A VR-Based Multi-Camera Learning System Integrating Drone for Spatio-Temporal Decision-Making in Cultural Live Production

Drone-assisted visual inputs introduce significant complexity in operator decision-making within multi-camera live production systems, particularly due to their dynamic and time-dependent characteristics. While prior studies mainly focus on drone autonomy or VR-based training, the impact of dynamic visual variability on operator decision-making remains underexplored. This work proposes a VR-based multicamera learning system integrating three static cameras and one drone operating along orbital trajectories with waypoint-based idle states. A controlled experiment involving 10 participants performing tasks (T1-T10) demonstrates that all tasks were successfully completed, while completion time analysis reveals a clear distinction between static (13 s) and drone-based tasks (35-40 s, with peaks > 100 s), and continuous switching (260 s). These results indicate that drone-induced temporal variability introduces spatio-temporal decision-making complexity in multi-camera learning environments. This reflects how dynamic drone visuals influence time-critical decision-making in cultural live production. Future work will explore learning effectiveness, multi-drone coordination, enhanced realism, and AI-assisted decision-making.