25.01.2014 Views

Intelligent Selection of Distributor by Manufacturer in Supply Chain ...

Intelligent Selection of Distributor by Manufacturer in Supply Chain ...

Intelligent Selection of Distributor by Manufacturer in Supply Chain ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Gurmeet Kaur et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 1002-1007<br />

ISSN:2229-6093<br />

<strong>Intelligent</strong> <strong>Selection</strong> <strong>of</strong> <strong>Distributor</strong> <strong>by</strong> <strong>Manufacturer</strong> <strong>in</strong> <strong>Supply</strong> Cha<strong>in</strong><br />

Management us<strong>in</strong>g Multi Agents<br />

Gurmeet Kaur<br />

Department <strong>of</strong> Computer Science<br />

and Eng<strong>in</strong>eer<strong>in</strong>g<br />

Lovely Pr<strong>of</strong>essional University<br />

Phagwara (Punjab)<br />

gurmeetkaurvirdi@gmail.com<br />

Dr. Babita Pandey<br />

Department <strong>of</strong> Computer Science<br />

and Eng<strong>in</strong>eer<strong>in</strong>g<br />

Lovely Pr<strong>of</strong>essional University<br />

Phagwara (Punjab)<br />

Abstract<br />

<strong>Supply</strong> cha<strong>in</strong> management (SCM) consists <strong>of</strong><br />

various components such as: supplier,<br />

manufacturer, factories, warehouses, distributions<br />

agents etc. These components are <strong>in</strong>volved <strong>in</strong><br />

various activities such as: supply<strong>in</strong>g raw<br />

materials, manufactur<strong>in</strong>g and distribution <strong>of</strong> the<br />

f<strong>in</strong>ished product to the f<strong>in</strong>al customer. The<br />

selection <strong>of</strong> right distributor is very important<br />

activity <strong>in</strong> SCM. This paper focuses on the Multi<br />

Agent technique for the selection <strong>of</strong> suitable<br />

distributor for the manufacturer. Different<br />

manufacturers have different preferences on the<br />

attributes like payment delay, market coverage,<br />

technical ability etc. The supply cha<strong>in</strong> consists <strong>of</strong><br />

an <strong>in</strong>telligent Broker agent which selects the best<br />

distributor accord<strong>in</strong>g to the requirement <strong>of</strong> the<br />

manufacturer. The multi agent system <strong>of</strong>fers a<br />

flexible solution for the selection as well as<br />

negotiation. The agents are implemented through<br />

the Jade environment.<br />

Keywords<br />

<strong>Supply</strong> Cha<strong>in</strong> Management (SCM), Agent, Java<br />

Application Development Environment (Jade),<br />

selection, negotiation<br />

1. Introduction<br />

The distributor’s selection has a large impact <strong>in</strong><br />

the supply cha<strong>in</strong> management (SCM). SCM<br />

consists <strong>of</strong> various components such as: supplier,<br />

manufacturer, factories, warehouses, distributions<br />

agents etc. These components are <strong>in</strong>volved <strong>in</strong><br />

various activities such as: supply<strong>in</strong>g raw materials,<br />

manufactur<strong>in</strong>g and distribution <strong>of</strong> the f<strong>in</strong>ished<br />

product to the f<strong>in</strong>al customer. The selection <strong>of</strong><br />

right distributor is very important activity <strong>in</strong> SCM.<br />

Very few researchers have worked on the selection<br />

<strong>of</strong> distributor <strong>in</strong> SCM such as: Zou et al. [1]<br />

deployed rough set theory to deal with the<br />

qualitative data and modified it for preferred<br />

distributor selection. They derived certa<strong>in</strong> decision<br />

rules which facilitated distributor selection. They<br />

generated several rules for distributors’ evaluation<br />

and selection. They showed the importance <strong>of</strong> the<br />

relationship <strong>in</strong>tensity, market<strong>in</strong>g experience and<br />

the management ability <strong>in</strong> select<strong>in</strong>g the<br />

distributors. Reddy et al. [2] deployed fuzzy goal<br />

programm<strong>in</strong>g approach for quota allocation to<br />

distributors <strong>of</strong> the supply cha<strong>in</strong>. Many <strong>of</strong> the<br />

researchers have also worked on the selection <strong>of</strong><br />

supplier <strong>in</strong> SCM. Mian [3] divided the multiattributes<br />

<strong>of</strong> supplier <strong>in</strong>to two categories <strong>by</strong> fuzzy<br />

IJCTA | MAY-JUNE 2012<br />

Available onl<strong>in</strong>e@www.ijcta.com<br />

1002


Gurmeet Kaur et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 1002-1007<br />

ISSN:2229-6093<br />

and certa<strong>in</strong>ty properties, and used multi-Agent<br />

technology to make supplier selection decision<br />

<strong>in</strong>telligent. Kumar et al. [4] <strong>in</strong>troduced a method<br />

<strong>of</strong> supplier selection based on multi agent and<br />

fuzzy decision mak<strong>in</strong>g techniques <strong>in</strong>clude with<br />

risk factor for the selection <strong>of</strong> supplier. L<strong>in</strong> and He<br />

[5] def<strong>in</strong>ed the concepts <strong>of</strong> construction supply<br />

cha<strong>in</strong> and construction supply cha<strong>in</strong>. They<br />

proposed to resolve the problem <strong>in</strong> SCM <strong>by</strong><br />

construct<strong>in</strong>g an agent-based framework for<br />

construction supply cha<strong>in</strong> coord<strong>in</strong>ation based on<br />

the agent technology and multi-attribute<br />

negotiation and utility theory.We propose agent<br />

based approach for the selection <strong>of</strong> distributor. The<br />

distributor is selected on the basis <strong>of</strong> multi<br />

attributes. Different manufacturer have their<br />

different needs. We are us<strong>in</strong>g these attributes to<br />

match the manufacturers’ and distributors’<br />

pr<strong>of</strong>iles. The broker agent is responsible for<br />

match<strong>in</strong>g the pr<strong>of</strong>iles <strong>of</strong> manufacturer and<br />

distributor. We aim to implement a method <strong>of</strong><br />

selection <strong>of</strong> distributor us<strong>in</strong>g the multi agent<br />

approach. The agents are developed us<strong>in</strong>g JADE<br />

(Java Agent Development Environment).<br />

The organization <strong>of</strong> the paper is as follows:<br />

Section 2 gives the overview <strong>of</strong> the methodology<br />

used <strong>in</strong> the development <strong>of</strong> proposed multi agent<br />

system. In section 3, we have expla<strong>in</strong>ed the<br />

different agents <strong>in</strong> the multi agent system. Section<br />

4 represents the implementation details and section<br />

5 <strong>in</strong>cludes the conclusions.<br />

2. Methodology<br />

The selection <strong>of</strong> distributor consists <strong>of</strong> three<br />

steps: registration, Evaluation and Filter<strong>in</strong>g.<br />

In the first stage registration <strong>of</strong> the distributor,<br />

manufacturer and Broker agent to the DF agent is<br />

done <strong>in</strong> the jade environment. After that the<br />

manufacturer and the distributor send their<br />

advertisements to the broker agent. The<br />

advertisement consists <strong>of</strong> a list <strong>of</strong> specific<br />

attributes and each attribute has a<br />

weight attached to it.<br />

The attributes are [1] Payment delay (PD),<br />

Market coverage (MC), Order Quality (OQ), Cost<br />

Control (CC), Technical Ability (TA), Shipments<br />

(SH), Market<strong>in</strong>g Experience (ME), Relationship<br />

Intensity (RI), F<strong>in</strong>ancial Strength (FS), Logistic<br />

1.Request<br />

DF Agent<br />

1.Request<br />

2.Inform<br />

4. Call for proposals from<br />

the <strong>Distributor</strong> agents for<br />

the product.<br />

2.Inform<br />

1.Request<br />

2.Inform<br />

<strong>Distributor</strong><br />

Agent<br />

5.Send attributes value<br />

7.accept/reject proposal<br />

Broker Agent<br />

(6. do the match<strong>in</strong>g <strong>of</strong><br />

attributes <strong>of</strong> the<br />

<strong>Distributor</strong> agent and<br />

manufacturer agent)<br />

3.Send product id<br />

and attributes<br />

<strong>Manufacturer</strong> Agents<br />

8. <strong>in</strong>form agent<br />

Figure 1: <strong>Selection</strong> <strong>of</strong> distributer<br />

IJCTA | MAY-JUNE 2012<br />

Available onl<strong>in</strong>e@www.ijcta.com<br />

1003


Gurmeet Kaur et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 1002-1007<br />

ISSN:2229-6093<br />

Capability (LC), Infrastructure and Equipment<br />

(IE) and Market<strong>in</strong>g capability (MR) with values<br />

from 1 (low) to 3 (high).<br />

The broker agent f<strong>in</strong>d the difference between the<br />

manufacturers’ attributes and various distributers’<br />

attributes accord<strong>in</strong>g to equation1.<br />

D j =<br />

…….(1)<br />

Where M i is the ith Attribute <strong>of</strong> the<br />

manufacturer and D ij is the ith attribute <strong>of</strong> jth<br />

Distributer. Then it selects the best distributor<br />

(BD) which has m<strong>in</strong>imum sum <strong>of</strong> difference (D j )<br />

with the attributes wanted <strong>by</strong> the manufacturer<br />

(us<strong>in</strong>g equation2).<br />

BD= m<strong>in</strong> (D 1 ,D 2 ,......,D n ).........(2)<br />

All the work <strong>of</strong> match<strong>in</strong>g <strong>of</strong> pr<strong>of</strong>iles is done <strong>by</strong><br />

the Broker. The whole process is summarised <strong>in</strong><br />

the figure 1.<br />

<strong>in</strong>terface. The follow<strong>in</strong>g Graphical user <strong>in</strong>terfaces<br />

are made to enter the attributes <strong>by</strong> the<br />

manufacturer and <strong>by</strong> the <strong>Distributor</strong>. The<br />

distributor is chosen on the basis <strong>of</strong> these<br />

attributes.<br />

3. Multi agent System<br />

We have made three types <strong>of</strong> agents:<br />

3.1. <strong>Manufacturer</strong> Agent: The manufacturer<br />

agent does the work on behalf <strong>of</strong> the<br />

manufacturer. It tries to supply the various<br />

products to the <strong>Distributor</strong>s.<br />

3.2. <strong>Distributor</strong> Agent: The distributor agent is<br />

responsible for distribut<strong>in</strong>g the product. It gets<br />

the desired product from the<br />

manufacturer.<br />

3.3. Broker Agent: This agent is responsible for<br />

the overall selection <strong>of</strong> the best distributor for<br />

the manufacturer <strong>by</strong> match<strong>in</strong>g their pr<strong>of</strong>iles.<br />

4. Implementation<br />

4.1. Development <strong>of</strong> the <strong>Manufacturer</strong> and<br />

<strong>Distributor</strong> Gui:-<br />

The manufacturer and the distributor agents can<br />

enter the value to the attributes <strong>by</strong> a user<br />

Figure 2: <strong>Distributor</strong> GUI<br />

Figure 3: <strong>Manufacturer</strong> GUI<br />

4.2. <strong>Selection</strong> <strong>of</strong> Best Distributer<br />

The Broker Agent get the attributes from the<br />

<strong>Manufacturer</strong> agents and the distributor agents.<br />

IJCTA | MAY-JUNE 2012<br />

Available onl<strong>in</strong>e@www.ijcta.com<br />

1004


Gurmeet Kaur et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 1002-1007<br />

ISSN:2229-6093<br />

Then after tak<strong>in</strong>g all the <strong>in</strong>puts the broker<br />

evaluates the Differences between the pr<strong>of</strong>iles <strong>of</strong><br />

a particular manufacturer Agent and the Various<br />

<strong>Distributor</strong> Agents. The broker then compares the<br />

attributer provided <strong>by</strong> the manufacturer and the<br />

<strong>Distributor</strong> and selects the most appropriate<br />

<strong>Distributor</strong>. From the table1 as the sum <strong>of</strong><br />

differences <strong>of</strong> <strong>Manufacturer</strong> and distributer1 is 5<br />

which is less than the sum <strong>of</strong> difference <strong>of</strong><br />

manufacturer and Distributer2, so the<br />

<strong>Distributor</strong>1 is selected <strong>by</strong> the broker agent.<br />

The figure 4 shows the complete<br />

communication is done between the<br />

manufacturer, <strong>Distributor</strong> and the broker<br />

Agents dur<strong>in</strong>g the selection <strong>of</strong> the distributer.<br />

Table 1: Comparison <strong>of</strong> <strong>Distributor</strong> and <strong>Manufacturer</strong> Agents<br />

Attributes PD MC OQ CC TA SH ME RI FS LC IE MR SD<br />

<strong>Manufacturer</strong> L M H M H H M H M M L M<br />

Distributer1 M M M M H M M L M M L M<br />

Distributer2 L M L M M H H M L M M M<br />

D1(<strong>in</strong>teger value) 1 0 1 0 0 1 0 2 0 0 0 0 5<br />

D2(<strong>in</strong>teger value) 0 0 2 0 1 0 1 1 1 0 1 0 7<br />

L=Low, M=Medium, H=High and SD=sum <strong>of</strong><br />

difference<br />

In the table, the value high is taken as 3,<br />

medium=2 and low =1.<br />

4.3. Negotiation among the broker and the<br />

distributor:<br />

After selection <strong>of</strong> the most suitable distributor,<br />

now the broker does the negotiation on behalf <strong>of</strong><br />

the manufacturer. As the broker has the price that<br />

the manufacturer expects from the distributor and<br />

also it provides the limit (here we have taken the<br />

limit 20% <strong>of</strong> the price given <strong>by</strong> manufacturer) up<br />

to which it can <strong>in</strong>crease its range <strong>in</strong> accordance to<br />

the best distributor.<br />

Broker compares the prices <strong>of</strong> distributor and the<br />

manufacturer and aga<strong>in</strong> proposes the distributor<br />

to lower its price (negotiation) if it is greater than<br />

the price + limit provided <strong>by</strong> the manufacturer.<br />

When the negotiation is complete, the broker<br />

send accept proposal message to the distributor<br />

for the f<strong>in</strong>al acknowledgement the distributor<br />

Figure 4: Sniffer Agent (<strong>Selection</strong> process <strong>of</strong> distributor)<br />

IJCTA | MAY-JUNE 2012<br />

Available onl<strong>in</strong>e@www.ijcta.com<br />

1005


Gurmeet Kaur et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 1002-1007<br />

ISSN:2229-6093<br />

send a <strong>in</strong>form message.<br />

The flow diagram(figure 5 ) shows the algorithm<br />

used for negotiation process. When the best<br />

distributor comes to know that it has been<br />

selected <strong>by</strong> the broker for the distribution <strong>of</strong> a<br />

particular product (<strong>by</strong> gett<strong>in</strong>g a propose message<br />

from the broker) then it reply with price that he<br />

wants for keep<strong>in</strong>g the product.<br />

The figure 6 shows the messages exchanged<br />

between <strong>Distributor</strong>, manufacturer and the broker<br />

Start<br />

Get mprice<br />

Get dprice<br />

Calculate limit=mprice*(20/100)<br />

Calculate difference=dpricemprice<br />

False<br />

If<br />

D


Gurmeet Kaur et al ,Int.J.Computer Technology & Applications,Vol 3 (3), 1002-1007<br />

ISSN:2229-6093<br />

References<br />

[1]Zhonghai Zou, Tzu-Liang Tseng, Hansuk Sohn,<br />

Gu<strong>of</strong>ang Song and Rafael Gutierrez (2011) “A<br />

Rough Set Based Approach to <strong>Distributor</strong><br />

<strong>Selection</strong> <strong>in</strong> <strong>Supply</strong> Cha<strong>in</strong> Management” Elsevier<br />

Ltd. ©2011<br />

[2]B.Chandra Mohana Reddy, K. Hemachandra<br />

Reddy , C. Nadha Muni Reddy and K.Vijaya<br />

Kumar Reddy “Quota Allocation to <strong>Distributor</strong>s <strong>of</strong><br />

the <strong>Supply</strong> Cha<strong>in</strong> under <strong>Distributor</strong>s’ Uncerta<strong>in</strong>ty<br />

and Demand Uncerta<strong>in</strong>ty <strong>by</strong> Us<strong>in</strong>g Fuzzy Goal<br />

Programm<strong>in</strong>g” Jordan Journal <strong>of</strong> Mechanical and<br />

Industrial Eng<strong>in</strong>eer<strong>in</strong>g Volume 2, Number 4,<br />

December. 2008.<br />

[3]Wang Mia (2011) “Comprehensive evaluation<br />

and selection model <strong>of</strong> supplier under E-bus<strong>in</strong>ess”<br />

IEEE 978-1-4244-8673-0 @ 2011<br />

[4]Vivek Kumar , Dr. S. Sr<strong>in</strong>ivasan and Dr.<br />

Subhra Das (2011) “A Multi-Agent System for<br />

Management <strong>of</strong> Supplier <strong>Selection</strong> Process <strong>in</strong> a<br />

Fuzzy <strong>Supply</strong> Cha<strong>in</strong>” International Journal <strong>of</strong><br />

Computer Applications (0975 – 8887) Volume 23–<br />

No.6, June 2011<br />

Computer Applications (0975 – 8887) Volume 23–<br />

No.6, June 2011<br />

[7]Yang L<strong>in</strong> and YaBo He (2011) ”<strong>Supply</strong> Cha<strong>in</strong><br />

Coord<strong>in</strong>ation Of Construction Management Based<br />

On Multi-Agent ” IEEE©2011<br />

[8]Vivek Kumar and Dr. S. Sr<strong>in</strong>ivasan (2010) “A<br />

Review <strong>of</strong> <strong>Supply</strong> Cha<strong>in</strong> Management us<strong>in</strong>g<br />

Multi-Agent System” IJCSI International Journal<br />

<strong>of</strong> Computer Science Issues, Vol. 7, Issue 5,<br />

September 2010.<br />

GURMEET KAUR, She is<br />

persu<strong>in</strong>g M.Tech.(CSE) from<br />

lovely pr<strong>of</strong>essional university,<br />

phagwara (Punjab) after do<strong>in</strong>g<br />

M.Sc. (comp.sc.) from guru<br />

nanak<br />

dev<br />

university,amritsar(Punjab).Her research area<br />

<strong>in</strong>cludes multi agent system applications <strong>in</strong> supply<br />

cha<strong>in</strong>. email-id- gurmeetkaurvirdi@gmail.com<br />

Dr.BABITA PANDEY, She has done PHD. Now<br />

work<strong>in</strong>g as assistant pr<strong>of</strong>essor <strong>in</strong> lovely<br />

pr<strong>of</strong>essional university, Phagwara (Punjab). Her<br />

research area <strong>in</strong>cludes supply cha<strong>in</strong> management.<br />

email-id- shukla_babita@yahoo.co.<strong>in</strong><br />

[5]Wang Mia (2011) “Comprehensive evaluation<br />

and selection model <strong>of</strong> supplier under E-bus<strong>in</strong>ess”<br />

IEEE 978-1-4244-8673-0 @ 2011<br />

[6]Vivek Kumar , Dr. S. Sr<strong>in</strong>ivasan and Dr.<br />

Subhra Das (2011) “A Multi-Agent System for<br />

Management <strong>of</strong> Supplier <strong>Selection</strong> Process <strong>in</strong> a<br />

Fuzzy <strong>Supply</strong> Cha<strong>in</strong>” International Journal <strong>of</strong><br />

IJCTA | MAY-JUNE 2012<br />

Available onl<strong>in</strong>e@www.ijcta.com<br />

1007

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!