IJACT899PPL

A Supply Chain Simulation Model with Customer’s Satisfaction

Jianfeng Li, Yan Lin, Feng Jin Management Science and Engineering , Dalian Maritime University

Dalian city, PR. China E-mail [email protected]

Abstract

According to the customer’s satisfaction, this paper establishes a two level supply chain simulation model through system dynamics (SD) approach, and then, Simulink tool is adopted for that model. Through the simulation, it’s shown that improvement of supply chain process’s promptness can bring out the more customers’ satisfaction, and then, the sale amount is increased. The indirect profit is gained by means of improving operation’s efficiency in supply chain. In this way, the simulation about customer satisfaction is reflected clearly, and that system dynamics simulation model can be helpful for the supply chain management project.

Keywords: Supply Chain Management, Supply Chain Simulation, System Dynamics

1. Introduction

With the development of economic globalization, supply chain management (SCM) becomes an important management conception for enterprises in the current vehement environment [1,2]. As what Wood has stated, since the supply chain represents 60 to 80% of a typical company’s cost structure, a 10% reduction can yield a 40 to 50% improvement in pre-tax profits [3]. For SCM, on account of avoiding large expense for the failure of SCM projects, simulation is a widely and useful method, many scholars have worked over the supply chain in order to achieve some value things:

Kang Bokyoung applied an new social network analysis method to simulate supply chain integrating agent-based modeling[4]. Minegishi and Thiel made a system dynamics simulation for a food supply chain system, that work sheds lighted on the complex nature of this specific type of supply chain and in particular on the coordination of variables controlling the food production[5]. Adhitya, Arief integrated dynamic simulation and LCA indicators to bring forward some decision support for green supply chain operation[6]. Gavirneni, from the viewpoint of information distortion, simulated an overall supply chain model, that emphasizes the value of information and extended existing inventory theory[7]. Towill from system dynamics perspective also demonstrated that supply chain integration with exchange of information was as beneficial as lead time reduction throughout the supply chain via JIT [8]. Dejonckheere examined the beneficial impact of information sharing in multi-tier supply chains and discovered that information sharing helped to reduce the bullwhip effect in the chains with different inventory policies [9]. Kumar used six sigma simulation and designed experiment to quantify supply chain trade-offs and developed a flexible distribution networks[10].

In current simulation literature for SCM, the research on the reflection of customer’s satisfaction in the supply chain is very little, in this way, this paper establishes a two level supply chain simulation model with customer’s satisfaction through system dynamics, and showing some little improvement of operation’s efficiency in supply chain can bring out the change of main customer’s satisfaction, which also lead to alternation of other factors, such as sale amount.

The rest of the paper is organized as follows. In section 2, two level supply chain simulation model using SD method is established. In section 3, Simulink tool is adopted for the above simulation model, and then, the results expected from the simulation are described. In section4, the Simulation analysis on customer’s satisfaction is given out. In section 5, the conclusions are presented finally.

A Supply Chain Simulation Model with Customer’s Satisfaction Jianfeng Li, Yan Lin, Feng Jin

International Journal of Advancements in Computing Technology(IJACT) Volume4,Number10,June2012 doi:10.4156/ijact.vol4.issue10.15

125

2. Supply chain system dynamics model

Through the SD approach, a simple two-stage supply chain model is established, which is shown in Fig.1. There are four basic roles in supply chain: customer, retailer, manufacturer and supplier. The retailer orders the goods from the manufacturer and then sales them to customer. The manufacturer orders relative materials from the supplier. The activities such as ordering, transporting, producing, stocking and selling are included in that procedure basically. The variables in that model are following:

Market: the stochastic total demand in the market. Demand : the demand of customer in that supply chain. DemRate: the rate of total demand for that customer. CuRateC or CuRateG: the increasing or decreasing change rate of demand on account of

customer’s satisfaction. SumWeek: the sum of demands for customer in a week InventR or InventM: the inventory level for the retailer or manufacturer. OnR or OnM: the stock of goods on the route for the retailer or manufacturer. PM: the stock of goods produced by the manufacturer. GoR or GoM: the goods flow from manufacturer or supplier. ComeR or ComeM: the goods flow coming to the retailer or manufacturer. ProdM: the goods flow that manufacturer produces. GMT, CMT or PMT: the delay of ordering, transporting or producing for manufacturer. GRT or CRT: the delay of ordering or transporting for retailer. CML or PML: the limit of transporting or producing for manufacturer. CRL: the limit of transporting capability for the retailer. FR or FM: the forecasting value of selling goods for retailer of manufacturer. SafR or SafM: the safety coefficient for the retailer or manufacturer. GR or GM: the attending ordering amount for retailer or manufacturer. Sale: the selling goods flow to the customer.

Here, the demand of customer in the market (Market) is stochastic, and the demand in that supply chain is some part of the total demand in the market (Market*DemRate). The sale amount fulfilling customer’s satisfaction depends on the retailer’s inventory level, in this way, the sale amount is the minimum between the demand and the inventor level of retailer(InventR) in a special time t (fundamental simple time). Supposing when the sale amount can meet the demand of the customer, the number of customers buying in that supply chain will increase next time (CuRateC), otherwise the number of customers will decreases (CuRateG).

The retailer makes ordering decision through exponential smoothing method according to the summation of sale value over some time before, such as a week(SumWeek), What’s more, in order to

GoR Sale GoM ProdM ComeR

GM GR

FM FR

PML

PMT

Demand

ComeM

InventM PM OnR InventR OnM

CMT CRT

CML CRL

GMT GRT

SafM SafR

Manufacturer Retailer

Supplier Customer

SumWeek

GRVGMV

Market

DemRate

CuRateC or CuRateG

Figure1. Supply Chain Simulation Model with Customer’s Satisfaction


126

guarantee the enough inventories to fulfill the customer satisfaction, the retailer multiplies the forecasting value through exponential smoothing method by a safety coefficient(SafR) and considers that multiplied value as the forecasting sale amount next period(FR). Then, the retailer subtracts amount of the goods on the route(OnR) and inventory amount(InventR) from that forecasting sale amount to order from the manufacturer. If the result is larger than zero, the retailer orders that amount, else doesn’t order. The real ordering goods amount (GoR) is depended on the inventory level (InventM) of manufacturer. The same is the manufacturer. The variable FM is the forecasting value through exponential smoothing method, variable SafM is the safety coefficient, and GM is the ordering value for the manufacturer, in which, goods on the route(OnM), inventory amount(inventM), and the goods produced by manufacturer(PM) is also needed to subtract. However, the ordering amount from the supplier is not under the control (GoM). i.e. the supplier can supply any ordering amount that the manufacturer needs.

When the material goods flows in the supply chain, it will cost some time (CRT, CMT or PMT), so is the ordering information exchanges (GRT or GMT), which‘s more, the amount flow of transportation (ComeM or ComeR) and production (ProdM) are limited by their ability (CML, CRL and PML) on account of some reason, such as manufacturer can’t produce enough goods by machine in a fundamental simple time. In this way, the material is flowing over the supply chain continuously, and a supply chain simulation model with customer’s satisfaction is established. 3. Simulation through simulink

Simulink tool is adopted for the above SD simulation model (Fig.2), which is an environment for multi-domain simulation and model-based design for dynamic and embedded systems.

There are mainly five kinds of subsystem:

(a) Subsystem for signal statistic

The Sub_SignalStatistic module in Fig.3 is this kind of subsystem, which is for calculating the demands in some time before. In this module, input is the customer’s stochastic demands and output is

Figure2. Supply Chain Simulation through Simulink

GM

GRMarkRateR

FRS

FMS

Sale

In Out

Sub_SignalStatistic

Limit

In

Lev el

Out

Sub_PM

Limit

In

Lev el

Out

Sub_OnR

Limit

In

Lev el

Out

Sub_OnM

Limit

In

signal

Out

Lev el

Sub_InventR

Limit

In

Out

Lev el

Sub_InventM

In Out

Sub_ExponentialSmoothing_R

InOut

Sub_ExponentialSmoothing_M

Market

Signal

Sale

Demand

Sub_Customer

-K-SafR

-K-

SafM

Random

PML

PML

0

0

-1Z

GRT

-1Z

GMT

Divide

CRL

CRL

CML

CML

FR

GR

ComeM

PM

PM

Sale

FRS

ComeR

ProdM

GoR

Inv entM FM

Inv entR

FMS

OnM

OnM

SumWeek

SumWeek

Market

Demand

GM

GMZ

GMZ


127

the summation of demands in some periods. Supposing the retailer orders the goods from the manufacturer every week, it must calculate the summation of demands this week, and then consider that statistic amount as some basis to forecast the probability sale amount next week. A summation pulse signal every 7 day is produced in this subsystem.

(b) Subsystem for transportation or production

The modules of Sub_OnR, Sub_OnM and Sub_PM in Fig.2 belong to this kind of subsystem, which is for describing output and level condition of transportation or production. In those modules, Inputs are some material inflow and capability limit, and outputs are material outflow and the stocking level, such as the goods on the route and the goods which is being produced. Here, there are some delays in transportation and production in supply chain and the FIFO rule is adopted to shown in that procedure, which is shown in Fig.4.

(C) Subsystem for signal statistic

The modules of Sub_ExponentialSmoothing_R and Sub_ExponentialSmoothing_M in Fig.3 belong to this kind of subsystem, which are made in order to forecast the sale amount next week. Here, exponential smoothing method is adopted. Here supposing the smoothing factor 0.6 .The exponential smoothing module is shown in Fig.5, through that, the forecasting sale value is output, and then, supposing safety coefficient SafR and SafM are both 1.05, the retailer and manufacturer considers that multiplication as the needed sale amount next week.

Figure3. Subsystem for Signal Statistic

1

Out

MATLABFunction

mod1Zero-Order

Hold SumWeek

Switch3

Sum

K Ts

z-1

Discrete-TimeIntegrator 0

Constant

== 0

CompareTo Zero

Clock1

1

In

Figure4. Subsystem for Transportation or Production

2

Out

1

LevelGoR

OnR

Switch2

Subtract3

Subtract2

In1

In2

Out1

OnLine1

Memory2

[0]

IC2

-1Z

CMT

2

In

1

Limit

1

Out1Subtract1

Memory1

[0]

IC1

2

In2

1

In1


128

(d) Subsystem for the inventory

The modules of Sub_InventR and Sub_InventM in Fig.2 belong to this kind of subsystem, which is mainly for showing the change of the inventory level and the selling status to customer or retailer. In those modules, the inputs are the coming goods and the capability limit, and outputs are inventory level and sale amount really. Here because there is no time delay for inputting, transmitting and outputting the goods in the storehouses of the retailer and manufacturer, the blending rule is adopted in those modules, i.e. any goods are picking up at the equal possibility only if they are inputs into the storehouse, which is shown in Fig.6.

(e) Subsystem for customer

Sub_Customer system is the most important and designed for customer’s satisfaction. (Fig,7). In this subsystem, the demand is changed in the percent of CuRateC or CuRateG next time with signal, which is the difference between inventory and customer’s demand. when the sale amount can meet the demand of the customer, the number of customers buying in that supply chain will increase next time (CuRateC), otherwise the number of customers will decreases (CuRateG). What’s more, the proportion of sale amount to the demand in supply chain is considered as the symbol of customer’s satisfaction in some degree. (SatiR).

Figure6. Subsystem for the inventory

3

Level

2

Out

1

signal

InventR

ComeR

Switch1

Subtract3

Subtract2

Memory1

[0]

IC1

2

In

1

Limit

Figure5. Subsystem for Exponential Smoothing

St 1

Out

Zero-OrderHold1

FR

Switch1(7)Subtract

-7Z

Integer Delay(7)

0.6

CofR

Clock1

1

In

Yt

Figure7. Subsystem for the customer

1

Demand

Signal

MarketDemand

SatiR

Product1Product

-1Z

Integer Delay2

-CuRate

IC1

Divide

-C- DemRate

-C-

1-CuRateG

-C-

1+CuRateC

3

Sale

2

Signal

1

Market


129

4. Simulation analysis on customer’s satisfaction

In this simulation system, the simulation start time is 0 and stop time is 70, the solver type is Fixed-step, the solver is adopted in discrete solver, and the fix-step size (fundamental sample time) is 1. The goods for transportation is stochastic between 0 and 1. Some other variables are determined below:

0 0 0 0 0 0OnM PM InventM OnR InventR (1)

0.8PML CML (2)

2CMT PMT CRT (3)

1.05SafM SafR (4)

0.5DemRate (5)

0.01CuRateC (6)

0.02CuRateG (7)

2GRT (8)

2GMT (9)

In the simulation, the random demand of customer in the market (Market) is shown as follow (Fig.8), which is between 0 and 0.1 (the sale amount unit is 10 thousand).

Some proportion (Market*DemRate) of that total demand in the market is considered as the demand

in the supply chain. What’s more, it can be changed with the need gratification or not. Supposing there are some methods of SCM for improving information exchange quickness in the supply chain, in which the GRT and GMT both decrease from 2 to 1 (line’o’ to line’*’), the relative changes for two enterprises are shown as follows:

(a) The Change of Sale

The total sale amount in the supply chain increases a little, which is up from 1.1924 unit to 1.2585 unit in the simulation time (0~70). It means that improvement of information exchange can promote the cycle in supply chain and facilitate the sale amount (Fig9).

Figure8. The random demand of customer in the market


130

(b) The Change of Signal

The Signal is the difference between inventory and demand. When the information exchange is improved in supply chain, it’s shown clearly that the inventory of retail can be more enough to fulfill the demand of the customers (Fig10).

(b) The Change of Satisfaction

The satisfaction (SatiR) is simply expressed by the proportion of sale amount (fulfilling demand) to the demand in supply chain. It’s shown clearly that quickness of information exchange can make enough good sold to promote the customer’s satisfaction (Fig11).

Figure9. The sale amount in the supply chain

Figure10. The signal in the supply chain

Figure11. The satisfaction in the supply chain


131

In a word, it’s shown clearly from the above, that data represented by the line ‘*’ is better, in which, the retailer can sell more goods(Sale), the signal reflecting shortage is less(Signal), and the customer’s satisfaction is higher(SatiR). That means the improvement of supply chain process’s promptness can bring out the more customer’s satisfaction, and then, the sale amount is increased. The indirect profit is gained by means of improving operation’s efficiency in supply chain. 5. Conclusion

This paper establishes a supply chain simulation model with customer’s satisfaction through the system dynamics approach, and then, Simulink tool is adopted for that model. Through the simulation, it can be seen clearly that the value of information is emphasized and customer’s satisfaction can be increased by improving supply chain process’s promptness, and the information exchange’s quickness can bring out large profit indirectly. 6. Acknowledgements

The authors would like to thank peer reviewers for commenting this article. This work is supported by the National Natural Science Foundation of China (70801007, 71072124), the Fundamental Research Funds for the Central Universities (2011QN034, 2011JC008 and 2011QN158), DLMU Outstanding scientific and technological innovation team training fund (2012TD019) 7. References [1] HsuHao Tsai, YenPing Chi, "Trend Analysis of Supply Chain Management by Bibliometric

Methodology", JDCTA, Vol. 5, No. 1, pp. 285-295, 2011 [2] Xing Zhang, Qiuhong Zhao, Guoping Xia, "Research on Integrated Optimization Problem in a

Multi-product Supply Chain Based on Markov Decision Processes", JCIT, Vol. 7, No. 1, pp. 45 -53, 2012.

[3] Amrik S Sohal, Damien J Power, Mile Terziovski, “Supply chain management in Australian manufacturing—two case studies”. Computers & Industrial Engineering, Vol. 43, Issues 1-2, 1 July, pp.97-109, 2002.

[4] Kang Bokyoung, Kim Dongsoo,Kang SukHo, “Supply chain simulation integrating agent-based modeling with social network analysis: A conceptual framework”, ICIC Express Letters, v 6, n 4, pp 1115-1120, April, 2012.

[5] Shotaro Minegishi, Daniel Thiel, “System dynamics modeling and simulation of a particular food supply chain”, Simulation Practice and Theory, Vol.8, No 5, pp.321-339, 2000.

[6] Adhitya Arief, Halim Iskandar, Srinivasan Rajagopalan, “Decision support for green supply chain operations by integrating dynamic simulation and LCA indicators: Diaper case study”. Environmental Science and Technology, v 45, n 23, pp 10178-10185, December 1, 2011.

[7] Srinagesh Gavirneni, Roman Kapuscinski, Sridhar Tayur, “Value of information in capacitated supply chains”, Management Science. Vol.45, No.1, pp.16–24, 1999.

[8] Denis Towill, Mohamed M Naim, Joakim Wikner, “Industrial dynamics simulation models in the design of supply chains”, International Journal of Physical Distribution and Logistics Management, Vol.22, No.5, pp.3–13, 1992.

[9] Jeroen Dejonckheere, Stephen M Disney, Lambrecht Marc, Denis Towill, “The impact of information enrichment on the Bullwhip effect in supply chains: a control theoretic approach”, European Journal of Operations Research, Vol.153, No.3, pp727–750, 2004.

[10] Sammeer Kumar, Marietsa L McCreary, Daniel A Nottestad. “Quantifying Supply Chain Trade-offs Using Six Sigma, Simulation, and Designed Experiments to Develop a Flexible Distribution Network”, Quality Engineering, v 23, n 2, p 180-203, April-June 2011.


132

IJACT899PPL

Documents

Transcript of IJACT899PPL