AN EXPERIMENT TO MEASURE SIMULATOR-BASED AIRPLANE UPSET RECOVERY TRAINING TRANSFER: PRELIMINARY RESULTS OF PHASE TWO FLIGHT TESTING

摘要

Airplane upsets leading to loss of control (LOC) are a leading cause of air transport as well as general aviation fatal accidents. As a consequence, air transport pilots receive ab initio and recurrent upset recovery training in Level D simulators. However, little evidence exists to support the hypothesis that simulator training improves a pilot's ability to recover an actual airplane from a serious upset. We report on an FAA funded research experiment to evaluate the effectiveness of low-cost simulator-based upset recovery training. A 2×4 mixed factorial design allows us to compare in-flight upset recovery performance of trained and control group pilot participants over a variety of dependent variables. During Phase One flight testing, trained pilot outperformed control group pilots in areas such as time to recover from an upset, proper throttle management, roll control input, and appropriate application of G forces. However, there was no statistical difference between trained and control group pilots in the most important performance parameter-altitude loss during upset recovery maneuvering. In Phase Two flight testing, completed in April 2008, a new group of trained pilots outperformed new control group pilots in altitude loss as well. Nevertheless, trained pilot performance with respect to altitude loss was not close to optimal. We present and interpret preliminary results of Phase Two testing and discuss their implications with respect to air transport upset recovery training programs.