Motivation

Small unmanned aerial vehicles (UAVs) promise to revolutionize inventory management and materials handling in a variety of industries, a multi-billion-dollar market potential. However, their perceived noise and safety issues present an important barrier to delivering on this promise. Applications of UAVs in construction & maintenance, warehouse management, materials handling in factories, and localized freight transport often require UAVs to fly in close proximity to human agents (co-workers/end-users), over substantial periods. A majority of these applications prefer multi-rotor UAVs due to their low cost, maneuverability, and small spatial footprint. Anybody who has operated these small UAVs will attest to the fact that their noise signature (mainly propeller noise) is far from a pleasure to work with. While a few investigations of UAV acoustics exist, the critical understanding of 1) how humans perceive and respond to UAV noise, and 2) the nature of human–UAV interactions in these unstructured environments, is practically non-existent. Moreover, the few existing solutions to mitigating UAV noise impact often compromise their aerodynamic/structural performance.

Research Objective

Our long-term goal is to analyze and effectively regulate the “impact on surroundings” (e.g., noise and safety) of small UAVs, which otherwise present roadblocks to their integration with and adoption by humans in co-robotic applications. Our immediate objective is to reduce the acoustic signature of multi-rotor UAVs (to within human comfort thresholds), at a minimal compromise of the UAV’s performance, for application domains involving varying levels of interaction with humans.

Research Approach

To address these technical gaps, and thereby facilitate safer and ergonomic integration of UAVs in co-robotic work environments, this project brings together experts from vibration/acoustics, design-optimization/autonomous-systems, audiology, and human-cognition. Our approach is to optimally re-design the protective rotor-envelopes using optimized anechoic material and geometry for noise attenuation. This project combines physical and simulated experiments to characterize the acoustic transmission properties, model the UAV acoustic environment, conduct human-subject experiments to construct cognitive models of human perception of UAV noise, develop a virtual robotic simulation environment to study human–UAV interactions, and optimize the design.

Expected Outcomes

• Identification of optimal noise treatments for multi-rotor UAVs;
• Data-driven models of human perception of UAV noise;
• Virtual robotic environment for simulating small UAV/human interactions.

Acknowledgement of Funding

This project is funded by the 2016-2017 SMART Exploratory Funding program on “Reasoning and Interoperability in Unstructured Environments” at U Buffalo.

Preliminary Findings

Indoor measurement spectrogram for close range UAV noise	Frequency spectrum for indoor preliminary measurements

Virtual warehouse environment: simulating human-UAV interactions