Bionic modeling for knowledge-based guidance in automated underwater vehicles

This paper describes the auditory localization behavior of the barn owl as an example of how bionic systems may be modeled for application to underwater vehicles. The nocturnal barn owl (Tyto alba) is unique in its passive localization ability, having vertically displaced ears with different peak sensitivities. It has an audible range of .1 kHz to 12 kHz and a sensitivity of -18 dB SPL. Despite its small size, it can get bearing, azimuth and elevation without leaving its perch. It can localize pure tones almost as well as noise. Based on currently available data on the owl´s sensory and structural capacities, a model of its localization behaviors is developed to identify the decisions and information needed in localizing prey. The informational requirements at each stage of its mission are projected from environmental and behavioral data. Guidance is shown as an automatic system based on nulling mechanisms; after the identification of the prey, decision making is a go-no go election. Sensory information is continuously updated by the owl while in motion until the final attack phase. The use of intelligence is discussed. This model is offered as an illustration of how bionic systems may serve to identify the kinds of information required by an autonomous underwater vehicle (AUV). It also suggests a model of sensors and decisions whereby localization occurs. Application for AUV tasks are suggested.