Hacker News Reader

> clearly that wasn't so: they were just making things up as they went along.

Where did you get your information? I have experience in the industry and scheduling logistics is clearly not how you describe it. The issue is that to optimize for profit you sacrifice the ability to maintain service through catastrophic events and can end up in a bit of a dominoes situation.

Which information? That they belated realized they'd run out of duty hours and had to call in new crew, after the plane with the soon-to-expire crew had landed? They told us that while we waited at the gate, including updates about the expected time that the newly-called-in crew would arrive. They were quite transparent about everything.

Or are you asking why I think they're making it up as they go along? That is my conclusion, to be sure. It's one thing for dominoes to fall, but if you are in a situation where the dominoes are falling and you are not able to predict which dominoes will fall next and respond accordingly, you are making things up as you go along. I'd have expected that almost any decision they could even hypothetically make would run through a system that checked it for violations of constraints, which would have told them way ahead of time that they needed a fresh crew, and they'd have had the entire flight of the replacement plane to get one in (IIRC it flew Miami->Boston just to get us the plane).

Regarding your first paragraph, rarely are the gate agents telling the public the entire story. The truth is typically more complex and basically not of concern to the average passenger.

As for your conclusion, it's just not accurate. Systems _are_ used for scheduling and allocation of equipment and crew resources. You just have to realize that getting you, as an individual, to your destination on time on any given day isn't the number one priority of the company. There are a lot more concerns they are considering. If the airline industry is really as dysfunctional as you think, with such shortcomings in areas as important as equipment and crew scheduling and allocation, you should drop whatever you're doing and start an airline. You'll soon be filthy rich and do the world a great service.

That’s just hand-waiving. Sure, they could have been lying to us and had some hyper-competent, complex reason that delayed us and then made up some malarkey about duty limits. But you don’t know either, and the explanation they presented seems vastly more likely.

Also your generalizations fly directly in the face of the information emerging from this Southwest fiasco, which indicates they don’t have systems in place to even track where their crews are (just where they are scheduled to be), so the shortcomings seem to be quite real.

Couldn't the emergency flight have run into some problems, making it take longer and that led to the crew being out of duty hours? Like the flight was supposed to take 3-4 hours but wind and some airport issue caused it to take 5? Something that may not have been obvious until very late in the flight too.

>The issue is that to optimize for profit you sacrifice the ability to maintain service through catastrophic events and can end up in a bit of a dominoes situation.

Sounds like an argument for nationalization.

I don't follow. Nationalization might add more slack, and more resilience, but it'll be so much more costly to operate.

Airlines run on very thin margins, so need to optimize heavily.

We just need to accept that we can either get "affordable" tickets with occasional meltdowns or we go back to 30 years ago and pay double price for tickets but have more resilience.

The market as a whole has already resoundingly chosen the former.

Safely flying humans in metal structures is just really expensive and the current ticket prices pretty much reflect the real cost (and even ignore the climate externalities)

> I don't follow. Nationalization might add more slack, and more resilience, but it'll be so much more costly to operate.

Who cares? The point of nationalization is costs don't matter, we just see it as a cost we bear because transportation is important for civilization. The primary mode (for better or worse) in the US is public (roads) and that "cost" is not covered by anyone.

> Sounds like an argument for nationalization.

No, it sounds like an argument to never fly SWA again.

I’ve seen it happen and it’s always strange that they seem to not realize the crew will be unusable until they arrive … I assume they had been trying to get another crew at the same time.

There are several time limits, but I believe that one of them stops when pulling back from the gate. i.e. once the the plane is moving, they're granted several X more hours, but if they haven't left yet, then they've timed out. So the arriving crew may plausibly be usable upon arrival, but a few minutes late may be all it takes to time them out.

I've been on a delayed plane where one of the pilots timed out while sitting at the gate, so we had another second delay finding another pilot.

There's also the fun wrinkle that not all crews or crew members are current on ratings (ie training courses) to operate all types/versions of planes the airline as in service.

You also need sufficient open gates of the right type in the right place during the right time window or you idle a plane (which can then idle a crew). It's pretty easy to imagine how such a system which may work at a certain level of load can be tipped into a runaway scenario.

As a systems person I'd be an interested reader if Southwest were to someday release an incident analysis and post-mortem of the type that good IT organizations do.

The first is why Southwest (and others) only fly one type of plane.

Sometimes the dispatchers expect that the crew will still be legal, but then the incoming flight gets delayed just enough to push them over the limit.

Once you realize how many people in critical industries (power grid, telecom, global cargo/logistics) are in fact making things up as they go along, and there aren't really any highly organized and responsible people running the show, you start to worry.

Reminds me of this Onion headline: "Smart, Qualified People Behind The Scenes Keeping America Safe: ‘We Don't Exist’"

https://www.theonion.com/smart-qualified-people-behind-the-s...

What is the alternative? Extensive playbooks for every possible scenario that may or may not make sense in the actual scenario because a few critical details are different? A "mission-control center" is really no different: it's just people making stuff up on the spot.

In the end, there is no substitute for human judgement in the situation itself.

> What is the alternative? Extensive playbooks for every possible scenario that may or may not make sense in the actual scenario because a few critical details are different?

Yes? I suspect they have playbooks and/or run gameday exercises for other aspects of disaster preparedness and business continuity. Why not do the same for these sorts of issues that affect the travel network?

The network of airports, planes, and crews is a graph (or multiple graphs). What happens when a node disappears?

That may be a hard question to answer, but it seems cheaper to answer, and plan around, it before the situation occurs rather than after.

Or you realize a lot of things involving real world physical systems are hard and customers aren't fine with paying "whatever."