Stochastic analysis and design of CONWIP controlled production systems.

Purpose - This paper aims to present a model of a multi-phase multi-product manufacturing system considering a CONstant work-in-process (CONWIP) control mechanism and using continuous-time Markov chain modelling approach. Design/methodology/approach - The model includes defining a state space then constructing the rate matrix, which contains the transition rates, followed by formulating the transition matrix. The time-dependent probabilities that a product is in a particular state at a certain time are characterized. Performance measures related to the statistics on the waiting time and average number of work-in-process in the production system have been determined. Consequently, a numerical example is presented to illustrate the computations of different model aspects. Findings - The analyses explain a foundation needed for analyzing the steady state behavior of manufacturing systems. Results have shown how production data can be easily modified for what-if analyses by the use of Excel add-in tool. Practical implications - The multi-level model outlines a framework that provides a practical tool for production engineers seeking to enhance the performance of their production system by selecting the best order release mechanism. Originality/value - A novel aspect of the work reported in this paper is the application of Chapman-Kolmogrov mathematics and CONWIP ordering theory, which is developed for evaluating and managing CONWIP controlled production systems. [ABSTRACT FROM AUTHOR]