Coming Up Short

The subsequent crashed killed three and injured 180 of the 307 passengers and crew on board.

Soon after a chorus of experts and commentators immediately and emphatically dismissed cultural factors as a likely cause.

But the subsequent investigations have proven otherwise, raising challenging questions of the aviation industry. Testimony by trainee captain Lee Kang-Kuk, who was pilotflying OZ214 from the left hand seat (under the supervision of training pilot, and pilot-incommand, Captain Lee Jeong-min in the right seat), to a US National Transportation Safety Board hearing in December raises serious concerns, despite his 9,684 hours in a variety of aircraft types.

Captain Lee told the NTSB that he was “very concerned” about landing the 777 visually at San Francisco International Airport without the aid of a glide-slope indicator (ILS).

At the time of the accident the airport’s ILS was not operative but its PAPIs (precision approach path indicators) were working. He added that because other pilots were doing visual approaches, “he could not say he could not do” one.

But it appears that Lee is not alone in his fear of visual approaches.

In an interview with Bloomberg, former Delta Airlines and Asiana Airlines (2006-11) Captain Victor Hooper said that on one approach to LAX his copilot “froze” when asked to perform a visual approach.

According to the Bloomberg report Captain Hooper had to take the controls to get the aircraft back on track, and he recalled the copilot saying “I don’t need to know this.

We just don’t do this. ” Captain Hooper was also interviewed by the NTSB and said that “while all Asiana pilots he flew with were extremely competent at executing the training they were provided, there was minimal training in how to do a visual approach. ” He added they had limited opportunities for stick and rudder skills training but he “was sure that, given the opportunity, they would be excellent pilots. ” He added, “they really needed to add in the syllabus more visual approaches where they just go down and fly around a local area and fly without auto throttles, because they count on the auto-throttles and assume they’re going to work right. ” According to the Bloomberg story, Ross Aimer, a retired United Airlines captain, who trained crews at Korean Air Lines for Boeing subsidiary Alteon Training in 2008 and 2009, and Kenneth Musser, a former Delta pilot who flew 777s for Asiana for almost four years until 2009, both said they also noticed that many Korean pilots struggled with visual approaches.

Bloomberg also spoke with David Greenberg, the retired Delta Training Captain who was hired by Korean Air in 2000 to turn around the airline’s then disastrous safety record that had made it a pariah in the industry. On the deficiencies in hand-flying he said “I observed it,” but added in a portent of a much wider problem that it was not worse than with pilots elsewhere in the world.

On hand-flying, Boeing’s 777 Flight Level Speed Change (FLSC)mode has come in for scrutiny.

According to one 777 captain the issues revolve around the interaction between the 777’s auto-flight and auto-thrust. The FLSC is not recommended for the final phases of an ILS approach but there is no guidance on a visual approach.

“The FLSC allows the auto-thrust system to enter a “sleep” state if the throttles are not moved for more than 1. 2 seconds,” said the Captain.

“The problem is that in all other auto-flight modes the auto-throttle will re-engage. ” The pilots of OZ214 did retard the throttles after FLSC was engaged.

The FAA became aware of this issue during the 787 certification but after consideration of all factors deemed it not a safety issue, while Boeing has noted it in its training manuals for 15 years. And in the NTSB’s investigation an Asiana Airlines ground school instructor who trained Captain Lee told investigators he emphasised this issue because he had “personally experienced it. ”

Candid admissions

As disturbing as the fear of flying a visual approach admission was, there was more to come in the testimony from Captain Lee, who said that he was blinded while making the approach by something shiny. But when asked why he did not wear sunglasses, he said it would have been impolite for the training captain not to see his eyes. He added it was very important in their culture.

Asked why he did not break off the approach as he was pilot in command he told investigators: “That’s very hard because normally only in our Korean culture the one step higher level the final decision people he did he decide the going around thing.

It’s very important thing. As a first officer or the low level people they dare to think about the go-around thing. It’s very hard. ” Responding to this answer an NTSB investigator asked: “In your mind, then, and I don’t want to put words in your mouth, you tell me, did you feel that as the pilot in the left seat flying the airplane that youhad the authority to do, commence a go-around yourself?” Captain Lee responded: “Goaround thing. That is very important thing. But the instructor pilot got the authority. Even I am on the left seat, that is very hard to explain, that is our culture. ” He added that it was company policy after a FO-initiated go-around – overriding the flying captain – in an A321 which resulted in a tail strike. As it happened, Captain Lee, at the time an A320 captain, was part of the investigating team that compiled the report. For him not overriding the captain or pilot monitoring was indeed – overriding.

Asked whether he had ever had to take the controls away from a captain because they were not seeing something or were confused or disoriented, Captain Lee said “no, no way. ” Asked if he could imagine a situation where he would ever do that, he said “no way. ” However, he said that Asiana encouraged junior pilots to “speak up if they felt uncomfortable about something”.

Watershed

For Najmedin Meshkati, Professor of Civil/Environmental Engineering and Industrial and Systems Engineering, who also conducts and teaches human factors in aviation safety at the University of Southern California, OZ214 is a watershed in aviation.

“We cannot change national cultures and their emitting behaviors. We have to better respect them, better understand them and try to delineate their implications on human-machine systems integration,” Professor Mashkati told Australian Aviation.

He added: “In the whole context of human-systems integration paradigm, we need to try to understand and cater to cultural differences and use design for adjustability principles which we use for anthropometric (human body size) differences and use the same principles for accommodating cultural differences, if we want to be successful and get out of our denial mode. ” Professor Meshkati warns that the “denial of existence and impact of cultural factors and differences in aviation systems design has been a routine practice in the past. Its time is over, thanks to the rude awakening of the Asiana 214 [accident], and has to be changed. ” Professor Meshkati is one of the aviation industry’s foremost experts on human factors machine interface.

He has taught at the renowned 62-year old USC Aviation Safety Program since 1989 and was its director for seven years from 1992 to 1997. Meshkati is also a Jefferson Science Fellow and was a Senior Science and Engineering Advisor to the US Secretary of State. He is the author of one of the original and pioneering articles on the role of cultural factors in aviation safety, which was published in the October 1996 special issue of the International Civil Aviation Organization (ICAO) journal.

On national culture, Professor Meshkati says that “national culture, according to anthropologists, is the way of life of a people – the sum of their learned behavior patterns, attitudes, customs, and material goods. And, according to famous Dutch social psychologist, Professor Geert Hofstede, culture

Cultural factors have a significant effect on the realities of operating a complex technology.

in this context can operationally be characterised as ‘collective mental programming of peoples’ minds’. ” He adds: “According to (late) Professor Azimi, the culture of a society consists of a set of ideas and beliefs. These ideas and beliefs should have two principal characteristics or conditions: First, they should be accepted and admitted by the majority of the population; and second, the acceptance of these beliefs and ideas should not necessarily depend upon a scientific analysis, discussion, or convincing argument. ” Professor Meshkati adds that national culture affects not only the safety, but also the success and survival of any technology.

“National cultures, according to Hofstede’s monumental research, differ on at least four primary dimensions: Power Distance, Uncertainty Avoidance, Individualism-Collectivism, and Masculinity-Femininity. All of these play a role in safety in the cockpit. ” For Professor Meshkati there are two key issues that cultural factors impact – the interaction with others in the cockpit and the human/ machine interface.

“The first was very clear and played out very strongly [in relation to the Asiana 214 incident] and is a text book example of cultural factors. The latter of automation interaction is less well understood or discussed. ” In 1996 the US Federal Aviation Agency (FAA) conducted a comprehensive review of human factors in aviation and identified issues with automation and cultural factors. One of its recommendations was: “The FAA should ensure that research is conducted to characterise cultural effects and provide better methods to adapt design, training, publications, and operational procedures to different cultures.

The results of the research should also be used to identify significant vulnerabilities, if any, in existing flightdeck designs, training, or operations, and how those vulnerabilities should be addressed. ” Professor Meshkati questions: “What has been the response of the FAA and industry to this recommendation?” He knows of none. The FAA report found issues across all designs from all manufacturers but did not identify specifics.

Professor Meshkati highlights a 1998 study by Helmreich and Merritt of flightdeck automation, which surveyed 5,705 pilots across 11 nations and reported that “the lack of consensus in automation attitudes, both within and between nations, is disturbing. ” In the book Macroergonomics: Theory, Methods, and Applications (2005), Meshkati authored a seminal chapter “Macroergonomics and Aviation Safety: Why your flight safety is at the mercy of cultural factors”. He starts the chapter with a strong and relevant epigraph, a quote from (the late) Professor Stephen Jay Gould, the renowned Harvard University professor of geology, biology and the history of science, who contended that even hard sciences theories are strongly culturally-based: “Facts are not pure and unsullied bits of information; culture also influences what we see and how we see it. Theories, moreover, are not inexorable inductions from facts.

The most creative theories are often imaginative visions imposed upon facts; the source of imagination is also strongly cultural. ” This was written in Gould’s book, The Mismeasure of Man.

Professor Meshkati wrote in 1996: “Research has demonstrated that technology utilisation, without the incorporation of the necessary human factors and cultural considerations, is doomedto failure. It is incumbent upon the world’s airlines and aviation industry (aircraft and equipment manufacturers, air traffic controllers, and civil aviation authorities) to systematically take into account the physical and psychological factors, as well as the cultural attributes of their user populations, in the design and operation of passenger aircraft and aviation systems.

Cultural factors have a significant effect on the realities of operating a complex technology such as modern aviation, and the nature of these effects must be understood and accommodated if aviation systems are to operate safely. ” In a paper entitled Design Philosophies in Flight Deck Automation, Stephanie Chow and Meshkati quoted a study by Hutchins, Holder and Perez about culture and flightdeck operations at the University of California- San Diego, which found that there is no link between the aircraft manufacturer’s national culture and safety of the flight.

However, the paper by Chow and Meshkati points out that the same report elaborates on how culture plays a role in the organisation of behaviour in the flightdeck. French pilots are described as ‘simple executants’, whereas the American pilots are portrayed as pilots who like to overcorrect and take control of the aircraft.

The paper states that: “Airbus and Boeing will naturally manufacture and design flightdecks illustrating their respective culture. Even though both designs are the most advanced and safe systems in aviation history, pilots of various cultures will operate these jetliners differently.

Aircraft manufacturers and airline companies need to consider national cultural influences and human factors in designing the flightdeck and in how they train pilots. ”

Back to the past

The admissions of Captain Lee Kang-Kuk are a disturbing flashback to a special edition of the ICAO journal of October 1996 headlined ‘Human Factors in Aviation’. In that edition a survey of 13,000 pilots from 25 airlines from 16 countries found agreement ranged from 15 to 93 per cent to the statement: “Crew members should not question the decisions or actions of the captain except when they threaten safety of the flight. ” The same survey found a range in agreement of between 36 to 84 per cent to the statement: “If I perceive a problem with the flight I will speak up regardless of who might be affected. ” The study found that pilots from Australia, New Zealand and Ireland were the most assertive and likely to speak up, while the pilots from Japan and South Korea were the least likely.

It was noted in the ICAO journal that many nations had developed CRM training programs congruent with their own culture and these had been well accepted.

But the situation Captain Lee found himself in – of having to land for the first time at San Francisco, while under training with only 33 hours in the 777 and with the ILS inoperative – may have prompted him, when under stress to fall back on traditional values and beliefs of the Korean culture.

This is precisely what happened on July 1 2002 when a DHL 757-23APF and Bashkirian Airlines Tu-154M collided taking the lives of all 71 aboard both aircraft. After an ATC human failure due to high workload and lack of staff, the two aircraft were on a collision course and their respective TCAS systems gave opposite instructions to avoid impact. The DHL 757 was told to descend and the crew of the Tu- 154 were instructed to climb. Both also received further advisories to “increase descent” and “increase climb” respectively just 8 seconds before the crash.

But in the critical last seconds the Tu-154M captain, Alexander Gross, also heard Zurich ATC, which had realised the error instructing a descent not once but twice. Captain Gross ignored TCAS in favour of the instruction given by ATC at the critical moment and descended into disaster.

His country’s cultural tendency, some speculated, owed to a heritage where the command of the voice must be obeyed and in which the reliability of technology lagged well behind the West. Captain Gross at 57 would have lived through some of the worst of the Communist years where not to obey the voice of authority meant imprisonment. In an ironic tragic cultural twist, what well may have influenced the captain in the final seconds was that the human instruction was coming from Swiss ATC – known for its precision.

And cultural differences between societies are many and varied, points out Professor Meshkati.

It is not just a “Korean problem”. Some are major challenges for aviation.

Another cultural twist comes from a 1986 study of colour perception between populations of the US and China. Of 784 volunteers across both sexes and varied occupation groups in Kunming, China less than 50 per cent associated red with stop and green with go. The ratio for the US participants was almost 100 per cent. Red is a fundamental warning colour in aviation. The pilots’ saying of PAPI lights sums it up: “Red on white you’re alright, Red on Red and you’re dead”.

Even something as fundamental as training manuals require special attention. One study found that a poor quality translation – not conceptually adjusted – resulted in performance that was 14 per cent below those using the original manuals.

Asiana Airlines says that its “training previously exceeded all Korean and international standards and we have [now] instituted even more demanding requirements across the board for basic, initial and transitional training for flightcrew and strengthened flight instructor training. ” It has hired Akiyoshi Yamamura, formerly of All Nippon Airlines and the International Air Transport Association, to a new senior executive level position to oversee safety operations and it has enhanced training on automationIt is not just a “Korean problem”.

Some are major challenges for aviation. logic and the capabilities and limitations of the auto-throttle stall protection system.

In a statement for Australian Aviation the airline says: “Asiana has a state-of-the art training program that consists of ground school, simulator training provided by Boeing Korea LLC, and operator experience training (OE). All of our training exceeds International and Korean standards.

“We also provide special training for all airports that may provide special challenges for our pilots.

SFO often involves high-energy approaches, and while it is not the most difficult airport in our system, we do provide specialised training on the SFO approach.

“As the statements of Captains Jung Tai Soo and Kim Je Yeol indicate, the pilot flying was instructed on the specific effect of using FLCH (flight level change) mode on the auto throttle system in his transition training class. ” On its CRM program the airline states: “Asiana Airlines has spent years studying other airline crashes, and has created training programs designed to make sure its crews work together as a team in the interests of safety. That includes making sure junior officers understand that not only are they allowed to speak up when they notice a problem in the cockpit, they are required to – even if that means challenging a more senior officer.

“Our crew resource management (CRM) program specifically takes into account the role that cultural issues have played in other aviation accidents. We train our pilots that any member of the flight crew can – and must – speak up when needed.

All of our pilots understand this principle, which has long been a major focus of our training. And, importantly, no crew member is ever punished for speaking up, since we train that it’s a critical safety factor.

Indeed, on the accident flight, the most junior officer of the crew spoke up when he saw something wrong, according to documents released at the hearing, calling out ‘sink rate’ – indicating that the airplane was descending too quickly. ”

OZ214 raises many questions but no answers – as yet!

Just as automation has given less experienced pilots a safety cocoon and the industry its best ever crash record, has it also lulled many into thinking nothing more needs to be done in its pursuit of the holy grail of zero accidents? Professor Meshkati warns much more needs to be done. He cites the late Nobel physicist Richard Feynman (1986) in reference to the Space Shuttle Challenger explosion: “For a successful technology, reality must take precedence over public relations, for nature cannot be fooled. ” He stresses “when it comes to vexing and serious issues of cockpit automation, cultural factors, and their complex and mostly unknown interactions, we need a paradigm shift in our thinking and should diligently be addressing their paramount influence on aviation safety. ” Perhaps Albert Einstein sums up the problems best in this quote: “We cannot solve our problems with the same thinking we used when we created them. ”