Space is now a warfighting domain. But right now, space-based technology capabilities, such as compute and data storage, are constrained. What exists is highly function-specific and limited in application. These purpose-built capabilities can’t easily be repurposed or expanded, and they don’t support an efficient software-factory approach.
Most critically, data generated at the extraterrestrial edge has to be transmitted over low-bandwidth, high-latency connections to ground-based systems, and then sent to datacenters for analysis. Often the output of that analysis must then be communicated back into space. The process is slow and cumbersome, and it doesn’t support real-time situational awareness and decision-making.
That’s about to change. A modern space race – involving commercial, government, and military organizations around the world – is underway.
SpaceX has launched nearly 5,000 satellites into orbit, with approval for another 12,000. China plans 70 commercial and military space launches this year. NASA has selected Axiom Space to deploy the first commercial module to attach to the International Space Station (ISS). And within the next few years, several companies plan to build out commercially owned and operated space-based datacenters and micro-datacenters.
For the DoD, such space-based datacenters present opportunities for technological advances and stronger offensive and defensive systems. They also present serious risk of falling behind adversaries. It’s essential that military leaders take action now to specify the standards and architectures that will enable space-based datacenters to support military use cases and protect national security.
Constellations of Open Datacenters
With the right technology, space-based datacenters can store and analyze data at the orbital edge. This will allow data-driven insights to be generated closer to where data is captured.
Extending cloud-to-edge computing into space will require a new generation of space-hardened technologies that replicate server solutions on Earth. These technologies will need to offer high performance and scalability while meeting the unique size, weight and power (SWaP) and thermal demands of outer space.
The good news is that the foundational technologies for this orbital infrastructure are available today. For instance, modern microservices-based computing environments are well-suited to the extreme edge. Microservices architectures deploy each software application as a collection of fine-grained, loosely coupled digital services. They can be optimized to the extreme edge, because they can run on small and embedded devices while also being continually updated in a modular way.
What’s crucial is that these architectures run on open systems and open-source software. Why? Because, given the five- to 10-year life expectancy of assets, you don’t want to be locked into a proprietary architecture in space, where making changes is extremely difficult, costly, and time-consuming.
An effective strategy would be to combine Modular Open System Approach (MOSA) and Sensor Open Systems Architecture (SOSA) methodologies. MOSA and SOSA are design and acquisition strategies, used in industries such as technology and defense, which emphasize open standards and modular components. The goal is to create systems that are adaptable, interoperable, and can easily incorporate new technologies. These are established standards supported by both military and industrial organizations.
The DoD and federal government have extensive data on the challenges of migrating on-prem heterogeneous data centers to modern platforms. Space represents a greenfield opportunity to build open, modular architectures that provide the flexibility to support ever-expanding use cases and mission-partner needs.
Countdown to Orbital Edge Ignition
The first micro-data centers should reach operational capability by late 2025 or early 2026. Full-scale datacenters should be in orbit within five to seven years. These timelines aren’t the distant future, especially given the realities of military budget processes, manufacturing lead times, and payload scheduling.
Space Force and other DoD agency leaders should take steps now to ensure the military can take full advantage of these emerging capabilities. Specifically, they should develop courses of action (COAs) that define requirements, prioritize initiatives, establish governance, and include relevant consortiums. They should also accelerate collaboration with industry partners through roundtables and technology-discovery events to co-innovate best practices, policies, standards, processes, and technological advancements.
Most importantly, they should insist on open systems and open-source software that’s proven, secure, and reusable. They should leverage the innovation of the open-source community, with its steering committees and working groups that collaborate toward common goals. Cooperative research and development agreements (CRADAs) used by federal laboratories offer a useful model to co-innovate toward desired outcomes.
Space-based datacenters will provide commanders and operators with substantially faster access to data capture, storage, processing, and analysis. The result will be faster, more accurate situational awareness and increased ability to take the right actions at the right time.
Now is the time for the military leaders to specify the DoD’s requirements and collaborate with industry to define space-based data centers and the extraterrestrial edge. They should seize the opportunity to leverage open systems architectures and open-source technology to protect national security and dominate the space domain.
The author, Travis Steele, is Chief Architect at Red Hat.