DREAMIT - Design, Reconfigure and Evaluate Autonomous Models in Training

Our Phase I work focused on improving modeling and simulations so that the impact of autonomous systems in the battlespace could be better understood. As we have trained our attention on Phase II, it has become increasingly clear that the work we are doing to improve the modeling and simulation of autonomous systems also provides significant leverage for the development of the intelligent behaviors that will underpin a pilots associate. We see a significant opportunity to advance the state of the art in human-machine systems, including the scale of intelligent agent models, the efficiency of constructing them, and the speed of integrating and validating them with simulation environments. We call our solution DREAMITDesign, Reconfigure and Evaluate Autonomous Models In Training. Our driving goal in Phase II is to demonstrate DREAMITs capabilities by using it to create an operational prototype of a pilots associate. DREAMIT represents a significant advance in the development of intelligent agents. DREAMIT advances intelligent agent development with best practices in software developmentby promoting modularity, generality, and re-usein the design of such agents, DREAMIT will lead to efficiency gains in agent development and validation, and ultimately reduce costs.