Assessment of Human Interaction with Virtual Environment Training Technology

摘要

This research investigated the evidence of performance improvement of piloting skills while using an immersive virtual environment (IVE) versus a nonimmersive virtual environment (NIVE) to train instrument pilot skills. The general hypothesis tested was whether there is equal improvement for people trained under IVE and NIVE. Subjects were tested in IVE and NIVE flight scenarios using three flying tasks normal crosswind approach and landing (NCAL), go-around (GA), and constant speed during climbing and descending (CSCD). Data were analyzed for two measures errors and error rate, for four dependent variables: altitude control, heading control, airspeed control, and vertical airspeed control. Overall, results failed to demonstrate enhanced training effectiveness for an immersive VR training environment compared to a desktop (nonimmersive) environment. These results indicate that the cost tradeoff between the uses of IVE over NIVE are task dependent and influenced by the fidelity of training environments. The results obtained from the current experiment do justify some potential cost-saving advantage of IVE over NIVE on selected task. For example, NIVE seems to provide training advantages on error rate reduction on control of vertical airspeed and altitude under NCAL tasks. Similarly, IVE seems to offer training advantages of error rate reduction on airspeed control and heading control under NCAL, and heading and vertical airspeed controls under GA tasks. However, the fact that either IVE or NIVE provides an increase in piloting task performance in some tasks needs to be considered in any training investment decision.