What Makes a Metaverse? The U.S. Army Meets the Requirements
Why does the US Army fit all the requirements to be considered a metaverse?
What is the metaverse, actually? Pete Morrison, CPO at BISim, unpacks the criteria that need to be fulfilled for the full-vision metaverse. Moreover, he explores how the US army fits the brief and suffices most of these stated criteria.
Most people likely developed at least a vague idea of the concept when Meta rebranded. Some may think of the film Ready Player One or video games like Roblox, Second Life and Meta’s own Horizon Worlds. Still, these popular iterations of the metaverse do not produce a comprehensive definition. These games, in particular, have different purposes than the metaverse. And the technology behind them is different.
A recent McKinsey study explored conceptions of the metaverse and its business potential. The report argues that a “full vision” of the metaverse must contain the following:
- Interoperability allowing different game engines and devices to connect
- Concurrency with thousands of people interacting simultaneously
- Use cases spanning human activity well beyond gaming
No current popular fits this description. Second Life and Roblox do not offer use cases outside of gaming. Horizon Worlds currently requires a Meta headset to play, and none of the above supports thousands of players in a high-fidelity environment.
Perhaps surprisingly, the organization closest to meeting McKinsey’s metaverse definition is the U.S. Army. Its Synthetic Training Environment (STE) converges all live, virtual and constructive training components into one environment. And these immersive scenarios and mission rehearsals support thousands of entities. In fact, I believe the project may be the first actualization of a metaverse.
See More: How AR and VR are Transforming the Automotive Industry
Interoperability
The McKinsey report states we are “likely a few years from overcoming interoperability hurdles.” Hundreds of virtual worlds exist, but generally speaking, there needs to be more interest in building connections between worlds that run on different game engines. Commercial worlds keep within their servers and hardware, which has resulted in multiple siloed metaverses. Case in point, the Meta metaverse, Microsoft metaverse and Roblox metaverse are all under separate development. Beyond desire, interoperability requires a complete overhaul of each simulation’s server architecture. A true metaverse depends on a new, open web architecture — and the military is building this architecture.
As far back as 2003, military organizations tasked me with connecting different training simulations. For example, an organization may want to integrate a top-down constructive simulation with a virtual systems trainer. In these simulations, each person in the exercise operates an icon to rehearse their movement, and these icons move around the digital map. This constructive simulation could connect to a virtual simulation, like a gun galley trainer or flight simulator, that looks like a computer game.
Each of these simulations uses different software and hardware and runs on a different server architecture. Initially, these systems could only talk through the lengthy process of transferring data between different architectures. While this approach worked, it was limited and expensive. Each simulation stems from the same web architecture and isn’t bound to specific projects. For example, a tank developed for one particular exercise no longer lives in that single simulation but on the web, where administrators can seamlessly pull it for any number of different simulation environments.
Because multi-domain operations (MDO) are the reality of modern warfare, branches have less need for siloed training. In fact, mission rehearsal regularly involves pieces moving on the ground, sea, air, cyber and even space. Any successful simulation must practice orchestrating all these moving parts. If a platoon of ground troops requires air support, a soldier’s combat simulator and pilot’s flight simulator need to have the same “out of window” view, as well as the same method of communication. The pilot and ground troops also must connect to a JTAC (Joint Terminal Air Control) simulator that shares the same environmental experience.
Beyond just one country’s interoperability between branches, different allied countries are looking to orchestrate their operations too. NATO agrees, asserting that alliance member training software should combine into a “common architecture.” As Robert Siegfried, chair of the NATO modeling and simulation group, stated, “What we want to do is connect the national synthetic training systems so that we can train together daily.” This level of daily compatibility requires all platforms and devices to connect to the same source, regardless of differences in hardware and simulation tech.
In short, military organizations need metaverse-level interoperability to train for modern warfare’s realities. Commercial metaverses are built, above all, to make money, and this reliance on consumer interest forces them to throw stuff at the wall and see what sticks. The military approach is different. Their focus on achieving true interoperability is why the military will complete the first true metaverse.
Concurrency
The McKinsey report envisions the metaverse containing thousands of simultaneous entities. It asserts that concurrency limits either “cap the number of players” or “space players across a map” to avoid overloading their server processing. These steps diminish the environment’s capabilities, particularly its immersion. How do we improve concurrency limits? Once again, web architecture. The military has designed metaverse architecture to support thousands of live players and AI entities.
Mission necessity drives military investment in concurrency, as with interoperability. Operations could require hundreds of simultaneous live players and thousands — if not millions — of AI entities. But the current complexity of entities, behaviors and terrain is nowhere near where it needs to be.
Imagine developing a simulation exercise for the 1993 Battle of Mogadishu, the subject of the popular movie Black Hawk Down. The initial raid included 160 U.S. forces, 19 aircraft and 12 ground vehicles. Any simulation of the full-scale operation would require at least this many live entities working together. Meanwhile, Mogadishu, Somalia, is home to over 2 million people. Many of these residents were hostile toward the raid, so each would require complex, AI-based decision-making and combat abilities.
Beyond just hostilities, the behavior of the city itself must also be accurate. Traffic patterns, weather, conditions of buildings and streets and civilian behavioral patterns must represent potential conditions of the actual raid. If the simulation were to take place in 2022, it should represent potential communication challenges and cyber threats to reflect the scope of danger in the environment. The level of detail and complexity leaves little wonder why no server architecture supports the required processing power. However, that is not to say we are far away from this technology.
At the moment, the U.S. Army utilizes dozens of Battle Simulation Centers — warehouse-sized computer labs that house hundreds of desktop PCs and large-scale tank and aircraft simulators. Training exercises simulated in these centers can support around 250 live players and 3,000 AI units concurrently. While some commercial games support this number of live players, none do so at the same level of environmental fidelity, game complexity and performance. Still, even this concurrency scale does not match the McKinsey definition, nor does it allow for a realistic creation of every potential scenario.
The STE aims to achieve the level of scale needed for a metaverse through investment in cloud architecture. Instead of running the simulation through siloed, location-based local servers, the STE will use software like Mantle ETM to synchronize data between servers. Cloud-based architecture can exponentially scale simulation concurrency, expanding the number of simultaneous players to the thousands and the number of AI entities to the millions.
The U.S. Army thought of concurrency limits well before commercial metaverse hype. Early on, it invested in cloud scalability and synchronization, which is a massive paradigm shift requiring a lot of time for stakeholder buy-in and implementation. The Department of Defense’s attentiveness to concurrency is yet another reason they will get to metaverse-level scalability first.
See More: A Quick Guide to Smart Manufacturing
Extensive Use Cases
McKinsey’s final requirement for a metaverse is offering use cases “well beyond” the scope of gaming. While the report showcases some emerging commercial use cases, these still link to gaming. For example, Wendy’s made headlines for opening a restaurant in the metaverse earlier this year. However, the “restaurant” merely housed several sponsored virtual games, such as basketball or shuffleboard. Its underlying purpose did not extend into food service or retail.
Gaming is not an inherently wrong place to start in developing a metaverse. Computer games have always had enormous potential for cognitive learning and training transfer — but that’s only when skills practiced in a computer game actually transfer into real-world skills. The military has used games to increase soldier training transfer for decades. But, unlike commercial metaverses, getting users to spend time (and money) in the simulation is not their end goal. Everything practiced and rehearsed in the military metaverse is designed to increase real-world skills and understanding.
The U.S. Army is creating the STE for specific use cases: weapon and vehicle training and experimentation, mission rehearsal and planning, wargaming and scenario prediction. Each of these use cases requires very different software and hardware. Vehicle training, for example, involves custom VR/AR tech that accurately replicates the real-world equivalent and a detailed first-person, out-of-window perspective. Scenario planning needs sophisticated AI entities and an up-to-date rendering of terrain details and enemy positioning.
A single virtual world that supports each of these use cases will demand advanced terrain development and management levels, enormous capacity to scale, interoperable systems connection and the ability for users to access and interact with the world in numerous constructive ways. The Army’s goals are lofty, yet it’s made incredible progress. I predict that by 2030, soldiers will perform daily training within the STE. If this is the case, they’ll train in nothing short of the world’s first complete vision of the metaverse.
What are your thoughts on the widespread acceptance and use of the metaverse? Share with us on Facebook, Twitter, and LinkedIn.
MORE ON METAVERSE
