| | Last Update: Monday, Jul 29th 2024 08:17Z 30666 Articles available Events from Mar 23rd 1994 to Jul 28th 2024 | |
| | |
|
|
The human factor named "Simon Hradecky" and the team of man and machineBy Simon Hradecky, created Sunday, Aug 19th 2012 14:36Z, last updated Sunday, Aug 19th 2012 14:45ZIn July 2012 I gave an e-mail interview in German to Austrian Aviation Net, an independent Internet Portal with the focus on aviation in Austria (Europe), my home country. A number of readers of both Austrian Aviation Net and The Aviation Herald challenged me to provide an English version to the AVH readers as well.
When I received the questions by Roman Payer, editor in chief of Austrian Aviation Net, I found them highly interesting and challenging and thus immediately engaged in answering the questions. After I had completed the script, questions arose however whether it would be wise to release the interview given that The Aviation Herald's editorial line, without exception, is to provide facts while my own personal opinion has nothing to do in the stories. After consultation with friends and lengthy considerations I decided to release the interview, and in the aftermath also decided to grant the request to have the interview published on AVH as well.
Interview conducted by Roman Payer (RP), answered by Simon Hradecky (SH)
The Aviation Herald daily reports about incidents and critical situations in the aviation industry. The service is recognized as accepted industry service for safety relevant occurrences. Behind the English language service stands Simon Hradecky of Salzburg, Austria - an expert in Aviation Safety.
On the occasion of this year's summer talks Austrian Aviation Net talked to Simon Hradecky about aviation safety and fatigue in the cockpit, but also queried whether the recent conflict within Austrian Airlines would affect safety of the airline. SH also explained why he considers the final report into the crash of AF447 a pioneering milestone, how dangerous (oil-) fumes are and in which country flying is safest. And why one can expect with certainty in some countries, that it won't last too long until the next serious accident.
RP: Why do you deal only with incidents? And why in English?
SH: I am technical software developer. Following my beginnings with networking, robotics, tooling machines and flexible manufacturing I began, not in the least because of personal interest as result of my military service, to extensively study aerodynamics and thermodynamics and later on develop a mathematical model to compute aircraft and engine responses in real time. When it became clear that the mathematical models would result in plausible algorithms and software, the question arose whether such computations would match real aircraft handling especially in abnormal flight scenarios.
Unusual flight scenarios do not occur in normal operations according to experience, even full flight simulators simulate aircraft behaviour realistically only within the boundaries of flight data packages that have been generated by test pilots, hence my interest in abnormal situations of flight, incidents and accidents, led me to compare the data of such occurrences with the results of my own mathematical computations. My international friends showed considerable interest also in this part of my work, hence I began routinely report about such occurrences in a small circle. English as the only world spanning language of aviation was the logical choice. RP: What made you think to found The Aviation Herald? How long ago?
SH: Thirteen years later the decision was born to professionalize the so far private reports and make them available to a larger circle of readers. The idea of The Aviation Herald was born in April 2008 and following the technical implementation made live on May 12th 2008. At that time I had hoped to be able to perhaps attract about 1000 regular readers per month and was completely amazed when The Aviation Herald exceeded that number a few days later already, in January 2009 jumped above 100,000 visitors and in the meantime regularly attracts about 1.3 million readers per month.
RP: Why Aviation?
SH: During my military service I first got in contact with aviation when I was used to monitor air traffic control radio communication and to plot the actual situation within air space surveillance. I began to develop an interest in how every cog intermeshes with the next - so that despite lively traffic even at that time there were no occurrences worth mentioning -, what sense was behind the various definitions and rules and from there, finally developed an interest in the physics of aircraft. During university and the first years of professional life I lost sight of aviation again, however, one of my clients, an enthusiastic glider pilot, managed to transfer his enthusiasm again and taught me the basics of flying. Flying became a hobby afterwards that led me to many full flight simulators of MD-83s - to date my favourite aircraft - Boeing 747-400, DC-10s, A300s, Concorde and others around the globe, subsequently aviation became a real task of my company.
RP: What makes the difference between The Aviation Herald and classic media?
SH: In the common press the impression often arises that even the most tiny technical disturbance is capable of creating a catastrophe. Lack of knowledge of how an aircraft and its systems function and lack of opportunity to take active action but be helplessly exposed to the situation as "pillion rider" reinforce such fears and understandably lead to panic too. Press such creates fear of flying with many human beings, who then report their fears in such minor incidents, often enough without background knowledge and without knowing if there were any dangers, and so reinforce this type of reporting. Thus a vicious circle has been created. Logically, a story "Escaped death by a hair's breadth" sells better than "Precautionary Landing", so that the allegation that media are more interested in their own income and the own agenda than truth and providing facts cannot be brushed aside.
It was also my intention to help our readers appreciate the background of such occurrences and to demonstrate that in the vast majority of occurrences no risk of a crash exists as enormous safety margins have been built into every aircraft and there are reserves like engine power, reserve systems, procedures and other performance reserves useable in nearly all issues. We regularly receive feedback from passengers who thank us for our reports as these enabled them to better assess a disturbance of their flight, and we even have feedback stating that passengers were able to overcome their fear of flying through the knowledge carried by our reports.
I am pleased that several media react to our reporting and improve their own coverage - even though the current horror stories about the United Boeing 757-200 over the Atlantic "fall 20,000 feet", which is completely and utterly contradicted by facts, seem to prove the opposite and seem to be evidence of the vicious circle described above, especially with the background that a single English-language tabloid press story was taken over by all other media obviously without cross checks.
RP: Where do you get all these information from?
SH: As result of the consequent editorial line The Aviation Herald has gained great trust by the industry, except the press offices of a number of airlines. We therefore get many hints directly out of the industry, but also from passengers, observers in the air and on the ground and from our readers. Of course we also utilize official sources and our extensive radar and flight plan databases.
RP: What was the most read story and why?
SH: Logically the loss of Air France 447 has moved our readers the most and resulted in more than a million accesses of our coverage of the AF-447. Next is the crash of the Sukhoi 100 in Indonesia this year followed by the emergency landing of the Qantas A380-800 in Singapore following the uncontained engine failure, the crash of the aircraft of Poland's President in Smolensk and the water landing of the US Airways A320 in the Hudson River.
RP: You are surely in contact with many pilots. How do you feel about the current conflict at Austrian Airlines and Tyrolean Airways?
SH: Operational decisions are not being commented by The Aviation Herald and myself.
As an Austrian I am worried however whether I shall be able to experience the incredible feeling upon boarding an Austrian Airlines Aircraft after a long trip abroad, when upon climbing the stairs I am received by the smell of delicious coffee and the sounds of the Blue Danube.
RP: Many say that such a merger poses a security risk. What is your opinion?
SH: These are operational decisions, I don't want to comment those.
RP: Incidents are often quite delicate. Airlines fear a negative image and false reports can quickly result in protests. How do you decide what goes online and what not? Does the occurrence need to be officially confirmed? How do press offices react to the reports?
SH: I am more than well aware of the risks of false reports and their consequences for the affected, but also for The Aviation Herald.
Therefore I never rely on a single source, except for official sources although even those are being cross checked wherever possible. In many stories published by us there is no official confirmation of the occurrence. But we always have sufficient evidence in our hand to prove our story in case a suit is filed. The perpetuation of evidence which is integral part of our researches is by far the most extensive part of our work.
This effort pays off, although despite all these measures I am not immune to errors. As in aviation itself the human factor named "Simon Hradecky" plays a decisive role in my work. In addition, The Aviation Herald can only be as good as the sources I trust in and only be as good as the information that become available to us, my sources and myself. I can say with pride, that The Aviation Herald has so far not been confronted with threats of legal action or lawsuits.
We frequently receive pointers to quite credible stories, which cannot be verified. A few completely incredible cock-and-bull stories however were proven correct in research and have resulted in reports. Some airlines try to influence our coverage or even attempt to suppress a story, sometimes in a stridency that comes close to a threat of legal action. If I ever needed to give in to such cases of taking influence, for example due to doubts on quality of research or insufficient evidence to prove our case, my credibility would be lost in my own eyes just as well as I'd rate my own credibility lost if legitimate evidence would not be given proper coverage.
RP: How many percent of all daily occurrences are covered by The Aviation Herald? In other words, how much else happens in the air without coverage?
SH: The exact number of global occurrences is not known, therefore is it is difficult to estimate. We could take Canada for comparison, which has an outstanding reporting system: Each day there are about 50 occurrences reported for Canada alone. In order to not depict Canada as the most unsafe country of the world in the view of readers, who do not (yet) know that background, I can report only the most serious incident per day in average - a higher frequency of reports by experience immediately results in the question, why Canada would be that unsafe. Hence The Aviation Herald only reports the "peak of the iceberg" in maximum.
RP: Is fatigue in the cockpit a real security problem in your neutral view?
SH: A tired human being reacts slower and tends to make more errors than if well rested. Therefore fatigue is a problem. Fatigue however is only part of the human factors.
It is my view that the "human factor", even though every accident report pays great attention to the human factors, does not find proper recognition in the technical, operational and legislative rules.
The human being has been provided with enormous intelligence; by definition and in practice this means that the human often finds solutions, now and then seemingly paradoxical solutions, in unexpected situations and is capable of putting those solutions into practice. The human being however also has limitations, set by nature in millions of years of evolution and not removable despite the most beautiful theories: for example errors in routine tasks, that even checklists can not entirely remove. The machine however copes extremely well with routine tasks but cannot exceed, what an engineer designed into the machine, and thus fails in unexpected situations. Man and Machine do thus complement ideally, not just in aviation, if both are used properly and the flow of information between man and machine is optimally ensured.
This however requires interaction between man and machine, in which the machine must adjust to the human and human possibilities and limitations and not vice versa as is seen too often in practice, unfortunately. This means that precise and complete status messages must be brought to the attention of the human in such a way, that even under stress and time pressure the human can recognize, understand, verify, accept and act upon such information. Questions like "What is she doing now?", quite a frequent question heard in the cockpit, are high warning signs that the interaction of man and machine does not work. If such questions already arise in the cockpit during normal operation and the machine does not sufficiently explain to the human what she is currently doing, how should this work in a stress scenario within a technical fault? Especially in this area there is little research or legislative approach.
Ergonomy, the research of interaction between man and machine, is a huge part of the software industry. Software Ergonomy even became its own science. In the context of this discipline it was proven, that bad presentation of information to the human can substantially increase the burden and thus the fatigue of a human being, but a good presentation can substantially reduce that burden. Equally it was proven, that the way of how information is transferred to the human, can substantially increase or decrease the rate of errors.
I am missing a similar importance of ergonomy in aviation. In this sense the final report by the BEA into the crash of AF-447 has very positively surprised me, I did not expect such a detailed analysis of the events in the cockpit with the focus on interaction of man and machine, the essentiality of information, that was displayed or not, as well as the credibility of displays and alarms from the human point of view, in short I did not expect such a detailed analysis of cockpit ergonomy. In my view therefore this report is a pioneering milestone which opens hope in improvement of ergonomy in aviation.
From this point of view I am appalled by every report citing pilot error as cause of an accident. This falls short of the goal of a report, cannot eliminate the causes and cannot prevent a repetition of such accidents as is being impressively demonstrated by the very first ever accident report released by a UK aviation accident commission 100 years ago and current final reports. The decisive question reads: "Why has the human acted in a way that we assess as an error?" We'll find the true cause in the answer to this question.
Only when we cease to idealize human beings and to make demands, that a human being can not comply with long term and especially under pressure, only when we cease to make the human in the cockpit a scapegoat, who often cannot defend against such accusations anymore and thus is the easiest target, only when we consequently research, depict and correct the causes of their mistakes, we'll be able to reduce the number of accidents because of "human factors".
RP: Out of your many years of experience: what are the major risk factors?
SH: Professional blinkers, routine, reliance on technology and disregard of human factors in the design of aircraft, cockpits, training, maintenance, procedures and laws as well as disregard of human factors in everyday life in cockpit.
The idea, that a human in the cockpit can only make mistakes and thus needs to be monitored by technology and needs to be overruled where necessary to prevent human mistakes, is flawed in itself three times:
First, technology is designed and produced by humans, the human factor therefore not only exists with the operator of the technology, but in the technology itself. With this approach a human being, commonly an engineer who never flew himself and has no knowledge of cockpit environment or aerodynamics, intercepts control inputs of a trained specialist in the cockpit, even though the current situation of flight is not known to that engineer and quite possibly never occurred in any analysis before.
Second, this approach robs the human in the cockpit of any usefulness, if unexpected solutions which would permit to maintain or regain control of the aircraft are being prevented. This approach thus downright avoids cockpit resource management (CRM), which is taking advantage of all resources available in the cockpit. In my view CRM not only concerns the humans on board, but has to include the aircraft systems as well. It's of no use if the machine takes control according to the motto "I am more clever than the captain" and thus performs like the proverbial patriarch who ignores the co-pilot - which was one of most common causes of accidents in the past and led to the introduction of CRM.
Third, this monitoring and intrusion into the authority of the pilot leads to faith into technology and reliance on technology, however does not cater for the possibility that exactly this technology performs erratically or fails and thus plunges the human into a scenario, that the human has no experience with and possibly has no time to adjust to.
I do not at all rule out however that such intrusions into the control inputs of the pilot can prevent or at least reduce consequences of accidents as long as the technology works and the aircraft remains in a condition that was covered by developer analysis.
I believe the approach is much better to have the machine support the human to the best knowledge and technical possibilities but grant the human absolute authority and final decision on board. This approach ensures that the intelligence of the human being is utilized optimally while at the same time the machine by continuing routine tasks reduces workload of the human, so that the human has time to think and doesn't get into extreme stress, in short, this approach ensures the ideal cooperation of the team consisting of man and machine, even in crisis.
I do not at all rule out that within the context of this approach the machine could intercept control inputs of the human with appropriate information about the fact of that intercept, the reason for the intercept and the possibility to reject the intercept. Such an intercept can offer assistance as well.
Both approaches can only work if initial and recurrent training as well as daily work ensure, that pilots know their aircraft and her behaviour in manual control within normal flight envelope and are able to recognize and correct aircraft responses near the boundaries of the flight envelope.
I know source codes of software in use on real aircraft. I was shocked by these source codes which essentially consist of nothing more than huge, deeply nested decision trees "if that input is that, and this input that, and this input ..., then set output to that", thousands and thousands of lines of nested if statements. It is easy to predict that in such huge decision trees exactly that one condition is overlooked and not checked that later becomes the fate of a crew. In my professional experience as software developer such decision trees represent the admission, that the developers (analysts, mathematicians, programmers, testers) have no knowledge of the subject and just program according to the symptoms described by the customer.
My professional experience also tells me that complex terms may require a year or more to work into the matter and transfer the customer requirement specification into more universal mathematics and algorithms without such decision trees, that complete the assigned tasks safely and completely in a few, manageable branches. My 35 years of professional experience tell me as well, that despite this extreme learning and analyzing effort the development will be completed more quickly because the test effort (usually about 80% of the total design time) substantially reduces. Consequently, later maintenance and enhancement of the software will be easier and quicker.
I'd like to illustrate these rather dry theoretical remarks on the example of the laws of falling developed by the philosopher, mathematician, physicist and astronomer Galileo Galilei at the leaning tower of Pisa in the years of 1589-1592. Before Galilei conducted his experiments and his analysis, the question "how long does the fall of a ball take from 5, 10 or 15 meters" could only be answered by empirical observation as follows:
If height of drop equals 5 meters, then the fall takes 1.0 seconds if height of drop equals 10 meters, then the fall takes 1.4 seconds if height of drop equals 15 meters, then the fall takes 1.7 seconds
What however, if we now need the time of fall for 20 meters? Our program above is entirely useless, we could as well ask "does the ball now arrive as Sushi" - but at least we had a laugh on our side. Galilei was not satisfied with these empirical results, analysed the results also with the help of an inclined plane and developed a mathematical model, which was able to reproduce the empirical results by computation but also predict falls from other heights. This mathematical model was subsequently scientifically verified, recognized as correct and taken into knowledge of mankind as the "laws of falling". Our "program", without taking drag into account, now becomes a one liner instead of a three liner and reads:
the fall takes square root of (height * 2 / 9.81) seconds
Short summary from my point of view: ergonomy is not just limited to interaction of man and machine in operating the machine, but is also indispensable in the development of technology because in this case, too, interaction between man and machine takes place. The lack of ergonomy not only in the operation of the device, but also in its development, constitutes a significant risk for the safe operation of an aircraft.
RP: Oil fumes on board are considered as big, invisible danger within the industry. Is this correct? Do you see many incidents?
SH: TCP is well known as neurotoxin which can cause serious and permanent damage to the health. In the meantime, following yearlong battles in front of courts, a causal link between oil fumes, TCP and illnesses of flight crew, cabin crew and passengers is more and more often recognized. Oil fumes containing TCP thus pose a significant risk.
The Boeing 787 is therefore a first step into the right direction, cabin air is no longer taken from the bleed air off the engines, which necessarily need oil to operate and cannot be made 100% tight even with best technology, but is taken from independent supply, the risk of contamination with oil fumes therefore appears largely reduced. At the same time, and this is the true irony of it, this separation enhances the efficiency of the engine, reduces the fuel consumption, enhances cabin safety and thus renders the original approach at the time of development of jet engines absurd, that bleed air would be an easy and efficient way of cabin air conditioning by removing the necessity of having a separate compressor.
RP: Do you have fear of flying because of your daily writing about accidents?
SH: No. I enjoy flying without any weak feelings, especially because I know and can assess as result of the many reports, that in the vast majority of problems the aircraft continues to fly safely. In the common press however the false impression often arises that a little problem is enough to force an aircraft into catastrophe.
RP: What are the most common occurrences?
SH: Any statistical evaluation of our coverage would be grossly misleading because we only see the peak of the iceberg. Therefore we have consequently refrained from any statistical evaluation of our data from the very beginning, and we have declined many such requests.
RP: Last question: which airline did you never need to write about? In other words, which airline is the safest?
SH: This question does not arise, at least not in this form. That airline is safest, who put safety and economy, both are necessary and relate to each other, into the centre of their activity and deals with safety aspects openly and honestly.
For me the question actually arises in the opposite direction: those airlines, I write about most often, probably are the safest. It is an open secret that Canada has the most extensive and best reporting system on the planet and its government takes safety matters very seriously. Despite the enormously difficult geographical and meteorological conditions Canada thus belongs to the safest countries on the globe.
Where I have never or almost never reported occurrences, one can expect with certainty that it doesn't last all too long to the next serious accident.
|
By Flugcast on Friday, May 19th 2023 07:22Z
By Dennis B on Tuesday, Sep 14th 2021 20:49Z
By Enzo Panico on Thursday, Sep 5th 2019 14:27Z
By Mohsin Khan on Wednesday, May 22nd 2019 09:28Z
By Mike on Saturday, Jan 7th 2017 23:20Z
By Mike on Saturday, Jan 7th 2017 23:19Z
By John Wollak on Tuesday, Jan 3rd 2017 10:42Z
By D on Tuesday, Aug 16th 2016 12:58Z
By ikm on Thursday, Jul 28th 2016 21:17Z
By Jorg on Saturday, May 21st 2016 05:00Z
By balu on Wednesday, Sep 9th 2015 10:31Z
By Ray Dahl on Sunday, Jul 5th 2015 10:52Z
By Carlos Pera on Friday, Apr 3rd 2015 19:00Z
By Colin on Tuesday, Jan 27th 2015 11:20Z
By propdriver on Thursday, Mar 13th 2014 00:30Z
By Canadian Frequent Flyer on Wednesday, Mar 12th 2014 11:14Z
By New Reader on Tuesday, Mar 11th 2014 22:49Z
By enzo on Sunday, Feb 16th 2014 19:15Z
By cyyz on Thursday, Jan 16th 2014 20:58Z
By Rem�gio Chilaule on Thursday, Dec 5th 2013 23:53Z
By AleCapece on Friday, Nov 15th 2013 22:52Z
By (anonymous) on Tuesday, Nov 5th 2013 07:47Z
By messner on Saturday, Oct 19th 2013 19:44Z
By messner on Saturday, Oct 19th 2013 19:34Z
By No Fly Zone on Thursday, Sep 26th 2013 00:39Z
By mahdi on Saturday, Sep 14th 2013 20:34Z
By Tango Sierra on Monday, Jul 8th 2013 14:06Z
By Paolo on Thursday, Jul 4th 2013 06:13Z
By Jimon737 on Tuesday, Apr 23rd 2013 09:02Z
By Stuart on Thursday, Jan 10th 2013 04:45Z
|
The Aviation Herald Apps Android and iOS
Support The Aviation Herald
Euro
US$
Interview:
|
|