SBIR-STTR Award

To address the Army's need for an Artificial-Intelligence-based Automated Fire Control System, Physical Optics Corporation (POC) proposes to develop a new Aggressive AI Self-Learning Defense Network (ASSAILNET) technology. Specifically, the innovation in the network architecture, the network node interface design, and the network training strategy will enable each AI node in the network to interface seamlessly with a wide range of attached sensors, to share its sensor data with other nodes, to correlate shared data from other nodes with its own data, and then, together with other nodes, make coordinated collective decisions and, as a whole network, stage concerted effective actions to mitigate identified threats. As a result, this technology offers a solution to provide fire control awareness to weapons and other systems, which directly addresses the Army requirements for an AI-based Automated Fire Control System (AFCS). In Phase I, POC will demonstrate the feasibility of this technology and provide a detailed engineering design study that explains how the new technology will lead to an AI system that operates the AFCS. In Phase II, POC plans to complete the software/algorithm, update the source code, and demonstrate an AI prototype in the laboratory.Deep Learning,reinforcement learning,Automated Fire Control System,artificial intelligence,self-learning network,sensor interface

To address the Army’s need for an artificial-intelligence (AI)-based Automated Fire Control System (AFCS), Physical Optics Corporation (POC) proposes to advance the new Aggressive AI Self-Learning Defense Network (ASSAILNET) technology proven feasible in Phase I. The innovation in ASSAILNET enables each AI node in the network to interface seamlessly with a wide range of attached sensors, weapons, and other AI nodes; to share its sensor data with all nodes; to correlate shared data from other nodes with its own data; and then, together with other nodes, make coordinated collective decisions and, as a whole network, stage concerted effective actions to mitigate identified threats. As a result, this technology offers a solution to provide fire control awareness to weapons, sensors, and other systems, which directly addresses the Army requirements for an AI-based AFCS. In Phase I, POC successfully met all the objectives and conclusively demonstrated the feasibility of ASSAILNET. Key technical risks associated with the technology were reduced by designing, developing, and testing a limited-scale prototype in a controlled virtual environment. As a result, Phase II development is now a straightforward process of engineering, optimization, and risk-reduction. In Phase II, POC plans to mature this technology and conduct a field test.