SRK as a framework for the development of training for effective interaction with multi-level automation

This paper examines the development of training for effective interaction with multi-level automation, i.e., a system that switches between functions and roles corresponding to different levels of automation. These interactions need to span not only the nominal procedures and skills expected of the operator, but also effective reasoning about when and whether the automation should be employed. This suggests framing the operator's tasks using Rasmussen's classical categorization of human behavior as skill based, rule based, or knowledge based (SRK), thereby providing appropriate insights into appropriate training objectives and methods. This paper uses the aircraft traffic alert and collision avoidance system as a case study, demonstrating the application of the SRK framework to develop pilot training. Comparison of pilot behavior with and without this modified training highlights the training's ability to improve interaction with the automation, leading to recommendations for broader application of the SRK framework in training development.