If FedEx and Amazon set up private airports dedicated to these machines, only put cargo in them, and make their flight routes distinct from other aircraft, I imagine this could be up and running faster than people think.
Edge cases shouldn’t result in loss of life, so getting through regulations and insuring should be much easier than other situations.
The private airport idea is interesting. From a pilots perspective, one of the major roadblocks to these systems being widely integrated into the current infrastructure is managing traffic. If they’re largely operated separately from manned traffic that eliminates nearly all of those difficult problems.
The comment about edge cases causes me to think about weather. I don’t think many non-pilots have a good appreciation of how much complex decision making goes into operating in and around inclement weather. Fuel constraints, airport constraints in terms of equipment, weather, and traffic, unreliable forecasting, etc. Theres a lot that goes into it and each aircraft begins making decisions that involve several people, including the on-board pilots.
It seems like you can have one operator for 5-10 aircraft but then suddenly the weather rolls through and every single one of those aircraft needs individual attention and planning. Did you save much by still having an operator/craft? Or do you need more stringent weather requirements to operate these aircraft? Or do you try to have the best of both worlds and eventually the autonomous systems just crash themselves in non-populous areas absent other guidance.
If these systems are largely grounded every time there’s inclement weather I can’t imagine the people who advertise “the world, on time” are going to be too keen on operating them at scale. Dispatch-ability is a major factor in these types of operations.
Let’s say you have one virtual pilot for 10 planes. They won’t all be landing at the same time. So that pilot could guide each one down successively during inclement weather.
I bet that would be way more fun than the bunches of sitting around that I imagine is currently part of the job.
There's a lot more to managing planes in bad weather than landing them: you can't exactly fly through a thunderstorm.
Furthermore, sometimes go/no-go decisions have to be made a long way out. If it looks like there might be lots of thunderstorms near your destination, it's sometimes prudent to divert immediately rather than risk being near an airport with weather too bad to land AND low fuel.
A great example is a youtube video by stevo1kinevo: he's flying a TBM 850 (fast turboprop) from Florida to Boston International. While checking the weather enroute he decides the chance of storms near Boston is too high, and with the complexity of getting in and out of a major airport, he decides well in advance to divert to an airport on Long Island. This sort of stuff happens all the time and is the very definition of good airmanship.
You can't, but can a UAV? I'm curious how much more tolerant of bad weather a plane can be if there are no people inside. Are storms often strong enough to knock a plane out of the sky?
> Are storms often strong enough to knock a plane out of the sky?
Simply put: yes. Not just any thunderstorm. All thunderstorms.
All aircraft avoid thunderstorms. It's not a case of discomfort for humans inside it; it's about the risk to aircraft themselves.
Thunderstorms contain severe updrafts and downdrafts and huge windshear between them. Aircraft cannot outclimb the downdrafts so cannot maintain altitude inside them. And severe windshear will tear them apart.
The only reason you don't hear about aircraft lost in thunderstorms is because pilots avoid them. All of them.
A friend of mine who is a glider pilot once compared thunderstorm to a steamroller.
If you get hit by the leading edge your plane will get smashed into the ground. If you somehow get to the back part, your plane might end up in the stratosphere, which might be just as fatal due to very low temperature and lack of oxygen.
He mentioned some early gliding pioneers who were insane and lucky enough managed to "ride" a thunderstorm and achieve record breaking flight distance - and somehow not getting killed in the process.
The addition of people to the plane makes zero difference to storm survivability, why would it?
Storms are definitely regularly large enough to bring down even large aircraft, it doesn't happen often because planes of systems on board to ensure they avoid such weather, but it can happen!
Flying by instruments, I wonder how high the risk truly is. One of the bigger concerns might be an instrument failure, like the failure of the airspeed sensors on Air France 447.
Very very high. What fuels a thunderstorm is large amounts of air rising, and falling very rapidly. When the aircraft encounters those air movements it can very quickly exceed the structural limits of the aircraft.
Some more detail: flying in rising or sinking air isn't a problem in itself, as long as you can tolerate the change in altitude. But flying from one into the other, fast, is a sure way to exceed structural load levels.
Yeah, definitely. And thinking about it more I'm sure there are some air traffic controllers that manage 10+ airplanes fine (although they aren't as involved as the pilots, so a remote operator would have a higher workload).
I bet a fleet of commercial autonomous aircraft operation out of private facilities would also have a lot more common procedures (all their airspeeds, climb/descent rates, etc would all be the same), so it might even be easier.
Also maybe they could pool resources? Like pilots in an area where everything is fine leave their planes on full auto and go take over another plane from another pilot?
Edit:
Or maybe, expect the planes to be fully auto, then have a bank of pilots available to take over any plane that is having troubles.
The sitting around can be boring but there’s little that’s more fun than turning off the autopilot and hand flying an Expressway Visual 31 at La Guardia in gusty winds with traffic departing runway 4.
Do it at a workstation in a desk chair with none of the “stress” of the situation and I’m going to find another career. The combination of skills needed and the threat of life and limb just make stuff like that exciting in a visceral way that would never be possible with an RC plane.
As to your argument, yes the load would have to be balanced across flights at various stages. You’d be surprised how many planes are all in a very similar boat though when they’re all converging on a hub in SE USA and a summer thunderstorm comes booming through. Spend an evening watching MEM on FlightAware and LiveATC as FedEx does their thing. What seems like the whole fleet comes marching in through a very narrow window of time.
I don’t think that the same population who currently flies would be the ones who would fly pilotless planes from desks.
I’ve flow personally for 22 years. Other than simulator-based training, I have no interest in flying a flight simulator. Why do that when the real thing is available and so much more compelling as you describe? No way I’d fly from a desk either, but that doesn’t mean that someone else wouldn’t be happy to have that job.
Agreed, it’s not in the interest of my career. I was more just commenting about the GP’s thought that it might be more fun to get to only intervene for landing. Ten simulator landings will never outweigh one real landing, even if the price is butts in seats for 10 hours.
If the system is designed right, they could be even safer than piloted planes in the case of an emergency. Currently, if a pilot is flying and, say, the engines stop and can't be restarted, they're going to be thinking about trying to survive and get to a place where they can land or at least ditch the plane survivably.
An autonomous system doesn't have to do that - if it decides that the situation is unrecoverable, it can spend its remaining time in the air finding and navigating to the most sparsely populated point of impact within its range. Something like "Engines out and won't restart? Find the nearest/largest body of water and ditch into it".
You'd be surprised. Typically, you'll find that jetliners are older and therefore have an older tech stack, but also that they're redundant. Multiple pitot tubes, for example. On newer GA planes, there may be less redundant sensors, but the technology running them can be more modern and more powerful. The tools on the jet get the job done with two pilots and are there to make money, while the GA airplane is going to be piloted single pilot, and more likely for pleasure.
Sure, depending on the environment. But when an aircraft is in one of those environments where reflections and line of sight to satellites are an issue, GPS accuracy is the least of your concerns.
This is in regards to an emergency response system, so an attacker would have to wait for an emergency to attack and well, you may as well just see how it pans out naturally rather than risk jail time to try and coerce an accident that might already be happening.
It’s not (just) about attack though. If you have your emergency in a place where the surrounds cause GPS to be inaccurate you cannot rely on GPS. Ergo since you cannot plan where to have an emergency relying on GPS is not sufficient.
> they're going to be thinking about trying to survive and get to a place where they can land or at least ditch the plane survivably.
I think you immediately have to question how often this has been the source of a disaster. To me, this seems highly unlikely to have been a problem, or to be a problem at all.
Most accidents happen near airports. Most airports are in populated areas, because an airport far removed from a populated area is next to useless in terms of freight or passenger transport.
> An autonomous system doesn't have to do that
Well.. it can't. Which is part of the problem:
> it can spend its remaining time in the air finding and navigating to the most sparsely populated point of impact within its range.
That's a whole separate automation and computer intelligence problem on it's own. There's also no guarantee that a sparely populated area isn't completely empty or even currently unused for any civil or special purpose. Expecting a computer to "solve" this problem is asking a lot.
Just look at the Gimli Glider as an example[1]. Choosing a "safe" place to have an emergency isn't at all obvious and assumptions based upon outdated information are dangerous in and of themselves.
> Find the nearest/largest body of water and ditch into it
Finally.. even if it could make this determination, safely, and with enough sensor observations to reduce human loss of life on the ground, you're presuming a plane without engines can always reach this apparent planned crash site. If your plane can't add velocity, your ability to plan _anything_ is severely curtailed.
> Most accidents happen near airports. Most airports are in populated areas
Are they?
I'd argue that they generally aren't... except that, over time, cities may grow until they are right next to the airport again. And then new airports get constructed - usually far away.
> terms of freight
Only for certain classes of freight. You distribution center needs ready access to transportation, it doesn't have to be next to your consumers. We are not delivering packages to their destinations with planes, after all. Usually trucks do that.
> If your plane can't add velocity, your ability to plan _anything_ is severely curtailed.
Not an issue with pre-planned routes. There can be contingencies in place for malfunctions at any stage in the flight.
I wouldn't worry too much with engine failures and the like. Worry about sensor problems.
> I'd argue that they generally aren't... except that, over time, cities may grow until they are right next to the airport again. And then new airports get constructed - usually far away.
That sounds backwards. Airports get built as close as possible to the cities they serve. For newer airports that ends up being some distance away because they need to avoid existing homes.
> You distribution center needs ready access to transportation, it doesn't have to be next to your consumers. We are not delivering packages to their destinations with planes, after all. Usually trucks do that.
The point of the plane is to save on trucking time though, so you want a distribution center fairly close to population centers.
Newer airports almost as a rule get built far way not only to avoid homes . Homes get built near airport, then airport gets sued anyway. It is more that it very difficult and very very expensive to get the size of land a large airport needs close to the city. It is also the politics of everyone wanting a piece of development.
Airports get built "close" in the sense that most international airports are named after a city, and the distance to that city is generally minutes rather than hours of travel.
Airports get built "far away" in the sense that, between land prices and noise abatement, airports are generally on the very outskirts of a city at the time they are built and the city's central train station will be much closer to the city centre.
But apart from military bases on remote islands, no airports are so remote that "don't worry where you ditch the plane if there's a problem, there's nothing near the airport" would be an option :)
That's assuming you can anticipate the problem. The obvious edge case is that something goes badly wrong but the software does not recognize it as outside the design envelope.
Computers do some pretty "dumb shit" when they're flooded with data and running bad algorithms. I mean 737MAX didn't need a pilot error in order to do something critically bad.
> Edge cases shouldn’t result in loss of life, so getting through regulations and insuring should be much easier than other situations.
I would hope there is some regulatory concern about flights over populated areas. Yes, there are a number of ways this concern might be addressed, but like everything else, I expect the FAA to be really anal about it.
I also hope they are concerned about potential remote hijacking. It is probably desirable that these things can be overridden by a human operator, rather than solely relying on the flight computer. But that opens the possibility to attackers as well.
I don't think distinct flight routes would be necessary or even preferable. They'd have to integrate these planes into the national airspace system, and at that point you're using the routes (airways) that everyone else uses.
My worry is that once these things are fully automated and planes are no longer manned this will also create an infrastructure for future 9/11 like attacks.
The issue is finding land that is effective for the transition points. You need these private airports to have easy access to highways/existing cargo/delivery infrastructure.
It is usually cheaper to just build your fulfillment center next to an existing airport which already meets these requirements.
To be fair, maintaining straight and level flight— i.e., not falling out of the sky— is the easiest part of flying an airplane. Computers have been good at that for a while; in fact, that’s how airline flights are controlled these days during basically all of cruise flight (by autopilot).
Takeoff and landing are the tricky parts.
(I say this as someone who’s firmly on the skeptical side of the spectrum regarding the near-term prospects for fully autonomous vehicle control.)
Garmin has already gained FAA approval for an emergency autoland system that integrates with most modern private aircraft. It is designed to be used in case of an emergency where the pilot has "stopped functioning". It finds the nearest appropriate airport, announces it's intentions to ATC and then flies to and lands at that airport.
There are landing procedures where the pilot simply follows the machine instructions for a long while already. And it works perfectly well where it's available.
Also, the take-off being easier, it's not conceivable that a machine couldn't do it. The reason there is no such machine today is exactly because it's easier, so pilots never need any help.
Instead, the hard part isn't flying. it's all the things an airplane does while landed. It's taxing, interacting with other planes, interacting with other vehicles (there are a lot of them in an airport), parking, etc. If the airport is entirely built for automation, I imagine that will become much easier.
That’s one thing that boggles my mind about airports: it seems like they’re all designed from the ground-up as if it’s the first airport ever built. Terminals too.
I get that the biggest airports have grown organically, but I’m surprised a city building a new terminal or airport (Berlin anyone?) can’t order a standard proven design like they’re building a McDonald’s or Holiday Inn.
Airports are built to fit the space they're allocated. And runways have to be sited accounting for approach patterns that can be many hundreds of miles. Given the variation in geography of the planet Earth there's zero one-size fits all construction that can be done.
Airports are standardized to the degree they can be standardized. There are entire sections of the Human Factors and Ergonomics Design Handbook devoted to airport terminals and other large transportation spaces. But these spaces have to be designed to fit the place that they are constructed in.
Berlin is disaster and running joke on so many levels.
If the city was competent about the project, they would have actually hired people who knew what they were doing. While new terminals and airports are not as standard as Holiday Inn, there are plenty of constraints in most cities [1], that do not allow a lot of room for radical change, and design usually come out not widely different.
> Also, the take-off being easier, it's not conceivable that a machine couldn't do it. The reason there is no such machine today is exactly because it's easier, so pilots never need any help.
I don't think it's just that take off is easier, but that it's also more dangerous than landing. If something happens during take-off and the computer needs help, it needs the pilot to step in quickly so you don't have time to do a computer -> person hand-off.
I don't think it would affect this proposal (no passengers on board, and package-only airports can be located far away from things to hit) but worth noting.
> If something happens during take-off and the computer needs help, it needs the pilot to step in quickly so you don't have time to do a computer -> person hand-off.
The same happens at landing, but the airplane is moving faster.
Over simplifying, there are a lot of situations during take-off and landing where stopping isn't an option. During landing, the extra speed helps you take-off and try again (go around) with minimal risk. A computer can pretty easily handle "something's weird, TOGO and have the pilot take over." Take-off doesn't have that safety margin.
> Also, the take-off being easier, it's not conceivable that a machine couldn't do it. The reason there is no such machine today is exactly because it's easier, so pilots never need any help.
It exists:
"Airbus demonstrates first fully automatic vision-based take-off"
Pilots have to learn the schematics of electrical wiring behind the dashboards. They are expected to do problem-solving in case of malfunction, and this happens really often. The first time asymetric thrust was used to pilot a plane (ailerons and elevators didn’t work), it was done without being in the books. I wonder how this can work with computers, given they have very fee sensors compared to humans (« Do you smell smoke coming from the vents? ») and they are not yet capable of reasoning. They need to be trained millions of time for each failure mode. Granted, thanks to 70 years of experience with MD, we have listed those failures quite extensively and this part is well documented.
This is not the most common way to detect a fire in an aircraft (unless GA). The most common way is "FIRE" being indicated in the annunciator panel.
> Pilots have to learn the schematics of electrical wiring behind the dashboards.
No they are definitely do not. They will only know the wiring if they are also mechanics.
They need to know which fuses control what, but that's a much higher level.
> They are expected to do problem-solving in case of malfunction
Also no. The first thing to be done is triggering the memory items. Those are memorized. Zero thought here. Pilots have to (and generally will) fall back to what they have been instructed to do.
Next there's the checklists. They will execute the checklists line item by line item. No thinking happens here either, nor it is supposed to.
If all those fail... usually you'll read the resolution in some NTSB report.
> The first time asymetric thrust was used to pilot a plane (ailerons and elevators didn’t work), it was done without being in the books
Not sure if that was the first time, maybe first time in an _airliner_. Also in that particular incident, they had an experience commander, first officer and an instructor for that specific model of aircraft working together (the instructor was handling the thrust levers).
That was also a very extraordinary incident.
> flight crew performance was highly commendable and greatly exceeded reasonable expectations
> I wonder how this can work with computers
During the Columbia incident, the computers didn't know the wing was melting. But they detected a deviation from the flight path, which they tried to correct with control surfaces. When that failed, they added RCS thrusters too. The 'wing melting condition' wasn't programmed, but still they managed to use all available resources.
For more on how we could allow craft to 'intelligently' respond to failures, take a look at the Remote Agent, from NASA's DS-1 mission.
That depends on your definition of "tricky". If you mean something that requires human intelligence, I would say it's the complete opposite. More or less the first thing pilots start doing during an emergency is to follow lists, it's all very procedural and most failures have been predicted in advance, as well as what would need to be done to resolve them, or at least eliminate the immediate threat -- and if these procedures do not help, you are in a very bad place anyway. There are exceptions, of course.[1] I would say that the "tricky parts", as always, are very generally the situations that we did not predict and did not program for.
There is one benefit of having people in the loop in that they come with a completely independent set of sensors for flying the aircraft. Look at the recent issues where the single sensor on the Boeing Maxs crash the aircraft. They're far from perfect for sure though(see the times people did the opposite and ignored the correct sensors causing a crash) but there's a definite benefit to having two completely different systems monitoring the data. It's unusual to see that level of safety differentiation, not only having multiple copies of a sensor to take measurements but multiple different designs and types of sensors measuring the same thing to avoid correlated failures.
Having multiple instances of the same type of sensor doesn't necessary help a lot because they tend to have highly correlated failure modes. A human pilot contains entirely different sensor hardware which, although unreliable, tends to fail in different ways.
What's the advantages of a human compared to a computer for landing in rough weather? I would think it's a fairly logical problem. Wind is rotating the plane left, need to move the stick right. Wind is pulling the plane up, need to slow down or push the stick forward, etc. Need to make adjustments past safe levels or can't keep the plane level enough means it's too rough and the landing needs to be aborted. In my mind, it seems like the thing that's perfect for a computer. It can instantly take into consideration hundreds of precise measurements, sensors, and history, and make corrections for the precise amounts. A human has a much slower reaction time and can't make adjustments with the same degree of precision. I feel like a computer could correct a quick gust of wind before a person even has time to move their hand.
Compared to a self-driving car where you have all sorts of situations that require complex analysis, like trying to determine if that's a person lying on the highway or an empty garbage bag, this seems like it should be a walk in the park.
Autoland systems and humans have different limitations. Humans are limited by visibility, because we look out the window and control the final parts of the landing by what we see. The autopilot is limited by wind restrictions, because the control input rates of an airport are limited.
That limitation is there because the sensor inputs aren't as good as you think. Looking out the window reacts much faster to wind gusts than receiving a radio signal to determine runway line-up.
If you want to have a computer execute a full autoland in varying and gusting crosswind you'll probably need it to use cameras and computer vision, but then it becomes limited by the same visibility issues humans have.
You would think we wouldn’t have an endless string of 0-day exploits for operating systems as well but here we are, for the last 25 years. It should be a simple and logical problem as well.
An operating system has to be able to handle arbitrary data, arbitrary network traffic, and arbitrary user input, with the goal of enabling its operator to perform arbitrary computation.
That might be a slightly less tractable problem than controlling a fixed set of flight systems to maneuver an aircraft with a fixed set of degrees of freedom.
Definitely not to understate the complexity of flight systems software at all, but the analogy to operating systems seems unhelpful.
I think one of the key things computers aren't as good as people are is communication. There is a lot of info that a human can convey with a short sentence that a computer would have no chance of doing. The ability for a human to effectively communicate with ground, tower and other aircraft in inclement weather is key.
> To be fair, maintaining straight and level flight— i.e., not falling out of the sky— is the easiest part of flying an airplane.
Sure. But if it does happen, the self-driving airplane needs to have a landing plan that isn't "fall on a population center." It's not insurmountable, but does need to be baked in.
>But if it does happen, the self-driving airplane needs to have a landing plan that isn't "FALL* on a population center."*
Planes don't simply "fall" though, or more specifically it's not a limiting factor for autonomous flight because if they do Fall (due to total structural breakup) there is absolutely nothing a human pilot can do about it either. The more likely worst problem that is still within the realm of "anything can be done about it" is total engine failure (and with it the potential for failure of systems that depend upon engine power as well). Commercial jet aircraft tend to have glide ratios (how far they can go horizontally for each unit they drop vertically) of high single digits to low teens, though some go farther (the 747 for example is 17:1). So from 35000' say they'll typically be able to travel at least 60-110 miles (with some bonus for their initial velocity).
The power problem has also needed to be considered long since due to all the fly-by-wire systems, though for an autonomous cargo aircraft this may be another area worth a second look since some weight savings from normal life support might be well spent on additional redundancy.
As other posters have pointed out a fully autonomous aircraft just needs to aim for the most sparsely populated area of land or body of water. No need to worry about human life on the aircraft dramatically opens up potential ok ground areas. So a failure at cruising altitude that a piloted aircraft might survive anyway should leave a sizeable radius for an unpiloted aircraft to find a ditch point with low risk to human life. So ultimately while yeah sure, of course this should all be formalized, it shouldn't be a real blocker for this application.
So your aircraft is stricken and finds the nearest local field (putting aside commercial desires to protect the cargo too). The aircraft data tells it the field is empty but actually it’s filled with cattle/machinery/a circus/state fair/Trump boat flotilla or otherwise.
Isn't turbulence hard to predict? Could something cause an issue. I always assume pilot take over or at least monitor things more aggressively when the plane starts shaking a lot.
Turbulence doesn't mean loss of control. It's just rough air for a bit, and computers can manage through it. Modern passenger/cargo planes are designed to be naturally stable (positive static stability) in flight so that disturbances are dampened out.
Bigger issues like bad weather and other traffic would likely be programmed well in advance rather than leave that to real-time decisions.
Correct me if I'm wrong, but I was under the impression that most pilot intervention for turbulence was for passenger comfort. If your cargo is secured properly, you should be able to fly through rougher stuff and navigate around the real bad weather that's easier to detect.
In general you're right, planes falling on people is rare, but IIRC sadly one person on the ground died when the Airbus flight PK8303 flight crashed this year in Karachi; some others were injured and most people on the flight were killed.
We built avionics, software, mechanisms and a comm system that enables automated operation of the plane from a remote command station. The remote operator has a GUI where they can command the aircraft.
I know a lot of the Reliable Robotics team from past projects, and they're an especially talented group of people. I would be working there now if I wasn't committed to my current project.
What I haven't seen is any mention of what happens to all the pilots out there. Between COVID and now this, it seems like the path of a great middle/upper middle class job that's lasted for decades is doomed. For UPS/Fedex, I would also guess that the salary of the pilot is effectively a rounding error on all of the total costs of transporting objects around the world.
I think it's unfair to ask a technology startup to answer a public policy question. Economic forces will drive automation of these sorts of tasks one way or another, which will shift income away from labour and towards capital.
It's up to governments to ensure the winners compensate the losers. In the long term, this may be an UBI, but for now it should take the form of retraining and economic adjustment packages (similar to what might occur when tariffs are repealed). This should be funded by a super-profits tax on heavily automated industries.
We all are responsible for our actions on the world. The government doesn’t set policy that says pilots should be automated, and I wouldn’t necessarily assume this is inevitable at all. Government can only be reactionary, and these things are happening without a larger societal conversation that says yes we want this. So I do believe this is a fair question to ask. It would be unethical to not consider ones actions.
One must also consider the ethics of wasting human life and skill on something that can be efficiently automated while other tasks lie undone.
Either way I agree with the parent comment, this is a public policy question about how to make sure people are able to maintain lucrative lives (education policy, monetary policy, social policy) not a question for the business making an existing task more efficient and expecting them to answer what society wants to do with the freed time.
I can't speak for them, but as a private pilot who has become friends with lots of folks who are not career airline pilots, I can tell you that the delusion among that group of folks is quite high that things like this aren't coming for their jobs -- and fast.
The self-flying use case is, at least from a technical standpoint, far more straightforward than self driving, and the cost incentives are pretty huge. Not to mention the fact that some of this tech has already made its way into private aircraft, with Garmin already having a production system in place where non-pilot passengers can simply push a button on an aircraft and it will completely autonomously find a suitable airport to divert to, communicate with ATC, and land -- even taking adverse weather into account.
There is also a lot of regulatory and tech work at the FAA to also try to ulitize drone-like systems to help operate aircraft. The reality is that the vast majority of time that a pilot is in a flight it is quite low-intensity work (if there is any work to do at all). The thought here is that remote pilots could control the machine (or assist a single pilot in the cockpit) during critical phases of flight. The advatange here, of course, is that a remote pilot could do many such flights over a short period of time, whereas the one on board is stuck to the one flight even in moments where he/she has absolutely nothing to do.
Whichever way you slice it, if you think there is some argument for fully or partially autonomous driving, all of those arguments are orders of magnitude more compelling in flying. Timing, of course, is the real question mark.
You don’t need to deal with making sure pilots are all at the right airports. You don’t need to provide them lunch. You don’t need a bunk for them. Pilots are limited in hours of piloting time and sleep requirements, but they can be instantly transferred between planes instead of walking.
I find it strange that autonomous car driving innovations are done by specialized companies and not the dominant car makers. The same for aircraft.
At least the technology news bubble makes it seem so, it’s like they are waiting for someone to eat their lunch, hello Walkman and camera
Frankly the tech for autonomous aviation has been around for a while now. I expect even strong crosswind landings will soon be better handled by a computer than a pilot. But then here’s the problem...
Autonomous systems are great when things are going well, but at the end of the day they fare very poorly in edge cases sometimes with catastrophic results. When you’re driving a car you usually follow traffic rules and stop lights to get to your destination safely. In an airplane there’s constant communication and things change frequently based on weather, traffic, emergencies, etc. I can’t count the number of times ATC has repeated back an incorrect call sign. As a human we know it was probably meant for us and call back to confirm. Would a computer do that? Even if we used datalink communications then ATC would now have a much higher burden. Also human pilots would lose the situational awareness of hearing instructions for other aircraft on frequency.
Then there’s the question of— in an emergency situation is the computer going to be able to make the right judgement calls? What if its sensors are iced over/inop?
All in all, while technologically I feel this could have been pulled off yesterday, I doubt we will see fully autonomous aircraft operating long routes in controlled airspace anytime soon.
So, NASA, is in the process of working on some of this. They’re trying to come up with a new ATC model which will work better for drones / eVTOL air taxis / etc.
Some of the ideas are around “carving” out corridors, where you would be able to fly, and this corridor could be reasonably large (across multiple towers today). The other interesting thing is hierarchical ATC automation.
So, normal ATC would still exist, but they would interact with a tier of controllers who are responsible for them interacting with UAVs.
A lot of this apparently came from the idea that eVTOL is the future, and we will all be commuting to work in an electric airplane.
It's not the mistakes that matter, it's the proportion. Pareto says most cases should be ordinary, so looks like we got that covered. The question is given x amount of flight hours under likely ranges of events, who does better and by how much.
If the human can perform better in niche situations but not better in ordinary then that may lead to an overall decrease in crashes or loss of cargo (the metrics we actually care about). If the human only performs slightly better, well it makes sense to switch to autonomous because humans are expensive, especially when things go wrong.
The point is that no one expects the machine to be perfect. You can get caught up in the "well it can't handle x well" all day, but that's missing the forest for the trees. A robot isn't going to get frustrated if you only fly it on VFR days and ground it for 20 hours in Little Rock because weather changed.
IMO, the critical thing the human crew did in US1549 was the decision to take the Hudson; once they decided that, most crews could pull off the execution of that decision. (Then the presence/absence of boat traffic comes down to luck.)
> once they decided that, most crews could pull off the execution of that decision
That's not correct. Ditching is very hard. Mess up your angle or sink rate or roll even slightly and you'll be in pieces. Same if the water doesn't cooperate.
One interesting thing is that it happened with an Airbus. The flight computers actually modified Sully's inputs on pitch by about 2 degrees, which ensured a softer splashdown than what Sully was commanding.
But that was a very subtle adjustment. Sully was an experienced pilot, with Glider training too.
Yes. The aviation system produced the human pilot who was able to transfer their skills and experience from all previous landings to this new unusual one. I do believe another equally as well trained pilot would have been able to pull it off as well. What would a computer program have done?
Perhaps pilots of previous generations could do this but, having trained numerous pilots for international airlines, as well as pilots for "fee-for-departure" (regional) airlines in the US, I wonder how many crews are up there that could competently handle most catastrophic events. Though I didn't train him directly, the pilot who crashed the Atlas 767 was pushed through the training program I worked for. Described as a "train wreck" in articles, his ability level was characteristic of others I trained (which I regret). Major US carriers do well screening their pilots but, at lower tiers of US aviation employment, I wonder if more automation would be the safer choice.
Human pilots are trained in how to ditch, but the decision to ditch in that scenario was, as far as I'm aware, an instance of creative problem solving.
There are endless "what if" type scenarios where the answer boils down to simply detecting that this is one of those "what if" scenarios and then initiating the appropriate emergency procedure so it can be figured out properly. E.g. an emergency landing is always an option.
That's basically what pilots do as well. They spent most of their time assessing whether everything is OK via checklists and when it is somehow not OK there's not enough time to think and they reach for some emergency procedure instead. Mostly this never happens and there are whole sets of procedures that most pilots never even handle outside a simulator. E.g. landing a 737 with all engines out is a thing but obviously not something most 737 pilots would ever have to do in a real plane.
ATC obviously needs to be automated too. Automating routine interactions like exchanging altimeter settings, agreeing on what the current weather report is, what the active runway is, etc, would eliminate a lot of need for humans to talk to each other and pass the information on to dedicated computer systems on both ends (which is largely what pilots and controllers do). These systems could, and should, communicate directly. Any autonomous plane would have to be integrated this way. Situational awareness is currently bottle necked on the ability to pass information around, which happens to be something computers are good at.
A stop gap solution is putting a remote human pilot in between to enter information relayed by human atc. The main challenge is not just replacing the pilots but also ATC and standardizing the interactions between them.
This is just as true for Human pilots as well, there are plenty of judgement calls and scenarios average pilots also will fail .
Are the autonomous systems better than our best, most experienced pilots? not yet. They are however likely already better than the average pilot. Updating the autonomous software to factor in new scenario is easier than grounding pilots and getting them retrained on the simulator
The expense and availability of type certification and simulators is one of the reasons the 737 MAX was designed to handle like the older model, although it was fundamentally different. After the fixes, if the FAA require separate type rating it will be very expensive and slow process for all airlines.
The autonomous systems will behave consistently under a given scenario, human responses and errors may vary. Fixing the former is easier than quality checking the later.
Planes already fly a lot on their own, they also crash on their own already too, due to software bugs or faulty sensors, 737 MAX is software bug from a faulty sensor reading on the autonomous system after all. It is a blurry line between what we have and a fully autonomous system. The risks are already there, might as well embrace it and test the software lot more rigorously than depend on the human to intervene and override as we do now.
Tech for automating ATC has been available for decades. The biggest resistance has come from the major US airlines, and associated regulatory capture. Automating ATC would significantly increase efficiency and landing slots. Increasing landing slots at major US airlines would create more competition for the majors. Monopolizing landing slots helps majors suppress competition on key routes and hubs. Wonder if these dynamics are moot now however post covid...
>Autonomous systems are great when things are going well, but at the end of the day they fare very poorly in edge cases sometimes with catastrophic results
I think the best case scenario is to have an autonomous system that only steps in when the human pilot is doing something wrong. Then the pilots stay well-trained, while getting the benefits of autopilot. You could then manually give autopilot control for short periods in times when it is too dangerous for human pilots, ex. windy dark landings.
This has sort of been tried before for some time with occasional disastrous results. When the pilot was trying to do one thing to save the aircraft and the autonomous system was trying to do another, crashes have resulted. It's likely that either one would have been ok on their own, but the combination can be bad.
I know the rebuttal to this is "override the pilot" but it's really not so simple. Sometimes the pilot has switched off part but not all of an autopilot system, and it makes sense to maintain that capability.
I believe this is speculated to have happened with an Airbus doing a low pass or touch-and-go at an airshow. The aircraft seemed to believe it was landing, the captain was requesting TOGA thrust.
Airbus auto throttle is a nice idea when it works (once you understand it), but there's certainly something uncanny about the idea of not having direct control of the primary and secondary flight controls.
If I was coming in with one engine out, the last thing I'd want is the aircraft trying to out-think me and tell me what's right, or decide I'm at risk of stalling (2 feet above the runway), and apply power (which with only one engine available would risk skewing the landing).
Well this is where I think autonomous planes would be superior. First off they could make determinations on their own, the old have three or more computers making decisions and voting, as well as be constantly updating and receiving from ground stations. This can be vital in a situation where the computer decides it no longer trusts its own inputs.
It all comes down to this, it won't be solved quickly, but with proper flight routing and use of specialized airports it can all be resolved without endangering the public. For the most part planes are not falling out of the sky on a regular basis but it would be interesting to see statistics on the number of registered events per miles flown.
Human ATC could still be involved, the words used for events can be very specific and the system can know how to respond to all of this. Automated flights would have special call signs and even may end up handled by assigned ATC to keep them separate.
>Autonomous systems are great when things are going well, but at the end of the day they fare very poorly in edge cases sometimes with catastrophic results. [...] All in all, while technologically I feel this could have been pulled off yesterday, I doubt we will see fully autonomous aircraft operating long routes in controlled airspace anytime soon.
You're correct that edge cases are tricky, but you've made the mistake of forgetting the other possible half of dealing with them that full robotics makes possible: simply sufficiently reduce the cost of edge failure. An aircraft with humans on board crashing is a tragedy, even if it's in the middle of nowhere. In general a cargo aircraft carrying mass produced goods crashing in the middle of nowhere is merely a financial statistic. By far the trickiest part of flying is takeoff/landing and navigation on the ground. But players with shipping volume at the level of FedEx, UPS, USPS or probably even Amazon/Walmart alone can flat out afford their own dedicated airports. With no existing path dependency, they can easily build them quite a ways from population.
This completely changes the calculus and possible responses. If something goes badly wrong on takeoff/landing or at the automated airport itself, it just costs money. There is financial incentive for the players to figure it out, but it doesn't involve anyone else. And once at flight given modern aircraft glide ratios, there is no place on the planet that doesn't have ocean/empty land within glide range. For human flights, ditching the plane as safely as possible, or hopefully making it all the way to an airport, are critical considerations. For a robot plane though experiencing an all-engine out just specifically ordering it to crash somewhere without anyone around is mostly okay (there will be some environmental pollution, but it'd be trivial on humanity's list of environmental damage, and the operator could be fined for cleanup). Robot aircraft could even flat out be equipped with self-destruct systems like the range safety systems rockets have.
So for the specific domain of dedicated cargo craft it seems reasonable that fully autonomous control could exist far ahead of human transport simply by virtue of being more expendable and having infrastructure purpose built exclusively for it in ways that are impractical elsewhere.
Autopilots also rely on the instruments working perfectly. A human can usually safely land a plane even when all the instruments (and even the engines) fail.
My understanding was that the autonomous part is not really the barrier at this point. After all, military drones fly around the world already (for 10 years) and there are strong analogues in civilian flight/aircraft.
The problem is "airspace integration", i.e. how does an unmanned aircraft communicate with other pilots, ground, etc. and what are the rules (right of way, visual separation, etc) with planes in lower classes of airspace.
That problem has also been around for 10 years, and I would say nearly no progress has been made on that front.
If there were progress on that (some kind of regulatory solution/certainty), it would unlock a whole lot of activity here.
However, how will the robotic plances integrate with existing voice-based air traffic control? Would there be a new digital channel, or the planes would it use an on-board voice recognition and synthesizer?
If the airplane isn't in the clouds and is below 18,000 feet MSL the pilots are still responsible for seeing and avoiding other traffic. There are a lot of places in the USA that don't have ATC radar coverage, so traffic separation is based almost entirely on see-and-avoid and cloud distance laws for VFR pilots.
The ATC problem is probably the easiest one to solve, and there are already some solutions in use like D-ATIS and automated pre-departure clearance (PDC).
If that airplane isn't in clouds and is at/above FL180, pilots are still required to maintain vigilance to see-and-avoid other traffic. It's true that positive separation is also provided by IFR rules in class A airspace, but any time you're in visual conditions, you're expected to see-and-avoid.
That would be quite a feat without updating the quality of the voice transmissions. When you listen to some recordings of ATC, it becomes clear pretty quickly that context clues is paramount to being able to communicate with ATC. Some of the transmissions are so difficult to hear it amazes me that humans can even communicate through it.
Perhaps a lot of that is cabin noise, and without it, the transmissions would become a lot clearer.
For what it’s worth the radio transmissions sound quite a bit more clear (usually) in the plane. LiveATC uses ground stations which don’t always have great reception and they compress the audio like crazy. I can barely make out what they say on LiveATC sometimes but in the air it’s rarely a problem.
If I was doing this I would start with remotely "piloting" (sending instructions to) them and having the remote pilot communicate with ATC via the traditional radio channel.
That would require radar contact along the whole route. Otherwise if a transponder stops working you suddenly have a ghost aircraft capable of causing massive amounts of damage.
Very cool! I wonder what the real economic benefit of removing a pilot from an airplane is, and what the dominant factors are?
Assuming that a human pilot is better in some emergency scenarios, and might be able to save you from catastrophic loss: these events are exceedingly rare nowadays, and the savings from liability and scheduling constraints must be massive without a human in the pilot seat.
Feeder aircraft suffer from poor pilot / aircraft utilization. Once you go with remote operators you can increase operator utilization significantly. A computer can also fly the aircraft much more efficiently than a human can allowing for significant fuel savings. Other benefits include lifting of crew rest time requirements and improved dispatch rates. Some of this allows for better asset utilization. Finally, you can also re-balance a fleet across geographies more easily to address shifting demand.
Who do people think will have their jobs replaced first, cargo pilots or truck drivers?
One big factor is it often comes down to which worker group has a stronger union and likewise entrenched political interests at local levels, as we've seen in technological progress elsewhere.
Second to that is probably the safety regulatory process. 'AI' or computer automation has also been used heavily in airplanes for a long time so that regulatory process is probably going to happen much more advanced than cars.
Finally, driving regulations often happen at a much more local level (cities can ban AI drivers in their borders for ex) but flights are divided over much wider geographical lines. And there are far fewer airline companies than truck companies to deal with.
The "cost of the pilot" / "cost of the operation" metric gives us this order: urban transportation, inter-city ground transportation, air transportation, water transportation.
The difficulty of automating it metric gives us this order: air transportation, water transportation, inter-city ground transportation, urban transportation.
So I'm personally expecting a mix: air transportation, inter-city ground transportation, water transportation, urban transportation.
There is more "noise" to deal with on the water than on the air. For a start, there is the entire ocean floor, that is relevant on many places where ships usually go into, and isn't as simple to map as a runway.
That problem can be avoided in the beginning. Ships could pilot themselves in deep waters between ports and leave docking / undocking to local human port pilots.
There needs to be a pollution (including noise) tax collected to counterweight the inevitable of potentially non-stop swarm of buzzing that will be making 10,000s to 100,000s of deliveries in cities everyday.
I would prefer if it was legally required to be below a certain amount or you can’t do it at all. Perhaps you can purchase decibel credits. If I can’t sleep at night it doesn’t make me any happier to know that the city is getting more revenue.
Shout out SNA, which has a spectacular departure flight path to ensure compliance with back bay NIMBYS who showed up after the airport was already built.
It will take long for this to change the industry and there are a lot of planes that will be flown manually for decades to come. But obviously it doesn't hurt to have a plan B.
I actually expect that short term there will be an increased demand for pilots as smaller electrical airplanes start becoming more popular. Most of these will be single pilot operations doing short hops. If there are going to be lots of them, there will be a need for lots of pilots too. This too will ramp up over a few decades and eventually that whole business will be autonomous. But not short term.
Autonomous flight is a great example of the 80/20 rule. We've basically had "self flying" airplanes for decades, but to get them to 100% autonomous we need to spend all of the extra engineering effort into covering the remaining edge-cases + landing.
this statement belies a lack of understanding of just how long autoland has existed. Guess which decade autoland was invented... and what decade commercial service flights began making fully autonomous landings...
It is not, on the whole, that difficult a problem to actually land an aircraft autonomously. Now that planes can autonomously fly the rest of the route too, it makes sense to just take the pilot out of the cockpit.
> It is not, on the whole, that difficult a problem to actually land an aircraft autonomously.
Sure, if you have ILS and it's operational.
Large aircraft will not rely on GPS.
It's a far more difficult problem to integrate such aircraft among the existing traffic. You are underestimating the effort of autonomous flight gate to gate.
Edge cases shouldn’t result in loss of life, so getting through regulations and insuring should be much easier than other situations.