An Experimental Comparison of the Effectiveness of Various Levels of Simulator Fidelity on Ab Initio Pilot Training
Loading...
Date
2023-05-17
Authors
Paul, Naomi
Journal Title
Journal ISSN
Volume Title
Publisher
University of Waterloo
Abstract
Flight simulators have long been used to great levels of success with studies demonstrating that pilots
who first train using a simulator reach acceptable levels of competence in less time than pilots who start
in the cockpit of an actual aircraft. However, the high costs of simulators are not economically
sustainable for the aviation industry and is one of the barriers to entry for those of diverse backgrounds.
To address these challenges, there is hope that medium- and low- fidelity simulators including virtual
reality head-mounted displays and desktop simulators may be approved for ab initio pilot training as a
supplement to real flight training. Ab initio pilot training is pilot training “from the start”, training pilots
with no experience until achievement of their commercial pilot’s license.
Natural tactile interaction is traditionally associated with effective pilot training tasks. Despite the lack
of natural tactile interaction in virtual reality, some tasks, such as procedural tasks, simply require
exposure to the aircraft environment. This exposure may be achieved through 3-dimensional simulated
models of the aircraft in virtual reality.
My research proposes a between-subjects experiment, with 10 participants per group (30 participants
total), to quantitatively address this question by analyzing the improvement of pilots completing simple
procedural and aircraft handling tasks using either a high-fidelity flight training device, medium-fidelity
desktop simulator, or low-fidelity virtual reality simulator. All participants are student pilots at the
University of Waterloo with under 20 hours of flight experience, participated in 5 consecutive days of
the study, with training effectiveness evaluated through the improvement in performance from Day 1
to Day 5. This research collected objective flight performance assessed by both a flight instructor and
through flight data, as well as subjective rating data regarding mental workload, stress level, and
experience of simulator sickness. One-way and repeated measures ANOVA analyses were used to
analyze the improvement in participants’ performance for ab initio procedural and aircraft handling
tasks, comparing three between-subject conditions of using a high-fidelity flight training device,
medium-fidelity desktop simulator, or low-fidelity VR simulator.
It was found that virtual reality and desktop simulators are as effective as high-fidelity flight training
devices for pilot training of procedural tasks, without increasing the risk of experiencing simulator
sickness. However, for aircraft handling tasks, participants training using virtual reality or desktop
simulators did not improve to the same degree as those training on the high-fidelity flight training
device. This provides evidence for the argument towards approving the strategic allocation of virtual reality for pilot training in combination with existing simulator methods for handling and other tasks,
leading to potential cost savings which make training less expensive and more accessible to future
pilots.
Description
Keywords
aviation, simulator fidelity, virtual reality