Using reliability indicators to explore human factors issues in maintenance databases

Purpose – The aim of this paper is to explore the usefulness of repairable parts simple historical databases in assisting the human factors experts to identify candidate areas for applying human factors methods. Therefore, also contributing to the search for maintenance quality improvement. Design/methodology/approach – The study was based on the failure history of part fleets installed on the same type of jet engines, and used mean time between failures (MTBF) and failure rates plots, the Laplace trend test, the AMSAA‐Crow‐Duane model and serial correlations. Findings – Increasing and decreasing trends in failure rates indicated factors that cause deflection from the literature assumptions of constant failure mode and “as good as new” maintenance philosophy. Further statistical calculations revealed patterns between MTBF and frequency of maintenance, specific serial numbers (SN) vulnerability to replacement and depot maintenance tasks, correlations between MTBF and number of both installations and maintenances, and influence of the maintenance month on the maintenance‐failure hours' interval. Practical implications – The literature refers to the relation between the parts reliability and the human factors in the maintenance domain. The research confirmed the literature references in data collection problems coming from human factors interferences; the patterns found were attributed to system deficiencies related to workload management, parts configuration management, supervision and manufacturing problems. Originality/value – The application of this research in combination with methods such as field observations and interviews of personnel involved in the maintenance domain can uncover specific maintenance working environment weaknesses and lead to suitable remedies.