Toward Realism in Human Performance Simulation

This chapter focuses on challenges to improving the realism of socially intelligent agents and attempts to reflect the state of the art in human behavior modeling with particular attention to the impact of values, emotion, and physiology/stress upon individual and group decision-making. The goal is to help those interested in constructing more realistic software agents for use in human performance simulations in both training and analysis settings. The first two sections offer an assessment of the state of the practice and of the need to make better use of human performance moderator functions (PMFs) published in the behavioral literature. The third section pursues this goal by providing an illustrative framework for integrating existing PMF theories and models, such as those on physiology and stress, cognitive and emotive processes, individual differences, and group and crowd behavior, among others. The fourth section presents asymmetric warfare and civil unrest case studies to examine some of the concerns affecting implementation of PMFs such as verification, validation, and interoperability with existing simulators, artificial life emulators, and artificial intelligence components. The final section of this chapter concludes with lessons learned and with some challenges if the field is to reach a greater level of maturity