Fully automated landings have been here since the late 1970s. It’s called “autoland”. Once programmed, the plane can fly toward the runway, capture the glideslope, descend to the runway, flare (ie pitch up to arrest the descent), automatically retard the throttle, automatically apply aerodynamic speed brakes, automatically apply wheel brakes after touchdown, and automatically track the runway centerline (auto rollout guidance).
The pilot is still required to lower the landing gear and flaps.
For low-visibility autolandings, most critical systems are triple-redundant. Autopilots, radar altimeters, etc. If the autoland system fails, the pilot must be prepared to land manually.
If you have flown on a jet airliner much the past 30 years, you have probably made an automatic landing. Some carriers require or strongly recommend their pilots make automatic landings in instrument conditions because the plane generally does a better job.
Re automatic takeoffs, the basic technology exists but there is less need for this for various reasons. Takeoffs require instant decisions about various ground events, a takeoff can be infinitely deferred, whereas a landing cannot, etc.
There are some issues with this, I think partially due to how that article is defining “auto-land”. They seem to be including basic ILS approaches when it’s claimed that “Fifty years later, virtually every modern airliner from the smallest regional plane to the largest wide-body jet has autoland capability.”
While it’s true that most every commercial aircraft can fly an ILS and other types of precision approaches with vertical guidance, that’s not at all the same thing as “auto-land” as I understand the term. The article does reference “Cat III” approaches, which is more accurate. But it’s absolutely not true that most planes are landing “hands-off”.
Here are two articles from Patrick Smith’s “Ask A Pilot” blog that discusses this (both from a few years ago, but I believe still pretty true):
In the first link Smith says, "Meanwhile fewer than one percent of commercial aircraft landings are “automatic.”
The business jet I fly does not have Cat III capability, and not many do. Full autoland is mostly installed on long-haul planes, and it also involves some ground equipment. So even if installed in the plane, it isn’t always usable. And the crew has to be qualified. The guys I know who are tell me they execute very few auto-landings.
This goes to the larger issue that the general public has very little understanding of what cockpit automation really does. While it’s true that pilots do less flying by hand these days, we are actively monitoring the automation most of the time. I’ve posted elsewhere the many steps and advanced planning it takes to accomplish actions such as descending via a standard arrival. It’s rather involved, and believe me it takes more than pushing the big red button that says “autopilot” (that doesn’t exist).
I don’t think there would be any airliners manufactured today that weren’t fitted with an auto-land system, they are all “technically capable” of doing it. It is not always in the airline’s interest to maintain that technical capability to the extend that the aircraft are legally able to do it though, they don’t all have “certified capability”.
Furthermore, as you say, the number of flights that actually end in an auto-land is very small. Where I currently work I’ve done one for real and two for training / systems testing. Our procedures have just changed to reduce the number of systems test auto-lands from one per month per aircraft to a lower on demand number dictated by specific maintenance requirements. This procedure change was made due to the risk of practice auto-lands ending badly (see Singapore below).
As you’ve noted, the pilots have to be trained for auto-lands, the aircraft systems have additional maintenance requirements, and the airport itself has to have the correct infrastructure. Additionally, low visibility procedures have to be in place at the time the auto-land is done (unless it is for practice in visual conditions).
It is quite common for aircraft on the ground to interfere with the localiser signal of an ILS and that can cause problems when carrying out an auto-land. What a lot of people probably don’t realise is that when airliners are landing themselves there are severe restrictions placed on the number of movements that airport can handle. Some of that is because auto-landings implies poor visibility which means ground movements are slow and tedious, but some of it is because of the interference of the landing system by other aircraft. Having manually flown landings allows for a huge increase in the rate of landings and take-offs an airport can handle.
Flying an Aircraft vs Managing a Flight
Agreed. I suspect that what pilots think is hard and easy about their job is not the same as what the public think would be hard or easy. The actual manual flying isn’t particularly hard, it just uses brain capacity and this makes it hard to do all the other stuff, planning, monitoring, managing. It’s a physical skill that can be learned by most people (not all) and becomes “easy” with practice in the same way that playing piano becomes “easy” with practice (but try holding a conversation while playing a musical instrument! It may seem easy to play a piece of music but it locks away vast portions of your brain like a computer using most of its RAM.)
What most people probably never quite understand is that regardless of whether the pilot or autopilot is physically flying the aircraft, the human pilot is always mentally flying the aircraft and doing that effectively takes a certain amount of training and skill. Continually mentally flying the aircraft then means we can intervene when the autopilot, or more often the flight management system, is cocking it up. What the public further fail to realise is that intervention of some sort is required on almost every flight. If I was to leave the autopilot to blindly follow its little descent plan each time I flew we would invariably end up 3000 feet high at the start of the approach and unable to land off it.
I came up with the Flying vs Managing heading after reading this in post in preview, I think it gets to the guts of the issue but it’s a bit “tacked on”.
I think most people only consider the flying aspect of flight. They are not generally aware of the extent that a flight must be managed. They know that a human can fly a plane and that a machine can fly a plane therefore if the machine is flying the plane there is little else that needs to be done. In reality the biggest challenge involved in a flight is managing it. That entails ensuring the correct aircraft takes off with the correct load of people and fuel from the correct runway using the correct flap and thrust setting and flies the correct vertical and lateral path to the correct destination runway using the correct speeds and configuration for the approach and landing.
Autopilots are very good at flying. They are very good at using a feedback loop to compare an existing state (wings banked 5º left) with a desired state (wings level) and making control inputs to match them up (right roll until wings level). But they are currently rubbish at managing a flight. They have very limited capacity to decide what a “desired state” actually is.
Current auto-flight technology offloads the relatively tedious task of physically flying the aircraft from the pilot so they can concentrate on the higher level tasks such as if / when / how to deviate around this weather paint on the radar that may or may not be a thunderstorm.
To go back to the music analogy, the autopilot plays the instruments but the pilot conducts the orchestra.
Thanks for that correction. To a non-pilot, a fully coupled approach to low minimums might be called an “automatic landing”, since the plane captures the glideslope/localizer by itself, tracks it down to the runway environment while autothrottle controls thrust. In this case the pilot (or the OP’s aunt) would have to flare the aircraft but the difficult part of maintaining a stable descent is handled automatically. I think some Primary Flight Displays can provide “flare guidance” even when autoland is not armed.
But as you said that’s not a real “autoland”, which requires additional hardware and redundancy to allow automatic flare, speed brakes, wheel brakes and possibly rollout guidance, plus recurrent certification of plane, pilot and airport equipment.
However for the OP’s question of whether his aunt could fly an Airbus, this need not be an FAA-sanctioned event. It’s just a thought experiment. It could take place in a Level D full motion simulator: Full flight simulator - Wikipedia
In fact this type of experiment has already been performed several times. Depending on the subject, test conditions, use of automation, amount of coaching, etc, a non-pilot can successfully fly and land an airliner: https://www.youtube.com/watch?v=lw6mjVIdbbc
But the OP’s question asked could his aunt “learn to fly…an Airbus”. This implies not a single contingency event like taking over for a dead flight crew, but could she learn (evidently over time) to operate the plane more comprehensively? That involves more than stick-and-rudder handling but academics, operating procedures, etc. He implied her motor skills operating a car made this questionable but probably of greater importance is whether she could master the details of the complex machinery and operational procedures.
Flying a coupled approach can be a lot less “automatic” than it sounds. The business jet I fly is pretty high-tech, but here are a few things going through my mind and the tasks I have to accomplish to execute an ILS or WAAS-enabled GPS approach (meaning there is vertical guidance to a decision altitude). And this is with all the available automation working, including auto-throttles:
Hopefully we’ve programmed the approach correctly. There may be multiple initial points. More likely, we’re being vectored, but still must have programmed it correctly with regard to our distance from relevant fixes. Even with a good briefing between the pilots, minor changes in ATC handling can necessitate quick fixes.
If we are doing a visual approach backed up with an instrument procedure, have I put the plane at the proper height to utilize the glide slope? In some ways, visual approaches present problems that aren’t generally there in instrument conditions.
Does everything I’m seeing match up with the chart? Are we passing fixes on the approach as predicted and at the right altitudes? I’m going to lower flaps and gear as we go, and I had better be planning that in advance.
Does the automation successfully grab the localizer (lateral guidance) and glide slope (vertical guidance)? This is much more iffy than you might think. As Richard Pearse mentioned, airplanes on the ground can momentarily blank out ILS signals and I’ve seen my plane fail to pick up one or both on a number of occasions. They are also separate events - it could get one and not the other, requiring me to intervene. GPS tends to be more reliable in this way, which is why I’m coming to prefer those approaches to the traditional ILS.
Assuming the automation has captured both the localizer and glide slope, complete the before landing checklist and verify that the auto-throttle is trending our speed down toward Vref.
If we’re in IMC I’m thinking about the decision height and making sure my first officer is calling it out in reference to mean sea level, not confusing it with the height shown on our radar altimeter (which displays height above ground).
If we’re in wind I’m paying even closer attention to the speed. In really rough wind I may be using an alternate mode for the autothrottle which will require me to manually adjust the speed.
According to company and aircraft rules, I need to turn off the auto-throttle and auto-pilot at a certain height, at which point it’s all me. In visual conditions I’ll frequently turn it all off well in advance.
Whether flying by hand or with automation on the approach, I have in the back of my mind the steps for a go-around. These vary a bit depending on how low we are. If we are asked to go around early you don’t want to hit the TOGA button (take off / go around) because that will automatically kick off the autopilot and engage takeoff power through the auto-throttle. I know a crew who got themselves in trouble doing that. Rather, I’ll intervene through the automation and systematically cancel the approach and reconfigure the plane. Down low, it’s TOGA, flaps 12, positive rate - gear up, then follow ATC instructions.
So… automation brings its own complications and utilizing it correctly and efficiently is a task. Not nearly as simple as it sounded to me when I was a Piper Cherokee guy with a wing leveler.
Bringing this back to the OP, I’ll say again most people can learn to fly. When I was an instructor I had to counsel out two or three people. They each had something “wrong” with their thought processes, beyond what an instructor could address. But outside that or serious physical issues, I believe most people could do it if they are motivated. But as one progresses into more complex planes it takes more and more commitment to the craft.
I’ve read of one person who learned to fly before he learned to drive. For him the hard part was apparently suppressing his instinct to pull back on the steering wheel and accelerate when the car in front of him slowed down.
My father flew for airlines for a living. They had a lot of check-ups, check-outs and various training they had to keep current. The way you fly is, to professionals, noticeable and important.
Contrast that with the misbehaviors you see people try to get away with when driving, deliberately, or plain bad driving by the oblivious. He summarized this as most drivers of cars are “unprofessional” and certainly act that way.
I believe that a skilled, competent and aware driver of cars could transition to airplane piloting given the motivation, and provided they pass physical checks and did not, for instance, get airsick or have trouble seeing or staying awake.
However the vast majority of drivers would need a lot of work to get their skills to pilot level. Or it simply may be impossible to teach some people.
…which is why the dream of “flying cars” will have to wait until we have a system for foolproof autopilot flying from takeoff to landing. And we don’t even have that yet for ground cars (a much simpler 2D problem).
As a private (small plane) pilot, I cannot imagine anyone running into a situation like that enough while flying for it to become an instinct. Maybe the kid got that from simulators/games though.
My father, who was a private pilot, pointed out that Flight Simulator was completely unrealistic. Light planes are designed to be hard to crash, but the software made the slightest error into a disaster.
He could have been joking, but he certainly didn’t pick up anything from simulators - he was a WWII bomber pilot. Here’s a rendition of the story http://fancyclopedia.org/hal-clement
“He once told friends that he had learned to fly before he learned to drive. After the war he was taking driving lessons in one of the dual control cars where the instructor also has a brake pedal. The car in front stopped abruptly and Harry’s natural instinct was to step on the gas and pull back on the wheel to fly over it. Luckily, the instructor stopped the car.”
However, the actual number of HOURS probably isn’t all that different. A Part 61 training course for a private pilot’s license in the US requires 40 hours of flight time; most people probably have 50-70 hours before they take the tests. Teen drivers in my state (Kansas) are required to have 50 hours of supervised driving before they can have an unrestricted license; most states have some similar requirement for young drivers, although exact details vary (and those learning to drive as an adult may have lower or no minimums). It’s just that most people have a car in their driveway, but but not their own plane, so flying lessons get spread out over a longer period, perhaps an hour or two a week rather than multiple hours in a day.
My brother was a pilot, light planes only and I sat right seat on a number of flights. He was the kind who checked all the lists but still faced some hairy situations. The worst was when it suddenly clouded up below us and he wasn’t instrumented rated. A hole opened up and he was able to spiral into it. He told me that he had to bank at 60 degrees which means we were under 2g acceleration.
One thing that was only briefly touched on was communication with ATC. Although I had earphones on and could listen in, I found it quite hard to understand the instructions. I suppose they are mostly boilerplate and you get used to them.
This morning we have 45 knot winds gusting to 60 and planes seemed to be landing normally (I live under the glide path about 4 km from the near end of the runway and enjoy watching the planes on final approach, wheels and flaps down), but I would have been scared to land a light plane under those conditions, but would not have hesitated one second over driving.
Almost everyone is confused and a bit intimidated when first exposed to ATC comms. But, while they aren’t quite boilerplate they do follow established forms in fairly predictable ways. With any sort of regular exposure, most pilots get used to them quickly.
The difficulty of landing, at least WRT most small planes, is greatly overestimated by many. When I was learning to fly, I was completely comfortable with the entire landing procedure on my own within a few hours of touch-and-go flight exercises. One day I was landing in a strong crosswind and it was just second nature to come down with the plane nicely banked into the wind, landing on one wheel before gently settling on the other, and it was only later that I thought, “hey, first time I did that – that was pretty cool!”. This was a Piper Cherokee and my understanding is that small Cessnas are even easier to handle. Conversely, the other night I was driving on the freeway on a pitch-black night in a pouring rainstorm, through a long construction zone with narrow lanes marked in orange paint which is practically invisible on a dark rainy night, with no shoulder and concrete barriers running right alongside the rightmost lane. That was far more stressful than any airplane landing I’ve ever made.
That was my experience, too, but I think it’s a combination of interface and the way that some aircraft are (poorly) simulated. I found it relatively easy with a good well-calibrated joystick (you can’t fly a plane very well with a keyboard and/or mouse) and the right plane. My favorite was the DC-3, a very nicely done user-provided add-on to an older version of Microsoft Flight Simulator. I once loaded it up with “unlimited fuel” and flew in real time from Toronto to London Heathrow. It took a while (I let the autopilot do a lot of the work), but it was pretty cool finally seeing distant lights and being more or less on track – I felt like Charles Lindbergh!
Yes, plenty of people take driving lessons for a year. Me for example.
For the rest of us I think it’s partly a question of how much experience you’ve had in related activities: I could drive a (manual shift) truck before I could drive a motorcycle, and a motorcycle before I could drive an automatic car. I was used to riding in a truck, I already knew how they behaved long before I started driving. I couldn’t handle the car partly because it was so similar but different to the truck.
I too have heard the story about trying to pull up when the car in front slows, but in my case from someone I’ve met, not a WWII vet.
And from a flying school where my boss flew 30 years ago, the girl who was told that she was wasting her time trying to learn to fly. She was upset, partly because her driving instructor had just told her pretty much the same thing…
Here’s one of my favorite YouTubers, Mentour Pilot, an airline pilot and 737 flight instructor, explaining what a civilian with no flying experience should do to land a 737 NG (using autoland).
(That technology would sure take a lot of the excitement out of the movies and TV shows whose plot centers on a novice having to land a plane!)
Yes. My dad’s take on flight sims is that you really don’t have a feeling for what’s going on because the visibility “out the window” in the simulations was so lacking and the controls so unrealistic. The newer sims are getting a lot better in this regard tho.
My mom, who also was a pilot, didn’t want me developing “bad habits” from my time flying flight sims, assuming I was also going to go for my real pilot’s license someday. She needn’t have worried: I didn’t get my license anyway, having run out of time / money and later on, interest.
