| dc.description.abstract | The advancements in vertical take-off and landing (VTOL) aircraft have rapidly increased in the past few years, and there are working prototypes with human pilots already tested. With the current state of progress, provided the battery technology and automation level catch up to the required standard, VTOL cars could come to the market soon. However, the regulatory bodies are still working on the policies for automated cars and are far from their end goals. Given the scenario, it would be extremely beneficial to have empirical data to inform engineers, designers, and policymakers about what could be an intuitive controller from the existing hardware widely available in the market. This study investigates the ease of use of flying a VTOL aircraft between three of the most widely used controllers, namely the driving wheelset, drone radio controller, and joystick, backed up by performance data, EEG data, and the NASA-TLX survey.
A case study was conducted for 30 participants with a G2/G license aged 18 to 64 years. Each participant tried all three controllers in a randomized order to fly through a standard track in Virtual Reality (VR). Performance data and EEG signals were recorded in real-time, and a NASA-TLX survey was conducted after the user tried each controller. After they tried all three controllers, an overall survey was given to rank the controllers from the most preferred to the least preferred and to reason their choices. Finally, the users were asked to fly through the track with the driving wheelset one more time, where the randomized wind was introduced to see if that could affect their performance and overall workload.
The results of the experiment are compared among the three controllers using the three different types of datasets. The result shows that the joystick controller was the most preferred controller among the three controllers, backed up by the user survey, EEG data and performance data. If we compare just the statistical performance and not the surveys, the result is not significant enough to be reported. It is also found that weather conditions can significantly affect performance for the users.
For future work, the experiment should be carried out with varying weather conditions. Also, to improve among the existing controllers, the joystick could be modified to include pedals for rotation instead of the twist, reducing the chances of unintentional input from the user. | en |
