In 2015, Allegiant Air jets made unexpected landings at least 77 times | Hacker News Reader

In 2015, Allegiant Air jets made unexpected landings at least 77 times

tampabay.com

65 pointsmartin_31 comments

Here's why this article is not FUD.

First of all, everything with mechanical parts needs regular maintenance, and almost always will reach a point where it should be replaced. If you aren't maintaining it, you are opening yourself up to unnecessary risk.

Second, these are maintenance-related issues which are required by law and are not being followed up on, so they're not following the law, even if the FAA claims they're hunky dory. They're clearly not.

Third, when a vehicle like a car has a maintenance or human operator error, a small handful of people's lives at risk. With planes, it's a couple hundred people's lives, and you can't just pull off to the shoulder of the highway to change a flat. Not to mention the things you may hit when you "unexpectedly forceably land".

Planes are designed to withstand multiple serious malfunctions, because there is no alternative: you're 30 thousand fucking feet in the air going 500 miles per hour. So just because there are major failures that have not yet killed anyone is not because there is no problem: it's because planes were designed to be bomb-proof and they are barely holding on to their shakey safety record.

Airline safety is incredibly important because even a single safety issue can cause a large impact in the public's perception of the industry's safety, which affects how many tickets are sold. So from a selfish perspective, you want them to be safe so your ticket price doesn't go up.

Finally, there's the whole point of the article: THE AIRLINE IS TWO TO FOUR TIMES LESS SAFE THAN ALL THE OTHERS, WHAT THE FUCK ARE THEY DOING?!

Is that a lot? Nobody has actually died. Nobody has crashed.

US airlines are incredibly safe as a form of transit. An airline flying older planes will have more issues.

Someone will always have the most issues. I think it's cheap journalism to call them out without an actual compelling reason to think lives are at risk.

They're way outside the expected amount for a US airline. I'm curious what the comparison would look like internationally. But still, part of keeping air travel safe is enforcing safety and maintenance standards.

I'm really glad this newspaper wrote this article! It's exactly the type of investigative story I think can be really valuable.

The "how we did the story" section at the bottom was great. Nice framing of the problem and corresponding data analysis.

The thing most major disasters involving human built things have in common is that it is often a series of failures. An 'alignment of the planets' where everything happens to fall in to place and that string of 1-s comes up on the cosmic die to let you die.

What this article is covering is how much closer, by a wide margin, Allegiant's running to that result than other airlines in the US.

The 'alignment of the planets' is known in certain industries as the Swiss cheese model[1].

[1]https://www.wikiwand.com/en/Swiss_cheese_model

That is an interesting way to describe "defense in depth" to non-technical people that ought to be easy to understand. Thank you.

On certain studied construction sites, it was observed that number of reported incidents correlates negatively with fatality rate. Fig 7.1 on Sidney Dekker's _Field Guide to Understanding Human Error_.

I don't have the book any more, but I believe the idea was that emphasizing number of incidents only led to incident suppression.

If you read the article it seems Allegiant MD-80s have problems 2x times more often than American MD-80s and 3x more than Delta MD-80s. Assuming they did the math right it's a concerning apples to apples comparison.

Also according to this article Allegiant's emergency landing rate is 10x Southwest, which was accounted safest, and over 2x that of American, which had the highest rate other than Allegiant. Again, seems like a disturbing difference regardless of the airplane involved.

Add in the fact that Allegiant's current CEO was the former CEO of ValueJet [1] and this does not seem like a good airline if you want to get home in one piece. They sound like Spanair or Arrow Air. I love aircraft but flying on those two was creepy. They both had serious crashes, as did ValueJet.

[1] https://en.wikipedia.org/wiki/Maurice_J._Gallagher,_Jr.

Wow the landing rate is similar to ValueJet too

"The Federal Aviation Administration's (FAA) Atlanta field office sent a memo on February 14, 1996, to Washington, D.C., stating that 'consideration should be given to an immediate FAR 121 re-certification of this airline' - in other words, the FAA wanted ValuJet grounded. ValuJet airplanes made 15 emergency landings in 1994, 57 in 1995, and 57 from January through May 1996."

source: https://en.wikipedia.org/wiki/ValuJet_Airlines

> Is that a lot? Nobody has actually died. Nobody has crashed.

"Not killing passengers": that seems a rather primitive metric, which for air transport I think we left behind in the 1950s when turbine engines brought a magnitude leap in reliability and power.

Nowadays the primary differentiators amongst ( FAA and JAA-certified ) airlines are punctuality, service, schedule and price. An airline that consistently and significantly fails on one or more of those, as Allegiant appears to be doing, will not last long in the deregulated aviation market. So in that sense it is valuable journalism to bring their poor reliability to light.

> Is that a lot? Nobody has actually died. Nobody has crashed.

So we should wait until ~100 people die in a plane crash before addressing the obvious problem? I'm not following your logic at all.

Oh, and actually, thousands of people have died in plane crashes before. Just not on this airline (yet). But there's no reason that the same shoddy maintenance practices that killed thousands of people on other airlines can't also kill people on this one.

The article makes it sound like their planes are failing all the time, while the statistics say that Allegiant has 12 failures per 10000 flights. That sure is more than any other American airline, but it is not nearly as alarming as this article is describing it.

Add to that that there was not a single fatality, and you can safely conclude that this article is alarmist FUD, nothing more.

Also the repeated use of "break down in mid-flight" irritates me. To me it reads too much like "break up in mid-flight", as in, the plane breaks into small pieces.

> there was not a single fatality

This strikes me as the same reasoning that Feynman criticized in the Rogers report:

> The argument that the same risk was flown before without failure is often accepted as an argument for the safety of accepting it again. Because of this, obvious weaknesses are accepted again and again, sometimes without a sufficiently serious attempt to remedy them, or to delay a flight because of their continued presence. [1]

Just because no one died doesn't mean that the practice is safe. It might be safe, just like it could have been with the shuttle, if the engineers had designed the o-rings to withstand erosion. In that case you'd have tests and a spec for how much erosion is safe.

Similarly, unless there's a guideline on acceptable rate of mechanical failures (which the article suggests there isn't) "there was not a single fatality" is not an acceptable proof of safety.

[1]: http://science.ksc.nasa.gov/shuttle/missions/51-l/docs/roger...

The airline industry functions as a whole. One bad actor can seriously hurt the whole industry. With such a poor safety record (2 - 3 times worse than any of the others), they are not credible. There is a danger that if this is not taken care of, others may follow suit and this will become the new normal.

It is very strange that the FAA is so serious about on time statistics and enabling comparisons between airlines on that basis, but not on this basis.

If we are looking at the airline industry as a whole, wouldn't the worst offender always be a certain number of standard deviations from the norm? I don't think we can make a judgment unless we can see a graph that represents failures per 10000 flights amongst the industry as a whole, and see if Allegiant falls within the bell curve that produces.

I don't think so. It's a dynamic system, not a static one. If one actor gets away with it and is more successful, then others will follow. The regulator's responsibility is to minimize safety critical differences between the carriers by getting the worst performers to improve or leave the market.

If you think about it, there's a really good reason why you might not want to punish airlines for making emergency landings.

Not one by one, no. But if your airline makes emergency landings at more than twice the rate anyone else does, it's hard to complain when someone asks questions about that.

> But industry observers say there’s a reason most air travel is so expensive. It’s difficult both to offer great deals and spend the money needed for a reliable fleet.

In Europe it seems to work ok for low-cost airlines. Compared to state-airlines the aircraft are newer and cleaner, and they aren't know for reliability issues. As a customer that suggests they are better maintained - and nobody is going to prop them up if something goes wrong, so they have more incentive to do so.

You can check here how they manage to do that https://www.youtube.com/watch?v=069y1MpOkQY

It says they fly the MD-80, also known as the mechanical disaster.

This is off topic, but A fascinating thing about these planes is the control surfaces are aero powered through tab surfaces. They hang and move freely when the plane is stopped and only become controllable after the plane starts rolling down the runway.

I've looked at several sources now, and still haven't got a clear mental picture of exactly how that works. I suppose it must, since the aircraft don't usually fall out of the sky, but it's hard to see a reason why such complexity would be necessary.

Tabs are a way to reduce the amount of stick force (i.e. force applied by the pilot) required to deflect a control surface in a mechanically linked system. In most modern aircraft hydraulic or electric systems fill the same purpose

To give a simplified explanation, a small tab located at the trailing edge of the surface is deflected by the control linkage. The deflection of the tab generates an aerodynamic force which then creates a moment around the control surface hinge, moving the entire surface.

So if you deflect the tab upwards, that air flow over the tab generates a downwards force that acts to move the control surface down.

And another layer of simplification: the pilot controls small control tabs which fly the main control surface into the position the pilot wants it to be in, which then moves the wing.

Usually the connection between pilot intent and wing movement is:

Pilot moves the stick->Aileron moves, changing lift on the wing->wing moves

In the MD-80:

Pilot moves the stick->Tab moves, changing lift on the aileron->aileron moves, changing lift on the wing->wing moves

This is probably the worst thing to read the day after buying plane tickets for my family to fly Allegiant Air down to... Tampa.

Allegiant.

Thank you for this. I went and read the article anyway, but this title is absolute garbage.

I didn't make anything of it but after reading your comment I'm guessing Allegiant is an airline with frequent trouble? Because the word doesn't mean anything by itself in this context that I can see.

4 downvotes, so apparently it means something to everyone but me.

Reply on news.ycombinator.com