Autonomous Logistics Goes to War: How Delivery Drones Are Crossing Into Defence Resupply

€2.5 million. That's what it takes to build a drone that flies like a jet, carries 100 kilograms, and lands without a pilot, a runway, or a helicopter's price tag. Copenhagen-based Acodyne raised exactly that in June: a pre-seed round, oversubscribed, backed by a Swedish defence VC and Denmark's state export fund. It tells a story larger than one startup.

The cargo drone market sits at roughly $2.1 billion today. PitchBook projects $103.7 billion by 2035. What changed between 2024 and 2026 to make that trajectory plausible is the question worth asking.

The Money Behind the Drones

Start with the macro number. PitchBook's 2025 robotics report logged $27.6 billion across 1,009 deals, more than double 2024's $13.7 billion. Of that, defence and security robotics alone accounted for $8 billion in 234 deals, a 138.8% year-on-year increase with no precedent in the sector's history.

Unmanned aerial vehicles drew $6.2 billion of that. Autonomous land vehicles pulled $512 million. Unmanned maritime systems, $944 million. The common thread: every service branch across NATO is trying to solve the same problem: how to keep forward-deployed units supplied without exposing aircrews to missile and drone threats.

The Acodyne Bet

It is a 10-person company based at the Technical University of Denmark. Founded in late 2023 by Mads Schnack, Jasmina Pless, Claes Nicolaisen, and Martin Arndt, Acodyne builds unmanned eVTOL cargo aircraft using ducted-fan propulsion, the same core technology found in commercial jet engines, scaled down and electrified.

The first model, the E100, carries 100 kilograms at 450 kilometres per hour. Range is 500 kilometres on battery power, extending to 1,000 in a hybrid configuration. The aircraft transitions from vertical take-off to fixed-wing cruise mid-flight, controlled by an AI autonomy stack called eTHOR, developed in collaboration with DTU Compute. The entire system packs into a standard 20-foot shipping container.

"For so many years, the fastest and most efficient way of transporting heavy goods has been with a ducted-fan motor," Pless told TechFundingNews in June. "We're betting that's still the winning technology, instead of just scaling up propellers."

The bet comes with caveats. It has yet to complete its first flight test. Those are scheduled before the end of 2026, with field trials in Greenland following in spring 2027. Scaling from 100 kilograms to the planned 500-kilogram variants within two years requires solving engineering challenges that it describes as unproven at this stage.

But Acodyne is not alone in this corridor. Three other companies are pursuing the same thesis with different technical answers.

Four Approaches to the Same Problem

Enigma Aerospace emerged from stealth in March 2026 with $7 million in venture and Department of War funding. Its Phoenix aircraft carries 1,000 pounds over 2,000 nautical miles using precision airdrop or short take-off rather than vertical flight. In May, Enigma announced a manufacturing partnership with Re:Build Manufacturing targeting production of thousands of units per year.

Circus Defence, a German firm, took a ground-based approach in April: it secured a public procurement contract from the Lithuanian Armed Forces to deploy an autonomous AI-powered sustainment robot on NATO's eastern flank. The system is integrated into existing barracks infrastructure and will be tested in multinational exercises. Lithuania borders Belarus and sits near Russia's Kaliningrad exclave. The operational requirement is immediate, not speculative.

Forterra, a US developer of autonomous military vehicles, closed a $238 million Series C in late 2025. Its platform is purpose-built for logistics convoys in contested environments. The scale of the round, one of the largest in defence robotics, signals that institutional capital now treats autonomous military logistics as a category, not an experiment.

Forterra's approach uses wheeled and tracked ground vehicles rather than aircraft. The trade-off: lower speed, but higher payload, longer endurance, and no airspace integration complexity. Each of the four companies has bet on a different mobility path: ducted-fan eVTOL (Acodyne), fixed-wing airdrop (Enigma), wheeled ground robot (Circus), armoured ground convoy (Forterra). Their target customer is the same battalion commander trying to keep forward positions supplied without risking helicopter crews.

The Economics of Replacing the Helicopter

A single CH-47 Chinook flight hour costs roughly $12,000 to $15,000 in fuel, maintenance, and crew. A military helicopter mission in a contested zone also carries the cost of the aircrew's risk premium: insurance, training replacement, and the operational gap if the aircraft is lost. Autonomous cargo platforms eliminate the crew cost entirely and reduce the per-sortie expense by an estimated 40 to 60 percent, per defence logistics cost analyses from PitchBook and industry reports.

That math is driving procurement decisions. When the Lithuanian Armed Forces contracted Circus Defence for an autonomous sustainment robot, the operational calculus was about reducing the logistical footprint on NATO's eastern flank while maintaining 24-hour resupply capability without exposing drivers to drone surveillance or ambush. The same logic applies at scale across every service branch with forward-deployed units.

The commercial sector provides the volume that brings unit costs down. RIVR Robotics raised $22 million in seed funding in late 2025 for wheeled-legged delivery robots that navigate urban terrain at 15 km/h. Starship Technologies has completed millions of commercial deliveries with its sidewalk robots. None of these companies market to defence customers, but the sensor stacks, autonomy software, and supply chains they are building are identical to what a military ground-logistics robot requires. The crossover is structural, not incidental.

The Dual-Use Engine

What makes this moment different from earlier waves of defence robotics investment is the dual-use structure of the capital. Acodyne's round included Gungnir Capital, a dedicated defence VC, alongside PSV Hafnium and EIFO, Denmark's state export fund. The latter typically backs commercial deep tech, not weapons systems. Circus Defence's contract is military, but the underlying AI stack is the same perception and navigation software used in commercial autonomous warehouses. Enigma's Phoenix aircraft shares subsystems with the commercial cargo drone industry.

"A lot of the funded start-ups are developing platforms capable of operating in both contexts," a PitchBook analyst noted. "Drones for monitoring critical infrastructure, autonomous vehicles for logistics in conflict zones as well as in hazardous industrial environments, computer vision for target recognition as well as for quality inspection."

This dual-use aperture broadens the total addressable market and reduces perceived investor risk. A defence logistics platform that can also service offshore oil rigs, mining operations, or disaster relief zones does not live or die on the defence budget cycle.

Former Marine Corps Commandant General David Berger put the operational logic bluntly in August 2025: "If you can't communicate and you can't support yourself, you can have the best systems and the best trained force in the world — you're not going to last very long." Berger argued the US is "way behind" in using autonomy for logistics, a gap that the 2026 investment data suggests is now closing from the private side rather than waiting for procurement reform.

What Still Needs to Prove Itself

Every company in this space faces the same chasm: transitioning from a validated concept to a flight-tested or combat-tested platform. Acodyne has not flown yet. Enigma has a manufacturing partner but no delivered units. Circus's system will be evaluated in exercises, not combat. Forterra is the furthest along, but its $238 million round was a bet on future production, not current revenue.

The certification path for unmanned cargo aircraft in civilian airspace remains unclear. The European Union's U-space framework is progressing, but it was designed for small drones, not 500-kilogram cargo aircraft flying at 450 km/h. Military certification bypasses civilian airspace rules but introduces its own procurement timelines, which can stretch years beyond private-market expectations.

The cargo drone market projection of $103.7 billion by 2035 assumes regulatory and certification bottlenecks are resolved within this decade. That is not a given.

Sources

Scar: As we wrote in July, the broader defence robotics investment landscape grew to $27.6 billion in 2025. This article zooms into the autonomous logistics sub-segment within that surge.