Downloaded from https://www.everyspec.com ADS-79D-HDBK However, even the most urgent development process should follow an organized implementation to ensure that the results are effective. The processes related to identifying candidate CIs and HIs should be guided by performance of the results. Since the process of CI and HI development is data driven, there are a number of proven methods to assess the fault detection, isolation, and RUL estimation performance as defined in the following paragraph. Determining the CI and HI capability to discover the fault early and with high confidence, as well as providing an estimate of RUL with high confidence, is essential to success for CBM.
The indicator will show significant separation between faulted conditions and healthy conditions as defined by Receiver Operating Characteristic (ROC) curve analysis or other comparable analysis. The indicator should be physically meaningful, designed to detect specific fault conditions that are named in the FMECA. The indicator should be designed to operate in an aircraft environment taking into account aircraft noise and components that would not be installed on laboratory test stands. The indicator‘s response should be unique for the fault mode(s) that apply to it. The indicator should not respond to external noise or other fault modes.
D.4 SPECIFIC GUIDANCE D.4.1 Condition indicator (CI) and health indicator (HI) behavior. CIs and HIs included in the CBM system for a particular Army aircraft are based on the following criteria: a. They are identified through RCM methods including FMECA and may be categorized as: i. Category 1–Catastrophic: Faults that could result in death or loss of the aircraft. All Category 1 faults identified in RCM analysis should have CIs/HIs developed, unless the forecast rate of occurrence is less than 1 per 10 million flight hours and selected by the AED. ii. Category 2–Severe: Faults that could lead to severe injury or damage to the aircraft.
All Category 2 faults should have diagnostic coverage unless the forecast rate or occurrences is less than 1 per 1 million flight hours. The diagnostic coverage should be allocated to the most frequent faults to the least frequent faults. iii. Category 3–Major: Faults that may result in damage or injury. Included only in cases where the degradation in readiness or cost exceeds thresholds determined by the Program Manager (PM) for the aircraft. May also be included if the fault leads to cascading failures of Categories 1 and 2. Coverage for Category 3 faults should be determined from analysis of maintenance costs and readiness and selected by the PM. b.
The CIs/HIs should be explainable in physical terms, such as bearing failure, shaft misalignment, or high temperature. c. The CI/HI is identified by analysis that considers its functional role in the system as well as its physical properties. The functional analysis describes the impact of degradation or loss of the function on the rest of the component or system. This analysis may include Principal Component Analysis (PCA), a technique that reduces multi-sensor data or data from correlated variables into a smaller set of data which optimizes CI/HI performance. d.
The CI/HI is analyzed with respect to the feasibility of detecting the fault, the repeatability of gathering accurate fault data through the sensor, and the relative cost or effort required to obtain the 77