Airplanes Can Now Take Off and Land With GPS. Will Flights Be Faster? | Hacker News Reader

Airplanes Can Now Take Off and Land With GPS. Will Flights Be Faster?

mashable.com

35 pointshoag28 comments

A machinist I knew for an airline told me a story about how the avionics in plains in the late 80's could land autonomously in fair weather. The problem wasn't that the systems weren't smart enough, but that the tarmacs weren't robust enough to handle 'perfect' performance every time. After thousands of planes touched down within feet of each other, the pavement would crack and the systems couldn't be configured to inject variability to compensate.

If this were true, then surely chucking a tiny bit of randomness in there would solve the issue.

There's a very similar problem in maritime navigation. Ships used to steer their course along shipping lanes by hand, but now they're automatically steered by GPS. Everyone is crowded right into the middle of the lane, greatly increasing the likelihood of collisions. After a fatal accident caused by this issue[1], the British Coastguard specifically instruct ships not to follow the exact line of the shipping lanes in the English channel.

[1] http://www.maib.gov.uk/cms_resources/SB_%202_2001_%20Ash_and...

Planes have been able to land in poor visibility automatically since 1968 using the ILS (instrument landing system) radio waves. There are various categories with category 3 being supported in 1968 (low decision height and visibility).

Autoland systems: http://en.wikipedia.org/wiki/Autoland ILS Categories: http://en.wikipedia.org/wiki/Instrument_landing_system#ILS_c...

Great news, though the headline is kind of weird.

GPS-only RNAV STARs (Standard Terminal Arrival Routes) have been in use for some time.

The "next gen" technologies referred to in the article involve a lot more than just GPS improvements (such as WAAS and publishing more RNAV GPS approaches annually). They include new ground radar for automated taxi operations, a replacement of traditional transponder-based aircraft separation, redesigning ATIS and weather reporting, and an overall push to standardize all ATC data -- ultimately getting more real-time data into cockpit avionics.

It's about freaking time. I remember back in May 2009 I was on a UA flight from HNL-SFO and if I'm not mistaken, they were testing SFO's GPS approach system: IIRC, we descended from cruising altitude to wheels down in about 10 minutes. Awesome.

That's quite a common time (10-13 minutes) from cruising altitude to landing at most airports in East Asia. America has other issues that interfere with more sensible landing procedures that need to be addressed, but are unlike to be.

This is very interesting, if you have a few minutes, can you please explain briefly why this is the case to someone who doesn't know much about aviation?

From what I understand, its mostly due to the landing approaches being more straight forward, the lack of stacking of planes to land (less circling) and timing between planes is usually slightly longer than in the US.

An extra ten or twenty seconds between planes taking off and landing can have a large impact over the course of a day. In the US the majority of flight delays are in the late-afternoon and into the evening because of the knock-on effect of delays (at that airport or others) and the lack of spare capacity within the timings to adjust the scheduling of takeoffs and landings.

That's interesting. I didn't know that. I would assume, therefore, that domestic operators in East Asian countries tend to average at least slightly better fuel economy? Would be interesting to see some numbers on this. A 10-13 min descent is a huge improvement over the 20-25 min descent commonly used today elsewhere.

Today, it seems like the bulk of the pilot's job is to get the plane in the air and then get it back on the ground. In 10 years, are we going to be at the point where a pilot's job is completely redundant? (Or maybe a fleet of planes is handled by a small group of "ground" pilots?)

Didnt the FAA just state that unmanned aerial vehicles will be flying around general airspace by 2014?

AFAIK "unmanned" at this point doesn't mean "pilotless" - you still have a full-time pilot, just not on board the aircraft proper. It seems pilot jobs are safe... for now.

No Pilot = No Hijakers

... unless your a good hacker of course ;-)

That's actually a damn good reason to avoid remote control of passenger airplanes. All it takes is a nice worm in the network...

By comparison, maintaining physical cockpit security is extremely easy: lock the door, and shoot anyone who tries to force it.

I hear Iran has a few.

Go read the AF 447 accident report and you'll see why humans are both the weakest link and still completely necessary. Sensors fail.

one of the articles covering it here.

give me chills every time i read it

Correction: crappy, inadequately-redundant sensors fail.

Everything fails, if given enough chances to do so.

Just curious: What happens if some dude in a truck drives around an airport with an illegal GPS jammer? These things are everywhere: http://www.economist.com/node/18304246

Couldn't someone do the same thing with existing radio navigation tech? I suspect they just track the source of the signal down and arrest them.

Not a big deal. The pilots would either switch to using other instruments / visual flight rules or would abort a landing and go around for another pass.

Hasn't the FAA seen Die Hard 2?

This works great until some foreign power takes out several of our GPS birds or a CME fries some important stuff, then what? We give up IMC aviation? Can't go back to ILS -- I hear they're taking those out. NDBs? Gone. LORAN? Are those even in existence any more? Can the big airports supplement with WAAS units enough for some kind of commercial aviation to continue? Is that possible?

I love GPS, and I'm for using it more and more, but I have a philosophical problem with putting all of our eggs in one basket. Just like the military's reliance on stealth, this is really cool stuff -- until it stops working. Then you're worse off than before.

Maybe I'm just old cranky guy today.

GLONASS (and eventually Galileo and Compass) have put us in something of a cold war situation. All the superpowers will have full GNSS systems in operation by 2020. Crucially, Galileo is effectively the "non-aligned" GNSS system, with diverse investment from outside Europe.

If we start shooting at satellites, there'll be much scarier things to worry about than how aviators will plot a course for landing. For better or worse, satellite warfare is the new Mutually Assured Destruction.

Thank you. Okay, I'll buy the MAD analogy, although we were very lucky that MAD worked. No guarantee it's going to work in a multi-party situation. But fine.

Sill, where's the dual technology system? When I got my IFR ticket, I shot an NDB and an ILS plus I had to know about GPS approaches (they weren't certified then). If I couldn't shoot an ILS at an airport, maybe I could pull off an NDB at my alternate. If your GPS broke you could use the NDBs to get a fix, your pitot freezes over you use pitch and power, your vacuum gyro fails you have an electric backup, etc. There was always a mix of methods. The beauty of the system was that you never depended on just one thing.

Is that gone now? It's all GPS or nothing? A solar flare takes out our GPS capability, and simply because "there'll be much scarier things to worry about" that means we're not as hosed?

Thanks again for the information. Just trying to sort out in my head what risks are being taken.

I believe the plan is to leave some ILS systems in place as backup. Honestly, they're in a bad spot. I'm sure all their engineers want to keep all current ILS around, but money is on the line. Sometimes you got to make trade offs.

Reply on news.ycombinator.com