The U.S. Navy's Surface Warfare Officers School (SWOS) in Rhode Island is pioneering the use of a low-cost simulation-based intelligent tutoring system (ITS) as part of its Tactical Action Officer (TAO) training program to train Navy officers in high-level tactical skills. This software was designed and built for SWOS for use on standard PCs, was introduced to the School in early 1999, and the Navy has a royalty-free license to use it. The software can be used both as a classroom aid and by individual students. A key objective of the software is to increase the active training that officers receive to improve their ability to apply their conceptual knowledge of tactics. Early results from its use with two classes are encouraging and indicate that the software will succeed in this goal by enabling as much as a ten-fold increase in hands-on training.

The software has three parts. First, there is a scenario generator, with which instructors, with limited assistance from a programmer, can create any number of simulated scenarios. These can be set in any part of the world, and populated with different surface and air platforms. Each individual platform is implemented as an "intelligent agent" and can be given its own performance characteristics and behaviors so it will act realistically. For example, a hostile plane will have its own mission (which the student can only surmise) but will react to challenges from the student's friendly platforms. Second, there is the intelligent tutoring system, which presents selected scenarios to the student to practice different tactical concepts. The software will adaptively select scenarios for individual students, that practice concepts he or she hasn't yet practiced or has recently failed, or enable a student to pick any scenario. As well as the intrinsic feedback that free-play simulations naturally provide a student, the TAO ITS provides detailed, useful extrinsic feedback to the student once a scenario is finished, which reviews all the concepts attempted and whether they were passed or failed. At this point, the student can review multimedia material about any concept, or see a replay of the scenario to review errors. The third part of the software is an instructor interface tool for instructor to review all the students' work with the tutoring system to assess their progress in detail.

This paper describes the Tactical Action Officer Intelligent Tutoring System as an example of what ITSs can do and the benefits they can provide. It also includes an explanation of why the case-based reasoning technique was used in the software to reduce three problems commonly associated with intelligent tutoring systems: effective incorporation of subject expert knowledge in the software, cost, and development time. It also reviews SWOS's experience with the software since its introduction, students' opinions of the software, and suggests ways in which future simulation-based intelligent tutoring systems might be improved based on SWOS's experience.