Unlocking the buried margins entrenched in test data

What information is critical in test and diagnosis? Limits for measurements are certainly a critical aspect of the diagnosis of a Unit Under Test (UUT) fault. Specific measurement deviations can be extremely telling when it comes to potential Unit Under Test (UUT) problems which can have an adverse consequence of lowering the UUT Mean Time Between Failure (MTBF). Even though the measurement might be within tolerance, borderline or out-of-the-normal median range readings tend to be an indication of a future unit problem. A potential future problem in the UUT can be evident by unlocking specific measurement data and making the appropriate information known to the user. Limit parameters are always critical during the test and diagnosis process. Any UUT will fail or show a weakness at some point. There is usually a point between a hard functional failure and a degraded function or pending failure. The UUT will fail, it is just a question of when. The effect of the failure or change in performance interferes in different degrees with missions that depend on that system. In critical missions a system failure can lead to catastrophic consequences. Significant changes in the performance of equipment may occur as a result of aging and other factors within systems and their associated support equipment. Current methods do not allow for the effects of these factors in the determination of equipment performance tolerances or test limits, resulting in apparent and actual decreases in equipment readiness and test program precision. The overall cost of a failure or malfunction, measured in any standard, is always higher than the preventive action. Therefore, system users are interested in knowing when a system or part is about to fail in order to take preemptive actions. Diagnoses in general are complicated, but the test results coming from Fiddly, Fuzzy, Finicky (F³) failure events offer an extra layer of confusion. F³ failures could be classified as a natural sequence of events that are eventually going to happen. On the other hand, it should be stated that some failures never occur even though there is a potential they might occur. Moreover, the test program set (TPS) or built-in test procedures anticipate only catastrophic and stable faults that will always result in a predictable manner². It would not be feasible for TPS developers to plan for exotic failure modes that would result from F³ events. As a result repair technicians are often left in a state of confusion about test results in general and diagnostic help in particular. It is important to know that test technology is not limited in its extent and growth. New hardware and software techniques are continually evolving and we must pursue optimal repair by utilizing saved test data to increase the MTBF and the overall reliability of the UUT. This paper will be a discussion of evaluating test data with the intent to add additional sets of limits so as to provide the repair technician with varied choices depending on the actual measurement for specific UUT tests. The System Test Specifications provide the Upper Limit and the Lower Limit for each measurement. These limits must be followed, but additional limits can be employed which provide the repairer with additional information which can optimize the repair process.