Milind Tambe, Paul Rosenbloom and Karl Schuamb
This paper focuses on our recent research effort aimed at developing human-like, intelligent agents (virtual humans) for large-scale, interactive simulation environments (virtual reality). These simulated environments have sufficiently high fidelity and realism[l 1,23] that constructing intelligent agents requires us to face many of the hard research challenges faced by physical agents in the real world -- particular, the integration of a variety of intelligent capabilities, including goal-driven behavior, reactivity, real-time performance, planning, learning, spatial and temporal reasoning, and natural language communication. However, since this is a synthetic environment, these intelligent agents do not have to deal with issues of low-level perception and robotic control. Important applications of this agent technology can found in areas such as education , manufacturing , entertainment [2, 12] and training [7, 24].