Volume 3, Issue 11, November – 2018 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Supplier Selection Criteria Using the Analytical

Network Process Method
Erry Rimawan, Yayan Saputra
Mercu Buana University

Abstract:- Decision making is an important thing that Supplier selection is a very critical activity in purchasing
must be done by a company, because companies that do management in the supply chain, because supplier
not carry out decision-making activities appropriately performance plays an important role in cost, quality,
can result in high maintenance costs, causing complaints delivery and services in achieving the objectives of a supply
from customers. There are many decisions made by a chain (Amiri et al. 2018). According to (Blocher et al.
company, one of which is making decisions regarding the 2002), purchasing managers of an industrial company more
selection of suppliers of goods to be used by a company. often use consideration of the lowest price supply factor to
Selection of suppliers is important, where the selection of choose suppliers among suppliers. One other factor that is
the right supplier of goods, not only suppliers who can also dominantly used is the desire to immediately own and
provide quality material, on time, and affordable prices meet the raw material inventory needs. Decision making in
but also must provide quality, optimal service both in supplier selection like this can be at great risk for raw
terms of responsiveness, flexibility, smooth material use when the company will start its production
communication and information .The method used to process activities.
determine supplier alternative solutions based on several
factors can be done using ANP. ANP or Analytic PT. UT is a distributor company engaged in mining
Network Process is a mathematical theory that allows a that prioritizes the quality of its products. According to
decision maker to face factors that are interconnected observations in the service section, besides good quality,
(dependence) and systematic feedback (feedback).This unit work performance is also an important thing to get a
research was conducted at PT. UT. This study begins quality unit. Therefore, PT. UT has a cylinder component
with a literature study first, and then proceed with supplier that can supply components to mining production
giving a questionnaire to the company, where the results units managed by PT. UT service. Some of the products in
of the questionnaire are used as a reference factor of this the UT Company are dump trucks, excavators, and Scania.
study. After distributing the questionnaire, it can be seen In the service process several parts of the company bring in
that there are several criteria, namely price, delivery, 70% of components from suppliers. That is the cylinder and
quality, flexibility, responsiveness, service, guarantee engine. In fact, it often experiences delay in cylinder
and performance history and obtained sub-criteria component supply, PT. UT has sent a plan for changing
namely giving discounts, bid prices, payment grace, parts to suppliers 3 months before the change is made. So
component strength and others . Supplier cylinder itself that PT. UT has a loss on unit service costs which is caused
has 3 choices of suppliers, namely PT. DU, PT. HT and by the low lifetime of the cylinder component. In addition to
PT. TR. After doing calculations using the ANP method the surcharge costs incurred to rebuild the used components,
with the help of Super Decisions software and manual the prices offered are careless and do not follow the price of
calculations, the results of the selected supplier are the current agreement. From the side seal side, the supplier
obtained from the supplier of PT. TR with manual uses a part that is not genuine from Komatsu. cost loss
calculation priority value 0.4561 and super decision experienced by PT. UT can be shown by the table below:
software 0.43612.

Keywords:- Decision Making, Selection of Supplier, ANP.

I. INTRODUCTION

 Background
In the competitive era, the selection of effective
suppliers plays an important role in the success of an
industry. Choosing the right supplier helps an industry to
control costs and improve the competitive side of the
industry. In addition to the decisions taken are logical,
decisions must also be based on certain theories. We often
encounter supplier selection problems, but often we are
confused to choose the best object, if all the demands of the
criteria appear to be equally important. Even from several
objects have the opposite criteria. For example for the best
price criteria is object A, but for the best quality criteria is
object B and for the best function criteria is object C. Fig 1:- Maintenance Cost Graph

IJISRT18NV85 www.ijisrt.com 216

Volume 3, Issue 11, November – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
 Limitation of Research
The scope of the study is limited to:
1. Research was taken at the cylinder component supplier
selection activity by a multi-supplier service division.
2. Determination of criteria and sub-criteria obtained
from literature studies that are tailored to company
policy.
3. Solving the problem is limited only to provide
suggestions that can be implemented by the company.

II. LITERATURE REVIEW

A. Analytic Network Process (ANP)

The Analytic Network Process (ANP) method is the
development of the Analytical Hierarchy Process (AHP)
method. The ANP method is able to improve AHP
Table 1:- Average Cylinder Substitution Lifetime weaknesses in the form of the ability to accommodate the
Performance interrelationships between criteria or alternatives. Linkages
to the ANP method are of 2 types, namely the
Data from Table 1 can be concluded the phenomenon interrelationships in a set of elements (inner dependence)
of the problem that occurs is that there are many and the interrelationships between different elements (outer
components supplied by supplier A who cannot reach the dependence). The existence of this link causes the ANP
lifetime performance targets that have been agreed at the method to be more complex than the AHP method (Saaty,
beginning of the contract, which affects service costs and 1998). Weighting with ANP requires a model that represents
unit performance. The observation results provide the interrelationships between the criteria and the subcriteria
information that PT. UT needs to rearrange the supplier it has. There are 2 controls that need to be considered in
selection evaluation system. The selection of suppliers also modeling the system that you want to know the weight of.
functions as an evaluation material that can later be used to The first control is a hierarchy control that shows the
increase supplier selection or as a consideration of whether relevance of the criteria and sub-criteria. In this control does
or not to find alternative suppliers. The criteria used must be not require a hierarchical structure as in the AHP method.
adjusted to the needs, objectivity and existence of company Other controls are related controls that indicate the existence
resources. Criteria used by PT. UT has an interrelated of interrelationships between criteria or clusters (Saaty,
relationship with each other so the right method to use is 1996).
ANP. In addition ANP can also be used to predict the form
of relationships between the criteria reviewed. The Weighting with models that represent the
Analytical Network Process (ANP) method is a sequence of interrelationships between the criteria and the subcriteria it
the Analytical Hierarchy Process (AHP) method. ANP's has. There are 2 controls that need to be considered in
strengths from other methodologies are its ability to help us modeling the system that you want to know the weight of.
measure and synthesize a number of factors in a hierarchy or The first control of hierarchical control that shows the
network. The advantages of ANP over AHP are that this relevance of the criteria and sub-criteria. This control does
method is more detailed in analyzing criteria, with more not require a hierarchical structure as in the AHP method.
objective results, more accurate predictions, and more stable Other controls and criteria for interrelationships between
results. criteria or clusters(Saaty,2001). This method is a
development of the AHP method, which allows for
 Research Objectives dependencies between criteria and alternatives that are not
The objectives in conducting this research and writing are: available in the AHP method. With feedback (feedback), all
1. Get the supplier cylinder criteria. alternatives can depend on the criteria, and depend on each
2. Knowing the order of priority factors that influence other between these alternatives.
supplier selection at PT. UT.
3. Determine supplier indicators that best meet supplier B. Basic Principles of the Analytic Network Process
selection criteria that should be chosen by PT. UT based (ANP)Method
on the Analytic Network Process (ANP) method. Decision making using the ANP method is based on several
basic principles, namely (Yurdakul, Mustafa, 2002).
 Benefits of Research
The benefits of research and writing are:  Structure of a network shaped problem.
1. As input for the company in an effort to increase The structure of the problem in ANP does not
productivity through the perspective of company have to have a linear structure from top to bottom
resources, namely material that is good from suppliers. as well as hierarchy, but it is more similar to a
2. As a reference for companies in evaluating increasing network, with a cycle of relationships from its
supplier selection or as a consideration for finding clusters. Comparison of hierarchical and network
alternative suppliers structures is shown in the following figure :

IJISRT18NV85 www.ijisrt.com 217

Volume 3, Issue 11, November – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
component (level), the value of the comparison matrix is
searched for the eigenvalue of the vector. Then the
vector eigenvalue is input into the supermatric. If the
matrix is multiplied by the supermatric itself (raised)
until a stable weight is obtained, the Steady state matrix
will be obtained, where the values of each element show
priority weights that have accommodated all interactions
between components (levels).

 Steps in the Analytic Network Process (ANP)Method

Fig 2:- Differences in Linear and Nonlinear Network The following are the steps for making ANP according
Structures(Saaty, 1996) to Saaty (Saaty, 1999) :
1. Step 1: Construction of the model and structuring the
From the picture above, it appears that the structural problem The main objective is to identify alternatives
differences occur between linear network models
that will be most significant in decision making. For
(hierarchy) and non-linear networks. The hierarchy model more details, the sequence of model development can be
only has a one-way functional dependency relationship, described as follows (Sarkis, Joseph, 2003) :
namely the dependence of the component (level) on the a) Describe the elements of a problem (system). The
bottom of the component (level) at the top. The network
principle of decomposing and defining elements is the
model is able to accommodate reciprocal functional
same as AHP, which is minimum, complete and
dependence (two-way), namely interdependent relationships
operational.
between components (levels) up and down.
b) Formation of components (levels). If there are elements
 Determination of the weight of the element against the that have equal quality grouped into a component (level
reference component or cluster) of the same.
Determination of weights is done using a paired 2. Step 2: Pairwise comparison matrix that shows
comparison matrix. With this comparison matrix, we relevance. In ANP the approach in making permanent
will get the weight of the comparison between the decisions is based on the decision to get priority as well
elements in a component (level) of the element that as the AHP method. A group of experts develop a scale
becomes the reference for assessment. As by using the that can describe a decision process so that it can
AHP method, with this comparison matrix can be produce the best decisions. Saaty (1980) sets quantitative
tracked the consistency of the assessment of a user. To scales 1 to 9 to assess the comparison of the importance
get the priority sequence between elements of a of an element to other elements (Saaty, 1996).

Interests Definition Explanation

1 Just as important Both elements have the same effect
Experience and judgment favor one element
3 A little more important
compared to the other
Experience and judgment favor one element
5 More Important
compared to the other

One element is very popular and practically its

7 Very Important
dominance is very real compared to its partner

One element proved to be absolutely preferred

9 Absolute more important compared to its partner, at a high level of
confidence

This value is given if there is doubtful judgment

2,4,6,8 Middle value
between two adjacent assessments

If activity i obtains a number when compared to

The opposite Aij = 1/Aij activity j, then j has the opposite value when
compared to i
Table 2:- Guidelines for Granting Values in Pairwise Comparison

Score 1 Indicates two choices have the same or no component) of the comparison component (column
difference of interest and score 9 shows a very large component). If a component has a weak level of influence,
dominance of a component being considered (line the range of scores ranges from 1 to 1/9 (one ninth), where 1

IJISRT18NV85 www.ijisrt.com 218

Volume 3, Issue 11, November – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
shows no difference and 1/9 shows the strong dominance of What is stated by:
the column element against the row element. When rating a AW = nW
score is done for a pair, an inverse value is automatically the
inverse ratio in the matrix. The order of the pairwise In the matrix theory the formula above is a
comparison matrix is described as follows (Saaty, 1996): characteristic equation with W which is an Eigen vector of
1. Compare all elements for each level in pairs. The matrix A with an eigen value of n. When written in full, the
comparison is transformed into a matrix. equation will look like in the following equation:
2. Comparison is done based on "judgment" from the
experts of the parties or who have an interest in decision
making.
3. Conducted directly (with discussion) or with a
questionnaire the total amount of judgment amounts to n
x. [N - I] / 2, n is the number of elements compared.

C A1 A2 A3 ... An
A1 a11 a12 a13 a1n
A2 a21 a22 a23 a2n
A3 a31 a32 a33 a3n
.... ...
An an1 an2 an3 ... ann In general there are several corresponding eigenvector
Table 3:- Pairwise Comparison Matrix values that meet the above equation. The variable n in the
above equation can be replaced with a vector A, as follows:
The matrix above is a pairwise comparison matrix.
The matrix is generated from a comparison between Aw = 𝜆𝑤
elements of certain criteria (in this case C). Aij value is the Where 𝜆 = (𝜆1, 𝜆2, … , 𝜆𝑛)every𝜆 the one that fulfills the
comparison value of the Ai element to the Aj element which above equation is called the eigen value, while the vector
states the relationship: that satisfies equation 2.30 is called the eigenvector. If the
a. How far is the level of interest of Ai when compared to matrix A is known and wants to be obtained W, then it can
Aj, or be solved through the following equation:
b. How much is Ai's contribution to criterion C compared [𝐴 − 𝑛] 𝑊 = 0
to Aj, or This equation can produce a solution that is not zero
c. How much characteristic C is found in Ai compared to (if and only if) n is an eigenvalue of A and W is the
Aj or
eigenvector. After the eigenvalue matrix A comparison is
d. How far is the dominance of Ai compared to Aj.
obtained. For example :
If the value of aij is known then theoretically the value 𝜆1. 𝐴2, 𝜆𝑛and based on matrix A which is unique,
of aij = l / au 'While the value of aij in situation i = j is that is𝑎𝑖𝑗 = 1. Dengan 1=1.2, ….,n,than :
𝑛
mutak 1. The numerical value imposed for the comparison
above is obtained from the scale of the comparison made by ∑ 𝜆𝑖 = 𝑛
Saaty. 𝑖=1

3. Step 3: Comparison of element weights Here all eigenvalues are zero except for one that is not
The weight sought is expressed in vectors W = [WI, zero, that is maximum eigenvalue, then if the assessment
W2, W3, ..., Wn]. The Wn value states the relative weight of carried out consistently will get the maximum eigenvalue of
An criterion for the entire set of criteria in the sub-system. A which is worth n. to get W, it can be done by substituting
In a perfect (theoretical) assessment situation a relationship the maximum eigenvalue price in the equation:
is obtained: AW = 𝜆𝑚𝑎𝑥 𝑊
𝑎𝑖𝑘 = 𝑎𝑖𝑗 . 𝑎𝑗𝑘 for all i, j, k Then the equation can be changed to:
A- 𝜆𝑚𝑎𝑥 1 𝑊 = 0
The matrix obtained is a consistent matrix. Thus the
To get a zero price, what you have to do is :
comparison value is obtained from the participant based on
the table, ie aij can be expressed in vector W as: A- 𝜆𝑚𝑎𝑥 𝐼 = 0
𝑎𝑖𝑗 = 𝑤𝑖⁄𝑤𝑗 . 𝑖𝑗 = 1,2,3, … , 𝑛
Based on the equation, the price of λmax can be obtained by
From the above equation can be made the following entering the equation λmax and adding the equation
equation: ∑𝑛𝑖=1 𝑊 1 2 = 1, then we will get the weight of
𝑎𝑖𝑗 . 𝑤𝑖⁄𝑤𝑗 = 𝑙, 𝑖 = 1,2,3, … . 𝑛 each element of the Wi operation, with i=1, 2, …,
𝑛
n) which is an eigenvector that corresponds to the
∑ 𝑎𝑖𝑗 . 𝑤𝑖⁄𝑤𝑗 = 𝑙, 𝑖 = 1,2,3, … . 𝑛 maximum eigenvalue.
𝑗=1
𝑛
4. Step 4: Calculation of the consistency ratio level of
∑ 𝑎𝑖𝑗 . 𝑤𝑗 = 𝑛𝑤𝑖 𝑖 = 1,2,3, … . 𝑛 inconsistency
𝑗=1

IJISRT18NV85 www.ijisrt.com 219

 Volume 3, Issue 11, November – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
the response is called the inconsistency ratio (CI) whose CI = consistency index
calculation is as follows:
𝜆 −𝑁 Based on Saaty calculations using 500 samples.
CI = 𝑚𝑎𝑥𝑛−1 Judgment matrix is taken randomly from scale, 1/9. 1/8, ...,
where :
1, 2, 9 will get the consistency average for different size
𝜆𝑚𝑎𝑘𝑠 = eigenvalue maximum
matrices, as follows :
𝑛 = 𝑚𝑎𝑡𝑟𝑖𝑥 𝑠𝑖𝑧𝑒

Matrix Size 1,2 3 4 5 6 7 8 9

Random Index 0 0.58 0.9 1.12 1.24 1.32 1.41 1.45
Table 3:- Random Index Value

A comparison between CI and RI for a matrix is identified

as a consistency ratio (CR)
CR = RI / CI
Vector calculation results are accepted if CR is around 0.1
or less (0.2 can be tolerated, but not more). If the CR is not
less than 0.1 the problem is studied again and a
reassessment is carried out.

e. Step 5: Choose the best alternative

III. RESEARCH METHODS

In conducting this research, several stages of

research methodology are needed from the initial to the final
stages, along with the stages of the research methodology:

Fig 3:- Research Method

IJISRT18NV85 www.ijisrt.com 220

Volume 3, Issue 11, November – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

IV. RESEARCH RESULTS AND DISCUSSION

Determination of Criteria, Sub Criteria, Alternative

Criteria Sub Criteria

Discounted rates (C2)
Cost (C) Bid price (C3)
Surcharge according to damage (C4)
Payment period (C5)
Delivery period (D1)
Delivery (D) Delivery of goods on time (D2)
The accuracy of shipping parts (D5)
Suitability of goods with specified specifications (Q1)
Ability to provide consistent quality (Q2)
Quality (Q) The completeness of document (Q3)
Provision of goods without defects (Q4)
Component strength (lifetime) (Q5)
Ease of changing defective products (R1)
Responsive (R)
Speed in responding to customer desires (R2)
Convenience to contact (S1)
The ability to provide information clearly (S2)
Service (S)
Speed in response to customer requests (S3)
Respond quickly in resolving customer complaints (S4)
Provide a guarantee or guarantee to the customer (W1)
Warranty (W)
Can provide assistance in an emergency (W2)
Ability to maintain contract agreements (P1)
Ability to fulfill the number of orders (P2)
Performance Ability to fulfill scheduled schedules (P3)
History (P)
Total supply in one year (P4)
Experience related to claims or claims (P5)
The frequency of company failures in fulfilling orders on time (P6)
Table 4:- Criteria and Sub Criteria

Criteria Supplier Name

PT. TR
Alternative
PT. HT
PT. DU
Table 5:- Supplier Data of PT. UT

IJISRT18NV85 www.ijisrt.com 221

Volume 3, Issue 11, November – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Fig 4:- Network ANP supplier cylinder selection

Analysis of Weight of Criteria, Sub Criteria, and Alternatives

Criteria Cost Delivery Quality Responsivenes Service Warranty Performance

(C) (D) (Q) s (R) (S) (W) History (P)
Cost (C) 1 4 1/3 2/7 1 1/2 5/6 2 1/3
Delivery (D) 2/9 1 1/2 4/5 3/4 3/5 2 2/5
Quality (Q) 3 2/3 2 1 4 4/5 4/5 1 1/3 2 1/2
Responsiveness (R) 1 1 1/4 1/5 1 5/6 1 3 2/3
Service (S) 2 1 1/3 1 1/4 1 1/5 1 1 1/7 2 1/3
Warranty (W) 1 1/5 1 5/7 3/4 1 7/8 1 4
Performance History (P) 3/7 2/5 2/5 2/7 3/7 1/4 1
Table 6:- Calculation of criteria weight values

Number of Rows Priority Results Priority to

Cost (C) 2.31896E+28 0.141660 4
Delivery (D) 1.51598E+28 0.092608 6
Quality (Q) 4.29933E+28 0.262636 1
Responsiveness
1.97931E+28 0.120912 5
(R)
Service (S) 2.91374E+28 0.177994 2
Warranty (W) 2.48036E+28 0.151520 3
Performance
8.62208E+27 0.052670 7
History (P)
1.63699E+29 1
Table 7

Discount Bid Surcharge Payme Number of Priority Prio

ed rates price according to nt Rows Results rity
(C2) (C3) damage (C4) period to
(C5)
Discounted rates 1 2 6 6 5/7 9.7254E+19 0.5470 1
(C2)
Bid price (C3) 1/2 1 4 1/3 2 3/8 4.928E+19 0.2772 2
Surcharge according 1/6 1/4 1 2/3 1.3154E+19 0.0740 4
to damage (C4)
Payment period (C5) 1/7 3/7 1 1/2 1 1.8115E+19 0.1019 3
1.778E+20 1.0000
Table 8:- Weighting sub Cost criteria

IJISRT18NV85 www.ijisrt.com 222

Volume 3, Issue 11, November – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Number of Priority
A B C Priority to
Rows Results
PT TR 1 1 1 5/8 53830158 0.37843 2
PT HT 1 1 2 57359905 0.40324 1
PT DU 3/5 1/2 1 31057635 0.21833 3
142247698 1
Table 9:- Alternative weighting of cost criteria (C)

PT Number Priority Priority

PT HT PT DU
TR of Rows Results to

PT TR 1 1 1/5 1 2/3 9193.095 0.404294 1

PT HT 5/6 1 2 1/7 8804.12 0.387187 2
PT DU 3/5 1/2 1 4741.439 0.208519 3
22738.65 1
Table 10:- Weighting alternative criteria for discounting (C2)

Test ForConsistency Ratio (CR)

Suitability of goods with Amount in Eigen

Sub Criteria Discounted rates
specified specifications column vector

Discounted rates 0.86690

1 6 1/2 1.1535
(C2)) 7
Suitability of
goods with
0.13309
specified 1/7 1 7.5136
3
specifications
(Q1)
Table 11:- Average pairwise comparison of Cost criteria (C) with sub quality criteria (Q)

Based on the comparison table above, eigenvalue CR = CI/RI (table

vector can be calculated, maximum lamda (λmaks), Wharton)
consistency index (CI) and index ratio (CR). Here's the 0
= 0,58 = 0
calculation:
Value of λmax =
Data is said to be consistent if CR is around 0.1 or less
(1,1535*0,8669
(0.2 can be tolerated, but not more). If the CR is not less
0) +
than 0.1 the problem is studied again and a reassessment is
(7,5136*0,1330
carried out.
)
=2 ANP Method Weighting Analysis
2−2
CI = 2−1 = 0 The weighting results for each criterion, sub-criteria and
alternatives are as follows:

IJISRT18NV85 www.ijisrt.com 223

Volume 3, Issue 11, November – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
Alternative Weight Alternative Level 2 (Sub Weight (Sub Alternative (Sub Weight Alternative
Level 1 Criteria Weight (Criteria)
(Criteria) Criteria Criteria) Criteria) Criteria) Sub Criteria
PT TR 0,404293705
C2 0,54700 PT HT 0,387187391
PT TR 0,378425514
PT DU 0,208518904
PT TR 0,395672033
C3 0,27720 PT HT 0,360816901
PT DU 0,243511067
Cost ( C) 0,14166 PT HT 0,403239602
PT TR 0,23621329
C4 0,07400 PT HT 0,379063144
PT DU 0,384723566
PT TR 0,416196002
PT DU 0,218334885
C5 0,10190 PT HT 0,261199677
PT DU 0,322604321
PT TR 0,417814724
PT TR 0,245118669
W1 0,63397 PT HT 0,315230525
PT DU 0,266954751
Warranty (W) 0,15152 PT HT 0,66384831
PT TR 0,662971462
W2 0,36603 PT HT 0,191175078
PT DU 0,091033021
PT DU 0,14585346
PT TR 0,21815609
Q1 0,20572 PT HT 0,369158658
PT TR 0,416327319 PT DU 0,412685252
PT TR 0,452465601
Q2 0,33221 PT HT 0,371463809
PT DU 0,17607059
PT TR 0,572067808
Quality (Q) 0,262636 PT HT 0,424359398 Q3 0,07083 PT HT 0,273001066
PT DU 0,154931125
PT TR 0,580914921
Q4 0,14367 PT HT 0,226670253
PT DU 0,192414826
PT DU 0,099313282 PT TR 0,413953987
Q5 0,24757 PT HT 0,351525862
PT DU 0,234520151
PT TR 0,667836954
S1 0,06793 PT HT 0,199345816
PT TR 0,463936244
PT DU 0,132817229
PT TR 0,284461562
S2 0,20226 PT HT 0,371841093
PT DU 0,343697345
Service (S) 0,177994 PT HT 0,412778135
PT TR 0,52681279
S3 0,20226 PT HT 0,313539714
PT DU 0,159647496
PT TR 0,469943106
PT DU 0,123285621
S4 0,49463 PT HT 0,205749003
PT DU 0,324307891
PT TR 0,550491576
PT TR 0,548442117 D1 0,11514 PT HT 0,255946093
PT DU 0,19356233
PT TR 0,367593333
Delivery (D) 0,092608 PT HT 0,359164104 D2 0,61193 PT HT 0,414290588
PT DU 0,21811608
PT TR 0,547786829
PT DU 0,092393779 D5 0,27294 PT HT 0,28090735
PT DU 0,171305821
PT TR 0,489175377
PT TR 0,340639146
R1 0,41590 PT HT 0,338592967
PT DU 0,172231656
Responsiveness ( R) 0,120912 PT HT 0,50373763
PT TR 0,550256693
R2 0,58410 PT HT 0,274914625
PT DU 0,155623224
PT DU 0,174828682
PT TR 0,24300112
P1 0,39822 PT HT 0,251094882
PT DU 0,325903998
PT TR 0,625391814
PT TR 0,519189385
P2 0,18297 PT HT 0,289217323
PT DU 0,191593292
PT TR 0,413841253
P3 0,17815 PT HT 0,276855074
Performance History PT DU 0,309303673
0,05267 PT HT 0,240415754
(P) PT TR 0,459555048
P4 0,09115 PT HT 0,289586232
PT DU 0,250858329
PT TR 0,365075712
P5 0,08494 PT HT 0,332355828
PT DU 0,302568461
PT DU 0,134192432
PT TR 0,567322755
P6 0,06456 PT HT 0,220614882
PT DU 0,212062364
Tabel 12:- Analisis Pembobotan Metode ANP

IJISRT18NV85 www.ijisrt.com 224

Volume 3, Issue 11, November – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
V. CONCLUSION 2. Service S (0.177994)
3. Warranty W (0.15152)
Based on the final project research that has been done, 4. CostC (0.14166)
conclusions can be taken as follows: 5. Responsive R (0,12091)
1. There are several criteria used as a basis for 6. Delivery D (0.0926) and
determining the alternative cylinder component supplier 7. Performance history P (0.0526)
selection carried out at PT. UT. These criteria include :
A. Cost (C) B. The priority order of the sub quality criteria is:
This criterion has several sub criteria, namely: 1. Ability to provide consistent quality
a. Discounted price (C2) 2. Strength of components
b. Bid price (C3) 3. Suitability of goods with specified
c. Compatibility (surcharge) (C4) specifications
d. Payment grace period (C5) 4. Provision of goods without defects and
5. Completeness of documents
B. Delivery (D)
This criterion has several sub criteria, namely: C. The priority order of the service sub criteria is:
a. Delivery period (D1) 1. Respond quickly in resolving customer
b. Delivery of goods on time (D2) complaints
c. Part delivery accuracy (D5) 2. Speed in response to customer requests
3. The ability to provide information clearly and
C. Quality (Q) 4. Convenience to contact
This criterion has several sub criteria, namely:
a. Suitability of goods with specified D. The priority order of the guarantee sub criteria is:
specifications (Q1) 1. Provide guarantees or guarantees to customers
b. Ability to provide consistent quality (Q2) and
c. Completeness of documents (Q3) 2. Can provide assistance in an emergency (2).
d. Provision of defective goods (Q4)
E. The priority order of the Cost sub criteria is:
D. Responsiveness (R) 1. Giving discounts
This criterion has several sub criteria, namely: 2. Bid price
a. Ease of changing defective products (R1) 3. Payment period and
b. Speed in response to customer desires (R2) 4. Compliance with surcharge

E. Service (S) F. The priority sequence of responsive sub criteria is:

This criterion has several sub criteria, namely: 1. Speed in responding to customer desires and
a. Convenience to contact (S1) 2. Ease of changing defective products
b. The ability to provide information clearly (S2)
c. Speed in responding to customer requests (S3) G. The priority order for the delivery sub criteria is:
d. Respond quickly in resolving customer 1. Delivery of goods on time
complaints (S4) 2. The accuracy of shipping parts and
3. Delivery period
F. Warranty (W)
This criterion has several sub criteria, namely: H. The priority sequence of the performance history
a. Provide a guarantee or guarantee to the sub criteria is :
customer (W1) 1. Ability to maintain contract agreements
b. Can provide assistance in an emergency (W2) 2. Ability to fulfill the number of orders
3. Ability to fulfill the scheduled schedule
G. Performance History (P) 4. Total supply in one year (Production)
This criterion has several sub criteria, namely: 5. Experience related to claims or claims and
a. Ability to maintain contract agreements (P1) 6. The frequency of company failures in
b. Ability to fulfill the number of orders (P2) fulfilling orders on time.
c. Ability to fulfill the scheduled schedule (P3)
d. Total supply in one year (Production) (P4) 3. The chosen alternative between PT. DU, PT. HT and
e. Experience related to claims or claims (P5) PT. TR is PT. TR is the priority value of manual
f. Frequency of company failure to fulfill orders calculation 0.4561 and super decision software 0.43612.
on time (P6) While PT DU's manual calculation value is 0.21635 and
super decision software 0.21396, PT HT calculation is
2. The order of priority criteria and sub criteria in the 0.32754 and super decision software is 0.34993.
selection of suppliers at PT. UT, namely:
A. The order of priority criteria is:
1. Quality Q (0.2626)

IJISRT18NV85 www.ijisrt.com 225

Volume 3, Issue 11, November – 2018 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
REFERENCES

[1]. Amiri, M., S.S. Jamshid, P. Nafiseh and S.

Mahdi,(2011) Developing a EMATEL method
toprioritize distribution centers in supply
chain.Manage. Sci. Lett., vol.10, no.3-4, 279-288.
[2]. Blocher, Edward J., K. H. Chen, T. W. Lin, 2002,
Management Compensation on Cost Management 2e:
A Strategic Emphasis, Mc. Graw-Hill Irwin.
[3]. Marimin. (2004), Teknik dan Aplikasi Pengambilan
Keputusan Kriteria Majemuk, Cetakan Kedua, PT
GRASINDO, Jakarta.
[4]. Ngatawi. Setyaningsih, Ira. 2011. “Analisis Pemilihan
Supplier Menggunakan Metode Analytic Hierarchy
Process (AHP)”. Semarang.UIN Sunan Kalijaga.
[5]. Pratama, Septian R.A. dkk. 2015. “Penentuan Supplier
Resin Dengan Menggunakan Analytical Network
Process”Probolinggo: Universitas Panca Marga
Probolonggo.
[6]. Rusydiana, Aam Salamet., Devi, Abrista. 2013.
“Analytic Network Process: Pengantar Teori dan
Aplikasi”. Bogor: smart publishing.
[7]. Saaty, T. L. (1996). Decision Making With
Dependence and Feedback: The Analytic Network
Process. Pittsburgh: RWS Publications.
[8]. Saaty, T. L. (1994). Fundamentals of Decision Making
And Priority Theory With The Analytic Hierarchy
Process Vol. VI. Pittsburgh: RWS Publications.
[9]. Saaty T.L. 2008. The Analytic Network Process.
Pittsburgh: Ellsworth Avenue.
[10]. Saaty, T. L. and Vargas, L. G. (2006). Decision
Making With The Analytic Network Process:
Economic, Political, Social and Technological
Application with Benefits, Opportunities, Costs and
Risks. New York: Springer.
[11]. SuperDecisions. 2017.
anuals.(online):https://www.superdecisions.com/manu
als/ (22 November 2017)
[12]. Wibisono, Yogi Y. Gondo, Kristi D.A. 2013.
“Pemilihan Pemasok Dengan Metode Analytic
Network Process (ANP): Studi Kasus Di PT.AI”.
Banding: Universitas Katolik.

IJISRT18NV85 www.ijisrt.com 226