Differentiating Cloud Capabilities for JADC2

The Joint All-Domain Command & Control (JADC2) concept is the DoD's north-star aspirational vision to achieve decision dominance against our adversaries. Senior leaders recognize the critical need to accelerate the application of data, edge, cloud, and other technological solutions to achieve the JADC2 vision. Leaders now understand the technology stack complexities and appreciate the distinctions between the app layer, the data/AI platform layer, and the infrastructure layer. For decades, the DoD acquisition process only focused on requirements at the app layer and as a result, capabilities became siloed vertical stack solutions that cannot integrate with other capabilities. Leaders now recognize the importance of defining requirements at the infrastructure and data/AI platform layers to allow for portability, better integration, and cloud-native capabilities. The DoD wants to create the technology stack so that there is a hard decoupling between the app layer and the data/AI platform layer as well as between the data/AI platform layer and the infrastructure or hardware layer. This desire creates a complex decision space in a multi-cloud, hybrid, and disconnected edge environment involving three tradeoff variables: generalizability, complexity, and usability.

Generalizability is about creating a competitive environment among cloud solution providers, the Defense Industrial Base (DIB), and the rest of the industry to prevent vendor lock by decoupling the dependencies between the technology stack layers. Complexity increases when the Government inherits the management of the decoupled layers while minimizing the use of Platforms as a Service (PaaS) and relies more on multi-cloud Infrastructure as a Service (IaaS) solutions to manage common services and third-party platform capabilities. Usability is about maximizing the benefits of the collective cohesion of multiple PaaS services working together in one cloud solution provider. These benefits include ease of use, integrated monitoring, zero-trust security, serverless computing, abstracting away server management, and more. When we maximize generalizability, we increase our complexity and lose out on the benefits of PaaS services. The DoD wants the ability to host data anywhere and allow applications to seamlessly move between multi-cloud, hybrid, and disconnected edge environments while minimizing refactoring. Cloud providers should compete their IaaS services as a priced commodity, with security functions as their differentiator, while providing differentiating performance capabilities with their collective PaaS services to compete inside the data/AI platform layer. In the spirit of "customer obsession," cloud providers should recognize the DoD's desire to maximize their options inside the data/AI platform layer to allow for the best technological innovation that will protect the warfighter, our Country, and our mission partners. The purpose of this article is to highlight some differentiating capabilities needed to achieve the JADC2 vision in light of the complex decision space of DoD platform development.

The Multi-Horizon Framework

The multi-horizon framework is an excellent mental model we can use to articulate the differentiating cloud capabilities that will help realize the JADC2 vision. Horizon One outcomes will optimize our daily lives by eliminating inefficient, manual, and error-prone practices with scalable analytical pipelines. Horizon Two outcomes will address the DoD's technical debt by "pulling forward" the use of modern data platforms to catch up with common industry practices and adopt the governing policies needed to become AI-Ready. Horizon Three outcomes will create revolutionary change with new experiences that will digitally transform JADC2 functions while enhancing interoperability and information sharing with our mission partners.

Horizon One - Optimize Analytical Pipelines

During Project Convergence (PC), the Army's premier JADC2 demonstration, government integrators developed several disparate products to manage the complex PC system of systems architecture. These disparate products are in the form of Excel files, PowerPoint, Visio, and pdf files created by several organizations and are saved onto local computers and emailed or posted to share sites that the collective whole may not have access to. Additionally, these products lack consistent naming conventions and key identifiers to properly aggregate the essential information needed to manage complex system configurations. These disparate products make it nearly impossible to propagate architectural changes resulting in outdated, disconnected, and potentially contradictory information. The impact of these manual, non-scalable information flows prevent our ability to observe our networks and make sense of sensor-to-shooter threads. The effective implementation of AI will be unattainable if we continue to let humans manually perform tasks that machines excel at. To address these challenges, the DoD needs tools that lower the skill barrier of entry for pipeline development. Low-code platforms that facilitate citizen developers to create scalable analytical pipelines with embedded AI that integrate with team collaboration apps will optimize the DoD's processes and achieve tractable Horizon One outcomes.

Horizon Two - Eliminate Technical Debt

Within the Industry, data platforms have evolved from data warehouses with ETL pipelines to data lakes with ELT pipelines to a new paradigm known as the data mesh. However, the DoD continues to store data locally, share data via email, and snail mail SQL databases with portable hard drives. There is a tremendous technical debt within the DoD where the technical rework and cultural change necessary for digital transformation have a high implied cost. Digital transformation starts with upskilling our people, empowering them with edge and cloud technologies, developing advanced digital capabilities, and employing innovative digital processes (for more information see this link: ATEC Digital Transformation Strategy). The DoD is making excellent strides in developing digital engineering roadmaps, unified data reference architectures, data domain management processes, and modern software delivery processes. Differentiating data platform cloud capabilities are key enablers for delivering digital capabilities and employing digital processes. There are several Horizon Two outcomes needed to address the DoD’s technical debt including zero-trust security, modern data governance and management, and software factory processes. Two of these outcomes include the need for digital engineering environments and the extension of cloud capabilities to the edge. This article discusses the digital engineering ecosystems that will deliver JADC2 capabilities at the speed of relevance: The Data Mesh and Model-Based Systems Engineering Nexus. This article discusses DoD relevant edge workloads and workflows: Common Disconnected Tactical Edge Workloads and Modern Workflows for JADC2.

Horizon Three - Transform JADC2 Functions

The JADC2 strategic approach describes three functions that will shape the delivery of capabilities. The Sense function integrates information across all domains (land, sea, air, space, cyber, and the electromagnetic spectrum). Data from a host of sensor and weapon platforms, social media, open-source data, and host-nation data that range from National to tactical collection assets must be processed at the edge and distributed efficiently in a DDIL environment. The Make Sense function heightens our awareness of a situation. Understanding our environment, our adversary's intentions and our friendly situation in a rapid manner provides the contextual understanding needed to arrive at optimal decisions. The Act function recommends courses of action for multidomain convergence that achieve optimal effects while efficiently disseminating decisions to the Joint Force and our mission partners.

The technological innovations of 5G spectrum and space transport, foundation large multimodal models, and autonomy will create disruptive transformations of the JADC2 functions. JADC2 networks need self-healing features that can seamlessly transition communication channels to expedite data transfer with advanced compression algorithms while minimizing electromagnetic signatures. The 5G spectrum and space will transform our transport capabilities to enable self-healing network features at the edge, continuous deployment of software and machine learning model updates, distributed edge computation, and connections to hyper-scale computing. Generative AI models have the potential to combine electro-optic, infra-red, cyber data with historical textual mission reporting and other forms of data to produce outputs that act as an AI-assisted Commander’s copilot for recommending tactical decision options. In the future fight, hypersonic, directed energy, and cyber effect weapon systems will move faster than human decision-making. The DoD will need autonomous systems to react at machine speed to defend against these future threats.

Conclusion

Empathy towards the DoD’s desire for the generalization of app and data/AI platform hosting is important. Rather than attempting to dominate the entire vertical stack or one horizontal layer in the stack, vendors should compete their differentiating capabilities that are focused on solving clearly defined problems. In this article, we discussed the differentiating cloud capabilities needed for JADC2 using the multi-horizon mental model. The pursuit of the three horizon outcomes is not linear. The opportunity space for each horizon depends on the organizational culture, commitment to modernization, workforce data literacy, and resource availability. Simultaneously accelerating innovations across all three horizons will create design patterns that will scale to other organizations. Scaling efficient analytical pipelines will optimize our daily lives and reducing the DoD’s technical debt will establish the foundation we need to become an AI-Ready Joint Force. This foundation will open the aperture toward the disruptive transformation of the JADC2 functions. Advancements in emergent technologies are democratized and shared across the globe making it difficult to gain a competitive advantage against our adversaries. Therefore, the DoD must consider the technology opportunities across all cloud solution providers and vendors to assemble the best-in-breed innovations that will achieve our aspirational JADC2 vision.