Statistical Control Process Application to Test Failure Information

Performance characteristics of weapon system components and the associated support equipment may change with age and other factors. Current test programs have no means to evaluate the effects of these changing factors on equipment performance tolerances. The poblem is exacerbated by the lack of archived performance data and diminishing sources of parts to repair legacy systems.The goal of this task is to investigate and apply statistical process control (SPC) and artificial intelligence (AI) technologies in combination to detect trends in failure data, capture and interpret the trends and then provide information for decision support. The proposed solution envisions a three layer model. The first layer being a management data base which accepts test failure data and organizes it for viewing, evaluation and archiving. The SPC layer interprets the failure data and creates performance characteristic patterns, which are then archived in the management layer. The AI layer utilizes an expert system (ES) shell to integrate the statistical interpretation of thest data from a unit under test with prior interpretations of historical data, thereby generating appropriate decision support information. By continually updating historical data, the system performance will improve with time.