From CJADC2 to Decision Advantage.

Combined Joint All-Domain Command and Control, or CJADC2, is often summed up in three words: sense, make sense, and act. That idea has not gone away. What has changed is the language around it. Across the Department of War and the services, the center of gravity is shifting from CJADC2 as a headline program to decision advantage as the operational objective.

Five years ago, the message was often described as connecting every sensor to every shooter. Today, the message is more operational: create decision advantage by connecting data, AI, mission partners, commanders, and effects. The emphasis has moved from building networks to enabling outcomes: faster understanding, better choices, and more resilient execution across contested, coalition, and data-driven mission environments.

The public record still shows real progress. Exercises such as Project Convergence and service-led efforts such as the Advanced Battle Management System (ABMS) are proving that data can move faster, decision cycles can shrink, and partners can be brought into more complex operational experiments. But those demonstrations also reveal a harder truth: the missing pieces are no longer just radios, cloud platforms, or more compute. The challenge is building mission infrastructure: a single architecture that unifies compute, networking, security, AI, data movement, and mission applications around operational outcomes.

In this latest blog, we discuss why CJADC2 must move beyond a collection of tactical networking, edge compute, security, AI, and application projects. The strategic question is whether these capabilities can be engineered together into mission infrastructure, an operating architecture that keeps data usable, decisions moving, and partners connected under real mission pressure.

CJADC2 is the framework. Decision advantage is the objective.

The U.S. Department of War describes CJADC2 as a way to help commanders sense, make sense, and act across every domain, throughout the electromagnetic spectrum, and with mission partners. It depends on automation, artificial intelligence, secure infrastructure, resilient communications, and better information sharing.

That remains the right operational frame. But CJADC2 is increasingly becoming the enabling framework rather than the headline capability. Its implementation now spans multiple paths: Mission Partner Environment, data-centric architecture, AI-enabled decision support, joint fires, coalition integration, edge compute, zero trust, and cloud-to-edge operations.

For mission infrastructure, four things matter most: data needs to be usable, communications need to hold up under pressure, systems need to keep working when disconnected, and information must be shared securely with approved partners.

In plain terms, CJADC2 is not just about connecting more systems. It is about creating an architecture where compute, networking, security, AI, data movement, and mission applications work together so the right data reaches the right people and machines at the right time, in a form they can use, under conditions that rarely cooperate.

That is where implementation becomes difficult. A 2025 GAO review found that the Department of War still lacks a comprehensive framework to guide CJADC2 investments and measure progress. In practice, that means progress can happen in pockets, but lessons are harder to scale across services, commands, and partners.

That gap matters because CJADC2 is not a single system that can be bought, fielded, and declared complete. It is an operating model. It links people, platforms, data, networks, security rules, mission partners, applications, and decisions across the force. Mission infrastructure is the practical architecture that makes the operating model real.

Service implementation is moving beyond the CJADC2 label

Recent exercises show that CJADC2 is moving from theory into practice, but the service-level language is becoming more specific. Project Convergence has brought U.S. services and multinational partners together to test faster data movement, tactical battle management, and more connected decision-making. ABMS has explored cloud computing, AI-enabled command and control, data sharing, and ways to connect systems that were not designed to work together.

The Army shows where implementation is heading. CJADC2 is becoming the enabling framework, while NGC2 is emerging as the visible modernization effort that turns those principles into mobile, software-defined, data-driven command and control.

What the demonstrations show

Taken together, Project Convergence, ABMS, NGC2, and related efforts show that the force can move more data, fuse more information, and bring more partners into a shared operational picture. They also show that tactical command and control is becoming more mobile, more data-driven, more software-defined, and more dependent on edge processing.

But technology is only part of the answer. Some of the biggest gains come from simplifying procedures, reducing overclassification, and improving how information is shared with partners. That is a critical lesson. CJADC2 does not slow down only when networks fail. It can also slow down when policy, process, and data rules are not built for mission speed.

Why the tactical edge is the real test for mission infrastructure

The edge is where conditions break first

In headquarters or well-connected cloud environments, data sharing is difficult but manageable. At the tactical edge, everything becomes harder. Bandwidth is limited. Links are intermittent. Power, size, and weight matter. Units may need to operate disconnected from the wider enterprise. Coalition partners may need access to some information, but not all of it.

This is why mission infrastructure matters. If every decision depends on sending data back to a central cloud, waiting for analysis, and pushing the answer forward again, the model breaks under contested conditions. The edge needs integrated compute, storage, security, AI, networking, and mission applications to keep decisions moving locally when the wider network is degraded.

What is still missing from mission infrastructure at the edge?

• A common mission infrastructure architecture. The strategy describes the desired outcome, but there is still no widely visible model for how compute, networking, security, AI, data products, and mission applications should be organized across echelons, services, and partners.
• Graceful degradation across the whole stack. CJADC2 assumes that communications will be contested or degraded. Mission infrastructure needs systems that can prioritize data, keep essential applications alive, preserve security policy, and switch between available links without manual workarounds.
• Operational edge AI. The missing piece is not simply putting AI models closer to the mission. It is managing the full lifecycle: data tagging, model distribution, provenance, fallback modes, and policy-compliant sharing with partners.
• Engineering for SWaP. Size, weight, and power are not side issues at the tactical edge. They shape what can be carried, powered, cooled, protected, and sustained in real operations.
• Mission-ready integration. The tactical edge needs more than tactical networking. It needs a unified architecture where compute, networking, security, AI, data movement, and mission applications are designed to operate together from the start.

Networking is one layer of mission infrastructure

Resilient communications are necessary, but not sufficient

Most CJADC2 discussions recognize the need for resilient communications: satellite links, 5G, software-defined radios, mesh networks, and hardened interconnects. These are important ingredients, but they are not the whole architecture.

The harder problem is making applications, data flows, security policies, and AI-enabled services work across whichever connection is available. The network should not be treated as a standalone product layer. It should be part of mission infrastructure that adapts around the mission, not the other way around.

Bandwidth still shapes the operating model

Bandwidth at the edge cannot be treated as unlimited. The force will still need to decide what data moves, what stays local, what gets compressed, what gets prioritized, and what can wait. That makes local processing, policy-aware routing, and application resilience essential parts of the mission infrastructure model.

If bandwidth were abundant and reliable everywhere, the architecture could remain more centralized. But contested environments demand a hybrid model. Some decisions and data products must survive locally when the wider enterprise is unreachable.

Classification and releasability remain major blockers

One of the most important CJADC2 lessons is that classification policy is now part of the network architecture. If data cannot be released quickly to the people who need it, the system slows down, even if the technical connection exists.

This is especially important in coalition operations. Approved partners need data that is routable, releasable, understandable, and usable from the start of the mission. Releasability cannot be bolted on after the fact. It has to be designed into the data, the workflow, and the security model.

Coalition interoperability is a trust problem, not just a technical one

NATO’s Federated Mission Networking approach is a useful lens. It recognizes that coalition networking is about people, processes, standards, validation, training, and technology working together. Nations keep their own systems and authorities, but they need common arrangements that allow them to operate together quickly.

That is why CJADC2 should not be judged only by whether one system can technically connect to another. The real test is whether approved partners can use the right data at mission speed without manual stitching, policy delay, or bespoke integration every time.

What industry is solving, and what still falls short

These are meaningful advances. But many of them solve parts of the problem rather than the whole operating model. CJADC2 needs more than strong components. It needs mission infrastructure built in from the start: shared data standards, common implementation patterns, policy-aware networking, coalition-ready releasability, edge AI lifecycle management, and mission applications that are designed to work as one architecture.

• Edge platforms and mobile battle management are helping bring compute and decision support closer to the mission.
• Data fabrics and edge cloud approaches are improving how information is organized and moved between systems.
• Multi-bearer networking is improving resilience across satellite, terrestrial, mesh, 5G, and line-of-sight links.
• Zero trust and controlled access are improving how users with different permissions access data securely.
• AI-enabled decision support is beginning to shift the conversation from moving data to turning data into operational choices.

So, what needs to change?

To make CJADC2 useful at the tactical edge, the focus needs to shift from proving individual capabilities to building repeatable mission-infrastructure patterns that can withstand real-world conditions.

• Publish and enforce a common mission infrastructure architecture that defines compute tiers, network behavior, data contracts, degraded-mode operations, model distribution, application hosting, and coalition release paths.
• Make releasability an engineering requirement by treating markings, metadata, release rules, and partner access as part of the data product itself.
• Build for contested mission environments by enabling applications, data flows, security policies, and AI services to operate across available bearers without service-specific rework.
• Turn exercises into reusable implementation patterns so lessons from Project Convergence, ABMS, NGC2, and related efforts can scale across the enterprise.
• Move edge AI from demo to lifecycle by focusing on deployment, provenance, monitoring, fallback modes, and trusted use under degraded conditions.
• Move decisively from tactical networking to mission infrastructure by designing compute, networking, security, AI, data movement, and mission applications as one architecture rather than a collection of products.

The bottom line

CJADC2’s hardest problem is no longer proving that data can move once across a demonstration network, or that one sensor can connect to one shooter in a controlled scenario. It is making compute, networking, data, AI, policy, security, and mission applications work together at the tactical and coalition edge, reliably, securely, and under degraded conditions from day zero.