Avionics health management: searching for the prognostics grail

The focus of this paper is to present the advances, benefits and challenges in measuring, monitoring and managing the health of aircraft avionics systems as well as the support equipment used to test these systems. Most people are skeptical when avionics and prognostics are used in the same sentence. For the purpose of this discussion, we will grant that most electronic discrete parts fail randomly. However, in the aircraft maintenance arena that point is moot because the lowest repairable level is the assembly on which those discrete parts reside. When an assembly fails or has an intermittent fault (resulting from aging solder connections or other environmental or mechanical factors), it is manifested as a system fault or failure. As degraded performance trends occur over time, there is an increased probability of predicting with reasonable confidence, when a given assembly is likely to experience an insipient fault or a cause a mission failure. While many of the current USAF maintenance metrics add no apparent value to prognostics capability, a few critical data elements are discussed. An optimum set of metrics is proposed through which the performance of avionics assemblies can be monitored. Considerable insight into the relative performance of a wide range of avionics assemblies has been gained through analysis of test parameters and failure information (typically not archived) that have been captured from automated test equipment (ATE). The automated methodology that captures statistically processes and archives test data from both the units under test (UUT) and the ATE instruments is described. The insight gained from this technique has led to cost avoidance in the tens of millions of dollars by reducing no faults found (NFF) occurrences, which, in turn, improves mission capability rates and reduces logistics support cost. Specific examples of these benefits to the USAF F-16 fleet are provided.