Innovative Uses of Artificial Intelligence and Machine Learning in Scenario Planning and Design

RT&L FOCUS AREA(S): Autonomy; Artificial Intelligence/ Machine Learning TECHNOLOGY AREA(S): Information Systems OBJECTIVE: Develop innovative scenario planning technologies utilizing novel Artificial Intelligence (AI) and Machine Learning (ML) methods to augment and assist Subject Matter Experts (SMEs) developing test cases within a missile defense Modeling and Simulation (M&S) architecture. DESCRIPTION: This topic seeks innovative AI/ML techniques that can provide SMEs with the information they need to quickly identify text-matrix gaps and build new and novel test cases to exercise the missile defense architecture. Current missile defense scenario planning involves scenario designers manually producing test cases to be executed. These scenarios are based on various factors including system requirements, test objectives, venue capabilities, and past data analysis and can take upwards of two years to develop and run. Scenarios include a physical representation of the threat (Red Force), Blue Force, and environments. This is a time-intensive manual process where an integrated product team coordinates assessment requirements that are fed into a Rapid Scenario Prototype (RaSP) team that creates scenario requirements to meet engagement and test objective goals with the least number of test cases and scenarios possible. The scenario requirements are then incorporated into a test case description document that is used by the community to create the scenarios to be ran. This scenario generation process is a cumbersome, sequential, resource and time intensive effort requiring work by every component of the integrated simulation team plus supporting organizations, which impacts the responsiveness of the M&S to government needs. With the amount of data exponentially increasing, faster and smarter scenario generation methods based on requirements and past data sets are desired. These technologies should enhance the credibility of the integrated M&S; shorten integration time enabling the government to gain efficiencies, reduce event schedules, and produce greater quantities of credible decision quality data. PHASE I: Design and develop improved solutions, methods, and concepts for applying ML and AI in scenario generation. The solutions should capture the key areas where new development is needed, suggest appropriate methods and technologies to minimize time intensive processes, and incorporate new technologies researched during design development. Define the architecture and data structures required to support the missile defense M&S enterprise. PHASE II: Complete a detailed prototype design incorporating government performance requirements. The contractor will coordinate with the government during prototype design and development to ensure that the delivered products will be relevant to ongoing and planned missile defense projects. This prototype design will be used to form the development and implementation of a mature, full-scale capability in Phase III. PHASE III DUAL USE APPLICATIONS: Scale-up the capability from the prototype utilizing the new hardware and/or software technologies developed in Phase II into a mature, fieldable capability. Work with missile defense integrators to integrate the technology for a missile defense system level test-bed and test in a relevant environment. REFERENCES: 1. G. Deshpande, C. Arora and G. Ruhe, "Data-Driven Elicitation and Optimization of Dependencies between Requirements," 2019 IEEE 27th International Requirements Engineering Conference (RE), Jeju Island, Korea (South), 2019, pp. 416-421, doi: 10.1109/RE.2019.00055. https://ieeexplore.ieee.org/document/8920671 ; 2. C. Wang, L. Ma, R. Li, T. S. Durrani and H. Zhang, "Exploring Trajectory Prediction Through Machine Learning Methods," in IEEE Access, vol. 7, pp. 101441-101452, 2019, doi: 10.1109/ACCESS.2019.2929430. https://ieeexplore.ieee.org/document/8766820 ; 3. R. M. M. Vallim, A. C. P. L. F. de Carvalho and J. Gama, "Data Stream Mining Algorithms for Building Decision Models in a Computer Role-Playing Game Simulation," 2010 Brazilian Symposium on Games and Digital Entertainment, Florianopolis, 2010, pp. 108-116, doi: 10.1109/SBGAMES.2010.14. https://ieeexplore.ieee.org/document/5772278 ; 4. Y. T. Demey and M. Wolff, "SIMISS: A Model-Based Searching Strategy for Inventory Management Systems," in IEEE Internet of Things Journal, vol. 4, no. 1, pp. 172-182, Feb. 2017, doi: 10.1109/JIOT.2016.2638023. https://ieeexplore.ieee.org/document/7778999