Hacker News new | past | comments | ask | show | jobs | submit login

Garmin's got a relatively new system called Autoland that is designed for this exact situation. It allows a plane to land simply by pushing one button. It will pick the nearest appropriate airport, communicate with ATC, and land the plane entirely on its own. It's pretty amazing technology: https://www.planeandpilotmag.com/article/garmin-autoland-thi...



I've always said the reason we don't have flying cars yet is because people can't drive in 2D much less 3D. This gives me hope.


"Flying cars" made sense to me until I started flying. Now I think "driving airplanes" is the more appropriate phrase for what might be in the realm of possibility.

For example: here's how you'd prepare to visit distant relatives with each vehicle:

Car: load up however much weight you want, turn the keys and start driving. Low on gas? Just turn off at the next exit. Weather looks bad? Just drive slowly and carefully and you'll be fine.

Airplane: visually inspect your vehicle, be careful distributing limited weight around the cabin, get a weather briefing and accept that many days you just can't fly, break out your slide rule (literally!) and plot a course between waypoints, with calculations accounting for wind deflection, magnetic variation, fuel burn, and various other factors. And don't forget to plan out refueling stops and emergency airfields too. Then run through your checklist and (once you get permission from the tower, if any) take off.

I never appreciated how user-friendly modern cars are until flying. And air travelers are spoiled by all-weather jetliners piloted by the pros.


Couldn't most of that be done by computer? Weight sensors in the wheels can determine weight and balance, the computer can evaluate the weather and plot the course taking into account all of the factors you noted, including refueling stops and emergency airfields, and it can even run through the checklist for you.

About the only part of that it can't do is the visual inspection.


Sure, if you trust the correctness of the computer's software with your life and the lives of any passengers. There are popular apps now that do some of what you suggest, especially trip planning.

I've seen enough bugs in my day job to want to at least verify the computer's work and have backup instruments, even if just my own senses. Aerospace software is known for relatively low bug counts[1] but also causing fatal crashes[2].

I trust my car controls to be correct and if they aren't I can brake to a stop if anything else seems off. (Unless of course the brakes stop working!) But since you have to take off to fly, you might not realize you can't control the plane until it's already at dangerous speeds.[3]

1: https://www.bugsplat.com/blog/less-serious/why-nasa-code-doe...

2: See https://en.wikipedia.org/wiki/List_of_software_bugs sections on Space, Military, and Transportation for examples.

3: Amazing story about an airliner that took off not knowing its controls had been reversed (a maintenance mistake). Pilots declared "MAYDAY plane is completely uncontrollable we plan to ditch" but eventually figured out how to fly the reversed controls and landed it safely—super impressive. https://www.youtube.com/watch?v=kIc8Rr-cKd8


I think most of that is done automatically by computers in the case of airline jets. For single engine aircraft it's mostly DIY. Though with the jets the pilots get many hours of training on what to do if the various systems / computers go wrong.


Believe it or not, weight and balance calculations are not automatic even in the most advanced airliners. The airlines have algorithms they use which spit out numbers that are provided to the pilots, which they enter into the aircraft's computer. These numbers are vitally important, and on one flight I was on we had actually taken the runway but had to taxi back off because Delta was too slow at getting the numbers calculated.


Not that your general point is wrong, but I don't think "flying cars" are intended to replace airliners. They're supposed to be for travel within a city. Complicated calculations for route planning and refueling seem much less relevant for that case.


How much weight do you think it would realistically take to alter flight in a plane of this size? For instance, if both the pilot and passenger weighed e.g. 200lbs and sat toward the left of the plane, would that considerably (or perhaps even just perceptibly) impact flight? Same for some of the other variables, is there an appreciable different for things like magnetic variation? Wind, of course, seems reasonable - the others I've heard less about. I don't fly, and have never been in a small engine craft.


Fore and aft weight distribution matters a lot in terms of how the airplane flies. Side to side not as much. (Most single-engine airplanes have more moment arm of fuel weight change in the wings than moment arm of two people sitting on the left side of the airplane.)

Fore and aft changes in center of gravity affect how far the center of mass is from the tail control surfaces and the amount of negative lift the tail has to contribute. I can feel the difference when my (fairly small framed) family moves around in the back cabin.

Aft weight distribution makes the airplane far more sensitive in pitch, reduces drag from the tail negative lift, which slightly increases climb rate and/or forward speed for a given power setting.


Not sure about the plane from the original post; it looks pretty hefty. And lateral weight is so close to the center of mass that it's unlikely to have much effect. What's more of a concern is having a bunch of weight far from the plane's center of mass, where the weight tries to lever the plane end over end, increasing the risk of a stall. I heard about a crash where the pilot's seat adjuster didn't lock, so when he took off his seat slid all the way back (just a few inches) but that was enough to cause a crash. Could theoretically happen to any size plane, but matters more with little light planes like those in general aviation.

I have no idea if "Spirit Airlines weight distribution issue" actually happened, but it's funny so I'll share: https://youtu.be/YvfYK0EEhK4

Magnetic variation in my area is +20° (west) off true north. So if I want to follow longitude line true north I need to fly such that the compass reads 20° NEN. And don't forget to account for the hunks of metal inside the airplane, which can affect the compass differently depending on your heading.


The real reason is that no one wants thousands of flying cars buzzing everywhere over a city.


The real reason is that we don't have one-person aircraft that can stop and hover with <300ft stopping distance and is under 10 ft wide. Helicopters come close but they're expensive and I'd hate to see parking for such a model of transportation.


Helicopters aren't really allowed to fly over most areas in cities though, except for police and other emergency helicopters. The noise and danger make it a no go, even if they were smaller.


Helicopters can fly wherever they want, barring the same restrictions that apply to fixed-wing planes. For the most part, there aren't really any helicopter-specific restrictions.


Yes, but this doesn't really help the case for flying cars either way.

Edit to clarify: Presumably you'd want to land your flying car almost anywhere in the city; this is not going to happen anytime soon, for the same reasons that helicopters and planes can't.


Buzzing should emphasize the loud freaking noise these things would create landing, taking off, and flying at low altitude. Annoying to say the least. What we don't need more of is noise pollution.


The autopilot declares an emergency via transponder code and radio calls, which means suddenly everything else in the sky is moved out of your way for you and you're given near-exclusive use of whatever airport the autopilot picks.

This does not bring us closer to "flying cars."


You mean like when cars turn on emergency blinkers any everything moves out of your way?


A pilot declaring an in-flight emergency gets near authority over the airspace they are in. After confirming the declaration fuel/passenger count, the next question a controller usually asks is "what are your intentions?", ie "what are you going do, so we can get everyone out of your way?"

I've never heard a controller get more authoritative than politely ask the pilot if they'll consider something or are able to do something.


Isn’t 3D easier? There sure is a lot more room.


It's not the 3D that kills you; it's the sudden 2D at the end.


Yup but I always thought what if we do like a 1 feet above the ground? We avoid death by plummeting from the sky.

There are other advantages like

- we don't need roads and we save a lot of energy and money on their construction and maintenance.

- we also save energy by going from point A to point B directly instead of following the road.

- No friction between tyre and ground

I guess the amount of energy required to keep vehicle above ground by 1 feet is more than all the savings combined.


You’re not going point to point a foot above the surface unless you’re over water (calm seas at that). Soviets built a concept plane that did just this, though. Airplanes are more efficient near the ground than a little bit away, but not as efficient as at normal cruising altitudes.


You're probably thinking of Ground Effect Vehicles (sometimes Ekranoplans). The Russians built lots, not just as a concept, if you own a huge lake (not an ocean) they're somewhat practical for crossing it quickly. The Americans and Canadians (who also own some large lakes) have likewise built some of these.

The Ground Effect, as its name suggests, only exists near the ground, so in one sense you're "flying" but if the surface drops away you will fall too. Hence it's good on a lake or possibly open plains, but won't work on normal ground with rises and hills and so on, never mind buildings and trees.


I'd thought that there were only a few prototype Erkanoplans built, though it seems there actually was an operational fleet.

There's a particular larger model intended for defence purposes which saw only a single prototype, now abandoned.

The Soviet Navy ordered 120 Orlyonok-class ekranoplans, but this figure was later reduced to fewer than 30 vessels, with planned deployment mainly in the Black Sea and Baltic Sea fleets.

https://en.wikipedia.org/wiki/Ground-effect_vehicle#Soviet_U...

There is one specific prototype which has seen some attention (article and videos), the so-called "Caspian Sea Monster":

https://en.wikipedia.org/wiki/Caspian_Sea_Monster

On HN a couple of years back:

https://news.ycombinator.com/item?id=24857096

For water use, hovercraft tend to be more flexible. They still experience issues in rough water, and though they can cross flat unimproved terrain (beaches, swamps, snow, meadows), they perform quite poorly on slopes, particularly laterally, and generally have poor lateral stability, notably with high winds.

Hydrofoils accomplish much the same capability on water. Tracked vehicles on land. Not having to support your mass dynamically also helps. That said, hovercraft remain useful for military marine beach landings. And eels.


I was thinking of the single giant transatlantic troop carrier they built. You’re right, there are others.


3D has the problem is that you can't just stop, wait and think if anything is totally weird. A self-driving car can pull over if it's confused, a self-flying plane can't.


Then again, self flying planes have been a thing since the early 1910s. It's far easier when you can just pick and altitude and heading and you're basically 99.9% in the clear that you won't hit anything.

In a car that approach will get you roughly a few meters forward, so it's incredibly hard to make a working car autopilot in comparison.


Well, usually it can stop, but if it gets confused during road construction where there's no shoulder to pull off to and it just stops in the road, you run the risk of a truck rear-ending you.


Empty air, a 2d empty plane would be easy too. If it was 3d tunnels it would be hard, those Descent videogames were tricky.


It is considerably more complicated. Apart from added axes of movement (not just vertical, but pitch/roll), you also have to contend with the fact that air is the only source of friction. Acceleration and breaking are much slower than something with wheels on the ground. There is a reason why flying a plane in Grand Theft Auto is much harder than driving a car, even with the simplified physics and vehicle controls.

If we could make aircraft that handle in our atmosphere like the spaceship in Descent, then that would close the gap a bit. But I'm not holding my breath.


The more axis you have to account for, the more complicated it gets.

Go play a car simulation game. Done? Now go play Descent and tell us if it's any easier.


Except when something breaks, and you need an equivalent of parking by the side of the road.


Some factors that complicate the job: Higher speed, more kinetic energy when crashing, aerodynamic lift, uncontrollable regimes...


It is definitely not easier. I mean avoiding a mid air collision might be but that’s rarely what kills a pilot.


no : stalls are deadly in 3D, send the ground at you with no control unless one recovers


Hasn't autopilot had "auto-land" technology for many years now? How is Garmin's different?


Garmin's is literally "oh shit the pilot passed out!" and an entirely inexperienced passenger can push a single button, ending up with the plane stopped on the nearest appropriate runway with the engine off. The autoland system in airliners requires a lot of setup by the pilots and continuous monitoring all the way down, plus working ILS (instrument landing system) equipment at the airport.


When flying private, my pilot told me: “just press this button when I have a stroke or something”.

Almost 20 years ago.

Ignorance is bliss I guess.


Yeah, I don't know what button he would have told you to push, but it definitely wouldn't have landed the plane. I don't think there were even automatic wing-leveling systems that far back.


Not really. You need to intercept an ILS beacon and not all airfields have them and you still need to find the intercept.

Garmin allows true auto landing without ground equipment


US military drones have been doing it for decades now.


How successfully? An bad landing for a drone is much less costly than for a plane with people in it.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: