Loyal wingman

The principal application is to elevate the role of human pilots to mission commanders, leaving AIs as "loyal wingmen" to operate under their tactical control as high-skill operators of relatively low-cost robotic craft.[3][4][5]

Loyal wingmen can perform other missions as well, as "a sensor, as a shooter, as a weapons carrier, as a cost reducer".[6][7]

A loyal wingman is expected to cost significantly less than a manned fighter, and will typically be considered vulnerable to attrition.[8] It would have sufficient intelligence and onboard defence systems to survive on the battlefield. The United States Secretary of the Air Force Frank Kendall has described them as remotely controlled versions of targeting pods, electronic warfare pods or weapons carriers to provide additional sensors and munitions; to balance affordability and capability,[9][10]

Boeing Australia is leading development of the MQ-28 Ghost Bat loyal wingman for the RAAF, with BAE Systems Australia providing much of the avionics.[11] The MQ-28 was first flown in 2021.

China has been studying the loyal wingman concept since at least 2019 and has shown off some concept airframes. However, although China already manufactures drones and has well-developed swarming technology, the planned level of autonomy or even AI for these systems are not known.[12][13]

The HAL CATS Warrior is an AI-enabled wingman drone under development by Hindustan Aeronautics Ltd. (HAL) for the proposed Combat Air Teaming System (CATS).

Japan announced a development programme for a loyal wingman drone in 2021, issuing the first round of funding in 2022.[14][15] The drone is intended to be carried for deployment by a proposed F-X fighter, also under development.[16]

Russian projects for wingman-class drones are thought to include the Sukhoi S-70 Okhotnik and the Kronstadt Grom. However, although Russia already manufactured drones, the planned level of autonomy or even AI for these systems are not known.[17][18]

The RAF in the UK has been developing the Loyal Wingman concept since 2015, with the Spirit Mosquito technology demonstrator flying in 2020. Programme funding was cancelled in June 2022 because the Ministry of Defence felt that it was better spent on less ambitious advances.[19]

Collaborative combat aircraft (CCA) is the official USAF designation for an AI combat drone, and is broadly equivalent to the loyal wingman.[20]

XQ-58A Valkyrie unmanned combat aerial vehicle

The USAF Next Generation Air Dominance (NGAD) program was initiated in 2014. It includes the development of CCA.[21][22] Up to five autonomous CCAs could operate with a manned fighter.[23]

The Skyborg programme, going back at least to 2019, is developing the systems to operate wingman drones alongside advanced manned fighters. Of four contenders, the most public is the Kratos XQ-58A Valkyrie. The Air Force Research Laboratory (AFRL) will test their Skyborg manned-unmanned programs such as Autonomous Air Combat Operations (AACO),[24] and DARPA will test its Air Combat Evolution (ACE) artificial intelligence program.[6][25] The System for Autonomous Control of Simulation (SACS) software for human interface is being developed by Calspan.[25]

DARPA AlphaDogfight Trials (ADTs),[3] 20 August 2020

In 2020, DARPA AlphaDogfight predicted that AI programs that fly fighter aircraft will overmatch human pilots.[26][27]

The General Dynamics X-62 VISTA is a modified F-16 which can fly autonomously, with a test pilot to take over if necessary

Two alternative autonomous AI systems have been installed in a General Dynamics X-62 VISTA at the Air Force Test Pilot School. The two systems flew the aircraft in turn, on 9 December 2022. By 16 December 2022 the X-62 Vista had flown eight sorties using ACE, and six sorties using AACO, at a rate of two sorties per day.[24]

The General Atomics Longshot is intended to be carried for deployment by the manned aircraft, and is air-launched when needed. This allows a shorter range for the drone, while maintaining advanced protection for the manned aircraft. DARPA adopted the General Dynamics design for its Longshot programme in 2022.[28]

In 2022 Heather Penney identified five key elements for the proactive development of autonomous CCA, remote pilots of UAVs and pilots flying separately in manned aircraft (also called crewed-uncrewed teaming, or manned-unmanned teaming).[29][30]

1. Create concepts that will maximize the strengths of both CCA and piloted aircraft working as a team.
2. Include operators in CCA development to ensure they understand how they will perform in the battlespace.
3. Warfighters must be able to depend on CCA autonomy.
4. Warfighters must have assured control over CCA in highly dynamic operations.
5. Human workloads must be manageable.

A typical CCA is estimated to cost between one-half and one-quarter as much as an $80 million F-35.[31] US Air Force Secretary Frank Kendall is aiming for an initial fleet of 1,000 CCAs.[32]