CASTT: Coaching Assistant for Scenario-Based Teamwork Training

Effective team training is at the heart of enhanced mission effectiveness for military operations. However, despite numerous advances in multi-player simulation-based training environments and intelligent tutoring system (ITS) technologies for individual skills, these systems have yet to fully address the automated instruction of teamwork skills. Accordingly, we propose to build CASTT (Coaching Assistant for Simulation-Based Teamwork Training), an Intelligent Teamwork Tutoring System (ITTS) for training small Army dismounted infantry teams within massively multi-player (MMP) MOUT environments. Our primary development objectives are: (1) a theory-driven training curriculum - missions, environments, and events that provide opportunities for practice, assessment, and feedback on teamwork-specific skills; (2) automated performance assessment capabilities - rules for mapping simulation data to representations of situations, tasks, and competent/incompetent trainee actions, as well models of trainee performance over time; and (3) feedback components - After Action Review (AAR) functionality that directs annotated replay of scenario events during which errors were detected, as well as additional training aids that guide the selection of training events in order to target skills identified as most in need of practice across team members. Collectively, this effort will result in the development of automated coaching technology that is uniquely tailored to the requirements of simulation-based teamwork training.