SBIR-STTR Award

As a company experienced in passive acoustic sensor and signal processing design and development, Analysis, Design & Diagnostics, Inc. (AD&D) understands that the ability to detect and classify sea-born vessels with a very high Probability of Correct Classification (Pcc) and a very low Probability of False Classification (Pfc) is an extremely challenging undertaking. The primary method currently used by AD&D is to employ optimal detectors designed to exploit traditional narrowband, broadband and active acoustic emissions from targets of interest. We recognize, however, that there is enormous potential in novel classification algorithms developed in other fields such as computational neuroscience, one of which is Cortical Processing. Under this SBIR we will investigate the feasibility of using Cortical Processing to classify sea-born vessels to a Taxon level.

Benefit:
The ability to automatically detect and classify submarines from surface ships with a very high probability of correct classification (Pcc) using in-node processing is critical to the success of the Advanced Undersea Weapons System (AUWS). If cortical processing can be shown to improve Pcc or assist classification using traditional methods it will provide the U.S. Navy with an unprecedented capability. Cortical Processing will help to ensure the performance capability is able to tolerate environment variability, multiple targets, and a diverse set of surface and submerged targets, as needed, for effective weapon performance.

Keywords:
Automated Detection, Automated Detection, Probability of Correct Classification, Probability of False Classification, Submarine, Cortical Processing, Automated Classification

The amazing ability of humans to recognize a wide variety of sounds with minimal exposure to the myriad of subtle nuances and possible variations of these sounds has driven intense research into modeling the neurobiological processes of hearing over the last 25 years. The field of study focused on processes in the early auditory system and primary auditory cortex has made substantial advancements over the last two decades as computers have finally begun to supply the necessary processing power to run realistic models. A model of hearing in the mammalian primary auditory cortex, called Cortical Processing, has been developed. We will expand upon the research performed in Phase I and apply Cortical Processing to classify threat targets to a taxon or class level.

Benefit:
The cortical processing being developed will go into the sensor node being developed for the AUWS program. The In-node processing for Low Power Target Detection and Classification being developed under this program will serve as the foundation for the Advanced Undersea Weapons System (AUWS) node. The AUWS program is a Future Naval Capability (FNC) being developed by the Office of Naval Research. The AUWS program will position and control sensor and weapons nodes to autonomously detect, classify, localize (DCL) and neutralize surface and subsurface threats in shallow/intermediate water depths.

Keywords:
Autonomous, AUWS, Automated Classification, low power, Cortical Processing, FNC, Submarine