Methodology for analysis of maintenance through graphical representation

Maintenance activity is usually captured in a database format that does not lend itself to analysis by graphical means. This differs from analysis of failure information which has historically used exploratory data analysis techniques to assist in detection of relevant contributing factors, outliers and grouping of information. This paper discusses an approach to graphical depiction of maintenance, the interpretation of results and the benefit to maintenance planning, troubleshooting and problem resolution. A simple graphic methodology is introduced to aid in the analysis of maintenance data by providing visualization of tabular data. By applying the techniques described herein, the relationships between the type of maintenance, maintenance action, and component function becomes immediately identifiable. In addition, clusters of maintenance activity are easily viewed and dissected. This automation of this methodology offers the analyst a reduction in the time required to evaluate aircraft maintenance history and a format that is easily presented to and digested by other disciplines.

1 Introduciton

Examination of tabular data associated with maintenance activities, while usually detailed, does not offer the benefits of visual representation such as the techniques employed by Exploratory Data Analysis¹ (EDA) practices. This paper will detail and propose a method for graphical depiction of maintenance activity using a modified Cartesian format supplemented by the overlay of complementary information as available in the data source.