They say you can’t get anything done by sitting on your rear end, but US autonomy and unmanned systems specialist Shield AI expects to prove that wrong as it enters the realm of large uncrewed combat air vehicles (UCAVs) with its X-BAT proposal ^[1].

This is a concept that could transform US and allied airpower in the Western Pacific, freeing it to some degree from reliance on air bases that China has well equipped itself to smash at the beginning of a war. This is because X-BATs would be capable of vertical take-off and landing (VTOL): they wouldn’t need runways.

The company’s objective for extremely fast development looks challenging, however.

Unveiled on 21 October with a barrage of publicity and a full-scale mockup, the X-BAT is a different kind of UCAV. It’s big, with an estimated maximum takeoff weight about half of that a Lockheed Martin F-16 Block 70 or Dassault Rafale. That’s well over twice the size of most UCAV designs, and more like Dassault’s new UCAV ^[2].

The target unit production cost is $27.5 million, around a third of what a crewed fighter costs, and most UCAV developers are aiming at well under a fifth. Shield AI says the X-BAT’s through-life cost will be only a tenth of a crewed fighter’s. And the X-BAT takes off and lands vertically, all the time.

1950s retro-tech is the key to X-BAT’s performance: it is a tail-sitter. All flights begin and end on a launch and recovery trailer that puts the recovered aircraft in a horizontal attitude, wings folded, for refueling, rearming and storage, then rotates it to the vertical for launch; the aircraft comes back with its nose pointed upwards, too.

An engine of a type used in the F-16—the Pratt & Whitney F100 or General Electric F110—will give more than 110 kilonewtons (25,000 pounds) of thrust in full afterburner to provide a thrust/weight ratio of more than 1:1, which is what you need if you want to go straight up. With heavy weapons and fuel gone, non-afterburning thrust is enough for landing.

The X-BAT has a lightweight, blended airframe that has a lot of internal volume for its size, and it is said to accommodate air-to-air weapons, a full sensor and communications suite, and enough fuel for a 1,850 km (1,000 nautical mile) combat radius. Weapons as large as Northrop Grumman’s AARGM-ER anti-radar missile can be carried externally.

Range and tail-sitting address the problems of deploying and sustaining large numbers of UCAVs without overburdening forward bases. X-BATs can be based well away from the war zone, with trailers stored closer to the theater. The aircraft can be dispersed around main bases to complicate targeting.

For aviation geeks, tail-sitting calls to mind videos of the US Navy’s Convair XFY-1, nicknamed Pogo, a small aeroplane with a massive propeller that was almost impossible to fly ^[3], the pilot recumbent with his feet in the air. (‘A good position, but not for flying,’ as Harrier designer John Fozard put it.)

But the jet-powered Ryan X-13, a near contemporaryof the XFY-1, was far more successful, performing dozens of transitions from an X-BAT-like trailer between mid-1957 and early 1958. In July 1957, the tiny jet flew a sortie from the shore of the Pentagon Marina ^[4], which nobody would have dared do with any other VTOL of the time.

Modest 1950s engine thrust/weight ratios limited the X-13 to lifting only itself and minimal fuel. That has changed. Flight-control systems have advanced, and the tail-sitter emerges as the simplest possible VTOL: downward thrust isn’t created with complex gear to vectoring the longitudinal engine output or by adding a heavy fan; instead, the whole aircraft is rotated so the engine is just pointed downwards. Using the trailer eliminates the cost and weight of landing gear and solves the problem of how to store or move a top-heavy 40-foot-high aircraft.

Shield AI’s schedule and cost targets hover (pardon the expression) between the aggressive and aspirational. The company has been working since early 2024 and plans a flight demonstration in the second half of 2026, full system flights in 2027 and production by 2029.

This is not much time unless an industry team and finance can be put together quickly, because there are multiple flight modes to be explored, and most flight or systems testing cannot be done until VTOL is mastered.

Shield AI is most likely banking on using F100 or F110 engines that are already in US Air Force F-15s and F-16s and won’t need overhauls for hundreds of running hours (all that a UCAV needs). The engines will become available as the F-15s and F-16s are retired. But they will need a vectoring nozzle—and this in turn raises data-access issues unless that job is done by the original engine manufacturer.

Hitting cost targets will require innovative approaches to avionics, particularly sensors, since Shield AI is promoting X-BAT as not only an approach to the Collaborative Combat Aircraft requirement for fighter-like drones that fly with fighters; the company also promotes it as an aircraft that can operate in autonomous formations, without a crewed aircraft to act as Nanny.

I produced a back-of-the-envelope study of an X-13-like UCAV in my last job, and British Aerospace looked at something similar in the 1980s. Mine was much smaller than the X-BAT (the carrier was designed to look like a commercial semi-trailer) but the study raised some important issues.

While long range, VTOL and the trailer address your strategic mobility problems, a trailer that needs a ten-wheel Oshkosh heavy tractor to pull it is not great for tactical mobility on Asia-Pacific road systems. And since an aircraft like an X-BAT can’t land without a trailer, you must have spare trailers.

A former BAE Systems engineer adds that the high-velocity outwash from an afterburning nozzle aimed at the ground can blow over things and people within a wide radius and that each trailer system needs a dedicated crew.

These limitations in my study led to another realisation, which was that, while the island chains were not well endowed with places to put long runways, they had plenty of sheltered water.

Multiple X-BATs and a launch-and-recovery platform, plus fuel and munitions, would fit on a ship of well under 10,000 tonnes displacement, and even within the dimensions of fast ferries and smaller roll-on, roll-off ships. This would be not a sea warfare ship but a readily relocatable platform that to complicated the enemy’s targeting.

And that raises the possibility of adding a small catapult or other take-off boost system: you could ditch the afterburner and use an engine that’s better for long range.

But Shield AI will surely persist with its current concept for now. The X-BAT has its challenges, but it can’t be faulted for original thinking, and a simple runway-free aircraft that’s effective in air combat and strike could be what’s needed to make sure that a Western Pacific allied force survives on the ground.