Human Factor Analysis is a staple of aeronautical studies, where the correct behaviour of people, during the operational phases (flight, maintenance, ground services), plays a key role in ensuring safety. The analysis of human behaviour has shown that individual actions are guided by a complex “operating system”, therefore the study of human factors aims to facilitate and optimize performance in relation to the system, reducing human error. Mistakes do not necessarily indicate negligence or incompetence. It may happen that even without violating established rules errors would occur, the norms themselves could subject to change in an uncertain scenario.
It has been established over the years that constant control of the human factor is essential to correctly steer the behaviour of individuals within small and large organizations. But, how can we identify the human factors that lead to human error?
In literature it is possible to find different techniques that classify and optimize human factors by connecting them to involuntary mental processes (e.g. inattention, disregarding the rules, habit, ignorance) or intentional mechanisms, like violating rules and procedures with the specific purpose to cause harm, or “in good faith” to expedite certain actions bound by existing procedures.
Other techniques, more difficult to implement, however, focus the analysis on the trigger cause and the conditions in which the error occurs. Often, human errors occur due to guidelines that are hard to adhere to, therefore the error is concealed within the rulemaking organization (e.g. the airline, in case of pilot error).
The consequences of a catastrophic event become evident only when all the internal (psycho-physical conditions) and external (interactions between human and organization) factors affecting efficiency and reliability are combined.
How can we extend on the concept of human factor outside of the aerospace sector?
Management processes play a fundamental role in understanding the causes of the mistakes, however, many catastrophic events are frequently attributed to the incidence of the human factor. A generic organization, if constantly improved by well-defined rules and processes, can perfect all aspects concerning the human factor. Therefore, if an organization is established, managed and functioning through human work, it is essential to investigate the risks aspects related to the human factors.
Is it conceivable to reduce the level of risk associated with human errors by optimizing our processes?
This is how neural networks can help, to optimize some human behaviours with an approach which, opportunely integrated with polyvalent logic methods (e.g. Fuzzy Logic), can deal with ambiguous, imprecise and not exactly defined aspects such as the human factors. It can be hard to understand how an individual can react in different situations and above all it is not a given that all individuals behave in the same way when facing the same work dynamics.
To perform such a study is essential to start collecting data to understand how the human factor can influence some choices in typical work situations. Expert opinion is essential to evaluate human behaviour with reasonable accuracy. A process of risk reduction, linked to the human factor and defined by a numerical approach, can be treated by means of neural networks to strive for an optimized solution.