Methodological Evaluation and Multilevel Regression Analysis of Manufacturing System Reliability in Ethiopia

"background": "The reliability of manufacturing systems is a critical determinant of industrial productivity and economic development. In many developing economies, systematic evaluations of these systems, particularly within operational plants, are scarce, leading to a lack of context-specific empirical evidence for improvement. ", "purpose and objectives": "This study aims to methodologically evaluate the reliability of manufacturing systems within operational plants and to develop a robust analytical model for identifying the key operational and organisational factors influencing system reliability. ", "methodology": "A cross-sectional field study was conducted across multiple manufacturing plants. Data were collected on system failures, maintenance logs, and operational parameters. A two-level hierarchical linear model was employed, with equipment nested within production lines. The core model is specified as \ (MTBF{ij) = \0j + \1Xij +, where \0j = \00 + \01Zj + u0j. Analyses used robust standard errors to account for heteroscedasticity. ", "findings": "The multilevel regression revealed that preventive maintenance adherence at the line level was a significant predictor of mean time between failures (MTBF). A one-standard-deviation increase in preventive maintenance compliance was associated with a 24% increase in MTBF (95% CI: 18% to 30%). Variability in reliability was substantially greater between production lines than within them. ", "conclusion": "The methodological approach confirms that manufacturing system reliability is influenced by factors operating at different organisational levels. The findings underscore the disproportionate impact of line-level management practices over equipment-specific variables in this context. ", "recommendations": "Manufacturing managers should prioritise and rigorously enforce scheduled preventive maintenance protocols at the production line level. Further research should integrate real-time sensor data into the proposed multilevel modelling framework. ", "key words": "system reliability, hierarchical linear modelling, preventive maintenance, manufacturing, industrial engineering", "contribution statement": "This