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

An AI/ML-enabled simulation optimization tool can exploit existing simulation models to streamline the scenario planning and analysis process. Our Phase II project will focus on using AI/ML techniques and simulation optimization to solve the problem of generating a minimal collection of test scenarios, using the minimum number of shared scenario components that meets a required set of test objectives. Each individual scenario taken through the end to end scenario planning process involves substantial effort including discussion, documentation, and manual scenario generation in high fidelity tools. Likewise, scenario components require a great deal of time and effort to prepare for the test event. We will explore the use of an optimization tool with medium fidelity digital simulation models early in the scenario planning process to minimize the required test scenarios and scenario components. Reducing the number of test scenarios and scenario components has the potential to significantly reduce the end to end scenario planning timeline and associated costs of both planning and execution. We will leverage the SBIR-derived OptDef simulation optimization and analysis tool to prototype our scenario planning optimization approach. We will explore ways for this technology to be used in digital event test planning to enable maximum benefit. Approved for Public Release | 22-MDA-11340 (16 Dec 22)