A new generation of autopilot is here. But can it fully replace human aviators?

It’s a scenario worthy of a horror movie: You’re flying as a passenger in a small aircraft when some emergency, perhaps a medical condition that wasn’t detected as part of an on-the-job physical, incapacitates the pilot. The plane’s autopilot kicks in as it should, but all this does is keep the plane headed in the right direction and cruising at the correct altitude. It is basically a holding pattern — except that, unlike a regular holding pattern, in this case, there is no one able to safely land the plane. With no trained pilots on board, it’s only a matter of time before the aircraft runs out of fuel and crashes.

Garmin, a company that manufactures marine, aviation, and consumer technologies that run on GPS, has come up with an answer. The company’s idea was simple: To develop a single button, the airplane equivalent of Amazon’s one-click purchase, that would allow pilots or passengers to land an aircraft autonomously. That concept, once a dream, was this month awarded FAA certification for general aviation aircraft, proving the concept to be effective.

Welcome to the next phase of autopilot technology.

The strongest link… and the weakest

“[The pilot is] both the weakest and the strongest link,” Phil Straub, executive VP and managing director of aviation at Garmin, told Digital Trends. “Nothing can really beat the human intuition, the brain, the sensory system, and all that. But we also have our weaknesses, if we’re honest with ourselves. What we saw is that [there are new technologies] that can come together to a point that we could take the place of a pilot if needed in an emergency situation. That’s what really inspired us to get to this point.”

Here’s how the Emergency Autoland system works: A person presses a dedicated button on the company’s G3000 flight deck controller to start things off. Once activated, this dedicated button triggers a search of suitable airports and runways, scouring the available landing spots for things like desired runway length and width, presence of airport towers, and the like. It balances these criteria with data such as distance, remaining fuel, and wind conditions. It then clears the flight plan, enters a new one, and calculates the best possible route to the destination, which involves working out how best to avoid dangerous terrain or adverse weather patterns. Once it is within range, the Autoland system then initiates the right approach and, finally, automatically lands the aircraft without any pilot or passenger intervention needed.

The system forms part of what Garmin calls Autonomí, a “family of solutions that adds automated, safety-enhancing tools for the aircraft, pilot and passengers.” Emergency Autoland is currently available on several smaller aircraft, including the Piper M600, Cirrus Vision Jet, and Daher TBM 940. Others will be available in the near future.

Planes without (human) pilots?

This is incredibly impressive stuff. It showcases just how far today’s autopilot systems have advanced. But it also hints at the really exciting possibility; the bit that goes far beyond potentially tragic edge cases. What Garmin’s Emergency Autoland suggests is that we are not too far away from fully automated airplanes — potentially even sans pilots.

Phil Straub stresses that this is not the purpose of the Autoland project. This is a safety net for human pilots; pilots, perhaps, like Straub himself, who is a keen amateur pilot himself. However, he concedes that there is certainly the possibility of fully automated flights in the future.

Garmin

“There is the desire to eventually be able to replace the pilot, or to not be dependent upon a pilot as we know it right now,” Straub said. “Pilots as we know them have fundamental airmanship skills, they know how to deal with crosswind landings and turbulence, and things like that. [But] those types of things are probably what automation can do very well.”

Garmin is far from the first company in history to think that there are certain aviation tasks machines might be better off at performing. Many of them have gained momentum in the past few decades. Ernest Hemingway once wrote of bankruptcy that it happens gradually, then suddenly. Technological change occurs in much the same way. A series of small changes accumulate, often behind the scenes. At some juncture, a tipping point occurs and everything changes. The “gradual” part of airplane automation has been taking place for years.

The glass cage of automation

As Nicholas Carr details in his 2014 book, The Glass Cage: Where Automation is Taking Us, the modern history of aviation has been a series of advances designed to take direct control of aircrafts’ out of pilots’ hands. Digital technologies have replaced the need for on-board navigators, radio operators, and flight engineers. So, too, have they transformed the cockpit into a futuristic command deck that wouldn’t look out of place on a space shuttle.

Beginning with the Airbus A320 series in the 1980s, direct mechanical control of aircraft has been superseded by digital fly-by-the-wire systems; installing algorithms and sensors in between the human control and machine response. Still other advances include ever-more-sophisticated autopilot systems and smarter assistive landing gear in larger aircraft. Fully automated flights are the next phase in this evolution — and probably not as far-fetched as you might initially think, either.

“The reality is that [our existing] automation can take off an airplane right now, climb it up to cruise altitude, do an instrument approach (read: The series of predetermined maneuvers prior to a landing), land, come to a stop, and even shut down the engine so that passengers can safely get out of an airplane,” Straub said.

For these aircraft to replace human pilots, however, isn’t so straightforward. Safety would need to be proven. Legislation would need to be argued and put in place. Perhaps more than anything, Straub said, we will need to rethink existing infrastructure for such technology to — no pun intended — take off.

The problem with infrastructure

“[Current infrastructure is] not designed around autonomous operation,” he said. “It’s designed around an air traffic controller verbally speaking to a pilot, providing voice instructions of what they should do, and the pilot responding. If you had an infrastructure designed around digital communications — and, by the way, that is evolving and coming — potentially, then you would have the ability to fully automate this operation.”

Autonomous planes are coming. Maybe not today, possibly not next year, but certainly shimmering somewhere on the horizon. After all, even challenging problems like air traffic control for robots are already being tackled by those working in the drone industry. For now, though, folks should probably be happy that we have systems like Emergency Autoland.

“I also say to our pilots, ‘remember, this isn’t for you. This is for your passengers,’” Straub said. “We put our family members on [planes that I fly]. I put business colleagues on airplanes that I fly. These are precious cargo. These are people that are very dependent upon that person in the front seat to get the airplane safely from point A to point B.”

In that light, techno-replacement has never looked so good.

Editors' Recommendations