Evaluation of High-Frequency Welding Machine Maintenance Schedule Using a Probabilistic Approach

This study evaluates an actual maintenance schedule in a steel pipe manufacturer in Indonesia, particularly in its welding process. In the December 2023-January 2024 period, there was a breakdown time of 36.8 hours on high-frequency welding machines, which indicates low reliability. This company adopts Corrective Maintenance (CM), which leads to frequent downtime due to the lack of preventive action. Therefore, this study aims to determine the optimal maintenance of high-frequency welding machines at a minimum cost. The corresponding machine has 638.4 hours of Mean Time To Failure (MTTF). This result corresponds to the Weibull distribution probabilistic model, where a potential failure above 50% occurs within t > 600 hours. This pattern clearly shows that the manufacturer should establish an optimal maintenance schedule around the MTTF. According to the result, the optimal maintenance policy is the Periodic Replacement Policy. This policy encourages the manufacturer to conduct preventive maintenance every 645 hours. The minimum expected maintenance cost rate has also been obtained through this policy, which is IDR 21,516.00. Keywords: Maintenance, Mean time to failure (MTTF), Repairable System, Repair Policy, Weibull Distribution