Reconsidering Human Error


2nd Lt. Seth Montgomery, a student pilot assigned to Training Squadron 28 (VT-28) straps himself into a T-6 II turbo prop plane. (Photo by Mass Communication Specialist 2rd Class Victor R. Navarrete)

2nd Lt. Seth Montgomery, a student pilot assigned to Training Squadron 28 (VT-28) straps himself into a T-6 II turbo prop plane. (Photo by Mass Communication Specialist 2rd Class Victor R. Navarrete)

This year marks the 70th anniversary of one of the most influential publications in human factors. In 1947, psychologist Paul Fitts, along with Air Force Capt. Richard Jones and support from Lt. Col. A. P. Gagge and Col. Edward Kendricks of the Aeromedical Laboratory at Wright-Patterson Air Force Base, researched results of hundreds of non-combat aviation accidents. The causes of the accidents were originally categorized as pilot error, but the authors were unsatisfied with these conclusions. The title of their report was “Analysis of 270 ‘Pilot-Error’ Experiences in Reading and Interpreting Aircraft Instruments.” The quotations around pilot-error were intentionally ironic, suggesting the authors would not use this term to classify the failures. They concluded that the actual source of failure was poorly designed instruments.

What makes The Fitts and Jones report so influential goes beyond design error being a substitute for human error. It was the scientific approach they applied to understanding performance. They gathered pilot performance data by studying various types of errors, interviewed pilots about their experiences and considered human strengths and weaknesses of how we process information. They analyzed errors as the result of interactions between multiple components in the cockpit and determined errors can result not from a single component working in isolation (i.e. the human), but by the interaction of multiple components. For example, the incongruent motion between the turn needle used in one instrument and the bank indicator appearing in a separate instrument was a source of confusion not previously discovered.

As a result of this lack of uniformity, the pilot must change his mental set each time he shifts his eyes from one instrument to another. He can undoubtedly learn to do this in time, as is shown by the skill attained by experienced instrument pilots. In fact, the shift in reference may become so automatic
that experienced pilots are unaware that it is happening.

But the necessity of constantly changing mental attitude certainly makes for more difficulty to learning instrument flying and may lead to occasional reversal during emergency conditions. It cannot be overemphasized that the pilot who must use his full set of instruments in critical maneuvers should have a panel in which he can shift from one instrument to another without conflict.

Paul Fitts went on to become a human factors pioneer. The influence of his work is still around today, whether in cockpit design or size and location of buttons on a smartphone. Wright-Patterson AFB still hosts research efforts advancing human factors in aviation, including the Navy Medical Research Unit-Dayton. Other areas of efforts include numerous divisions of naval aviation and the Department of Defense Human Factors Engineering Technical Advisory Group. Mishap investigators also use the Human Factors Analysis and Classification System, which intends to identify the latent conditions that put human operators in positions of failure. Each of these entities emphasize what Fitts initiated, which is understanding that performance of aviators, and all human users of technology for that matter, is the result of the relationship between the human and the work. Despite his legacy, and the ongoing efforts of many human factors researchers, human error is still a prominent causal factor discussed within safety communities.

The statistic often cited when discussing safety and human performance is that an estimated 80 percent of accidents are attributable to human error. If only it were that simple. The statistic implies that most accidents can be traced to the solitary human component within a complex system and that the remaining accidents are attributable to mechanical failure. Separating outcomes into distinct categories deprives us of a deeper understanding of the interaction between these components. Seventy years after Fitts and Jones, it is time once again to challenge the notion of human error as being a useful concept.

If Fitts and Jones had been satisfied with pilot error as the decisive factor in the accidents they analyzed, they may have recommended more training as an intervention to improve performance, or encouraged pilots to maintain adequate situation awareness and to avoid complacency. These solutions, however, focus on the human component alone and oversimplify the interactions occurring within complex systems. They often do not take into consideration strengths and weakness of human being as performers in such systems. Humans have natural tendencies and limitations, especially in demanding situations. But we are amazingly creative, adaptable, and resilient. The Bravo Zulu section of this magazine commemorates these very strengths when aviators overcome such difficulties.

The next 70 years will continue to present new challenges in aviation and human factors. If the progress made during the last seven decades is any indication, we will continue to embrace and conquer these challenges.

These are solvable problems. But we owe it to ourselves, and certainly to the future generations of aviators, to first and foremost correctly identify the problems.