The artificial intelligence ‘backseater’ in future air combat
5 Feb 2021|

Among the myriad technological developments in airpower—including advanced propulsion, better stealth, directed-energy weapons and hypersonics—perhaps the most important will turn out to be artificial intelligence. AI has the potential to transform air combat operations and the way airpower is conceived and used.

The US Air Force demonstrated the role of AI dramatically in December last year by flying a U-2 reconnaissance aircraft, which normally carries a single crew member, with an AI algorithm as a ‘virtual backseater’. As the pilot flew, the AI system—dubbed ‘ARTUµ’ in honour of Star Wars robot R2-D2—controlled the aircraft’s sensors and navigation.

On the test flight, ARTUµ was tasked with finding adversary missile launchers and was ‘solely responsible for sensor employment and tactical navigation’, while the human pilot concentrated on finding enemy aircraft and flying. ARTUµ ‘made final calls on devoting the radar to missile hunting versus self-protection’.

The flight demonstrated human–machine teaming at a new level and points to the potential to integrate AI backseaters into current fighters like the F-35 and future air combat aircraft such as the US Next Generation Air Dominance platform, or NGAD.

The F-35 has room for just one person in the cockpit, but with AI the pilot gets a backseater that can manage the complex process of integrating data from a multitude of onboard and offboard sensors and other information sources, freeing up the pilot to fly and fight. AI can also manage human–machine teaming between the F-35 and autonomous systems, such as the ‘loyal wingman’ autonomous platform being developed in Australia as part of the Boeing airpower teaming system.

So, AI dramatically reduces the aircrew’s workload and can process data and information faster than a human pilot, including rapidly assessing threats. But it also boosts the operational potential of autonomous platforms in crewed–uncrewed teaming because it can ensure greater oversight and control—‘on the loop’ for the autonomous platform—while giving the human pilot the big picture of tactical operations. That eases rules-of-engagement constraints on using autonomous systems, including future lethal autonomous weapons platforms.

For fast jets like the F-35, the benefits of integrating AI are obvious—it opens up new approaches to complex and fast-moving air combat operations.

For slower, heavier aircraft such as the E-7A Wedgetail and the P-8A Poseidon, integrating AI inside the cockpit is likely to enhance the operating efficiency of the planes and their crews, particularly as they begin to exploit offboard sensors based on drones such as the MQ-4C Triton, maritime platforms or space-based capabilities. AI can manage a diverse network of sensors and platforms and take the lead in transforming the flood of data gathered over the future battlespace into a knowledge edge—the key to getting the information advantage that’s critical for success.

The role of AI in the cockpit gives us a glimpse into the future of airpower. The Royal Australian Air Force needs to be thinking about where airpower goes next, even as the F-35 achieved initial operational capability in December and moves to final operational capability in 2023. That future is almost certain to be a mix of crewed and autonomous systems, working together in teams to deliver a range of effects as a system of systems.

We need not wait until the late 2030s for a sixth-generation platform to emerge to replace the fifth-generation F-35 in the 2040s. It would be a mistake to delay consideration of future air combat capability by maintaining a traditional acquisitions mindset. Yet, the 2020 force structure plan pushes back the start of the F-35 replacement process to the mid-2030s. Given the pace of technological change and the rapid deterioration in our strategic environment, that notional timeline needs to be challenged and capability acquisition needs to happen faster. With the US already flying a demonstrator for NGAD and considering a ‘digital century series’ approach to acquisition, it’s time Australia shook off the slow, generational acquisition paradigm. Investment in breakthrough technologies such as AI is part of that transformation.

A complementary next-generation air combat system spread across multiple platforms and technologies—crewed and autonomous—that has AI as a key component needs to be prioritised as an opportunity for common development with our Five Eyes partners, especially the US. The Australian Defence Force could work with its US counterparts, or with allies and partners in Europe and Japan, to develop future air combat capabilities that will complement and, ultimately, replace the F-35 sooner than the 2040s. Boeing’s ‘loyal wingman’ drone is a good first step and may be able to be evolved into a potent future air combat capability.

It’s time for the RAAF to be forward-looking and proactive in accelerating the acquisition of the next generation of air combat capabilities. That effort should incorporate the broad application of AI—inside the cockpit and across the battlespace—to allow our military to think and act at machine speed.

Of course, there are risks in moving fast. AI is still a relatively new technology, and its complexity needs to be mastered. Networking a multitude of offboard sensors that an AI algorithm can use to inform aircrew of the common operating picture will require significant investment in software development, supported by the establishment of research facilities and the training of personnel. It will require a substantial boost in digital high-speed communications bandwidth to work. That’s likely to require investment in other technology areas—notably, satellite-based laser-optical communications and quantum technologies.

Nor will the ADF operate in an uncontested electromagnetic domain. Potential adversaries such as China are already developing sophisticated capabilities in electronic and network warfare, so our tactical command and control, including elements controlled by an AI ‘backseater’, must be robust and resilient.

Meeting all of those challenges will take time, money and skilled personnel. The US Air Force’s U-2 demonstration of ARTUµ is one step down a path towards tomorrow’s airpower, and Australia needs to proceed with determination to achieve success.