Real-time autonomous multiple sound sources mapping for unmanned aerial vehicle embedded applications

Acoustic imaging is a powerful tool for noise source localization, and can be used to detect faults and gas leaks in industrial infrastructures. Over the past decade, unmanned aerial vehicles (UAVs) have been used in several domains for performing rapid and reliable monitoring tasks using a variety of sensors and methods. However, the noise generated mainly generated by rotors in UAVs poses a significant challenge to acoustic diagnostics. In this work, an experimental platform designed to evaluate the performance of autonomous sound-guided UAVs is presented. Therefore, in this way real-time tests are carried out to illustrate the capabilities of the platform. The proposed acoustic platform localizes efficiently a sound source despite significant UAV-generated noise and interference from a secondary, stronger sound source.