By: Tuva Siegel, Warrior Editorial Fellow

The U.S. Space Force is advancing human-machine teaming by testing how joint AI caabilities can increase operational speed, improve multidomain integratio  and enhance collaboration between Space Force Guardians and Airmen. At Shadow Operations Center-Nellis in Las Vegas, the Joint Force spent two weeks testing new approaches to human-machine teaming through the Multi-Decision Advantage Sprint for Human-Machine Teaming (MASH) experiment.

The urgency of learning how to make decisions faster than the adversary is driving this mission, placing U.S. Space Force Guardians "in the seat" alongside Air Force battle managers for the very first time, explains U.S. Space Force Col. Teina Stallings-Lilly, in a DVIDS release. By learning how these disparate tools can effectively integrate to solve complex problems across the air, space, cyber, maritime, and ground domains, the MASH experiment integrates an ensemble of artificial intelligence and automation software services from the first three Decision Advantage Sprints for Human-Machine Teaming (DASH) events.

The importance of MASH lies in its ability to accelerate better, faster decision-making. U.S. Air Force Col. John Ohlund describes the process of incorporating AI as “ensuring our operators can rapidly process vast amounts of data and deliver lethal effects faster than ever before." The modern battlefield demands a combination of domain focus and strength, with AI working to bridge these gaps and smooth out the human burden. This also means that roles are being reframed and evolving to become less rigid; for example, Space Force Guardians are moving beyond “observational roles”, claims DVIDS. U.S. Space Force 1st Lt. Abby Warner, explaining that that working alongside the Air Force opened her eyes to “how the air domain tackles these challenges. Their focus on tempo, synchronization, and rapid Courses of Action iteration mirrors what Space Force needs, especially when dealing with contested electromagnetic environments." 

The problems around decision-making were found to be similar across all domains, further emphasizing the need for a common solution to ease the collective burden. A shared AI process across domains could reduce differences in how information is handled, potentially allowing for faster processing and decision-making. Ohlund explains that Air Force air battle managers don't have the authority to execute a space or cyber effect; instead, it's their job to prepare the information and package the options for the general. “We want the computers to do that work, to ruminate over every possible multi-domain effect; that way we can present the highest quality menu of decisions to the right commander, faster than ever before."

This exercise was not solely U.S.- specific; in fact 4 allied nations observed MASH, learning how to integrate architectures and set the foundation for future interoperability. Not only are the software solutions helpful across domains, but they are "directly translatable" to Navy, Marine Corps, and Army partners. This collaboration is “the next pivotal step in providing 'combat multi-domain power' for the 'Total Joint Force,’” emphasized U.S. Air Force Lt. Col. Corey Ellsworth. 

During the experiment, MASH challenged six industry software development teams and the Shadow Operations Center – Nellis (ShOC-N's) own military software development team and had to create tools that support the three primary decision-making functions outlined in the Air Force’s Transformational Model. First, the aquisition system helps military planners decide what actions to take against a target. Then, it recommends possible actions, identifies which capabilities are best suited to achieve the desired effects, and ranks those options. From there, it creates possible battle plans by pairing effects with the most effective capabilities and adding any additional resources needed throughout the operation to support each option. Ohlund clarifies that "We are proving that a true plug-and-play, modular approach not only works, but it fosters continuous competition and allows the government to select the best-of-breed software services as they mature."

The focus on guardian and airmen-driven innovation appears to lie at the forefront of the Pentagon’s practice when dealing with new technology, especially AI. Officials highlight the unique operator and developer relationship, or as DVIDS outlines, both Airmen and Guardians stress-tested the AI's “decision logic, identifying limitations and providing immediate feedback to the developers sitting directly behind them.” Elizabeth Frost, AFRL MASH lead, said that “The teams are eager for feedback and implemented changes rapidly. This collaborative effort paid off during the second week of the sprint, as we saw a remarkable increase in the volume and quality of courses of action submitted."

The sprint produced measurable improvements in operational speed; before, a team would need approximately fifty minutes to an hour to get a single tasking done, now five or six tasking scans can be completed within that time frame. Additionally, Carlos Dye, the ShOC-N MASH software development team lead, explained that military developers used their firsthand operational experience to guide the software development process. This approach allowed the system to handle much of the complex data processing while keeping the human operator responsible for making the final tactical decisions. The MASH experiment demonstrated a path forward for multi-domain operations, validating the DAF’s Transformational Model by showing that “machines can process data at a speed unmatched by humans” when supported by a shared decision-making framework.

The Space Force has previously outlined its strategic framework when integrating AI to secure the high ground. The Space Force AI framework divides its approach into three tiers: Enterprise AI, Functional AI, and Mission-Specific AI, which consists of partnering with “the commercial sector to leverage a staggering $300 billion in annual private AI investment for national defense,” states an article from the Space Force. Enterprise AI consists of general-purpose tools such as Gen. AI, available across the Department of War. Functional AI is more tailor-made and specific to domain-specific data, including policy and regulatory documents for specialized workflows. Finally, mission-specific data is custom-built and includes tools for helping operators track missile threats and analyze space domain awareness. Bartley Stewart, Space Systems Command’s Data and AI officer, explains that decision accuracy is prioritized over the adoption of AI technologies, producing “asymmetric decision advantage.” AI remains confined to “three key factors in decision-making: accelerating response times, reducing cognitive burden and increasing confidence in data used to inform decisions,” said Stewart.

These frameworks outlined by the Space Force, paired with the MASH experiment, align directly with the Department of War’s “AI-first” warfighting force. The execution of “Pace-Setting Projects” (PSP) helps to demonstrate the accelerated pace of “execution, focus, and ethos we need to stay ahead,” states the Chief Digital & Artificial Intelligence Office. MASH appears to be one of these projects, ultimately mirroring Ohlund's "faster than any adversary" goal.

Tuva Siegel is an Editorial Fellow at Warrior Maven. She studies English at Kenyon College. Tuva is the author of Drömland, a fictional collection of short stories, and is currently studying weapons and military technology.