Global Journal of Pure and Applied Mathematics.
ISSN 0973-1768 Volume 13, Number 11 (2017), pp. 7879-7887
© Research India Publications
http://www.ripublication.com
S* Control Chart in Screw Quality Assessment
Shamshuritawati Sharif
1
, Suzilah Ismail
2
and Zurni Omar
3
1, 2, 3
School of Quantitative Sciences, UUM-College of Arts and Sciences,
Universiti Utara Malaysia, 06010 UUM Sintok, Kedah, Malaysia.
1
Orcid: 0000-0002-3305-9661
Abstract
Quality control chart is a powerful tool in assessing quality of products and
services. The assembled of control chart involves two phases. The first phase
function as the quality standard reference which consist of well behave samples
that are within the control limit. The second phase is the real samples that need
to be quality tested. In this study we constructed S* control chart based on the
two phases using screws data. The advantage of this chart is fulfilling the
requirement of high dimensional data set where the number of dimensions is
more than the number of sample sizes (p > n) as displays in screw. The samples
taken involved only three screws (n = 3) but seven variables (p = 7) measuring
the quality of the screws. S* control chart was successfully constructed and able
to act as warning signals in detecting defects screw.
Keywords : large dimension, multivariate control chart, screw production,
small sample size, statistical process control
INTRODUCTION
Statistical process control (SPC) is known to be effective methods for acquiring the
quality standard of products or services through monitoring and controlling the process
(Djauhari, Lee and Ismail, 2014). Originally, the application area of SPC is focusing in
the manufacturing sector. Later, the expansion involving other areas, such as
environmental science (Zimmerman, Dardeau, Crozier & Wagstaff, 1996); engineering
(Mason & Young, 2002) and health care (Myles, German, Wilson and Wu, 2011;
Smith, 2013).
One of the SPC methods is control chart and has a strong point due to the quality
controller is able to decide based on valid real time results instead of doing assumptions
using “last month” performance (Myles et al, 2011). A control chart can be constructed