Aviation Maintenance Human Factors and Performance Excellence

Introduction and Background

U.S. statistics indicate that 80% of aviation accidents are due to human errors, with 50% attributable to maintenance human factor problems. Current human factor management programs have not succeeded to the degree desired. Many industries today use performance excellence frameworks — such as the Baldrige National Quality Award framework — to improve overall organizational effectiveness, organizational culture, and personal learning and growth. A survey administered to a sample population of senior aviation maintainers in 18 countries revealed a consistent problem with aviation human factors and the need for a more integrated framework to manage human factor problems in aviation maintenance.

Consider a common scenario in an aviation organization such as the military: personnel are told they must work over a weekend because a fleet grounding issue requires all aircraft to be inspected by Monday morning. After nearly 60 hours of work that week, technicians are exhausted. Their organization has provided Human Factors (HF) training and management has told them to call for a time-out when tired — yet management simultaneously declares an urgent need to complete inspections over the weekend.

Although the body signals that it can no longer continue, the mind insists on being a team player or risk failing the mission. Reflecting on the year, one recalls multiple close calls with errors of judgment and the famous lecture from HF training: "the chain is only as strong as its weakest link." The next morning, before arriving at work, news arrives that a colleague the night before struck and damaged an aircraft nose landing gear with a Harlan tractor, prompting management to call an urgent safety briefing. This is immediately connected to an earlier complaint about the Harlan and Toyota tractors having opposite reverse gear configurations — push the lever forward in one and it goes forward; the same motion in the other reverses it. Does this sound familiar?

Today, more than ever, the aviation world faces the constant challenge of addressing human factors in maintenance. While there have been several advances in the study and implementation of human factor programs, there are still significant inconsistencies in how these programs are implemented, resulting in varied outcomes.

Aircraft maintenance encompasses fast turnaround, high-pressure environments in which hundreds of tasks may be performed simultaneously by large numbers of personnel on highly complex and technologically advanced systems in a confined area. It is very easy for information and tasks to fall through the safety net. Events around the world in the late 1970s, 1980s, and early 1990s — involving crashes or serious accidents — alerted the aviation world to the fact that although aircraft were becoming much more reliable, human performance had the potential to obliterate those technological advances.

This research analyzes the top human factor problems in aviation maintenance and evaluates a holistic solution through a performance excellence framework. It begins with a brief history of HF programs and the changes that have taken place over the years, then explores current HF programs adopted by various organizations and examines why HF errors occur and how comprehensive current solutions are. Finally, it considers the Baldrige National Quality Program and criteria for performance excellence to formulate a more comprehensive solution to managing HF in maintenance — in essence, a more systemic solution, since human factors are about more than just people.

In the late 1970s, Cockpit Resource Management (CRM) featured prominently in pilot training. The term was used to describe the process of training flight crews to reduce pilot error by making better use of resources on the flight deck. The name was later changed from Cockpit to Crew Resource Management (CRM) to shift the emphasis of training toward cockpit group dynamics. Some airline programs addressed specific topics such as team building, briefing strategies, situational awareness, and stress management (Byrnes and Black, 1993). In the early 1990s, CRM training began to reflect the many factors — including organizational culture — within the aviation system that can determine safety.

Similarly, but much later, it was not until the 1990s that Maintenance Resource Management (MRM) was made available to maintenance personnel. After years of accidents, many caused by HF errors, little was done to determine root causes. Unlike CRM, MRM was very new to aviation maintainers, and it was not until June 10, 1990 — when a cockpit window blew out at 16,000 feet and a pilot nearly went with it — that an in-depth examination of the contributing factors to a maintenance error was undertaken (System Safety Services, 2000). David King of the United Kingdom was one of the first to examine HF in the light it is understood today.

Review of Human Factors Literature and Statistics

The need for a change in approach to human errors and their reporting was reinforced during the CAA-sponsored 12th Symposium on Human Factors in Aviation Maintenance, held at Gatwick Airport, England, on 10–12 March 1998. It was the first of the international symposiums involving the CAA, FAA, and Transport Canada.

The foundation of Human Factors training as a modern aviation tool was probably initiated in the United States at a workshop sponsored by NASA in 1979. This workshop grew from NASA research into the causes of air transport accidents. The International Civil Aviation Organization (ICAO) now requires organizations to include HFIM training, helping maintenance personnel avoid errors they never intend to make (System Safety Services, n.d.).

U.S. statistics indicate that 80% of aviation accidents are due to human errors, with 50% due to maintenance human factor problems. Most programs currently implemented are designed to identify HF errors, educate personnel on their causal potential, suggest ways to contain and correct the problem, and create an HF error-free environment. However, the percentage of HF errors in aviation mishaps is on the rise. There is a need for a more integrated and holistic approach to HF management.

Due to limited funding and time, this researcher limited the surveys to the Senior National Representatives (SNRs) at OO-ALC/YPX to provide a summary of their countries' perspectives on HFIM management. Research results of the survey are assumed to be representative of possible results from similar units within their respective Air Forces.

In the United Kingdom between 1982 and 1991, there were 1,270 Mandatory Occurrence Reports (MORs) involving maintenance errors submitted to the CAA Safety Data Department (CAA, 1992). Of these, only 230 resulted in an unexpected or undesirable occurrence that interrupted normal operating procedures and could cause an accident or incident. The CAA concluded that there was no significant risk to the public. In the period 1992–1994, however, there were 230 MORs, and from 1995 to 1996 there were 534 — reported errors were occurring at a greater frequency. A study by Boeing in 1993 examining 122 occurrences between 1989 and 1991 revealed that 56% of human factor errors resulted in omissions, with a further 30% resulting in incorrect installations.

In a field test by Boeing conducted from 1994 to 1995 with nine maintenance organizations, the main types, causes, and results of errors were summarized as follows (Boeing, 1996): the top three operational events were flight delays (30%), aircraft damage (23%), and air turn-backs (15%). The top maintenance error types were improper installation (35%), improper testing (15%), and improper servicing (12%). The top contributing factors were information (50%), communication (42%), and job/task/environment conditions (40%).

In 1998, the Australian Transport Safety Bureau (Hobbs & Williamson, 1998) surveyed close to 1,400 Licensed Aircraft Maintenance Engineers (LAMEs). The most common outcomes for airline-related maintenance occurrences were: systems operated unsafely during maintenance; towing events; and incomplete installation. The most common outcomes for non-airline occurrences were: incorrect assembly or orientation; incomplete installation; and persons contacting hazards.

The most common causes of these unsafe acts revealed in the 1997 Australian Transport Safety Bureau survey included: pressure (21% airline, 23% non-airline), fatigue (13%, 14%), coordination (10%, 11%), training (10%, 16%), supervision (9%, 10%), lack of equipment (8%, 3%), environment (5%, 1%), poor documentation (5%, 4%), and poor procedures (4%, 4%).

A ground crew attitude survey conducted in the military in Asia (classified source, n.d.) revealed similar findings to those of the Australian Transport Safety Bureau. The surveys were conducted bi-annually from 1999 to 2003 on approximately 2,500 aviation technicians. In the survey conducted in 1999, the top three violations were: servicing without a checklist; speeding; and omitting job steps. Approximately 20% of those surveyed disclosed that they would violate rules daily or once a week. The top three reasons for these violations were: too much work and too little time; insufficient manpower; and time pressure to complete duties.

In 2003, when the survey was conducted again, several key initiatives had been implemented to address HFIM, including: implementing a Human Factor training program initiated by Gordon Dupont, CEO of System Safety Services, in 1999; training 100% of licensed aircraft engineers in Human Factors Management; implementing a MEDA-type Human Error Analysis Tool (HEAT); embracing a local version of the Malcolm Baldrige Performance Excellence Framework for the military over six years from 1998; and adopting additional performance excellence measurement tools such as the Balanced Score Card and Enhanced Value Organization principles.

The survey results comparing 1999 and 2003 revealed several significant improvements. A new segment on Safety Culture showed that 99% of personnel agreed the organization placed strong emphasis on safety and quality, and that management (96.43%), supervisors (97.30%), and personnel (94.38%) all took safety and quality seriously. Overall violations observed daily or weekly were reduced by 4%. The frequency of violations categorized as "very infrequent" improved by 22% (from 21% to 43%). There was an 11% reduction in holding back to call a time-out. Weekly overtime was reduced by 16%. Open reporting improved by 16% (from 66% to 82%).

Current Human Factor Programs in Aircraft Maintenance

Several UK maintenance organizations have pooled their Maintenance Error Management System (MEMS) data using a common MEDA taxonomy. Results presented at a MEMS-MEDA seminar in the UK in May 2003 showed that the top three items for improper installation were incomplete installation, wrong orientation, and system not re/deactivated. Individual performance factors and information were the most commonly cited contributing factors across all error categories.

The maintenance error trends in the U.S., Australia, Asia, and the United Kingdom from 1982 to 2003 are alarmingly similar and continue to plague the aviation industry. In some areas — particularly in military aviation — the trends in maintenance human factor errors have continued to increase. A closer examination of the statistics indicates that these trends are due mainly to lapses in organizational operational culture and business processes. Time pressure, stemming from lack of manpower and excess workload, appears to be the dominant factor. Recent HF training has focused on these lapses and their detrimental consequences, yet an uptrend in maintenance errors and violations persists.

Several HFIM courses have evolved since ICAO required HFIM training, including those developed by the UK CAA, FAA, and JAR-compliant courses designed to ensure consistency and conformance to minimum standards set by governing bodies. A typical HFIM course developed to comply with JAR145-12 includes: a general introduction to human factors; a safety culture and organizational factors overview; human performance limitations and human error models; environmental issues impacting human performance; procedures, information, tools and practices; professionalism, integrity, communication and teamwork; and organizational HF programs including the management of HF errors.

Gordon Dupont, formerly of Transport Canada and now CEO of System Safety Services, is a renowned human factors proponent who conducts HFIM courses around the world in the aviation sector. He is best known for his "Dirty Dozen" classification, which depicts the 12 most common human factor errors in maintenance. Gordon conducts three workshops — HFIM Parts One through Three — covering the background to HFIM behaviors and errors through case studies, organizational culture and risk management, and strategies for managing them.

As can be seen from the typical course structures above, current HFIM courses generally adopt the three E's: Educate personnel; Equip personnel with the tools necessary to contain, correct, and prevent HF errors; and Evaluate the management of HFIM programs. This is the core coverage for most HFIM courses and programs adopted by commercial airlines and other aviation industries.

Professor James Reason's "Swiss Cheese" model propounds that several latent conditions exist prior to an active failure or unsafe act. These failed or absent defenses align to cause a mishap or injury waiting to happen. Notably, one of the first lapses in defenses in his model begins at the level of organizational influences. Gordon Dupont modified Reason's Swiss Cheese model by incorporating his "Dirty Dozen" human error factors as the preconditions to unsafe acts that could eventually cause an accident or incident. Dupont attributes 70% of accident causation to fallible decisions by management, deficiencies in line management, and the "Dirty Dozen" preconditions (G. Dupont, n.d.).

While many of these programs have genuinely made the aviation work environment safer, most still view HF from a "people" perspective rather than an "organization" perspective. There is a need to develop programs that improve the performance of all areas of an organization as a whole, providing long-term solutions to HFIM.