SBIR-STTR Award

In a few years, the National Training Center (NTC) will have new simulation systems and upgrades to RDMA and RMCS. Given effective and efficient integration, tehse systems will increase, significantly, the amount and fidelity of data available to the training analyst. How effective these improvements are depends on how efficiently the man-machine combinations at training analyst work stations are able to collect, analyze, package, and distribute the desired performance data. In Phase I, we will conduct functional requirements analyses of each step of a proposed concept; investigate the feasibility of advanced processing and decision aid techniques; and develop a taxonomy that technically, as well operationally, will lead to automated integration of the data collected from simulation, observation (human and television) and instrumentation, and the application of decision aid functionality for organizing and analysing the real time training performance data. In phase ii, we will transition the taxonomy into its total capability context. We will design, model and implement the Phase I approach in sufficient detail to demonstrate the feasibility of the data collection and analysis concept.

Keywords:
Artificial Intelligence Automation Decision Aids Distributed Processing Combat Training Centers

LB&M will develop a prototype Automated Training Analysis and Feedback System (ATAFS). ATAFS is an expert system that automatically generates after action review (AAR) training aids for SIMNET collective training exercises based on answers to preexercise questions, trainer responses to interactive prompts, and rules embedded in ATAFS's knowledge base. ATAFS approach reduces the number and complexity of operator skills needed to operate the system, and compliments the trainer's level of technical and tactical competence and skills of observation. For the Phase II we will encode rules developed during Phase I for platoon task: "Execute Actions on Contact" and acquire sufficient hardware to demonstrate three capabilities. First, we will demonstrate the functioning of rules which trigger the generation of AAR aids based on the recognition of specific tactical events by the system or the trainer. Second, we will demonstrate ATAFS's ability to perform computations and construct AAR aide during and after the exercise. Third, we will demonstrate ATAFS's ability to arrange AAR aids generated from diverse sources and integrate them into a single medium.

Keywords:
Artificial Intelligence Automation Decision Aids Distributed Processing Combat Training Centers