Improving the Reliability of Repairable Systems Using a Combined Experimental Design Approach

Enhancing the reliability of repairable systems plays a crucial role in improving operational efficiency and reducing maintenance costs across various industries. Due to their functional complexity and exposure to diverse environmental and operational conditions, these systems are prone to frequent failures. Therefore, developing optimized methodologies for experimental design and reliability analysis is essential.

This study presents an innovative approach that integrates the Taguchi design with the I-optimal design to identify optimal operational conditions, minimize failure rates, and improve system reliability. The Taguchi method was first employed as an effective tool to reduce sensitivity to noise and enhance the robustness of experimental results. Its integration with the I-optimal design further enabled the identification of the best factor level combinations while reducing the number of required experiments. The efficiency and information richness of the resulting design were subsequently evaluated using the D-optimality criterion, which demonstrated high design performance.

Given the limited access to real-world failure data, time-to-failure data were generated through predictive modeling and simulation to evaluate the proposed methodology. For data analysis, parametric survival models were employed, providing accurate representations of system failure behavior and enabling the investigation of interaction effects among multiple factors.

The findings of this study revealed that integrating Taguchi with I-optimal design, followed by evaluation with the D-optimality criterion, significantly improved model accuracy while reducing experimental effort. Moreover, the proposed approach increased system resistance to environmental variations, thereby extending time-to-failure and enhancing overall reliability metrics.

By combining advanced experimental design techniques with robust statistical modeling approaches, this research provides a systematic and practical framework for optimizing the reliability of repairable systems. The results highlight the effectiveness of customized experimental designs in reducing failure rates, improving operational stability, and strengthening system robustness. This study represents an important step toward more efficient reliability optimization methodologies and offers valuable insights for industries such as manufacturing, energy, and transportation, enabling enhanced system performance and longevity with minimized maintenance costs and operational disruptions.