Methodology for assessing dependencies between factors influencing airline pilot performance reliability: A case of taxiing tasks

Pilot performance reliability is critical to civil air transportation safety. Internal and external factors influencing pilot performance reliability are important elements to estimate human error probability. However, dependencies between these factors, including state dependencies and effect dependencies, are always ignored or only partially considered in most human reliability analysis (HRA) method. This paper attempts to develop a methodology incorporating both state and effect dependencies between contributory factors into HRA for airline pilot tasks based on fuzzy logic and the Cognitive Reliability and Error Analysis Method (CREAM). Anchor points were determined for each common performance condition (CPC) state levels through airline pilot elicitation process and then corresponding state and effect fuzzy sets were constructed for each CPC. The CPC dependency fuzzy inference rules were constructed and transformed with reference to the CPC dependency assessing rule in CREAM. And a cognitive failure probability (CFP) calculation method based on the final CPC effect fuzzy sets and the extended CREAM was proposed. To illustrate the use of the proposed methodology, a case study about commercial airline pilot taxiing task was given. The results demonstrate the validity of the proposed methodology and its capability of representing the ambiguity and fuzziness of expert judgement.