Squad Cognitive Instructor for Performance of Intelligent OODA (SCIPIOÖ)

In this proposal, Discovery Machine, Inc. (DMI) proposes to leverage our Hybrid Artificial Intelligence (AI) system to support patterns of life (PoL) and intelligent tutoring agents within the ArmyÆs Synthetic Training Environment (STE).Ā DMIÆs objective is to enable intelligent, adaptive, and humanlike synthetic agents for Soldiers. To do so, DMI will utilize the Rapidly Encoded Synthetic Immersive Training Environment (RESITE«) to author complex immersive training scenarios that can be integrated with future STE TMT and TSS systems.ĀĀ DMI shall apply the Multi-level Universal Specification for Intelligent Constructs (MUSIC) to integrate the RESITE« Agent Manager (RAM) with selected STE simulation environments.ĀĀ DMI shall apply our knowledge capture methodology and authoring technology to acquire and deploy subject matter expertise in an intelligent tutor that: adapts training capabilities to adjust subsequent scenarios, automates speech analysis and natural language processing, provides analysis of the traineeÆs cognitive functions during simulation, automates feedback for trainees tailored to individual performance, and for leadership tailored to scenario efficacy, and provides ōAnywhere, anytimeö simulations that allow participants across the globe to collaborate in realtime.Ā In addition, DMIÆs technology enables large-scale and replicated PoL life in simulations and support for a high number of observer controller-trainers (OC-T), simulation pucksters, and human-controlled opposing forces (OPFOR) in the authored scenarios.ĀĀĀĀĀ DMI proposes to build an AI virtual instructor called the Squad Cognitive Instructor for Performance of Intelligent OODA (SCIPIO) that shall monitor and assess a Squad as it performs its missions.Ā SCIPIO shall leverage information from:Ā the simulation, other MUSIC agents in the simulation, past student and squad performance, as well as information from each squad memberÆs sensors within the simulation and student communications with each other and other agents.ĀĀ All of this data shall enable SCIPIO to assess the situation within the authored scenario, and either coach or assess the Squad based on this assessment.ĀĀ This data shall also be stored and analyzed via machine learning enabling SCIPIO to learn from each Squad that trains.ĀĀ Reports shall be provided at the end of each scenario, both to the trainee regarding individual performance and to leadership regarding overall scenario efficacy.