Spacecraft system-level integration and test discrepancies : characterizing distributions and costs

Abstract

The goal of this research is to characterize the distribution and costs of spacecraft discrepancies found at the system level of integration and test, as well as understand the implications of those distributions and costs for the spacecraft enterprise as a whole. If discrepancies can be better understood, they can potentially be reduced or even eliminated. Reducing discrepancies will result in cycle time reduction and cost savings, as well as increased product quality and reliability. All of these potential outcomes are indications of successful progress toward becoming a lean organization. Data on discrepancies at the system level of integration were gathered from spacecraft vendor databases, while interviews with key program managers and engineers provided perspective and insight into the data. Results are based on 224 spacecraft representing at least 20 different programs or product lines, and encompassing 23,124 discrepancies. The spacecraft date from 1973-1999, and represent different vendors as well as a mix of commercial and government spacecraft. Spacecraft discrepancies are analyzed in this work on the basis of ten categories: the spacecraft mission, the spacecraft subsystem where the discrepancy occurred, the date of the discrepancy occurrence, the discrepancy report open duration, the immediate action taken to fix the discrepancy (disposition), the root cause of the discrepancy, the long-term corrective action prescribed to prevent the discrepancy from happening again on future spacecraft, the labor time spent on the discrepancy, and the cycle time lost due to the discrepancy. Statistical measures of central tendency, correlation and normality are presented for each category. This statistical analysis forms the basis for research findings at the enterprise level in the areas of quality yield, resource utilization, stakeholder satisfaction and flow time. Recommendations to enterprise stakeholders for increasing the value derived from system-level integration and test follow from the enterprise-level findings.