Intelligent Tutoring Systems (ITSs) have proven to be very effective at transferring complex knowledge to a student. They have also proven to be difficult to create and maintain. To create Intelligent Tutors in a cost-effective and maintainable manner, a set of tools and processes called Adaptive Student Tutoring Environment (ASTUTE) is under development. ASTUTE is a composition of commercial off-the-shelf (COTS) applications and specialized software that will enable instructional designers or subject matter experts to create computer-based training that will transfer the proficiency of experts to a learner in a way that is very stimulating to the learner. The resulting tutors, in effect, emulate on-the-job training (OJT) with an expert who is dedicated to being a private mentor for each learner.

This paper presents the ASTUTE architecture and describes how tutors using that architecture simulate mentored OJT and train apprentice students to perform at expert levels. Included are discussions on ASTUTE's methods of: (a) molding instruction to students learning habits, capabilities, and skills; (b) exercising skills in an environment that best reveals the intended use of the knowledge; (c) providing external coaching support that either fades as the student's skill builds or is under control of the student, giving advice only when asked; and (d) providing reflective follow-up that allows comparison of a student's solution with an expert's solution.