Proposing a two-criterion quality loss function using critical process capability indices

Abstract

Purpose Average quadratic quality loss function (QQLF) measures quality of a given process using mean shift from its target value and variance. While it has a target parameter for the mean, it lacks a target for the variance revisable for counting any progress of the process across different quality levels, above/below the standard level; thus, it appears too general. Hence, in this research, it was initially supposed that all processes are located at two possible quality spaces, above/below the standard level. The purpose of this paper is to propose a two-criterion QQLF, in which each criterion is specifically proper to one of the quality spaces. Design/methodology/approach Since 1.33 is a literarily standard or satisfactory value for two most important process capability indices Cp and Cpk, its upper/lower spaces are assumed as high-/low-quality spaces. Then the indices are integrated into traditional QQLF, of type nominal the best (NTB), to develop a two-criterion QQLF, in which each criterion is more suitable for each quality space. These two criteria have also been innovatively embedded in the plan-do-check-act (PDCA) cycle to help continuous improvement. Finally, the proposed function has been examined in comparison with the traditional one in Feiz Hospital in the province of Isfahan, Iran. Findings Results indicate that the internal process of the studied case is placed on the lower quality space. So the first criterion of revised QQLF gives a more relevant evaluation for that process, compared with the traditional function. Moreover, this study has embedded both proposed criteria in the PDCA cycle as well. Research limitations/implications Formulating the two-criterion QQLF only for observations of normal and symmetric distributions, and offering it solely for NTB characteristics are limitations of this study. Practical implications Two more relevant quality loss criteria have been formulated for each process (service or manufacturing). However, in order to show the comprehensiveness of the proposed method even in service institutes, emergency function of Feiz Hospital has been examined. Originality/value The traditional loss function of type NTB merely and implicitly targets zero defect for variance. In fact, it calculates quality loss of all processes placed on different quality spaces using a same measure. This study, however, provides a practitioner with opportunity of targeting excellent or satisfactory targets.

The Standardization of administrative processes: a case study using continuous improvement tool

“…It is important to identify common and repeatable tasks since the purpose of this identification is to group the items with the highest repetition frequency and the shortest variation time (Liker and Meier, 2007). Geramian et al (2017) suggest that the PDCA methodology is the best tool to manage established procedures. The Planning, Do, Check and Action (PDCA) methodology intends to manage repetitive processes, establishing a working standard for each stage, from the initial stage of the project to its final product.…”

Fuzzy inference system application for failure analyzing in automobile industry

“…The word target was later interpreted as the target value for quality characteristics, by Taguchi (Phadke, 1995). In fact, Taguchi defined quality through the lack of quality and believed that any deviation of the quality characteristic from its target value causes a loss of quality for both the manufacturer and the customer (Chen et al , 2015; Geramian et al , 2017).…”