Air Force Gen. Mark Kelly, the head of Air Combat Command (ACC), envisions a future fleet of multi-role Collaborative Combat Aircraft (CCA) that build off initial fielding lessons learned for a small number of CCAs.

“If I was doing a clean sheet design, I would look for something you could iterate,” Kelly told reporters at the Air & Space Forces association conference this week at National Harbor, Md. “You could bolt on growth. You could get an order one day going, ‘Hey, these guys and gals at Nellis [AFB, Nev.] want an eight-ship, and they want them all to be X-Band radars.’ So, let’s shift to X-Band radars. Unlock a nose, bolt on another nose. The next day…they want all jammers. So really quickly, let’s take off the X-band radars and put on the jammers.”

“I would not lock myself into, ‘It’s a sensor and can’t do anything else, or it’s a jammer and can’t do anything else, or it’s not armed, and you’ll never arm it,'” he said.

Kelly said that he believes in the next two to three years the Air Force will conduct operational testing of several CCAs out of an Air Force location or locations that can launch drones–Creech AFB, Nev., Tonopah Test Range, Nev., Holloman AFB, N.M., and/or Tyndall AFB, Fla.

“We’ll put them on the red air side,” he said of the CCAs. “Our [Air Force] Weapons School students will fly against them, and one of those Weapons School students is gonna say, ‘You know. This problem set would be a lot easier, if we had them on our side.’ And maybe two rides later, they’re flying on the blue side…I see us iterating and testing our way into this [CCA concept].”

Beside autonomy, important features for CCAs to have will likely be communications systems that can withstand interference and have permissions for take-off, operations, and landing in the United States.

Air Force Secretary Frank Kendall has said that his top technology priority is “autonomous behaviors and artificial intelligence decision support,” and he has suggested that the Next Generation Air Dominance (NGAD) manned fighter could serve as a play caller for autonomous CCAs (Defense Daily, Aug. 11).

For CCA, the Air Force may glean lessons learned from other manned-unmanned teaming efforts, such as the Skyborg Vanguard program, and may take a step toward equipping fighter squadrons with some initial stage CCAs in next year’s fiscal 2024 budget request.

The Air Force has been developing an acquisition strategy and timeline for CCA, up to five of which may be employed by each “quarterback,” manned NGAD fighter, which Kendall recently said has entered the engineering and manufacturing development stage. Air Force acquisition chief Andrew Hunter has said that he envisions more than one supplier of CCA, rather than a “winner take all strategy.”

The Air Force’s Skyborg Vanguard program has yet to transition to production, but the service has said that the Kratos [KTOS] XQ-58 Valkyrie is an example of what Skyborg drones may look like.

Other drones, such as Boeing‘s [BA] Airpower Teaming System (ATS) under development as the MQ-28 Ghost Bat for the Royal Australian Air Force’s Loyal Wingman project, are also possible CCAs for the U.S. Air Force. ATS is a stealthy, multirole, unmanned aircraft system.

Among the largest defense contractors, Northrop Grumman [NOC], BAE Systems and Lockheed Martin [LMT], which has invested $100 million in distributed teaming technologies under the company’s Project Carrera to inform CCA and other efforts, are also expected to submit proposals for CCA.

Each of the manned NGAD aircraft may cost “multiple hundreds of millions of dollars,” Kendall has said, while a high number of CCAs operated by manned NGADs would help lower the cost of CCA. Because of the extended range and mission payloads needed for CCAs, however, the latter could be significantly more expensive than other drones and would have about half the unit cost of the manned NGAD, Kendall projected.

General Atomics is to offer its Gambit series of drones for Air Force CCA and is aiming for a significantly lower price point than that mentioned by Kendall.

“What makes this [Gambit] different than what I think you see from other folks is that we’re centered on what we’re calling the Gambit core–chassis, avionics, flight controls, landing gear in a unified body,” said C. Mark Brinkley. a spokesman for General Atomics Aeronautical Systems, Inc.

The Gambit core “represents about 70 percent of the price of the system,” he said. “That allows you to build at scale. You get an economy of scale across this collaborative system of aircraft. On the other 30 percent of the price, you decide what you want it to do.”

Gambit 1 is to be a long-endurance intelligence, surveillance, and reconnaissance CCA, while Gambit 2 would be a Loyal Wingman to manned fighters in an air-to-air role. Gambit 3 would serve as an adversary air trainer for the Air Force against F-35s and other fighters, while Gambit 4 would be a tailess, low observable, air-to-ground combat reconnaissance drone with swept wings.

“You have to figure out, ‘How do I regulate the price?'” Brinkley said of companies’ CCA offerings. “I think you regulate the price by having common systems, generate that across the full run of all your individual systems, and then you decide, ‘How many adversary air trainers do I need? How many ISR/no weapons do I need? How many Loyal Wingmen? How many silver bullet low-observable?”

General Atomics is aiming for the average cost of a Gambit to be “well below” $20 million, Brinkley said, perhaps $13 million to $17 million.

Kratos’ possible offerings for CCA include rail-launched and runway independent drones, including the Valkyrie, the Unmanned Tactical Aerial Platform (UTAP-22) Mako, and the Air Wolf.

Ensuring that CCAs do not make costly and fatal errors will be critical to building trust in them for their employment as DoD assets.

Steve Fendley, the president of Kratos’ unmanned systems division, said that “the technology level that’s required for an unmanned aircraft to go fly by itself and go do things that are predictable and deterministic, we’re there.”

“That exists today,” he said. “We fly them all the time. We’ve done a number of flights this year and dozens of flights over the past couple of years with a system that everything that system does is known, not necessarily because there’s a human on the ground telling it exactly what to do, but even the pre-programmed part for the autonomy part of the mission is all well-defined and structured and follows a basic rule set.”

“What’s really hard is the artificial intelligence [AI] piece where you have non-deterministic decisions being made by something that, if it’s true AI, you can’t predict it,” Fendley said. “Otherwise, it’s not true AI.”