Performance-Based Technique for Assessing Equipment Maintainability.

摘要

Previous research produced a computer-implemented model of corrective maintenance performance, based on a relatively simple maximum-productivity rule for selecting maintenance operations, and a relatively complex data base representing particular systems. The model has been expanded to recognize the impact of task-sequence context upon the actions necessary to accomplish tasks. Decisions at each stage of a simulated maintenance requirements now reflect the effects of previously performed work on the time and effort necessary to perform future tasks. Maintainability projections were generated for a digital infrared transmitter/receiver system, specially constructed to be configured in two functionally equivalent forms. Ten electronics technicians worked to identify and resolve eight inserted malfunctions each, using built-in indicators and standard test equipment. The overall projections of maintenance times compared well to the experimental data. A measure of design complexity is proposed for the evaluation of maintainability. This measure, mean number of indicators necessary to accomplish fault isolation, is sensitive to multiplicity of fault modes and to the extent to which fault symptoms are confounded at the maintainer interface.