Zhang, W., Feltner, D., Shirley, J., Kaber, D. B. & Swangnetr, M. (in review). Enhancement and application of an unmanned aerial vehicle supervisory control interface evaluation technique: Modified GEDIS-UAV. Submitted to Trans. on Human-Robot Interaction (2/11/18)
Supplemental information to the paper
Intro:
The Modified-GEDIS UAV (M-GEDIS-UAV) is an interface evaluation tool intended for supervisory control unmanned aerial vehicle (UAV) interfaces. The tool development and application instruction are documented in “Enhancement and Application of an Unmanned Aerial Vehicle Supervisory Control Interface Evaluation Technique: Modified GEDIS-UAV” (currently in review with ACM Transactions in Human-Robot Interaction). The tool is subject to further revision based on feedback from the journal reviewers and/or its users.
Application instruction:
The M-GEDIS-UAV has a checklist structure. An interface is evaluated based on 9 indicators, which describes various interface design features. Each indicator can be further assessed with sub-indicators, which are design characteristics of the design features. For ease of application, each indicator is presented with a separate worksheet in the attached file.
In order to evaluate an UAV interface, an analyst should compare actual interface design characteristics with related criteria and determines whether each criterion is satisfied or not. That is, the analyst assigns a binary score (“1” if satisfied, “0” if not) for each criterion. If the characteristic described by a criterion is not related to the interface under evaluation, the analyst assigns a response of not applicable (“NA”) and the criteria is not considered in scoring of the interface design. The score for each sub-indicator can be determined by averaging the scores for the criteria within the sub-indicator checklist. On this basis, an indicator score becomes the average of all relevant sub-indicator scores. That is, for each desired design feature, the degree of interface conformance (a percentage) with guidelines for specific characteristics of the feature/component is calculated. Subsequently, the GEI is calculated as the average percent satisfaction of desired design features for an interface under evaluation. The GEI is simply obtained by averaging all the indicator “scores”. The GEI can vary between 0 and 1, with 1 representing an optimal design.