Modeling Pilot Flight Performance on Pre-flight and Take-off Tasks with A Cognitive Architecture

Models of cognitive architecture can be used to simulate and forecast human performance in complicated human-machine systems. This thesis demonstrates a pilot model capable of performing and simulating pre-flight preparation and take-off duties. The model was developed using the Queueing Network-Adaptive Control of Thought-Rational (QN-ACTR) cognitive architecture and can be connected to flight simulators like X-Plane to create various data types such as performance metrics and mental workload estimates. Declarative knowledge chunks, production rules, and a collection of parameters all contribute to the model’s output. A human experiment involving pre-flight and take-off tasks was conducted to acquire the data required for the model’s development and validation. The model can generate flight operation behaviors that are comparable to that of human pilots, demonstrating the potential of cognitive architecture-based approaches for supporting pilot training and performance evaluation in aviation.

Master’s Thesis, University of Waterloo