Why the Army's $1.3 Billion Optionally Manned Fighting Vehicle Can't Decide If It Needs a Driver

Picture a vehicle so advanced it can operate with a crew or without one. That's the promise behind the Army's Optionally Manned Fighting Vehicle, the program meant to finally replace the aging M2 Bradley Infantry Fighting Vehicle. After decades of failed replacement attempts, the Army restarted the OMFV effort in 2019 with fresh optimism and a clean slate.

That clean slate now has a lot of eraser marks on it.

The OMFV is supposed to be the Army's next-generation infantry carrier: heavier protection than the Bradley, a bigger gun, and the ability to operate autonomously when the mission calls for it. Five companies competed in the early design phase. By 2022, the Army had awarded contracts to two finalists, General Dynamics Land Systems and American Rheinmetall Vehicles, for a competitive prototype phase. The goal was to field an initial fleet sometime in the early 2030s.

That timeline has already slipped. The competitive prototype contracts carry a combined value of roughly $1.3 billion just for the development phase. Total program cost projections have climbed past $45 billion if you count full production and sustainment over the vehicle's life. For context, that's enough to buy a small country a reasonably decent air force.

Here's the actual problem, and it's almost philosophical: nobody inside the Army can fully agree on what "optionally manned" should mean in practice.

Does the vehicle need to be capable of fully autonomous combat operations, making decisions about movement and fire independently? Or is it enough for the vehicle to drive itself in a convoy while a remote operator handles the weapons? Should the autonomous systems work without any communications link to a human, or does every autonomous action require a man-in-the-loop approval? These sound like engineering questions. They are, but they're also deeply doctrinal ones, and the Army's doctrine writers and the vehicle program managers have not always been reading from the same document.

The result is a requirements document that keeps shifting weight. Early specifications emphasized autonomous mobility. Then concerns about liability for autonomous lethal decisions pushed the requirements toward human-supervised modes. Then wargaming exercises suggested the Army actually needs vehicles that can push forward without putting soldiers at immediate risk. The requirements shuffled again.

Every time the requirements shift, the prototype designs have to absorb the change. Sensor suites get swapped. Software stacks get revised. Weight budgets blow up because adding autonomy hardware means adding processors, cameras, lidar units, and redundant communications gear, and all of that adds tonnage that threatens to exceed bridge-crossing weight limits the Army actually cares about.

There's also a software problem sitting underneath the hardware problem. The Army wants the OMFV to plug into its broader network of autonomous and semi-autonomous systems, the whole Project Linchpin artificial intelligence ecosystem it's been building. Linchpin is still maturing. Asking two competing vehicle contractors to design autonomy software that integrates with a network that isn't finished yet is a bit like building a house around a floor plan your architect is still sketching.

For a cleaner picture of how the moving pieces interact:

graph TD
    A[OMFV Requirements] --> B{Manned or Unmanned?}
    B --> C[Autonomous Mobility Mode]
    B --> D[Remote-Operated Mode]
    B --> E[Crewed Combat Mode]
    C --> F[Linchpin AI Integration]
    D --> F
    E --> G[Traditional C2 Systems]
    F --> H((Field Deployment))
    G --> H

Notice how both autonomous paths funnel through Linchpin before reaching the field. If Linchpin isn't ready, those top branches stall. That dependency is the quiet risk nobody in the press releases wants to highlight.

To be fair to the Army, the Bradley replacement program has a long and embarrassing history of failure before OMFV even started. The XM2001 Crusader. The Future Combat Systems manned ground vehicle. The Ground Combat Vehicle. Each effort burned through billions before cancellation. The Army restarted OMFV with a deliberately slower, more competitive approach specifically because it had learned those lessons.

Whether the lessons stuck is the open question.

What's frustrating is that the underlying technology is genuinely closer to ready than it's ever been. Commercial autonomous vehicle development has handed the defense industry a massive toolkit it didn't have in 2010. The Army isn't starting from zero on sensors or compute. The gap between "technically feasible" and "doctrinally agreed upon" is where the money keeps disappearing.

Somewhere, a Bradley is still rolling down a road, holding the line until its replacement figures out whether it wants to drive itself.