Intelligent Tutoring Methods for Optimizing Learning Outcomes with Embedded Training

摘要

The advancing state of the art in virtual and constructive computer based tools provides an opportunity to construct increasingly engaging embedded training. However, even with the benefits of training conducted while immersed in the operating environment and coupled with advanced simulation technology, if the key ingredient of tailored performance feedback is absent then learning outcomes may still be limited. This amounts to one of the central challenges for embedded training, where the goal is to deliver instructional benefits in a setting where the availability of human instructors is limited or none. Within such a framework, structured training methods provide a means to achieve learning objectives and concept retention, with minimal instructor involvement. Intelligent structured training applies real-time automated evaluation and feedback methods based on Intelligent Tutoring System (ITS) techniques. This paper reviews two case studies of embedded training prototypes developed for the U.S. Army which employ structured training methods to optimize learning without direct instructor involvement. These prototypes include a man-wearable trainer for dismounted operations, and a robotic vehicle control station trainer. This paper also summarizes results from selected preliminary experiments which give indicators for the promise of the structured training approach in the embedded setting, particularly with respect to