Using Simulation To Identify And Measure Pilot Skills For Automated Aircraft
Skill identification and measurement is a primary requirement for developing effective training and evaluation programs. For many years the airlines and other aviation organizations have been working to develop effective automation training programs without objective methods for observing and measuring the associated skills.
Pilot performance with automation is often difficult to observe and it does not fit into the traditional definition of skill. Our challenge was to usable definitions of the skills associated with using automation and to develop objective methods to identify and measure these skills.
Approach and Solution
We first used literature related to skill definition and identification to develop definitions of the general skills associated with the use of flight deck automation. The next step was to find methods to observe those skills. All airlines use flight simulators in their training programs. We chose to evaluate whether digital data available from the simulators could be used to measure the pilots' performance in using the automation. We concluded that to effectively identify simulator data patterns that reflect pilot performance using automation there must be a way to identify exactly when the pilots make use of the automation. This will require scenarios and their conduct in the training sessions to be standardized to allow the analyst to predict when pilots will be using the automation and the context in which they use it. We were not successful in performing this type of analysis with the data that we were using for this purpose. Future studies will be able to find more effective ways of observing automation skill performance. The final report is available. More info about Final Report...
Naval Air Warfare Center Training Systems Division and Federal Aviation Administration (FAA)
George Mason University (GMU), University of Central Florida (UCF), and Alaska Airlines