Automated Knowledge Discovery and Reliability Analysis for the F414 Engine

Impact Technologies, LLC proposes to further develop and demonstrate an automated knowledge discovery and reliability analysis tool for the F414 engine. The work is based on available data in NAVY databases such as DECKPLATE or DECKETR or potentially the Maintenance Data Warehouse (MDW) maintained by GE. Knowledge discovery techniques will be applied to maintenance records. Using that knowledge, innovative reliability analysis techniques will calculate component reliabilities using tracked life usage indicators and maintenance information. These techniques are well suited to addressing the competing risk problem as we deal with multiple failure modes and the right-censoring problem, as components are maintained before they fail. Once fleet-wide in-service component reliability is calculated, it will be trended and tracked. Users will be notified of any significant deviations in reliability early on. An opportunistic maintenance/component matching optimization module will also use in-service reliability to make recommendations on warranted maintenance and component replacement selection. Specifically the core innovations of the proposed work include: 1) a knowledge discovery module that processes maintenance records, trends reliability and detects outliers; 2) an in-service component reliability analysis module based on historical maintenance records; and 3) an opportunistic maintenance/component matching optimization module that would maximize the expected time on wing.