The Rise of AI in Aviation.

The aviation industry is undergoing a transformation unlike anything we’ve seen since the advent of the jet engine. Artificial intelligence (AI) is no longer confined to science fiction or experimental labs—it’s in the cockpit, and it’s here to stay. But what does this mean for pilots, passengers, and the future of air travel?

With so much turmoil in aviation over the last few months, are we on the brink of a safer, more efficient era, or is too much control being handed over to machines?

What can we expect to see?

Due to the anticipated shortage of pilots in the coming decades- airlines around the world are looking for solutions that allow their full fleet to be operational, rather than cancel flights & ground airplanes due to a lack of staff. One solution to this problem is autonomous aircraft. In the current day- AI is already a crucial part of aviation, assisting in so many aspects of a passenger’s experience, whether it be in; Air Traffic control- where AI collects data in order to optimise the timings of outbound & inbound flights, reducing congestion & therefore lowering the risk of collisions; or whether it be in predictive maintenance- where AI can analyse data from aircraft sensors to predict when an airframe needs to undergo service, subsequently reducing the risk of a failure in flight.

Now, AI is being used to enhance the quality of flight, in both standard and emergency situations. Airbus ‘UpNext’, a subsidy of the Airbus brand, has begun testing on a new, on-ground & in-flight pilot assistance program using the A350.

The system- known as ‘DragonFly’- involves three front-mounted cameras which collect data about external conditions (i.e., visibility, winds, etc), and cross-references these with local geographical data (e.g. suitable airports), providing the plane with the ability to perform actions such as; automated emergency diversion in cruise; automatic landing and taxi assistance, in the unlikely event that one, or both pilots become incapacitated. This program is both aimed at evaluating the feasibility of autonomous flight systems and improving the safety & efficiency of operations- allowing for a greater quality of passenger experience.

Whilst DragonFly™ are the one of the industry leaders in the development of commercial autonomy- AI has been a part of military operations for decades. In 2007, the RQ-170 Sentinel UAV (Unmanned Aerial Vehicle) was introduced into the US air force as a supplement for the aging U-2 reconnaissance jets.

The Sentinel is a high altitude, long-range pilotless aircraft designed by Lockheed Martin- that allows real time imagery of a battlefield to be captured, providing key information about any opposing factions that a country may be fighting against. The most notable use of the RQ-170 was its deployment in Pakistan- where it provided live coverage to the White House of Osama bin Laden’s assassination on the 2nd of May 2011.

Another example is the recent FAA approval for reliable robotics to begin installing autonomous flight systems on a USAF KC-135 refueler, an airframe that dates back to the early 1960s. Whilst initial testing will have pilots on board, the idea is that in time the tanker will be able to perform a full refuelling sortie (mission), and then flawlessly return to base, all without the assistance of a captain on the flight deck. Reliable Robotics state that they are looking to ‘enter several different markets, both commercial and military’ – hinting at the prospect we may see autonomous cargo flights within the next 5 to 10 years.

What’s the Opposition?

There are several reasons we haven’t seen pilot-less flights more frequently, the first of which is the attitude of the general public. Gross scepticism around full autonomous flights and single pilot operations (SPO’s) has prevented legislation to increase the scope of testing from being passed by governing bodies such as EASA and the FAA. In fact, In fact, in a survey by ALPA back in 2019- over 85% of the travellers who participated stated that they would feel ‘very uncomfortable’ not having two pilots in the cockpit, never mind having none- presenting a huge obstacle for AI in aviation- as you can develop this technology as much as you’d like- but if passengers don’t trust it, who’s going to use it?

Another significant obstacle is the way in which some of these AI models function. Many AI systems that would be used in autonomous flight operate as "black boxes," making it very challenging to understand what parameters the system used to make a decision. This becomes especially challenging during an incident (e.g. a crash), when investigators may struggle to full uncover what caused the plane to go down. Furthermore- this ambiguity raises the question of who to blame for such an incident. Is it the fault of the AI system and its decision making, or does the fault fall onto the designer of the system, and the algorithms that was developed during its creation.

A final reason (but by no means are these the only three issues) is the problem of cybersecurity. In an world that is becoming increasingly digital, the AI systems that would be implemented on aircraft are also vulnerable to adversarial attacks and exploitation if not properly secured. We have seen tragic hijacking events in the past- including the devastating loss in 9/11, so what happens if hackers had the ability to control a whole network of aircraft without moving from their desktops?

Whilst a hypothetical thought in principle- it is a real possibility that needs to be considered, especially when we see the high global tensions, both in Crimea and the Middle East.

What’s Next?

The future of AI in aviation is both exciting and uncertain. By 2030, some experts predict that fully autonomous passenger flights could become a reality, at least on short-haul routes. Others believe that AI will remain a tool to assist, rather than replace, human pilots.

But the question remains: Are we ready to embrace a future where the “pilot” is a machine? Will passengers trust AI to get them safely to their destination? And what does this mean for the hundreds of thousands of pilots whose jobs could be at risk?

As we stand on the cusp of this new era, one thing is clear: AI is reshaping aviation in ways we’re only beginning to understand. Whether it’s a revolution or a reckoning, our skies will never be the same.