Table of Contents

• Secretary of Transportation Federico Pena, Airline Safety Conference, January 9, 1995
• Putting Professionalism in the Cockpit, by: Dr. John Lauber, NTSB
• Commission of Inquiry into the Air Ontario Crash at Dryden, Ontario, by: Dr. Robert L. Helmreich
  NASA/University of Texas
• NTSB Releases it's study on: "Commuter Airline Safety"
• Organizational Culture and Conflict in Airline Operations, by: Maureen A. Pettitt, Western Michigan University

It is very heartening to see the tremendous turnout for this historic Airline Safety Conference. I thank everyone from airline management, labor, the NTSB, and the many fine FAA employees who have worked so hard to organize this effort.

From the first day that President Clinton took office and asked me to serve our country as Secretary of Transportation, we have recognized the vital importance of the airline and aerospace industries to our country and to the world.

We understood that airline and aerospace companies directly employ upwards of 1.5 million Americans in high wage, high skill jobs and have an economic impact measured in the hundreds of billions of dollars a year.

That's why we led the charge to create the National Commission to Ensure a Strong Competitive Airline Industry. And that's why we are already acting on 50 of that Commission's 61 specific recommendations.

Administrator David Hinson has committed to review all regulations that unnecessarily and negatively impact the airline industry.

I have taken steps to support competition and to constructively ameliorate labor management tensions.

We have fought hard to open world markets for U.S. carriers as we have announced through our recent breakthrough with Canada and our recent new international aviation policy-the first since President Jimmy Carter.

And as a passenger who has flown over 300 flights on your planes in the past two years - flying coach, by the way - I have acted to address the concerns of airline consumers - the nearly half a billion passengers who last year flew half a trillion miles on commercial flights in our country.

Like everyone else involved in American aviation from airline CEOs to flight attendants and mechanics, FAA inspectors and our colleagues in the NTSB - I have confidently entrusted my life and my family's lives to air travel.

Like you, I know that this industry, in partnership with government, has built the safest aviation system in the world. In fact, we have made such great strides together that air travel is no longer an elite privilege as it was little more than a generation ago - but a true mass market that the vast majority of Americans now rely on.

All through the explosive growth in air travel in the 17 years since deregulation, we have significantly improved the safety record of this industry - to the point where air travel is 30 times safer than travel on our interstate highways - themselves the safest highways in the world.

Despite these successes today, we've come together to face a new challenge - a challenge that I am confident we will meet.

It is absolutely imperative that we be direct and honest about this challenge and acknowledge some stark facts.

Fact one is that last year we experienced seven fatal crashes in commercial aviation - tragedies in which 264 people lost their lives.

I can never forget visiting those crash sites, or the experience of consoling families who lost a father, mother, brother, sister, relatives, or friends. Nor will I forget the shock of airline employees who struggled with the loss of a co-worker or a friend, or the despair of an air traffic controller who was tracking a flight.

And I will never forget the emotional trauma that communities, volunteers, and rescue teams have endured in responding to these tragedies.

A second stark fact that we must acknowledge is that airplane crashes consumed the attention of the national and international media and the general public in a unique way.

As a nation, we have become immune - almost numb - to the hourly crashes our nation's highways which take 40,000 lives a year. Sadly, people have become too complacent about these tragedies... while we all know that any fatal airline crash will always be a national story.

I only mean to emphasize to each of you in this room that this is the real environment in which we operate - every second of every day.

The third cold fact that we must face is that some difficult questions are being leveled at everyone in this room.

I tell you we must address these questions forth rightly and directly and, where appropriate, use these occasions to educate and to inform the American people.

Listen to some of the questions that the Administration and I get asked all the time by people around the country:

"Mr. Secretary, is it true that financial losses in the airline industry have compromised the commitment to safety?"

"Has deregulation caused such intense competition and turmoil in the industry that safety has suffered?"

"Are commuter airlines unsafe?"

"Are their pilots flying fatigued'?"

"Is the FAA too slow to respond because it has a 'tombstone mentality'?"

"And, by the way, Mr. Secretary, why do you still say that our airline system is the safest in the world in light of the accidents in the past twelve months?"

Whenever people or institutions are faced with these sorts of questions, there's a powerful temptation to just shrug them off - by suggesting that the inquiries come from naive, uninformed people who are "not experts. "

Another reaction is to develop a siege mentality and hide in the bunker while these sorts of charges keep coming.

I believe that a better response - the right response is what we are doing here today: that is, to have a ruthlessly honest self-evaluation, a gut-check if you will, and answer the following questions:

"Is there something more we should be doing?"

"Is there anything that we have not anticipated?"

"Are we applying our expertise and training and technology fully ?"

"Are training, supervision, and inspection pro-rams the very best that they can be?"

"Are we absolutely sure that we are flying by the book?"

In short, its UP to each of us in this room to take the lead - and to demonstrate to ourselves and to the American people that we will do everything possible not only to maintain - but to improve - our already high safety standards and performance.

And we will not settle for anything less than Zero Accidents!

From the perspective of the DOT and the FAA, we are committed to doing our job in this safety partnership even better - anticipating safety issues before they emerge - whether they relate to weather, human factors, aircraft design or certification.

This is a commitment that Administrator David Hinson and I share.

I commend the Administrator for the progress he has led in this direction - notably the recent FAA reorganization that creates a dedicated safety office reporting directly to him. And, to the FAA safety specialists who are here today, I say to you that we must do more.

We need to develop an even greater sense of urgency in deploying new technologies that improve safety from Doppler radar to surface radar ASDE equipment on Runways to Global Positioning Satellite systems.

We must move forcefully to reform the Air Traffic Control system - to give it the entrepreneurial flexibility to recruit and assign personnel, to purchase and install new technology on a timely basis, and to secure the long-term financing it needs to plan ahead for system upgrades.

If we are to efficiently and safely manage the 60 percent growth in air traffic - that means 300 million more passengers a year - that the FAA anticipates over the next decade, we must make this reform a priority.

But the responsibility to improve safety is a shared one. So I am also calling on industry to continue acting with real urgency on all new safety technologies and procedures.

I am asking industry to take your own high-profile safety initiatives - without regulatory prodding - to exceed your already high standards.

And let us, government and industry alike, communicate the seriousness of our efforts to the American people.

There are certainly costs to all these investments. But as we have learned from other parts of transportation industry - for example, the automobile industry safety pays.

Today's conference is part of a three-step safety initiative that I announced in the wake of last month's crash in North Carolina.

This initiative includes the special safety audit that we are performing to supplement the 400,000 safety inspections that the FAA already conducts each year. It also includes our comprehensive efforts to raise commuter airline safety standards to the level of major carriers.

I believe that these actions send a clear signal to the American people that each of us - in government and industry - do take their - and our - concerns for airline safety very seriously.

I commit to you that we will work in a constructive and cooperative partnership in every way possible to get the job done.

But whenever necessary, I will not hesitate to exercise our full legal - and moral - authority to take strong actions to ensure that all of our safety rules are fully met.

Over the past two years, I have talked to many airlines executives, to pilots, flight attendants, mechanics and maintenance people - and to many, many of your customers.

I am pleased to say that everyone involved in American aviation shares a genuine commitment to safety - and your attendance here reflects that.

Together, we have made tremendous strides over the years. Now, the American people expect us to do even better. And I am confident that we will.

I'll be here all day today and back again tomorrow.

I look forward to this shared undertaking and I thank you in advance for the contributions you are about to make during this historic conference. Thank you.

by: Dr. John Lauber, NTSB

How do you differentiate between a professional atmosphere in the cockpit where people feel free to speak up and there is an easy relaxed atmosphere that is conducive to safe flight operations, and an atmosphere that reflects laxity and non-attention to professional responsibilities and whatever else you want to call it? There is a fine line there.

Line pilots themselves bear a great deal of responsibility for action that needs to be taken in this regard... I strongly encourage and support further activities on the part of the pilots' associations to push for professional standards and to deal with these kinds of problems from within the ranks of line pilots. At the same time, it is quite clear that management has a major responsibility.

There is no accident that has ever taken place which involves individual or crew error which does not also illustrate errors in management and in the organizational side of the effort. No human performance takes place in a vacuum. It takes place in an environment that is created and engendered by management, and so any time we have a failure, it is never a failure that stops at the cockpit door. It extends all the way up to the top corporate offices.

And somehow, we are all going to have to work together to come to grips with this issue of professionalism, of professional standards, of cockpit discipline, of cockpit procedures, and whatever else we want to call it. In that regard, there is a related area, that I also want to mention, and one that I have not heard mentioned all that directly or all that frequently. Yet, it is a simple part of this issue: the whole question of leadership in the cockpit.

This is something that is fundamental to cockpit operations. It is one of the most fundamental roles that the pilot-in-command and the captain plays, and that is serving as the leader of the team, as the leader of the crew. It is an area that, in many cases, has been lost sight of.

But there is another area that I want to briefly mention with respect to the issue of leadership and it is of some concern. A lot of the effort of cockpit resource management g, in my mind, would be devoted to developing leadership skills, recognizing the importance of the leadership role of the captain, and in trying to develop those skills to impart effective leadership behavior to the captain and the other crewmembers.

There is the other side of the issue, of course, which is good followership. I am concerned at some efforts that I have seen involving cockpit resource management in which we lose sight of this leadership issue. You hear all these warm and fuzzy comments about interpersonal relationships and personality. But I sense a tendency, in some circles at least, to view cockpit resource management as being a matter of personalities and getting people to like each other.

I have called it "hot tub harmony," that comes from my days in California when we sat around in hot tubs and held hands and flew safely together. It does not work that way. I do not care whether people like each other or not. They can fly together very safely even without liking each other. It is possible to develop effective leadership skills on the part of individuals. Leaders can be effective in terms of setting the tone, of inviting assertiveness on the part of the subordinate crewmembers and yet retaining command and the responsibility that goes along with command in the cockpit.

We have seen a lot of incidents and accidents that have developed because of a lack of effective leadership on the part of the captain, and conversely, the lack of effective followership on the part of the other flight crewmembers. Therefore, it is an issue that we have to come to grips with.

There are a couple of interesting things in the (Les) Lautman and (Peter) Gallimore papers that I want to throw out for your consideration. (Lautman and Gallimore presented papers on "Control of Crew-Caused Accidents" at an earlier Flight Safety Foundation International Air Safety Seminar in Tokyo. They took a number of statistical data to examine why air accidents happen and what flight departments can do to prevent accidents.)

They had a database that consisted of some 93 major accidents during the period of 1977 to 1984, and fully one-third of those accidents were attributed to pilot deviation from basic operating procedures. The other factors along similar lines were right up there in the same ballpark. Clearly, the major consideration is pilot adherence to sound operating procedures, and again, it ties back to this whole issue of cockpit discipline and professionalism.

One of the interesting things that came out of their studies, and I should say that one of the questions that Lautman and Gallimore asked in their papers, is the whole issue of what is it that distinguishes an airline that has a good operating record, a good safety record, from the rest of the crowd? What makes them different? What do they do differently that might be related to their enviable safety record?

And I want to share just two of those with you. The number one factor in their list is management emphasis on safety.

It is quite clear that in those airlines which have the exemplary safety records, the commitment to safety begins at the top, in the corporate offices, and extends throughout the organization. Coupled with that management emphasis on safety most frequently was a proactive, visible, single, separate entity of flight safety organization. This pops out of the database. It quite clearly underscores the importance of a viable separate safety organization within a corporation. I too, have heard the arguments that have been posed in our organization (that) we are all responsible for safety. That may be true, but that also blithely ignores the organizational realities and the fact that you need someone to set the goals, the corporate safety goals, and you need someone to implement those goals into hard action.

So I would add, on the basis of Lautnian's and Gallimore's findings, and much other information that is available to us, my endorsement of the importance of an active flight safety department within any airline.

The second major factor that Lautmaii and Gallimore discuss in their series of papers has to do with this whole question of standardization and discipline. There is a lengthy discussion in there about the importance of developing standard cockpit operating procedures, of developing a standard procedural language for use in cockpits, about the importance of thorough training and checking in-flight standards program, about the importance of an independent line checking organization separate from the flight training effort.

I think along with the safety departments and safety programs, there is always, of course, the checking and flight standards and training elements that we simply cannot spend enough time fine tuning.

When we combine all of these tools together, we will be able to make some inroads on what continues to be the most significant cause of incidents and accidents in our industry. We will prove once again that in spite of all of the developments in technology, airplanes are still airplanes, and they are still going to have to be crewed by pilots who are professional, who are well-trained, and who have the support of the organization. That is necessary in order to conduct safe flight operations.

by: Dr. Robert L. Helmreich
NASA/University of Texas

At the request of the Commission of Inquiry Into the Air Ontario Crash at Dryden, Ontario evidence assembled in the course of investigation into the causes of the crash was examined in terms of human factors and organizational issues. Material reviewed included reports of the Operations Group and the Human Performance Group, interviews with relevant personnel, and sworn testimony presented before the Commission. When viewed a research perspective, the body of facts suggest an operational environment that allowed an experienced crew to reach a flawed decision regarding the safety of take-off during snowfall with accumulating contamination of the aircraft's wings.

The absence of direct evidence from voice or flight recorders initially seems to be a serious hindrance to the investigative effort. In fact, the lack of this type of evidence has resulted in a more extensive exploration of broader issues, including regulatory and organizational factors that might otherwise have been conducted. Because of the depths of the investigation, the lessons to be gained from this in-depth investigation may prove to be of value for the governance of flight operations and the training of crews.

It may be useful to outline the background for the author's opinions. They grow out of more than twenty years experience conducting research into the multiple determinants of human behavior and performance under the sponsorship of agencies such as the National Science Foundation, the Office of Naval Research, the National Aeronautics and Space Administration, and the Federal Aviation Administration. Current investigations are under the auspices of the NASA/University of Texas Aerospace Crew Research Project, directed by the author. Included in the project are investigations of personality factors relative to pilot and Astronaut selection group dynamics, aircraft characteristics such as automation, and organizational issues such as the development and influence of subcultures (Helmreich & Wilhelm, 1990; Helmreich, in press).

Another central element of the research is evaluation of the effectiveness of training in Crew Resource Management (CRM.- Helmreich, 199 1). CRM training is aimed at improving crew coordination, decision making, situational awareness, and interpersonal communications. It stresses the importance of utilizing all available resources inside and outside the cockpit and the development of an effective team including cabin crewmembers in the process. The concept of CRM is becoming widely accepted and is an integral part of training in many organizations. Only recently, however, has empirical research demonstrated that such training can affect flightcrew behavior (Helmreich, Chidester, Foushee, Gregorich, & Wilhelm, 1990; Helmreich, Wilhelm, Gregorich, & Chidester, 1990).

Underlying the research is the fact that the behavior of flightcrews in any given situation is determined by a number of simultaneously operating factors. These include: 1) the regulatory environment - operational standards and supervision; 2) the organizational environment - the culture and behavioral norms of the organization including morale, policies and standards, organizational stability and change, and available resources; 3) the physical environment meteorological and operating conditions and the aircraft, including its conditions and capabilities; 4) the crew environment - interpersonal coordination and communications including cockpit, cabin, and ground personnel, and individual characteristics of crewmembers - training, experience, motivation, personality, attitudes, fatigue, and stress both from the immediate operational situation and significant personal life events. Figure I shows graphically the environments surrounding flight operations. Events and circumstances exemplifying these categories will be discussed as they relate to the Dryden crash and possible reasons for the actions of the crew of Air Ontario Flight 363.

The results of this analysis suggest that the concatenation of multiple factors from each category allowed the crew to decide to take off with contaminated wings. According to this view, no single factor taken in isolation would have triggered the crew's behavior prior to and during take-off, but in combination they provided an environment in which a serious procedural error could occur. This array of contributory influences without a single, proximal cause warrants classification of the accident as a system failure. The analysis will attempt to define these influences and their inter-relationships. Observations and suggested counter-measures will also be provided.

Figure 1. Flightcrew Environments: Factors Influencing Behavior

History of the trip. The crew reported in at Winnipeg at approximately 0630 CST Monday, March 6, for a five day trip in Fokker F-28, registration C-FONF, involving six legs per day ending at 1530 CST. The trip schedule and crew pairings are shown in Figure 2. Captain George Morwood had flown with the two flight attendants before, but none had flown with First Officer Keith Mills. After flying the Monday, March 6 sequence, Captain Morwood was displaced Tuesday by Captain Robert Nyman and Wednesday by Captain Alfred Reichenbacher. He resumed the trip for Thursday, March 9 and Friday, March 10.

On March 10, the crew checked in at Winnipeg at approximately 0640 and discovered the Auxiliary Power Unit (APU) was inoperative. The aircraft departed for Dryden at 0749, approximately 10 minutes late after waiting for de-icing. It was further delayed at Dryden by poor weather at Thunder Bay. At Thunder Bay the flight was refueled on the basis of a passenger load of 55. However, an additional 10 passengers were added, placing the aircraft over the computed maximum allowable gross weight for take off. After some debate over course of action, the aircraft was defueled and the additional passengers retained. The flight departed Thunder Bay 64 minutes late and arrived at Dryden 1130 CST. The aircraft was refueled at Dryden with an engine running because there were no ground start facilities there. Contrary to Air Ontario policy stated in the cabin manual, passengers remained on board during refueling.

Figure 2. Trip Routing March 6-10, 1989
Air Ontario Line for Morwood/Mills

During the stop at Dryden snow was falling and accumulating on the wings. First Officer Mills commented on the radio to Kenora at 1200, "...quite puffy snow, looks like its going to be a heavy one." Shortly after beginning to taxi, a passenger asked Flight Attendant Katherine Say when the plane was going to be de-iced. The flight attendants did not inform the flightcrew of these expressed concerns about the need to de-ice.

The flight was delayed for approximately four minutes while a light aircraft landed. At 1207 CST the flight was cleared to Winnipeg and at 1209 First Officer Mills transmitted that the flight was about to take off. The aircraft lifted off but never left ground effect and crashed into trees beginning 126 meters from the end of the runway. The aircraft was destroyed by impact and fire. Both pilots, one flight attendant, and twenty-one passengers were killed. Forty-four passengers and one crew member survived with injuries. The chronology for March 10 is shown in Figure 3.

Figure 3. Air Ontario Flights 362/363
March 10, 1989

1. The Regulatory Environment

The crew of Air Ontario 363 was governed by the regulations and practices of Transport Canada. Several aspects of the current regulations provided an indirect, deleterious influence on the crew's operational environment. These allowed the development of a situation which failed to provide safeguards in this case against flawed decisions concerning landing and take-off in Dryden under adverse weather conditions. The following issues are cited as relevant to the accident.

I(a). The failure to provide clear guidance for organizations and crews regarding the need for de-icing. The regulatory requirement in effect at the time of the accident prohibited aircraft from commencing a flight "...when the amount, of frost, snow, or ice adhering to the wings, control surfaces, or propeller of the aeroplane may adversely affect the safety of the flight". As noted in the Commission of Inquiry into the Air Ontario Crash at Dryden Ontario Interim Report (1989), "...there are no existing Transport Canada approved guidelines in which dispatchers or flight and ground crews may use to assist them in making a reasoned judgment as to what amount of contamination to an aircraft's lifting surfaces would adversely affect the safety of flight". In the absence of guidelines, idiosyncratic views of the degradation caused by differing amounts of contamination could prevail. There were also no formal requirements for training in the effects of icing contamination and associated phenomena such as "cold soaking", and the differential susceptibility of different aircraft types to icing effects.

I(b). A lack of rigor in regulating and monitoring the operations of Air Ontario, Inc., following its merger and during the initiation of jet service in the F28. Transport Canada allowed the F28 operation to continue passenger service for a number of months without an approved Minimum Equipment List and an accepted Aircraft Operating Manual specifying standard operating procedures. Closer monitoring of the initiation of this service would have revealed other Significant operational problems including inconsistent content in manuals (i.e., different manuals in the cockpit and conflicts between cabin and cockpit manuals) and problems in weight and balance computations. It would have been especially important at this time to conduct extensive line observations of crew performance in the F28. Testimony of Transport Canada witnesses identifies a lack of resources for the enforcement of safety regulations and monitoring of flight operations.

I(c). An audit of Air Ontario operations that was delayed and Incomplete in Scope. Evidence from several airline mergers that have been observed in the U.S. suggests that they create conditions' which warrant increased regulatory surveillance. There are always disruptions in operational effectiveness surrounding the joining of disparate operations that call for increased efforts directed toward monitoring operations and ensuring compliance with appropriate safety standards. Strikes have also been observed to create major operational problems, even after their settlement and to interfere with effective crew-management communications. A national audit of Air Ontario was scheduled for February, 1988. While the airworthiness, passenger safety and dangerous goods portion of the audit were completed as scheduled, the flight operations portion was postponed until July, 1988 and again until November, 1988, when it was completed. The combination of a merger, a strike, and the introduction of a new aircraft type, would seem to have mandated an extensive audit of the operation. It is noteworthy that the audit that was conducted failed to examine the most significant operational change in the organization, the initiation of jet service in the F-28. Testimony by the leader of the audit indicates that he was inexperienced in audit procedures, was directing his first audit, and had limited staff. The statement that examination of crew training records forms the heart of an audit certainly reflects an honest opinion. However, from the author's research experience, an alternative view can be proposed that the observable behavior of crews in line operations is the key to understanding the level of safety and effectiveness in flight operations.

I(d). The failure to require effective training and licensing requirements for flight dispatchers and to establish regulations governing dispatch and flight following. Transport Canada had no formal requirements for the training and licensing of dispatchers and allowed a carrier such as Air Ontario to operate with a pilot self-dispatch system. While the arrangement at Air Ontario was in compliance with regulations, it practiced much less rigorous control of operations than its parent organization, Air Canada.

I(e). The lack of clear criteria for the criteria for the qualifications and training of airline management, Check Airmen, and Air Carrier Inspectors. In times of rapid organizational change, frequent shifts in operational condition practices are common as is substantial turnover in managerial positions. While organizations normally strive to maintain the highest possible level of experience and competence, in the absence of formal rules, compromises are frequent. It is suggested that more clearly defined guidelines could help organizations recognize situations where they need outside expertise to increase the safety and effectiveness of operations. In evaluating personnel, both extent and quality of experience can serve as indicators of whether there are sufficient qualifications to direct and evaluate operations effectively. In the case of a new operation such as the initiation of F-28 service, such determinations may be difficult for those directly involved to make.

One persistent problem in the standardization of air carrier operations is the fact that regulatory inspectors and Check Airmen monitoring line operations are normally limited to working within a single aircraft type. The implication of this is that procedural variances that develop between the aircraft fleets of an organization fail to be detected by individuals who are restricted to dealing with a single component of the organization. Several airlines are adopting the policy of having evaluators monitored crew coordination and effectiveness across aircraft types to gain insight into type differences and developing subcultures.

II. The Organizational Environment

A number of factors surrounding the nature and operation of Air Ontario created an environment conducive to operational error. At the highest level, Air Canada, despite owning controlling interest, failed to require Air Ontario to operate to Air Canada standards and failed to provide resources to achieve these standards. Similarly, a number of decisions and practices at Air Ontario served to allow an operation with significant safety-related deficiencies to develop and continue. The focus of this discussion is not on faulting organizations for failing to go beyond regulatory requirements. Rather, it is to discuss the operational impact of the organizational setting and practices that were present at this time. The factors to be discussed have been observed to impact operations in other air carriers facing similar constraints. It should be noted, however, that organizations undergoing such transformations might not be in a position to recognize their safety implications from within.

II(a). Lack of operational support from Air Canada. During the period of initiation of F-28 service, Air Canada owned a seventy-five percent interest in Air Ontario which operated under shared ("AC") flight designator. Air Canada has long experience in jet transport operations and stringent requirements for dispatch and flight following. The resources of this organization would have been highly valuable in smoothing the transition to the merged carrier and initiating jet service in the F-28. According to testimony, there were financial reasons (maintaining independent operations and pay scales) for maintaining a separation between the two carriers and there was no regulatory requirement for sharing resources and standards.

II(b). The disruptive impact of mergers and strikes. Mergers among air carriers have become increasingly frequent in recent years. In the course of our investigations, research into crew attitudes and behavior has been conducted in several airlines which were the results of one or more mergers. As part of the research, crewmember, attitudes toward management of the flight deck are assessed using a survey instrument, the Cockpit Management Attitudes Questionnaire (CAMQ) (Helmreich, 1984; Gregorich, Helmreich, & Wilhelm, 1990). Attitudes regarding flightdeck management have been validated as predictors of crew performance and were derived from research implicating them as relevant in many accidents and incidents (Helmreich, Foushee, Benson & Russini, 1986). The data showed significant differences in attitudes as a function of previous organizational membership in each organization we have studied - in one case nearly a decade after a merger. The results clearly indicate the existence of enduring subcultures within organizations. It is our premise that when cultural factors support the maintenance of differing attitudes about the appropriate conduct of flight operations, the effectiveness of flightcrew performance is likely to be compromised. Degani and Wiener (I 990), in their study of normal checklist usage in air carrier operations, suggest that the stresses of mergers can result in crews retaliating against management by disregarding mandated checklist procedures. The process of combining seniority lists from merging organizations also frequently results in poor relations among crewmembers from the different airlines. We have found that pejorative nicknames are often employed to label crewmembers from the opposite side of mergers.

Similarly, our data indicate that labor-management strife can have a deleterious effect on crewmembers' morale and attitudes toward their organizations. While there is no evidence to suggest that a crash has resulted directly from the impact of a strike, there is no doubt that the negative climate fostered by poor pilot-management relations is not conducive to effective team performance. In several airlines, even some years after a strike, relations among pilots and between pilots and management remain poor.

Evidence form Air Ontario personnel supports the existence of differing sub-cultures in Austin Airways and Air Ontario with occasional categorization of former Austin Airways personnel as "Bush Pilots" who could be assumed to have informal, operational practices at variance with those of former Air Ontario flightcrews. The F-28 program was disproportionately managed by former Austin Airways personnel who could have influenced the operation in the direction of Austin Airways norms. The dominance of Air Ontario flight operations management by Austin Airways personnel also created ill-will among some former Air Ontario pilots. Morale problems and poor relations among crewmembers can interfere with effective teamwork and crew coordination.

One finding from our research into Crew Resource Management training is that it can serve to reduce differences in attitudes about flightdeck management between subcultures and between crew positions. Air Ontario management had looked into such training. Captain Robert Nyman, Director of Flight Operations, testified that the CRM courses available did not appear to fit the Air Ontario operation. Both the Chief Pilot and Chief Training Pilot attended a CRM course presented in Toronto by a major airline and reported it to be both of limited value and expensive.

II(c). High personnel turnover following the merger. In the period between the merger of the two carriers and the accident, there were substantial changes in personnel. Part of the operation was sold and the size of the combined organization was reduced from eight hundred to approximately six hundred. There was also turnover in two critical areas of management, Vice President of Flight Operations and Director of Flight Operations. Similarly, the position of Safety Officer was filled, became vacant due to a resignation, and subsequently re-filled. The lack of continuity in management could have impeded needed supervision of operational issues such as the introduction of a new aircraft type and standardization of operations following the merger. Programs such a CRM cannot alleviate operational problems associated with a lack of management stability and consistent direction.

II(d). Lack of organizational experience in jet operations. Air Ontario as an organization did not have experience in jet transport operations. At the time of the introduction of the F-28, efforts were made to acquire outside expertise in management and representations to this effect were made to Transport Canada. Ultimately, Captain Claude Castonguay, who had substantial jet transport operational experience (including in the F-28) was hired, but resigned after one month. Six months later he was called back to perform two line indoctrination's. In his letter of resignation, Captain Castonguay stated, "So much as I would like to keep working to establish your FK28 program, I have concluded that I cannot function in my duties as Check Pilot when I do not get the support I need." No one was subsequently hired from outside the organization to fill this role, leaving Air Ontario to manage the process with internal resources.

II(e). Deficiencies in System Operation Control (SOC) practices. Air Ontario operated with a dispatching system that consisted partly of full flight following and partly of pilot self dispatch. Although this system was permitted by current Transport Canada regulations, it failed to provide crews with the same level of support and resources given crews in the parent organization, Air Canada.

In the absence of regulations mandating formal training and licensing for dispatchers, Air Ontario primarily employed on the job training for dispatch personnel. For the introduction of the F-28, brief training in the operation of this type of aircraft was provided only for duty managers. In contrast, Air Canada provides its dispatchers with more formal training and operational guidelines - including rules that would forbid dispatching an aircraft with an inoperative APU into a station such as Dryden with no ground start capabilities. That the Air Ontario system was deficient is indicated by observed errors in flight releases such as fuel load calculations using wrong parameters. Indeed, the flight release for C-FONF contained errors on the day of the accident.

II(l). Lack of standard operating procedures and manuals for the F-28. Service was initiated without a specific Air Ontario operating manual for the F-28. There was also no approved Minimum Equipment List for some months after passenger service began. There were inconsistencies between cockpit and cabin manuals provided to crews. For example, the cabin manual required passenger disembarkation for refueling with an engine running while there was no parallel rule in the cockpit manual. Crews thus lacked formal organizational guidelines either from resources available on the flightdeck or from SOC.

II(g). Inconsistencies/deficiencies in training F-28 crewmembers. Initial training of F-28 crewmembers, including both ground school and simulator training, was contracted with Piedmont Airlines. Piedmont itself was involved in a merger with USAir which decided to achieve standardization of the merged operation by shifting all former Piedmont personnel to USAir procedures and manuals. There was several implications of this organizational environment for Air Ontario crews. The first was that some received training from the Piedmont F-28 manual while those training later, worked with the USAir manual. Since Air Ontario had not developed its own manuals, some individuals returned with the Piedmont Manual and others with that of USAir. While Air Ontario stated that the Piedmont Manual was its standard, this was not clearly communicated to crews and no efforts were made to provide all crews with the same manual. Air Ontario also failed to receive updates to the manuals it was using. Although the Fokker Aircraft Manual was carried in the aircraft, there was a lack of training involving this manual and there were discrepancies between the Fokker and Piedmont manuals, for example in computing corrections for runway contamination. A second result of the Piedmont merger was a scarcity of simulator time for completing the training of Air Ontario crews. Because of this, a number of pilots were trained in the aircraft by newly qualified Air Ontario pilots rather than in the Piedmont simulator. Even with highly experienced instructors, there is an industry consensus that simulator training provides broader and more effective training.

Crewmembers surveyed by the Safety Officer following the accident generally reported their Line Indoctrination at Air Ontario to be "fair" in quality. One deficiency noted was a failure to define clearly the duties of the pilot flying and the pilot not flying.

II(h). Leadership of the F-28 program. Captain Joseph Deluce was selected as Project Manager and Chief Pilot for the F-28 and Convair 580. Captain Deluce had numerous responsibilities including line flying during the strike which preceded aircraft delivery and conducting training and line indoctrination in the F-28 for new crewmembers. He also carried Chief Pilot responsibilities for both fleets. Captain Deluce had limited operational experience in both the F-28 and the Convair 580. Airlines typically choose individuals with substantial experience in an aircraft type to be Chief Pilot.

One incident that may have had a significant impact on crewmember attitudes was the removal of an F-28 crew from a line trip to meet with the Chief Pilot for allegedly writing up too many maintenance discrepancies on the aircraft. The perception of other crewmembers of such an event would likely be of a lack of leader support for optimal operating conditions and a strong pressure to operate at all costs.

II(i). The informal culture at Air Ontario. One of the more striking findings to emerge from our research into flightcrew behavior has been the discovery of significant differences between aircraft fleets within organizations in attitudes regarding flightdeck management and in ratings of behavior in both line operations and Line Oriented Flight Training conducted in the simulator (Helmreich, Chidester, Foushee, Gregorich, and Wilhelm, 1990; Helmreich, 1990). These have been observed even in organizations with a strong commitment to standardization and form one of the justifications for implementing CRM training to develop common standards and values. Informal subcultures frequently tolerate or encourage practices which are at variance with organizational policies or regulatory standards.

Conditions at Air Ontario during the period of initiation of F-28 service would appear to have been conducive to the development of a non-standard subculture. These include previously noted lax regulatory supervision, high management turnover, the self-dispatch system with SOC personnel who lacked knowledge of the F-28 and were generally inexperienced, and the lack of clearly specified and enforced standard operating procedures. The reputation of being "Bush pilots" was attached to former Austin Airways pilots who formed a large percentage of the leadership of the F-28 program. Evidence of procedural variance is found in several reported practices. An example is writing mechanical problems or snags on paper to be passed to relieving crews instead of entering them in the aircraft logbook, thus permitting deferral of maintenance and avoiding the grounding of aircraft a practice in violation of Transport Canada regulations. Others include the so-called "eighty knot check," a visual examination of the wing surfaces during take-off to ensure that contamination had blown off prior to rotation, and the practice of making overweight landings. A related fact is that Captain Deluce, the Chief Pilot, had been involved in at least two earlier, reported incidents involving take-offs with snow or ice contaminated surfaces. These suggest that the culture, at least among former Austin Airways crewmembers, may have allowed crews considerable leeway in making decisions about whether to take-off with surface contamination - a practice that was not proscribed by current Transport Canada regulations. It seems likely that the message communicated during training, and in the Fokker manual for the F-28, that no snow, ice, or frost should be present on wings may have been discounted to some extent by crews who had successfully operated (albeit in different types of aircraft) with some degree of contamination. Additionally, the Check Airmen appointed for the F-28 fleet were inexperienced in the aircraft and with jet operations and may not have been in a strong position to impose standards.

II(j). Maintenance problems with the F-28. A number of maintenance problems were encountered with the F-28. These were exacerbated by a lack of familiarity with the aircraft on the part of maintenance personnel and a shortage of spare parts. The Journey Log for the accident aircraft, C-FONF, listed a number of problems between June and December, 1988, many deferred for extended periods. These included earlier problems with the Auxiliary Power Unit (APU) in August and October 1988. On several occasions in 1989 the cabin filled with smoke with passengers aboard. On the day of the accident, C-FONF was dispatched with an inoperative APU and had three other deferred maintenance items including roll and yaw in the autopilot and a fuel gauge reading intermittently. Other discrepancies that were brought to the attention of the cockpit crew by the cabin crew prior to the first flight on March 10 included inoperative exit lights, dim cabin emergency floor lighting, missing oxygen masks, and problems closing the main door because of a missing clip.

II(k). Flight Attendant training. The practice of Flight Attendant training at Air Ontario discouraged flight attendants bringing operational issues to the attention of the flightdeck and questioning operations. Training stressed the competence of pilots and fostered a position of total reliance on the cockpit crew. Two examples of the results of this separation of cabin and cockpit can be seen on the day of the accident. These included the hot refueling of the aircraft in Dryden at variance with the cabin manual and the failure of the flight attendants to relay passenger concerns about de-icing to the flightdeck. In contrast to this lack of communication, the concepts taught in Crew Resource Management stress the importance of complete information exchange between the flightdeck and the cabin.

III. The Physical Environment

A number of negative factors were present in the operating environment facing the crew on March 10. These included an aircraft with mechanical problems including the inoperative APU and poor weather that had created an early delay for de-icing in Winnipeg and a subsequent hold in Dryden because of weather at Thunder Bay. Indeed the weather was unsettled in the entire region that day necessitating non-standard alternates at a greater than normal distance, thus increasing dispatch fuel requirements. There was also a change in the passenger manifest in Thunder Bay increasing the passenger load and necessitating defueling to meet weight restrictions for take off and landing at Dryden. At Dryden, there was no ground start equipment making it necessary to leave an engine running and forcing the Captain to hot refuel. Finally, snow was falling during the station stop in Dryden.

IV. The Crew Environment

A number of factors that were present in the crew environment of the accident flight have been identified through research in other organizations as significant stressors that can serve to reduce flightcrew effectiveness. These include both situational factors surrounding the operation and characteristics of individual crewmembers.

Situational Factors

IV(a). Crewmembers' unfamiliarity with the aircraft and their training experience. Both Captain Morwood and First Officer Mills were new to the F-28 and had fewer than 100 hours of operational experience in this aircraft type. After completion of ground and simulator training at Piedmont, Captain Morwood returned to flying the Convair 580 and his line transition to the F-28 was further delayed by the Air Ontario strike. First Officer Mills received his training in the aircraft rather than the simulator. For Captain Morwood, the delay in reinforcing his training on the line could have rendered him less effective initially. For First Officer Mills, the lack of opportunity to acquire skills and confidence in the simulator could have had a similar effect.

There is growing concern in the industry, based on several recent accidents in the U.S., about the safety implications of pairing crewmembers new to an aircraft soon after completion of line indoctrination, particularly under adverse weather conditions. There is obviously a significant learning curve in becoming comfortable with a new aircraft, particularly one substantially different from prior equipment. One of the basic premises of the crew concept of flight operations is that crewmembers support each other in service of the goal of safe and effective flight management. When both crewmembers are still acquiring familiarity with the aircraft, the margin of safety is reduced. Efforts are underway in the U.S. to set requirements for operational experience after initial training and to mandate scheduling of newly qualified crewmembers with those having substantial experience in the aircraft type.

IV(b). Organizational background and lack of experience working together. Several additional issues made the pairing of Captain Morwood and First Officer Mills potentially stressful. One was the fact that Morwood came for the Air Ontario organization while Mills' background was with Austin Airways. Additionally, both Morwood and Mills had been operating as Captains in their prior aircraft. Individuals accustomed to acting as pilot in command have been noted to function less effectively when paired. These factors, combined with lack of enforced standard operating procedures (including the noted failure to specify pilot flying pilot not flying duties in the F-28 line indoctrination), could well have reduced the effectiveness of this crew as a team.

This trip was also the first time that the crew had operated together and Captain Morwood was displaced for two days. Experimental simulation research conducted by NASA-Ames Research Center (Foushee, Lauber, Baetge, & Acomb, 1986) found that crew coordination and effectiveness is increased by the simple fact of working together as a team. In this study, crews who were fatigued (from a three day, multi-segment line trip) or not fatigued (coming from days off) flew an experimental simulation involving bad weather and mechanical malfunctions. The purpose of the study was to explore the effects of operationally induced fatigue on performance. The most surprising and serendipitous finding from the study was that crews who had flown together previously, performed better than crews paired for the first time whether or not they were fatigued!

IV(c). Delays and stresses imposed by the operating environment. The initial segment of March 10 was delayed because of a need to de-ice the aircraft in Winnipeg. As noted, there were also major (APU) and minor mechanical problems with C-FONF. In a radio communication, Captain Morwood commented "...everything else has gone wrong today." After the first leg, an additional delay was experienced because of poor weather in Thunder Bay. On arrival at Thunder Bay, additional passengers were taken aboard from a canceled flight after refueling, making it necessary to remove fuel to meet weight requirements and causing it to depart more than an hour behind schedule. On arrival at Dryden, it was necessary to refuel with an engine running because of the lack of ground start capability. At the same time, snow was falling. As the Captain had fewer than I 00 hours in the aircraft type, he required a higher RVR than a more experienced pilot would have. He may (or should have been) concerned that visibility would become below his minimum prior to departure. The flight was already running late and a number of passengers had tight connections in Winnipeg. A final delay of approximately four minutes was incurred to await the arrival of a Cessna 150 which was experiencing difficulties because of the poor weather.

Personal Factors

IV(d). Captain George Morwood. Captain George Morwood was 52 years old and had more than 24,000 hours flying time. His operational experience was entirely in Canadian operations. He had worked for the predecessor of Air Ontario and had served as a Check Pilot and Chief Pilot for the Convair 580 at Air Ontario. He trained on the F-28 at Piedmont Airlines in January and February of 1988, but did not begin line flying in the F-28 until December, 1988. At the time of the crash he had 81 hours in the aircraft. His jet experience included approximately 600 hours in the Gulfstream G-2.

According to his record and peer reports, Morwood was above average in ability. He had shown concern with safety issues in his prior management positions and was aware of icing effects, including those caused by differential temperatures of fuel and ambient air. According to his record, he had delayed or canceled flights because of icing. Probably based on his long experience as a Check Pilot, and Chief Pilot, Captain Morwood was reported to be in the habit of operating as an "instructor" while flying. In theory, this characteristic could be an annoyance to highly experienced junior crewmembers such as First Officer Mills who had considerable experience flying as a Captain.

Captain Morwood was reported to have a strong commitment to on time operations and a high level of concern for his passengers. There were a number of delayed passengers with connecting flights in Winnipeg on March 10. In addition, Morwood had a scheduled personal trip immediately following his last flight segment. These factors could have heightened motivation to complete the scheduled flying.

IV(e). First Officer Keith Mills. Keith Mills was 35 years old and had more than I 0,000 hours flight experience. He began flying for Austin Airways as DHC-6 Co-pilot in 1979 and became a Captain on the Hawker-Siddely HS-748 in February 1988. He completed F-28 ground training in January, 1989 and aircraft training at Air Ontario. At the time of the crash he had 65 hours in the F-28 and approximately 3,500 jet hours in the Cessna Citation.

Mills had some record of difficulties with "stick and rudder" aspects of flying, but he met all regulatory requirements for competence. His failure to receive simulator training in the F-28 and Morwood's long experience and reputation as a perpetual "instructor" may have made Mills somewhat reluctant to practice optimal crew resource management concepts and to provide operational suggestions to Captain Morwood. Mills also had a scheduled personal trip at the end of his last flight segment.

V. The Situation of March 10

The picture that emerges from examination of the regulatory and organizational environments to which this crew was operating is one of an array of factors which served to undermine their effectiveness and to increase the stress of flight operations. None of these factors taken alone is likely to cause an accident - as evidenced by the fact that the F-28 was operated without incident or accident for months prior to March 1 0. However, when these factors were combined with the particular conditions of the physical environment (the inoperative APU, lack of facilities at Dryden, weather conditions, pressures to take off, etc.) the margin of safety was clearly reduced. Factors in the crew environment such as the operational unfamiliarity of the crew with each other and the aircraft doubtless exacerbated the situation.

V(a). Environmental Stressors. In considering the crew's actions on March 10, the environmental factors that may have been perceived as stressors should be reviewed. Psychological stress can serve to reduce individual and team effectiveness especially in the areas of interpersonal communications and coordination and decision making. Relevant classes of stressors include time pressure, and frustrations associated with inadequate resources and sub-optimal operating conditions. Captain Morwood and First Officer Mills faced a number of these conditions throughout their day. It may provide a useful context for the situation at Dryden to summarize them chronologically.

1. On accepting the aircraft in Winnipeg, the APU was found to be unserviceable. As noted previously, there were three additional, deferred maintenance items and other items in the cabin reported by the flight attendants.

2. The marginal weather throughout the region forced an initial delay for de-icing and the adoption of a distant alternate with a consequent requirement to carry additional fuel.

3. It was necessary to plan for "hot refueling" in Dryden because an engine would have to be left running. This may have triggered additional concerns because of company policy (and a stated requirement in the Fokker Publication on Cold Weather Operation) that the aircraft could not be de-iced with the engines running. However, it is not clear whether Captain Morwood had received a company memorandum about de-icing policy for the F-28.

4. SOC dispatched the flight with a clearly erroneous Flight Release. Testimony from pilot witnesses indicated little confidence in the SOC operation. It may have been a source of frustration or concern for the crew on this date to have been dispatched with no explicit accommodation for the unserviceable APU under adverse weather conditions.

5. Both crewmembers had fewer than I 00 hours in the F-28. In addition to the stress imposed by lack of familiarity with the aircraft, Captain Morwood had more restrictive limits for visibility because of his low experience level in type. This could have added to his concerns about getting in and out of stations with poor weather.

6. The flight was delayed on its initial stop in Dryden because Thunder Bay weather was below landing limits.

7. There was considerable confusion surrounding the loading of additional passengers in Thunder Bay and the need to defuel the aircraft to meet weight restrictions. The crew had to communicate with SOC through a radio relay by Air Canada since there was no direct communication link from the flightdeck. This situation increased the delay of the flight to more than an hour on departure from Thunder Bay.

8. The fire trucks required for hot refueling were not in position on the aircraft's arrival at Dryden. This factor added to the accumulating delay and possible frustration of the crew over the disruption surrounding the day's operations.

9. The date of the accident was the beginning of the March school break. There were many passengers with connections to make. The crew expressed concern over this in radio communications.

10. As the flight landed in Dryden, it began to snow, with the fall increasing during the stop. While the reported visibility was above minima, the actual visibility may have been at or below the Captain's minima at the time of take off.

While none of these issues alone can be considered an overwhelming stressor, taken in concert they indicate a taxing operational environment.

From the perspective of hindsight, it seems likely that a change in any one of a number of conditions might have provided the extra margin of safety needed. For example, a more stringently regulated and managed dispatch system would probably have precluded operations into Dryden on the return from Thunder Bay. An effective training program in Crew Resource Management could have resulted in a review of the operational situation involving both pilots and led to a critical evaluation of the decision to take-off without de-icing. Similarly, training that encouraged cabin crewmembers to share operational concerns with flightcrews and pilots to listen to such concerns might also have triggered further consideration of the implications of accumulating contamination on the aircraft.

The issues discussed in preceding sections have an empirical basis as significant influences on flightcrew behavior, but a weighting of each as a determinant of the outcome of Flight 363 cannot be made from the available record. Nor can the decision processes surrounding the take-off from Dryden be specified in the absence of Cockpit Voice Recorder evidence. However, it is possible to envision a likely scenario for the crew's actions based on consideration of the four sets of determinants of crew behavior described previously. It must be stressed that this represents a post hoc reconstruction that may be erroneous in part or whole.

VI. A Scenario for Crew Decision Making in Dryden

In retrospect, the decision to operate into Dryden on the return from Thunder Bay without a functioning APU was questionable, but understandable. The initial stop in Dryden was uneventful, despite a delay because of weather conditions in Thunder Bay. Although the forecast for the region showed a risk of freezing precipitation, on approach to Dryden conditions were VFR. Making the stop would minimize passenger disruption. However, once on the ground in Dryden, the weather and operational situation deteriorated. At the same time, the crew had conducted a day of flying that must be considered stressful because of the mechanical problems with C-FONF, increasing delays, the changed passenger load resulting in additional delay, and the crew's relative inexperience in F-28 operations. While on the ground in Dryden, the following issues faced the Crew:

1. Considerations surrounding refueling with an engine running.

2. Pressures to get passengers to Winnipeg for connections.

3. The inconvenience of stranding passengers in Dryden with limited facilities.

4. Logistic problems surrounding de-icing with an unserviceable APU and no ground start capability.

5. The need to import ground start equipment if both engines were to be shut down and consequent long delay.

6. Snowfall during the stop causing both aircraft and runway contamination and deteriorating visibility that might be below minimums for the Captain.

7. The implications of contamination on the aircraft.

8. The implications of contamination on the runway (including conflict between Fokker and Piedmont manuals in this area).

9. The additional delay posed by the arrival of the Cessna 150.

10. Planned personal trips which would be impacted by long delay in Dryden.

One of the effects of psychological stress (including that imposed by time pressure) is an inability to process multiple sources of information as effectively an under more relaxed conditions. As listed in the previous section, a case can be made for the fact that the crew, and especially Captain Morwood as pilot in command, was under considerable stress by the time the flight stopped for the second time in Dryden. It may also be inferred that the operating standards of Air Ontario and the absence of formal training and organizational endorsement of crew coordination concepts, would have tended to preclude rigorous crew evaluation of the operational situation.

Surrounding the decision to take-off are several critical questions. One is whether the crew was aware of the safety implications of the accumulating snow. As noted, Captain Morwood had a history of concern and awareness of icing risks. He had delayed the initial flight of the day for de-icing. Testimony by a representative of Transport Canada included an incident when Captain Morwood insisted on going back to the gate in the Convair 580 for de-icing even though the Inspector had remarked that the snow seemed dry and the propellers were blowing it off the wings. Also, a 1983 letter from Air Ontario management endorsing the Captain's authority to de-ice when circumstances require was found in Captain Morwood's flight bag at the accident scene.

A second question is whether the crew was aware of the accumulation of snow on the wings at Dryden' The Captain visited the terminal during the stop in his shirt-sleeves and would have been aware of snow falling. During a conversation with SOC during this period, he commented to Ms. Mary Ward that the weather at Dryden was "going down." The cockpit crew also had the ability to observe the wings from the cockpit and the testimony of informed passengers indicated that snow was accumulating visibly there. It seems inconceivable that the crew would have been unaware of snow on the wings. The fact that Morwood inquired of the station manager at Dryden about de-icing facilities there also suggests awareness.

Despite his knowledge of icing and probable awareness of the snow gathering on the wings, it seems most likely that Captain Morwood weighed costs and benefits surrounding the issues listed above and concluded that the best course of action would be to take off expeditiously. Several things may have influenced this decision. One is that because of the multiple stressors involved in the situation and his focus on completing the trip, he failed to weigh the risks as heavily as the benefits from getting out before the weather deteriorated further. The ambiguity of regulations regarding icing could also have influenced his decision. Although it was noted that emphasis was placed in training at Piedmont on taking off with no wing contamination, he may not have felt that the issue was as serious in the F-28 as other aircraft given higher rotation speeds and additional opportunity to blow the accumulation off during take-off roll.

The role of First Officer Mills in this decision is, of course, indeterminate. However, based on considerations regarding experience and status it is not likely that he was heavily involved by Captain Morwood.

There was probably a misperception about the nature of the contamination as it relates to "cold soaking", the situation when portions of an aircraft are at a temperature below the ambient temperature because of having descended from altitudes where ambient air is colder or from heat transfer to areas containing fuel colder than the ambient temperature. Pilots interviewed by the author were primarily concerned with heat transfer at high altitudes and less aware of the phenomenon occurring on the ground due to cold fuel in wing tanks. The Piedmont manual which was used at Air Ontario addresses this phenomenon in a section on Cold-Weather Operations. It states:

"When the tanks contain sufficient fuel of sub zero temperatures as may he the case after long flights at very low ambient temperature, water condensation or rain will freeze on the wing upper surfaces during the ground stop forming a smooth, hardly visible ice coating. During take off this may break away and at the moment of rotation enter the engine causing compressor stall and/or engine damage. " (Piedmont F-28 Manual, Exhibit 307 3A-24-1)

A decision could well have been reached that the snow would blow off, given the large fluffy flakes coming down and the lack of accumulation on the tarmac surrounding the aircraft. The possibility that a layer of rough ice caused by cold soaking extended to the leading edge was probably not entertained by either Morwood or Mills.

Psychological pressure to complete the trip as scheduled, commonly referred to as "get home-itis", cannot be ruled out. Captain Morwood was clearly concerned about holiday passengers with connecting flights in Winnipeg and both he and Mills had personal trips planned after completion of the trip. Had the flight been canceled in Dryden, it would have been necessary to fly in ground start equipment causing a lengthy delay and disruption of crew and passenger plans. Once on the ground in Dryden, the implications of a long delay doubtless had a subtle influence on the decision process.

A final chance to re-evaluate the situation was probably missed when the flight took its final delay for the landing of the Cessna 150. However, the accumulation of stress and frustration surrounding the day's operations had probably reduced the crew's effectiveness and decision making capabilities by this time.

While the Captain as Pilot in Command must bear responsibility for the decisions to land and take-off in Dryden on the day in question, it seems equally clear that the aviation system failed him at the critical moment by not providing effective management, guidelines, and procedures that would assist him in such decisions.

In the following section, observations and suggested corrective measures are offered in the hope that they may provide greater resources for future crews who find themselves in stressful situations trying to evaluate multiple pieces of information and having to make choices among unpleasant, alternative courses of action.

VII. Observations

The following are corrective measures that could be taken to increase system safety and effectiveness. It is noted that the first recommendation of the Commission to Transport Canada was to remove the ambiguity from regulations surrounding wing contamination and that this was favorably received.

VII(a). Monitoring of air carrier operations. It would be valuable to establish guidelines for air carrier management in terms of qualifications needed for effective job performance. A similar set of standards could be established for Air Carrier Inspectors and others involved in surveillance of airline operations. Requirements for inspectors and check airmen could include training in the evaluation of human factors aspects of flight operations.

Training in the conduct of air carrier audits and requirements for qualification of audits could be strengthened. In particular, emphasis in audits should be on observation of line operations evaluating both human factors and technical proficiency.

Strengthened requirements for flight dispatch and the training of dispatchers should be developed for all airline operations.

VII(b). Winter operations. Yearly training and review of Winter operations procedures should be conducted. This should include not only general issues regarding icing, cold soaking, and de-icing procedures, but also information specific to particular aircraft types as needed.

VII(c). Common standards for major airlines and their feeder operations. Airlines operating under a common designator should maintain the same standards of training, dispatching, and performance. The need is probably greater for effective training and organizational support in smaller carriers that operate into secondary stations with fewer facilities. In many cases, pilots in regional carriers may have had less experience and less formal training. The resources of the major carriers could be highly beneficial for the safety and effectiveness of these regional carriers and could allow them to establish levels of training that they could not effect independently.

VII(d). Formal training in Crew Resource Management for all crewmembers. Accumulating experience in the U.S. and many other countries has demonstrated the importance of CRM training. The U.S. has encouraged this training through an Advisory Circular and it is a requirement for operating under a new Special Federal Aviation Regulation called the Advanced Qualification Program. Efforts are underway in the U.S. to initiate a regulatory requirement mandating CRM training for all air carriers operating under Parts 121 and 135 of the Federal Aviation Regulations. A premise of the CRM Advisory Circular, supported by empirical research, is that a single training experience in CRM concepts is insufficient to provide long term changes in crew coordination and performance. Such training must be accompanied by opportunities to practice the concepts and to receive reinforcement for their use. Check Airmen and Instructors have been identified as critical to this endeavor and should be given training in the evaluation and reinforcement of human factors issues as an extension of their traditional role (Helmreich, Chidester, Foushee, Gregorich, & Wilhelm, 1989). This type of evaluation and reinforcement can and should occur both in ground training and during line checks and should center on clearly understandable exemplars of effective and ineffective performance that have come to be called behavioral markers of crew performance. There is a growing belief that this training can be effectively extended to cabin crews and other operational personnel. One can speculate that had both the flight attendants and cockpit crew completed CRM training and accepted its concepts, there might have been an exchange of information that would have precluded the take off.

VII(e). Crew oriented training and evaluation. The historical emphasis in aviation has been on individual, technical proficiency and both training and evaluation have centered on the performance of the individual pilot. However, data from accidents and incidents suggest that the CRM-related issues isolated in accidents and incidents involve failures of crews to operate effectively as teams. Many airlines and military units have reacted to this by increasing the emphasis in training and checking on crew-level performance. In checking line operations this is accomplished by including the performance of the crew as a unit as part of the evaluation and debriefing (for example, using the CRM/LOS Checklist as a template for evaluation). Another approach being used increasingly (and required in the U.S. for carriers that will operate under the Advanced Qualification Program) is the use of Line Oriented Flight Training (LOFT) which involves complete crews training in simulators under realistic operating conditions including flight releases, air traffic communications, and facing a variety of operational problems including inflight emergencies. A key to the success of this training it that it is non-jeopardy meaning that crews are allowed to experiment with a variety of behaviors and approaches without placing their licenses at risk. Events are allowed to proceed without intervention by the Inspector and are usually recorded on videotape for subsequent review and debriefing. In its early development, LOFT required access to high fidelity simulators placing this form of training out of the reach of many organizations, especially regional and commuter airlines. However, recent research and theorizing (Franz, Prince, Salas, & Law, 1990; Helmreich, Kello, Chidester, Wilhelm, & Gregorich, 1990; Helmreich, Wilhelm, & Gregorich, 1988) suggests that low fidelity simulators and training devices may provide excellent settings for training in crew coordination and should make the technique available to almost all organizations.

VII(f). Establishment of a Safety Office in all air carriers. In addition to regulatory monitoring of air carriers, an independent Safety Office can serve an important function in isolating potential threats to safety. A Safety Officer with direct access to top management is in a position to initiate corrective action when threats to safety are uncovered. In addition to training in investigative techniques, training in human factors, database management, and analysis would also be highly desirable for Safety Officers and their staffs.

A safety study was recently conducted by the National Transportation Safety Board on the safety of commuter airlines due to the fact that accident rates continue to be higher than those for domestic Part 121 airlines. Issues discussed in the study range from: the need for sweeping regulatory action to the adequacy of Part 135 pilot training. In the section on Pilot Training and Qualifications, Crew Resource Management training (CRM), availability and use of flight simulators and advanced training devices, and the Advanced Qualification Program (AQP) were discussed. The Safety Board agrees with the FAA's move toward uniformity in pilot training requirements for Parts 121 and 135 but further urges the FAA to proceed with: mandatory CRM training programs; the continued promotion and development of the AQP; the certification and operation of Part 142 training centers; and issues addressing pilot operating experience and crew pairing. For a copy of "Commuter Airline Safety" (NTSB/SS-94/02), published by the National Transportation Safety Board, please write or call: Public Inquiries, NTSB (RE-51), 490 L'Enfant Place East, S.W., Washington, DC, 20594,(202) 382-6735.

by: Maureen A. Pettitt
Western Michigan University

In the October issue of the CRM Advocate, John Lauber discussed the notion of "safety culture" as a powerful determinant of human performance in airline operations. As Dr. Lauber points out, one implication of the concept of a safety culture is "that organizations, like individuals, can cause accidents' The National Transportation Safety Board now includes organizational factors as part of its investigation findings and safety recommendations. I would like to expand somewhat on the concept of a safety culture by a discussion of organizational cultures, the conflicts that can arise within an organization, and how CRM training might be used to develop constructive conflict management skills.

Dr. Lauber notes that Crew Resource Management (CRM) training is founded in the recognition that "the safe and efficient operation of the modem aviation system depends on well-designed ... aircraft operated by capable, well-trained individuals operating as highly integrated, well coordinated teams." CRM programs utilize case studies, teamwork activities, personality assessments, and simulations to examine communication skills, interpersonal styles and leadership, planning and coordination, situational awareness, and decision making. The purpose of familiarizing pilots with these principles is to encourage crew members to perform more effectively as a team.

Organizational structure and relationships are discussed within the context of the formal and informal systems, roles and status, group norms, crew coordination, communications, and leadership. Pilots become aware of the negative consequences associated with poor leadership and interpersonal relationships and with group norms that are counterproductive, for example: pressure for conformity, reluctance to ask questions or advocate for a course of action, overemphasis on the relative importance of roles, overestimation of the competence of high status members, underestimation of the competence of low status members, and a sense of helplessness among low status members.

Less attention, however, has been given to issues related to recognizing the cooperative and competitive aspects of organizational and group goals or to the skills required to confront and manage conflicts between goals. Yet conflicts between group and organizational goals or conflicts between the goals of group members are frequently part of the decision process in airline operations. When there is no system for resolving conflicts within the decision context, several negative consequences are likely: the wrong issues or items receive attention, decision authority becomes concentrated, there is restricted sense of inquiry, the quality of decisions becomes less important than resolving the conflict.

Ultimately, the decision makers become susceptible to erroneous judgment.

Organizations And Conflict

Almost all discussions about organizations make an assumption: organizations are oriented toward a specific goal. Yet the goals pursued by an organization are multiple, and frequently in conflict. The stated goal or objectives are almost always highly ambiguous and almost never operational. They do not identify specific steps or tasks required to accomplish them, so there can be many routes to goal achievement. The route prescribed by the formal system may differ from that actually taken by the informal system. Further, it is difficult to know if the goals were achieved in the most effective manner. Satisfactory profits this year may be earned at the expense of satisfactory profits three years from now.

Consider this hypothetical example. Aircraft damage during ground operations can be a major source of frustration and cost. An airline might respond to this problem by developing an accountability system, a system which generates a multiplicity of forms and a rash of statistical data. This system also results in a high level of defensiveness between groups (maintenance, ground service, etc.) and blame avoidance becomes common behavior.

High on management's list of organizational goals, however, is on-time departure. Rewards and discipline within the formal system reflect this priority. Each responsible group works diligently to meet the airline's criteria, to the point that incidents of aircraft damage -- or conditions that could easily result in aircraft damage -- are ignored or suppressed.

Management's efforts to find a solution to this problem are not likely to include the recognition of the conflicts created by formal organization's goals and decisions. The formal system regards on-time departures as an organizational and operational priority. Further, management decisions are based on the belief that safety requirements necessitate a high degree of functional accountability and, further, that negative sanctions are necessary to ensure standard performance.

The unforeseen consequences of the emphasis on on-time performance criteria could be disastrous for the organization: some necessary preflight functions are neglected, the responsible groups expend too much energy on blame avoidance and defensive behavior, groups and individuals start to perceive that there is little reward for competence.

Culture And Conflict In The Cockpit

The formal organization within which flight crews function has, as noted above, publicly stated goals and an administrative system to ensure that those goals are met. These goals reflect the culture of the organization, that is, the underlying values and beliefs that inform organizational life. The attitudes and behaviors which stem from this culture establish, at least in part, the organizational climate (or environment). The culture and climate provide members with their understanding of the purpose and meaning of the organization and of their work.

The reality is, however, that there are also informal systems within the formal organization. These informal systems are usually accompanied by a culture and norms distinct from the formal system. The hypothetical example above points out several negative consequences of conflict between formal and informal systems within the organization. Equally negative outcomes are the result of within-group conflict.

Cockpit crews function, in general, as what has been described by Richard Hackman as a "self-managing work group." That is, they:

(a) are an intact and identifiable social system (even if they are small or temporary),

(b) have a defined piece of work to accomplish, and

(c) have the authority to mange their own task and interpersonal processes as they carry out their work.

In other words, not unlike the formal organization, the cockpit has both an administrative system and a cultural system which enables the "production" of available passenger miles.

The primary task of the crew is to fly the aircraft in a safe and efficient manner. Today, the crew also has a secondary task--interacting as a team. This task is essential to safe, effective flight operations: the group interaction process indirectly affects operational effectiveness because the process directly affects the strategies used by the crew for problem solving and decision making.

The cockpit social or cultural system is determined by such factors as the flying task, cockpit technology, personality of the crew members and expectations each has regarding roles and task performance norms and includes traditions, ritual, and a specialized language associated with the cockpit and its membership. The effective and efficient conduct of activities requires an administrative system through which tasks are coordinated and resources are managed. The role of each crew member is, to varying degrees, prescribed by Federal Aviation Regulations and the airline's flight operations manual which define the duties, responsibilities and authority assigned each functional role.

Both the administrative and the cultural system provide the infrastructure for role differentiation in the cockpit. These differences are maintained and protected through group norms and symbols, and are reinforced by the pilot seniority system which provides senior crew members with both higher pay and better work schedules. While such a system may be appropriate or even desirable for the airlines and their pilots, it also legitimates role differentiation and a hierarchical role structure.

The cockpit climate and group norms which result from the administrative and cultural system will depend on the personality characteristics of the crew members, their experiences and capabilities, and the leadership style of the captain. The cockpit climate can, however, create the potential for intergroup conflict and for conflict between organizational and group goals. Conversely, cockpit climate and norms can also encourage constructive conflict management.

Conflict Management

What I am suggesting, of course, is that an airline's "safety culture" should include recognition of intragroup and intergroup conflict and a commitment to developing conflict management skills. While the term conflict usually conjures up visions of disputes, chaos, or violence, it can be constructive as well as destructive. Conflict can serve as the source of personal, organizational, and social change.

The concern here is the training and development of conflict management skills which, in turn, lead to constructive strategies for handling the conflict. Conflict strategies are generally categorized, more or less, as one of the following:

1. Forcing - the most powerful get their way.

2. Avoiding (or Accommodating) - people pretend there are no differences or minimize their importance.

3. Negotiation - people bargain in an attempt to get their way or to maximize their personal gain.

4. Confronting - people open all issues and data to inspection in an effort to initiate a solution.

The worst situation for managing conflict is when there is high interdependence among the parties, bad relationships, and when smoothing or avoiding differences are used to manage conflicts. The management of conflict is most effective if:

1. Conflicts are brought out into the open.

2. The conflicting positions are discussed in a reasoned and non-threatening manner.

3. The issues are resolved and a commitment made in a timely fashion.

This approach calls for a high order of interpersonal skills and a willingness to take risks. These skills are not too different from those which characterize most CRM training programs: active listening; developing a cooperative, problem-solving relationship with others; and, establishing a climate that encourages group decision making and expands the options available for problem solving.

As with other CRM skills, there are certain educational principles which apply to developing conflict management skills:

1. A significant change in behavior is unlikely to occur from conflict management training unless there is emphasis on skills.

2. The social and cognitive skills involved in constructive conflict management are different from those involved in technical operations.

3. The use of constructive management skills is more likely if the organizational culture is one that acknowledges both cooperative and competitive interests and encourages collaborative problem solving at all levels.

The material contained in The CRM Advocate back issues is the property of the contributing editors. No duplication of any kind is authorized without the express written permission of the editor. All rights reserved. For training and information purposes only. The intent of the editors is shared information, through controlled distribution to the benefit of the safety of flight.

[ Home ] [ Up ] [ 93.1 ] [ 94.1 ] [ 94.2 ] [ 94.3 ] [ 94.4 ] [ 95.1 ] [ 95.2 ] [ 95.3 ] [ 95.4 ] [ 96.1 ] [ 96.2 ]