An application of diagnostic inference modeling to vehicle health management

In this paper we discuss the approach we have taken to applying diagnostic modeling and associated reasoning techniques to the problem of diagnosing and prognosing faults as part of vehicle health management systems. We present a brief background of diagnostic fault modeling based on lessons learned from ongoing research by ARINC in cooperation with NASA Langley Research Center, as part of the NASA/FAA Aviation Safety Program. We discuss the application of these techniques and possible implementation scenarios to commercial aircraft health management. We identify information sources available on a typical commercial transport and discuss methods for evaluating them, either singly or in combination, to establish knowledge of the current or predicted health state of the aircraft.