Summaries - Office of Research & Innovation
Research Summaries
Back Improved Methods of Combat Identification: A Data-centric Approach
Fiscal Year | 2023 |
Division | Research & Sponsored Programs |
Department | Naval Research Program |
Investigator(s) | Das, Arijit |
Sponsor | NPS Naval Research Program (Navy) |
Summary | Complex sensor networks are an enormously large and expensive set of systems, software, sensors, and emitters that are used to generate data contributing to battlespace awareness. Many sensor network configurations can collect so much data that a 'Data Rich, Information Poor' (DRIP) situation results. The desired solution is the coordination of high value units (HVUs) in support of actions such as identification and targeting. Timely detection of unknown signals is currently inadequate to maintain situational awareness at a tactical level. Few analysts have the experience and data accesses to make good use of the available data. Those who do are not closely placed to support the tactical and operational levels of warfighting. Automating the workflow and processes of well-seasoned analysts using new AI and modeling technologies would enable improved and more timely extraction of essential information from the data and support better situational awareness and tactical decision making. The motivation for analysis stems from the significant pressure that is usually applied to rapid, effective, and accurate decision making. The benefits of leveraging these methodologies include greater situational awareness with automation tools distilled to the tactical operator level with fully informed data sources released at the appropriate classification level. In this research, we propose utilizing multiple approaches with a focus on graph theory methodologies, such as machine learning, probabilistic modeling, and clustering techniques, that serve to formulate an experiment and execution to discover, hypothesize, and demonstrate through simulations a self-learning knowledge graph environment to explore the effectiveness of automatic knowledge creation for combat identification. These methodologies are an emerging enabler for rapid signal associations required for targeting and situational awareness. Combat identification procedures and planning can benefit from leveraging the power of graph machine learning and probabilistic modeling for automatic entity creation for both unstructured and structured data sets. |
Keywords | AI/ML, deception, data fusion, confusion matrix, classification, AIS |
Publications | Publications, theses (not shown) and data repositories will be added to the portal record when information is available in FAIRS and brought back to the portal |
Data | Publications, theses (not shown) and data repositories will be added to the portal record when information is available in FAIRS and brought back to the portal |