Improving the performance of manufacturing systems with continuous sampling plans

Abstract

Continuous sampling plans are commonly adopted in various manufacturing industries to improve quality. Quantitative performance measures such as throughput and Work-In-Process (WIP) are highly coupled with quality and are also affected by sampling plans. However, quantitative and qualitative issues have long been studied separately in the literature. This article proposes an integrated quantity and quality model for manufacturing systems with sampling plans. A continuous-time and discrete part flow Markov model is first proposed for a single-stage system. Unlike previous Markov models, this model is capable of calculating various performance measures; e.g., throughput, quality, average fraction of inspection and WIP. A method for performance analysis of two-stage systems is presented. The methods satisfactory accuracy is demonstrated with experimental results. Additionally, quantitative analysis is performed to investigate the effect of sampling fraction and clearance number (parameters of a sampling plan) on the various performance measures. Industrial practitioners may benefit in improving the performance by varying sampling plans. Finally, a decomposition model of multistage systems is studied based on which a method to determine the best sampling plans for manufacturing systems is developed. Numerical experimental results demonstrate the effectiveness of the proposed method in finding appropriate sampling parameters for profit improvement. © 2013 Copyright Taylor and Francis Group, LLC.