Taking Technology to New Heights: Artificial Intelligence in Aviation

Global Aerospace Editorial Team, January 29, 2018

As defined by research firm Gartner, artificial intelligence (AI) is “technology that appears to emulate human performance typically by learning, coming to its own conclusions, appearing to understand complex content, engaging in natural dialogs with people, enhancing human cognitive performance (also known as cognitive computing) or replacing people on execution of nonroutine tasks.” Many applications of AI, which would have been considered science fiction in the not too distant past, are becoming commonplace today—things like self-driving cars, digital home assistants, and autopilot systems that can manage an entire flight from takeoff, climb and cruise to descent, approach and landing, all without human intervention. However, as AI researchers around the world know, those simple tasks are just scratching the surface of what computer “neural networks” will one day handle.

aviation ai

The goal of creating machines that can think like humans has been pursued since the computer was first invented. Will this latest surge in interest be a passing fad? According to most experts, AI is here to stay, in large part because new technology like parallel processing, cloud computing, and advanced “learning algorithms” has removed the roadblocks that stymied many past initiatives.

Safer Flights

In its ongoing quest to make flight safer and more efficient, the aviation industry has always been an early adopter of new technology. A simple autopilot, that keeps an airplane flying straight and level is just one example, and already there are systems that make that impressive feat of engineering seem downright crude. For example, Garmin’s Telligence Voice Command system can manage many basic cockpit tasks including reading wind forecasts, changing radio channels, and providing details on current position on demand. It’s not hard to imagine a whole host of additional functions that systems like this will be taking on in the near future.

The big leap will come when we begin trusting AI not only to provide data or respond to simple commands, but to make decisions, and in particular, decisions in scenarios that fall outside the “norms” of aviation operations. We don’t ask a great deal of today’s autoflight systems (relative to what we believe they can achieve one day), but even at that they are quick to surrender and will return control to the flight crew in challenging situations like excessive turbulence.

Watching to Learn

But it appears that the “big leap” is just around the corner, and the ability of AI systems to learn will be the key. For example, researchers are creating systems that can develop skills for handling in-flight crises by “watching” how well-trained and experienced flight crews respond in similar scenarios. Soon, an autoflight system won’t be switched off during an emergency, but instead will be actively assisting the crew precisely where and when its input is needed.

Will AI replace pilots? Not anytime soon. As observers point out, there are a number of fields in which AI is now playing an important role—everything from railways to healthcare—and in none of them have humans been displaced. Instead, all signs point to AI systems functioning as increasingly capable digital assistants that work in concert with their flesh and blood counterparts. With all its potential to improve aviation, the “rise of the machines,” portrayed in fiction as something we would be wise to fear, is actually something we would be foolish not to embrace.