Randy Fields, Vice President, Edge Compute and Mission Technology
The Air Force is putting open architecture at the center of how it plans to build, upgrade and sustain mission systems. In July 2026, the service released two foundational government reference architectures, the Agile Mission Suite Government Reference Architecture (AMS GRA) and the Autonomy Government Reference Architecture (A GRA), reinforcing a broader move toward modular, government-owned standards that make it easier to integrate capabilities across platforms. The goal is faster capability delivery, greater competition and more flexibility as mission needs change.
That shift has a direct impact on edge processing. As more sensors, applications, and autonomous systems connect across the battlespace, processing has to happen as close to where data is created and decisions are made.
The edge has to handle more complexity
As the Air Force opens mission systems to more applications, sensors and vendors, the edge environment has to keep up with what gets added. New capabilities bring different demands for processing, storage, networking and security, often on platforms where space, weight and power are already limited.
That changes the role of edge hardware. It has to support a growing mix of workloads without forcing programs to redesign the underlying system every time requirements change. Compute resources need to be configurable, networking has to move large volumes of data quickly, and the architecture needs enough flexibility to accommodate capabilities that don’t exist yet.
For open architecture to work properly, that adaptability has to extend all the way down to the hardware.
Software portability depends on capable hardware
The Air Force’s open-architecture push is built around the ability to integrate software and algorithms across different platforms and vendors. The Collaborative Combat Aircraft program is already demonstrating this approach through the Autonomy Government Reference Architecture, which is being integrated across multiple aircraft and autonomy providers.
Portable software still needs a consistent place to run. Edge hardware has to provide enough processing power, networking capacity and orchestration support for applications that may include autonomous functions, sensor fusion, command software and AI and machine learning.
That becomes increasingly important as more decisions move toward the tactical edge and autonomous decisions. Sending every data point back to a distant data center adds latency and depends on communications paths that may be denied, degraded or disrupted. The focus on ensuring an operator can operate independently through processing data locally gives operators and autonomous systems the immediate access to the information they need.
High-speed data movement becomes part of the mission
Modern manned and unmanned platforms and systems generate huge amounts of data. Open architectures can make it easier to add sensors and software, but every new capability also adds demand on the system moving information between them. As a Field CTO, I see that integration challenge firsthand. Bringing together tech from vendors, universities and other partners can create real friction, particularly when large volumes of data need to move quickly enough to give warfighters an operational advantage.
That puts greater demand on the edge. Processing platforms need high-speed networking that can move information efficiently between sensors, processors, storage and mission applications while supporting the standards that allow tech from multiple different sources to work together.
Sensor fusion and autonomy make that essential. These applications rely on multiple data sources coming together fast enough to build an accurate operational picture and support decisions at mission speed. The processing power matters, but so does the architecture responsible for moving that information where it needs to go.
Knox provides a modular compute foundation
Ultra I&C’s Knox family was built around many of the same principles driving the Air Force’s open-architecture strategy. Knox uses a SOSA-aligned, 3U OpenVPX architecture with configurable slots for compute, storage, switching, multilevel security and cross-domain solutions.
The platform supports VITA 49.2, OpenVPX and other modern frameworks while providing native 100 GbE networking. It can host cloud-native workloads, AI and ML models, sensor fusion, digital twins and mission applications with built-in orchestration and DevSecOps integration.
That composable approach gives integrators a hardware foundation that can change with the mission. Knox can scale from size-constrained unmanned aircraft to larger airborne and expeditionary systems, allowing processing capabilities to be configured around platform needs while maintaining a common architectural approach.
Building for the next upgrade
The Air Force’s move toward open architecture is changing how capability gets onto the aircraft. Mission systems are becoming more modular, software is becoming more portable and upgrades are expected to happen on timelines closer to modern software development.
Edge processing has to support that pace. The value comes from giving programs an adaptable compute foundation that can accommodate new workloads, faster data flows and evolving mission requirements without creating another closed system inside an open architecture.
Knox brings that flexibility to the edge with a commercial modular hardware approach, providing high-speed networking and the processing environment needed to support modern mission software. As the Air Force continues opening its architectures, that kind of adaptable edge compute will be central to turning open standards into operational capability.