Confidence-Based Reliability Evaluation of Multistate Production Network with Process Improvement

Abstract

Effectively assessing and improving production systems in different scenarios are critical for the management of production processes. This paper proposes a new approach that takes process improvement into consideration to evaluate system reliability. The system reliability is defined as the probability that the volume of input processed successfully through the individual workstations within a production system. This system reliability evaluation approach models the production system as a confidence-based multistate production network (CP-MPN). The discrete-time Markov chain is adopted to feature the CP-MPN, and the target success rate (TSR) is set as a reference to achieve the purpose of process improvement of each workstation. Our results indicate that, when the TSR is improved, the CP-MPN can meet both the given demand and confidence level with less input. This curtailed input volume diminishes the loadings of workstations so that the system reliability of the CP-MPN can be enhanced. Two examples (a tile production system and a printed circuit board (PCB) production system) are presented to demonstrate the confidence-based reliability evaluation approach and prove our results. This paper provides not only a practicable method to evaluate system reliability that includes the process improvement perspective but also a workable indicator to guide production managers to improve the quality of workstations.