Of all the tantalizing futuristic possibilities on the horizon, self driving cars is clearly the topic people love to talk about these days. But fully autonomous cars are going to take a long time to actualize.
Passenger aircraft could be a different story, with profound implications at least for small plane travel.
Not only do we have plenty of drones patrolling the sky without pilots, but commercial airplanes have been largely autonomous for a long time. The newest models of Boeing and Airbus Jets take no more than a few moments of actual live pilot intervention to fly each flight, and my understanding is even that isn’t really strictly necessary.
But there doesn’t seem like there is much pressure to go the last mile and fully take the pilots out of the cockpit.
I assume this is for two reasons. First, just as it took a generation for people to trust elevators that didn’t have manual operators, I bet that many or most of your average modern air-passengers wouldn’t be super psyched about getting on a plane without pilots.
Second, for big commercial airlines the cost of having a pilot or two along for the autonomous ride is a tiny fraction of the flight operation's overall cost. Unlike for autos, all the hassle of moving away from human pilots probably doesn’t really change the economics of the airline business meaningfully.
No Advantage
Perhaps the big airlines could do slightly better with delays if the planes were autonomous and pilots never needed to be moved and never "timed out" on legal flight hours. But we’re in a world where all the airlines follow each other making big changes, like introducing bag fees. That means there is no competitive advantage in dropping live pilots, especially if you still need live flight attendants (a role so fundamentally human it is unlikely to be automated in our lifetime).
All this math changes in my mind when you start talking about small private planes. Steps towards autonomous flying would dramatically change that business.
First, imagine a world where small planes still had a single pilot, but there was a ground-based backup human pilot that could take over in case of emergency. One of the worst parts about many small planes is that because people want to save money, they frequently don’t fly with co-pilots. Having a co-pilot service on the ground, monitoring the pilot’s vital signs and providing on-call backup, could dramatically improve safety.
One step forward from that, once you can fly small private planes in ground-command mode like a drone, you might as well drop the pilot in the airplane entirely. If central ground-based pilots were flying small craft remotely with good on-board autopilots, then in theory each primary pilot could simultaneously fly a great number of planes at the same time. They’d only have to focus on an individual plane on takeoff, landing and at other key moments. You could also do things like switch over to experts like Sully Sullenberger if you got into real trouble.
Puddle Jumpers
Dropping the cost of pilots might not matter for your Airbus, when the cost of the pilot is divided by 300 passengers. But if you are shuttling people around in a four-seat plane then the cost of the pilot is a significant part of the overall cost. And if you are puddle jumping around with a pilot on-board you need to manage flight time, pay for their return trip, etc. By putting all the pilots on the ground and allowing each to fly more planes simultaneously, the cost of hopping around would drop a lot. It would make flying more accessible to many.
The last step, just like with cars, is full autonomy. With cars, that will take a while. It is one thing to figure out how get a car to drive itself from point A-to-B on well mapped roads. It is quite another to get cars to navigate private driveways and turn around. Even if you get all that working, replacing the existing fleet of cars in the U.S. and figuring out how to fuel, maintain, and manage a fully autonomous fleet of autos is going to be quite a task.
Cars need to be fully autonomous for people to be able to share them effectively and thereby get the full benefits. But the transition to full autonomy isn’t that important for the small airplane world. For planes, we could use existing technology and add ground control, achieving most of the efficiency benefits we’d get from full autonomy. The last mile of getting airplanes to taxi autonomously probably isn’t worth the cost and complexity. You can just let a human park the plane.
The upshot? If people would trust the technology to work (which should be possible since it’s the technology used by drones today), I believe moving the piloting of small aircraft to the ground would have an immediate impact on use of small planes.
If the price of flying reflected just the fuel cost (which can be as low as $100 an hour) plus maintenance and some other fees, I suspect many more people would take advantage of the country’s air infrastructure.