Simulation Based Predictive Analytics (SimPA) for Enhanced Autonomy and Artificial Intelligence Collaboration (AAIC)

RTSync proposes to develop a Model-based System Engineering (MBSE) framework based on DEVS Modeling and Simulation (DMBSE) that supports predictive analysis to enhance design, development, testing, and evaluation of complex system of systems with mission threads and workflows. RTSync’s simulation based predictive analytics (SimPA) platform will offer predictive capabilities to reduce development, integration, and deployment time and risk of new mission critical applications on autonomy and artificial intelligence (A2I) and lay the foundation for new business models. To achieve this objective, we aim to provide a prototype environment that employs our previously developed Collaborative Architecture for Model Level Integration (CAMLI) to effectively construct, reuse, and adapt families of mission thread and work flow models to evolving requirements. We plan to demonstrate that the prototype DMBSE predictive analytics platform to enhance A2I for better mission-tailored decision support for researchers at Air Force Research Laboratories (AFRL) and other department of defense, government, and commercial customers. Deliverables for the Phase II include a prototype to demonstrate functionality, software design, and web interfaces documents, as well as, demonstration of the prototype and design components integrated into AFRL environment. Our major milestones for this project include: selecting a Mission Thread as the basis for prototype demonstration, developing the model components required for the selected mission thread, extending the theory of modeling and simulation to help locate the most suitable model component or composition in the repository for a given such use. The team will specify metrics to measure relevant aspects of the intended workflow and the results of applying the proposed methodology and software system.   The particular goal of the application to AFRL is  to improve the pace and rigor at which  air force and collaborating forces share and deploy Artificial Intelligence and Machine Learning (AI/ML) technologies in training and operational exercises. This will involve helping AFRL to develop a novel infrastructure that will enable coalition forces to collaboratively discover and reuse AI/ML for dynamic mission requirements.   RTSync’s MBSE-based simulation predictive analytics framework will enable AFRL to formalize mission and infrastructure requirements for planned exercises, discover where agent interactions can be improved, and design better system-integration tests. The Air Force currently lacks engineering methods for the development of large-scale AI/ML systems that involve a number of different stakeholders with different requirements and policy restrictions, and which span multiple organizations. The proposed framework by RTSync will advance the Air Force’s agility to rapidly train, test, and integrate state-of-the-art AI/ML into operations at the speed and scale necessary for 21st century warfare.