Paper Presented at the 22nd International Seminar
International Society of Air Safety Investigators
Canberra Australia, November 1991

DOES COCKPIT MANAGEMENT TRAINING REDUCE AIRCREW ERROR?

Alan E. Diehl, Ph.D.

Technical Advisor for Human Performance
US Air Force Safety Agency
Norton AFB, California

Introduction

Human factors problems continue to be involved in majority of mishaps. Thus, the causes and cures for aircrew "error" are widely discussed topics. Several experts (e.g. Bruggink, 1978, Miller, 1979 and Nance, 1986) have noted that labels like pilot error" are often misapplied in describing ergonomic, management, regulatory or systems design shortcomings. This paper uses the generic term "cockpit management" when referring to the wide variety of programs which are designed to reduce aircrew errors. In fact, recent evidence suggests Cockpit Resource Management (CRM) and Aeronautical Decision Making (ADM) training may help reduce aircrew error accident rates by as much as 81 percent.

These programs have only emerged in the last decade largely because of the fundamental problems associated with detecting and controlling human error. Many of us have lamented the greater difficulties of accurately documenting human vis-a-via mechanical failures, (e.g. mental fatigue is usually tougher to prove than metal fatigue). It is also often harder for us as air safety investigators to specify effective countermeasures in the human factors domain. Thus, some organizations unfortunately have assumed such errors are "the price of doing business."

In these times of tight budgets, management and government authorities can be expected to demand proof that preventive measures are in fact working. Here again, proving the effectiveness of human factors initiatives is very difficult, (e.g. crews can always comply with unpopular standard operating procedures during check-rides). Lastly, experts such as Dr. Clay Foushee (1987) have aptly noted that accidents, because of their relative infrequency, make poor scientific criteria. It is also axiomatic that proving the negative (accidents which were prevented) is even more difficult.

This paper addresses these issues by: 1) Examining the prevalence of major types of contemporary errors and reviewing the traditional methods which have been used to improve human reliability, 2) Discussing how innovative cockpit management training programs were developed and implemented, and 3) Describing the current evidence on the effectiveness of such programs.

Taxonomy of Errors

In their classic study, Jensen and Benel (1977) noted all aircrew errors could be classified into one of three major categories based on behavioral activities: Procedural, Perceptual motor, and Decisional Tasks. Examples of procedural tasks include management of vehicle subsystems and configuration, while related errors would include retracting the landing gear instead of flaps or overlooking checklist items. Perceptual motor tasks include manipulating flight controls and throttles, while errors would include over shooting a glide-slope indication or stalling the aircraft. Decisional tasks include flight. planning and in-flight hazard evaluation, while errors would include failing to delegate tasks in an emergency situation or continuing flight into adverse weather. These researchers also noted while the term 'judgment" is sometimes used with both perceptual motor and decisional processes, it should be more closely associated with the complex cognitive processes involved in human decision making.

They analyzed all US general aviation accidents occurring from 1970 to 1974 using the National Transportation Safety Board (NTSB) computerized data base. Their analysis of the fatal accidents involving pilot error indicated that 264 were procedural, 2496 were perceptual motor, and 2940 were decisional in nature.

My recent paper (Diehl, 1991b) analyzed military and airline accident data, for comparison purposes, using the Jensen and Benel (1977) taxonomy. The NTSB computerized accident data base was examined for US airline (and scheduled air taxi) accidents occurring during 1987, 1988, and 1989. This data indicated that 24 of the 28 major accidents (those resulting in destroyed aircraft and/or fatalities) involved aircrew error. In these accidents there were 16 procedural, 21 perceptual motor, and 48 decisional errors. The relative percentages of these errors are depicted in Table 1, along with the previously discussed general aviation data and that of another study involving military accidents.

The computerized data base was examined for US Air Force (USAF) Class A flight mishaps. These were mishaps involving the destruction of the aircraft or over one-million dollars in damages and/or fatalities. The period reviewed included the data for fiscal years 1987, 1988 and 1989. Note US Government fiscal years begin on first day of October.

Here 113 of the 169 mishaps involved some type of aircrew error. Included were 32 procedural, 110 perceptual motor, and 157 decisional errors. These types of errors were labeled 'slips", "bungles" and "mistakes' respectively (Diehl, 1989). Thus, this data collectively reveals the significance of Decision Making to all segments of aviation.

Table 1

Types of Aircrew Errors in Major Accidents

Improving Aircrew Reliability

Over the years great strides have been made in improving the mechanical reliability of aircraft and systems, while various means of enhancing human reliability have also been undertaken. Some concepts, designed to prevent aircrew error, were undertaken even before the phenomena was thoroughly understood. As noted above, oftentimes a priori scientific proof was unavailable. A proverbial "Catch 2211 may have existed such as ' We can't use it until we know it works, and we'll never know it works until we try it." Other times innovations may have just been institutionalized common sense, (e.g. the development of written checklists as aircraft became increasingly complex). Such countermeasures have focused on improving six human 'faculties" fundamental to flying.

These-faculties constitute the "right stuff" and include: 1) abilities, 2) motivations, 3) knowledge, 4) procedural techniques, 5) perceptual motor skills, and 6) decisional judgment. As I recently noted, Diehl (1991b), these six items are part of a "Hierarchy of Aeronautical Faculties' with abilities on the base and decisional judgment at the apex, as the highest faculty. Figure 1 depicts this relationship along with the role played by traditional and modern preventative measures.

Obviously, the three categories of errors discussed above are failures associated with the three higher faculties (procedures, perceptual motor, and judgment behaviors). While the other faculties (abilities, motivations, and knowledge) are enabling factors which underlie such behavioral tasks.

Interestingly, an individual student aviator generally undergoes screening and training processes which proceed systematically from the base to the apex of this hierarchy. Not surprisingly, this is also the fundamental sequence in which the aviation industry attacked the problems associated with human error.

Abilities: Medical screening tests were successfully used in the -First World War. Early testing focused on easily measurable items such as cardiovascular health and visual acuity. Basic mental capacities were also measured. Such procedures were, of course, improved over the years by various military and civilian organizations.

Motivations: Mental tests for screening aviation candidates' personalities and interests have been widely used since the beginning of the Second World War. These psychological instruments, like the medical screening protocols, have been constantly refined. In addition, modern aviation organizations use a variety of measures such as employee assistance programs to enhance the physical and mental faculties of their employees.

Knowledge: In the post-Second World War era, greater emphasis was placed on screening the prospective airman's general knowledge, (e.g. a college degree became a pre-requisite for many military and commercial pilot training programs). Imparting the vast amounts of specialized knowledge (consisting of information and data as well as rules, concepts and principles) has been a major function of aviation training.

Great strides have been made in effectively imparting knowledge. For instance, standardized formats in flight manuals were in wide use by the 1950s. Instructional Systems Design concepts have been used since the 1960s to systematically identify "need-to-know" versus "nice-to-know" information. In recent years, computer based training applications have increased the efficiency of teaching aeronautical knowledge. This knowledge was always regarded as a prerequisite for learning procedures, skills, and judgment tasks.

Hierarchy of Aeronautical Faculties

Figure 1

Procedural Techniques: These "finger faculties' were necessary to manipulate the switches, buttons, and knobs of aircraft subsystems. As on-board equipment became more complex, these techniques took on expanded importance. By the 1950s, cockpit design standards and crew checklists were in widespread use. The decades which followed saw increasing use of devices such as cockpit procedures trainers to enhance the mastery of such tasks.

Perceptual motor Skills: The importance of these "stick-and-rudder" tasks have always been recognized. Good hand-eye coordination was a prerequisite for the timely maneuvering of an aircraft through three-dimensional space. Although, automation and stability augmentation systems have decreased somewhat the amount of time crewmembers now spend on basic aircraft control activities. Control-display integration and fused sensors have further decreased this type of workload. The use of modern digital simulators has facilitated the efficient acquisition of these skills, especially since the 1960s.

Decisional Judgment: "Headwork" or cognitive tasks were also regarded as vital. But such abilities were historically assumed to be a by-product of flying experience, or taught only informally. Little was known about this faculty until the 1970s. In that decade, cockpit voice and flight data recorders, as well as more systematic accident investigation methods, had revealed the magnitude of judgment, crew management and situational awareness problems. Moreover, human factors research began suggesting the possibility of formally teaching cockpit management tasks.

In that decade, the Federal Aviation Administration (FAA) had initiated the aforementioned study into potential methods of teaching judgment concepts to general aviation pilots (Jensen and Benel, 1977). At this time, USAF was examining methods to improve the workload management techniques of fighter pilots during emergencies (Thrope, Martin, Edwards and Eddowes, 1976). Meanwhile, the National Aeronautics and Space Administration (NASA) had simultaneously undertaken several comprehensive programs focusing on methods of reducing crew errors in transport aircraft (e.g. Ruffle-Smith, 1979). Several airlines we ' re then in the process of developing important training innovations. Line Orientated Flight Training (LOFT) was pioneered by Northwest Airlines to improve crew coordination in simulators. United Airlines had also initiated simulator research into subtle incapacitation recognition, while KLM was developing a course to teach leadership skills of their line captains. The latter program was, of course, undertaken as a result of their 1977 Tenerife accident.

Implementing Cockpit Management Training

This then was the state of affairs in our industry when a United Airlines DC-8 crashed into a suburb of Portland Oregon, on the evening of Dec 28, 1978. I was dispatched to this, now well known, accident as the NTSB Human Factors Group Co-chairman. The circumstances of this mishap were quickly established: The highly experienced crew became distracted' by a landing gear problem and ran out of fuel. After reviewing the reports of similar accidents and the existing research, I drafted the first recommendation calling for the operational implementation of cockpit resource management programs by US airlines (NTSB, 1979).

"Selling" this recommendation to the leadership of the NTSB was not difficult. For they were easily persuaded that CRM was "an idea whose time had come." These programs have gone into widespread use in the last dozen years. But, not-,without much debate about their effectiveness, which continues to this day.

Since writing that first CRM recommendation, I have found myself immersed in the continued advocacy, development, and evaluation of such programs. These endeavors have included helping extend the application of cockpit management concepts to general aviation and military users. For example, my first assignment after becoming the FAA Program Scientist for Human Performance was to monitor the initial experiments on the effectiveness of the prototype judgment training courses (Buch and Diehl, 1984). When the first airline accident occurred to a crew which had received CRM (NTSB, 1983), the FAA asked me to examine methods of improving the certification of these programs (Jensen, 1987). The USAF Inspection and Safety Center (now redesignated as the Safety Agency) recently tasked me with analyzing data on the effectiveness of civil and military cockpit management courses. The results of that research (Diehl, 1991b) will be discussed below.

This past year has seen a accelerating interest in such training. For instance, the NTSB as part of a major general aviation accident investigation, has recently recommended that Aeronautical Decision Making training be implemented among all categories of pilots in the civil aviation community (NTSB 1991). Similarly, all the USAF major commands now have adopted some type of cockpit management training program (Diehl, 1991a).

Components of Cockpit Management Training

Much has been written in recent years about methods for enhancing the collective Decision Making in multi-place aircraft through the use of CRM techniques (Lauber, 1984, Nance, 1986, Alkov, 1988, Foushee and Helmreich, 1988, and Helmreich, 1991). In contrast, less information is available on programs aimed at improving the Decision Making abilities of individual pilots (Buch and Diehl,

1984) Note that the term "aeronautical Decision Making" (ADM) has become synonymous with "judgment training" in recent years. Furthermore, categorical distinctions between CRM and ADM are disappearing in that today most comprehensive versions of these programs have several common functional components dealing with: attention, crew, stress, mental attitude, and risk issues.

The role which the five components or 'cockpit management tools" play in the hierarchy of aeronautical faculties is depicted in Figure 1. These tools may, in effect, provide 'synthetic experience" for neophyte airman while offering a structured system for assisting the Decision Making of their experienced counterparts. These five interrelated concept areas furnish 'rules and tools" to help prevent common errors. For instance:

1. Attention management issues include understanding how distractions and "error chains" can be avoided.

2. Crew management issues teach the importance of proper communications, division of responsibilities, leadership, and teamwork.

3. Stress management concepts focus on understanding the effects of life stress events as well as providing in-flight stress coping strategies.

4. Attitude management concepts describe the methods of recognizing and controlling certain hazardous attitudes and behavioral styles.

5. Risk management issues focus on the rational evaluation of qualitative and quantitative information related to operational hazards.

Cockpit Resource Management: Note that Crew Resource Management is another popular label for such courses. The conceptual basis of these programs was largely social psychology and management theory. Many of these programs were developed and refined with the data and expertise from NASA (Lauber, 1984, and Foushee and Helmreich, 1988).

Most contemporary CRM programs utilize training manuals, interactive classroom lectures, with audio-visual aids followed by LOFT sessions which are video taped for critique purposes. The courses provide a wealth of techniques to enhance flight deck communication. For example, avoid 'excessive professional courtesy": If the captain is two dots low on the glide-slope, tell him so in unequivocal terms. Don't say, ' You're a little low, Sir".

United Airlines has arguably fielded the most widely used of these courses, although KLM launched the first such course aimed at captains. Both airlines have successfully marketed their programs to other aviation organizations, and they continue to refine and enhance their respective programs (e.g. Freeman and Simmon, 1990 and Siemons, 1991). Other airlines (e.g. Quantas) have also independently developed highly innovative programs (Beaumont, 1989).

The USAF Military Airlift Command and the US Naval Safety Center have pioneered militarized CRM programs (Alkov, 1988), labeling them Aircrew Coordination Training (ACT). The USAF Strategic Air. Command has recently fielded a very comprehensive CRM program under contract to Hernandez Engineering. This course has, in turn, been adapted for training USAF fighter crewmembers stationed in Europe.

The USAF Inspection and Safety Center initiated the development of a prototype course focused on single-seat fighters in 1990. It was intended to improve intercockpit as well as intracockpit Decision Making. This program was a modified and enhanced version of the successful US Navy ACT program developed by CAE-Link. Units of the US Army have applied this same course to utility and attack helicopters. The USAF and USN training commands have also integrated these materials into their respective undergraduate pilot and navigator training. In 1992 the USAF T-1 "Jayhawk" trainer will become the first operational system procured with CRM specified in its design.

Aeronautical Decision Making: The conceptual basis of the ADM or judgment training programs was cognitive psychology, for this type of training was aimed at the attitudes and behavior of the individual pilot. Most of these programs originally focused on students pilots (Berlin, Gruber, Holms, Jensen, Lau, Mills and O'Kane, 1982), but the concepts were later applied to a advanced training including commercial, instrument, and helicopter pilots (Diehl and Buch, 1986). The FAA, Transport Canada, the Australian Aviation Department, and the USAF have sponsored the development and evaluation of ADM programs. These materials typically consist of training manuals and audio-visual products which explain fundamental concepts related to error causation and prevention.

The way these materials work is illustrated in Figure 2 from the student pilot training manual, (Diehl, Hwoschinsky, Lawton, and Livack, 1987). This figure depicts the Decision Making process as a series of feedback loops in which the pilots must manage his/her attention in a timely manor and sequentially employ the other cockpit management tools (for controlling stress etc.). The text describes how one does these things. Interestingly while this figure is somewhat simplistic in ' comparison with flow charts in more sophisticated texts (e.g. Reason, 1990), it does comport with ideas on pilot information processing offered by experts like Lee (1990).

AERONAUTICAL DECISION MAKING PROCESS

Figure 2

Situational Awareness: other similar cockpit management training programs have focused on enhancing attention and task management issues. For instance, when the USAF began replacing its two place F-4 with the single-seat F-15 in the 1970s, concerns were raised about pilot workload in emergencies. Situational Emergency Training was undertaken using cockpit procedure trainers. Thus, pilots could practice diagnosing typical emergencies, while maintaining aircraft control, rather than relying on memorized "bold face" procedures.

In the early 1980s, the US Air National Guard became concerned about the ability of their A-7 pilots to maintain proficiency in the low-altitude tactics. Their Low Altitude Training program was undertaken to teach pilots techniques for overcoming the unique hazards of operating in this highly dangerous and time critical environment (e.g.-the tendency to fly lower over small desert bushes because, at high speed, they appear to be the same size as the larger trees which one is used to). This program included academics, simulator, and flight training.

Another important situational awareness training effort was just announced by the USAF Tactical Air Command. Their Aircrew Attention Awareness Management Program is designed to acquaint fighter pilots and weapon systems officers with physiological and psychological factors affecting their performance. These concepts are being taught in part by specially trained physiologists (familiar with CRM, ADM, etc.) who have been assigned to each fighter training unit.

Measuring Training Effectiveness

These training programs have all been generally well received by the individuals and organizations which have used them. Much contemporary research has described the improvement the attitudes of people enrolled in such programs. But, for the reasons noted earlier, proving that the programs have prevented errors or reduced mishaps rates is difficult.

Fortunately, there is anecdotal information that such programs have helped prevent mishaps in a wide variety of civilian and military aircraft. For instance, a US Navy A-6 crew experienced a total hydraulics failure, but was able make a safe landing. The investigation concluded that this was a 'first-ever' in that type aircraft, and their aircrew coordination training was a factor in this "save" (Alkov, 1991b). The NTSB came to basically the same conclusion regarding the value of CRM training in a similar incident involving the United Airlines DC-10 at Sioux City, Iowa (NTSB, 1990).

Preventing Error: One of the best methods of examining the effectiveness of training programs is to perform empirical tests to document whether crewmembers who receive such training make fewer errors. ADM programs have been extensively tested in this way. That was partly because, unlike CRM which was primarily employed for airline and military operations, ADM was initially applied to general aviation student training situations. The latter environment obviously involves relatively high error rates and low costs, thus permitting the use of controlled experiments.

Worldwide there have been six government sponsored, independent, evaluations of the ADM training programs. A detailed description of this research was recently completed (Diehl, 1990). These evaluations were performed to ascertain the effectiveness of such materials under differing conditions. The basic criteria were errors made during short, seemingly routine, cross-country "observation flights." On these flights, specially trained observers surreptitiously placed subjects in a series of specific Decision Making situations (e.g. rushing preflight inspections, or suggesting steep maneuvers at low altitudes). Observers then unobtrusively recorded the errors on these judgment items. In these rigorous 'double-blind" experiments, the observers were not informed which subjects had received ADM training, while subjects were unaware of the real purpose of the flights beforehand (e.g. subjects might be lead to believe they would be evaluating new map designs).

As expected, the effectiveness of the ADM materials varied widely depending primarily upon the comprehensiveness of the training (see Table 2). For the six studies, the improvement ranged from 8% in a voluntary, minimally structured, situation to 46% for a well structured, comprehensive, ground school environment with simulator training. Note that all six tests were statistically significant at or beyond the .05 level of confidence.

Table 2

ADM Training Experimental Evaluations

Preventing Accidents

These experimental evaluations provide strong statistical evidence such training can change behavior, and thereby reduce errors in low-time general aviation pilots. But the fundamental criteria for evaluating the effectiveness of cockpit management training programs is their ability to reduce the accident rates in the broader "operational world."

Airlines are obviously the most numerous users of these programs. But, their accidents occur very infrequently, (perhaps once every few years for a particular operator). Thus, it would be very difficult to prove that an individual airline, which had adopted CRM, in fact had experienced a significant decline in their aircrew 'error accident rate.

Fortunately, the FAA and Transport Canada have developed versions of the ADM training manuals for helicopter pilots (Adams and Thompson, 1987). This manual became widely used by a number of major rotorcraft organizations (Adams and Diehl, 1988). Because the accident rates for rotorcraft (and military aircraft) are normally orders of magnitude higher than those for airliners, one should be able to more easily detect improvement in their records. This was, in deed, the case.

Bell Helicopters Textron Inc. (BHTI): This major rotorcraft manufacturer provides extensive initial and recurrent training in both the US and abroad. They have utilized the ADM materials (Adams and Thompson, 1987) in their training programs since they were first published (Fox, 1991). BHTI particularly targeted their popular Bell Model 206 "Jetranger" because that craft generated about 46% of the total US civil helicopter flying hours.

They have also developed a "Cockpit Emergency Procedures Expert Trainer." Fox (1991) described this system as an artificial intelligence based software package which allows a pilot to use a personal computer as a Decision Making simulator.

The results of BTHI ADM training efforts are impressive, especially when their accident rates are examined. Fox (1991) compared the 1983-1986 period (before training was begun) with the 1987-1990 period. The world-wide human error accident rate (per 100,000 hours) declined by 36% for the Jetranger. Note for comparison purposes, the rate for mechanically caused accidents declined by only 8%.

Fox (1991) also stated the US Jetranger human error accident rates declined by an even more impressive 48%. Here the comparison periods were 1984-1986 and 1987-1988. He notes that many Jetranger pilots attending their training also fly other single engine helicopters, which may partly explain the more modest 25% improvement in those rates during this period.

Petroleum Helicopter Inc., (PHI): This organization is the largest commercial helicopter operator in the US with approximately 300 helicopters and fixed-wing aircraft. The company historically has had an excellent accident rate, well below the industry average. Their chief pilot (Mr. Vern Albert) reported the results of the using of ADM/CRM training in Rotor & Wing International (1989), p. 65: "From 1980 through 1986, we had an accident rate of about 2.3 accidents per 100,000 flight hours. In mid-1986, we started ADM training, and the rate in 1987 was 1.86 and then dropped to 1.05 in 1988. The only thing we changed in our training syllabus was adding ADM and cockpit resources management." This translated to a 54% reduction in their overall accident rate.

US Navy: In 1986, the Naval Safety Center reviewed the CRM programs which were underway at several airlines and the USAF Military Airlift Command (Alkov, 1988). They began formal CRM training at all Navy and Marine Corps helicopter training units in 1987. CRM was then initiated in their A-6/EA-6 'Intruder" fighter-bomber training units in 1988. As noted earlier they labeled these CRM materials Aircrew Coordination Training.

Alkov (1991a), p. 25, stated 'Aircrew error mishaps rates for helicopter and the A-6/EA-6 communities have declined dramatically since the introduction of Aircrew Coordination Training." Comparing the data for the fiscal year before the CRM training began with the most recent fiscal year data supports this statement. For these fighter-bombers, their 1990 aircrew error rate for all mishaps was of 1.43. Compared with their 1986 rate of 7.56, this . represents an 81% improvement. Similarly, for their helicopters, the 1990 rate of 5.05 versus-the 1986 rate of 7.01 represents a 28% improvement. Incidentally, their tentative figures for 1991 suggest that both helicopter and A-6 mishap rates continued to decline by another several percent (Alkov, 1991b).

USAF Airlift Command: In 1985, MAC became the first military organization to adopt CRM training. This program was initiated by the MAC Commander (General Thomas Ryan, Jr.) and labeled Aircrew Coordination Training. MAC has several thousand aircrew members who operate almost 1,000 transports and helicopters and fly approximately 700,000 hours annually. Thus it undoubtedly qualifies as one of the larger organizations to embrace CRM concepts. The CRM materials were developed by individual MAC units and several contractors including: United Training Systems, Flight Safety International, Hughes, and CAE-Link.

The MAC safety record for the five fiscal years before CRM (1981-1985) was compared with the five years (1986-1990) after they adopted this training. The total number of aircraft destroyed dropped from 21 to 10 (a 52% improvement). Similarly their Class A and B operations-related flight mishap rate dropped by 51% (from .679 to .333 per 100,000 hours). Note that these events currently include mishaps involving aircrew errors which the damages exceeded $200,000. These improvements far outpaced the rest of the USAF which saw the number of aircraft destroyed decrease by 18% while their aircrew error mishaps dropped by 21%.

This ten year period was relatively stable operationally, although longitudinal comparisons can be fraught with dangers. There obviously were a number of other training related developments occurring at this time (such as the simulator upgrades made throughout the USAF). However, the one major difference between MAC training and that of the other commands during this period was the use of CRM. It is also possible that the MAC Commander's bold decision to undertake this type training for all aircrew members produced other desirable side-effects.

The two earlier USAF situational awareness programs have also been found to work well. For instance Situational Emergency Training was judged more effective than conventional training it replaced (Thorpe, et.al.t 1976), while the manager of the Low Altitude Training program stated that of the approximately 400 graduates from their course, only one had been involved in a collision with the ground mishap (Thomae, 1991).

Table 3

ADM/CRM Operational Evaluations

Discussion

Table 3 summarizes the evidence reported to date on the effects of using ADM and CRM training by several large organizations. In all six instances, the use of the training was followed by major reductions in their accident rates. Collectively, these improvements were statistically significant at beyond the 0.02 level of confidence. These results agree with the six controlled experiments which focused on low-time general aviation pilots. Furthermore, this operational user data covers a large variety of civil and military flying from light helicopters, and medium fighter-bombers to heavy transports. It is, of course, impossible to conclude that CRM/ADM training was solely responsible for all the improvements observed. However, the inclusion of such training by management may act as a catalyst for other beneficial behaviors which, in turn, can reduce mishaps.

Povenmire et al. (1989), in their classic simulator study, noted that those crews who "innately" know how to use CRM-like techniques are more effective. Similarly Helmreich (1991) shows those who have had CRM training outperform their untrained counterparts. These innovative concepts have been endorsed by a host of prestigious organizations (e.g. Flight Safety Foundation, Aerospace medical Association, Air Line Pilots Association). Furthermore, the International Civil Aviation Organization has recommended their use for training all newly licensed pilots, (ICAO, 1989). Ambitious research and development efforts are also underway at various organizations such as the US Naval Training Systems Center (Hartel, Smith, and Prince, 1991) and the FAA and NASA (Kayten and Foushee, 1991). This work will undoubted further enhance these programs.

These concepts may also have applications beyond flightcrew training, witness the fact that many airlines now include such materials in flight attendant courses. The FAA has also published Decision Making materials for air ambulance program administrators (Rotor & Wing International, 1989), while Continental Airlines has applied the concepts to maintenance personnel, (Fotos, 1991). Not surprisingly, the US Army provides such training to their safety officers (Lofaro, 1990), while the US Pacific Air Forces now includes this training for new squadron commanders (Diehl, 1991a). Lastly, Kayten and Foushee (1991) correctly note the similarities between CRM and industrial total quality management programs.

New versions of this training have been labeled 'third generation CRM programs" (Johnson, Shroyer, and Grew, 1991). In fact, I believe the more sophisticated programs have several characteristics:

1) focusing on enhanced effectiveness as well as safety,

2) targeting broader audiences (e.g. instructors, flight attendants, etc.), 3) having built-in update mechanisms, 4) using computer-based training, and 5) employing humor and aesthetics.

In the Air Line Pilots Association 1991 annual workshop it was interesting to hear candid discussions of still unresolved issues associated with cockpit management. It was reminiscent of my earlier concerns about how to evaluate or certify the people enrolled in such programs, as well as how to rehabilitate those crewmembers who reject or "fail" this training, or who have exhibited poor judgment in an incident, accident or violation (Diehl, 1982). I still feel that such issues may be more recalcitrant than our basic concerns about cockpit management education and training.

Conclusions

The results of these six empirical and six operational evaluations provide strong-evidence that these training programs can help reduce aircrew errors and thereby prevent accidents. Furthermore, categorical distinctions between CRM and ADM training are becoming blurred in that most current versions of these programs have five . common elements, which provide tools to deal with attention, crew, stress, attitude and risk management issues. While additional research and development continues there is a growing realization that these programs ideally need to be introduced early in flight training, reinforced during-upgrade training, and reviewed during recurrent training

Bibliography

Adams, R.J. and Diehl, A.E. (1988). The Human Factor in Helicopter Safety. Vertiflight, July/August Vol 34 No 4.

Adams, R.J. and Thompson, J.L. (1987). Aeronautical Decision Making for Helicopter Pilots. DOD/FAA/PM-86/450, Washington, DC.

Alkov, R.A. (1988). The Naval Safety Center's Aircrew Coordination Training Program. Approach, September Vol 34 No 3.

Alkov, R.A. (1991a). A Progress Report on ACT. Approach, January Vol 36 No 7.

Alkov, R.A. (1991b). Personal Communications.

Beaumont, G. (1991). Personal Communications.

Berlin, J.I., Gruber, J.M., Holmes, C.W., Jensen, P.K., Lau,

J.R., Mills, i.W., and O'Kane, J.M. (1982). Pilot Judgment Training and Evaluation. DOT/FAA Report CT-82-56, Vol I, II & III, Washington DC.

Bruggink, G.M. .(1978). Assessing the Role of Human Performance in Aircraft Accidents. In Proceedings of the Ninth International Seminar of Air Safety Investigators, Winter, Vol 11 No 4.

Buch, G.D., and Diehl, A.E. (1982). Pilot Judgment Training .Validation Experiment. Unpublished Transport Canada Report, Ontario.

Buch, G. D., and Diehl, A.E. (1983). Pilot Judgment Training Manual Validation. Unpublished Transport Canada Report. Ontario.

Buch, G.D., and Diehl, A.E. (1984). An Investigation of the Effectiveness of Pilot Judgment Training. Human Factors. October, Vol 26 No 5.

Connolly, T.J. and Blackwell B.B. (1987). A Simulator-Based Approach to Training in Aeronautical Decision Making. In Proceedings of the Fourth International Symposium of Aviation Psychology, Ohio State Univ., Columbus, Ohio.

Diehl, A.E. (1991a). Cockpit Management Training Programs: Who, What, Why, When, and Where. Flying Safety, December Vol 47 No 12, In Press.

Diehl, A.E. (1982). Developing a Comprehensive Program to Improve Pilot Judgment. Unpublished FAA Report, Washington, DC.

Diehl, A.E., and Buch, G.D. (1986). Developing an International Program to Improve Pilot Decision Making. In Proceedings of the Flight Safety Foundation, 39th International Air Safety Seminar, Vancouver.

Diehl, A.E. and Lester, L.F. (1987). Private Pilot Judgment Training in Flight School Settings, DOT/FAA Report AM-87/6, Washington, DC.

Diehl, A.E. (1991b). The Effectiveness of Training Programs for Preventing Aircrew "Error". In Proceedings of the Sixth International Symposium of Aviation Psychology, Ohio State Univ., Columbus, OH.

Diehl, A.E. (1989). Understanding and Preventing Aircrew judgment and Decision Making "Errors". Unpublished Staff Study, USAF Inspection and Safety Center, Norton AFB, CA.

Fotos, C.P. (1991). Continental Applies CRM Concepts To Technical, Maintenance Corps. Aviation Week.& Space Te 10 August 26, Vol 135 No 8.

Foushee, H.C. (1987). In Why Planes Crash. Nova T.V. Program No 1403, WGBH Educational Foundation. Boston, MA.

Foushee, H.C. and Helmreich, R.L. (1988). Group Interaction and Flight Crew Performance. In Human Factors in Aviation. Academic Press, San Diego CA.

Fox, R.G. (1991). Identifying the Hazard is Only the Beginning. Hazard Prevention. Vol 27 No 3.

Freeman, C. and Simmon, D.A. (1990). Crew Resource Management Model. In Proceedings of the 43rd International Air Safety Seminar, Flight Safety Foundation, Rome.

Hartel, C., Smith, K. and Prince, C. (1991). Defining Aircrew Coordination: Searching Mishaps for Meaning. In Proceedings of the Sixth International Symposium of Aviation Psychology,'Ohio State Univ., Columbus, Ohio.

Helmreich, R.L. (1991). Does CRM Training Work? Air Line Pilot. May, Vol 60 No 5. ICAO (1989). Human Factors Digest No 1. Fundamental Human Factors Concepts, International Civil Aviation Organization, Circular 216-AN/131, Montreal.

Jensen, R.S. (1987). Feasibility Study for the Assessment of Cockpit Resource management Training. FAA Contracted Study No DTFA01-85-P-08000/Q-09028, Ohio State Univ. Columbus, OH.

Jensen, R.S. and Benel, R.A. (1977). Judgment Evaluation and Instruction in Civil Pilot Training. DOT/FAA Report RD-78-24, Washington DC.

Johnson, N.A., Shroyer, D.H. and Grew, J.B. (1991). A New Generation of Crew Resource Management Training. In Proceedings of the Sixth International Symposium of Aviation Psychology, Ohio State Univ., Columbus, OH.

Kayten, P.J., and Foushee, H.C. (1991). The National Plan.for Aviation Human Factors. Aviation, Space, and Environmental Medicine. May, Vol 62 No 5.

Lauber, J.K. (1984). Resource Management in the Cockpit. Air Line Pilot. September, Vol 54 No 9.

Lee, R.B. 1991). Pilots Have a One-Track Mind. Flight Deck, Autumn, Vol 1.

Lofaro, R.J. (1990). Personal Communications.

Miller, C.O. (1979). Human Factors in Accident Investigation. Legal Eagle News. Nov, Vol 22.

Nance, J. (1986). Blind Trust. Morrow Publishers, NY, NY.

NTSB, (1979). Aircraft Accident Report, United Airlines, Inc., DC-8, Report No AAR-79-7, Washington DC.

NTSB, (1983). Aircraft Accident Report, United Airlines Inc., DC-8, Report No AAR-83-07, Washington, DC.

NTSB, (1990). Aircraft Accident Report, United Airlines Inc.,

DC-10, Report No AAR-90-6, Washington, DC.

NTSB, (1991). Aircraft Accident Summary Report, Mid-Air Collision Involving Lycoming Air Service PA-60 and Sun Oil Bell 412, Report No AAR-91-01, Washington, D.C.

Povenmire, H.K., Rockway, M.R., Bunecke, J.L. and Patton, M.W. (1989). Cockpit Resource Management Skills Enhance Combat Mission Performance in a B-52 Simulator. In the Proceedings of the Fifth International Symposium of Aviation Psychology, Ohio State Univ., Columbus, OH.

Reason, J. (1990). Human Error. Cambridge Univ. Press, Cambridge.

Rotor & Wing International, (1989). Vol 23 No 11.

Ruffell-Smith, H.P. (1979). A Simulator Study of the Interaction of Pilot Workload with Errors, Vigilance, and Decisions. NASA TM-78482, Moffett Field, CA.

Telfer, R.A. and Ashman, A.F. (1986). Pilot Judgment Training an Australian Validation Study. Unpublished Research Report, Univ. of Newcastle, New South Wales.

Thomae, C.J. (1991). Personal Communications.

Thrope, J.A., Martin, E.L., Edwards, B.J. and Eddowes, E.E. (1976). Situational Emergency Training. USAF HRL Report TR-76-47, Williams AFB, AZ.

Footnote

Dr. Diehl has been the Technical Advisor for Human Performance at the USAF Safety Agency since 1987. He has over twenty years professional experience with aircraft manufactures, the USN, NTSB and the FAA. He holds academic degrees in psychology, management and engineering, as well as an Airline Transport Pilot and Flight Instructors licenses. The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of the Department of Defense or the US Government.

[ CRM-DEVEL Home ] [ What's New ] [ Resources ] [ CRM-DEVEL Mailing List ]
[ Neil Krey's Professional Aviation Instruction Forum ] [ Neil Krey's FLIGHT DECK ] [ Aviation Instruction Bookstore ] [ Aviation Instruction Career Center ]

Copyright © 1996-2005 by Neil C. Krey unless otherwise indicated.
Non-commercial reproduction rights granted if the following notice is included:
"Source: Neil Krey's CRM Developers Forum, http://www.crm-devel.org"