Big Data Ml and AI for Combat ID and Combat Systems - Design, Demonstrate and Proof of Concept

Abstract

Project Summary: The project concerns two concepts for naval warfighting: common tactical air picture (CTAP) and combat identification (CID). The CTAP process collects, processes, and analyzes data from a vast network of sensors, platforms, and decision makers and provides situational awareness to air warfare decision makers in a kill chain process. The CID process locates and labels critical airborne objects (as friendly, hostile, or neutral) with high precision and efficiency based on a CTAP as part of the core kill chain process. The existing methods of CTAP and CID involve wide ranges of participating platforms, such as destroyers, cruisers, carriers, fighter attack aircraft and tactical airborne early warning aircraft; participating sensors, such as Radar, Forward Looking Infrared (FLIR), Identification Friend or Foe (IFF), Precision Participation Location Identifier (PPLI), and National Technical Means (NTM); and Participating Networks and Systems, such as the Aegis combat system, Cooperative Engagement Capability (CEC), and Link-16. CID is notoriously a very difficult function, often more art than science process is still very manual, and decision makers can experience cognitive overload, so analytics is just one aspect of CID. Challenges in the CID process include (1) extremely short time for fusion, decision-making, and targeting; (2) uncertain and/or missing data outside sensor (e.g. radar, radio) ranges; (3) manual decision-making; (4) heterogeneous data sources for fusion and decision-making; and (5) multiple decision-makers in the loop. The CTAP and CID problems can be seen as both Big Data and no data. On one hand, the data used for CID comes from a combination of massively cooperative and non-cooperative sensors, organic sensors, and non-sensor information (where, typically, each sensor collects certain attributes). The big CID data then needs to be fused over time and space since they are collected in a distributed fashion. On the other hand, adversaries often conceal and change their true intentions, therefore rare or no data are observed for analyzing anomalous and adversarial behavior. Therefore, the CTAP, CID, and kill chain problems are challenging application areas for analytic, machine learning (ML), or artificial intelligence (AI) methods. n the past, we demonstrated that Big Data techniques and deep analytics can provide potentially useful information assisting the CTAP and CID. Big Data (BD) techniques allow the distributed acquisition and fusion of disparate and crucial combat data in real-time or near real-time. This year, we made advancements as follows: - We investigated specifically the Defense Information Systems Agency (DISA)'s Big Data Platform (BDP). BDP is designed for real time processing of Big Data beginning at ingestion and ultimately presenting useful data visualizations that may alert decision makers for anomalies. - We analyzed the Automatic Dependent Surveillance-Broadcast (ADS-B). ADS-B functions with Global Positioning System (GPS) satellite, rather than radar technology, more accurately observe and track air traffic. Aircraft equipped with an ADS-B Out transmitter sends position, altitude, heading, that relays the information to air traffic controllers and other aircraft. Our work in progress applies various Big Data, and deep learning including ML/AI tools to predict if a flight is commercial or military.