We’re still feeling our way with this.

The concept of fighter-like drones, called collaborative combat aircraft (CCAs), holds much promise to air forces, notably to the US Air Force as it contemplates war with China. But maintaining CCAs in operation isn’t looking cheap and simple.

Air forces may instead drift towards getting some CCA effects with expendable drones that can be treated much like rounds of ammunition.

The Royal Australian Air Force has been working for more than six years with Boeing on a CCA design, the MQ-28 Ghost Bat, but the US Air Force, having started from behind, is now well on its way to launching large-scale production of similar aircraft.

The test aircraft for Increment 1 of the USAF’s CCA program, the General Atomics YFQ-42A and Anduril YFQ-44A, are starting their ground tests and are expected to fly this summer. From what we know, their principal mission is demonstrating the use of CCAs to carry air-to-air missiles and engage targets that are tracked and identified by pilots in crewed aircraft.

Supporters say simulations and wargames have shown the value of CCAs, which look like small fighters and are faster and more manoeuvrable than other drones. Part of their value is to put more missile shots in the air, from more places and closer to the enemy. They would present the enemy with a denser mass of threats and targets and increase confusion, reducing losses among crewed aircraft.

Speaking at an 8 May Mitchell Institute webinar, USAF Director of Force Design Major General Joe Kunkel said the service planned to buy 1,000 CCAs. ‘We’ll see sizable investment in the 2026 budget and pretty heavy investment in the FYDP,’ the future year defense plan, he said.

The webinar was the launch of a new Mitchell study, based on a tabletop exercise last year that focused on sustainment of a CCA force in the Western Pacific and was premised on the force mixes that emerged from a 2023 wargame.

The challenge of sustainment emerged from the earlier study. CCAs may be relatively small, but the Increment 1 aircraft are about the size of the smallest crewed jet fighter ever used operationally, the British Folland Gnat of the 1950s. They are aeroplanes that need runways, fuel, parking and maintenance. Let’s call them robo-gnats.

If they were much smaller they would have less range and less payload. Range is ‘critically important, and always an issue,’ Kunkel says. Higher payload translates to an ability to put more weapons in the air, another vital attribute.

Although the USAF plans to stand-up an ‘aircraft readiness unit’ at Beale AFB in California to rehearse CCA operations, it’s not likely to be practicable to treat several hundred robo-gnats as a US-based deployable force: they’d lack range and inflight refueling capability. So they’d have to be kept on forward bases.

But on those bases they’d need protection, and their operators would have at best a tedious time, training to launch and recover missions that mostly would not take place (because there’d be no pilots who’d need to fly to train).

The Mitchell study had a few palliatives, such as designing the CCAs to use shorter runways and dispersing them around more bases, which could be a partial substitute for protection. The USAF’s target for maintenance is that the aircraft should be able to operate for 60 hours with only refueling and rearming.

Logistics could be eased by planning and design. The USAF expects to acquire and operate several different CCA designs, but it is simplifying things by specifying common families of parts, from actuators to tyres, and common interfaces for sensors and weapons, Kunkel says. Another recommendation from the report was developing smaller, lighter weapons so more could be carried.

Meanwhile, on the other side of the world, another announcement pointed to the possibility that we’re looking at CCAs the wrong way. The Royal Air Force unveiled a working capability for the  Storm Shroud, a box full of defence-suppression. This is a Portuguese-designed and British-built Tekever AR 3 drone— driven by a propeller, stored and transported in a container, launched from a portable catapult, and fitted with a payload based on Leonardo’s Brite Cloud active-radar decoy.

It doesn’t look like much until you realise that a flight of such aircraft can be launched towards a radar far over the horizon, arriving just before the main attacking force, and delivering debilitating jamming effects just kilometres from the radar. The drones may well all get shot down, but if the main force meanwhile punches through the defense line unscathed, that’s a win.

A few lines in the Mitchell report pointed in that direction. ‘We steered clear of the term “attritable”,’ said study director and author Mark Gunzinger, because the willingness to lose aircraft depended so much on the scenario. It wasn’t just a matter of aircraft design.

The other pointer was that the Blue teams requested far more expendable systems than recoverable ones and highlighted their logistics advantages. ‘Designing CCA to be more like fighters increases their logistics footprints,’ the report states, whereas ‘designing CCA to be more like weapons significantly reduces their logistics footprints—it’s like a sliding scale.’

Runway-independent CCAs ‘helped operators to remain unpredictable’ and complicated Red’s airbase attacks. And, clearly, expendable systems can be much smaller than CCAs that must make a return trip, and, like munitions, they need no maintenance.

In short, expendable CCAs launched without runways are a different approach to the game of adding mass to the force, one that doesn’t involve buying a lot of logistical problems.

The conclusion is that powerful operational incentives are pushing towards one-way vehicles. Technology is helping them: small, cheap but effective payloads such as Brite Cloud, generic software-defined radio systems for communication, and unjammable optical navigation.

This may turn out to be the right solution for a lot of what air forces want CCAs to do.

Robo-gnats might be useful and initially necessary, and it’s entirely natural for air forces to like things that look like traditional aircraft. But it could be that money should instead be going into driving down the costs of small airframes, small engines, and flight controls then buying them in big numbers and expecting to lose them as we use them.