Fly Away: Evaluating the Impact of Motion Fidelity on Optimized User Interface Design via Bayesian Optimization in Automated Urban Air Mobility Simulations

TL;DR

This study investigates how motion fidelity in VR simulations influences passenger responses and UI design for automated air taxis. Using a between-subject design (N=40) and Multi-Objective Bayesian Optimization, the authors optimize a 12-parameter UI across six objectives to identify Pareto-optimal designs under motion versus no-motion conditions. The results show that motion fidelity lowers trust, understanding, and acceptance, while differences in optimized UI features are largely subtle, suggesting a need for personalized interfaces. The findings offer practical guidelines for designing UAM interfaces and highlight motion fidelity as a crucial factor to consider in future UAM user studies and simulations.

Abstract

Automated Urban Air Mobility (UAM) can improve passenger transportation and reduce congestion, but its success depends on passenger trust. While initial research addresses passengers' information needs, questions remain about how to simulate air taxi flights and how these simulations impact users and interface requirements. We conducted a between-subjects study (N=40), examining the influence of motion fidelity in Virtual-Reality-simulated air taxi flights on user effects and interface design. Our study compared simulations with and without motion cues using a 3-Degrees-of-Freedom motion chair. Optimizing the interface design across six objectives, such as trust and mental demand, we used multi-objective Bayesian optimization to determine the most effective design trade-offs. Our results indicate that motion fidelity decreases users' trust, understanding, and acceptance, highlighting the need to consider motion fidelity in future UAM studies to approach realism. However, minimal evidence was found for differences or equality in the optimized interface designs, suggesting personalized interface designs.

Figures (7)

• Figure 1: Design parameters to be optimized. The figure shows the values of 0, 0.5 and 1 for each design parameter
• Figure 2: Optimization Process of the averaged normalized objectives during the 30 runs
• Figure 3: Exemplary optimized design parameters of Participant 40 (no Motion) and Participant 1 (with Motion). The concrete design parameters of these participants are plotted in \ref{['fig:design_parameters']}. Strong to extreme evidence of differences was found for the boundary box and chevron size of other air taxis. All other differences of design parameters, such the additional information at the display are due to personal preferences but no evidence was found to support this difference between groups
• Figure 4: Comparison of the design parameters for both groups, using a Bayesian t-test. For both groups, the IQR is plotted. Further, the exemplary Pareto front designs of Participant 1 (with Motion) and Participant 40 (no Motion) are depicted (see \ref{['fig:optimized_parameters']} for the visualization. The annotations are as follows defined by lee2013bayesian: "$<<<$" for extreme evidence for equality (BF < 0.01), "$<<$" for strong or very strong evidence for equality (BF < 0.1), "$<$" for moderate evidence for equality (BF < 0.3), "$=$" for inconclusive or anecdotal evidence (BF between 0.3 and 3), "$>$" for moderate evidence for difference (BF > 3), "$>>$" for strong or very strong evidence for difference (BF > 10), and "$>>>$" for extreme evidence for difference (BF > 100). The dotted line indicates the threshold for the boolean design parameters to either be smaller or greater than 0.5
• Figure 5: Rating for the subjective questionnaires comparing both groups: no Motion and with Motion of all Pareto optimal values. The Bayes factor shows the trend towards equality (<1) and difference (>1)