Data-driven predictive maintenance planning using bootstrap Kolmogorov–Smirnov validated three parameter Weibull model for maintenance decision support
Year: 2026
Type: article
Abstract: Industrial plants often need to plan maintenance using routine logs and few repeated failures per component. This study proposes a transparent, uncertainty-aware predictive-maintenance workflow that converts maintenance records into time-between-failures data and fits a location-shifted three-parameter Weibull model via Benard median ranks and Weibull probability-plot regression. Parameter uncertainty is quantified with 95% bootstrap confidence intervals, and model adequacy is verified using a bootstrap-adjusted Kolmogorov–Smirnov test. Applied to two sewing-machine components (tension wire and hook), the validated models indicate decreasing hazard rates (β<1) and provide reliability metrics that guide inspection scheduling, preventive replacement windows and spare-parts provisioning. The results underline the value of reporting uncertainty and validating fit when decisions rely on sparse field data.
DOI: 10.17531/ein/237615
Author: Abir Mouldi
Institution: King Khalid University
Language: English
Subfield: Safety, Risk, Reliability and Quality
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $442
Funder: King Khalid University
Award: G2956939541
OpenAlex ID: W7234258631
