On 2 August, Charles River Analytics (CRA) announced a $1 million award of additional funding for its efforts in building behaviour modelling capabilities for the US Missile Defense Agency (MDA). The two-year contract for Modeling Operator Reasoning and Performance for Human-in-Control Simulation (MORPHIC) calls for the company to provide tools for missile defence analysts to create models of human behaviour for simulation.
CRA will use the Hap architecture to build the MORPHIC human behaviour modelling features. Hap defines a language for describing agent behaviour and provides an execution engine for running these agents. The company has incorporated Hap into several efforts, including the creation of behaviour that models cyber adversaries, or that can be moderated by physiological parameters, and developing intelligent tutoring algorithms.
“The missile defence community relies on using models and simulations for a variety of applications such as testing, training, exercises, wargaming, and future concept analysis,” explained CRA’s Vice President, Decision Management Systems, Brad Rosenberg. “However, it is challenging to model human behaviour in those systems. Human operators are core to the integrated ballistic missile defence system, so it is vital to understand how human behaviour impacts system performance in various situations. Under the MORPHIC effort, we are providing tools to construct, adapt, and execute human behaviour models within simulations.”
“Modeling behavior is incredibly complex,” adds the company’s Vice President, Human Effectiveness division, Dr. Peter Weyhrauch. “People often manage multiple, conflicting goals, process information in parallel, and are constantly taking in new information and evaluating it against what they currently believe about the world. Describing that behaviour requires fundamental concepts that are not readily available in most programming and modelling languages. The Hap architecture builds in those concepts from the start.”
MORPHIC is one of several efforts CRA has developed for the MDA. Other efforts include a game-based training tool (RAMPART), a simulation optimisation engine (SIMON), and real-time, sensor network optimisation software (SNOMAN).