A sensorless wind-aware multimodal feedback framework for human-drone interaction in urban environments

Abstract

Unmanned aerial vehicles (UAVs) are increasingly applied to urban delivery, infrastructure inspection, and emergency response, yet operation in dense built environments remains difficult because building-induced turbulence introduces invisible aerodynamic disturbances that impair control performance, increase cognitive workload, and weaken trust in autonomy. This study presents a sensorless wind-aware multimodal feedback framework for human-drone interaction in urban environments. The framework combines a Unity-based urban simulation environment, a sensorless wind estimation and wind-field reconstruction method, an immersive visual interface, torso-mounted haptic feedback, and an adaptive auditory warning channel. Instead of relying on a dedicated wind sensor, the proposed method derives a simulation-grounded estimate of local wind conditions from the drone’s dynamic response and uses this inferred wind representation to drive multimodal feedback for environmental awareness. A controlled experiment with 30 participants evaluated four feedback conditions: Control, Visual, Visual + Haptic, and Visual + Haptic + Audio. Results indicate that Visual + Haptic feedback provided the most effective trade-off among navigation efficiency, safety, workload, and trust, while the tri-modal condition further improved task efficiency at the cost of higher cognitive demand. These results suggest that integrating sensorless environmental estimation with multimodal interface design is a promising direction for improving human-centered UAV operation in complex urban environments.