Exploring the Relationship Between Accumulated Departures from Specifications and Associated Casualties and Mishaps Open Access
Downloadable ContentDownload PDF
The Systems Engineering community spends considerable effort developingsystem specifications during the design phase. Yet during the operational and supportphase, there is a potential degradation of those specifications in the form of delayed,missed, or insufficient maintenance (i.e., maintenance that does not restore the system tothe design specifications), which are commonly called departures (e.g., Structure/WeldJoint - Incorrect Electrode Usage, Valve Timing, etc.). While the impact of eachdeparture on the system is reviewed as part of the current approval process, there is noevaluation to the equipment and/or personnel from the accumulated number ofdepartures.The impact of these accumulated departures is analyzed to determine if there is acorrelation between these accumulated departures and casualties to equipment(documented on casualty reports that impact system availability and operationalreadiness) and/or mishaps to document a safety event and/or damage to property. Theanalysis required the development of a framework to systematically store and catalogU.S. Navy data on a select set of hulls from 2004 to 2016 specifically addressing data on6,810 departures, 4,808 casualty reports, and 6 mishaps. A series of analyses wereconducted to include (1) ensuring the hulls met the criteria of in-service (i.e., operationalready for deployment), (2) test for trends together as a class followed by this sameanalysis on a per hull basis, which helped define the correct correlation method, (3)Spearman’s rank-order correlation analyses as a class followed by this same analysis on aper hull basis and (4) regression analysis to determine if departures could be used topredict future casualties. The correlations and regression results suggested meaningful,statistically significant at the 0.01 level, for a majority of the relationships between theaccumulated number of departures determined at the class level and for each individualhull. The framework and process that is described in this paper can be used to track andinfluence the number of casualty reports that are predicted to occur by controlling thenumber of accumulated departure from specifications. There also was no correlationdetermined between the accumulated departures from specifications and subsequentMishaps and thus there was no regression analysis conducted.