Optimization of Maintenance Inspection Scheduling Decision by Integrating Predictive Maintenance and Operational Management Constraints: Case Study of Storage Tank Maintenance in Central Java

This study aims to optimize the cleaning and inspection schedule of storage tanks by integrating predictive maintenance with operational decision-making. The research addresses the challenges faced by the Oil and Gas Company in balancing asset reliability with continuous operational demands, using a mixed methods approach. Predictive modeling, conducted using RapidMiner, classified tank conditions based on technical data such as bottom plate thickness, corrosion rates, and coating conditions from previous reports (2022–2024), achieving a model accuracy of 94%. Simultaneously, the Analytical Hierarchy Process (AHP) was employed to evaluate operational criteria, including product demand, stock availability, redundant tank presence, overdue cleaning, vendor readiness, and alternative supply plans. Seven regional Terminal Managers participated as respondents to determine the prioritization of these factors and the weighting of each criterion. The results of AHP indicate that the technical condition of tanks (24.4%) and overdue cleaning (18.4%) are the most influential criteria in the decision-making process. By integrating AHP with predictive maintenance simulations, the study proposes a structured, proactive, and risk-based model to guide tank maintenance decisions. This model enhances reliability, minimizes operational disruptions, and ensures optimal resource utilization.