SINERGI Vol.
No.
June 2018: 113-119 DOAJ:doaj.
org/toc/2460-1217 DOI:doi.
org/10.
22441/sinergi.
SUPPLIER SELECTION BASED ON CAPABILITIES INDEX FOR
MULTIPLE CHARACTERISTICS WITH ONE SIDED SPECIFICATION
Erik Bagus Prasetyo.
Nani Kurniati Industrial Engineering Departement.
Faculty of Technology Industry.
Institut Teknologi Sepuluh Nopember Jl.
Arif Rahman Hakim Sukolilo.
Surabaya 60111 Email: erikbagusp@gmail.
com nanikur@gmail.
Abstract -- Raw materials are a significant requirement in the production process for manufacturing In meeting the needs of raw materials for the production process, most manufacturing firms rely on suppliers.
Supplier selection is an essential part of manufacturing companies.
From several supplier selection criteria, quality is one of the fundamental standards and is used in supplier Selecting suppliers based on the quality of their products will have a positive impact on manufacturing companies, such as increased profits through reduced operational costs and increased market share.
The problem faced is the lack of accuracy in choosing qualified suppliers.
In this study will compare two suppliers at manufacturing companies and pick one that has a higher capability Supplier selection is made by using multiple characteristic capability index C pl es.
The Supplier will be selected by comparing the ratio of two suppliers.
Numerical calculations are performed on leather suppliers in shoe companies based on bursting quality, tear strength, tensile strength and The result of the calculation can be seen that supplier B is chosen as a better supplier.
Characteristics of quality will affect the production process and application of shoes.
Keywords: Supplier selection.
Quality.
Capability index.
Multiple characteristics.
One-sided Received: January, 18 2018
INTRODUCTION
Manufacturing relationships with many parties, one of which is a The Supplier is a company that provides material that cannot be provided by the manufacturing company itself (Mitrega, et al.
Santoso and Besral, 2.
Manufacturing company must have the ability to choose the right supplier for succeeding.
Supplier selection is a fundamental and critical decision for companies (Kuo and Lin, 2012.
Rezaei and Davoodi, 2012.
Wu et al.
, 2.
The decision in choosing a supplier impact directly on the competitiveness of the company and accelerates the company's response to market Of the various criteria, quality is considered the most essential factor for supplier assessment (Liao et al.
, 2.
The problem faced is the lack of accuracy in choosing qualified suppliers.
The process capability index provides a numerical measure of the ability of a process to produce goods that meet specified quality requirements.
The advantage of using index capability processes is more accurate and reliable when compared to traditional methods (Pearn and Wu, 2.
Some authors have used index capability process with multiple quality.
Pearn et al.
, .
considers the supplier selection problem for a Revised: March, 13 2018 Accepted: March, 14 2018 normally distributed process with some independent characteristics based on the process capability index C pu .
LITERATURE REVIEW
Process Capability Index Single Characteristic Process capability index has been widely used to measure process capability and is essential for quality improvement activities.
Some process capability index has been developed such as Cp.
CPU.
CPL, dan Cpk (Kane, 1.
USL Oe LSL
USL Oe A C pu = 3A
A Oe LSL
C pl = 3A
EUSL Oe A A Oe LSL E
C pk = min E
3A 2 E
E 3A
Cp = .
where USL and LSL respectively are upper and lower specification limits.
AA is the process mean.
E is the standard deviation of the process.
Index Cp only measure the distribution of distribution .
rocess precisio.
, which only reflects the Prasetyo and N.
Kurniati.
Supplier Selection Based on Capabilities Index SINERGI Vol.
No.
June 2018: 113-119 consistency of product quality characteristics.
Index Cpk taking into account the magnitude of the process variance as well as the level of the average specification limits.
Cp and Cpk used to measure the process with two sides of the specification, i.
Lower Specification Limit dan Upper Specification Limit.
Cpu and Cpl designed specifically for processes with one specification that only requires USL or LSL only.
Cpu is an index that measures the ability of a process smaller-the-better with Upper Specification Limit (USL), while Cpl is an index that measures the ability of a process larger-the-better with Lower Specification Limit (LSL).
Process Capability Index Multiple Characteristics for One Side Specifications Wu and Pearn .
discusses multiple one-sided specifications with upper specification limits and proposes a process capability index for smaller the better as.
EE v C Tpu = A Oe1 EEi A ( 3C puj ) E EE j =1 EE where Cpuj show the value Cpu of characteristics jth for j = 1,2, .
, v and v is the number of The relationship between the index C pu and overall process yield P can be defined as.
P = Ei Pj = Ei A ( 3C puj ) = A 3C Tpu j =1 j =1 .
Overall process yield in parts per million (PPM) can be given as follows, yield = 106 C A 3C Tpu Pearn CI Tpu Taylor expansion for the following multiple variables.
E T 1 C Tpu 2 E C pu C N E C pu , .
The above method can be used for processes that only have many lower specification limits (LSL) with exact mathematical The previously mentioned results can be implemented to compare two suppliers with index values CI pu1 and CI pu 2 .
Then Pearn & Wu .
shows the ratio of 2 .
natural estimators as follows.
CI Tpu 2
CI T
Thus, the test statistic distribution R is the result of two normally distributed random variables and therefore is related to the Cauchy distribution.
Using the Jacobian transformation and the convolution approach, the probability density function R can be obtained as.
E A32 E
E A3 E E EE
1 EE 2
E2A 3 exp E Oe 2 E A3A 3 2A E1 Oe 2A E E E E
2AA 1A 2 EE
E 2A 3 E
E A 3 E EE EE
f R .
) = E 1 E A 2 A 2 A 2 EE
C exp E Oe E 12 22 Oe 32 E E
EE 2 E A 1 A 2 A 3 E EE
A1 = C Tpu1 .
A2 = C Tpu 2 .
A 12 = A3 = .
A1 r A2
A 12 A 22
A 22
T 2
T 2
1 C pu1
1 C pu 2
A 22 =
9n1 2n1
Oe1
r A 2A 12 A1A 22 2 E 1 r 2 E
A 2A 2
A 3 = E 2 2 E = 2 12 2 2
r 2A 12 A 22
1 A2
E 1
For every single characteristic, the value Cpuj can be estimated using natural estimator.
CI puj (USL Oe x ) , j = 1, 2,.
, v where x j = mean sample characteristics jth, s j = standard deviation of sample characteristics to jth and estimators of CI pu defined as.
RESEARCH METHODS The histogram is made with a one-sided specification limit and a normal probability plot of skin quality data collection with bursting, tear strength, tensile strength and elongation characteristics for supplier A and supplier B to determine the position and distribution of data.
The next step is to calculate the value of C pl from EE v CI Tpu = A Oe1 EEi A ( 3C puj ) E EE j =1 EE each supplier C pl1 and C pl 2 .
EE v CI Tpli = A Oe1 EEi A ( 3C puj ) E , i = A,B EE j =1 EE Prasetyo and N.
Kurniati.
Supplier Selection Based on Capabilities Index
ISSN: 1410-2331
A Oe LSL , i = A,B C pli = i 3A i LSL lower specification limit mean sample Ai standard deviation sample To compare the yield process of two suppliers, a hypothesis test was performed for the ratio of two indices yields as follows.
C Tpl 2
H0 =
H1 =
C Tpl1 C Tpl 2 C Tpl1 After the hypothesis then calculate the ratio of statistical tests R based on the standard approach to the distribution C pl C Tpl 2 C Tpl1 Analyze is done to test ratio and critical value with supplier requirement C pl = 1.
For supplier A and supplier.
B is calculated the mean sample, standard deviation sample, an index C Tpli for each of the characteristics obtained from Hold the test specimen flat between the jaws of the tensile testing machine so that the slit is aligned and parallel with the axis of the Clamp one of the legs in the lower jaw and then fold the other leg upwards through 180A and clamp it into the upper jaw.
In each case ensure that the end of the leg is parallel with the clamping edge of the jaw and that the slit is positioned in the axis of the tensile tester.
Tensile Strength Cut the specimen with a dumbbell shape.
Attach it to the tensile test machine.
Elongation The procedure follows a tensile test by installing an extensometer.
RESULTS AND DISCUSSION
In this study using bursting quality data, tear strength, tensile strength and elongation of two suppliers.
Each supplier has 150 data for each quality characteristic.
The minimum specification limit of each quality characteristic for bursting = 20 kg/cm2, tear strength = 10 Newton, tensile strength = 60 Newton, dan elongation = The processing of the leather itself determines the quality characteristics of the Characteristics of quality bursting, tear strength, tensile strength and elongation will significantly affect the production process and application of the use of shoes.
In this study using the limit with Lower Specification Limit (LSL).
The following is presented data from each the data with n1 = n2 =150 Histogram of Bursting Supplier A The following hypotheses can be used to select suppliers, .
H1 : C Tpl 2 A C Tpl1 So that can be obtained which supplier is better and will be prioritized.
The step of the experiment:
Bursting Test Cut test specimen with a diameter of 4.
5 cm.
Attach it to the lastometer testing machine.
Observe the specimen until it cracks.
Tear Strength Cut test specimens with the slit parallel to the long direction of the material .
ackbone direction for leather and selvage .
or machine direction.
Mark the along direction of all the test specimens.
Zero the tensile tester
force measuring system and move the jaws together to enable the test specimen to be Mean
StDev
26,46
0,9511
Frequency CC
H0 : C
Normal
LSL
Bursting Supplier A Figure 1.
Histogram of Bursting supplier A Prasetyo and N.
Kurniati.
Supplier Selection Based on Capabilities Index SINERGI Vol.
No.
June 2018: 113-119
Probability Plot of Bursting Supplier A Histogram of Tensile Strength Supplier A Normal
Normal
Mean
StDev
P-Value 26,46 0,9511 0,272 0,667 Mean
StDev
62,99
1,888
Frequency Percent
LSL
Bursting Supplier A Tensile Strength Supplier A Figure 2.
Probability plot of Bursting supplier A Figure 5.
Histogram of Tensile Strength
Histogram of Tear Strength Supplier A Probability Plot of Tensile Strength Supplier A Normal
Normal
LSL
Mean
StDev
15,71
0,6796
Percent Frequency Mean StDev P-Value 62,99 1,888 0,485 0,224 Tear Strength Supplier A Tensile Strength Supplier A Figure 3.
Histogram of Tear Strength supplier A Figure 6.
Probability plot of Tensile Strength supplier A Probability Plot of Tear Strength Supplier A Histogram of Elongation Supplier A Normal
Normal
Mean
StDev
P-Value LSL
Mean
StDev
75,72
0,9275
Frequency Percent
15,71
0,6796
0,265
0,691
Tear Strength Supplier A Figure 4.
Probability plot of Tear Strength Elongation Supplier A Figure 7.
Histogram of Elongation supplier A Prasetyo and N.
Kurniati.
Supplier Selection Based on Capabilities Index
ISSN: 1410-2331
Probability Plot of Elongation Supplier A Histogram of Tear Strength Supplier B
Normal
Normal
Mean
StDev
P-Value LSL
Mean
StDev
Frequency Percent
75,72
0,9275
0,497
0,209
15,73
0,6405
Elongation Supplier A Figure 11.
Histogram of Tear Strength supplier B
Histogram of Bursting Supplier B
Probability Plot of Tear Strength Supplier B
Normal
LSL
Mean
StDev
26,52
0,9131
Percent Mean StDev P-Value Frequency Figure 8.
Probability plot of Elongation Normal
Tear Strength Supplier B
15,73
0,6405
0,348
0,473
Bursting Supplier B Tear Strength Supplier B Figure 9.
Histogram of Bursting supplier B Figure 12.
Probability plot of Tear
Strength supplier B
Probability Plot of Bursting Supplier B
Histogram of Tensile Strength Supplier B
Normal
Normal
Mean
StDev
P-Value 26,52 0,9131 0,580 0,130 LSL
Mean
StDev
63,11
0,8295
Frequency Percent Bursting Supplier B Gambar 10.
Probability plot of Bursting Tensile Strength Supplier B Figure 13.
Histogram of Tensile Strength Prasetyo and N.
Kurniati.
Supplier Selection Based on Capabilities Index SINERGI Vol.
No.
June 2018: 113-119 capability index value C pl for each supplier can Probability Plot of Tensile Strength Supplier B Normal Mean StDev P-Value 63,11 0,8295 0,470 0,244 Percent be calculated.
For supplier A, the sample mean, the sample deviation standard of each is already known, and the first step is to calculate the value C pl for bursting, tear strength, tensile strength, dan elongation use the following formula.
C pl = Tensile Strrength Supplier B A Oe LSL The result of the calculation C pl for supplier A as Figure 14.
Probability plot of Tensile Strength supplier B follows, bursting = 2,2654.
tear strength = 2,8004.
tensile strength = 0,5287.
2,0563.
Then calculate the value for Histogram of Elongation Supplier B
Normal
LSL
Mean
StDev
75,68
0,8544
Frequency EE v
CI TplA = A Oe1 EEi A ( 3C puj ) E
EE j =1 EE
= 0,5286
As for supplier B, the sample mean, the standard deviation of each sample is known, and the first step is to calculate the value of Cpl for bursting, tear strength, tensile strength, and elongation using the following formula.
Elongation Supplier B Figure 15.
Histogram of Elongation supplier B
Probability Plot of Elongation Supplier B
Normal
Mean
StDev
P-Value Percent
75,68
0,8544
0,501
0,204
A Oe LSL The result of the calculation Cpl for supplier B as follows, bursting = 2,3815.
tear strength = 2,9828.
tensile strength = 1,2505.
2,2147.
Then calculate the value for C Tpl as EE v
CI TplB = A Oe1 EEi A ( 3C puj ) E
EE j =1 EE
C pl = Elongation Supplier B
= 1,2505
Figure 16.
Probability plot of Elongation From Fig.
1 to Fig.
16 shows the histogram with lower specification limits and standard probability plots with Anderson-Darling test from each supplier A and supplier B data.
From histogram data, it can be seen that some quality tensile strength values in supplier A do not meet the requirements .
Based on the above measurements, the sample means, sample deviation standard and C Tpl as The hypothesis used to compare from two suppliers is as follows.
H0 = H1 = C TplB C TplA C TplB C TplA While the calculation of R statistic test ratio based on the normal approach to the distribution C Tpl as follows.
Prasetyo and N.
Kurniati.
Supplier Selection Based on Capabilities Index
ISSN: 1410-2331
C TplB
C TplA
1, 2505
0,5286
= 2,3656
The critical value of = 0.
05 with the sample value n = 150 is 1.
8047 (Pearn et al.
From the above ratio, values note that R > R = 2.
3656 > 1.
8047 then H0 will be rejected, and it can be concluded that the capability index of the process differs significantly by = 0.
Then the hypothesis to select a supplier based on the capability index C pl .
H 0 : C TplB C C TplA H1 : C TplB A C TplA From the results of the capability index pl obtained C TplB = 1,2505 < C TplA = 0,5286, then H0 rejected, so from these results can be derived a better supplier B and will be prioritized.
CONCLUSION
Effective supplier selection will significantly determine success for manufacturing companies.
From several supplier selection criteria, quality is one of the preferred criteria in supplier assessment.
Selecting a supplier based on the quality of its product will have a positive impact on the manufacturing Capability process index is an important criterion used in the manufacturing industry to measure process performance.
Capability C Tpl provides a measure with a process index one-sided specification limit for normal processes and provides a precise numerical measurement of process performance on suppliers.
Higher accuracy to assess the two suppliers is obtained using the capability index C pl .
From the above C TplA = 0,5286 and C TplB = 1,2505.
Process capability index C TplB greater than C TplA , so that supplier B is better to choose.
REFERENCES