Airline Safety Report

[Parker_Woodruff]

https://www.iata.org/en/pressroom/pr/2021-03-25-01/

With a fatality risk of 0.13 for air travel, on average, a person would have to travel by air every day for 461 years before experiencing an accident with at least one fatality. On average, a person would have to travel every day for 20,932 years to experience a 100% fatal accident.

 

Enough said.  This is quite incredible.

[aviatoreb]

1 hour ago, Parker_Woodruff said:

https://www.iata.org/en/pressroom/pr/2021-03-25-01/

With a fatality risk of 0.13 for air travel, on average, a person would have to travel by air every day for 461 years before experiencing an accident with at least one fatality. On average, a person would have to travel every day for 20,932 years to experience a 100% fatal accident.

 

Enough said.  This is quite incredible.

Maths police here.  Pull over sir.  Can I see your license and registration for operating those statistics?

There's a few things wrong with that sentence you cut from the article.

Not least of which is the English.  What is a 100% fatal accident?  Is that better or worse than a 200% fatal accident?  Is 50% fatal accident one where you walk away only half dead?

Besides that...

In this kind of computation you never ever REACH 100% fatality expectation over the time epoch.  It is more like you asymptote toward 100% ever closer but never reach it.   That is them most serious problem with the statement.   A bit more nitpick - And besides that - this is in expectation across a population, but for an individual person its Boolean - either you experience a fatality or you don't.

Finally I can't do the computation myself to check easily what they mean by what is happening at 20,932 years since I don't know what are the units on that number 0.13.  Is that probability of fatality per flight?  (no way that's too high), or probability of fatality per year? (still must be too high).  Per million hours of exposure?  I bet that's it? Per million miles?  

[ZuluZulu]

1 minute ago, aviatoreb said:

Maths police here.  Pull over sir.  Can I see your license and registration for operating those statistics?

There's a few things wrong with that sentence you cut from the article.

Not least of which is the English.  What is a 100% fatal accident?  Is that better or worse than a 200% fatal accident?  Is 50% fatal accident one where you walk away only half dead?

Besides that...

In this kind of computation you never ever REACH 100% fatality expectation over the time epoch.  It is more like you asymptote toward 100% ever closer but never reach it.   That is them most serious problem with the statement.   A bit more nitpick - And besides that - this is in expectation across a population, but for an individual person its Boolean - either you experience a fatality or you don't.

Finally I can't do the computation myself to check easily what they mean by what is happening at 20,932 years since I don't know what are the units on that number 0.13.  Is that probability of fatality per flight?  (no way that's too high), or probability of fatality per year? (still must be too high).  Per million hours of exposure?  I bet that's it? Per million miles?  

What are you, some kind of math professor? 

[aviatoreb]

Just now, ZuluZulu said:

What are you, some kind of math professor? 

I'm not a real doctor, but I play one on tv.

But I am some kind of math professor.  Good, bad, or ugly, I shall let my teacher evaluations speak for themselves.  Just don't forget your units because they are important.  And label the dang axes on your graphs or no one will know what you are trying to say!

[carusoam]

Not so long ago... airliners used to drop from the skies on an annual basis..

• Drop an engine from your DC10 on T/O... WnB problem in Chicago...
• Have your engine not produce full power with ice on your wing and land in the river.... in DC...
• have the rear engine self destruct and wipe out the hydraulic system... somewhere around Kansas...
• Fly on the east side of Lake Erie on an icy day.... make the news in Buffalo...
• Have your center fuel tank explode after leaving NYC...
• Run out of fuel over Queens on your way into NYC...
• Depart NYC and some how knock the rudder/tail off your plane...
• Land long in Boston on an icy runway... slide into the ocean... the cockpit broke off and a few first class seats went missing...
• Load old O2 generators in the baggage area near the cockpit... have them start a raging fire... Lawn dart in the Everglades...
• terrorism world wide, under various circumstances, including pilot suicide in the alps... bombing over Lockerbie 
• Somebody parked a jet on a levee near New Orleans... after twin engine outs...

 

Aviation Memories that weigh on my mind...  From which we have gained a ton of knowledge about what to avoid... what to do next... and why we train... and discuss the wackiness of aviation.... 

All stuff I remember from the days when it happened... none of it looked up... so it may have some errors...

 

Procedures improved... piloting and maintenance....

Training improved...

Much more is known about the weather... from inside the cockpit...

Planes have improved...

Conditions have improved, pilots are actually people...

Instruments have improved...  we know where all the mountains are now... and where we are...

The latest airbus and Boeing sensor issues are quite the reminder of how a simple sensor error can cause a disaster...

Terrorism has been cut way back...

The upside... Sully vs. the Geese... even when the plane gives up...some pilots have the skills to make a safe outcome... 

Safety culture has come a long way...

PP thoughts only, not a statistician... a great memory exercise....

Best regards,

-a-

[1964-M20E]

6 hours ago, carusoam said:

• Somebody parked a jet on a levee near New Orleans... after twin engine outs...

 

 

The good thing about this one is everybody walked off the plane and they flew the plane out a few weeks later.

[Mooneymite]

15 hours ago, ZuluZulu said:

What are you, some kind of math professor? 

.    The rest of us just round-up, or round-down at the end of the asymptote.

When I board an airliner, I figure zero chance of a fatal crash is a good number.

Does anyone remember the flight insurance booths that used to be in every airline terminal?  They, like the chance of crashing, are now "approaching" zero.

Edited March 26, 2021 by Mooneymite
Punctuation. Spelling.

[jlunseth]

While it is true that the commercial airlines have come to a point where fatal accidents - indeed any kind of accident - practically never occurs in the US fleet, that is not us, not GA.  They now go for years without a fatal accident. GA is lucky to go a week without one, or even a day. It also is not true of the non-US commercial fleet.

[steingar]

They used to say the most dangerous part of the flight is the drive to the airport.  They should keep saying it, because it's true.  I have to give it to them, the airlines are really astounding.

[0TreeLemur]

2 hours ago, steingar said:

They used to say the most dangerous part of the flight is the drive to the airport.  They should keep saying it, because it's true.  I have to give it to them, the airlines are really astounding.

Totally agree.   The convergence of materials knowledge, reliable manufacturing, quality control, maintenance, and safety culture has made fatal 131 ops a near impossibility in a statistical sense.  Until it isn't.

[MooneyMitch]

2 hours ago, Mooneymite said:

Does anyone remember the flight insurance booths that used to be in every airline terminal? 

Gee whiz...... you must be as old as I am!   

Yes, as a kid, I do remember seeing those machines.  I’d try to stop and read the information on the machine and my mom would drag me away....... funny, I thought back then......this is going to be really thrilling, but it must be really dangerous if they’re selling you life insurance just before you board the plane ! 

[aviatoreb]

2 hours ago, steingar said:

They used to say the most dangerous part of the flight is the drive to the airport.  They should keep saying it, because it's true.  I have to give it to them, the airlines are really astounding.

That's right - the drive to the airport is the highest risk phase of the trip.

It is really remarkable with good procedures, practices, engineering on so on that hurtling through the air at 500mph on contraptions full of stressed parts, complicated interactions and so on, is indeed this safe, but it is.

[aviatoreb]

To my point - I wasn't just being pedantic - does anyone know the units on that 0.13 number?

[Mooneymite]

30 minutes ago, MooneyMitch said:

Gee whiz...... you must be as old as I am!   

Yes, as a kid, I do remember seeing those machines.  I’d try to stop and read the information on the machine and my mom would drag me away....... funny, I thought back then......this is going to be really thrilling, but it must be really dangerous if they’re selling you life insurance just before you board the plane ! 

I remember those booths being staffed by "concerned professionals".  They watched us pass by like we were on our way to the gallows.  Little did we know that the insurance booths would soon be gone, but the TSA would be with us forever!

Don't we all feel much safer now?  

Edited March 26, 2021 by Mooneymite
punctuation.

[MooneyMitch]

7 minutes ago, Mooneymite said:

  Little did we know that the insurance booths would soon be gone, but the TSA would be with us forever!

Don't we all feel much safer now?  

Nothing lasts forever............I hope!

I was on a wonderful motorcycle cross country trip after September 11.  I was in Washington DC, visiting friends, the Smithsonian Air and Space Museum [and the museum's Garber restoration facility.......fascinating place!], and our Capital [back when we could].  I happened to be sitting upstairs in the house chamber when a congressman [don't recall which one] took the floor and suggested the government should take over airport security...................oh my, I thought !

[whiskytango]

30 minutes ago, aviatoreb said:

That's right - the drive to the airport is the highest risk phase of the trip.

It is really remarkable with good procedures, practices, engineering on so on that hurtling through the air at 500mph on contraptions full of stressed parts, complicated interactions and so on, is indeed this safe, but it is.

Your chances of dying in an airliner crash are almost the same whether you get on the plane or just watch from the terminal as they pull away from the gate.  Kind of like winning the lottery jackpot - your chances of winning don't change too much if you don't buy a ticket 

[Hank]

1 hour ago, aviatoreb said:

To my point - I wasn't just being pedantic - does anyone know the units on that 0.13 number?

One thing I've noticed about Safety professionals:  they calculate and post lots of numbers, almost always with no units. I asked once about the Accident Rate at a previous employer. They posted the monthly accident rate, and graphed it against our goal. With no units. Upon enquiry, the lenghty explanation of why there were no units and what the number actually meant when applied across 1400 people working 40-50 hours per week left me confused and unsatisfied. It boiled down to the obvious, the more accidents we have, the higher the number goes. But they could not explain the relationship between the "accident rate" number and how many man hours were worked between accidents, or how many accidents we had per 100,000 [or 1,000,000] man hours . . . . .

[Mooneymite]

Just now, Hank said:

.... left me confused and unsatisfied. . . .

But, Hank, statistics don't lie!

[Hank]

6 minutes ago, Mooneymite said:

But, Hank, statistics don't lie!

"There are lies, damned lies and statistics." --Samuel Clemens

The fuel burn rate of my Mooney is only 0.04. What's yours?

[Mooneymite]

9 minutes ago, Hank said:

The fuel burn rate of my Mooney is only 0.04. What's yours?

Much less than yours according to MY statistics!  Try some MMO.   Run lean of peak and install bladders.  

[aviatoreb]

21 minutes ago, Mooneymite said:

But, Hank, statistics don't lie!

Its true - statistics DO NOT lie.  But people often misunderstand and misuse statistics and sometimes they even do it deliberately but more often inadvertently.  Often what is suggested by showing statistics is not what the data is really showing or what the statistics is really showing.  That is human incompetence but the maths fault.

Samual Clemens was a very very clever writer and there is a germ of truth in this specific quote "There are lies, damned lies and statistics." --Samuel Clemens - and it is not utterly the deep truth.  It is almost true but it is actually false.  Statistics are not lies.  They are truth but it still requires a skillful statistician or data scientist to tease the truth out of the data, and a skilled presenter to properly present that truth.

Presenting numbers without units is utterly useless.

Statements like "20,932 years to experience a 100% fatal accident" is a instantly self evidently false and erroneous statement with a basic misunderstanding of the entire concept.

[aviatoreb]

16 minutes ago, Hank said:

"There are lies, damned lies and statistics." --Samuel Clemens

The fuel burn rate of my Mooney is only 0.04. What's yours?

I am going to go to the Moon.  It is only 1 away.  And I should get there in about 1.  And It should cost about 1.  (Nondimensional units - which still have the physical units attached to have meaning) (1 Moon to Earth distance unit = ~250,000 miles) (1 time to get to moon unit = 2 months in a 172 flying nonstop at 60miles per hour). (1 is the cost of the project which converted to dollars - I dunno).

[aviatoreb]

29 minutes ago, Mooneymite said:

But, Hank, statistics don't lie!

37% of them do.  37.6% to be precise.

I am a mathematician and we are much more honest.  Taller, smarter, nicer, funnier, handsomer, but not as well paid. Go figure.

[aviatoreb]

34 minutes ago, Hank said:

One thing I've noticed about Safety professionals:  they calculate and post lots of numbers, almost always with no units. I asked once about the Accident Rate at a previous employer. They posted the monthly accident rate, and graphed it against our goal. With no units. Upon enquiry, the lenghty explanation of why there were no units and what the number actually meant when applied across 1400 people working 40-50 hours per week left me confused and unsatisfied. It boiled down to the obvious, the more accidents we have, the higher the number goes. But they could not explain the relationship between the "accident rate" number and how many man hours were worked between accidents, or how many accidents we had per 100,000 [or 1,000,000] man hours . . . . .

Here is my favorite gag on the concept of units conversion:

 

[whiskytango]

Or this: