This paper introduces a benchmarking technique we call Predictive Analytics (PA). Using a large set of data from FMEDA results in combination with validated field failure data, upper and lower reasonability bounds for instrument device failure rates can be established. These bounds can be used as a benchmark to help validate any data collection process. This benchmark represents the constant total failure rate (λ) inherent in the device during its useful life. For any given set of field failure data (FFD) for a device, a λ of the device is estimated and compared to the benchmark. It is not uncommon for the benchmark λ and estimated λ to differ considerably. PA provides a procedure for exploring explanations of these differences and assessing the accuracy of the estimated device λ with respect to the benchmark λ of the device. PA can often determine the source of that portion of the estimated λ value not inherent to the device but likely due to random failures of infant mortality, wear out, or initial failures, to systematic failures, or to application or site specific issues. This site specific device λ is the portion of the estimated λ the end user needs to address to improve operational reliability and safety. PA can also assess the quality of FFD and can facilitate the discovery of previously unknown device failure modes.
