AbstractStatistical process control (SPC) is one of the

important tools in quality control (QC). Commercial software

packages are available to implement SPC. However, they are costly
and not necessary customize to specific need of each production
process. Moreover, they might be difficult for the field operator with
limited knowledge of SPC to use the commercial software. In this
paper, a visual basic application was developed specifically to meet
the demand of a case study company in order to assist the
implementation of SPC. The case study company is the manufacturer
of metal frame for actuator. The software are able to construct control
chart, both x -R chart and x -S chart. These charts are common
tools used to determine whether the production process is in control
i.e. no assignable causes exist. Moreover, the program is able
construct histogram. The software is designed to store process data in
SQL database which can be access online. The result indicated that
as the program is configured specifically to the process, it can help
operator to implement SPC easily even with limited knowledge of
SPC.

KeywordsStatistical process control, Control chart, x - R chart
and x - S chart
I. INTRODUCTION
UALITY control (QC) is an important function in factory
as it deals with product inspection before the product was
shipped to customers. Statistical process control (SPC) is
one of the tools widely used in QC to monitor whether the
production process is in control through the use of statistical
control chart. This paper focused on the implementation of
SPC in the case study company which is a manufacturer of
electronics parts. A particular product investigated was the
metal frame for actuator which is a critical part in hard disk
drive product. SPC in the case study company is currently
implemented using commercial software package. This
software is used to analyse production process data with
control chart and histogram. However, only a few computers
are installed with this software because software license is
expensive. Moreover, the commercial software is not design
specifically to the process hence sometimes difficult for the
operator with limit knowledge of SPC. This research aims to
develop software for control chart and histogram analysis that
designed specifically to previously mentioned process by
using Visual Basic. This software helps operators in quality

1
The author is with the Industrial Engineering Department, Chiangmai
University, Chiangmai, Thailand (e-mail: [email protected]).
2
The author is with the Industrial Engineering Department, Chiangmai
University, Chiangmai, Thailand (e-mail: [email protected],
[email protected]).
control division to construct and analyze control chart and
calculate process capability. The software also provide
warning if the process is out of control.
Visual basic has been used widely by researchers to develop
various applications. Visual basic program is a computer
language in visual basic studio 2005. It is a tool for software
development from Microsoft. The advantages of visual basic
are for example, visual basic is an advance object oriented
programing, it use less code for writing and it support for
Microsoft Window and SQL server. Moreover, there have
been a number of reports on the success of software
development using Visual Basic. For example, Patane and
Ferrari [1] use a visual basic computer program for interactive
and automatic earthquake waveform analysis. Carr [2]
presents a visual basic program for principal components
transformation of digital images, Argyris et.al. [3] present a
Computer program in Visual Basic language for simulation of
meshing and contact of gear drives and its application for
design of worm gear drive. Nelson [6] presents a visual basic
program for flexible and interactive processing of ion-
microprobe data acquired for quantitative trace element,
Ekneligoda and Henkel [5] presents a visual basic program to
calculate spatial properties of lineaments, Wang et.al [7]
present a visual basic program for simulating distribution and
atmospheric volatilization of soil fumigants applied through
drip irrigation, Poppe and Eliason [8] presents a visual basic
program to plot sediment grain-size data on ternary diagrams,
Balbas et.al. [9] present a visual basic code for optimal
forecasting of financial time series and Kumar et.al. [10]
present a program in visual basic for predicting haulage and
field performance of 2WD tractors.
Due to the previously mentioned advantages and also a
number of successful reported Visual basic studio 2005 was
chosen as a development tool for the SPC software of the case
study company. This software is capable of construction
control chart and histogram for quality inspection data
graphically. The software will be linked to the case study
company database system. SQL server was used for database
design.
II. BACKGROUND
A. X

- R Chart
The x -R control chart is a common tool used in industry to
control process mean. As a result, the x -R control chart has
become a topic of interest to researcher. For example, Senturk
et.al. [14] use fuzzy set theory to set control limit of the x -R
Visual Basic Application for Statistical Process
Control: A Case of Metal Frame for Actuator
Production Process
Wimalin Laosiritaworn
1
, Tunchanit Bunjongjit
2

Q
Proceedings of the International MultiConference of Engineers and Computer Scientists 2010 Vol III,
IMECS 2010, March 17 - 19, 2010, Hong Kong
ISBN: 978-988-18210-5-8
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)
IMECS 2010


control chart. Fuzzy control limits provide a more accurate
and flexible evaluation.
x -R control chart should be used with data that are
normally distributed. Parameter x is the average of sample and
x is the grand average of sample. They can be calculated as
follows;

x =
x
1
+x
2
++x
n
n
, x =
x
1
+x
2
++x
m
m
(1)

Parameter n is a sample size, parameter x is the value of
sample measurement. Parameter m is the number of sample
measurement. Then the range (R) is calculated from the
different between largest and smallest value of sample using
the following formula;

R = x
max
x
min
, R

=
R
1
+R
2
++R
m
m
(2)

The control limits of the x chart are calculated from

uCI = x + A
2
R

, CI = x , ICI = x - A
2
R

(3)

The control limits of the R chart are calculated from

uCI =
4
R

, CI = R

, ICI =
3
R

(4)

Parameter A
2
, D
3
and D
4
are the constants for construction
of x and R charts.
B. X

-S Chart
The x -S control chart is calculated from standard deviation.
It is popular for control process mean as reported by a number
of publications. For example, Takemoto et.al. [12] proposed
the multivariate x -S control chart to monitor process mean
vector and covariance matrix simultaneously by using
approximation procedure for determining the control limit of
multivariate x -S control chart and for evaluating its power.
Shamsuzzaman et.al. [13] proposed the x -S control chart
apply to the x -S M control chart for optimizing total cost
because x -S M control chart decrease sample size. x -S control
chart should be used when sample size is more than 10.
Parameter S is the standard deviation and S

is the average of
standard deviation. Their formulate are;

s =
_
(x
i
-x )
2 n
i=1
n-1
, s =
s
1
+s
2
++s
n
m
(5)

The control limits of the x chart are calculated from

uCI = x + A
3
s , CI = x , ICI = x - A
3
s (6)

The control limits of the s chart are calculated from

uCI = B
4
s , CI = s , ICI = B
3
s (7)

Parameter A
3
, B
3
and B
4
are the constants for construction
of x and s charts.
C. Histogram
Histogram is a bar chart display the frequency of data. The
horizontal axis of histogram is the data bin of observed value
arranged from minimum bin to maximum bin. The vertical
axis is the frequency of collected data. Histogram can be used
to analyse the shape and distribution of data of known
variance. Histogram can be used to check abnormal data,
check out of specification, calculate process capability and
monitoring process. In order to draw histogram, firstly the
range of data has to be calculated from the different between
maximum and minimum data. Then number of data(a),
observed bin(b) have to be specified. Then the number of
bin(K) and bin width (W) can be calculated as follows;

K = a, W =
Rangc
K
(8)

Then the upper class interval and lower class interval are
calculated from;

Upper class interval of b= [maximum value of (b-1)
+ minimum value of b]/2 (9)

Lower class interval of b= [maximum value of b
+ minimum value of (b+1)]/2 (10)

The Final Stepis to draw histogram by disperse frequency of
data.
Process capability can be measure from process potential
capability (C
p
) and process performance capability (C
pk
) which
can be calculated using the following formula;

o =
R

d
2
, C
p
=
0SL-LSL
6c
, C
pk
= min{
0CL-X

3c
,
X

-LCL
3c
] (11)

Where, is the standard deviation. Both process potential
capability and process performance capability with value more
than 1.33 indicated good capability of the process.
Some recent research on histogram are for example, the
work by Strelkov [16], where a new measure to calculate the
similarity between the histogram and the peak matching
measure was proposed. It describes the closeness of the
positions of the peaks in the compared histograms and
similarity of the peak shapes. Zhen He et.al. [17] developed a
novel multi-dimensional histogram method called PR
Histogram. It is the first method that incorporates proactive
histograms for selectivity estimation. Xin lu et.al. [15]
investigated histogram construction for selectivity estimation
based on a new criterion aggregate error. The histogram
construction algorithms were developed based on relative
aggregate error minimization and demonstrate the
effectiveness and efficiency of the algorithm by extensive
experiments over synthetic and real world datasets. Strauss
[18] proposed the construction of histogram from a set of
observations on a fuzzy partition of the real line. Such
histogram is able to account for precise and imprecise
observations. A known imprecision in observation can easily
be transferred to the histogram. This counting transfer
combined with the distributed vote is called the quasi-
continuous histogram technique.
Proceedings of the International MultiConference of Engineers and Computer Scientists 2010 Vol III,
IMECS 2010, March 17 - 19, 2010, Hong Kong
ISBN: 978-988-18210-5-8
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)
IMECS 2010



D. Visual Basic Application
Visual Basic language evolves from Beginners All-
Purpose Symbolic Instruction Code. Later on, Microsoft
developed Visual Basic from Basic and makes it more
graphical. Visual basic language has the same syntax as visual
basic.NET. Graphic of visual basic is the coordinate system
about x and y. Many researcher use visual basic application
for building specification software. For example, Chen [4]
present MRPM program, which is an integration of three
Visual Basic program for implementing mineral resource
potential mapping with the above three models. The software
is a user-friendly interactive computer program executable
directly on Window system. Wang [7] use a graphical user
interface program in visual basic to simulate distribution and
volatilization of soil fumigants when applied through
subsurface drip irrigation and the program provides an
intuitive user interface by linking databases with default soil
and chemical properties to predict subsurface distribution
patterns.
III. METHODOLOGY
Figure 1 is the flow chart of SPC software First, inspector
input data into the database manually. Then, this data will be
retrieved from the database, which can be access from all
computers in the factory. The software process is as follows:
Step one: An operator inputs measured data into the
database.
Step two: An operator input key data, such as the data
analysis date, the arm number, sampling size, type of
measurement, type of machine used for measurement and the
type of product, plus a code number, into a control chart and
histogram software for call data from database.
Step three: The software retrieves data from the database. If
the data is shown as zero then the operator must input the
key details again.
Step four: The software retrieves data from the database.
The data shown is: an average of x
i
i, an average of R
i
, an
average of S
i
, a maximum of x
i
, a minimum of x
i
and the
measurement data. If the data is shown as zero, the operator
must key in the important data analysis data again.
Step five: The software calculates the statistical process
control, which consists of an upper control limit, a center line,
a lower control limit, a process potential capability and a
process performance capability.
Step six: The software calculates a value in order to plot a
control chart.
Step seven: The software detects abnormal data by using the
laws within the control chart.


Fig. 1 Flow chart of control chart software

IV. RESULT AND DISCUSSION
A. Database Design with SQL server
The software must transfer data to the database when the
operator measures the product, so the database is very
important for this research. Each industry has a different
system. This research is about metal frames for an actuator
factory, so the programmer must understand the system in
place at this factory.
Yes
No
Yes
No
Data from
measurement
Input Data
Database
START
Calculation average Ri, Si, Xi,
maximum of Xi, minimum of Xi and
Data measurement
Enter Date for analysis data, Arm number,
sampling size, type of measurement, type of
measurement machine, type of product.
Receive process and
customer specification
Plot control chart and histogram by click button
Detect abnormal data by laws of control
chart by click x chart button
Show UCL, CL, LCL, , Cp and Cpk
Number of data > 0
Receive average Ri, Si, Xi, maximum of Xi,
minimum of Xi and Data measurement
Number of data > 0
Calculation UCL, CL, LCL, , Cp and Cpk
STOP
Proceedings of the International MultiConference of Engineers and Computer Scientists 2010 Vol III,
IMECS 2010, March 17 - 19, 2010, Hong Kong
ISBN: 978-988-18210-5-8
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)
IMECS 2010



Fig. 2 Database design of metal frame for actuator factory by SQL
server

Figure 2 shows the design of the database. The
XbarRchart table shows the data inputted into the database
such as the: data number, lot number, product ID, date of
measurement, data sampling, measurement ID, employee ID,
sampling size and arm number and preliminary calculations
such as the: average of x
i
, maximum of x
i
, minimum of x
i
, R
i

(minimum and maximum range), S
i
(standard deviation). The
Employee Table shows a list of employees that measured the
product in another measurement machine. The Management
User Table is a list of those highly responsible employees
who have access to and the ability to correct data in the
database. The Measurement Detail Table contains the types
of measurement including the identification number and name
of the measurement. The Measurement MC Detail Table
shows the type of machine or instrument used for
measurement and includes an identification number and the
name of the measurement device. The Product Detail Table
contains details about the type of product, including the
identification number and the name of the product. The Spec
Detail Table shows the specification of the product each
customer required.
B. Input Data Form
The Data Input Form is designed for operator use. The
user must input the arm number, sample size, type of
measurement and the type of machine into the software.
Before storing the data on to the database, the program
calculates the maximum and minimum value, the range value,
average value and standard deviation. They had kept in
database for manage database easily and for the next
calculation.


Fig. 3 Input form can input six arms and six sample sizes.

Figure 3 shows the Data Input Form. In this window the
sample size is between two and six. This sample size depends
on the type of product.
C. Output Data and Analysis Form
This software must create a Data Output and Analysis
Form in order to calculate a x - R chart, a x - S chart and a
histogram. It must also create a database, identify abnormal
data in the control chart and show the lot number for items
outside of control; these being the abnormal items within the
control chart.

Fig. 4 Example of histogram from the software

A histogram is important for analysis of distribution of the
data. Figure 4 is an example of histogram from the software.
This data has normal distribution. The highest frequency is
between 0.4394415 and 0.4394935. The histogram has 12
levels. A histogram is also used for frequent analysis of data.
The benefit it provides is related to process improvement. For
example, an engineer can adjust the numerical control
machine, the calibration measurement machine, the
measurement check instrument and increase the steps within
the quality control division process, in order to improve
product quality.
Proceedings of the International MultiConference of Engineers and Computer Scientists 2010 Vol III,
IMECS 2010, March 17 - 19, 2010, Hong Kong
ISBN: 978-988-18210-5-8
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)
IMECS 2010



Fig. 5 Example of Control chart from the software
Figure 5is an example of control chart obtained from the
software. This control chart have upper control limit at
0.4408451, centerline at 0.4394702 and lower control limit at
0.4380953. It shows no data that is out of the control limits.
The 30/9/2552 refer to November 30 in 2009, the 1:43:19
refer to time, the Lot1 refer to lot number one. The control
chart also shows the date of quality control measurement, and
it measures four times per lot. It can show the customer
specification and the production process specification in a
graph. Another chart can be opened by clicking a button
within the software.
D. Laws of control chart
The Control Laws Chart is important for process control,
because it can analyze the data and detect lots with low
product quality. The old process of quality control analyzes
abnormal data by looking at the control chart, something
which is time-consuming. Therefore, this software makes a
code from the basic visuals to automatically detect abnormal
data. Parameter z1to z9 is the average data of the process,
UCL is the upper control limit, LCL is the lower control limit
and CL is the center line. The Basic Visual Code is shown in
Table 1.
TABLE I
CODE IN VISUAL BASIC FOR LAWS IN CONTROL CHART
Laws CODE OF VISUAL BASIC
6 points Increase z1<z2 and z2<z3 and z3<z4 and z4<z5
and z5<z6
6 points Decrease z1>z2 and z2>z3 and z3>z4 and z4>z5
and z5>z6
9 points
Center line to +3
z1>CL and z2>CL and z3>CL and
z4>CL and z5>CL and z6>CL and
z7>CL and z8>CL and z9>CL
9 points
Center line to -3
z1<CL and z2<CL and z3<CL and
z4<CL and z5<CL and z6<CL and
z7<CL and z8<CL and z9<CL

Out of control z1>UCL or z1<LCL
From Table 1, an increase or decrease of six points is
important, because it means the data may be outside of
control. The law of nine points is important for this factory,
because it reflects customer demand. Being out of control is
dangerous for the production process. Sometimes, the value of
the data may show it at out of control, but it is actually still
within the specification of the customer, so the product is not
in actual fact a defect. Therefore, this depends on the decision
of an engineer.
V. CONCLUSION
A control chart and histogram software was developed from
Visual Basic to calculate and show graphically the control
chart and histogram. This software is used in the quality
control division. The weak point of this software is that the
data must be less than one hundred. The specification of the
software is to retrieve data by the date of measurement range.
This software can make an x -R chart, an x - S chart and a
histogram to calculate C
p
and

C
pk
for the process efficiency to
be used by the engineer in a metal frame actuator factory. The
program receives data by linking databases. The database used
for this research was developed from an SQL server, because
it supports Visual Basic program development and Microsoft
Windows. Another reason of using an SQL server is that this
factory has a license for an SQL server 2000 and for Microsoft
Windows XP. The objective of this research is to make
software that can be used and operated conveniently. Potential
benefit of the software to the case study company includes;
decrease cost for buying commercial software package
because software license is expensive. Moreover, a control
chart and histogram software is design specifically to the
process hence easy for the operator with limit knowledge of
SPC to use.
ACKNOWLEDGMENT
This project is financially supported by the
Industry/University Cooperative Research Center )I/UCRC (in
HDD Component, the Faculty of Engineering, Khon Kaen
University and National Electronics and Computer
Technology Center, National Science and Technology
Development Agency.
REFERENCES
[1] Domennico Patane and Feruccio Ferrari, Seismplo A Visual Basic
computer program for interactive and automatic earthquake waveform
analysis, Computer & Geosciences Vol.23, No.9, pp. 1005-1012, 1997.
[2] Jame R. Carr, A Visual Basic program for principal components
transformation of digital images, Computer & Geosciences Vol.24,
No.3, pp.209-218, 1998.
[3] J. Argyris, M. De Donno, F.L. Litvin, Computer program in Visual
Basic language for simulation of meshing and contact of gear drives and
its application for design of worm gear drive, Coput. Methods Appl.
Mech. Engrg. 189 pp.595-612, 2000.
[4] Yongliang Chen, MRPM three visual basic program for mineral
resource potential mapping, Computer & Geosciences 30 pp.969-983,
2004.
[5] Thushan C. Ekneligoda and Herbert Henkel, The spacing calculator
software A Visual Basic program to calculate spatial properties of
lineaments, Computer & Geosciences 32 pp.542-553, 2006.
[6] David R. Nelson , A Visual Basic program for interactive process of
ion microprobe analytical data, Computers & Geosciences 32, pp.1479-
1498, 2006.
Proceedings of the International MultiConference of Engineers and Computer Scientists 2010 Vol III,
IMECS 2010, March 17 - 19, 2010, Hong Kong
ISBN: 978-988-18210-5-8
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)
IMECS 2010


[7] D.Wang, J.M. He and J.A. Knuteson, A Visual Basic program for
simulating distribution and atmospheric volatilization of soil fumigants
applied through drip irrigation, Computer and Electronics in
Agriculture 56 pp.111-119, 2007.
[8] L.J Poppe and A.H. Eliason , A Visual Basic program to plot sediment
grain-size data on ternary diagrams, Computer & Geosciences 34
pp.516-565, 2008.
[9] Alejandro Balbas, Beatriz Balbas, Inna Galperin and Efim Galperin,
Deterministic regression model and visual basic code for optimal
forecasting of financial time series, Computers and Mathematics with
Applications 56 pp.2757-2771, 2008.
[10] Ranjeet Kumar, K.P. Pandey , A program in Visual Basic for predicting
haulange and field performance of 2WD tractors, Computers and
Electronics in Agriculture 67 pp.18-26, 2009.
[11] C.R. chen, H.s. Ramaswamy, Visual Basics computer simulation
package for thermal process calculation, Chemical Engineering and
Processing 46 pp.603-613, 2007.
[12] Yasuhiko Takemoto and Ikuo Arizono, A study of multivariate x, s
control chart based on Kullback-Leibler information, Int J Adv Manuf
Technol 25 pp.1205-1210, 2005.
[13] M. Shamsuzzaman, Zhang Wu and M. Reaz-us Salam Elias, Design of
x&s control chart with optimal manpower deployment, Computer
&Industrial Engineer 56 pp.1589-1596, 2009.
[14] Sevil Senturk and Nihal Erginel, Development of fuzzy x- r and x- s
control chart using cuts, Information Science 179 pp.1542-1551,
2009.
[15] Xin Lua, Jihong Guan, A new approach to building histogram for
selectivity estimation in query processing optimization, Computer and
Mathmatics with Application 57 pp.1037-1047, 2009.
[16] V.V. Strelkov, A new similarity measure for histogram comparison and
its application in time series analysis, Pattern Recognition Letters 29
pp.1768-1774, 2008.
[17] Zhen He, Byung Suk Lee and X.Sean Wang, Proactive and reactive
multi-dimensional histogram maintenance for selectivity estimation,
The journal of systems and software 81pp.414-430, 2008.
[18] Olivier Strauss , Quasi-continuous histograms, Fuzzy Set and System
160 pp.2442-2465, 2009.

Wimalin Laosiritaworn, PhD, She received her PhD in Manufacturing
Engineering and Operations Management from The University of
Nottingham, United Kingdom. She received her MSc in Engineering Business
Management from The University of Warwick, United Kingdom. She is
currently an assistant Professor in the Department of Industrial Engineering at
Chiangmai University.

Tunchanit Bunjongjit is a master student in the department of Industrial
Engineering from Chiangmai University, Thailand. He obtained his BS in
Industrial Chemistry from Chiangmai University, Thailand. He is interested in
pursuing research in the Quality Control, Statistical Process Control, Artificial
Neural Network and Software Engineer.


Proceedings of the International MultiConference of Engineers and Computer Scientists 2010 Vol III,
IMECS 2010, March 17 - 19, 2010, Hong Kong
ISBN: 978-988-18210-5-8
ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)
IMECS 2010