Human Factors still the current challenge of the aviation industry
No Limits for the Understanding of Human Factors
Ten Questions for the Author of “Ten Questions About Human Error: A New View of Human Factors and System Safety”
Everyday error cause most airplane crashes
Why are Routine Flight Operations Killing Pilots and their Passengers?

Following numerous unexplainable accidents during the 70s (examples: Miami in 1972 or Tenerife in 1977) the industry reacted with a new solution: the CRM training course.
The first definition of CRM means Cockpit Resources Management. A singular approach and vocabulary was born reagrding safety, but through the new light of human (operator) behavior.

A lot has changed: culture, spirit, behavior, management, training...

If CRM has changed a lot, more has still to be done, at least for two reasons:

1. We always have to learn. Nothing is definitely graved especially regarding safety and human behavior. There is one common point between the recent accidents and those during the 70s: all are human factor events.
2. The only reason today for aviation incidents or accidents is human factor, just because human being are at each level of the system, at every level of each process.
   

But, guess what did happen during the 80s... The unexplainable accidents did not cease. The CRM concept seemed to restrictive and did receive a new accronym: CRM for Crew Resources Management. Instead of the sole cockpit crew, CRM was including the cockpit and cabin crew.

During the 90s, the concept appeared still restrictive.
CRM evolved to Company Resources Management, including this time the operator departments.

Few years ago appeared a certain infatuation for human factors as the industry were discovering something new. In order to give more sense to this, two other important aviation organisations moved to develop their own CRM:

1. TRM for Traffic Resources Management, the Air Traffic Control CRM
2. MRM for Maintenance Resources Management, the CRM for the maintenance organisations.

Today these three trainings (CRM, TRM, MRM) are required by the regulatory authorities in most of the countries were aviation is an important transportation tool.

But, the last step for CRM is far to be reached... The current (aviation) industry appears to be a highly complex system, involving different organisations (manufacturer, state authorities, airline, maintenance, airport, ATC, numerous other service providers as cleaning, fueller, etc.). If it is essential every organisation develop his own CRM training, these organisation must also work together inside this highly compex system. And here, everything has to be done.

Now bear in mind that CRM is nothing more than the practical aspects of The Human Factors. Since the early CRM and its vocabulary was focused on behavior, there is no doubt Human factors era started 30 years ago.

We just have to expand the approach we have about this concept in order to obtain a deeper and closer understanding of the industry mishaps.

This new approach, must be offered to the protagonists, mainly through education and training courses.

1. What could North Americans learn from Europeans in managing safety?

The safety management system (SMS) is about a partnership between the industry and the regulator. Those partnerships, and the lack of an adversarial relationship, are something that by very nature have already existed in Europe. You’ve already taken the safety management idea from the natural European interaction. It seems to be more accepted in Europe that systemic doesn’t necessarily just refer to a static conglomerate of stakeholders, but rather it refers to a completely new way of systems behaving. Systems behave in a certain way that requires a new set of models, a new set of ideas, a new set of indicators to monitor and manage. In Europe, systemic means “a new way in which a system behaves.”

2. Why do we have the same accidents over and over again?

Are we really having the same accidents? I would say yes, some. So yes, we’ve seen them before, but now they have been exported to other parts of the world, where regulation is not as strong. When you look at our part of the world, have we really seen these accidents before? When we look at our part of the world we are having accidents where failures, in really safe systems, are preceded not by component failures, but by normal work. Organizations are having accidents by drifting into failure (e.g. Alaska 261), when there are goal conflicts between production and safety resulting from resource scarcity, for example. We should not be surprised to see the leakage from these pressures. The goal is to figure out how to help organizations acknowledge, work on, and resist these pressures.

3. What is the “new view” of human error?

The new view of human error sees human error as a consequence, not a cause; it is a start, not a conclusion. The sources of mistakes are structural, not personal. The other part of the definition says that accidents are a structural by-product of people doing normal work; the systems are functioning normally.

4. The new view sounds fine, but we live in the real world, and when people make mistakes, there are consequences. In this new view, what happens to responsibility?

Responsibility is an important part of the new view. The new view says you cannot hold someone responsible if they do not have the requisite authority. As soon as you begin discussions about responsibility, you begin discussions about organizations. Responsibility cannot be spoken of in a vacuum.

5. How would you suggest commercial aviation move from the old view to the new view of human error, given the current safety programs we have [crew resource management (CRM), threat and error management (TEM), line operations and safety audit (LOSA), safety management systems (SMS), etc.]?

These initiatives, in principle, are not old view. They want to take people’s working conditions seriously; they want to take behaviour in context, which is, therefore, a new view. They are very much about understanding the conditions in which people work, and how they create safety. However, the risk in many of these programs is that they seem to see concepts, like error violation, as conceptually non-problematic. In these programs, we count errors and violations, and we use this information to determine how safe an operation is. The assumption that by counting errors and violations, you can measure safety, is problematic, as the real data lies much deeper. What do these errors and violations really mean?

6. The new view may be fine for big operators, but what could small operators do?

Small operators could learn to ask the right questions. When they see a human error problem, they could see it as an organizational problem. How does one go about asking a good question? Are you asking why from inside the tunnel (from the operator’s perspective, during the sequence of events before the negative outcome occurred)? Are you probing what the operator saw? What the operator heard? These questions work in reactive situations. What about in the proactive
sense? What are good questions to ask? It costs a lot of resources to ask good questions. Another thing small operators could do is freeze old view countermeasures—don’t knee jerk, pull licences, punish, write letters, etc. We need to step back and look forward.

7. What human factors training/education do inspectors/industry need to operate in an SMS world?

If you want to educate industry, and you want the regulator to collaborate in creating safety first, you need an organizational safety vocabulary so you can talk about the major risks. This may be very contextual. We need to turn people into system thinkers. Some of our models of accident causation are old. We need to shift our thinking and metaphors to understand that a system is something that lives, it can get sick from harmful pressures. We need to teach people how to look for other things—higher variables, such as: are they taking past experience as a measure of future success? You cannot see the universal but in the particular—but the particulars quickly stop making sense if you have no general concepts to relate them to. We need to invest in facilitating discussions between generalists and specialists. Technical people need education and updating, as do generalists. This will ensure they are capable of questioning their own assumptions. There should definitely be an opportunity for interactions where specialists and generalists learn from each other.

8. What qualities do aviation managers need to possess to be more proactive in managing safety?

Take domain expertise seriously. If you don’t, you do so at your own peril. Technical expertise alone does not qualify you to be a manager. You have to learn some skills that apply to running a group of people.

9. How do you detect and mitigate drift (the slow incremental departure from initial written guidance on how to operate a system)?

Get in fresh perspectives. Never stop asking questions. Ensure your people have a constant sense of unease. Recognize that that which is acceptable or normal, is not necessarily safe.

10. What’s next after SMS?

What you have to watch out for is that SMS does not become the nuts and bolts of the 21st century, where all we do is check whether documentation and processes meet specified criteria of quality because safety is an emergent property—it is more than the sum of quality parts. We have to go beyond SMS as a set of separate components, and learn more about how our people can get to see the big picture, because it is in the big picture that big accidents happen—not in the breakdown of any one component.

Printed in the Transport Canada Aviation Safety Letter, Issue 3/2006. This article was from the 2006 Canadian Aviation Safety Seminar (CASS 2006) which theme was “Human and Organizational Factors: Pushing the Boundaries.” To set the stage, Sidney W. A. Dekker, Professor of Human Factors at Lund University in Sweden, opened the plenary session with a provocative discussion on the new view of human factors and system safety.

Experts who study airplane accidents say the mistakes that lead to crashes are the same kinds of mistakes people make all the time, akin to locking keys in the car or forgetting an item on a shopping list. “In an airplane, it gets you in trouble,” said Scott Shappell, a professor at Clemson University who studies aviation accidents and the human errors that cause them.

Scholars say that as many as 80 percent of airplane accidents are caused by human error — it’s what experts in aviation call “human factors” and will probably be deemed one of the leading causes of the 27/08/2006 crash of CRJ100 Comair Flight 5191, which killed 49 people. The plane crashed after trying to take off from the wrong runway at Blue Grass Airport (pictures).

Human factors are an entire field of study in aviation. It embraces not only how a cockpit is run but also how mechanical repairs are made.

Specialists in the area work not only for the FAA but also as investigators for the National Transportation Safety Board and as consultants to airplane manufacturers to make sure planes are designed in a way that does not encourage human errors.

Its practitioners want to prevent mistakes. But they also recognize that, sometimes, mistakes get made.

Techniques and procedures should create a series of filters to catch mistakes before they cause a problem.

And they should have prevented the Aug. 27 crash, said Jim Hall, a former chairman of the NTSB. “Almost every safety net that was in place was blown through,” Hall said. Hall thinks the Comair crash will be studied by experts for years to come. “It is the most significant accident of this decade,” he said. “It has a wealth of safety dos and don’ts in it that need to be examined.”

On the morning of Aug. 27, the one air traffic controller on duty at Blue Grass Airport cleared Comair Flight 5191 onto Runway 22 and then turned his back to catch up on paperwork. According to FAA policy, two controllers should have been on duty .

The plane’s captain, Jeffrey Clay, who died in the accident, taxied the plane into position, but instead of turning onto Runway 22, which is used by commercial flights, the plane turned onto Runway 26. That runway is unlit and only 3,500 feet long. It’s meant for smaller panes.

Recent construction had changed the approach to Runway 22, and the construction was not reflected in the chart of the airport given to the pilots.

Because of those problems, James Clay, the brother of Jeffrey Clay, does not believe that blame for the crash should rest on his brother and First Officer James Polehinke, who was the only survivor of the crash.

To figure out what went wrong, human-factors investigators will want to know what the pilots were talking about in the cockpit, Shappell said. They will also want to know when they went to bed, how they slept, what they ate and drank and if they were having difficulties at home.

Investigators will also want to interview Polehinke, who remains in serious condition at a Lexington hospital. He remembers nothing about the crash.

Things like Comair’s bankruptcy proceedings and company staff morale may also be important.

On the day of the crash, the pilots initially boarded the wrong plane. That misstep could have put them behind schedule, said Paul Czysz, a retired aeronautics professor at St. Louis University. The air traffic controller could also have been in a hurry, which might explain why he had his back turned, Czysz said. “It was just a series of mistakes that when you add them all up, it’s a tragedy,” Czysz said. “At any one time, if somebody would have said, 'Where am I?’ it wouldn’t have happened.”

But once the pilots were on the wrong runway, another factor would have come into play, what experts refer to as “confirmation bias.”

“Once you make a decision, you seek out those things that confirm your original decision and ignore everything else,” Shappell said.

By SARAH VOS
McClatchy Newspapers
Published: September 26, 2006 in Anchorage Daily News
http://www.adn.com/news/alaska/aviation/story/8236439p-8133279c.html

Abstract
Routine flight operations present pilots with a myriad of latent threats. Three accident scenarios are presented that exemplify how a routine flight operation can end in disaster. The pilot’s complex and dynamic psycho-cognitive behaviors are analyzed and show that satisfactory technical training alone does not make a safe pilot.

More emphasis needs to be put on the "human system" the most likely system to fail in flight. Recommendations address the areas where intervention and education may mitigate some of these issues.

Scenario 1
The crew had just finished recurrent training. The instructor praised both pilots for exemplary performance in the simulator, and attested to that fact with positive comments on both pilot’s grade sheets. Both pilots had thousands of hours of flight experience and thousands of hours of combined time in the particular make and model they were flying. They were back on the line the following day.

Their first leg back on the line proved tragic, as both pilots, and 27 passengers were killed when the aircraft descended prematurely on a non-precision approach at night. As usual, the first question asked was "what happened?" How could such an experienced and well-trained crew commit this type of error, especially the day after they received recurrent training and were commended on their skills?

This is but one example of a routine flight operation gone terribly wrong. The pilots had flown into this airport on numerous occasions, albeit during daylight hours. The weather was reported to be good Visual Flight Rules (VFR), the wind was calm, and the runway was 10,000 feet long. Visual Approach Slope Indicators (VASI’s) were available to establish a proper glide angle to the runway threshold. But for some reason, the crew descended below the VASI’s prematurely, causing the aircraft to impact the ground a few miles from the end of the runway. Another classic Controlled Flight Into Terrain (CFIT) accident has occurred. A perfectly airworthy airplane, under complete control, was flown unintentionally into the ground without any prior awareness by the flightcrew.

This example shows us, in its purest form, where technical training ends and human factors begin. This type of accident occurs more frequently than one would be led to believe. The pilots assumed this was a routine flight. After all, the weather was good and there was nothing wrong with their aircraft just minutes before landing.

As it turns out, the captain, who was the pilot flying, was compelled to attempt a night visual approach to the runway, even though the VOR Runway 17 instrument approach was briefed and set up earlier. When the first officer queried the captain on this discrepancy, the captain replied that he "wanted to shoot the visual approach since the weather was good and it would save some time." That was the last discussion recorded on the Cockpit Voice Recorder (CVR) before the sound of impact, approximately two minutes later.

In a macro-analysis of this accident, it was concluded that the aircraft impacted rising terrain approximately 2.3 miles from the runway threshold. Additionally, the aircraft was 800 feet lower than it should have been at that point if the pilots had executed the VOR Runway 17 instrument approach. For a technically proficient crew, which this crew was, the instrument approach alternative would have been routine, and the outcome would likely have had a more successful result.

"Why," as it relates to aviation accidents, is a very complex and challenging question. The attempt to analyze a pilot’s cognitive thought processes extends far beyond the scope of this paper. After all, only the pilot can really answer the question, "what were you thinking?" We can however, use deductive reasoning to look at where some of the problems manifest themselves.

For the sake of simplification, we will look at only two distinct areas, (1) Training facility weaknesses, and (2) Psycho-cognitive threats during routine flight. A breakdown in these areas can pave the way for the highest and most undesirable event; an accident.

Training Facility Weaknesses
Not enough emphasis put on the most unreliable system in the aircraft, (the pilot): Pilot training on a specific aircraft can last anywhere from a few days, up to a few months, depending on the type of aircraft. Training facilities put a large amount of effort into teaching systems in the shortest amount of time possible. And while the importance of good systems knowledge is undeniably important, the most failure-prone system, the pilot, is often overlooked or disregarded.

Crew Resource Management training is weak or non-existent at many facilities: Although many training facilities have begun to incorporate a fair amount of CRM training into their programs, some facilities do not have the time or properly trained facilitators to make a significant impact during a normal training period. After a 2 hr training period, a single CRM debriefing comment by the simulator instructor to the affect of "you should speak up more next time," does not adequately address the problem.

Simulator training time is too compressed. Many emergency/abnormal scenarios that are combined to save time are unfounded and are extremely unlikely to occur in real life: Some facilities, in the interest of time, will combine multiple emergency/abnormal scenarios. It is extremely improbable that a modern airliner or business jet will experience an engine failure and a total hydraulic failure at the same exact time, and that the pilots will have to execute a circle-to-land approach with the weather right at landing minimums. Yet, these are the types of scenarios that some facilities are training and testing pilots on.

"Routine" flight operations are under-emphasized. Yet, routine flight operations claim many more lives than non-routine operations: Inasmuch as the previous topic depicted an overdose of non-realistic scenarios, this topic highlights a relatively untouched realm of training: Routine flight operations. Realistically, engine failures, hydraulic failures, and popped circuit breakers are not killing pilots and their passengers. The largest number of crashes and fatalities occur when nothing is mechanically wrong with the aircraft.

Psycho-Cognitive threats during routine flight
The next level picks up where the training ends. At this point, the crew has satisfactorily completed recurrent training and is back on the flight line. All incidences referenced from this point forward are considered "in-flight."

Keep in mind that all three accident scenarios in this report were due to a failure in human performance, and not a mechanical malfunction. In other words, the problems were not easily identifiable in training, but they became blatantly clear later on.

During flight, the pilot’s psycho-cognitive system performs like a computer, inputting thousands of bits of information, with the associated action commands performed as an output. Occasionally, there is a "short circuit" in these processes and the stage is set for problems.

The following list breaks down the events for Scenario One into CRM marker clusters, as defined in FAA Advisory Circular 120-51D. The author has incorporated additional clusters for clarity. Refer to the figure below for a graphical flow of the Captain’s behavioral patterns.

Proficiency Training- The crew was proficient with no training weaknesses noted.

Illness/Medication- Neither pilot tested positive for alcohol or drugs, including over-the-counter medication.

Fatigue- The crew was well rested

Distractions- Distractions were not considered a significant factor in the accident.

Stress- Stress was low. During the approach phase of flight, stress levels will normally be somewhat elevated.

Workload- Workload was considered routine. During the approach phase of flight, workload will normally be highest.

Task Management- Management of tasks became somewhat ambiguous. A last minute change of the approach procedure by the Captain was a factor.

Communicative Ability- The Captain’s decision to change the approach procedure and not re-brief was the beginning of the "red zone."

Complacency- The Captain displayed signs of complacency. He considered this a routine approach and the weather was good. He had also been into that airport many times before.

Decision Making- Complacency likely influenced the decision making misjudgment.

Personality Traits- Ingrained and hard to change. The Captain’s personality included a large amount of machismo, according to pilots who had flown with him in the past.

Risk Taking- This is the area where decision making and machismo converge. The Captain had decided to "take the risk."

Assertiveness- The First Officer may have had the last chance to trap the Captain’s bad judgment. However, the F/O did not speak up and challenge the Captain.

Situation Awareness- Due to all the previous unmitigated behavior problems, the crew experienced a loss of situation awareness. A perfectly airworthy aircraft was flown into the ground without any prior awareness by the flightcrew.

Scenario 2
The crew successfully completed all of the training requirements required for an annual pilot proficiency exam. As in the first scenario, the crew, as a whole, was highly experienced both in total flight hours and in the specific aircraft type.

This particular leg was a repositioning flight. Only the pilots were onboard, who were flying the aircraft to another airport to pick up their passengers and fly them to their destination. The repositioning flight was very short in duration, approximately 15 min. There were no known aircraft anomalies and the weather was considered good for the short flight. The airport facilities were well suited for the type of operation being conducted by this flight crew.

For some reason, the crew decided to perform a circle-to-land maneuver (which requires excess maneuvering to line up with the landing runway, as opposed to straight-in). During the circle-to-land, something had gone terribly wrong. A post crash investigation revealed that the aircraft experienced a loss of control at a low altitude and airspeed, indicative of an accelerated stall. Both pilots were killed. The investigation also verified no defects to the aircraft that may have contributed to the loss of control.

The following list breaks down the events for Scenario Two. Refer to the figure below for a graphical flow of the Captain’s behavioral patterns.

Proficiency Training- The crew was proficient with no training weaknesses noted.

Illness/Medication- Neither pilot tested positive for alcohol or drugs, including over-the-counter medication.

Fatigue- The crew was well rested

Communicative Ability- The crew had very good communicative ability. Checklists were called for and accomplished at the right time, briefings were thorough, and there was no ambiguity as to what actions were going to be taken next.

Risk Taking- The Captain was not a known risk taker. In fact, interviews with other pilots had indicated that he was rather "passive" and "conservative" in his flying skills.

Complacency- It did not appear that complacency played an active role in this accident. The fact that a circle-to-land maneuver is not executed that often would likely have raised the pilots awareness of the extra vigilance required for that particular maneuver.

Stress- Stress levels were in a high caution area for two reasons: 1) Pressure for a timely arrival to pick up paying passengers, and 2) An approach that requires a higher amount of monitoring due excessive maneuvering at low altitudes.

Workload- The combination of maneuvering the airplane, scanning for other traffic, and properly aligning the airplane with the landing runway significantly increased the workload. This was the beginning of the "red zone."

Task Management- Task management became an extremely high threat as the Captain, who was the non-flying pilot, became task-saturated in both monitoring the F/O’s performance, scanning for other traffic, and ensuring proper alignment with the landing runway.

Distractions- Pervasive distractions during the most critical phase of the approach drew the Captain’s attention away from the F/O’s dangerously slow airspeed and steep bank angle.

Personality Traits- The Captain had a Type-B Personality. Immediate and corrective actions, such as taking the controls of the airplane, may have been hampered by this personality trait.

Decision Making- This is the area where distractions and Type-B Personality converge. The Captain decided to let the F/O continue the approach with no apparent intervention.

Assertiveness- Assertion on the part of the captain in the form of "callouts" or other verbal warnings were lacking. Additionally, the Captain never made an attempt to take over the controls.

Situation Awareness- Due to all the previous unmitigated behavior problems, the crew experienced a loss of situation awareness. A perfectly airworthy aircraft was flown into a dangerous flight regime…an accelerated stall…close to the ground…from which recovery was not possible.

Scenario 3
The third and final scenario once again involved a highly experienced crew. In this case a commuter operation. Their last proficiency check included positive remarks on the captain’s training record reflecting "good use of CRM," and "excellent Situational Awareness." The first officer’s training record had included a notation "excellent flying skills, should be considered for upgrade shortly."

Both pilots were well rested after having the previous four days off. This was their first flight of the day. A routine flight was forecast for the short leg with a passenger manifest of 16 and one flight attendant. The weather would not be a factor. As a matter of fact, it was a remarkably clear spring day. Mechanically, there were no open airworthiness items that could affect the safety of flight.

Approximately one and a half minutes after takeoff, the commuter plane collided with a small general aviation aircraft. Both pilots, all 16 passengers, and the flight attendant were killed on the commuter. Additionally, the pilot and two passengers perished in the Piper Cherokee single-engine general aviation aircraft.

The following list breaks down the events for Scenario Three. Refer to the figure below for a graphical flow of the Captain’s behavioral patterns.

Proficiency Training- The crew was proficient with no training weaknesses noted.

Illness/Medication- Neither pilot tested positive for alcohol or drugs, including over-the-counter medication.

Fatigue- The crew was well rested

Communicative Ability- The Captain’s communicative ability was cause for a moderate amount of caution. In general, he tended to be very impulsive and act without regard to input from previous F/O’s that had flown with him.

Situation Awareness- Situation awareness was also in a moderate caution area. Pushback and taxi to the active runway was routine and the crew was cognizant of their taxi route, departure clearance, and other potentially conflicting ground traffic.

Stress- Stress levels became elevated due to the crew’s acceptance of an immediate takeoff clearance before completion of the Before Takeoff Checklist.

Complacency- The Captain’s decision to be rushed into a takeoff clearance may have been partially influenced by complacency. This can be corroborated by the Captains comment of "I’ve done this a thousand times before," captured on the Cockpit Voice Recorder. This was the beginning of the "red zone."

Workload- During the rushed takeoff procedure, workload increased to a point where there were so many tasks (checklist procedures) that needed to be accomplished in such a short period of time, that some tasks were omitted.

Task Management- Task management was severely degraded due to the reasons stated in the previous cluster.

Personality Traits- The Captain’s personality traits included a Type-A Personality, which can be classified as a very high caution area. Other pilots who had flown with the Captain described him as a person "always in a rush to get things done."

Decision Making- This is the area where task management and Type-A Personality converge. The combination of the Captain’s impulsive behavior, coupled with poor task management, led to a spontaneous and risky decision (acceptance of a premature takeoff clearance).

Risk Taking- Both the Captain and F/O were willing to assume the risk of an immediate departure based on a repetitive and successful schema ("we’ve done this so many times before").

Assertiveness- The First Officer may have had the last chance to trap the Captain’s bad decision. However, the F/O did not speak up and challenge the Captain.

Distractions- One of the items omitted on the before takeoff checklist was turning on the Traffic Collision Avoidance System (TCAS). Shortly after becoming airborne, the crew continued to try to "catch up" on the checklist items that should have been completed prior to takeoff. Both pilots were working inside the cockpit and neither was looking outside for conflicting traffic. Due to all the previous unmitigated behavior problems, a perfectly airworthy aircraft was involved in a mid-air collision.

On a research level, both NASA and FAA have stepped up investigation into this area. NASA’s research on Cognitive Performance in Aviation Training and Operations, and FAA’s AAR-100 Human Factors Division, continue to provide valuable data for incorporation into aviation training programs at all levels.

References

• Federal Aviation Administration (2001). Crew Resource Management Training. Advisory Circular 120-51D. Retrieved January 10, 2004, from http://www1.faa.gov/avr/afs/acs/120-51d.pdf
• Federal Aviation Administration. AAR-100 Human Factors Division. Available at http://www.hf.faa.gov
• National Aeronautics and Space Administration. Research on Cognitive Performance in Aviation Training and Operations. Available at http://olias.arc.nasa.gov/flightcognition/index.htm

By Robert Baron
Presented at the Human Performance, Situation Awareness and Automation Technology Conference (HPSAA), March 24, 2004, Daytona Beach, FL.