Salvo Management Using Artificial Intelligence

RT&L FOCUS AREA(S): Artificial Intelligence/ Machine Learning TECHNOLOGY AREA(S): Battlespace The technology within this topic is restricted under the International Traffic in Arms Regulation (ITAR), 22 CFR Parts 120-130, which controls the export and import of defense-related material and services, including export of sensitive technical data, or the Export Administration Regulation (EAR), 15 CFR Parts 730-774, which controls dual use items. Offerors must disclose any proposed use of foreign nationals (FNs), their country(ies) of origin, the type of visa or work permit possessed, and the statement of work (SOW) tasks intended for accomplishment by the FN(s) in accordance with section 3.5 of the Announcement. Offerors are advised foreign nationals proposed to perform on this topic may be restricted due to the technical data under US Export Control Laws. OBJECTIVE: Design a technology to serve as a battle management aid, in particular inform salvo management. DESCRIPTION: Although system-wide battle management handles tasking and assignments to sensors and weapon systems, each weapon system then must assign weapons from its inventory and request tasking. As new versions of weapons come online, mixed field inventory assignment and management can become difficult to optimize in light of fundamental uncertainties. Not only may the salvo composition need to be constructed from more than one weapon system, but modest but significant differences in ability and reliability should also be considered. Additionally, salvo timing and management after launch in light of potential battlespace updates should inform the battle management. Input will consist of threat state from the centralized battle manager, but considerations need to be made concerning the type of threat system being engaged in terms of fundamental capabilities matching interceptor capabilities to threat capabilities. With fixed inventory sizes, and uncertain threat inventories, the goal is to define how highly capable and reliable weapons should be partitioned to address the threat. Salvo composition, timing and management, before as well as after launch should be optimized with respect to available, and dynamic, threat information. PHASE I: Define a test set of two major interceptor versions, with modest but significant variations in capability and reliability among the sets. Define a small set of threats that represent variety in maneuverability, countermeasure configuration, and warhead number. Define an artificial intelligence system, e.g. reinforcement learning, that produces a small set of battle plans from which the warfighter can select. Demonstrate how battlespace updates effect the plan. Develop prototype code with an algorithm description document with test results. Include particular input information that would benefit the operation. PHASE II: Using high fidelity weapon and threat information, including throughout the timeline, upgrade the fidelity of the system to handle realistic data. Ensure the developed plan is consistent with commander intent as well as defined tactics, techniques and procedures. Define how the system would respond to data updates, and define requests for more information (sensor tasking.) Submit a list of preferred system information to support optimal engagement. Develop and deliver code with an algorithm description document with test results and analysis of government delivered scenario data as well as own data. Demonstrate through analysis robustness of the system. PHASE III DUAL USE APPLICATIONS: This task involves system automation and could learn from, as well as inform, automated systems such as vehicle driving, plant operation, and inventory management for emergency response. REFERENCES: 1. Alpha C2-An Intelligent Air Defense Commander Independent of Human Decision-Making, Qiang Fu et al, IEEE Access, May, 2020, DOI 10.1109/Access.2020.2993459. ; 2. An Introduction to Deep Reinforcement Learning, Vincent Francois-Lavet et al, December 2018, ArXiv:1811.12560v2. ; 3. AI-enabled wargaming in the military decision making process, Peter Schwartz et al, Proc. SPIE 11413, Artificial Intelligence and Machine Learning for Multi-Domain Operations Applications II, 11413H (April 2020) doi 1117/12.2560494. ; 4. PEORL: Integrating Symbolic Planning and Hierarchical Reinforcement Learning for Robust Decision-Making, Fangkai Yang et al, Proceedings of the twenty-Seventh International Joint Conference on Artificial Intelligence (IJCAI 18).