A FUNCTION-BASED EXPLORATION OF JPL'S PROBLEM/FAILURE REPORTING DATABASE

摘要

This paper presents the exploration of a failure database derived from a Problem and Failure Reporting (P/FR) database for NASA space missions. The overall goal is to analyze an existing database of problems and anomalies recorded at the Jet Propulsion Laboratory (JPL) for multiple missions, and classify a subset of them into distinct failure modes from a predetermined set of failure modes. Results are presented to describe the failure modes and functional descriptions derived from the PF/R database, using observed in-flight failure reports for six unmanned space missions. The challenges and issues in mining this information are presented, as well as comments on the adequacy of the existing failure mode and function taxonomies to describe the subset of observed failures reported in the PF/R database, and on the utility of such large databases.