Machine Learning (ML) to Develop Capabilities to Track AIS Ships Worldwide and Detect Anomalous Behavior to Impact Mission Success

Phase II will continue to develop two potentially significant breakthrough capabilities commenced by Jove Sciences, Inc. (JOVE) in Phase I:  The work performed in Tasks 1, 2, and 3 of this Phase II proposal is to continue the development of JOVE’s Dynamic AIS Ship Density/Lane  (DASD/L) Model (with automatic) AIS ship anomaly detection integrated), and will fill a capability gap that has existed for decades. DADS/L will improve the Situational Awareness of the Cooperative Engagement Capability (CEC), AEGIS Combat System (ACS), and Ship Self Defense System (SSDS) when transitioned . The DASD/L Model will significantly improve the performance of NAVSEA’s Undersea Systems enhancing ASW/USW acoustic performance for three specific applications to be discussed in this proposal. DASD/L will also significantly improve PMS-485’s SURTASS, Fixed Bottom Array, and Special Array acoustic performance for the same three applications. Several other DoD Programs of Record (PoRs),. Such as Data Control Ground Systems – Navy (DCGS-N), MQ-9 Maritime Wide Area Surveillance (MWAS), and many more. DASD/L will improve the performance of Non DOD programs, such as USCG Situational Awareness and acoustic detection programs, commercial ship routers, the NATO anti Pirate Task Force, countries fighting illegal fishing, and many more non DoD applications. The work performed in Task 4 will integrate DASD/L into the TRL-8 Advanced Correlator –Navy (ACOR-N) data fusion processor currently being evaluated for transition to the COCOMs by COMPACFLT N2/PACOM JIOC to improve detection, tracking, classification, and identification (DTC&I) of adversary naval combatants (surface ships and submarines) and EMCON Silent or “dark” contacts. This technology can be tested in Phase II during a Navy Exercise of opportunity to assess an OPAREA to which a Battle Group is heading hours, days, or weeks before the BG arrives in the OPAREA. This would benefit the CEC, ACS, and SSDS combat systems, as well as other NAVSEA PoRs. Tasks 1 through 4 are: (1) Track AIS ships worldwide using unclassified commercial and classified non-commercial AIS (NCA) sources, (2) ingest AIS data to develop a real time DASD/L Model that will improve Situational Awareness for the Battle Group (BG). DASD/L will run on a low cost, small footprint, high performance shared memory (RAM-Dense) ACOR-N AIS Processor (AAP) processor to display AIS ship tracks and their metadata for large Areas of Interest (AoIs). The Advanced Correlator-Navy (ACOR-N) software will be installed on the unclassified RAM-Dense ACOR-N AIS Processor (AAP) to track AIS ships worldwide in real time. 3) Use Machine Learning to detect anomalous behavior from AIS ships, and integrate into DASD/L and (4) Integrate DASD/L with ACOR-N being evaluated for transition to the Fleet by COMPACFLT N2/PACOM JIOC. The objective of this task is to transition the capability to DTC&I adversary naval combatants to CEC, ACS, SSDS and other NAVSEA PoRs.