Representing Entity Actions and Cue Taxonomies (REACT)

A major requirement for effective weapon system evaluation, construct development, and training in constructive simulations is the ability to adequately represent behaviors in response to suppressive fire during urban combat. Current simulations lack adequate representation of the behavioral effects of suppressive fire as well as methods for using available data to quickly update these representations to reflect current realities. We propose to address the need for realistic, flexible, easy-to-update representations by designing and prototyping a framework for Representing Entity Actions and Cue Taxonomies (REACT). Our human centric approach is characterized by a formal knowledge elicitation process, which will identify the requirements for modeling and simulation, including a schema for categorizing suppression cues and their associated behaviors. We will then focus on identifying candidate information sources and framework modules that will provide an enhanced representation of the role of suppression in decision making and performance. Finally, we will prototype a conceptual model, which will focus on incorporating perceptual, cognitive, and behavioral inputs, and a system for populating and including the framework in the IWARS. We will utilize experts within our team to develop the REACT framework, populate it with realistic data, and review its potential to support other DoD simulations.