The Soar Training Expert for Virtual Environments (Steve), is a pedagogical agent for virtual environments. Steve's objective is to help students learn to perform physical, procedural tasks, such as operating or repairing complex devices. Students are immersed in a 3D computer simulation of their work environment, where they can improve their skills through practice on realistic tasks. Steve physically cohabits the virtual environment with them, continually monitoring the state of the environment and periodically controlling it through virtual motor actions. Steve helps students in various ways: he can demonstrate tasks, answer questions about the rationale behind task steps, and monitor students while they practice tasks, providing help when requested (c.f., [Rickel & Johnson1997]).
Although Steve is independent of the domain in with which he interacts, in this article we refer to elements of the domain in which he is currently operating, which includes items such as compressors, dipsticks, surge tanks, and the like.
In this section we give a sampling of how affective reasoning can be integrated with Steve. For several sample cases we give first the pedagogical design goals in integrating the affective component into the agent's design, and second an idea of how this would be implemented.