Big Data Tools for High Speed Threat Detection and Classification

Interference from merchant ships and marine animals often obscure threat signals evident on sonar display surfaces. Experienced sonar operators can distinguish signals of interest (SOIs) such as torpedoes and rogue surface craft from clutter in beam-level displays, but this manual search leads to high operator workloads and unacceptably long detect-to-engage (DTE) timelines. To address this problem, Black River Systems will develop FRONT ROW; Fast Recognition Of Naval Threats for Reducing Operator Workload to be used for the Navy’s AN/SQQ-89 undersea warfare system. FRONT ROW will utilize a deep-learning approach to automatically detect and classify torpedo-like threats present in passive sonar data.  FRONT ROW is an operator aid that displays beam-level data of interest for the operator to review. The proposed approach will implement a multi-layered approach using a hybrid of Machine Learning (ML) techniques to maximize the usage of the limited tactical training data that is available.  FRONT ROW’s short frame-based approach will utilize a simpler network, require less training data, and readily implement into existing Navy systems. We will also work with the Navy to develop strategies for collecting future training data to increase the accuracy of the system.