Investigation of the Effect of Time-Dependent Covariates on Maintainability Analysis

Abstract

Conventional techniques for the assessment of maintainability often focus primarily on repair time as a key determinant factor. The consideration of only the “time to repair” variable in maintainability performance evaluation can be restrictive, as several other important covariates can be influence the maintainability performance of the system. Ignoring these covariates may result in an inaccurate quantification of maintainability. Therefore, a more refined model is essential for precisely evaluate the influence of critical operational factors or covariate on maintainability, particularly for industries where system availability is of utmost importance. In maintainability analysis, the Proportional Repair Model (PRM) is widely used. This model operates under the assumption that covariates influencing maintainability, are time-independent. However, this assumption may not always hold true. Factors such as spare part availability and equipment aging can be time-‑dependent covariates. Failing to account for these time-‑dependent covariates can introduce bias into maintainability estimates. Therefore, failing to account of these time-‑dependent covariates can lead to biased estimates of maintainability. The purpose of this study is to explore the significance of time-dependent covariates in modelling maintainability.