Flight simulation is a primary and critical form of training; however, simulators have not been integrated with the technology of computer-based training. This project proposes to integrate an existing computer-based training system with current flight simulators. As a first step, an expert system will be applied to the evaluation of a Shuttle Mission Simulator lesson using that simulator's capability to store data.The results of the Phase I work demonstrated the feasibility of an intelligent, computer-assisted instruction system to evaluate lesson performance on simulators. The evaluator would be quite useful to improve the overall training operations especially for complex lessons like rendezvous that require a great deal of instructor expertise or for lessons that include specialized assessment as in fuel usage for maneuvering spacecraft. Phase I produced the following: a procedural-network, knowledge structure for representing both an expert and a student model of the simulator lesson; a strategy for evaluating a student or crew's performance based on (mocked-up) data generated by the simulator; and an initial methodology to explain student errors. Furthermore, a great deal was learned about knowledge engineering in this domain and the most pertinent issues for the NASA simulator environment and training.Potential Commercial Application:A list of potential commercial applications includes: airline and aerospace simulators, military operations using simulators (e.g. missile simulators), and commercial space operations.STATUS: Project Proceded to Phase II
