Abstract
Current advanced commercial transport aircraft rely on Flight Management Systems. The increasing complexity of these systems has caused an increase in errors in interaction with aircraft automation. Previous research has focussed on identification of the elements of automation (mode structure, consistency, command languages and other) which may lead to faulty human-automation interactions. These approaches require the complex system to have underlying structure in an available and communicable form.
In contrast, this paper discusses a more easily testable “end-to-end” metric which can be used independent of knowledge of this structure. The concept of predictability is presented as a candidate metric of the complexity of automation and is defined as a measure of how well an operator can anticipate what the system will do at some point in the future. In essence, this is a measure of the complement of how often a system will “surprise” an operator by acting in an unanticipated manner. The goal of this work is to identify areas of Flight Management Systems which have a strong impact on predictability to provide guidance in future designs and in current pilot training.
Get full access to this article
View all access options for this article.