The aircraft are already here. Pilot unions are preparing for battle. And the FAA is playing it cool. Autonomous flight is coming to civil aviation sooner than anyone thinks.
In January, Boeing CEO Dave Calhoun revealed an open secret in the world of aviation. “I think the future of autonomy is real for civil,” he told Bloomberg TV, before quickly offering some qualifiers. “It’s going to take time. Everyone’s got to build confidence. We need a certification process that we all have faith and believe in.”
The U.S. military has been flying autonomous planes for decades, of course, but always in a segregated airspace. Now it’s becoming increasingly clear that self-flying planes are coming to commercial aviation, and not in some distant Jetsons future world. Aircraft manufacturers are working toward it. Airlines are eager for it. The Federal Aviation Administration (FAA) is preparing for it. And pilot unions acknowledge the threat is looming on the horizon.
A decade ago, the conversation was largely speculative. But today, many in the aviation industry believe that small, self-flying planes could be carrying passengers by the end of this decade. Then, barring no major safety incidents, it could take as little as another decade before larger passenger jets operate without a pilot on the flight deck.
“It’s all about money,” says Dennis Tajer, a pilot for 35 years and the spokesman for Allied Pilots Association, which represents 15,000 American Airlines pilots. “Manufacturers are looking for the next innovative technology to deploy so that they can sell it and make money, and airlines are looking at how they can do this more cheaply.”
It’s a charge that’s difficult to rebut. Six years ago, a report from the Swiss bank UBS estimated that autonomous planes could save the air transportation industry more than $35 billion per year. Still, the same report flagged a bright red public perception problem. A 2017 global survey found that a majority of people would be unwilling to fly in a plane without a pilot, even if the airfare were cheaper. The next year, a public survey from Ipsos found that 81% of Americans would not be comfortable traveling on a self-flying plane. Notably, that survey was sponsored by the Air Lines Pilots Association (ALPA), whose 65,000 members make up the largest pilot union in the country.
The introduction of autonomous aircraft into the civil aviation mix will begin with small cargo planes, led by companies like Xwing, a Northern California-based startup. “We took an existing Cessna airframe,” says Xwing CEO Marc Piette, “which is the most widely used express cargo airframe, and we’ve been modifying that vehicle to convert it to a remotely-supervised vehicle. We think the cargo market is the best first place to deploy this. And we’ve been very deliberate.”
For the past few years, Xwing has been running automated test missions, mainly in California. A flight plan is submitted, just as if there were a human pilot, and the flight’s parameters are pre-programmed before takeoff. “It’s really a one-click thing,” Piette says. “You engage the system and it runs its mission.”
Until the technology is certified by the FAA, however, there will need to be a safety pilot on board. This allows Xwing to fly without jumping through regulatory hoops. “The safety pilot can disconnect a system and revert the aircraft to manual flying, but otherwise doesn’t do anything but monitor the system. It’s a very boring job,” Piette explains. Meanwhile, the Cessna is operated from the ground, with two human controllers watching a moving map on screens and interfacing with air traffic control.
“All of these companies are really looking forward to the day where there will not be a pilot on board.”
Once the technology is certified, Xwing plans to introduce and operate these vehicles by late 2025 and then make it available to other operators. “To give you an example, FedEx has about 240 Cessna 208s for its U.S. network,” Piette says, alluding to the scalability of the venture. He expects his autonomous aircraft to be transporting human passengers by the end of this decade.
Xwing is certainly not the only manufacturer working on autonomous cargo aircraft, but it has a secret weapon in its chief compliance and quality officer. Earl Lawrence knows a thing or two about FAA regulations, having recently left the agency, where he was the head of aircraft certification. Prior to that, Lawrence had started the FAA’s unmanned office. “One of the key things of bringing this category of aircraft to the cargo market is that we are not changing the rules. We are following the regulations,” he says, noting that some companies in the space have made proposals that do not comply with FAA regulations. “That’s what significantly slows things down.”
Few people know more about autonomous aircraft than Stephane Fymat, who heads up Urban Air Mobility (UAM) and Unmanned Aerial Systems (UAS) at Honeywell, which has a long history producing autopilot systems for Boeing Dreamliners, Gulfstream and Embraer jets as well as other aircraft.
Fymat shared with Forbes a PowerPoint deck for a speech he has given at invitation-only industry conferences. After a brief introduction, a slide appears with six images laid out in a grid. Each shows an aircraft produced by one of Honeywell’s clients. Some are designed for cargo and others for passengers. In one photo, men are loading boxes into a cargo plane made by Slovenian light aircraft manufacturer Pipistrel, which Rhode Island-based Textron acquired last year. The plane does not even have a cockpit.
“Basically, most of them will start with a pilot now and move to having no pilot on board,” Fymat says of Honeywell’s partners. “A few of them want to do it within four or five years and some think it’s more like a 10-year range.”
One immediate advantage of pilot-free small planes is the increased capacity. Autonomy can immediately turn a six-passenger plane into seven-seater. “All of these companies are really looking forward to the day where there will not be a pilot on board,” Fymat says. “They’re all planning for it, in fact, and we’re helping them get there.”
“Cargo is an entryway drug for them,” Tajer says. “The script is to start with something that doesn’t sound like it’s going to hurt people. But the reality is that it’s still the same sky, and it’s still a metal tube in the air and passenger jets will be sharing the sky with them.”
Readying for Take Off
How self-flying passenger jets will be airborne by the 2040s.
Various aircraft manufacturers are working with the FAA on certifying autonomous aircraft.
The first self-flying cargo planes will enter civil aviation, sharing the skies with piloted airplanes.
Small, self-flying planes will begin carrying passengers on short, regional flights.
Larger passenger jets will begin operating without a pilot on the flight.
For manufacturers, bringing new aircraft into the mix requires navigating the regulations of the FAA and the world’s other aviation authorities. “We have an application on an all-electric autonomous airplane that we refer to as Wisk,” Boeing’s Calhoun told Bloomberg. “That application is in and the FAA will begin working with us today on building out a certification program for autonomy and for Wisk. There’s so much to learn, so much to do, but it is being done.”
Some worry that the FAA may be too accommodating. “The FAA ebbs and flows between being a monolithic oversight authority to slowly, over time, becoming too cozy with the manufacturers and the airlines,” Tajer notes. “That’s a funding issue, it’s experience, and then it also can become a culture issue, which you don’t have to look too far back [to the 737 Max crashes], to see what happens when there’s too cozy a relationship. The FAA just needs to stand firm on delivering the safest possible system out there.”
During the pandemic, some airlines quietly lobbied the agency to amend what’s known as Part 121 of the Federal Aviation Regulations, which requires scheduled carriers to have at least two qualified pilots on the flight deck, arguing that a single-pilot approach would help alleviate the pilot shortage. The FAA’s European counterpart recently ruled against making such a change until 2030 at the earliest.
It’s a rationale echoed by manufacturers. “Look at the airliners that we fly in every day,” Fymat says. “They don’t yet take off by themselves. They don’t taxi by themselves. But once they’ve taken off, they’ll do the entire flight by themselves, and they will land by themselves if you wanted them to. Airplanes have been doing this for years.” He adds that it’s just a matter of time before jets can do it all. “Adding in the ability to divert and redo a flight plan because of an emergency or whatever, communicating with air traffic control, those are the next pieces. But the basics already there.”
Naturally, the idea of single-pilot flight decks doesn’t sit well with pilots. “They’re talking about taking out that backup human system,” Tajer says. “Having that second pilot can be the difference between people getting hurt or them getting through an incident safely. And that’s because no matter how much technology you have, the importance of having another human being who has as much at risk and is committed to protecting those passengers in the back is what makes our safety system so successful here in the U.S.”
Meanwhile, automation is advancing. “AI technology is a big opportunity to establish the next generation of products in the industry, the epicenter of really, really powerful core technologies,” says Arne Stoschek, the head of machine learning and autonomy at Acubed, an Airbus innovation center. At Toulouse Airport, near Airbus headquarters in southwestern France, the manufacturer is testing a suite of pilot-assistance technologies in a demonstrator it calls DragonFly. Much like how actual dragonflies can recognize landmarks while in flight, the aircraft’s systems use artificial intelligence and sensors on the outside of the plane to “see” features in the landscape and safely maneuver autonomously within its surroundings. To date, the capabilities include automated emergency diversion in cruise, automatic landing and taxi assistance.
In the near term, the greatest challenge for self-flying planes may be getting the public to accept them. Xwing’s Lawrence expects public perceptions to evolve with time. “As the technology gets rolled out, people will understand it better and trust it more,” he says.
Advocates point out that among the first beneficiaries will be people who live in rural locales far from major airports. “This isn’t just about getting somebody from downtown Wall Street to JFK in eight minutes flat, which is very valuable by the way,” Honeywell’s Fymat says. “It’s also about the rural community, who right now has no air service. Or the essential air service that they have is, like, once a week.”
Xwing’s Piette agrees. “The impact is going to be dramatic,” he says. “Among the reasons [regional airports] are losing commercial traffic today is that there’s a significant pilot shortage, which makes it very challenging to staff some of these routes. If you’re in places like Wichita or Louisville, good luck trying to find direct routes to wherever you want to go.” With a large volume of smaller aircraft, there can be a higher frequency of flights between any small-city destination pair, according to Piette. “And we can reinstate access to flights to small localities.”
Ultimately, Tajer argues, “it’s going to come down to whether people eventually get used to the idea. And I don’t know, but when you’re in the sky, and it’s noisy, and the aircraft is doing things you’ve never felt before, and nobody’s in the cockpit, you might be thinking, ‘I sure hope that kid down on the ground at the computer has got this stuff figured out, because my family’s lives depend on it.’ That’s a big hump to get over.”
This story was first published on forbes.com