LEGACY PLATFORM DATA VALIDITY

摘要

This paper addresses a disconnect between modern data analysis programs and the data gathering processes used for legacy DOD platforms. 1 Analysis shows unless the data collection processes are improved, component prognostic and diagnostic programs will fail in their designed purpose; actually decreasing platform reliability while increasing operating costs. Reviewing current processes and specific cases demonstrates how the processes, coupled with human factors, work in concert to invalidate raw failure data. The section concludes with how these processes affect the different DOD data collection and management programs. How a modern analysis system with AI capabilities using the current failure data would increase operating costs and reduce platform reliability is shown using cases and scenarios based on an existing platform. The impacts on safety, mission completion rates, availability, platform management efforts and operating costs are discussed to demonstrate how the combination of new technology and invalid data would have the opposite effects than those predicted or desired. The final section explains how data validity can be improved using solutions that are neither costly nor complex. A comparison is made between outcomes if the raw data remains invalid and the results if the raw data is valid. From this comparison, conclusions are drawn and recommendations made.