A day after Anduril Industries pulled the sheet off Thunder — a hybrid-electric armed reconnaissance tiltrotor built to fly as a loyal wingman alongside crewed Army aircraft — Archer Aviation showed the other half of the program. On 22 July 2026, at the Farnborough International Airshow, the company unveiled Halo, an autonomous commercial cargo VTOL jointly developed with Anduril that shares Thunder's series hybrid-electric powertrain and core systems. Same bones, different mission set: instead of Hellfires and loitering munitions, Halo is pitched at offshore energy resupply for oil, gas and wind platforms, heavy time-sensitive freight between hubs, humanitarian and medical cargo, wildfire logistics, hospital deliveries and maritime missions.

It is an unusually clean example of the dual-use pitch that has become fashionable in defense tech over the past few years, and an unusually informative one — because the military sibling was announced first, and with far more detail. If you want to know what Halo can actually lift, the most concrete public numbers are the ones Anduril gave for Thunder.

What Archer actually said

Archer describes Halo as a clean-sheet design combining a series hybrid-electric powertrain with dual optimum-speed tiltrotors, aimed at long range and high speed. The "optimum-speed" part is the technically interesting bit: rotor RPM varies rather than staying fixed, which the company says holds efficiency in cruise and reduces noise for suburban and noise-sensitive operations. For an aircraft meant to operate into hospital pads and offshore platforms — and to do so repeatedly, at volume — acoustic footprint is not a marketing detail, it is a licensing and community-acceptance problem.

The rest of the design brief reads as a logistics argument as much as an aviation one. Archer says the tiltrotor configuration, powertrain and rotor technology give the platform enough range for global self-deployment, and that the aircraft can also be packed into a standard shipping container for transport by air, road, rail or sea when flying it there is not the answer. The company says it is designing for low-cost, high-volume production using commercial supply chains, building on the manufacturing base developed with Stellantis in Covington, Georgia, with heavy modular payload capacity and full autonomy. Archer frames Halo as the product of eight years of eVTOL development and flight testing, originally advanced through its Midnight passenger eVTOL program.

The launch partner is Marubeni Aerospace Corporation, the aerospace arm of the Japanese trading house Marubeni; Archer says additional partners are expected later in 2026. Marubeni Aerospace president and CEO Satoru Nakagawa tied the interest specifically to the uncrewed piece, noting that cargo flights that do not require onboard crew are still flown with human pilots today: "Halo's unmanned, autonomous flight capabilities can address these challenges," he said. Archer founder and CEO Adam Goldstein went further: "Over the past eight years we've engineered, tested and scaled the technology this platform is built on... I believe this is the most sophisticated vertical lift aircraft platform ever developed."

What Archer did not publish on 22 July is a payload figure, a range figure, an endurance figure or a service-entry date for Halo. Those are the numbers that determine whether an offshore resupply business case closes. For now they are absent.

The defense twin fills in the blanks

Thunder, revealed a day earlier at the same show on 21 July and detailed by Aviation Today on the 23rd, has been in development for three years. It is a Group 5 unmanned aircraft system — the largest UAS category — and Anduril positions it as an uncrewed loyal wingman supporting Army maneuver operations, flying vertical takeoffs and landings, transitioning to fixed-wing cruise, and then hugging the terrain in nap-of-the-earth profiles rather than loitering at altitude where modern air defenses are waiting. Higher-flying aircraft such as the MQ-1 Gray Eagle are being shot down and are "not relevant anymore," said Shane Arnott, Anduril's senior vice president for programs and engineering, adding: "that's why we've gone back to the surface." The aircraft mission software, including that for nap-of-the-earth autonomous flight, has been in testing for two years, he said.

The weapons load is where the shared airframe starts to speak for itself. Thunder's nose bay holds 12 counter-UAS effectors or other defensive payloads. The main body bay accommodates 10 Hellfires or Joint Air-to-Ground Missiles, 16 Altius-600 loitering munitions, or 76 Advanced Precision Kill Weapon System rockets. Maximum altitude is around 18,000 ft. It has landing gear, and it fits within 40-foot shipping containers — the same containerisation logic Archer is advertising on the commercial side. Flight testing is expected in 2027, with low-rate production targeted for 2029–2030.

Anduril president and chief strategy officer Chris Brose framed the value proposition bluntly, describing Thunder as priced at a "fraction of the cost" of the Army's future MV-75 Cheyenne II long-range assault aircraft and the Apache — attritable, even though it is meant to survive the battlefield and come home.

None of those payload numbers are Halo's cargo capacity — internal bays sized for ordnance are not the same as a freight hold, and Archer has not translated one into the other. But 76 APKWS rockets or 10 Hellfires in a single main bay, plus a separate nose bay, plus 18,000 ft and 40-ft container transportability, describe an aircraft in a very different weight class than the sub-100 kg delivery drones that dominate commercial cargo UAS coverage. That is the useful read-across.

The credibility question

Here is the part that does not appear in either announcement. Archer's federal prime-contract record is, by the government's own accounting, almost nonexistent.

A query of USAspending.gov for Archer Aviation prime contract awards from FY2023 through FY2026 returns exactly two records. The larger is award FA228023F0008, under parent IDV FA228023D0001, from the Department of the Air Force: $1,293,077, start date 26 July 2023, described as "MIDNIGHT AERIAL AND GROUND TERRAIN AVOIDANCE AUTONOMY." The other is FA864922P0797, also Air Force, for $744,796, starting 24 March 2022, for precision landing localization technology for autonomous eVTOL aircraft in GPS-denied environments. Total federal prime-contract footprint: roughly $2.0 million.

Set that against the partner. Anduril's recent records in the same database include a $362,974,500 Department of Homeland Security border tower purchase (70B02C26F00000035, awarded December 2025), a $93,994,917 Navy order for Altius 600M-V loitering munitions (N0016425CJR94, February 2025) and an $87,713,060 Army Lattice platform order (W9128Z26FA005, March 2026). Three awards, roughly half a billion dollars, versus two awards and two million.

That asymmetry is not a knock on Archer's engineering — the terrain-avoidance autonomy contract is directly relevant to what Halo has to do, and eight years of eVTOL development and flight testing through the Midnight program is real hardware experience. It is a statement about where the program's institutional gravity sits. Thunder is the anchor. Halo is the commercial derivative that gets to ride an airframe whose development risk is being retired substantially on defense money and defense schedules.

Why It Matters

Heavy uncrewed VTOL cargo has been perpetually five years away for a decade, and the reason is rarely aerodynamics. It is that nobody can get a large, complex, low-volume aircraft to a unit cost that closes a logistics business case. The Archer–Anduril structure is an attempt to solve that with production economics rather than clever design: one airframe, two customer bases, a commercial automotive manufacturing base in Georgia, and a defense program that supplies the volume floor.

For operators, the practical implication is that the first credible signal about Halo will not be an Archer press release. It will be whether Thunder flies in 2027 and reaches low-rate production in 2029–2030. If the defense timeline slips, the commercial variant slips with it, because they are the same aircraft.

For the wider UAS sector, this is also a marker of where Group 5 is heading. Thunder's design logic — go low, hug the terrain, accept that high-altitude loiter is now contested — and Halo's design logic — self-deploy, containerise, fly cargo into places helicopters currently serve expensively — converge on the same conclusion: the next generation of large uncrewed aircraft is being built to move itself and its payload without a runway, a hangar or a crew, and to be cheap enough to build a lot of them.

The open questions for Halo remain the boring, decisive ones. How much can it carry, how far, and what does an hour of flight cost against a chartered heavy helicopter running crews to an offshore platform? Until Archer publishes those, the strongest evidence for Halo is the contract file of the company standing next to it.

Sources