Determine component probability from NASA ground system Problem Reports

During the shuttle era, National Aeronautics and Space Administration (NASA) used a failure reporting database called the Problem Reporting and Corrective Action (PRACA). Its purpose was to identify and track system nonconformance. The PRACA database over the years changed from a nominal way to identify system problems, to a complex tracking and report production database. The PRACA database became the primary method to categorize any anomalies, from corrosion issues to catastrophic failures. The systems documented in the PRACA database range from flight hardware to ground or facility support equipment. While the PRACA database is complex, it does possess all the failure modes, times of occurrence, length of system delay, parts repaired or replaced, and corrective action performed. The difficulty is mining the data to then utilize that data in order to estimate component, Line Replaceable Unit (LRU), and system reliability analysis metrics. In this paper, we identify a methodology to categorize qualitative data from the ground system PRACA database for common ground or facility support equipment utilizing a heuristic method to identify credible failures. These data are then used to determine interarrival times to perform an estimation of a metric for repairable component or LRU reliability. This analysis is used to determine failure modes of the equipment, probability of the component failure mode, and to support various quantitative differing techniques for performing repairable system analysis. Components or LRU´s are evaluated in the same environment and condition that occurs during the launch process.