Autonomous at Scale: Reading the Market Map.

The defence and intelligence, surveillance, and reconnaissance segment is the most mature, most liquid, and most capital-intensive part of the autonomous systems market. It has programme-of-record procurement, multi-year government contracts, and an established industrial base — and it is the segment that has driven the largest transactions of the current cycle.

AeroVironment is the clearest example of what a scaled defence drone business looks like: its $874.3 million DoD IDIQ contract, covering Jump 20, P550, Puma, Raven, and Titan counter-drone systems, represents production and sustainment revenue locked in over five years with foreign military sales provisions. That is not a speculative market opportunity. It is a contracted revenue base with a clearly defined industrial footprint.

The attritable and expendable UAV segment — low-cost autonomous systems designed for single or limited-use missions, drawing direct lessons from the Ukraine conflict — is the fastest-growing sub-segment within defence. The U.S. Replicator programme, designed to field thousands of attritable autonomous systems across domains, is the most significant programmatic demand signal in this space and has triggered a set of procurement competitions that will define the supplier landscape through 2030.

The key competitive dynamic in defence autonomy is the tension between established primes — AeroVironment, Northrop Grumman, General Atomics — and new entrants with software-first architectures: Anduril, Shield AI, and Joby’s defence division. The new entrants are winning individual competitions; the primes are winning sustainment and foreign military sales. Both will coexist, but the integration and M&A pressure between these two populations will be the dominant structural dynamic of the next three years.

Commercial drone logistics is real. It is operating. It is generating revenue. Wing has completed more than 450,000 operations across three continents. FlyingBasket is running 30 logistics trips per day on established offshore and alpine routes. Dronamics has self-authorised cross-border cargo operations across EU member states from its Malta base. These are not pilots. They are production operations.

The constraint on the segment is not technology and it is not demand. It is regulatory geography. As established in our EMEA analysis, BVLOS authorisations are accumulating corridor-by-corridor rather than being granted at scale. The FAA’s Part 108 rulemaking — with a final rule expected in spring 2026 and implementation 6–12 months thereafter — is the single most consequential regulatory event for U.S. commercial logistics drones in the current cycle. Its passage will expand the addressable operating envelope for routine BVLOS delivery, inspection, and autonomous logistics in the U.S. market more than any other single development.

In the interim, the commercial logistics market is segmenting predictably. Last-mile consumer delivery — Wing’s model — is scaling where regulatory permission exists and where density economics work. Middle-mile industrial logistics — FlyingBasket’s model — is scaling where the operational case is clear and the regulatory hurdle, while significant, is tractable for a well-resourced operator. The market opportunity in drone logistics is not evenly distributed: it is concentrating in the operators who have already paid the regulatory authorisation cost, and in the enabling technology vendors — detect-and-avoid, UTM, ground infrastructure — that every operator needs.

Across both defence and commercial segments, the assets that demonstrate the most consistent capital attraction are those that sit at the enabling layer of the stack rather than at the platform level. Auterion — the open-source flight software company — raised $130 million at a valuation that reflects software margins and platform network effects, not hardware unit economics. Unifly’s acquisition of EuroUSC Italia in May 2025 represents UTM software consolidation in Europe. The U-space deployment mandate coming in 2026 across EU member states will create mandatory procurement demand for UTM services that is entirely regulatory-driven.

The enabling stack has three characteristics that make it more attractive than platform hardware as a sustained investment position: it is hardware-agnostic (it serves multiple manufacturers), it has network effects (more operators on the platform increases data value and switching costs), and it converts more directly to recurring software revenue rather than lumpy hardware sales cycles.

The market map, properly drawn, shows a sector in the process of following the pattern of every previous technology market: the platform layer commoditises over time, the software and services layer concentrates value, and the infrastructure layer — U-space, droneports, corridor certification — becomes a regulatory moat with durable barriers to entry.