Accident investigation and PPRuNe (Discussion)

[bsieker]

Originally Posted by chuks

Even the very high-profile Concorde crash led to allegations that maintenance errors caused a critical mis-alignment in the landing gear that went ignore

Have you read the report? It's still on the BEA website. It's rather long, though.

The incorrect re-assembley of the landing gear bogie is discussed in detail, examined as to its causes (not following maintenance procedures, incomplete work-shift handover, sloppy management and non-use of specialised tools, ...), recommendations are given about improving maintenance procedures, etc. But it was ultimately found not to have been a cause to the accident.

This is confirmed by my Why-Because-Analysis (after a lengthy discussion with a professional expert, correspondence in the appendix) which I performed for my Diploma thesis (to be found at http://www.rvs.uni-bielefeld.de -> Publications -> Theses -> "Visualisation Concepts and Improved Software Tools for Causal System Analysis", Part II. Direct Link to the PDF.)

There are no indications in the flight data recorders that there was any undue influence to aircraft performance from the landing gear. There was no unusual sideways force, and acceleration was normal, until it hit the titanium strip.

[...] in the case of the Challenger accident. But bear in mind that there you had a brilliant mind at work [...]

I'm not saying there is "another Feynman" on this thread, but some of the people so indiscriminately labeled (or should I say "libeled"?) as "amateurs" may be brilliant minds, and perhaps also backed up by years of analysing aviation accidents and socio-technical avionics problems, having written peer-reviewed and -acclaimed papers about the matter.

[...] widget factory goes bankrupt, is it entertaining to try and guess what mistakes the Chairman made to cause that to happen?

I guess it might be, although perhaps "entertaining" is the wrong word. Interesting or enlightening might be more like it. But it it will be a lot harder, because people are so much less deterministic than technical systems. We have speculated about the idea of applying the Why-Because-Analysis concept to purely social/socio-psychological phenomena, but never pursued it wholeheartedly.


Bernd

[MrNosy2]

I see nothing wrong on suggesting possible scenarios about what might have happened based on known facts if the person doing it knows what they are talking about. An official report (if any) might not normally be released for, say, 18 months after the event so it is useful to discuss what is known about accidents in advance of the final report. Its not just this forum where this happens - reviews of recent accidents are carried out by a number of organisations, e.g UK Flight Safety and, especially IATA who host a meeting of about 15 to 20 airline safety managers every year at the start of January to review the accidents during the previous year and try to see what lessons might be learned.

The main problem with this accident is that, as far as I'm aware, the Turkish authorities have not issued any statement of facts (unless its tucked away in Turkish somewhere?). Compare this, for example, with the situation in the UK where the AAIB might well have been expected to have issued a Special Bulletin on such an accident if it happened in the UK by now containing the known facts.

[lomapaseo]

also give a thought that it is rare for a detailed discussion or analysis to be published where operational errors may have contributed to the accident chain.

After all the most important things to the rest of us are product updates , or manual updates. These are almost always put forth even without the benefit of a long drawn out report process. Things that are unique to an internal process (regulatory compliance, operator training sylabus, etc. etc.) rarely get published or at least take a much longer time to appear in the public domain.

From my view, most accident investigations are well served to get the right words out quickly to prevent future accidents if you read between the lines of the first months information following an accident.

Now if it's blame or finger ponting that you are looking for, don't hold your breath.

[Clarence Oveur]

Now if it's blame or finger pointing that you are looking for.....

It seems to me, reading this and other well visited threads, that there have been a recent influx of people who see a purpose in just that. Included are, rather disappointingly, people who's credentials initially suggested otherwise.

[twistedenginestarter]

One of the most fascinating aspects pf PPRuNe is the flood of input and analysis after a major accident. Sometimes the likely causes are identified fairly quickly. In the case of the MD-83, it's not so obvious. However you always get people saying we are not allowed to speculate. As with this thread, often there's a theme that what you say is automatically worthless if you are not a certain type of professional. But that's not the real issue. These people have a sort of religious aversion. Nobody needs to read these threads. They know that there's going to be no official enquiry findings published 12 hours after a crash, so why are they browsing? Also the curious thing is these discussions are mainly credible. You can't expect 100% but most people are making rational contributions or asking pertinent questions. Of course there's a lot of missing facts especially in the first hours and days, so it is speculation, but what is wrong with that? We'd all be very dull if we ignored things we didn't know were 100% factual, and what exactly would be achieved?

[lomapaseo]

... You can't expect 100% but most people are making rational contributions or asking pertinent questions

subjectively from my view, way too many are more willing to expound than to ask questions that might provide a tad more credibility to their contributions.

Those that do offer contributions with pointers at appropriate followup questions for fact finding are the easiest to read and ponder as well as to evaluate as new facts are developed.

[DingerX]

I had a pompous, academic-sounding dull contribution, complete with references to the 1854 amateur investigation into the Soho cholera outbreak, but thought better of it.

90% of all contributions are questionable, and they don't discriminate by profession. Every job involving professionals has non-professionals who go about telling the pros how they should do their job. That Widget-factory owner has his fan club too -- just check the minutes of the annual meeting, or his stock forum. If you want to know how bad it can really get, listen to a Sports radio talk show.

Any intellectual endeavor that involves sifting through data should welcome additional (and non-standard) sources of data as well as interpretations, as long as it bears in mind that most data are ultimately meaningless, and most interpretations wrong. But even the act of refuting the most absurd WAG is useful: it forces a reconsideration of the basic situation. So you can change the composition of the debaters, but they'll still be spouting crap most of the time.

Pros do lots of things well, but they don't have all the resources amateurs have. While pros must work systematically through a field, the amateur only works on the stuff s/he finds stimulating. When that "stimulating stuff" is not equal to "Working systematically through the field", the results are necessarily skewed, and the amateurs may not have the "big picture". But that's not necessarily bad. Any archaeologist commencing survey of an area knows that, in addition to lining up everybody for a grid-by-grid study, it helps to ask the locals if there's anything around they should know about.

What doesn't help is when someone makes an appeal to their own authority where a) that authority does not give them competence to speak on the case at hand, or b) how that authority gives them competence to speak on the case is not clear or specified, or c) they don't have the authority they claim for themselves.
a) "I've flown Cessna 172s in MSFS. There is no way MD-83 landing gear will fail to come down"
b) "As a professional pilot, I can say unequivocally that no pilot would land in those conditions."
C) "I have flown the Concorde on transatlantic sectors, and once to the Antarctic Research Station and Alien Detention Facility in McMurdo Bay"

The Latin word for professional guild survives, although its scope has been reduced to only one institution: universitas

[hetfield]

What a great posting!

Congratulations.

hetfield

[alf5071h]

Speculate to accumulate

I have previously argued in this forum about the advantages of informed speculation about incidents and accidents. Also there has been extensive debate on why and how to control the unwarranted contributions from the non-professionals. However, there still remains opportunity for well meaning contributors to reflect on the manner and content of their posts.

Speculation can provide both opportunities to learn new or refresh aspects of flight safety, but also practice individual thinking and analysis skills which are essential components of operating safely (via CRM), i.e. keep an open mind, see the other persons point of view, awareness of human error and limitations, etc.
Both learning and depth of knowledge thrive on experiencing situations first hand; incidents and accidents is one area that this should be avoided, thus we have to learn from others. ‘Day dreaming’ or visualisation is also a powerful learning tool associated with speculation, this enhances experience by adding patterns of thought and solutions for use in future unexpected situations by considering ‘what if’, but supported by sound knowledge.
‘What if’, is a powerful investigative tool for incidents and accidents, and also for examining our own thinking and behaviours.

Many posts engage in meaningless exchanges of ‘accuracy’ or ‘I know better’, whereas the effort might be better spent in expanding knowledge or researching issues particularly for posts which commence ‘I believe … ’.
In aviation, if you don’t know then your belief is very tenuous (often worthless), and thus it is essential to find out, gain greater understanding, double check, or compare data, before pronouncing. Similarly, Captains of the industry (crews and in forum threads) should lead through example by encouraging and educating with measured posts and responses – some of the other contributors might be crewing with you shortly.

Forum speculation offers opportunity to gain experience and practice skills of thought and communication. This opportunity should not be abused. All of the above requires self discipline, which is of course is the foundation of airmanship.

What Can You Learn from Accident Reports?
A tool for Human Factors Accident Investigation, Classification and Risk Management.
Systems safety in the wake of the Cave Creek disaster.

[twistedenginestarter]

Isn't It Ironic?

a) "I've flown Cessna 172s in MSFS. There is no way MD-83 landing gear will fail to come down"
b) "As a professional pilot, I can say unequivocally that no pilot would land in those conditions."
C) "I have flown the Concorde on transatlantic sectors, and once to the Antarctic Research Station and Alien Detention Facility in McMurdo Bay"

I don't recall anything like this being said and I didn't find anything via Search. Ironically you're criticising others yet justifying yourself with non-facts. If some such such non-sequitur appeared in one of these threads, you should quote that. This does two things - i) it potentially proves such a thing happened, whilst b) allows us to challenge your analogy.
Whilst it's true that professional airline pilots will know some things that armchair enthusiasts won't, the same is not true of logical analysis. I have two science degrees - Am I better than you at analysing theories? I don't know. You can't call it. However people like you want to. You want to say that only certain unspecified 'professionals' are allowed to theorize. It would be interesting to know what defines those professionals, incidentally. Being a pilot is not a qualification for analysing accidents although obviously it helps.
Try this analogy: pilots are like criminals, they have to use skill, experience and daring to do things in real time. Investigators are like policemen. They are slow and careful and have all the time in the World to think things through. Now tell me the only people who can be policemen are criminals or ex-criminals.

[DingerX]

Those are intentionally not statements actually made, but examples of fallacious appeals to authority, and we could dig up on these threads examples of all three forms quite easily.

My argument is quite the opposite of what you suppose: logical arguments are valid or not regardless of who makes them. That hasn't changed for a couple thousand years.

And sure, I agree that when it comes to "accident investigations", pilots are amateurs. They have (or should have) an interest in them, and they are extremely familiar with a segment of it. But I think you've got the temporality of cops and criminals mixed up. The analogy fails.

[Clarence Oveur]

At the risk of sounding all Rumsfeldian, there are those who do not know, how much they do not know. Consequently they believe themselves more knowledgeable than they actually are. Add to that a need, or desire, to be seen as an equal or accepted to a certain group, and you have the recipe for conflict.

The sound advice offer by alf5071h should be kept in mind by all.

[AirRabbit]

Whoa there ...

Originally Posted by Mad (Flt) Scientist

Study the composition of the NTSB - that is the BOARD MEMBERS - and you'll find that ultimately, although they rely on specialists to make the investigation, the actual board members who have the final say are, in fact, "amateurs" - and often bring an outside perspective to the final report.
edit: took some finding...
Current Members of the Board are:
* Mark V. Rosenker, Chairman
* Robert L. Sumwalt , Vice Chairman
* Deborah A. P. Hersman, Member
* Kathryn O'Leary Higgins, Member
* Steven R. Chealander, Member
Links to their bios are here
None are "trained accident investigators" by any stretch of the imagination (two were pilots, but that's not the same thing, obviously)
As to "why is aviation picked on" - it isn't. Ask any, say, nuclear industry employee if he thinks they are immune to "armchair experts". Or the military in general. Or the police forces. Or ....

In view of your statement “None are ‘trained accident investigators’ by any stretch of the imagination (two were pilots, but that’s not the same thing, obviously);" here is a quote from the “bio” of Robert Sumwalt.

A trained accident investigator, Mr. Sumwalt participated in several NTSB investigations including the crash of USAir flight 427 in 1994 near Aliquippa PA, and USAir flight 861 near Birmingham Alabama in 1998. He also participated in the Transportation Safety Board of Canada's investigation of the accident involving Swissair flight 111 off the coast of Nova Scotia in 1998.
From 1991 to 1999, Mr. Sumwalt conducted aviation safety research as a consultant to NASA's Aviation Safety Reporting System, studying various issues including flight crew performance and air carrier de-icing and anti-icing problems.
Mr. Sumwalt has co-authored a book on aircraft accidents and he has written extensively on aviation safety matters, having published over 85 articles and papers in aviation trade publications and he has broad experience in writing aircraft operations manuals and airline and corporate aviation policy and procedure guidelines. He has been a regular contributor to Professional Pilot magazine.
In 2003, Mr. Sumwalt joined the faculty of the University of Southern California's Aviation Safety and Security Program, where he was the primary human factors instructor.

[lomapaseo]

Lastly, we've never established any "qualification" for an "investigator": everyone shows up, with their own perspective and background. Often the paid "investigator" is the least qualified of the group (workers from the industry).

This is difficult to understand

Who is the "we" in the above

What's a "paid" investigator vs an "unpaid" investigator

If you refer to ICAO annex 13 for guidance than "expertise" is the only guidance that I can think of and even there it does not distinguish what kind of expertise. Mostly what is available are a collection of enough expertise to get the job done and in that the most critical expertise are those that know what they do know vs what they don't know.

Recognizing what you don't know and and then seeking help as needed is what gets the job done. If by industry folks you mean the manufacturers then they often serve as conduits via telephone to sources of specific expertise rather than as all around investigators themselves.

[Mad (Flt) Scientist]

My apologies to Mr Sumwalt; I admit I scanned the bios quite quickly.

My statement was based on knowing the background of some of the others on the board, having noted their various comments during the Pinnacle/Jefferson City board meeting ... the quality of which didn't seem particularly dependent upon their specific background.

In addition, I think it's quite important that the board be seen not to be a creature of the industry; it is, after all, one of the mechanisms by which the general public assures themselves of the safety of the transportation system, and if it were staffed entirely by industry insiders, there'd be even more people willing to cry "conspiracy" at every turn. It does, after all, fulfill a quasi-judicial/legal role, and we don't apply any expertise qualification to a jury on a capital trial....

[krujje]

It does, after all, fulfill a quasi-judicial/legal role, and we don't apply any expertise qualification to a jury on a capital trial....

MadFS, are you saying that the quality of NTSB board members is reflected by their inability to avoid doing that particular duty?

[krujje]

PJ2:

Notwithstanding Feynman's example of "amateur sleuthing", the best example of a non-specialist doing superb investigative aviation work is the Moshansky Commission and it's Report in the accident at Dryden, the recommendations of which were largely though not wholly adopted by Transport Canada, (which was heavily criticized in the Report, and to which strong criticisms have since been directed regarding SMS (which I have called "the de-regulation of flight safety"), by the same man). Here was a man who knew nothing about aviation yet had tremendous access to wide-ranging information, all with good results.

Moshansky was not a man who knew nothing about aviation... he was in fact a licensed pilot and aicraft owner. Furthermore, I would argue that this was in fact an example of an expert doing a superb investigation... maybe not an expert in aircraft accident investigation, but certainly an expert at reasoning and sifting through tremendous amounts of information to get to the truth, which is what was needed.

[AirRabbit]

Originally Posted by Mad (Flt) Scientist

My apologies to Mr Sumwalt; I admit I scanned the bios quite quickly.

In addition, I think it's quite important that the board be seen not to be a creature of the industry; it is, after all, one of the mechanisms by which the general public assures themselves of the safety of the transportation system, and if it were staffed entirely by industry insiders, there'd be even more people willing to cry "conspiracy" at every turn. It does, after all, fulfill a quasi-judicial/legal role, and we don't apply any expertise qualification to a jury on a capital trial....

First - sir, you are, indeed, a gentleman with character. Thanks.

Second - I think your statement about the NTSB is on target. The role they fill in accident investigation is not only valuable to the aviation industry, i.e., by pointing out the "probable cause" of accidents and incidents, but they also serve as a quasi-watch dog for the FAA. The Board isn't always right ... and they are often victims of political issues (as most persons are) ... but they are right a lot more than not and generally do a monumental service to the aviation idustry.

[ICT_SLB]

Probably should point out that the NTSB like CASB also deals with rail, maritime, road & pipeline accidents so Board members may have, & been appointed for, expertise unrelated to aviation. This is of course unlike BEA or the AAB.

[PJ2]

Reply to post #39;

krujje;

Thank you for the important correction.

to get to the truth, which is what was needed.

...and is still needed, apropos his comments early last-year regarding SMS and what I would term, "the de-regulation of safety oversight in Canada": Safety cuts risk air disaster: Judge - Toronto Star, March, 2007

SMS in and of itself is a good safety system because, in part, it supports monitoring and change through the intense collection and use of data.

SMS however, does not provide a robust external audit process and instead puts airlines in charge of both the collection and use of safety data AND the audit of such safety processes to determine "effectiveness". The process can work successfully so long as data is first, taken seriously and second, acted upon. That is not always the case when short-term commercial goals are the usual priorities. Regulatory oversight is often required to counter such commercial pressures and it is unclear how SMS will address the issue.