Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

‘Confirming we are cleared to land?’ Who said what at Austin

fallows.substack.com

Post reply on HN
Print view View original

231–240 of 402 posts

Re: ‘Confirming we are cleared to land?’ Who said what at Austin

#231

Post by HPsquared »

zaarn wrote: post #157
HPsquared wrote: post #155

Earlier quoted context omitted.

"Deterministic" only insofar as it's repeatable, but their behaviour is not predictable. If the behaviour is not predictable, how can it be validated as being correct?

If for a given set of inputs there is a deterministic output, then the overall behaviour to a series of inputs is just as deterministic and predictable. I'm not sure what you mean with "is not predictable" when you also admit that it's repeatable.

Predictable as in, can say in advance what it will do.

Repeatable means you get the same output a second time, given the same input. Predictable means you can say in advance what it will do, and can then check the output against your prediction.

If you can't predict the outcome, you can't validate the process, and can't guarantee its performance.

Re: ‘Confirming we are cleared to land?’ Who said what at Austin

#232

Post by zaarn »

HPsquared wrote: post #231
zaarn wrote: post #157

Earlier quoted context omitted.

If for a given set of inputs there is a deterministic output, then the overall behaviour to a series of inputs is just as deterministic and predictable. I'm not sure what you mean with "is not predictable" when you also admit that it's repeatable.

Predictable as in, can say in advance what it will do. Repeatable means you get the same output a second time, given the same input. Predictable means you can say in advance what it will do, and can then check the output against your prediction. If you can't predict the outcome, you can't validate the process, and can't guarantee its performance.

That much is clear but any remotely continuous AI model is very predictable.

Re: ‘Confirming we are cleared to land?’ Who said what at Austin

#233

Post by dsfyu404ed »

mcv wrote: post #223
0xFF0123 wrote: post #199

Earlier quoted context omitted.

I'd be curious to know how much it actually improves throughput. It seems strange to use optimistic locks in such a safety critical process, when a more fail-safe approach is available.

Sounds like a money problem to me. Less time between planes means more planes means more money. It sounds to me like far too many critical safety decisions are being made with an eye on money rather than safety. There were reports warning that Fukushima wasn't safe in case of a tsunami, but the company in charge of the plant decided to ignore that. Deep Water was caused by cost cuts. The Boeing 737 Max was made for f…

> Sounds like a money problem to me. Less time between planes means more planes means more money.

Everything is a money/greed problem if you look at it through the lens of motivated reasoning.

The usage of infrastructure is mutually beneficial. The airport, the various businesses operating at it, the customers, everyone benefits per flight. Likewise you want to make that infrastructure get used as much as possible.

Would you be complaining about greed if a subway was running trains close together?

Re: ‘Confirming we are cleared to land?’ Who said what at Austin

#234

Post by dsfyu404ed »

SQueeeeeL wrote: post #156
bilekas wrote: post #144

Earlier quoted context omitted.

I don't knock unions but I will add : > The median annual wage for air traffic controllers was $129,750 in May 2021 SOME unions do take advantage, given the traffic controllers' 'single point of failure' it can be very attractive for some unions who are greedy. Again I reiterate, unions are not a bad idea, just not all of them are solely in the interests of the actual employees.

I personally think stability and having some bureaucrat monitoring hours is good. I don't want my air traffic controllers popping pills and pulling double shifts, or showing up to work drunk because they're worried about getting fired for calling in sick. People with dangerous jobs need to be kept safe from themselves

CDL drivers have all of this without a union.

Re: ‘Confirming we are cleared to land?’ Who said what at Austin

#235

Post by hef19898 »

piva00 wrote: post #230
hef19898 wrote: post #139

Earlier quoted context omitted.

No idea about Austin airport. Regarding professionalism and failure, the Austin incident is actually proof of the system working as intended. There was, by the looks of it, only one real mistake made, maybe two. Training of everyone involved, including the ATC, prevented those mistakes from creating a cascade of failures that could have let to two hull losses and 100+ dead passenger on the Southwest flight. For fatal…

I don't think Austin is proof of the system working as intended, the only thing that saved them was the FedEx flight being able to take action at the last moment during their approach, if the fog was a little thicker and they couldn't see the Southwest flight on the runway as early as they did it would be a catastrophe. That's just luck, not a system that's working as intended.

No, that is good training on behalf of FedEx, Southwest (knowing you have a plane above your own during take off and still follow ATC guidance is no easy feat) and the ATC (knowing you fucked up, staying calm and giving the correct instructions and keep everyone alive is again no easy feat). And training of operators absolutely is part of the overall system.

Re: ‘Confirming we are cleared to land?’ Who said what at Austin

#236

Post by paranoidrobot »

bagels wrote: post #90
paranoidrobot wrote: post #86

Earlier quoted context omitted.

The issue was not "Where were the aircraft". The controller knew where they were at all times. (And they do have GPS, They also have ADSB, and ground based position detection, and, and, and..)

Time 1:54, Tower: “Southwest 708, confirm on the roll?” They don't know if Southwest was moving?

The aircraft were at all relevant times broadcasting their position using ADSB.

The controller had at least one screen displaying their positionx probably several. It may fuse multiple data sources. They may also have been able to see the Southwest aircraft, although maybe not clearly or only as lights.

Despite this it can be difficult to judge the speed of an object at long distances under less than ideal circumstances and at night without a long observation period.

So yeah they knew where they were but not the acceleration.

It's also worth pointing out that the controller may have been in sudden stress due to realising the aircraft were too close, and so may have made a call that they already knew the answer to.

Re: ‘Confirming we are cleared to land?’ Who said what at Austin

#237

Post by mhandley »

helsinkiandrew wrote: post #209
mulmen wrote: post #187

Earlier quoted context omitted.

Europe uses pessimistic runway locks but the US prefers optimistic locks. This improves throughput at US airports with the risk of a rollback on a dirty runway.

I'm not sure why this is a problem with airports with two runways (doesn't Austin have two?). I've been in a helicopter hovering over Heathrow airport where you can see the lights from the line of planes coming in to land every 45 seconds that looks like it goes off to infinity. Whatever technique they use there seems to work.

For the most part, Heathrow doesn't do mixed-mode operation (where you have takeoffs and landings on the same runway). However, my understanding is that this is not for safety reasons, but mostly to give people living under the flight paths some periods of respite from the noise. They did use mixed-mode single-runway operation for a while during covid.

https://www.heathrow.com/company/local-community/noise/opera...

Re: ‘Confirming we are cleared to land?’ Who said what at Austin

#238

Post by dsfyu404ed »

sfeng wrote: post #182
blitzar wrote: post #177

Earlier quoted context omitted.

As a passenger that flys almost exclusively long haul from very busy airports, it is rare that my plane will come to rest on the runway, generally accelerating in the turn onto the runway and taking off.

Did they come to rest on the runway, or did they wait a minute at the side of the runway and then proceed to takeoff? I don't think you would notice the difference as a passenger, unless you were listening in on the ATC transmissions.

As a passenger there's a pretty obvious difference between "swing a turn, stand on the brakes, floor it and full send" and "swing a turn, come to a stop, sit there a minute or so, stand on the brakes, floor it and full send"

Re: ‘Confirming we are cleared to land?’ Who said what at Austin

#239

Post by jnsaff2 »

hugh-avherald wrote: post #228
0xFF0123 wrote: post #199

Earlier quoted context omitted.

I'd be curious to know how much it actually improves throughput. It seems strange to use optimistic locks in such a safety critical process, when a more fail-safe approach is available.

It's worth noting that the ICAO system is not perfect from a safety point of view: it often results in the controller saying 'Expect late landing clearance' instead, and clearance to land very close to the ground. As an ICAO-based pilot, I genuinely don't know which is safer. (And I also don't think the difference was important here.)

I tend to disagree that it was not important here.

At least it would have forced the controller to verify that SW is clear of the runway before clearing fedex.

As it was, it cleared both and then did something else until the situation was over.

Post reply on HN