Pilotless planes are on the horizon, but are we ready for them?

Uber recently announced its plan to offer a pilotless flight taxi service in 2023, though it still faces major hurdles. Boeing and Airbus are also among aircraft companies working on the new technology for commercial airliners.

"Ten or 15 years time, artificial intelligence will be better than human drivers and human pilots," Barrie Kirk, the executive director for Canadian Automated Vehicles Centre of Excellence, a consulting company for automated vehicles, told The Current's guest host Megan Williams.

The first fully automated airliners will likely carry freight, as "parcels don't care whether there's a human in the cockpit or not," he said.

But not all experts are as optimistic as Kirk. From debate about the relative safety of unmanned cockpits to concern about the technology's lift-off among passengers, the future of pilotless planes remains cloudy.

Kirk is confident unmanned planes will soon be safer than those with pilots, stressing that the majority of traffic collisions are a product of human error. 

He added that a 2015 joint study with the Conference Board of Canada predicted that driverless cars would prevent up to 80 per cent of collisions and traffic deaths.

However, the safety of driverless cars has come under scrutiny after a self-driving Uber SUV fatally struck a woman in Tempe, Ariz, in March. 

Sharing data

But Kirk says having a network of self-learning artificial intelligence that can exchange data with one another will improve flight safety.

"When you have a close call in a plane, which is driverless, you can get a lot of information about what did happen and what should have happened. You can share that information with everybody and all planes."

While sharing data can help reduce the number of accidents, in-flight recorders already function to that end, said John Cox, CEO of Safety Operating Systems, an aviation consulting firm.

Instead, Cox says removing the human element from data sharing between planes may actually reverse gains made in that area.

"[AI] doesn't take into account the vast differences in the airplane types. Pilots will find it easier to take the information from previous events and accidents and apply it properly," he told The Current. 

Cox believes a human's capacity to come up with creative solutions to unexpected events stands as "a margin of safety" that automated technology is unlikely to achieve in the near future.

He refers to 2016's Hudson River landing or the United Flight 232 crash-landing in Sioux City as examples of life-saving adaptive thinking on the part of pilots.

"Those were very unanticipated events that required a lot of skill and experience from the pilots. Artificial intelligence does not do nearly as well with that," said Cox. 

'Inherent scepticism'

But both Kirk and Cox recognize that an acceptance around the idea of a pilotless flights is a significant barrier to getting this technology off the ground. Still, Kirk remains optimistic.

"My grandfather was born in the 1900s in England. He never owned a telephone, and he was afraid if he did something wrong the phone would blow up ... It seems rather quaint nowadays."

"The 'extraordinary' very quickly becomes 'ordinary.'"

But Cox says the "outstanding success" of aviation technology will make a transition to a pilotless future particularly challenging.  

From 1903, "where we saw one major plane accident every four flights," to 2017, "where we flew 4.5 billion passengers on 45 million flights without a single fatality on a commercial scheduled jet. We've come such a long way. So, there is an inherent skepticism when you start talking about brand new technology."

Listen to the full conversation near the top of this post. 

This segment was produced by The Current's Danielle Carr and Nick Wapachee.

This article originally appeared in CBC Radio