Skip to Main Content
The weighted repair assistance program learning extension, or WRAPLE, is a maintenance aid to help technicians isolate faultsÂ¿a necessary step in system repair. We derived the fault isolation strategies used in WRAPLE from a knowledge-based information-theoretic approach that incorporates fixed experience. This approach contrasts with fault isolation strategies, which are developed from estimates and remain fixed. A fault isolation strategy that Â¿learnedÂ¿ would adapt to field conditions and change on the basis of actual failures and use of maintenance resources, permitting more efficient maintenance and reducing mean time to repair. In this article, we describe WRAPLE and the learning algorithms by which it modifies its output. We also present an example of a typical system to illustrate the technique.