T H E AERONAUTICAL JOURNAL JUNE 2002 321

Human factors in multi-crew flight operations

H. W. Orlady
Orlady Associates,
Los Gratos, CA, USA

L. M. Orlady
Orlady Associates,
Annapolis, MD, USA

ABSTRACT
This paper discusses selected aviation human factors challenges in Twenty-five years ago, when he was the International Air Trans-
the air transport environment. Because of the authors' background, port Association's (I AT A) Secretary-General, Knut Hammarskjold
much of this paper will be based upon the growth of human factors opened IATA's 20th Technical Conference by saying, among other
and airline training in transport aviation in the United States. How- things:
ever, parallel growth and advancements have been made in other
parts of the world from the beginning of air transport and are being "Analysis of our accident data clearly indicates that our historic
made in many parts of the world today. This is certainly true of trends towards reducing accident rates can only be resumed if we
Great Britain. One should never forget that air transport is a world address and solve the problem of why critical human beings fail
enterprise and has been since its beginning. While it is important to to fulfill expectations. These accidents-fifteen or twenty years ago
recognise that cultural and other differences should be considered if would have been superficially dismissed as due to 'pilot error' or
they are relevant, it is equally important in air transport to recognise 'controller error.' We now know beyond a shadow of doubt that
that common problems are solved by common solutions and can be these descriptions of accident causes are at best misleading and at
addressed by, but are not limited to, common approaches. worst irresponsible, (underlining supplied).
It is recognised that the behavioral aspect of aviation human fac- The issues we now call 'human factors' ... are the critical is-
tors deserves considerable attention, particularly in the area of train- sues ... I believe that at this point in the history of civil aviation
ing and evaluation. We have also learned that previous training crite- there is enough knowledge available to launch a serious and
ria are not satisfactory for operation with increasingly complex effective attack on this — the 'Last Frontier' — of the airline
aircraft being operated in an increasingly complex environment. problem."<0
Human errors occur with greater frequency than is often believed
and can be made by anyone — even an airline captain. An inevitable Knut Hammarskjold was, of course, right, although IATA was
consequence of this entirely human characteristic is that the role of forced to admit at the conclusion of the conference that many of its
the co-pilot has been considerably enhanced. Furthermore, the moni- airlines were not yet ready for a human factors-based review of their
toring function of the pilot-not-flying, whether he/she is the captain operations. It has taken us most of the intervening years to learn,
or co-pilot, is now recognised as a critical safety function by virtual- among other things, that a large part of our new frontier is really to
ly all serious operational experts. learn more about the operational behavior of the individuals who
work in our system.
The industry has made considerable progress. For example, the
1.0 AVIATION HUMAN FACTORS International Civil Aviation Organization, (ICAO), passed Resolu-
tion A26-9 (now combined with Resolution A29-7), which
Since the beginning of this industry, the growth of human factors in addressed Flight Safety and Human Factors. IATA has held numer-
aviation generally, and in multi-crew flight operations in particular, ous Global Symposia and Regional Seminars that focused specifical-
has been evolutionary rather than revolutionary. While the growth in ly on human factors issues. In addition, IATA, ATA (Air Transport
recognition of the importance of human factors has often been some- Association), IFALPA (International Association of Air Line Pilots)
what sporadic, overall it has been reasonably steady. It is worth and the regulatory branches and pilot unions of many countries now
remembering that a major reason for the rate of progress of our treat regulatory aviation human factors as a 'core technology' in air
industry and its highly successful present state has been the inherent transport. The industry continues to learn about its role and impor-
conservatism of its operations. tance with safe flight operations.

Paper No. 2707. Manuscript received 18 July 2001, first revised version received 15 November 2001, second revised version received 8 January 2002,
accepted 25 January 2002. Paper first presented at the Flight Simulation 2001 Conference, Aircrew Training — Time to take stock, 16-17 May 2001, London, UK
Opinions stated in this paper are those of the authors. They may or may not reflect the position of the airline or other organisations with which the authors have been or
are presently associated.
322 THE AERONAUTICAL JOURNAL JUNE 2002

1.1 Human factors today requires a broad view lenges to both ground and flight instructors in terms of curriculum
design and performance standards.
Several years ago the individual in charge of human factors in the
official analysis of accidents in the US carefully explained to a US
pilot representative that in accident investigations, his principal 1.3 The 'team concept' and CRM
human factors concern was restricted to the blood alcohol level of
the pilots that were involved and in the 'G' levels of the fastenings An emphasis on a new look in airline operations, at least in the US,
that secured passenger seats to the cabin floor. We, and certainly started about the time (1987) that T. Alan McArtor, as the US Feder-
most accident investigators the world over, have come a long way al Aviation Administrator, told a group of airline executives, "Indi-
from this narrow viewpoint. However, the operating, regulatory, and vidual pilots don't crash, flight crews do."<2> Up to that point regula-
training communities cannot wait for seemingly inevitable accidents tors had not spoken this way, and neither had most operational
or incidents to again show us that we must consider human factors experts.
and especially its behavioral aspects in the operation of our aircraft. It was an important point, and is even more important today. The
Managing risk, trying to trap human error and control its operational universality and inevitability of human errors, as well as recognition
consequences, developing and teaching cognitive maps of automated that even a captain can make an inadvertent mistake reinforce these
systems that are meaningful to the end user-the pilot, and continuing concepts. It is a simple fact that one person, no matter how talented,
to strive to make use of incident data in a non-punitive fashion are cannot routinely day-by-day fly today's aircraft in today's dynamic
major challenges today. environment without being susceptible to entirely inadvertent human
We believe in taking a very broad view of aviation human factors error. Recognition of this point has led to the "team approach" in the
and certainly one that includes and emphasises human behavioral operation of today's aircraft as well as to expanded roles for the
factors. Contemporary aviation human factors practitioners and con- copilot and other crewmembers, including the cabin crew. More
temporary operational experts must paint from a pallet with a wider recently, it has also led to a major effort at error recognition and er-
variety of colors than has been their custom in the past. And, we ror management.
must use this pallet in the design aspects of the new technologies as Crew resource management (CRM) programmes have existed in
well as in their training, checking and implementation. various forms since the early 80s. The concepts have received new
emphasis with both the transition to two-pilot cockpits and with the
advent of advanced technology. While automated systems have re-
1.2 New challenges today duced errors in many areas, they also have created some new prob-
lem areas such as mode awareness and mode confusion. Good CRM
Today's aircraft with their many changes bring new challenges. One programs not only help detect and control new problems and inad-
of the most noticeable involves crew composition and the change vertent human errors but also provide a method of flying that in-
from three-person to two-person cockpit crews regardless of the air- creases safety and is more efficient.
craft gross weight or the number of engines on the aircraft. Other
Developing CRM and the 'team concept' has not been without its
changes involve the increasingly complex aircraft with their embed-
problems. For example, there are many different training approaches
ded computers and associated flight deck displays and documenta-
to crew resource management (CRM) including differences in the
tion, and the increasingly complex environment in which all trans-
definition of the term itself. For a long period, such training was ei-
port aircraft must be operated.
ther not required by the regulators or not required to be evaluated in
The tremendous range of some recently built transports creates a the same manner as a technical skill. There is still far from complete
major challenge. This increased range makes schedules possible that agreement regarding the measurement and evaluation of the behav-
are desirable for passenger service but that create special challenges ioral factors that are involved. Establishing the validity and reliabili-
for the flight crew. These challenges include areas such as crew rest ty of these factors, as they apply to a particular operator, is a signifi-
areas, duty time, fatigue, crew augmentation, and takeoff and land- cant challenge although progress continues to be made.
ing currency.
While special rest areas for both flight deck and cabin crewmem-
bers are provided to deal with crew rest and general fatigue in 1.4 Training, safety and economics
aircraft such as the Boeing 747-400 and the Airbus A330/340, other
challenges are not as easily addressed. One of these is that the ex- There are at least four areas of training in transport operations that
tended range of long-range aircraft permit schedules that do not pro- must be considered up front. First, there is no question but that safety
vide enough takeoffs and landings for the flight crews to maintain will continue to be the prime objective of all operating parts of the
required qualification. industry. However, economic factors have also become a require-
For example, a Chicago-Hong Kong-Chicago flight has only two ment in the increasingly competitive environment that is an inherent
takeoffs and two landings for each round trip. On a typical US carri- part of air transport today. It is an absolute requirement that any sys-
er, this trip is flown with four pilots. Because of federal and contrac- tem, procedure, or aircraft be safe first, and then both economical
tual scheduling and flight time limitations, it is not possible for each and efficient.
pilot to fulfill the US FAR (Federal Aviation Regulation) require- Secondly, good training has always been expensive, and unfortu-
ment that requires each pilot to make three takeoffs and three land- nately, good training will continue to be very expensive. However,
ings every 90 days in aircraft in which they are qualified. In order to good training will always be cheaper than an accident, which is fre-
meet this very real problem of currency, many airlines provide addi- quently the cost of inadequate training. A very real problem arises
tional simulator periods that enable pilots to maintain currency and when a part or all of a company's training is done by either another
also help maintain important manual skills. For example, one major airline or by a specific training organisation. It is absolutely essential
US airline simply schedules all B747-400 pilots for an advanced that the trainee airline's procedures and operating philosophy be
simulator training session every ninety days. accurately reflected in the training that its pilots receive. This addi-
Scheduling additional simulator takeoffs and landings does not tional step can be expensive and has not always happened.
completely address the importance of maintaining pilot manual skills Thirdly, we all know but sometimes forget that airlines are highly
in a long-haul aircraft, or in any highly automated, advanced tech- unionised. There is little doubt that in the future airlines will remain
nology aircraft for that matter. This topic will be discussed later in highly unionised. This fact has training implications. Among other
this paper. things, it means that pilot representatives will participate actively in
Aviation human factors is a vital component to address in today's all aspects of training. While the level of this participation may vary
changing environment. These changes present challenges to opera- among different carriers, it provides great opportunities to both man-
tors, regulators, and pilots. They also present new and distinct chal- agement and the unions.
ORLADY AND ORLADY HUMAN FACTORS IN MULTI-CREW FLIGHT OPERATIONS 32.

For example, a co-operative process of developing training and requirement to stay over for an extra day of training or the need to
checking activities should result in pilot support for these activities. return for additional training has punitive connotations, regardless of
Furthermore, the active participation allows both the trainee and the the need, purpose, or stated intent of the additional training.
training organisation to receive meaningful feedback regarding all
aspects of training, checking, and procedures. Unfortunately, there
can be times when the free flow of information is 1.6 What level of knowledge is required
impeded on both sides by various collective bargaining considera- Another problem for today's operator and its pilots is to determine
tions. This is unfortunate but a 'fact of life'. It is important for both exactly what level of knowledge or understanding is required for a
sides to recognise that in training, both sides truly have common pilot to safely and efficiently fly a specific advanced technology or
goals although they sometimes disguise or disregard them. 'glass-cockpit' aircraft. Within the industry there is far from agree-
A fourth area is another 'fact of life'. Both many present airplanes ment on this basic question. There are several reasons for the dis-
and many present operating personnel (pilots, cabin crew, mechan- agreement, including differences in the operating environment and
ics, dispatchers, and managers) will continue to be an important part the background of the pilots.
of the airline operation. Aircraft are planned to have a useful life of For example, at one US airline the knowledge requirements for a
at least 25 years, and personnel careers lasting 40 years are not particular mode of vertical navigation (VNAV) on the Boeing 737-
uncommon. We cannot train only for the last technology and assume 300 were changed several times during a four-year period. The re-
that all trainees have just stepped out of the latest advanced techno- quirements varied from not teaching this particular mode at all to
logy or glass cockpit aircraft we have in service. teaching partial aspects, more aspects, and then returning to not
teaching any with the expectation that the pilot would pick them up
on the line. The industry continues to disagree as to how much sys-
1.5 Evaluating crew vs individual performance tem knowledge a pilot should have and, unfortunately and under-
Evaluation of the relevant behavioral issues can cause a very real standably, the flight manuals frequently have not been really helpful.
problem during training and proficiency checking, despite a genuine The manufacturers have made good-faith efforts to include more
aim and then claim by the operator of 'training to proficiency'. The information, but the operators are allowed to choose what informa-
dilemma is straightforward. Both the regulators and the operators tion they will use and may even develop their own system hand-
have a clear obligation to ensure that competent pilots fly the aircraft books on a particular aircraft. Ultimately, the FAA's Principal Oper-
in air transportation. ating Inspector (POI) for that airline must approve the required
information and all manuals and handbooks.
While it has been well-established that people learn better in a
non-punitive environment and many carriers try to create a non- The pilots themselves also question how much knowledge of the
punitive environment in training — for example with various ver- automation they should possess, or more importantly, how much
sions of LOFT or LOS (line-oriented flight training and line-oriented knowledge they will need on the line. For example, on some Airbus
simulation) — it is clear that the operators and the regulators have a aircraft, it is difficult to track down some anomalies because so
definite obligation to ensure that pilots maintain their competency. It many of the systems interface with different computers within the
is generally agreed that this obligation is only met by actual demon- airplane. A particular difficulty can be assessing mechanical prob-
strations of proficiency. Proficiency checks have been a part of this lems before departure and trying to understand what sort of redun-
business since its inception, and, even under the guise of training, dancy and how much redundancy one has with a particular anomaly.
are not seen by most pilots as being non-punitive. Researchers have often pointed out that pilots need to have a better
In the difficult mix between a punitive (or checking) and non- understanding of the levels of automation in their aircraft to more ef-
punitive (non-checking) environment, the trainer runs into at least fectively use them and to understand how to reduce the levels of au-
three problem areas. All of them address the dilemma of evaluating tomation when necessary or desired. These researchers sometimes
unsatisfactory individual performance that may be couched within voiced this concern by saying that the pilots do not have a good
acceptable crew performance. Here are three difficult scenarios that enough 'cognitive map' of the system. Many pilots would echo this
a trainer or evaluator might encounter: frustration.

1. A weak PF (pilot-flying) who receives no help from a stronger

1.7 Retention of manual skills
PNF (pilot-not-flying), who recognises the problems and knows
better but decides not to help the PF who really needs him/her. Retention of manual skills continues to be a troubling area. Training
programmes have varied in terms of how much emphasis should be
2. A weak PF, knowingly or unknowingly, depends upon the PNF put on pilots manually flying the aircraft. At one airline, for exam-
to help pass the check, which incidentally the cockpit team may ple, pilots are allowed to use as much or as little automation as they
pass satisfactorily because of good PNF performance. desire during a checkride. On another airline, they are required to
use all of the automatics virtually all of the time. And at yet another
3. A PF with very strong technical skills meets most of the criteria airline, the operator ordered their aircraft without autothrottles
in the training scenario but clearly does not believe in the crew although they have all of the other traditional automatics.
concept and does not use, or only minimally uses, the PNF and Most pilots have a keen interest in maintaining their manual skills.
other resources. They do not always have as much interest, or are not trained as well,
in understanding how to reduce levels of automation to practice
Clearly, some of these scenarios are avoided through very careful these skills on the line. Turning off some or most of the automatics
scenario design. For example, if a PF seems to be overly dependent can significantly increase the workload of the PNF if the PF has not
on the PNF, the trainer may be able to introduce an element to conducted an adequate briefing or is not meeting performance para-
address that dependence. While not always possible, a fairly effec- meters such as those specified in a stabilised approach. On the other
tive technique would be to 'incapacitate' the PNF where the trainer hand, continuing to fly with all of the automation can create last
would give such instruction to the PNF through a discrete headset minute programming problems if, for example, the pilots are given a
transmission. last minute runway change even in visual flight conditions. Some
Today most airlines simply give pilots deficient in these areas carriers address this potential distraction by developing standard
additional training. However, it is not always easy to decide who operating procedures that prohibit programming below 10,000 feet
should get the training, what training is required, or how much train- because it can interfere with a safe operation.
ing should be specified. From the viewpoint of most pilots, the Long-haul pilots have a special challenge maintaining skills —
324 THEAERC ., JOURNAL JUNE 2002
•HHMHIHHHHHHHHHHfHMHflHHMIIHNHMHMHHsiHHHHil

both with using the automatics and with manual skills. By the nature 1.10 Zero accidents, zero incidents and zero risk
of their flight mission, it can be difficult or impossible to get enough
takeoffs and landings to maintain either the competency they desire Although certainly not the originator of the concept, about ten years
or required qualifications. Complacency and over-reliance on very ago the then US Secretary of Transportation, Frederico Pena, stated
reliable systems require a special vigilance in a flight environment that in his department, air transport's official goal was "zero acci-
that is often cloaked in fatigue. dents and zero incidents". That goal, however, does not equate to
"zero risk" and as John Lauber has stated: "the talk of 'zero risk' has
done a disservice to the public... 'zero risk' is not possible. Aviation
1.8 Handling the rare but inevitable failure operations are always going to be risk-intensive — and we should
put that up front."*3) A more reasonable and pragmatic objective is to
Finally, a continuing general problem that is implicit in the training try and manage the risk.
business, is that the industry will always have to deal with the Today, the basic industry approach is to minimise the occurrence
extremely rare case of the pilot who, for any number of reasons, does of inevitable errors, to detect those errors that do occur, to minimise
not meet proficiency standards. In addition to the obvious humani- their operational consequences, and lastly, to learn from the errors.
tarian factors involving aspects such as the pilot's general well-being While it is not easy, we believe that it is the right approach. We also
and the prolongation of his or her career, there are substantial eco- believe that the probably unobtainable goal of "zero accidents and
nomic reasons to retrain or provide additional training to problem pi- zero incidents," especially on a world-wide basis, should always be
lots in a bona fide effort to save their careers. An entirely new type the goal of the air transport industry for as Benjamin Howard, an
of human factors can be involved, and this can mean another type of early airmail pilot, airline pilot, test pilot, and aviation safety expert,
task for the trainers. One must remember that while training is very said many years ago, "Man's greatest sin is the unnecessary taking
expensive, it has also become increasingly difficult to fire a pilot of human lives."
without the attempt of meaningful rehabilitation measures. This can
be for collective bargaining or simply legal reasons.

REFERENCES
1.9 Transitioning to, from and between glass cockpits 1. HAMMARSKJOLD, K. Secretary-General of IATA in his Opening Address
Within each airline there can be several different types of transition to IATA's 20th Technical Conference, Istanbul, 10-14 November, 1975
as cited in Human Factors in Multi-Crew Flight Operations, ORLADY,
training courses and each may require a different approach. Nearly H.W. and ORLADY, L.M. (1999), Ashgate Publishing, p 53.
all operators face the problem of giving standard technology pilots 2. Human Factors in Multi-Crew Flight Operations, p 268.
their initial glass cockpit experience. In the US, airlines typically 3. Human Factors in Multi-Crew Flight Operations, p 426.
now devote one or even two days to simply making the transition
from standard technology to advanced technology or glass cockpit
aircraft. Virtually all of these airlines believe that special training
for those making their first transition to glass cockpit aircraft is BIBLOGRAPHY
definitely worthwhile. Many of these carriers have developed an HAMMARSKIOLD, K, Secretary-General of IATA in his Opening Address to
automation philosophy specific to their operation that is also IATA's 20th Technical Conference, Istanbul, 10-14 November 1975, as quoted
included and communicated in the course. This practice is highly in Human Factors in Multi-Crew Flight Operations, p 53 (ORLADY &
commended. ORLADY, Ashgate, 1999).
Another type of problem concerns the glass cockpit copilot who HOWARD, B. The attainment of greater safety, 1954, presented at the 1st
goes back to fly a standard technology aircraft for his/her first at- Annual ALPA Air Safety Forum and later reprinted for presentation at the
tempt to be a captain. Many of the formerly glass cockpit copilots Aircraft Accident Prevention Course, University of Southern California, July
1957.
rely upon automatics that are no longer available in the standard
International Civil Aviation Organization. Human Factors Digest No. 13,
technology aircraft. For example, some performance calculations Proceedings of the Third ICAO Global Flight Safety and Human Factors
that were formerly instantly made by an embedded computer must Symposium, August 1996.
now be manually calculated. LAUBER, J.K. as quoted in Human Factors in Multi-Crew Flight Operations,
Another transition training problem arises when there is a change p 426 (ORLADY & ORLADY, Ashgate, 1999).
from one manufacturer to another or even when a single manufac- ORLADY, H.W. and ORLADY, L.M. . Human Factors in Multi-Crew Flight
turer uses different nomenclature for an operational device that per- Operations, 1999, Ashgate Publishing.
forms essentially the same function. In spite of attempts to minimise RASMUSSEN, J., DUNCAN, K. and LEPLAT, J. (Eds.) New Technology and
differences, and therefore training costs, pilots changing aircraft Human Error, 1987, John Wiley & Sons.
from different manufacturers and, in some cases the same manufac-
turer, have to forget old habits and learn new procedures, new
acronyms, and even new automation philosophies. This task is not
always easy to do and creates problems both for the trainer and the
student.
Such transitions can create problems for the trainer who is quite
familiar and competent with the new aircraft but may never have
been exposed to the older aircraft or to the other manufacturer. An
easy example is a pilot starting an Airbus transition program but
saddled with the knowledge and experience of previously having
flown Boeing glass cockpit aircraft. The Airbus instructor/trainer
may have no familiarity with Boeing and, therefore, unable to ad-
dress questions of similarities and differences — a situation which
can be particularly frustrating to the transitioning pilot.
Such problems, while difficult, are not particularly new because
air transport is not a static business. Newer aircraft invariably in-
clude advancements that must be a part of the training programme.
Special attention can be made, however, to instructor training and to
recognising the trainee pilot's previous background.