Unmanned aerial vehicle-based sounding of subsurface concrete defects

Abstract

A sounding technique that uses an unmanned aerial vehicle (UAV) equipped with two microphones can detect subsurface concrete defects. Use of flexural vibration frequency as a basis for defect depth estimation is evaluated. While many non-destructive tests for concrete can detect depth, current UAV-based inspection methods like optical and thermal imaging are typically limited to two-dimensional subsurface defect information. Acoustic signals from sounding and UAV noise are known to exist in similar frequency ranges. Accordingly, three noise reduction measures for this sounding technique are assessed. Given adequate distance between the microphones and UAV, a two microphone signal subtraction technique is slightly effective for some noise, but a spectral noise gating procedure is shown to substantially decrease noise in the frequency range of interest.