Task Taxonomies and Modeling for System Performance Prediction

Abstract

The problem of predicting performance of complex human machine systems is a very difficult one. The use of task-oriented simulation models makes this problem more tractable, but by no means converts it to a trivial exercise. The difficulty of determining an appropriate abstract model from analysis of the actual human machine system remains. An approach is suggested for developing a systematic method of analyzing human machine systems and for formulating a task-oriented model from the analysis results. The role of task taxonomies in systematic analysis is discussed.