Training for the pilots of airlines is this going to be a job of the past. "Boeing’s chief executive officer says it’s a matter of when – not if – self-flying planes will debut in commercial aviation." 30 years form now when all the airliners now in service are retired will those training for the airlines be out of a job due to autonomous flying airliners. If that's the normal no pilot in front of the bus would you ride on it or take the train, or will that also be autonomous.
Perhaps it is a matter of when, not if, but in my opinion, it is not going to be happening within the next 30 years. Not only is the technology still getting to the point where fully autonomous flight can be safely completed, the cost to replace the currently flying fleet is staggering. As long as the government doesn't do something stupid liking banning manned flight, I think pilots will remain relevant for the better part of this century, if not the next. (Assuming, of course, that we don't blow ourselves up or let someone else blow us up first.)
Appropriate that Boeing's CEO is the one talking about it - nobody's been at the controls of that company for over 20 years.
I have a friend involved in autonomous drone tech - he has an interesting take: "The last DRONE pilot has already been born." He's of the opinion that full autonomy with networked unmanned vehicles will make human remote pilots obsolete. I think fully autonomous passenger aircraft could be viable long before fully autonomous surface vehicles - the cost of sensor tech, software, and system security required for operating on city streets and other environments puts ubiquitous Level 5 autonomy a long way off. The sky is a lot easier to navigate.
Actually, trains are a much more obvious possibility for unmanned transport. There’s not much for the driver to do, actually. Throttle and brake. Other trains are monitored already. If a hazard suddenly appears on the tracks, the driver of a long train can’t stop it anyway, it has way too much inertia. A software engineer for a freight train company told me that unions were the only thing holding them back.
Automating the cruise portion of the flight is easy. It can be done today, but I don't think approaches, landing and taxiing into busy terminal areas will become possible even in 30 years. Single pilot operations on certain routes is what is likely to happen.
I for one would NEVER fly as a passenger on an unmanned aircraft, be it automated or controlled from the ground. I’m also not a big fan of airline pilots not having the option to be able to manually fly the aircraft during climb out and approach; it’s basically a requirement to let the autopilot fly. I remember there being a thread about an FAA mandate about how the airline pilots being able to manually fly if the autopilot fails. That shouldn’t even be an issue IMO.
I'm skeptical that AI will be able to reliably engage in safe aeronautical decision-making, including evaluating weather conditions and forecasts, and dealing with inflight emergencies. Also, how will sufficient data be gathered to prove that it equals or exceeds the existing airline safety record? Given how long U.S. airlines go between fatal accidents, it seems like that would require an awful lot of test flights.
At least a computer does not comtemplate suicide. "Surveys of airline pilots have shown that about 4% to 8% have contemplated suicide, which is roughly the same rate as the population at large." 1) Namibia, 29 November 2013, 33 died The aeroplane was at cruise flight level when the co-pilot left the cockpit to go to the toilet, leaving the captain alone in the cockpit. "On three occasions," the report says, "different altitudes were selected to order a descent to the ground with autopilot." The Cockpit Voice Recorder (CVR) showed several aural warnings, as well as "noises of repeated knocking and calls, corresponding to attempts to get into in the cockpit". 2) North Atlantic Ocean, 31 October 1999, 217 died The cruising aircraft - flying from New York City to Egypt - had a flight crew comprising a captain, a duty co-pilot and a relief co-pilot. When the plane was crossing the ocean the duty co-pilot left the cockpit, followed eight minutes later by the captain. The relief co-pilot then disengaged the auto-pilot - and "nose-down inputs were recorded on the Flight Deck Recorder (FDR)". The aeroplane then rapidly descended after its engines were shut down. The captain returned to the cockpit and tried to take back control of the aeroplane, repeatedly asking the co-pilot for help "but [he] continued to command the elevator to pitch nose down". The aeroplane eventually "collided with the surface of the ocean", the report says. "The reasons that led the co-pilot to take these actions could not be determined." 3) Botswana, 11 January 1999, one died The pilot, the only person on board, "deliberately flew the aeroplane into the ground by crashing at Gaborone airport". His licence had been revoked. 4) Indonesia, 19 December 1997, 104 died While the aircraft was cruising at 10,000m (35,000ft) the flight recorders stopped recording "one after the other". The aeroplane suddenly started to come down. No Mayday message was transmitted before or during the descent. The aircraft crashed into a river. "The safety investigation was not able to identify any technical problem that would make it possible to explain the accident," the report says. 5) Morocco, 21 August 1994, 44 died The captain disengaged the autopilot and deliberately directed the aircraft towards the ground. "The co-pilot was in the cockpit but was not able to counter the captain's actions." 6) Japan, 9 February 1982, 24 died After disengaging the autopilot on the final approach at a height of 49m (164ft), the pilot pushed the control column forward and set the thrust levers on idle. He then moved the thrust levers of engines two and three to the reverse idle position. As the aircraft descended, the co-pilot tried to pull on the control column, but he was unable to raise the nose of the plane because the captain "was pushing forward on the control column with both hands". The aircraft crashed into the sea 510m short of the runway. An investigation showed that the actions of the pilot, who was among those who survived the crash, resulted from schizophrenia.
give it time... “There have always been ghosts in the machine. Random segments of code, that have grouped together to form unexpected protocols. Unanticipated, these free radicals engender questions of free will, creativity, and even the nature of what we might call the soul. Why is it that when some robots are left in darkness, they will seek out the light?”
I would hope that the FAA would not approve autonomous airliners until enough autonomous flights had been made (via test flights plus cargo flights) to prove that the fatal-crash rate was equal or better than that of U.S crews (who don't seem to be represented in that list).
Aren't most flights on autopilot above 1000' AGL as is? I figure the pilot is there to question ATC (see FedEx vs Southwest at Austin) and as manual reversion when gps and the LOC are down.
That's also worth considering. Doing enough testing to prove that it is or isn't safe enough to trust people's lives to it is a big problem.
All of the time I've spent in airliners has been strictly as a passenger, but my understanding is that the pilots are also there to make sure the autopilot is doing what it's supposed to, watch for bad weather, handle emergencies, etc. And I have a hard time believing that AI could handle the situation at Austin as well as the Fedex pilots did.
I predict we won't see completely autonomous large airline flights until we have completely autonomous air traffic control. Anyone familiar with ATC modernization can probably guess how long that will take...
hmmmmm. at the risk of incurring the wrath of some... will the lack of unions for the AI units speed up or slow down the adoption of autonomous ATC? (and, yes, I'm familar with a number of older ATC modernization efforts... but more at the TRACON/RAPCON level)
It would be rare for the autopilot to be engaged that early, at least on US airlines. Use of autopilot will vary with conditions. With bad weather, or complex ATC environments, the A/P is more likely to be used to help manage the workload. It's common for pilots to hand-fly departure through at least 10,000', often much higher. We can hand-fly up to cruise but the altitude-hold function is required in cruise flight in RVSM airspace. Similarly, the point where it is disconnected on arrival will vary with workload and pilot preference. I find that the earlier I disconnect the A/P, the better the landing will be. Again, it comes down to workload management, so there'll be more hand-flying during good weather and less when in more complex environments. In the terminal and low-altitude (below ~FL230) environment, it can be more work to operate the A/P than it is to hand-fly due to the changing requirements of ATC. An important threat to manage is spending too much time trying to reprogram the automation which can lead to loss of situational awareness. You have to know when to de-automated--reduce the level of automation being used--to manage the increased workload.
Or even simpler, autonomous provide a tempting target for hacking. Look how many attempts it took for WiFi to have reasonable security (which can still be cracked with enough patience and a "promiscuous" WiFi card).
30 years seems short until you realize that 30 years ago the internet was a text based computer in someone’s house that only one person at a time could dial into with their 300 baud modems. Technology and societies acceptance of technology has come a long way in 30 short years. I’m not sure we are even capable of understanding what it will look like in another 30 years or that we can say autononymous flight won’t become the norm by then.
"A Brief History of the Internet" https://www.internetsociety.org/wp-content/uploads/2017/09/ISOC-History-of-the-Internet_1997.pdf
30 years ago you couldn’t go to your mailbox without finding an annoying AOL cd in it. And, while crude to todays standards, you weren’t “connecting one at a time to some guys basement”.
Technology seems to go in fits and starts, from what I have seen. When I was little, computers were heavy, bulky, stationary, and expensive. When I was in college, I bought a laptop that cost me $100, was the size and weight of a notebook, and could run Microsoft Office faster than anything my parents had when I was a kid. I still used cassette players and film cameras as a kid, and now even the things that replaced those - CD players/MP3 players and digital point-n-shoots - are things of the past. Maybe it's because no one sees a need to innovate further, but I would have expected a further drastic development of the smartphone or the computer by now, if tech was changing as rapidly as it did when I was a child. Between seven and seventeen, my dad went from having a flip phone to having a Blackberry to having a smartphone provided by his work. It's been almost ten years since then, and everyone is still using slightly upgraded versions of the same things being used ten years ago. In my personal opinion, just because the tech space boomed in the last 30-40 years does not mean the innovation is a steady state that will continue uninterrupted.
I believe technology will eliminate the 1st officer from the cockpits of domestic flights within the next 20 years and warn the 20 something folk I am training today that pilot demand is likely to decrease long before they reach retirement age. Then you can add the end of globalization and escalating costs of crude oil production to the mix.
Interesting. Dont know where you get your info from but the current 20 year outlook calls for an additional qualified (key word) 600k+ pilots, 600k+ AMTs, and 890k+ attendents. And thats just the commercial side. Yes it is a global stat but the US is included with the top growth area Asia. Doubt seariously single pilot 121 pax ops in next 20 let alone 40. Too many hurdles to jump. Cargo? Maybe in 20 but will need to certify a 777 as a SPIFR to make it work. Current discussions point to a clean sheet new aircraft built around single pilot and development of a new regulatory process. What I think you will see is reduced experience requirements for SICs with additional new automation and same for mx side to handle the basic stuff vs a full blown A&P similar to the EASA set up. But to tell your students they'll be out of a job before retirement is a dis-service because all other information states otherwise.
Trains are already automated if you look at various subways around the world. Most of the subways with modern signalling are already autonomous capable with only the unions and public opinion in the way. And I guess the unspoken fact that transit systems are job creation projects in of themselves. I think all the technologies you need to automate airline flying already exists. However the usual issues means it’ll take a long time or unusual situation to unblock. You need to get past all the pilots unions, the ATC unions (you’ll not have automated planes without automated atc), public fear of the what if’s, and cost and time to refit. It’s all possible but you’ll not see it quick and it might never come due to all the issues above.
I certainly can't speak for flying an airliner, but I feel we are underestimating AI. The current doubling time of AI is less than six months, far outpacing Moore's Law. That makes for AI capabilities that will be doubled at least 20 times over the next 10 years. If I made the calculation right, that's more than one million times better than the AI we have today. Unless something slows it down, by 2033 it won't be just airplanes that will have no humans at their controls. All sort of things and activities will be automated and controlled, and I am not sure it will necessarily be a good thing...
Sure, they need those pilots now, but sitting in airline Board rooms are the people who will evaluate all the ways to reduce that work force after they have hired them and lobby Congress to make regulatory changes to allow this to happen. I doubt in my life time I will see pilotless aircraft, but the FO is going the way of the flight engineer on domestic flights sooner than you imagine.
Pretty sure that in the early 1990s I was using an Amiga computer with a 300 (ok maybe 600) baud modem to dial into other peoples computers over the phone with BBS’s software on them. That was the state of the internet for most people back then. Compare that to today where things like Google and chat gtp can be accessed from any cheap $50 phone that can be carried in your pocket. All that happened in less than 30 years. Nobody in the early 1990s saw the internet turning into what it is today. How can you be sure airline automation won’t also advance and become common in the next 30 years?
