SPO (Single Pilot Operation) sounds like a good idea in theory, but in practice, it’s terrible. In an attempt to save money, the cost ends up goes up, and that cost is human lives, system failure, human fatigue, eventually it’ll add up to a situation that technology nor pilot couldn’t correct, resulting in an accident that’ll cost lives. As the old saying goes, “What goes around, comes around.” Captain Joe is right, there should always be two pilots in the cockpit, at all times.
Moreover, the more technology will replace human operators, the more we're going toward a total human unemployment leading to human starvation and extinction. Am I too pessimistic? I hope so, but I'm afraid I'm not
@@vannizaniboni3502That’s not what automation or increased efficiency has ever resulted in, from the invention of fire to the advent of robotic fabrication. Rather more automation results in more human prosperity and more people in the fields of innovation, where humans excel.
There's hundreds of smaller planes that operate single pilot, and there's no issues. There's never been a recorded incident of a single-pilot accident on a passenger-rated plane due to pilot incapacitation. The medical requirements to fly are very strict.
If an airline starts a "no pilot in the cockpit" or even a SPO (Single Pilot Operation), I will avoid that airline like the plague. As much I like technology, it can not replace having two pilots in the cockpit. Superb video as always.
Perfect example! If mankind needs 2 humans to control a vehicle in 1 dimension space, how will it control a vehicle in 3 dimension space, plus the limitation that the vehicles can not stop when airborne?
In some states in Europe this is not anymore case. In 'local' passengers trains the conductor is 'checking' that the train engineer is still 'alive' Which means that ticket checking is reduced... This is permitted where there is a dead-man switch mechanism (if I am not mistaken)
I am a trained computer engineer with a masters degree in automation and control + I also trained as a commercial pilot. Recently while working with AI and testing its code-generation ability, I provided the exact same request ( prompt ) and I got four different responses from the AI and the generated code was useful only in one of the cases! Furthermore if people >KNEW< exactly how the autopilots, various control systems and AI work on an airplane ( specially the AI ) they would not want to fly in an aluminum tube at near the speed of sound and 11km altitude possibly over a large body of water controlled only by some AI model. Invest the money and time intended to train and develop reliable AI ( that can match a good pilot's ability ) instead on selecting people with the correct aptitude and attitude ( to fly airplanes ) and then train and support them well in/on their job. Only non-technical people or those who haven't operated planes in serious weather would make such a suggestion. And YES the pilot's salaries are peanuts compared to the other airline operating costs. I am not claiming that pilots are flawless, there quite a few operators with questionable attitudes and abilities that cause problems ....agreed....BUT eliminating the human element is not the answer. We need the human in the automation loop ( Boeing's approach by the way ), apart from the question of whether the flying public would accept this. Companies like Airbus would of course have the incentive to push for this...so it really boils down to pilots-vs-engineers.
Where did use of AI in the cockpit come from? As you correctly pointed out, it's not reliable but software written and proven via formal methods is what's needed to operate an autonomous aircraft. Captain Joe mentions AI to garner commentary like yours knowing it supports his narative.
@@davidpalmer9780 One cannot write formal actions that respond to infinitely unknown future situations. Humans can CHOOSE their actions, programs can't, they are programmed.
@@aeomaster32 In software development, formal methods are mathematical approaches to solving software (and hardware) problems at the requirements, specification, and design levels. Formal methods are most likely to be applied to safety-critical or security-critical software and systems, such as avionics software. Formal Methods - Wikipedia
@@davidpalmer9780 I wrote SW for the ISS and the Shuttle...somewhat familiar with safety-critical SW architectures and their requirements. Formal methods are great as long as you can define precisely what safe piloting of an aircraft in flight encompasses ( in such a way that it can be couched unambiguously in mathematical formalism ). Anything you can describe precisely might be turned into a system of interacting algorithms to produce desired side-effects ( namely control of some aspect of a physical system according to some predefined outcome ). Alas mathematics is an abstraction of reality and many control loops are actually only effective around linearized regions of their respective operating points. Dissipative and other non-linear effects quickly take us into the realm of nonlinear dynamics which often also displays chaotic behaviours. Well trained humans are great at handling these control tasks which we cannot yet formalize mathematically. Then there is the fusion of experience and sensory inputs into judgment and action. You may have experienced riding jump-seat ( or left or right seat ) in the cockpit of a turbo-prop with PAX following a step-down procedure into an airfield in mountain valley at night, down to minimums and with moderate turbulence and icing....and that without any issues and failures ...in other words everything on the aircraft is churning and burning as it should....Human experience and judgment perform better in a variety of circumstances which are hard to mathematically capture, formalize and specify a priori. Pilot-in-the loop automation works best and we need more of that.
[I am trained in automation and I have programmed industrial computers] At the same time, while watching this and other channels, I wonder why the logic implemented in the computers of a plane seems not to be fully consistent... I am not talking of AI. I hope that by investigating single-pilot or no-pilot flying, all those missing if-cases are understood and corrected. BTW read ''humble pi" on the subject of human induced errors (on programs running on computers or mistaken screws).
I’m just gonna say it; no to both. Firstly, you’d have to get the aircraft certificated for Single Pilot operations not only by EASA but also TCCA and the FAA which is notoriously difficult to do or isn’t allowed by law for aircraft over specific MTOW or PAX. Secondly, people don’t seem to understand that in order to fly high UAVs beyond line of sight requires Ku band level receivers and transmitters, which will basically microwave you if you’re within 25 feet (if I recall correctly. Might be larger area) which means you’re looking at cargo only ops. Finally, even if you’re flying a UAV line of sight, you will have a Go-Around min altitude because by the time the signal goes to the CPU then to the FADEC you’re already on the ground. RQ-4s are notorious for those kinds of incidents mostly because ATC is rarely briefed on their min altitude for Go Arounds.
This is assuming that a fully automated system performs less safely than an human, which, I believe, will very be proven wrong in the not so far future
@@mrcat5508I think I don't agree. I just asked chatgpt "how to handle a broken pitot tube", and it looks to me that despite this AI being not at all specialised in plane piloting, it has a pretty accurate idea of what to do.
1. Like all essential systems, two pilots are a necessary safety redundancy in the face of emergencies and failures. They're also uniquely organic (unaffected by software errors, not interfaced with computers). The stakes carrying hundreds of passengers are much higher than a single-seater fighter jet with a seat ejection system. 2. Multiple crew system reduces fatigue and loneliness (not a joke; working in a team tends to make a person not disregard proper procedures), and helps prevent pilot s**cide plane crashes. 3. A two-pilot system makes an excellent training environment for the junior pilot, as well as providing a good opportunity for the senior pilot to develop leadership skills.
Sully started the RAT b4 the checklist and its on his testimony. there was no dual engine failure at low altitude in the manual. AI works upon learning models, as long as we are willing to have "Disaster in the Hudson" instead of "Miracle on the Hudson" and actuaries determine that payout is worth the savings, it willl come. Similar to outsourcing MCAS software dev to someone who had never seen a pilots seat in any aircraft.
Isn't that a bit of a survivor bias? You chose a flight where the situation was indeed saved by the pilot, but how many planes crashed due to human error from their pilots?
@@casilasgoaler If you had any idea about my actual background in Mathematics and Statistics, you'd understand how funny your comment is to me. 😂🤣Thank you for the laugh. I do understand how many planes are successfully flown everyday - it's part of my job actually. I was mentioning the number of catastrophic situations created by a human error by the pilots, versus the number of said catastrophic situations saved by the pilots. Have a great day buddy, love you.
Well, if you go in that direction. Now imagine it differently. The A320 is designed in the 1980s so yes, perhaps the aircraft didn't notice the engine to start stalling. But technology has advanced, a lot. Now imagine this same situation with a 2030s designed aircraft. And instead of having 'the miracle on the hudson' we have a basically unknown news article stating 'aircraft lost both engines after take-off and safely landed at 'initial' airport'. It was proven the aircraft could land but if action was taken immediately, something humans simply can't.
@@fnaeem77 The RAT starts automatically with complete A C Electric lost , he started the APU by hard , to get normal electric from the APU Generator , very good !
Excellent video Joe! My career has been in the railway industry, where automation has made a driver obsolete, with appropriate safety levels, since the 1970s. Why? Because trains only have 2 degrees of freedom, and they are on the ground. When the automation detects a potential unsafe condition, it simply stops the train. Unfortunately, airline/aircraft corporate executives and accountants, not skilled in the technical aspects, believe that the same logic can be applied to aircraft, which have 3 degrees of freedom, and you can't just stop the plane when something goes wrong. Joe briefly mentioned public acceptance of OPO and NPO. That has been the biggest hurdle for trains. Many rail transit authorities around the world keep a person on board, not doing much other than opening and closing doors, just to please the public. The "driver" you see sitting in the cab is not driving the train; there only to make you, the riding public, feel better. We're just getting to the point where autonomous cars are starting to become accepted by some members of the public. It will be a long time before the general public accepts flying with one or no pilot.
@@capitanshaf5751unless we realize that its the current economic model that promotes such profit-oriented suggestions in the first place... And strive to change it to something better, for our children's sake.
That would be an economic disaster.Nobody would want to fly in a large aircraft anymore. Do they really think we are that stupid ?And is there really a responsible pilot who would go along with that ? Could we still trust such a pilot ? I trust Captain Joe !
Dont be silly i would trust a computer over a person any day. If a computer has an accident it will be fixed and that accident can never happen again. People keep making the same mistakes.
@@womble321Like they have fixed the lane keeping in my car to properly recognise diverging lane arrows and not interpret them ase leaving the lane and trying to yank the steering wheel to "correct". You forget, that fixing an error like that takes time and money and companies will do all they can to avoid spending it. Look many of the design issues in aircrafts that don't get redone for years as the manufacturer argue over who's fault it is.
@@womble321so we will wait for accidents to happen first till we fix all software bugs. Sure thing Sherlock. Am amazed at your stupidity, such kind is rare
@@womble321 Don't. Basic English. Have you ever used a computer? 😂 Tgey have problems all the time and, at times, the "fix" causes more problems. You would not believe how many errors softwares have and are considered fully functional.
@@womble321 Yeah that's not how it works. When a computer fails you can't just "fix it" and then it's fine. A computer failure is not always due to a design flaw. When an airbus has failures, it literally shuts off automation because it can't operate accurately based off faulty data. That's what a computer is. It takes an input and gives an output. It doesn't have a brain. It has no decision-making ability. Anything that is unfamiliar to it will cause it to just shut off or be a sitting duck. You're so worried about the pilots yet there hasn't been a fatal airline crash in the United States in 15 years. You say people keep making the same mistakes when the industry quite literally changes regulations in training and operations for that exact reason, and it obviously has been working.
Agree 💯 Joe. My gut said a big NO when this idea was first mooted, and you have just articulated the details of my gut feeling in this video. Having watched a huge amount of aviation accident/incident videos where the cockpit crew performed unbelievably under the most challenging of situations, it is incomprehensible that AI could even reach that level of finesse. Your statement about technological advances being an amazing tool are spot-on. That brief shot of a pilot alone in a two-seater airline cockpit made me feel a depressing sense of aloneness - and I'm glad that was one of your points raised. Will be a very sad, and retrogressive path aviation takes if this becomes a reality. Thanks for a great video Joe.
Get it, you can automate flying with just the captain. What if he gets food poisoning, or simply wants to take a dump and there’s a decompression in the cabin. Who’s gonna make the decision to land, will the captain be able to clean his mess fast and understand the situation in 5 seconds all while he’s in a decompressed washroom of 2x2ft size smelling like rotten bacon and eggs???
The computers will detect the decompression before it occurs and commence the descent whist simultaneously determining where to divert or to continue to its destination, alert the airline, alert ATC and all whist the Captain is relaxed taking a dump.
@@davidpalmer9780 Cool story, so what happens if its a system failure that caused the decompression, and the computer just got the information that its now 10,000 ft lower than it actually is, and the airspeed is 50 knots faster than it should be? I'll answer it for you. A fatal crash that the pilot would never have time to correct, if he was off taking a dump, and yes system failures like this HAVE happened in the past, and can happen again. You need a human at the controls who can see what's happening and disconnect the auto pilot, and return the plane to safe flying status. People always talk about the "fully automated drones" but there are things people don't get about those.. 1) they are always connected to satellites, and their home base. If something goes wrong, they can be manually controlled. There's your human intervention again even with those. 2) Those bandwidths required to control those things would microwave any passengers 3) A commercial Airliner is an entirely different beast when it comes to flying parameters, systems, and flight controls than a drone. Vastly more complicated.
I was going to make a rather long winded comment here, but I think you hit most of the points I would make, so I will stick to the high points...Pilot incapacitation is a real thing, and perhaps more common than we'd like to believe. My recommended video list is currently showing several recorded ATC conversations involving an incapacitated pilot. Incapacitation doesn't always have to be severe. I case of the flu, or even food poisoning can render a pilot unfit, particularly if they have nobody to assist them or notice their mistakes.....nor can you fly a plane from the lavatory. Further, there have been several cases of pilot suicide. While having a second pilot hasn't prevented all of these tragic incidents, it has prevented some, and, in fact, we'll never know how many incidents were prevented by having another pilot handy. Finally, computers have always impressed me with their ability to take orders, I'm not as convinced with their ability to creatively give them. Most emergency situations are at least somewhat unique, and computers have never been particularly good at dealing with situations they weren't programmed for. Fortunately, major emergencies are rare, but when they occur, I doubt you'll find a pilot who isn't glad there are two sets of hands in the cockpit, and in certain cases, happy to have three or four. I'm sure the late Capt. Al Haynes would agree.
The pilots on UA232 couldn't control the phugoid path the aircraftbwas taking with their thrust adjustments. An autonomous autothrottle system would have predicted and reacted faster to flatten out the phugoid and coupled to an autopilot, would have made a successful landing at Souix City.
@@davidpalmer9780 Computers may react faster, but they can only accomplish things they are programmed to do, in advance. One in a million type failures are probably not on that list. Even now, while automation can assist during some emergencies, if things get too far out of whack, the computer gives up and puts the human back in control.
As a 30 year airline pilot let me tell you how Single Pilot Operations (SPO) will play out - the first place we will begin to see this is with cargo carriers because nobody cares if a plane load of Amazon boxes crashes. After a few years of this and very gradually we will see SPO slowly move into regional flying by way offering lower fares. Remember consumers vote with their feet and this move will largely be driven by consumers favoring the lower prices that SPO airlines will be able to offer. Just like how right now it is not cost effective for North American airlines to run four engine airplanes soon too it will not be cost effective for them to run two pilot cockpits.
I agree, SPO is only a matter of time. Passengers may feel uncomfortable for a while and opt for flights with two pilots, but then the cost of those tickets will get out of reach for most. However, fully automated flights, with no humans in the cockpit, will likely be a far harder sell and might not appeal to passengers in the foreseeable future. Fully automated cargo flights could become commonplace, though.
@@KukosEQ and still, the number of pilots in the cockpit has only gone down. It's not just the salary, it's also the training, hiring, evaluating, etc etc that comes when you have human pilots. Training pilots and then keeping their training up to date is quite costly to the airline, I would assume, so halving those costs must seem attractive. An AI doesn't need training (once you have one), doesn't sleep, eat or go on strike, it just works.
@@savagecub Cargo ?? Maybe it’s new to me ,if you kill 1 or 2 , it doesn’t matter ? The airlines save theoretical 2-4 % with out pilots , minus the huge infrastructure to make that happen , give 2-4 % to the good trained pilots or 30-50 % to the developer ??
1. Electronic warfare. A remote-controlled plane can be taken down without any missile. 2. Pilot fatigue. A Pilot alone will be much more fatigued than two pilots. 3. Emergencies. Emergencies bring a high workload. One pilot flies, the other one trouble-shoots. 4. Judgment redundancy. 4 eyes spot more than 2 eyes, a second brain can identify the mistakes of the first brain. Obviously, Garmin hat already implementes the autonomous emergency return. But that works only with an intact plane. No-pilot is ruled out. I can imagine single-pilot only for flights less than 1h, with a full cockpit data link to a set of emergency pilots and engineers on full standby.
You think? The aircraft will be autonomous without the support you mention. Supported by cargo operations, the autonomous system will be proven the same way jet engine reliability was proven which allowed going from 4 to 3 and now 2 engine operations across the Atlantic and Pacific oceans.
@@davidpalmer9780comparing 4 engines to 2 engines with autonomous aircrafts is a really wild comparison man! Proves you have zero experience in the Aviation industry or whatever you have is pretty useless. The reason 4 engine was changed to 2 because the engines became more efficient and less costly. Nothing to do with "proven this proven that" lol.
Define A.I. . If we define it as not self aware, intsruction following program, then it can already land an airplane if the conditions are right, example: ILS and GLS autoland features. If you mean pseudo self-aware software, capable of critical decision making and learning on the go, by A.I., then it is also possible, the technology is already here. Earlier this year or maybe last year the USAF's (I think) experimenting/developing division successfully completed a test of an unmanned, autonomous software equipped flight with a modified F-16 -I think-. The sofware also beat the pilot in dogfight. So the tech is here. Is it safe? Depends on the circumstances. Military and civil sectors and scope of usage are different.
I am in a flying club and also rent from time to time, so I fly several planes. Every time I get in the cockpit, I always make sure to know exactly where the autopilot and electric trim circuit breakers are just in case the those oh so helpful systems suddenly disagree with my intentions to have a safe flight.
... and I'm sure that autonomous flight software would do the same. Well done for doing this check as others who don't have lost their and others lives too.
Airlines: No more pilot! Swiss-Cheese model: Yes. A well trained crew backed up by autonomous monitoring flying the aircraft is the safest system practically achievable. Aircraft flown by computer servers and world engines is the literal stuff of dystopian movies, videogames, and nightmares.
When sh*t hits the fan, any plea by a distressed pilot to automation for assistance will be met with something along the lines of the famous quote by HAL 9000 in 2001 "I'm sorry, Dave. I'm afraid I can't do that.". On a side note, these CEO's that are pushing the no-pilot-in-the-cockpit idea, would they want their private jets to fly as such with their a$$ in them?
I think one of the most important factors is that pilots are continuously learning from each other. Young first officers learn from experienced captains, but even captains can learn from first officers too. How would that be possible in single pilot ops?
This subject crosses my mind as a student pilot with experience in A.I. @Captain Joe covers the core areas of concern really well here. I think single pilot ops is a scary prospect
I'm sure you realise that Captain Joe has a very biased viewpoint in this video as he wants to preserve the role of airline pilots. He neglected to show that radio operators were also carried on early airliners too and later removed with the navigator and clight engineer roles. Cargo aircraft will prove autonomous aircraft operations before being permitted for passenger transportation.
I totally agree with you, Captain Joe. Here in the southwest United States, I live in Phoenix, Arizona, there is a driverless taxi cab called WAYMO. Even Young college students want nothing to do with a driverless automobile! The problem with technology is, that something can always go wrong. Not only do I feel much safer, but it just plain is much safer to have a human behind the controls, especially when that human has a lot of experience operating the vehicle.
Airbus is really pressing their luck with this, “Single Pilot Operation”. What if in the future there’s a single pilot (on a shorter flight), and that pilot becomes incapacitated (like suffering a heart attack), then that flight is SOL. There should always be an emergency pilot on board just in case.
I vividly remember when I was in school for aircraft around 2000 era, all the talk was about how pilots in military and civil aircraft would become a thing of the past very soon. However, people feel alot more safe with an actual person flying and that's why it never happened from what I could always see.
The day they eliminate pilots I'm not flying ever again. I work in medical devices (robotic surgery) and there's no freaking way I would EVER trust a system without redundancy.
Of course there will be multiple redundancies built-in, just like there are 3 navigation systems in most airliners today, there will redundant, autonomous systems.
While almost all of your rant's arguments may be perfectly on point, I miss your reflection on an even greater timescale. Just about the same arguments were brought forward by yesterday's seasoned pilots when removing the *navigator* ; again decades later when removing the *flight* *engineer* . Today, you are that seasoned pilot, and I admire your knowledge, your dedication and your top-notch level of professionalism. Yet I didn't hear a single word about you wanting the flight engineer back. I'd be more than happy to learn about your thoughts on that.
This is a great point. As it stands today having a fully automated plane is not really possible. But it’s hard to know what will be possible in 10+ years. It will never be flawless but neither are human pilots. The continuance of CFITs and the increase in runway overruns show that this is certainly the case and humans flying a perfect functioning plane are still liable to crashing. Ultimately there are going to be situations where an autopilot will cause an accident that a human pilot would likely have avoided. And instances where an autopilot prevents an accident a human pilot likely would not have. If we get to the point where the latter happens more then the former self flying planes will be a net positive even if we haven’t quite perfected it
Don't forget the radio operator who also was no longer needed on the flight deck. Autonomous aircraft operations will be proven the same way the move from 4 engine operations to 3 engines and now to 2 engines when crossing vast oceans. It's the proven reliability of the engines that allowed this to happen. Autonomous cargo flight operations will prove the technology over the years, just like engine reliability did and will pave the way for autonomous passenger operations in the future.
So the only reason airbus wants to keep 1 pilot around is because if something goes wrong they have someone to blame. "PILOT ERROR" It's a huge risk for aircraft manufacturers to develop and sell fully autonomous aircraft because if/when they go down(and they will) , the lawsuits alone will shut them down for good as they bleed money.
...and still they continue to develop their autonomous flight programs with Airbus and Dragonfly and Boeing and their aquisition of Aurora Flight Sciences who specialise in autonomous flight operations.
Totally agree with your points, and I'd like to add one: latency. I work as a networking engineer and a commercial geostationary satelitte link can have, in some case, up to 0.8sec of delay, not sure about military tech latency. Doesn't seem much, but I'm quite sure a ~1sec delay reaction from a remote pilot on ground can make the difference in a critical situation.
The great thing is, once they get rid of the second pilot in the cockpit, they will be able to spare some more by removing dual controls and right-seat instruments : what a breakthrough !
IMHO one of the drivers for SPO is not as much cost reduction as it is about pilot shortage. The increasing need for pilots is obvious as many more are getting out of the profession than are entering while more and more planes are operating on more and more connections. SPO will not halve the need of pilots, but the mismatch between demand for pilots and available pilots is lessened. But all of your arguments against SPO or No Pilot scenarios are in my mind absolutely right and SPO is also incredibly shortsighted as SPO is basically requiring an all Captain rank pilot staff - How are younger pilots going to be climbing up the ladder? SPO will may be in short term fix the shortage issue, but I fear will ultimately doom the industry
I'm on flight school and I don't know where the shortage is happening cause flight instructors can't get a job, if you don't want to instruct you won't fly, but because instructors are not getting jobs, instructing is not an option anymore since the current experienced instructors are not getting into the airlines. This is a profit tactic, if pilot shortage was the issues there are many other (and better) ways to deal with it
Just don't have pilots. There are more savings to airline operators that just salaries as Captain Joe mentioned in this heavily biased video. Savings in transport (dead heading) , layover accommodation, training, simulator operations, pilot schedulling and more.
The transition from 2 to 1 pilot is the exact same technological change from 1 to 0 pilots, because the aircraft will always need to be able to fly the aircraft in the event the 1 pilot becomes incapacitated. The aircraft becomes the backup system to ensure there is no single point of failure (pilot). This is why reducing crews to 1 pilot is something we should fight tooth and nail.
I think that there is probably a generational difference in how this is viewed. As a boomer, I say, no way (I work in technology and with AI and know how problematic it can be, who wants the AI autopilot to hallucinate ). My youngest son however inherently trusts computers over humans. He says things like “they don’t get tired”, “they never miss a reading “… etc. even in the scenarios of handling the unexpected, computers are not subject to the startle factor and begin making corrections long before a human could. He and I often debate this (like I am certain many reading this will want to). My point isn’t who is right, but rather, different generations likely view this very differently (as a generalization).
Shows how young he is. Blindly trusting technology is wrong. Using technology as a tool is perfect exactly for the reasons you mentioned (no tiredness, precision, etc.).
It's Captain Joe mentioning the use of AI, not Airbus or Boeing or others as it's not reliable. Autonomous flight software proven by formal methods will be used and it will be used in cargo operations first to become trustworthy for passenger operations.
You said it, Joe! This one word: Germanwings. Makes the whole thing a terrible idea. Old joke: Next step is One pilot and one Dog. Why a dog? to byte the pilot if he tries to touch anything, and the pilot? There to feed the dog. I'm not setting foot on any airliner with less than 2 flight crew in the flight-deck.
It's about when things go wrong, not when things go right. A well trained human brain is no match for a computer. It's also the reason pilots should be well paid - you're paying them for when things go wrong, not for all the 'easy/boring' bits. Managers have NFI in this regard.
I am a tech student who was given course work about MCAS. a few things I learnt from the event is that, automation may not be a solution to everything. When you look deep in to that very system, the lack of proper knowledge on how to over ride it, led to loss of lives. Now imagine a malfunction of any of the system to be introduced in the ‘SPO’ aircrafts. Will there be chances to over ride or even make the machine to divert from its operation programming?
Did you know some airlines didn't choose redundancy for the number of sensors that delivered data to the MCAS and when that sensor failed it caused the problem you mentioned.
@ well that is true. You can right me. But weren’t those safety features made optional by the manufacturers? Hope such options don’t come up in the ‘SPO’ crafts
@@ssemandayahayakyeyune4294 Yes... redundant sensors are an option but with choice of a single sensor by those airlines comes their responsibility to train their pilots on the failure of such a sensor too.
1979 - 1980 war ich vor meinem Studium Purser bei Lufthansa. Auf der 707 wurden die Navigatoren eingespart, die 747 wurde gerade eingeführt und auf meinen Mustern hatte die 727 noch den Flugingenieur, aber die kleine „Bobby Boeing“ flog mit zwei Piloten. Da kam dann beim fliegenden Personal der folgende Witz auf: Lufthansa spart jetzt alle Piloten ein, ein Computer fliegt den Flieger alleine. Nach dem Start kommt die Ansage an die Passagiere: „Guten Tag auf unserem modernen Flugzeug ohne Piloten, ich bin Ihr Bordcomputer und werde Sie schnell und sicher ans Ziel bringen. Sie brauchen sich keine Sorge zu machen, ich bin so programmiert worden, dass ich keine Fehler machen kann……….. Keine Fehler machen kann……… Keine Fehler machen kann…….“
Is not only about safety if you ask me. Rich ceos are already rich, big airlines are already big, world is getting more connected than ever. When is gonna be enough? I mean, work gives proposit, goal, dreams; whats the point of making everything meaningless and unalive?
No country's legal system should approve the replacement of the Pilots with AI or even the Single Pilot operation. Its simply too dangerous !! Like you very rightly said Joe, the extra presence of the Co Pilot is absolutely Critical in the Cockpit, for a multitude of reasons ! Human safety, always 1st !
Yeah... this was said for the radio operator, navigator, flight engineer and will happen to the pilots. It's the future. Also, wasn't an Air France Airbus was brought down with passengers and crew through pilot incompetence?
Why do we have 2 pilots in the flight deck to begin with? For 2 key reasons. One of them being for the purposes of aeronautical decision making. You might not believe this but in most operations the captain has discussions or briefings as we call it with their first officer and vice versa to discuss some potential risks to a flight and what could be done to mitigate them. One pilot can’t do it alone. It takes a team to come up with a game plan for the flight and how to conduct it safely. But the main reason is in the event of emergencies. Take US Airways 1549. You think Captain Sully did that on his own? It was a team effort with both him flying the plane safely down to the river while First Officer Jeff Skiles was completing necessary checklists and other tasks to assist Sully and landing in the Hudson. Take FO Jeff Skiles out of the equation, Sully would have been so task saturated flying the plane plus the tasks the Jeff Skiles was completing and I don’t think Sully would have been able to accomplish what he did that day. Airlines need to stop being greedy. As much as I love flying and aviation as a whole, if airlines are going to get greedy and pull one or both pilots from the flight deck, I propose a boycott of every airline that goes through with this. One less pilot means less safety in the air.
My 2 cents, as an aeronautical engineer with experience on safety systems. The problem is not if technically SPO is feasible: it is, or will be shortly. The problem is: who is responsible when something bad will happen, e.g. crash due to pilot incapacitation and the protection system unable to cope with the situation. Who will pay? The manufacturer? The airline? All of us (via the insurances increasing their fees?)
No difference to the ETOPS changes that permitted going from 4 engin operations across vast oceans to 3 engines and now 2 engines. Not a peep from anyone about that. The technology will be proven in cargo operations.
@davidpalmer9780 ETOPS and SPO are totally different beasts. SPO is maybe barely comparable to single-engine commercial airplanes (of which there is only one in production, and used only for charter or EMT services)
@@luigitenneriello74 You misunderstood what I'm saying. My mention of ETOPS is analogous to the proving method required for proving autonomous flight operations given the jet engine proved itself over a long time to be quite reliable with quantitative statistics maintained of in-flight shutdowns. The increased very high reliability of those jet engines over time led to the ETOPS changes now allowing twin engine flights over vast oceans. Disregarding SPO, the same goes for the autonomous flight systems (no pilots) to be proven on cargo flights to ensure reliability of the system over many years just like the jet engines proved their reliability in operation. Until the risk assessments show an autonomous flight operations acceptable risk profile for passenger operations, cargo operations will continue until it does, just like the ETOPS experience. Does that help with comprehension on my previous post?
I agree with you on all points. They are exactly those issues I would think of too, when SPO or even non-Pilot Operation is done. When I would know, already during booking, how many pilots would be operating on the airplane, I would for sure eject those with less than 2. Even if those were offered cheaper. But I'm afraid, the majority of people will vote with their budget in mind. Most of them not even aware the risk they are taking. I'm not very optimisticly looking into that future. The development seems easy. From selfdriving cars to selfflying airplanes. Many will say, why not? It works, doesn't it?
I predict that we will move to fully autonomous airliners within 20 years, probably with drone pilots on the ground as backups for a decade or so until full confidence is gained in the fully autonomous airliners.
Captain Joe, many of the reasons regarding safety you mentioned are exactly why the airlines are implementing higher levels of automation. Think about it , the reason you stated an auto Pilot existed was for safety. The extension of this could be also the motivation behind single pilot operations. SPO. Being a pilot myself I feel a great unease with this philosophy. It would seem to me that cutting, 50% of the 4% budget is ridiculous. They should focus on cutting 50% of the 50% budget thereby saving 25%. Although that’s a big challenge but so is SPO.
My concern with any of these computer controlled vehicles is the edge cases. They can be programmed for lots of eventualities but no one can foresee everything. The one thing we humans are good at is reacting to new and unexpected situations. I also don't want a vehicle making the moral choice of who should die in the event of an accident.
Great one as usual Joe! Really hope airbus or any aircraft manufacturing company who's working on this rethinks it. They can work on thousands of other things so actually should let this aspect of aviation remain untouched
Of course the airplane can fly itself.... But it comes with a price and a high risk. When we have plane incidents, it is because things are abnormal, the holes in the cheese match each other. I think both AI and a remote "pilot" on the ground will create more and bigger holes in the cheese and increase the risk of accidents. Airplane incidents often have a prior history, which can be difficult to observe if "pilot" are not present personally in the cockpit and not have been involved in the entire flight envelop. A cockpit work environment can be busy, even during a full normal flight. Sudden incidents require two pilots present to handle this. There must be two pilots in a cockpit to catch dangerous routines, stupid decisions and to make the right decisions. If there are only a single pilot in the cockpit, then there is a high risk that the pilot will get tired and fall asleep, Pilots must have company / counterpart, How will the pilots' skills be if automatics take more and more of the flight?. Remember outcome of many Flight incident has had a lucky outcome because the pilots have good flying skills, it's that simple fly the plane with the remaning functioning systems availble, Do you think AI can figure this out?
Yes it can bse most advanced aircrafts are now flyby wire, no direct control to the aircraft systems AI pilot can fully control the aircraft. The only issues is job loss and it will cause unemployment. 😮 but operating the aircraft may AI pilot is far beyond a human pilot.
“You cannot program the fear of death into a machine” Thats why human pilots are more valuable. When something goes wrong. They will do _EVERYTHING_ in their power to save the passengers and plane
Great evaluation Capt Joe. In addition to this is the unruly passenger. It is only when the pilot engage with them they subdue to authority. For me, human factor is the most important issue to address where no computer or AI can replace. For sure, the communication link for this airline can be hacked and make hostage of the whole flight. Its will be a very lucrative for the hijackers, very appealing to invest.
The obvious problem with one pilot and a locked cockpit is what happens if the pilot falls ill/ drops down dead? This has happened before. In an emergency (many scenarios from bird strikes to false telemetry to doors falling off etc) two pilots working together operating the craft manually has got to be better than one or none.
Miracle on the hudson movie is a perfect example of why a non pilot operating system, and a 2 pilot operated system should never take place in the industry.
I knew it would be Airbus to push this first. My reasoning is the use of Fly-by-Wire. You get so much automation data from thousands of daily usage of it that it's natural to try to automate it even more. IMO I think it's a fair test, on long haul flights, to try 3 pilots instead of 4
it's perfectly agreeable. Nowadays almost every crash is caused by pilots'errors. Since air space is hugely regulated and not open like roads to anyone, I'm expecting a real advantage in eliminating the human errors. Nowadays planes are not designed with sufficient reliability, redundancy and communication capabilities to do so, so compare with nowadays events is unfair. Computers and sensors are so far betters in data analisys and evaluation than humans that with no doubts in a future planes will fly by themselves. Any transition and validation period? For sure. Technology is making incredible steps and in 20 years the world will be completely different
I think ghis would happen some fa away in the future. Technology nowadays has difficulties to make a driverless car, what about a plane. Don’t worry, Joe, you’ll keep flying a lot more than you think!
It's completely crazy when you think about it. A cylindrical object full of hundreds of souls flying through the air at 900 km/h without a pilot. If pilots can't/won't sit in something like that, why should passengers?
I'm currently halfway through my SPL (sailplane) and have wanted to become a commercial pilot since I was a very young child (basically an infant)... However if this actually becomes true, which I really hope it doesn't, I might have to reconsider things... I'm personally going to have a plan B if for whatever reason I do not continue my career and need to make money. However, I will remain a GA pilot so aviation will always be my passion
Unpopular opinion: I've read through all the comments about people swearing they'd never fly in an unmanned vehicle and then think nothing about getting into an elevator, theme park public transit and increasingly taxis. Furthermore, billions of people still eat at McDonald's when better options exist all around them even after being poisoned by their food WHY because it's *cheap*. Elevators once had elevator operators in them because the public had a lot of fear and anxiety of stepping into a small metal box to go up or down in a building. (also why elevator music was a thing to try and calm people) I'm not going to say airplanes should be flown with one or no pilots, but I am saying the cost vs death calculation is being made all the time by the public and the march towards cheaper everything is *LOUD*. I agree that safety is important, but the majority of the public regularly thinks of safety as an afterthought even after crashes if there's the perception of a "good deal". There will be flights flown without pilots faster than everyone on this channel would like, get used to it. 🥴
Seems to me the question of SPO is .. what is the redundancy for that pilot? All critical systems require redundancy, so unless they're saying the aircraft flys itself 100% and the pilot is the backup - twiddling their thumbs (or "monitoring") most of the time? Ground backup to a pilot requires reliable remote piloting, which seems complex and silly.
This is what happens when penny counters are in charge, all about cost cutting, the amount of workload no matter the level of automation present during an emergency is very high.
I will play Devils advocate here and say that the single operator will be inevitable. There will be heavy automation, ground monitoring / intervention AND for 'Break Glass in Emergency' situations there will be an onboard operator who is trained to take control, when authorised, and save the day. I believe the same operator will have multiple roles such as flight dispatcher (also known as Oscar, Red-Cap for the oldies, Traffic Load Officer etc) Cabin Manager aka Purser and maybe even refueling personnel. The technology is already here.....a few years ago Space X sent 4 people, of whom 3 were complete novices, into space and orbit the earth several times. They had some fancy touch screens, sure, but they were essentially no different to the test animals in the first space test flights of the 1950's/60's era. Dragon X navigated departure from Earth's gravity and weather, navigated a mine field of space debris and returned a completely reusable vehicle and occupants.........all remotely.The tech is more than capable. Finally as for the travelling public being accustomed to automated public transport - well there are plenty of driverless trains, buses, taxis and boats operating with full human passenger capacity. Mental conditioning and financial incentive will eventually cement the decision and community uptake. As Capt. Joe has alluded, the Cockpit Crew member reduction has been an evolution. The continued refinement and evolution is inevitable. I don't like it any more than the next person who disagrees with it, but sadly it will happen. Thanks.
In my opinion, they should keep the two-pilot system while having only one in the cockpit while the other rests. The autopilot does most of the flying, while the pilot monitors for unexpected events.
@@brandtbecker1810 Indeed I am a colleague on the 747. If we look at history similar a similar argument was made at the brink of the jet age as well as when aircraft with modern/integrated systems reduced the number of crew-member needed from 4 to 3 then to two. Robotic. surgery has reduced the number of surgeons required in an operating room. Taxi drivers and uber drivers will soon be replaced by self driving cabs/cars.Some military pilots have been replaced by airmen operating drones from remote locations. Robots have replaced many levels of workers at factories worldwide. I do not put my job before progress. Somehow we have managed to adapt.
Approaching retirement from what has felt like a life of flying airliners I’m somewhat pleased that none of my children have ever expressed any interest in the occupation.
I remember reading this saying as a child in the 80s, "To err is human. To really screw things up, you need a computer." Automated single/no pilot airplanes would be a hard pass for me. I'd need to have a lot more confidence in the programming.
So people didn't question when airlines went from 4 engines to 3 engines and now to 2 engines when flying across the Atlantic and Pacific oceans. The technology will be proven, just like the reliability of the engines were proven to permit the previously mentioned 2 engine operations. Cargo flights will do the proving for autonomous flights.
SPO looks like another step forward to fully remove the pilots and to remove the heavy cockpit from planes entirely. No wires/hydraulic lines to the pilot's controls and dashboard detectors, no windshield, no seats, space could be used for passengers.
I am not done watching the Video but already have some thoughts. 1. Two Pilots are not just redundant, they split or share the workload. 2. How do they Imagine FOs getting experience when only one pilot flies the plane … the answer should not be Simulator. 3. No automation can be intuitive Like a human. I want to See a Computer coming to the conclusion Sully got selecting the APU that early which were nowhere near Standard emergency procedure. Finally, Pilots you should Not allow this to happen! (Edited, only for typos)
And selecting the APU assisted Sullenburger how? Given it produces miniscule thrust to improve glide performance if that's where your coming from. He still ended up in the Hudson and made mistakes doing so too.
@@davidpalmer9780 The APU is important for a lot of things. first, it gives the plane energy back so that the instruments work and this allowed Sully to do this very good forced water landing leading to no casulties. And that he had chosen an important task against the available and allowed procedures shows also that there now is a Checklist for such water landings having the APU up very early. But also that is just a detail we discuss here. Also in this answer there are a lot of ammo against a computer flying the plane besides the autopilot. A Computer can't break the rules it is programmed to follow. And a developer (I am one) can't forsee all the possible scenarios might happening up in the air so even from that point of view it renders this impossible.
Love the picture with the four Joes and blinking of the captain Joe. People are almost certainly the least cost in most industries but they are the most vocal ones annoying the business owners. Devil's advocate given current birth rates, the resource called able people is going to be reduced significantly. So technology will have to make us less necessary to accommodate with the available numbers.
Although rare, there have been 'incidents' and instances that validate the necessity of a 2+ person flight crew (Pilot Flying and Pilot Monitoring). There was the infamous British Airways Flight 5390 where the pilot was sucked out of the flight deck. Had there not been a FO there that day, the problem may have ended a heck of a lot more worse off than it did. I also recall some incident where a Captain completely lost control of an aircraft (and due to their panicking, started making the situation even worse) but thanks to a calm first officer, the situation was resolved. My point being that the idea of having just one pilot flying and no FO or pilot monitoring should NEVER EVER even be considered, let alone approved. And removing them altogether and relying on "technology alone" is a recipe for a really tragic disaster. Technology should be considered an aid to people and not a replacement for people.
If or when the time comes that there are no human pilots at the controls of airplanes, I personally will never set my foot on board of an aircraft! People have gone nuts with their wet dreams of "artificial intelligence" which, in fact, is not an intelligence at all but rather sophisticated pre-programmed computer actions.
Yea but navigating through 2 or 3 dimensions is not what makes it difficult. It’s not really easier to design a self driving car then it is a self flying airplane. Airplanes aren’t looking at lane markers and trying to figure out where the lane is. Nor do they have to worry so much about a person jumping out in front of them. They don’t have to deal with situations where they are mere feet away from other planes. Atc keeps planes separated, tcas can figure out potential collisions way earlier than situations a driver has to deal with. When a tcas advisory is given to pilots they are to follow it even against atc directions because tcas has already figured out the actions both planes need to take to ensure a collision doesn’t occur. Pilot intervention in this case, by deciding they or atc knows better at how to avoid this collision then the automated systems, makes it more likely there will be a crash not less. It’s not really an issue to design a system to follow a path through 3 dimensions just like it’s not an issues to design a system to follow a 2D path. The difficult bit is making judgement calls and handling unexpected situations. And so long as we keep airspaces controlled like they are now it’s easier to deal with situations a plane has to deal with compared to situations a car does. In terms of path following at least. The fact that a plane can’t just apply the brakes and come to stop if it suffers a malfunction or catches fire certainly complicates things. But it complicates things for human pilots as well
I am a retired airline pilot and have been flying since 1971. I doubt less than 2 pilots will happen in my lifetime. But when it can be shown that it is safer to fly without pilots, we will have airlines flying with 1 or zero pilots. Today most accidents are caused by pilots and not airplanes. The biggest challenge will be public acceptance. People in the 1920s would think flying 400 people across the would be unthinkable. We cannot even imagine what we will be doing in 50 to 100 years. In 1970 we could not Imagine even smart phones. People might not even be able drive a car in 100 years. How many know how to drive a horse and buggy today? 16:24
Totally agree and in the same way engine reliability was proven to allow ETOPS changes which then permitted going from 4 to 3 to 2 engine operations across the vast oceans, so to will the autonomous software be written and proven by formal methods then further proven in cargo operations for many years before passenger operations. Nobody 'squarked' (excuse the pun) when those ETOPS changes were made... I wonder why?
It's all for profit. There's no reason to spend more time diving deeper because that's all it is: money. If airlines didn't lose so much money in payouts, PR, lost of hardware, etc in every crash, they would have been down to 1 pilot years ago
SPO (Single Pilot Operation) sounds like a good idea in theory, but in practice, it’s terrible. In an attempt to save money, the cost ends up goes up, and that cost is human lives, system failure, human fatigue, eventually it’ll add up to a situation that technology nor pilot couldn’t correct, resulting in an accident that’ll cost lives. As the old saying goes, “What goes around, comes around.” Captain Joe is right, there should always be two pilots in the cockpit, at all times.
Moreover, the more technology will replace human operators, the more we're going toward a total human unemployment leading to human starvation and extinction.
Am I too pessimistic? I hope so, but I'm afraid I'm not
@@vannizaniboni3502That’s not what automation or increased efficiency has ever resulted in, from the invention of fire to the advent of robotic fabrication. Rather more automation results in more human prosperity and more people in the fields of innovation, where humans excel.
There's hundreds of smaller planes that operate single pilot, and there's no issues. There's never been a recorded incident of a single-pilot accident on a passenger-rated plane due to pilot incapacitation. The medical requirements to fly are very strict.
@@PatrickDuffy-u3s ummmmm...... ua-cam.com/video/GrI0xH1rssc/v-deo.html
@@PatrickDuffy-u3s strict??? hahahah
Im telling you right now that if airlines start adopting planes without pilots, i will never set foot on a commercial plane ever again
It won’t matter as plenty of others will as airline travel gets both safer and less expensive with autonomous airliners.
Same
Planes are already 99% automated. You just didn't know it.
@ doesn’t matter, a plane with no human pilot is a software error away from disaster
@ And planes with human pilots are a pilot error away from disaster. What’s your point?
If an airline starts a "no pilot in the cockpit" or even a SPO (Single Pilot Operation), I will avoid that airline like the plague. As much I like technology, it can not replace having two pilots in the cockpit. Superb video as always.
I would feel much safer
@@womble321maybe you are dumb but thankfully the world still has sense
@@womble321 if you don’t know , you can’t be afraid!!
fighter pilots do single operations safely
"You spent so much time wondering if you could, you never stopped to think if you should." - Jurassic Park
In the United States, most trains are still operated by at least two people, and they are locked onto a track and the route set by the control center.
Perfect example! If mankind needs 2 humans to control a vehicle in 1 dimension space, how will it control a vehicle in 3 dimension space, plus the limitation that the vehicles can not stop when airborne?
Look at the unionized workforce for that answer otherwise it would have been done long ago.
In some states in Europe this is not anymore case.
In 'local' passengers trains the conductor is 'checking' that the train engineer is still 'alive' Which means that ticket checking is reduced...
This is permitted where there is a dead-man switch mechanism (if I am not mistaken)
I am a trained computer engineer with a masters degree in automation and control + I also trained as a commercial pilot. Recently while working with AI and testing its code-generation ability, I provided the exact same request ( prompt ) and I got four different responses from the AI and the generated code was useful only in one of the cases! Furthermore if people >KNEW< exactly how the autopilots, various control systems and AI work on an airplane ( specially the AI ) they would not want to fly in an aluminum tube at near the speed of sound and 11km altitude possibly over a large body of water controlled only by some AI model. Invest the money and time intended to train and develop reliable AI ( that can match a good pilot's ability ) instead on selecting people with the correct aptitude and attitude ( to fly airplanes ) and then train and support them well in/on their job. Only non-technical people or those who haven't operated planes in serious weather would make such a suggestion. And YES the pilot's salaries are peanuts compared to the other airline operating costs. I am not claiming that pilots are flawless, there quite a few operators with questionable attitudes and abilities that cause problems ....agreed....BUT eliminating the human element is not the answer. We need the human in the automation loop ( Boeing's approach by the way ), apart from the question of whether the flying public would accept this. Companies like Airbus would of course have the incentive to push for this...so it really boils down to pilots-vs-engineers.
Where did use of AI in the cockpit come from?
As you correctly pointed out, it's not reliable but software written and proven via formal methods is what's needed to operate an autonomous aircraft.
Captain Joe mentions AI to garner commentary like yours knowing it supports his narative.
@@davidpalmer9780
One cannot write formal actions that respond to infinitely unknown future situations. Humans can CHOOSE their actions, programs can't, they are programmed.
@@aeomaster32
In software development, formal methods are mathematical approaches to solving software (and hardware) problems at the requirements, specification, and design levels. Formal methods are most likely to be applied to safety-critical or security-critical software and systems, such as avionics software.
Formal Methods - Wikipedia
@@davidpalmer9780 I wrote SW for the ISS and the Shuttle...somewhat familiar with safety-critical SW architectures and their requirements. Formal methods are great as long as you can define precisely what safe piloting of an aircraft in flight encompasses ( in such a way that it can be couched unambiguously in mathematical formalism ). Anything you can describe precisely might be turned into a system of interacting algorithms to produce desired side-effects ( namely control of some aspect of a physical system according to some predefined outcome ). Alas mathematics is an abstraction of reality and many control loops are actually only effective around linearized regions of their respective operating points. Dissipative and other non-linear effects quickly take us into the realm of nonlinear dynamics which often also displays chaotic behaviours. Well trained humans are great at handling these control tasks which we cannot yet formalize mathematically. Then there is the fusion of experience and sensory inputs into judgment and action. You may have experienced riding jump-seat ( or left or right seat ) in the cockpit of a turbo-prop with PAX following a step-down procedure into an airfield in mountain valley at night, down to minimums and with moderate turbulence and icing....and that without any issues and failures ...in other words everything on the aircraft is churning and burning as it should....Human experience and judgment perform better in a variety of circumstances which are hard to mathematically capture, formalize and specify a priori. Pilot-in-the loop automation works best and we need more of that.
[I am trained in automation and I have programmed industrial computers]
At the same time, while watching this and other channels, I wonder why the logic implemented in the computers of a plane seems not to be fully consistent... I am not talking of AI. I hope that by investigating single-pilot or no-pilot flying, all those missing if-cases are understood and corrected.
BTW read ''humble pi" on the subject of human induced errors (on programs running on computers or mistaken screws).
I’m just gonna say it; no to both. Firstly, you’d have to get the aircraft certificated for Single Pilot operations not only by EASA but also TCCA and the FAA which is notoriously difficult to do or isn’t allowed by law for aircraft over specific MTOW or PAX. Secondly, people don’t seem to understand that in order to fly high UAVs beyond line of sight requires Ku band level receivers and transmitters, which will basically microwave you if you’re within 25 feet (if I recall correctly. Might be larger area) which means you’re looking at cargo only ops. Finally, even if you’re flying a UAV line of sight, you will have a Go-Around min altitude because by the time the signal goes to the CPU then to the FADEC you’re already on the ground. RQ-4s are notorious for those kinds of incidents mostly because ATC is rarely briefed on their min altitude for Go Arounds.
Profit over safety. Sad world we live in
Back in my day…
Its no different what the tech bros are doing with automated cars vehicles@@mrcat5508
This is assuming that a fully automated system performs less safely than an human, which, I believe, will very be proven wrong in the not so far future
@@flav6350 except if anything goes wrong. A pilot can handle a broken pitot tube but an ai would have no idea what to do
@@mrcat5508I think I don't agree. I just asked chatgpt "how to handle a broken pitot tube", and it looks to me that despite this AI being not at all specialised in plane piloting, it has a pretty accurate idea of what to do.
1. Like all essential systems, two pilots are a necessary safety redundancy in the face of emergencies and failures. They're also uniquely organic (unaffected by software errors, not interfaced with computers). The stakes carrying hundreds of passengers are much higher than a single-seater fighter jet with a seat ejection system.
2. Multiple crew system reduces fatigue and loneliness (not a joke; working in a team tends to make a person not disregard proper procedures), and helps prevent pilot s**cide plane crashes.
3. A two-pilot system makes an excellent training environment for the junior pilot, as well as providing a good opportunity for the senior pilot to develop leadership skills.
Sully started the RAT b4 the checklist and its on his testimony. there was no dual engine failure at low altitude in the manual. AI works upon learning models, as long as we are willing to have "Disaster in the Hudson" instead of "Miracle on the Hudson" and actuaries determine that payout is worth the savings, it willl come. Similar to outsourcing MCAS software dev to someone who had never seen a pilots seat in any aircraft.
Isn't that a bit of a survivor bias? You chose a flight where the situation was indeed saved by the pilot, but how many planes crashed due to human error from their pilots?
@@ShamWerksvery few. Very few compared to the thousands of flights happening every single day. It's not our fault you don't understand statistics.
@@casilasgoaler If you had any idea about my actual background in Mathematics and Statistics, you'd understand how funny your comment is to me. 😂🤣Thank you for the laugh. I do understand how many planes are successfully flown everyday - it's part of my job actually. I was mentioning the number of catastrophic situations created by a human error by the pilots, versus the number of said catastrophic situations saved by the pilots. Have a great day buddy, love you.
Well, if you go in that direction. Now imagine it differently. The A320 is designed in the 1980s so yes, perhaps the aircraft didn't notice the engine to start stalling. But technology has advanced, a lot.
Now imagine this same situation with a 2030s designed aircraft. And instead of having 'the miracle on the hudson' we have a basically unknown news article stating 'aircraft lost both engines after take-off and safely landed at 'initial' airport'.
It was proven the aircraft could land but if action was taken immediately, something humans simply can't.
@@fnaeem77 The RAT starts automatically with complete A C Electric lost , he started the APU by hard , to get normal electric from the APU Generator , very good !
Replace the CEO with an AI, win-win :)
That's called a DAO in AI speak and it may just bring autonomous flight operations sooner.
Think of 'SkyNet'.
Excellent video Joe!
My career has been in the railway industry, where automation has made a driver obsolete, with appropriate safety levels, since the 1970s. Why? Because trains only have 2 degrees of freedom, and they are on the ground. When the automation detects a potential unsafe condition, it simply stops the train. Unfortunately, airline/aircraft corporate executives and accountants, not skilled in the technical aspects, believe that the same logic can be applied to aircraft, which have 3 degrees of freedom, and you can't just stop the plane when something goes wrong. Joe briefly mentioned public acceptance of OPO and NPO. That has been the biggest hurdle for trains. Many rail transit authorities around the world keep a person on board, not doing much other than opening and closing doors, just to please the public. The "driver" you see sitting in the cab is not driving the train; there only to make you, the riding public, feel better. We're just getting to the point where autonomous cars are starting to become accepted by some members of the public. It will be a long time before the general public accepts flying with one or no pilot.
I think the 'fake' driver is more about jobs in a highly unionized workforce.
You are damn right! Safety and redundancy is the most important thing in Aviation and cannot be reduced by greed for profit.
Wow this infuriates me, future for our children looks dystopian
@@capitanshaf5751unless we realize that its the current economic model that promotes such profit-oriented suggestions in the first place... And strive to change it to something better, for our children's sake.
When there are no pilots up front, I wont be flying anymore.
No, we'll be taking the driverless bus instead.
If the driverless car or bus encounter an issue they can just stop. Apples to oranges comparisson imho.
That would be an economic disaster.Nobody would want to fly in a large aircraft anymore. Do they really think we are that stupid ?And is there really a responsible pilot who would go along with that ? Could we still trust such a pilot ? I trust Captain Joe !
Dont be silly i would trust a computer over a person any day. If a computer has an accident it will be fixed and that accident can never happen again. People keep making the same mistakes.
@@womble321Like they have fixed the lane keeping in my car to properly recognise diverging lane arrows and not interpret them ase leaving the lane and trying to yank the steering wheel to "correct".
You forget, that fixing an error like that takes time and money and companies will do all they can to avoid spending it. Look many of the design issues in aircrafts that don't get redone for years as the manufacturer argue over who's fault it is.
@@womble321so we will wait for accidents to happen first till we fix all software bugs. Sure thing Sherlock. Am amazed at your stupidity, such kind is rare
@@womble321
Don't. Basic English. Have you ever used a computer? 😂 Tgey have problems all the time and, at times, the "fix" causes more problems. You would not believe how many errors softwares have and are considered fully functional.
@@womble321 Yeah that's not how it works. When a computer fails you can't just "fix it" and then it's fine. A computer failure is not always due to a design flaw.
When an airbus has failures, it literally shuts off automation because it can't operate accurately based off faulty data. That's what a computer is. It takes an input and gives an output. It doesn't have a brain. It has no decision-making ability. Anything that is unfamiliar to it will cause it to just shut off or be a sitting duck. You're so worried about the pilots yet there hasn't been a fatal airline crash in the United States in 15 years. You say people keep making the same mistakes when the industry quite literally changes regulations in training and operations for that exact reason, and it obviously has been working.
Agree 💯 Joe. My gut said a big NO when this idea was first mooted, and you have just articulated the details of my gut feeling in this video. Having watched a huge amount of aviation accident/incident videos where the cockpit crew performed unbelievably under the most challenging of situations, it is incomprehensible that AI could even reach that level of finesse. Your statement about technological advances being an amazing tool are spot-on. That brief shot of a pilot alone in a two-seater airline cockpit made me feel a depressing sense of aloneness - and I'm glad that was one of your points raised. Will be a very sad, and retrogressive path aviation takes if this becomes a reality. Thanks for a great video Joe.
Get it, you can automate flying with just the captain. What if he gets food poisoning, or simply wants to take a dump and there’s a decompression in the cabin. Who’s gonna make the decision to land, will the captain be able to clean his mess fast and understand the situation in 5 seconds all while he’s in a decompressed washroom of 2x2ft size smelling like rotten bacon and eggs???
The computers will detect the decompression before it occurs and commence the descent whist simultaneously determining where to divert or to continue to its destination, alert the airline, alert ATC and all whist the Captain is relaxed taking a dump.
@@davidpalmer9780 Cool story, so what happens if its a system failure that caused the decompression, and the computer just got the information that its now 10,000 ft lower than it actually is, and the airspeed is 50 knots faster than it should be?
I'll answer it for you. A fatal crash that the pilot would never have time to correct, if he was off taking a dump, and yes system failures like this HAVE happened in the past, and can happen again. You need a human at the controls who can see what's happening and disconnect the auto pilot, and return the plane to safe flying status.
People always talk about the "fully automated drones" but there are things people don't get about those.. 1) they are always connected to satellites, and their home base. If something goes wrong, they can be manually controlled. There's your human intervention again even with those. 2) Those bandwidths required to control those things would microwave any passengers 3) A commercial Airliner is an entirely different beast when it comes to flying parameters, systems, and flight controls than a drone. Vastly more complicated.
I was going to make a rather long winded comment here, but I think you hit most of the points I would make, so I will stick to the high points...Pilot incapacitation is a real thing, and perhaps more common than we'd like to believe. My recommended video list is currently showing several recorded ATC conversations involving an incapacitated pilot. Incapacitation doesn't always have to be severe. I case of the flu, or even food poisoning can render a pilot unfit, particularly if they have nobody to assist them or notice their mistakes.....nor can you fly a plane from the lavatory. Further, there have been several cases of pilot suicide. While having a second pilot hasn't prevented all of these tragic incidents, it has prevented some, and, in fact, we'll never know how many incidents were prevented by having another pilot handy. Finally, computers have always impressed me with their ability to take orders, I'm not as convinced with their ability to creatively give them. Most emergency situations are at least somewhat unique, and computers have never been particularly good at dealing with situations they weren't programmed for. Fortunately, major emergencies are rare, but when they occur, I doubt you'll find a pilot who isn't glad there are two sets of hands in the cockpit, and in certain cases, happy to have three or four. I'm sure the late Capt. Al Haynes would agree.
The pilots on UA232 couldn't control the phugoid path the aircraftbwas taking with their thrust adjustments.
An autonomous autothrottle system would have predicted and reacted faster to flatten out the phugoid and coupled to an autopilot, would have made a successful landing at Souix City.
@@davidpalmer9780 Computers may react faster, but they can only accomplish things they are programmed to do, in advance. One in a million type failures are probably not on that list. Even now, while automation can assist during some emergencies, if things get too far out of whack, the computer gives up and puts the human back in control.
The day after I lock in pilot as my career this video comes out.... thanks Universe.
You can always fly FPV drones in the next world war..
Hahaha 😂.
That's messed up. I was in the same situation. At least I have flown FPV.
Itbwon't happen overnight... but it will happen.
As a 30 year airline pilot let me tell you how Single Pilot Operations (SPO) will play out - the first place we will begin to see this is with cargo carriers because nobody cares if a plane load of Amazon boxes crashes. After a few years of this and very gradually we will see SPO slowly move into regional flying by way offering lower fares. Remember consumers vote with their feet and this move will largely be driven by consumers favoring the lower prices that SPO airlines will be able to offer. Just like how right now it is not cost effective for North American airlines to run four engine airplanes soon too it will not be cost effective for them to run two pilot cockpits.
I agree with your projection.
Cargo ops would be the safest bet to prove the technology.
I agree, SPO is only a matter of time. Passengers may feel uncomfortable for a while and opt for flights with two pilots, but then the cost of those tickets will get out of reach for most. However, fully automated flights, with no humans in the cockpit, will likely be a far harder sell and might not appeal to passengers in the foreseeable future. Fully automated cargo flights could become commonplace, though.
@@vasilivhcost of 2nd pilot is nothing compare to cost of fuel, airpirt fees and leasing
@@KukosEQ and still, the number of pilots in the cockpit has only gone down. It's not just the salary, it's also the training, hiring, evaluating, etc etc that comes when you have human pilots. Training pilots and then keeping their training up to date is quite costly to the airline, I would assume, so halving those costs must seem attractive. An AI doesn't need training (once you have one), doesn't sleep, eat or go on strike, it just works.
@@savagecub Cargo ?? Maybe it’s new to me ,if you kill 1 or 2 , it doesn’t matter ? The airlines save theoretical 2-4 % with out pilots , minus the huge infrastructure to make that happen , give 2-4 % to the good trained pilots or 30-50 % to the developer ??
1. Electronic warfare. A remote-controlled plane can be taken down without any missile.
2. Pilot fatigue. A Pilot alone will be much more fatigued than two pilots.
3. Emergencies. Emergencies bring a high workload. One pilot flies, the other one trouble-shoots.
4. Judgment redundancy. 4 eyes spot more than 2 eyes, a second brain can identify the mistakes of the first brain.
Obviously, Garmin hat already implementes the autonomous emergency return. But that works only with an intact plane. No-pilot is ruled out. I can imagine single-pilot only for flights less than 1h, with a full cockpit data link to a set of emergency pilots and engineers on full standby.
You think?
The aircraft will be autonomous without the support you mention.
Supported by cargo operations, the autonomous system will be proven the same way jet engine reliability was proven which allowed going from 4 to 3 and now 2 engine operations across the Atlantic and Pacific oceans.
@@davidpalmer9780comparing 4 engines to 2 engines with autonomous aircrafts is a really wild comparison man! Proves you have zero experience in the Aviation industry or whatever you have is pretty useless. The reason 4 engine was changed to 2 because the engines became more efficient and less costly. Nothing to do with "proven this proven that" lol.
@@casilasgoaler
Wow... so you don't know the history of ETOPS?
A.I can't get my ocado shopping right, let alone land a jet. we need 2 professionals in the cockpit
Define A.I. . If we define it as not self aware, intsruction following program, then it can already land an airplane if the conditions are right, example: ILS and GLS autoland features.
If you mean pseudo self-aware software, capable of critical decision making and learning on the go, by A.I., then it is also possible, the technology is already here. Earlier this year or maybe last year the USAF's (I think) experimenting/developing division successfully completed a test of an unmanned, autonomous software equipped flight with a modified F-16 -I think-. The sofware also beat the pilot in dogfight.
So the tech is here. Is it safe? Depends on the circumstances.
Military and civil sectors and scope of usage are different.
u dont need an AI to land an aircraft, we already did that in 1970, it always cracks me up when people think we need AI to fly airplanes lmao
I am in a flying club and also rent from time to time, so I fly several planes. Every time I get in the cockpit, I always make sure to know exactly where the autopilot and electric trim circuit breakers are just in case the those oh so helpful systems suddenly disagree with my intentions to have a safe flight.
... and I'm sure that autonomous flight software would do the same.
Well done for doing this check as others who don't have lost their and others lives too.
Airlines: No more pilot!
Swiss-Cheese model: Yes.
A well trained crew backed up by autonomous monitoring flying the aircraft is the safest system practically achievable. Aircraft flown by computer servers and world engines is the literal stuff of dystopian movies, videogames, and nightmares.
When sh*t hits the fan, any plea by a distressed pilot to automation for assistance will be met with something along the lines of the famous quote by HAL 9000 in 2001 "I'm sorry, Dave. I'm afraid I can't do that.". On a side note, these CEO's that are pushing the no-pilot-in-the-cockpit idea, would they want their private jets to fly as such with their a$$ in them?
I think one of the most important factors is that pilots are continuously learning from each other. Young first officers learn from experienced captains, but even captains can learn from first officers too. How would that be possible in single pilot ops?
This subject crosses my mind as a student pilot with experience in A.I. @Captain Joe covers the core areas of concern really well here. I think single pilot ops is a scary prospect
I'm sure you realise that Captain Joe has a very biased viewpoint in this video as he wants to preserve the role of airline pilots.
He neglected to show that radio operators were also carried on early airliners too and later removed with the navigator and clight engineer roles.
Cargo aircraft will prove autonomous aircraft operations before being permitted for passenger transportation.
I totally agree with you, Captain Joe. Here in the southwest United States, I live in Phoenix, Arizona, there is a driverless taxi cab called WAYMO. Even Young college students want nothing to do with a driverless automobile! The problem with technology is, that something can always go wrong. Not only do I feel much safer, but it just plain is much safer to have a human behind the controls, especially when that human has a lot of experience operating the vehicle.
I second your idea Joe. It would be hectic if only one pilot or none is in control of the plane . Safety first!!😊😊
Airbus is really pressing their luck with this, “Single Pilot Operation”. What if in the future there’s a single pilot (on a shorter flight), and that pilot becomes incapacitated (like suffering a heart attack), then that flight is SOL. There should always be an emergency pilot on board just in case.
Hopefully there is a pilot among the passengers.
@@deltaskyhawk does MSFS experience count? 😂
@@drwhitewashHey hopefully xplane does too! Or maybe if they want we have our vatsim hours 😂
@@OnionAviation I have logs from different Virtual Airlines to prove my pilot hours count 😆
@@drwhitewash I should work on that too 😂
I vividly remember when I was in school for aircraft around 2000 era, all the talk was about how pilots in military and civil aircraft would become a thing of the past very soon. However, people feel alot more safe with an actual person flying and that's why it never happened from what I could always see.
I really feel like these corporations have nothing to do but cut costs and make more money. It's getting to a point beyond ridiculous
The day they eliminate pilots I'm not flying ever again. I work in medical devices (robotic surgery) and there's no freaking way I would EVER trust a system without redundancy.
Of course there will be multiple redundancies built-in, just like there are 3 navigation systems in most airliners today, there will redundant, autonomous systems.
no plane should fly with 1 pilot, what if a pilot becomes ill or we get a repeat of Germanwings 9525
This is what they seem to forget
Not having any pilot would solve Germanwings 9525.
@paulroling1781 until someone will try to hack the airplane's system and crash it into the side of a mountain
What if autopilot fails, who's gonna land the plane?@@paulroling1781
While almost all of your rant's arguments may be perfectly on point, I miss your reflection on an even greater timescale.
Just about the same arguments were brought forward by yesterday's seasoned pilots when removing the *navigator* ; again decades later when removing the *flight* *engineer* .
Today, you are that seasoned pilot, and I admire your knowledge, your dedication and your top-notch level of professionalism. Yet I didn't hear a single word about you wanting the flight engineer back.
I'd be more than happy to learn about your thoughts on that.
This is a great point. As it stands today having a fully automated plane is not really possible. But it’s hard to know what will be possible in 10+ years. It will never be flawless but neither are human pilots. The continuance of CFITs and the increase in runway overruns show that this is certainly the case and humans flying a perfect functioning plane are still liable to crashing. Ultimately there are going to be situations where an autopilot will cause an accident that a human pilot would likely have avoided. And instances where an autopilot prevents an accident a human pilot likely would not have. If we get to the point where the latter happens more then the former self flying planes will be a net positive even if we haven’t quite perfected it
Don't forget the radio operator who also was no longer needed on the flight deck.
Autonomous aircraft operations will be proven the same way the move from 4 engine operations to 3 engines and now to 2 engines when crossing vast oceans. It's the proven reliability of the engines that allowed this to happen.
Autonomous cargo flight operations will prove the technology over the years, just like engine reliability did and will pave the way for autonomous passenger operations in the future.
So the only reason airbus wants to keep 1 pilot around is because if something goes wrong they have someone to blame. "PILOT ERROR"
It's a huge risk for aircraft manufacturers to develop and sell fully autonomous aircraft because if/when they go down(and they will) , the lawsuits alone will shut them down for good as they bleed money.
...and still they continue to develop their autonomous flight programs with Airbus and Dragonfly and Boeing and their aquisition of Aurora Flight Sciences who specialise in autonomous flight operations.
Totally agree with your points, and I'd like to add one: latency.
I work as a networking engineer and a commercial geostationary satelitte link can have, in some case, up to 0.8sec of delay, not sure about military tech latency.
Doesn't seem much, but I'm quite sure a ~1sec delay reaction from a remote pilot on ground can make the difference in a critical situation.
The great thing is, once they get rid of the second pilot in the cockpit, they will be able to spare some more by removing dual controls and right-seat instruments : what a breakthrough !
IMHO one of the drivers for SPO is not as much cost reduction as it is about pilot shortage. The increasing need for pilots is obvious as many more are getting out of the profession than are entering while more and more planes are operating on more and more connections. SPO will not halve the need of pilots, but the mismatch between demand for pilots and available pilots is lessened.
But all of your arguments against SPO or No Pilot scenarios are in my mind absolutely right and SPO is also incredibly shortsighted as SPO is basically requiring an all Captain rank pilot staff - How are younger pilots going to be climbing up the ladder? SPO will may be in short term fix the shortage issue, but I fear will ultimately doom the industry
I'm on flight school and I don't know where the shortage is happening cause flight instructors can't get a job, if you don't want to instruct you won't fly, but because instructors are not getting jobs, instructing is not an option anymore since the current experienced instructors are not getting into the airlines. This is a profit tactic, if pilot shortage was the issues there are many other (and better) ways to deal with it
Just don't have pilots. There are more savings to airline operators that just salaries as Captain Joe mentioned in this heavily biased video.
Savings in transport (dead heading) , layover accommodation, training, simulator operations, pilot schedulling and more.
The transition from 2 to 1 pilot is the exact same technological change from 1 to 0 pilots, because the aircraft will always need to be able to fly the aircraft in the event the 1 pilot becomes incapacitated. The aircraft becomes the backup system to ensure there is no single point of failure (pilot). This is why reducing crews to 1 pilot is something we should fight tooth and nail.
Thanks for new updates Joe, I think you’re on point.
I think that there is probably a generational difference in how this is viewed. As a boomer, I say, no way (I work in technology and with AI and know how problematic it can be, who wants the AI autopilot to hallucinate ). My youngest son however inherently trusts computers over humans. He says things like “they don’t get tired”, “they never miss a reading “… etc. even in the scenarios of handling the unexpected, computers are not subject to the startle factor and begin making corrections long before a human could. He and I often debate this (like I am certain many reading this will want to). My point isn’t who is right, but rather, different generations likely view this very differently (as a generalization).
Shows how young he is. Blindly trusting technology is wrong. Using technology as a tool is perfect exactly for the reasons you mentioned (no tiredness, precision, etc.).
It's Captain Joe mentioning the use of AI, not Airbus or Boeing or others as it's not reliable.
Autonomous flight software proven by formal methods will be used and it will be used in cargo operations first to become trustworthy for passenger operations.
The future of aviation is bitter truth...
I haven't even started ground school...and now this is happening 😢
This gonna ruin a lot of people's dreams. Lemme tell you that right now.
Even without ruining people's dream, safety is going to be thrown out of the window. Wait, the front door actually.
Gonna ruin the dreams of people expecting to make money from SPO
You said it, Joe! This one word: Germanwings. Makes the whole thing a terrible idea.
Old joke: Next step is One pilot and one Dog. Why a dog? to byte the pilot if he tries to touch anything, and the pilot? There to feed the dog.
I'm not setting foot on any airliner with less than 2 flight crew in the flight-deck.
Please, let it always be 2 pilots in the cockpit! 🙏😬
I love seeing cockpit flying videos where both are working and interacting.
It's about when things go wrong, not when things go right. A well trained human brain is no match for a computer. It's also the reason pilots should be well paid - you're paying them for when things go wrong, not for all the 'easy/boring' bits. Managers have NFI in this regard.
I am a tech student who was given course work about MCAS. a few things I learnt from the event is that, automation may not be a solution to everything. When you look deep in to that very system, the lack of proper knowledge on how to over ride it, led to loss of lives. Now imagine a malfunction of any of the system to be introduced in the ‘SPO’ aircrafts. Will there be chances to over ride or even make the machine to divert from its operation programming?
Did you know some airlines didn't choose redundancy for the number of sensors that delivered data to the MCAS and when that sensor failed it caused the problem you mentioned.
@ well that is true. You can right me. But weren’t those safety features made optional by the manufacturers? Hope such options don’t come up in the ‘SPO’ crafts
@@ssemandayahayakyeyune4294 Yes... redundant sensors are an option but with choice of a single sensor by those airlines comes their responsibility to train their pilots on the failure of such a sensor too.
1979 - 1980 war ich vor meinem Studium Purser bei Lufthansa. Auf der 707 wurden die Navigatoren eingespart, die 747 wurde gerade eingeführt und auf meinen Mustern hatte die 727 noch den Flugingenieur, aber die kleine „Bobby Boeing“ flog mit zwei Piloten. Da kam dann beim fliegenden Personal der folgende Witz auf:
Lufthansa spart jetzt alle Piloten ein, ein Computer fliegt den Flieger alleine. Nach dem Start kommt die Ansage an die Passagiere:
„Guten Tag auf unserem modernen Flugzeug ohne Piloten, ich bin Ihr Bordcomputer und werde Sie schnell und sicher ans Ziel bringen. Sie brauchen sich keine Sorge zu machen, ich bin so programmiert worden, dass ich keine Fehler machen kann………..
Keine Fehler machen kann………
Keine Fehler machen kann…….“
einen kleinen
thank you captain Joe, you speak sense
Is not only about safety if you ask me. Rich ceos are already rich, big airlines are already big, world is getting more connected than ever. When is gonna be enough? I mean, work gives proposit, goal, dreams; whats the point of making everything meaningless and unalive?
No country's legal system should approve the replacement of the Pilots with AI or even the Single Pilot operation. Its simply too dangerous !! Like you very rightly said Joe, the extra presence of the Co Pilot is absolutely Critical in the Cockpit, for a multitude of reasons ! Human safety, always 1st !
Yeah... this was said for the radio operator, navigator, flight engineer and will happen to the pilots. It's the future.
Also, wasn't an Air France Airbus was brought down with passengers and crew through pilot incompetence?
Eloquently put. Thank you.
Full support!
10:40 This was powerful
Well said, Cpt. Joe 👍🏻
Various pilots in the cockpit is safer, for sure 💖 Great video
Why do we have 2 pilots in the flight deck to begin with? For 2 key reasons. One of them being for the purposes of aeronautical decision making. You might not believe this but in most operations the captain has discussions or briefings as we call it with their first officer and vice versa to discuss some potential risks to a flight and what could be done to mitigate them. One pilot can’t do it alone. It takes a team to come up with a game plan for the flight and how to conduct it safely. But the main reason is in the event of emergencies. Take US Airways 1549. You think Captain Sully did that on his own? It was a team effort with both him flying the plane safely down to the river while First Officer Jeff Skiles was completing necessary checklists and other tasks to assist Sully and landing in the Hudson. Take FO Jeff Skiles out of the equation, Sully would have been so task saturated flying the plane plus the tasks the Jeff Skiles was completing and I don’t think Sully would have been able to accomplish what he did that day. Airlines need to stop being greedy. As much as I love flying and aviation as a whole, if airlines are going to get greedy and pull one or both pilots from the flight deck, I propose a boycott of every airline that goes through with this. One less pilot means less safety in the air.
My 2 cents, as an aeronautical engineer with experience on safety systems.
The problem is not if technically SPO is feasible: it is, or will be shortly.
The problem is: who is responsible when something bad will happen, e.g. crash due to pilot incapacitation and the protection system unable to cope with the situation. Who will pay? The manufacturer? The airline? All of us (via the insurances increasing their fees?)
No difference to the ETOPS changes that permitted going from 4 engin operations across vast oceans to 3 engines and now 2 engines. Not a peep from anyone about that.
The technology will be proven in cargo operations.
@davidpalmer9780 ETOPS and SPO are totally different beasts. SPO is maybe barely comparable to single-engine commercial airplanes (of which there is only one in production, and used only for charter or EMT services)
@@luigitenneriello74
You misunderstood what I'm saying.
My mention of ETOPS is analogous to the proving method required for proving autonomous flight operations given the jet engine proved itself over a long time to be quite reliable with quantitative statistics maintained of in-flight shutdowns. The increased very high reliability of those jet engines over time led to the ETOPS changes now allowing twin engine flights over vast oceans.
Disregarding SPO, the same goes for the autonomous flight systems (no pilots) to be proven on cargo flights to ensure reliability of the system over many years just like the jet engines proved their reliability in operation.
Until the risk assessments show an autonomous flight operations acceptable risk profile for passenger operations, cargo operations will continue until it does, just like the ETOPS experience.
Does that help with comprehension on my previous post?
As long as trains have engineers, I’m not worried. They’re on a damn track and STILL NEED people to operate
I agree with you on all points. They are exactly those issues I would think of too, when SPO or even non-Pilot Operation is done.
When I would know, already during booking, how many pilots would be operating on the airplane, I would for sure eject those with less than 2. Even if those were offered cheaper. But I'm afraid, the majority of people will vote with their budget in mind. Most of them not even aware the risk they are taking.
I'm not very optimisticly looking into that future. The development seems easy. From selfdriving cars to selfflying airplanes. Many will say, why not? It works, doesn't it?
I predict that we will move to fully autonomous airliners within 20 years, probably with drone pilots on the ground as backups for a decade or so until full confidence is gained in the fully autonomous airliners.
Captain Joe, many of the reasons regarding safety you mentioned are exactly why the airlines are implementing higher levels of automation. Think about it , the reason you stated an auto Pilot existed was for safety. The extension of this could be also the motivation behind single pilot operations. SPO. Being a pilot myself I feel a great unease with this philosophy. It would seem to me that cutting, 50% of the 4% budget is ridiculous. They should focus on cutting 50% of the 50% budget thereby saving 25%. Although that’s a big challenge but so is SPO.
C'mon Joe! I was expecting a non-bias analysis! Good video anyway!
Very heavy bias with this one.
He's fighting for his future captaincy...
My concern with any of these computer controlled vehicles is the edge cases. They can be programmed for lots of eventualities but no one can foresee everything.
The one thing we humans are good at is reacting to new and unexpected situations. I also don't want a vehicle making the moral choice of who should die in the event of an accident.
Great one as usual Joe! Really hope airbus or any aircraft manufacturing company who's working on this rethinks it. They can work on thousands of other things so actually should let this aspect of aviation remain untouched
Of course the airplane can fly itself.... But it comes with a price and a high risk.
When we have plane incidents, it is because things are abnormal, the holes in the cheese match each other. I think both AI and a remote "pilot" on the ground will create more and bigger holes in the cheese and increase the risk of accidents.
Airplane incidents often have a prior history, which can be difficult to observe if "pilot" are not present personally in the cockpit and not have been involved in the entire flight envelop.
A cockpit work environment can be busy, even during a full normal flight. Sudden incidents require two pilots present to handle this. There must be two pilots in a cockpit to catch dangerous routines, stupid decisions and to make the right decisions.
If there are only a single pilot in the cockpit, then there is a high risk that the pilot will get tired and fall asleep, Pilots must have company / counterpart,
How will the pilots' skills be if automatics take more and more of the flight?. Remember outcome of many Flight incident has had a lucky outcome because the pilots have good flying skills,
it's that simple fly the plane with the remaning functioning systems availble, Do you think AI can figure this out?
Yes it can bse most advanced aircrafts are now flyby wire, no direct control to the aircraft systems AI pilot can fully control the aircraft. The only issues is job loss and it will cause unemployment. 😮 but operating the aircraft may AI pilot is far beyond a human pilot.
“You cannot program the fear of death into a machine”
Thats why human pilots are more valuable. When something goes wrong. They will do _EVERYTHING_ in their power to save the passengers and plane
Great evaluation Capt Joe. In addition to this is the unruly passenger. It is only when the pilot engage with them they subdue to authority. For me, human factor is the most important issue to address where no computer or AI can replace. For sure, the communication link for this airline can be hacked and make hostage of the whole flight. Its will be a very lucrative for the hijackers, very appealing to invest.
The obvious problem with one pilot and a locked cockpit is what happens if the pilot falls ill/ drops down dead? This has happened before. In an emergency (many scenarios from bird strikes to false telemetry to doors falling off etc) two pilots working together operating the craft manually has got to be better than one or none.
Miracle on the hudson movie is a perfect example of why a non pilot operating system, and a 2 pilot operated system should never take place in the industry.
As passenger, it's important to me to know that the captain handling a critical situation is sharing the same destiny than me.
I knew it would be Airbus to push this first. My reasoning is the use of Fly-by-Wire. You get so much automation data from thousands of daily usage of it that it's natural to try to automate it even more. IMO I think it's a fair test, on long haul flights, to try 3 pilots instead of 4
it's perfectly agreeable. Nowadays almost every crash is caused by pilots'errors. Since air space is hugely regulated and not open like roads to anyone, I'm expecting a real advantage in eliminating the human errors.
Nowadays planes are not designed with sufficient reliability, redundancy and communication capabilities to do so, so compare with nowadays events is unfair. Computers and sensors are so far betters in data analisys and evaluation than humans that with no doubts in a future planes will fly by themselves. Any transition and validation period? For sure. Technology is making incredible steps and in 20 years the world will be completely different
I think ghis would happen some fa away in the future. Technology nowadays has difficulties to make a driverless car, what about a plane. Don’t worry, Joe, you’ll keep flying a lot more than you think!
It's completely crazy when you think about it.
A cylindrical object full of hundreds of souls flying through the air at 900 km/h without a pilot.
If pilots can't/won't sit in something like that, why should passengers?
I'm currently halfway through my SPL (sailplane) and have wanted to become a commercial pilot since I was a very young child (basically an infant)... However if this actually becomes true, which I really hope it doesn't, I might have to reconsider things... I'm personally going to have a plan B if for whatever reason I do not continue my career and need to make money. However, I will remain a GA pilot so aviation will always be my passion
Unpopular opinion: I've read through all the comments about people swearing they'd never fly in an unmanned vehicle and then think nothing about getting into an elevator, theme park public transit and increasingly taxis. Furthermore, billions of people still eat at McDonald's when better options exist all around them even after being poisoned by their food WHY because it's *cheap*. Elevators once had elevator operators in them because the public had a lot of fear and anxiety of stepping into a small metal box to go up or down in a building. (also why elevator music was a thing to try and calm people) I'm not going to say airplanes should be flown with one or no pilots, but I am saying the cost vs death calculation is being made all the time by the public and the march towards cheaper everything is *LOUD*. I agree that safety is important, but the majority of the public regularly thinks of safety as an afterthought even after crashes if there's the perception of a "good deal". There will be flights flown without pilots faster than everyone on this channel would like, get used to it. 🥴
Seems to me the question of SPO is .. what is the redundancy for that pilot? All critical systems require redundancy, so unless they're saying the aircraft flys itself 100% and the pilot is the backup - twiddling their thumbs (or "monitoring") most of the time? Ground backup to a pilot requires reliable remote piloting, which seems complex and silly.
Well said Joe!!
This is what happens when penny counters are in charge, all about cost cutting, the amount of workload no matter the level of automation present during an emergency is very high.
I will play Devils advocate here and say that the single operator will be inevitable. There will be heavy automation, ground monitoring / intervention AND for 'Break Glass in Emergency' situations there will be an onboard operator who is trained to take control, when authorised, and save the day. I believe the same operator will have multiple roles such as flight dispatcher (also known as Oscar, Red-Cap for the oldies, Traffic Load Officer etc) Cabin Manager aka Purser and maybe even refueling personnel.
The technology is already here.....a few years ago Space X sent 4 people, of whom 3 were complete novices, into space and orbit the earth several times. They had some fancy touch screens, sure, but they were essentially no different to the test animals in the first space test flights of the 1950's/60's era. Dragon X navigated departure from Earth's gravity and weather, navigated a mine field of space debris and returned a completely reusable vehicle and occupants.........all remotely.The tech is more than capable.
Finally as for the travelling public being accustomed to automated public transport - well there are plenty of driverless trains, buses, taxis and boats operating with full human passenger capacity. Mental conditioning and financial incentive will eventually cement the decision and community uptake.
As Capt. Joe has alluded, the Cockpit Crew member reduction has been an evolution. The continued refinement and evolution is inevitable.
I don't like it any more than the next person who disagrees with it, but sadly it will happen.
Thanks.
7:16 friendly disagree with that. The SR-71, F-15 Tomcat and F-117 Nighthawk are just way more beautiful in my opinion.
In my opinion, they should keep the two-pilot system while having only one in the cockpit while the other rests. The autopilot does most of the flying, while the pilot monitors for unexpected events.
That wouldn't save any money, though.
The issue is saving the money. 😮
@@captsonko.9345 Human reason and judgment can never be truly replaced by AI, it ain't worth it.
My apologies Joe. I am a colleague. I very respectfully disagree with you. All the best, keep up the excellent work.
You sure don't sound like much of a colleague - so, basicall you want to see your job eliminated? That's logical - NOT
@@brandtbecker1810 Indeed I am a colleague on the 747. If we look at history similar a similar argument was made at the brink of the jet age as well as when aircraft with modern/integrated systems reduced the number of crew-member needed from 4 to 3 then to two. Robotic. surgery has reduced the number of surgeons required in an operating room. Taxi drivers and uber drivers will soon be replaced by self driving cabs/cars.Some military pilots have been replaced by airmen operating drones from remote locations. Robots have replaced many levels of workers at factories worldwide. I do not put my job before progress. Somehow we have managed to adapt.
Approaching retirement from what has felt like a life of flying airliners I’m somewhat pleased that none of my children have ever expressed any interest in the occupation.
Think about the positive side of it, no pilot will ever get killed in an air crash
I remember reading this saying as a child in the 80s, "To err is human. To really screw things up, you need a computer." Automated single/no pilot airplanes would be a hard pass for me. I'd need to have a lot more confidence in the programming.
So people didn't question when airlines went from 4 engines to 3 engines and now to 2 engines when flying across the Atlantic and Pacific oceans.
The technology will be proven, just like the reliability of the engines were proven to permit the previously mentioned 2 engine operations.
Cargo flights will do the proving for autonomous flights.
Totally agree with you Captain 🧑✈️
You drunk the KoolAid...
I totally agree with you on this
SPO looks like another step forward to fully remove the pilots and to remove the heavy cockpit from planes entirely. No wires/hydraulic lines to the pilot's controls and dashboard detectors, no windshield, no seats, space could be used for passengers.
I am not done watching the Video but already have some thoughts.
1. Two Pilots are not just redundant, they split or share the workload.
2. How do they Imagine FOs getting experience when only one pilot flies the plane … the answer should not be Simulator.
3. No automation can be intuitive Like a human. I want to See a Computer coming to the conclusion Sully got selecting the APU that early which were nowhere near Standard emergency procedure.
Finally, Pilots you should Not allow this to happen!
(Edited, only for typos)
And selecting the APU assisted Sullenburger how?
Given it produces miniscule thrust to improve glide performance if that's where your coming from.
He still ended up in the Hudson and made mistakes doing so too.
@@davidpalmer9780 The APU is important for a lot of things. first, it gives the plane energy back so that the instruments work and this allowed Sully to do this very good forced water landing leading to no casulties. And that he had chosen an important task against the available and allowed procedures shows also that there now is a Checklist for such water landings having the APU up very early. But also that is just a detail we discuss here. Also in this answer there are a lot of ammo against a computer flying the plane besides the autopilot. A Computer can't break the rules it is programmed to follow. And a developer (I am one) can't forsee all the possible scenarios might happening up in the air so even from that point of view it renders this impossible.
Love the picture with the four Joes and blinking of the captain Joe.
People are almost certainly the least cost in most industries but they are the most vocal ones annoying the business owners.
Devil's advocate given current birth rates, the resource called able people is going to be reduced significantly. So technology will have to make us less necessary to accommodate with the available numbers.
Although rare, there have been 'incidents' and instances that validate the necessity of a 2+ person flight crew (Pilot Flying and Pilot Monitoring). There was the infamous British Airways Flight 5390 where the pilot was sucked out of the flight deck. Had there not been a FO there that day, the problem may have ended a heck of a lot more worse off than it did. I also recall some incident where a Captain completely lost control of an aircraft (and due to their panicking, started making the situation even worse) but thanks to a calm first officer, the situation was resolved. My point being that the idea of having just one pilot flying and no FO or pilot monitoring should NEVER EVER even be considered, let alone approved. And removing them altogether and relying on "technology alone" is a recipe for a really tragic disaster. Technology should be considered an aid to people and not a replacement for people.
If they had no pilots flying back then and autonomous software in-place, there would not have been a windscreen to fall out.
If or when the time comes that there are no human pilots at the controls of airplanes, I personally will never set my foot on board of an aircraft! People have gone nuts with their wet dreams of "artificial intelligence" which, in fact, is not an intelligence at all but rather sophisticated pre-programmed computer actions.
It's Captain Joe who mentioned AI, not Airbus or Boeing so it's a mute point you make drive by Captain Joe to support his accepted bias narrative.
I have seen many vidéos of aviation 's accident and I don't believe that a single pilot could avoid that. Team work is crucial is a cockpit.
we are still having issues getting autonomous 2D operations (aka atuopilot car driving) but lets add another dimension and claim we can. Chapeau!
Yea but navigating through 2 or 3 dimensions is not what makes it difficult. It’s not really easier to design a self driving car then it is a self flying airplane. Airplanes aren’t looking at lane markers and trying to figure out where the lane is. Nor do they have to worry so much about a person jumping out in front of them. They don’t have to deal with situations where they are mere feet away from other planes. Atc keeps planes separated, tcas can figure out potential collisions way earlier than situations a driver has to deal with. When a tcas advisory is given to pilots they are to follow it even against atc directions because tcas has already figured out the actions both planes need to take to ensure a collision doesn’t occur. Pilot intervention in this case, by deciding they or atc knows better at how to avoid this collision then the automated systems, makes it more likely there will be a crash not less.
It’s not really an issue to design a system to follow a path through 3 dimensions just like it’s not an issues to design a system to follow a 2D path. The difficult bit is making judgement calls and handling unexpected situations. And so long as we keep airspaces controlled like they are now it’s easier to deal with situations a plane has to deal with compared to situations a car does. In terms of path following at least. The fact that a plane can’t just apply the brakes and come to stop if it suffers a malfunction or catches fire certainly complicates things. But it complicates things for human pilots as well
I am a retired airline pilot and have been flying since 1971. I doubt less than 2 pilots will happen in my lifetime. But when it can be shown that it is safer to fly without pilots, we will
have airlines flying with 1 or zero pilots. Today most accidents are caused by pilots and not airplanes. The biggest challenge will be public acceptance. People in the 1920s would think flying 400 people across the would be unthinkable. We cannot even imagine what we will be doing in 50 to 100 years. In 1970 we could not Imagine even smart phones. People might not even be able drive a car in 100 years. How many know how to drive a horse and buggy today? 16:24
Totally agree and in the same way engine reliability was proven to allow ETOPS changes which then permitted going from 4 to 3 to 2 engine operations across the vast oceans, so to will the autonomous software be written and proven by formal methods then further proven in cargo operations for many years before passenger operations.
Nobody 'squarked' (excuse the pun) when those ETOPS changes were made... I wonder why?
Before even watching, I’ll say better make sure Ted Striker is one of the passengers. Unless you are flying to WestWorld.
It's all for profit. There's no reason to spend more time diving deeper because that's all it is: money. If airlines didn't lose so much money in payouts, PR, lost of hardware, etc in every crash, they would have been down to 1 pilot years ago