Title: A new probability-based multivariate process capability index

Authors: Kailas Govinda Khadse; Ramkrishna Lahu Shinde

Addresses: Department of Statistics, School of Mathematical Sciences, M.J. College (Autonomous), Jalgaon 425-001, India ' Department of Statistics, School of Mathematical Sciences, Kavayitri Bahinabai Chaudhari North Maharashtra University, Jalgaon 425-001, India

Abstract: Multivariate process capability indices (MPCIs) MC_p, MC_pk and MC_pm have been proposed in literature to measure multivariate process capability. From a practitioner's view, the intention of this paper is to propose a new probability-based MPCI MS_pmk which takes into account process variability, departure of the process mean from target and proportion of conformity when assessing multivariate process performance. The MS_pmk is the extension in multivariate setup of univariate process capability index S_pmk. The S_pmk is smooth version of widely used univariate process capability index C_pmk. The MS_pmk is very easy to compute and combines the merits of MPCIs MC_p, MC_pk and MC_pm. Proposed MPCI assumes that production process is multivariate normal and specification region is rectangular. Proposed MPCI is recommendable with respect to its performance in various scenarios and its natural estimator statistical properties for real world usage.

Keywords: probability-based multivariate process capability index; multivariate normal process; rectangular specification region; conformity proportion; multivariate process performance; multivariate process capability measures.

DOI: 10.1504/IJQET.2021.116757

International Journal of Quality Engineering and Technology, 2021 Vol.8 No.3, pp.249 - 267

Accepted: 10 Apr 2021
Published online: 01 Aug 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article