet ambiguous description in research, this paper will illustrate the viability and interoperability of composable foundational building blocks for any organization’s mesh, proposing a bare minimum of additional mesh components to establish a data-mesh-worthy technical backbone. The paper will provocatively challenge any single data mesh instantiation, pushing for the ideal of fully-federated domain independence bridging proprietary instantiations and illustrating seminal proof of concepts being tested in real-world scenarios.
Ultimately, a mesh allows digital twins and other digital representations to scale. It knits the connective tissue among disparate systems, builds collective interoperability, and retains the privilege of local privacy and autonomy. This combination is critical in training and warfighting scenarios, during which multiple simulators built by various entities must interact and share namespaces with minimal to zero delays.
A sophisticated technical and strategic response with a well-formed and flexible implementation plan is necessary when linking IP-sensitive proprietary modules with context-tuned data management practices, compliant with regulations yet differentiated from peers. Adopting agile and continuous learning principles, this paper approaches the reference architecture, reference design, and implementation plan for building a data mesh enabling the modeling and simulation community to make seamless connections both 1) between humans and machines and 2) across technical and regulatory boundaries.