Advancing Reliability Centered Maintenance: A Robust Statistical Approach for MTBF Estimation in Aero-Engine Fleets via Subgrouping Strategy and Dual Weibull Modeling

This poster outlines a robust statistical framework designed to overcome a critical challenge in aircraft maintenance: the inaccurate estimation of Mean Time Between Failures (MTBF) for aero-engine fleets. Presented at the 2025 International Conference on Mechanical Manufacturing Technology and Material Engineering (MMTME 2025) at Northeastern University, Shenyang, China, this work addresses the issue of fleet heterogeneity, where manufacturing and operational variances create statistical noise that obscures true failure patterns. The proposed solution moves away from conventional data pooling and instead introduces a multi-stage data filtration method. Using a Randomized Block Design (RBD) and ANOVA, the framework systematically filters inconsistent data to isolate statistically homogeneous subgroups of engines with a consistent failure signature. This "purified" dataset forms a more reliable basis for maintenance planning. Key Findings Effective Filtering: From an initial 48 random subgroups, the methodology successfully identified and retained 17 statistically consistent ones, discarding over 65% of the noisy data. Validated MTBF: Reliability modeling on the filtered data yielded a primary MTBF of 171.16 hours (using a Weibull model) and a cross-validated MTBF of 176.35 hours (using an Exponentiated Inverse Weibull model). High Confidence: The close ~3% convergence between the two independent models strongly validates the framework's ability to extract a stable and reliable failure signal from a complex dataset. Ultimately, this work provides a dependable foundation for data-driven Reliability Centered Maintenance (RCM), empowering airlines to enhance safety, optimize maintenance schedules, and reduce operational costs.