Command agents with human-like decision making strategies

Human behaviour representation in military simulations is not sufficiently realistic,specially the decision making by synthetic military commanders. The decision makingprocess lacks realistic representation of variability, flexibility, and adaptabilityexhibited by a single entity across various episodes. It is hypothesized that a widelyaccepted naturalistic decision model, suitable for military or other domains with highstakes, time stress, dynamic and uncertain environments, based on an equally testedcognitive architecture can address some of these deficiencies. And therefore, we havedeveloped a computer implementation of Recognition Primed Decision Making (RPD)model using Soar cognitive architecture and it is referred to as RPD-Soar agent inthis report. Due to the ability of the RPD-Soar agent to mentally simulate applicablecourses of action it is possible for the agent to handle new situations very effectivelyusing its prior knowledge.The proposed implementation is evaluated using prototypical scenarios arising incommand decision making in tactical situations. These experiments are aimed attesting the RPD-Soar agent in recognising a situation in a changing context, changingits decision making strategy with experience, behavioural variability within andacross individuals, and learning. The results clearly demonstrate the ability of themodel to improve realism in representing human decision making behaviour byexhibiting the ability to recognise a situation in a changing context, handle newsituations effectively, flexibility in the decision making process, variability within andacross individuals, and adaptability. The observed variability in the implementedmodel is due to the ability of the agent to select a course of action from reasonablebut some times sub-optimal choices available. RPD-Soar agent adapts by using‘chunking’ process which is a form of explanation based learning provided by Soararchitecture. The agent adapts to enhance its experience and thus improve itsefficiency to represent expertise.