Trusted Artificial Intelligence Network for Communications and Radar

To address the Air Force need to significantly enhance rogue signal detection and classification using AI, Intellisense Systems, Inc. (Intellisense) proposes in Phase II, to advance the development of the Trusted Artificial Intelligence Network for Communications and Radar (TAINCOR) system based on a novel hardened and robust deep neural network (DNN) architecture which can be deployed on a low-size, -weight, and -power (SWaP) system and can process RF data at >200 Hz (99% accuracy; (4) demonstrate that TAINCOR algorithms can be run OTA in real time in a man-portable prototype. In Phase II, Intellisense plans to evaluate the proposed solution against the objectives and measurable key results defined in the Phase I feasibility study. Intellisense will apply the lessons learned investigating the art of the possible in Phase I to mature the technology to directly address Air Force requirements and objectives. Specifically, during the Phase II effort Intellisense will develop, install, integrate, and demonstrate the TAINCOR prototype system and apply this prototype to operational Air Force problems by: (1) training and validating TAINCOR algorithms on data from relevant Air Force systems including radar systems and tactical radios; (2) packaging and maturing the TAINCOR code base to be deployed at scale for both algorithm training and inference; (3) executing on a clear transition path for TAINCOR to be integrated with existing and future Air Force systems. The Phase II effort will prepare TAINCOR for transition in Phase III to Air Force and DoD customers, addressing their needs for signal detection and classifications for radar and RF communication applications.