This paper examines the theory, design, application and recommended standards associated with adaptive instructional systems (AISs) as drivers for tailoring military training in existing simulation environments. Adaptive instruction has been found to be significantly more effective than classroom or other non-adaptive methods. Most military simulations today are classified as minimally adaptive in that they modify content based solely on learner/team performance, and generally do this in a prescriptive way. AISs are artificially-intelligent, computer-based systems that guide learning experiences by tailoring instruction and recommendations based on the goals, needs, and preferences of each learner/team in the context of domain learning objectives. AISs come in several forms including intelligent tutoring systems (ITSs), intelligent mentors (e.g., recommender systems), and intelligent media. Adaptive instruction is desired as a military training tool to improve training efficiency (e.g., accelerate learning) by intelligently focusing tutoring resources where they are needed most - on gaps in learner knowledge and skill. Both US Army Synthetic Training Environment and My Navy Learning programs have design goals to enable adaptive instruction. Adaptive instruction has been a topic of research for decades and AIS architectures such as the Generalized Intelligent Framework for Tutoring (GIFT) have been used to demonstrate the efficacy of tailoring simulation-based military training (e.g., Virtual BattleSpace – VBS) to optimize learning outcomes. However, the principles needed to automatically tailor training within existing simulation architectures have not been fully described nor generalized. No interoperability standards yet exist. This paper ties together instructional theory and design principles needed to seamlessly integrate AISs with military training simulations with the goal of enabling AISs to automatically tailor instruction in real-time. To realize a fully enabled AIS for military simulations, we have identified components, models, functions, gateways, and data requirements to support both syntactic and semantic interoperability. Recommend practices, standards and use cases will also be discussed.
Simulation
Adapting existing simulation architectures to enhance tailored instruction
27 views
1 Downloads