Understanding and Mimicking the Fly's Directional Hearing: Modeling, Sensor Development, and Experimental Studies

Thumbnail Image


Publication or External Link







Microphone arrays have been widely used in sound source localization for many applications. In order to locate the sound in a discernible manner, the separation between microphones needs to be greater than a critical distance, which poses a fundamental constraint for the miniaturization of directional microphones. In nature, animal hearing organs are also governed by the size constraint; the smaller the organ size, the smaller the available directional cues for directional hearing. However, with an auditory organ separation of only 520 µm, the fly Ormia ochracea is found to exhibit remarkable ability to pinpoint its host cricket at 5 kHz. The key to this fly's phenomenal directional hearing ability is believed to be the mechanical coupling between the eardrums. This innovative solution can inspire one to find alternative approaches to tackle the challenge of developing miniature directional microphones.

The overall goal of this dissertation work is to unravel the underlying physics of the fly ear hearing mechanisms, and to apply this understanding to develop and study novel bio-inspired miniature directional microphones. First, through mechanics and optimization analysis, a fundamental biological conclusion is reached: the fly ear can be viewed as a nature-designed optimal structure that is endowed with the dual optimality characteristic of maximum average directional sensitivity and minimum nonlinearity, at its working frequency of 5 kHz. It is shown that this dual optimality characteristic is only achievable when the right mechanical coupling between the eardrums is used (i.e., proper contributions from both rocking and bending modes are used). More importantly, it is further revealed that the dual optimality characteristic of the fly ear is replicable in a synthetic device, whose structural parameters can be tailored to work at any chosen frequency. Next, a novel bio-inspired directional microphone with mechanically coupled diaphragms is designed to capture the essential dynamics of the fly ear. To study the performance of this design, a novel continuum mechanics model is developed, which features two coupling modules, one for the mechanical coupling of the two diaphragms through a beam and the other for each diaphragm coupled through an air gap. Parametric studies are carried out to explore how the key normalized parameters affect the performance of this directional microphone. Finally, this mechanics model is used to guide the development of a large-scale microphone and a fly-ear sized microphone, both of which are experimentally studied by using a low-coherence fiber optic interferometric detection system. With the large-scale sensor, the importance of using proper contribution from both rocking and bending modes is validated. The fly-ear sized sensor is demonstrated to achieve the dual optimality characteristic at 8 kHz with a ten-fold amplification in the directional sensitivity, which is equivalent to that obtainable from a conventional microphone pair that is ten times larger in size. To best use this sensor for sound source localization, a robotic platform with a control scheme inspired by the fly's localization/lateralization scheme is developed, with which a localization accuracy of better than ±2 degrees (the same as the fly ear) is demonstrated in an indoor lab environment.

This dissertation work provides a quantitative and mechanistic explanation for the fly's sound localization ability for the first time, and it provides a framework for the development of fly-ear inspired acoustic sensors that will impact many fronts.