The Ergonomics Open Journal

2015, 8 : 1-12
Published online 2015 January 22. DOI: 10.2174/1875934301508010001
Publisher ID: TOERGJ-8-1

Aviation Human Error Modelled as a Production Process

Dirk Pons and Karla Dey
Department of Mechanical Engineering, University of Canterbury, Private Bag 4800, Christchurch 8020, New Zealand.

ABSTRACT

As technology systems have become more complex, so it is increasingly difficult for human operators to comprehend how the system is behaving. There is a need to better understand the causes of human-error in the context of the situational variables. The specific case under examination is the process of landing an aircraft. This paper applies a production engineering perspective. By production-process, we refer to the set of actions that are necessary to move a system from one state to a desired new state. Such processes involve mechanisms for taking inputs and converting them to the desired outputs, under certain controls and constraints. Human error is thus treated as a constraint on the processes. A novel process methodology is developed to represent how technical systems, situational variables, external factors, and human error evolve over time in a process. The applicability is demonstrated with a dataset of case studies. In parallel, a new categorization of human errors is derived for this situation. The method permits cases to be examined individually, and also collectively for patterns. The results show that the initial landing approach is often the point of error initiation. For the specific case of landing accidents the methodology identifies that the failure causality was the lack of attention to establish procedure at the very start of descent, followed by persistent disregard of disconfirmatory evidence at the initial landing approach, especially under conditions of poor visibility. The paper makes a methodological contribution as well as suggesting new insights for how human error occurs at landing.

Keywords:

Aircraft pilot, production process, safety, sociotechnical interaction, systems engineering.