Supply chains kill armies. That's not rhetoric — it's operational history. Napoleon's Russian campaign. MacArthur's advance to the Yalu. Rommel in North Africa. In each case, tactical and even strategic failure traced back to an inability to move fuel, ammunition, and food to where the fighting was happening. Combat power is theoretical without logistics. And logistics convoys are among the most dangerous assignments in any ground campaign.
In Iraq and Afghanistan, improvised explosive devices targeted supply convoys relentlessly. Between 2001 and 2012, logistics vehicles accounted for a disproportionate share of U.S. casualties. The military's response — hardened vehicles, route clearance procedures, aerial resupply where feasible — addressed symptoms but not the underlying problem: human drivers in predictable patterns on predictable routes.
Autonomous ground logistics is the structural solution. And we're now far enough into the development cycle to see which companies are building durable technology versus which ones are selling demonstration vehicles.
Why Now
The enabling technologies crossed a threshold in the last four years. Lidar costs have dropped by roughly 85% since 2019. Edge compute capable of running real-time sensor fusion in a vehicle environment is now affordable at scale. And the autonomy software stacks developed by the commercial autonomous vehicle industry — however tortured that industry's path has been — have produced algorithms for unstructured environments that didn't exist a decade ago.
The Army's Robotic Combat Vehicle program and the Marine Corps' Autonomous Logistics program both have active solicitations. The question for investors isn't whether this market will develop — it will — but which technical approaches will survive contact with the operational environment, not just the test range.
The Hard Problems That Separate Serious Companies
Three challenges consistently differentiate companies doing real work from those doing demos.
GPS-degraded navigation. The commercial AV industry operates in environments with excellent GPS, good map data, and roads designed for vehicles. Military logistics happen in none of those conditions. Contested environments feature GPS jamming and spoofing as a matter of course. Companies that can navigate reliably using inertial measurement, terrain mapping, and visual odometry — without GPS — are solving a fundamentally different problem than commercial AV firms.
Convoy coordination under comm degradation. A logistics convoy isn't a single vehicle — it's 6 to 20 vehicles that need to maintain formation, respond to route changes, and survive the loss of communication nodes without the whole column stopping. Distributed autonomy at this scale, with intermittent communications, requires architectures that commercial applications don't need.
Interoperability with legacy fleets. The Army is not replacing its existing vehicle inventory. Autonomous logistics systems need to operate alongside human-driven vehicles, integrate with existing C2 software, and function within tactical operations centers that were not designed with autonomous systems in mind. This is an integration problem as much as a technology problem.
Investment Signals We Track
When we evaluate autonomous logistics companies, we pay attention to specific indicators beyond the standard deep-tech diligence checklist.
| Signal | What It Tells Us |
|---|---|
| IVAS / JLTV integration testing | Company is working within actual Army vehicle programs, not in isolation |
| GPS-denied test results | Technology has been validated in militarily relevant conditions |
| OTA contract awards | DoD has determined the technology meets a threshold of operational relevance |
| Cleared team members in GNC roles | Company can access classified operational data needed for real-world testing |
The companies that will win this market aren't the ones with the best demo videos. They're the ones that have spent time with actual operators in actual operational environments and built their systems around what those operators actually need.
Market Scale
The Army alone operates over 250,000 tactical wheeled vehicles. A 10% conversion to autonomous-capable platforms, over a procurement cycle of 15 years, represents tens of billions in addressable market. That math doesn't include the Marine Corps, Special Operations Command, or allied nation demand for interoperable autonomous systems.
We've seen this movie before with unmanned aerial systems. The RQ-11 Raven entered service in 2003. Twenty years later, the Army has over 5,000 small UAS in its inventory. The adoption curve for autonomous ground systems will likely be slower — the operational environment is more complex, and the integration requirements more demanding — but the end state is directionally similar.
The companies building the foundational autonomy stacks for ground logistics today are positioning for a market that, a decade from now, will be measured in tens of programs of record and hundreds of thousands of vehicles. We'd rather be invested in the stack than in any single vehicle program.
