Optimization of Production Processes in the Furniture Industry Using Semi-Markov Models

Purpose: This study aimed to develop a semi-Markov model for assessing the reliability of a machining centre in a furniture industry enterprise. The focus was on minimizing production downtime caused by technical failures, addressing the lack of industry-specific reliability models for medium-volume furniture production. The study seeks to support furniture enterprises in managing production continuity effectively. Design/Methodology/Approach: Empirical data from a furniture enterprise operating continuously in three shifts were analyzed. The semi-Markov model utilized technical documentation and records of 6005 downtimes and failures over two years, focusing on technical failures lasting over 10 minutes. Five key machine elements were modeled, with their times of fitness for use described by exponential distributions. Mathematica software was used to compute the density functions of these distributions, allowing for detailed analysis of the failure intensities of individual elements. Findings: The model predicts the probability of the machining centre transitioning from a state of fitness for use to unfitness, based on failure intensities. It determines the expected failure-free operation time and the frequency of failures, aiding in optimal production and maintenance planning. Practical Implications: The model helps furniture enterprises improve the reliability of complex technical systems by predicting failures and planning maintenance activities. This reduces unexpected downtime and supports production scheduling, particularly for large-scale contracts. Additionally, it can optimize spare parts inventory and maintenance strategies, lowering operational costs. The proactive approach facilitated by the model enhances asset management through condition-based maintenance policies. Originality/Value: The study is an innovative application of semi-Markov process theory to the reliability of technical systems in the furniture industry. Its originality lies in adapting the model to real-world operating conditions and analyzing the reliability of key machine elements using exponential distributions. The findings contribute to predictive reliability models with potential applications in other industries.