Reliability engineering applied to maintenance optimization: a case study of injection molding machines in a Brazilian pen manufacturing industry

Abstract Reliability engineering is essential for ensuring the performance, safety, and quality of systems and products across various branches of engineering. In manufacturing environments, maintenance represents a significant challenge due to cost control, operational complexity, and productivity requirements, often accounting for approximately 5% of gross revenue. This study investigates the maintenance policy adopted for injection molding machines in a Brazilian pen manufacturing industry, with the goal of designing a cost-effective preventive maintenance strategy. The proposed approach employs a flexible reliability model based on a mixed Weibull distribution incorporating a restoration factor, which allows accurate representation of different failure modes and post-maintenance behaviors, with parameters estimated from multiple machines. The model parameters and restoration factor were estimated directly from historical data using the Differential Evolution algorithm, a key methodological contribution that enabled robust maximum likelihood estimation under complex search spaces. The results identify optimal preventive maintenance intervals for the studied machines, balancing preventive and corrective costs. In practice, the proposed strategy demonstrates how data-driven maintenance policies can significantly reduce total maintenance costs and improve planning decisions, providing valuable guidance for industries seeking reliability-centered solutions.