A Guideline to Select Control Charts for Multiple Stream Processes Control

Abstract

Abstract. Multiple-stream processes (MSP) can usually be found in a factory. MSP control chart is one of the tools used to control these processes. It is required that MSP control charts are selected appropriately based on streams characteristics, in order to minimize false decision on process status and irrelevant action on the process. This paper presents the guideline and a case study of selecting MSP control charts by considering the major factors, which are correlation among streams, number of streams, limitation to use one chart per stream, difference among streams average, and shift size of streams average to be detected.

Monitoring of Multiple Binary Data Streams using a Hierarchical Model Structure

“…Very limited research exists on simultaneously monitoring multiple binary streams. Most existing works are not for binary streams, but rather for variable data with normal distributions . Very limited work on multi‐stream process monitoring involving binary data exists in literature.…”

“…Most existing works are not for binary streams, but rather for variable data with normal distributions. [3][4][5][6][7][8][9] Very limited work on multi-stream process monitoring involving binary data exists in literature. Most recently, Li et al 10 proposed a method to control the false discovery rate (FDR) of multiple cumulative sum (CUSUM) charts, where each CUSUM chart is designed to monitor an individual binary stream that follows binomial distribution.…”

Quality Assurance through Process Improvement - A Concise Review

“…The sensitivity to detect process shift is least for Shewhart when compared to EWMA and CUSUM charts. Jirasettapong and Rojanarowan in their work suggested that in broader terms, where the number of streams in a MSP are less than or equal to 5, then CUSUM charts are aptly suitable, whereas if the number of streams in MSP are greater than 5 then, Shewhart chart in combination with X -R charts is suggested [9]. It has been stressed that the type of control chart depends on the interrelation between the process parameters.…”

Statistical Control of Multiple-Stream Processes: A Literature Review

“…A methodological guide for such an analysis can be a very useful contribution. Jirasettapong and Rojanarowan (2011) is the only work I have found on the issue of selecting the most suitable monitoring scheme for an MSP. It considers only a limited number of alternative schemes and a few aspects of the problem.…”

“…Since this was not the case, the target, the upper specification limit (less solid a restriction than the lower one) and the upper control limit of the chart were redefined as a function of the desired false-alarm rate, the point (shift, power) specified, the lower specification limit and the variability between stream means. Jirasettapong and Rojanarowan (2011) were concerned with the selection of the appropriate control charts for monitoring an MSP, considering the degree of correlation among streams, the number of streams, the feasibility of using one chart per stream, whether the streams can or cannot be centred on the same target, and the relevant shift size. They provided a taxonomy, in the form of a tree diagram, to classify the MSPs according to these factors, where the leaves are the recommended control charts.…”