Cognitive modelling and constrained reasoning for intelligent agents

In the past decade, technological advancements have made it easier for computer simulations to include high fidelity human and environmental behaviour. Cognitive Work Analysis (CWA) has been applied in various domains such as medicine, aviation and military to help analysts develop an in-depth understanding of complex socio-technical systems. CWA may open up a new era for computer scientists who wish to model complex environments where human behaviours are the core elements. Conventional heuristics and reasoning algorithms fail at a point where uncertainties must be included in the simulated human decision-making processors. In order to achieve realism, it is important for such processors to be able to choose the most viable course of action after generating several plans that allow for uncertainty in the domain. Each plan can either be equally weighted, ranked with different importance level or ranked according to personality types (eg. high risk or conservative personalities). In this paper, our aim is to show what information should be fused from the cognitive model which is integrated into a BDI (Believe, Desires and Intention) intelligent agent framework, JACKm. In our case, we are modelling the command & control (C2) and decision-making of a company commander in the battlefield. A mechanised infantry, company attack scenario is used as a case study to demonstrate the functionality and capability of the intelligent agents.