Predicting Machine Failure in Maintenance Systems Using Markov Chain Analysis: A Case Study at the Mechanical and Mold Manufacturing Factory / Al-Zawraa State Company

الملخص

The purpose of this research is to study the problem of machine failures in production systems and their impact on operational efficiency and maintenance costs.  The main problem addressed in this research is the lack of an effective predictive model that can analyze machine condition transitions and estimate the probability of failure within maintenance systems at the factory studied.  The research method adopted a case study approach which combines historical maintenance data, probabilities, and the Markov chain analysis.  The Markov chain analysis was the major tool to represent machine states including Good, Degradable, and Failure.  The research sample consisted of operational time-series data collected from the states of two critical cutting machines experiencing failures at the Mold and Manufacturing factory from January to December 2025.   Key performance measures such as system reliability, availability, mean time between failures (MTBF), and expected downtime are calculated to evaluate system performance.  The findings reveal that the Markov chain model is effective in predicting machine failure patterns and provides valuable insights into system deterioration and recovery dynamics.  This study addresses a gap in predictive maintenance modeling by applying a probabilistic Markov framework to real industrial data.  The research came up with several recommendations the most important of which is the adoption of a predictive maintenance strategy rather than TPM.  The present research is significant since it contributes highly to the field of operations management in general and to the field of maintenance in particular.  In addition, this research can play a crucial role in improving the efficiency and sustainability of the studied factory.