A Flexsim-based Optimization for the Operation Process of Cold-

Chain Logistics Distribution Centre

X. Zhu1, R. Zhang*2, F. Chu3, Z. He4 and J. Li5

1, 4, 5
School of Mechanical, Electronic and Control Engineering
Beijing Jiaotong University
Beijing, 100044, China
2, 3
School of Economics and Management
Beijing Jiaotong University
Beijing, 100044, China
*rtzhang@bjtu.edu.cn

ABSTRACT
With people’s increasing concern about food safety, cold-chain logistics distribution centre is playing an important role
in preventing food from going bad. Now cold-chain logistics distribution centres have the problems of too much
transportation, low degree of automation, unreasonable layout planning, complex distribution process etc. It is
important to solve these problems in order to achieve efficient distribution. Firstly the modeling and simulation for the
operation process of a fruits and vegetables cold-chain logistics distribution centre by using Flexsim software is
realized. Then the paper analyses the preliminary output data and finds out the bottleneck and idle resources. Finally
this paper makes adjustments for the system to get a better result which hopes to give a reference for the modeling
and simulation for the operation process of other cold-chain logistics distribution centres.

Keywords: Cold Chain Logistics, Distribution Centre, Operation Process, Modeling and Simulation, Flexsim, Optimization.


1. Introduction

Cold-chain logistics is a complicated system is important to solve these problems in order to

engineering which keeps refrigerated and frozen achieve efficient distribution. The analysis for cold-
goods in specified low temperature environment chain logistics distribution centre is one kind of
during production, storage, transportation, process typical random discrete event system. Analytical
and sale in order to guarantee the quality and method can’t analyse and optimize such system
performance of goods [1][2]. In this supply chain completely.
system, cold-chain logistics products have features
of freshness, perishability [3], timeliness, large costs However, simulation technique can show specific
[4], logistics performance [5] etc. It is very important status and find out the bottleneck and idle resource
for cold-chain logistics distribution centre to reduce of the system which gives a detail description about
processing time in lowering the risk of food spoilage. actual process and information of this logistics
system. Now experts mainly focus on the inventory
At present, the research about cold-chain logistics layout of distribution centre [11], picking strategy on
distribution centre is still in early stages. Domestic simulation of logistics system. For example, Hassini
and foreign experts mainly focus on distribution [12] confirms an optimization algorithms for storage
mode selection [6] [7], routing selection based on and distribution of the goods, and the algorithms are
VRP method [8], storage and picking methods [9] successfully applied to a large supermarket
etc., which basically involves the subsystems of distribution centre; Chiun [13] simulates inventory
cold-chain logistics distribution centre. However, few layout and sorting system of a distribution centre by
studies on the overall process of cold-chain using WITNESS software; Zhao [14] builds a
distribution centre have been discussed. Meanwhile, simulation model of picking process and then
domestic cold-chain logistics distribution centres optimizes this model by using Arena software.
have the problems of too much transportation, low Fathollah.M [15] provides a simulation framework of
degree of automation, unreasonable layout a certain supply network in automotive industry in
planning, complex distribution processes etc. [10]. It order to study the impacts of commonality.

270 Vol.12,April2014
 AFlexsimͲbasedOptimizationfortheOperationProcessofColdͲChainLogisticsDistributionCentre,X.Zhuetal./270Ͳ278

By using modeling and simulation theory and of the system precisely (composition of system,
Flexsim software, this paper firstly analyses the system organization, suppose of parameters).
process of a fruits and vegetables cold-chain
logistics distribution centre and then builds a f) Simulation and running. In order to forecast actual
simulation model for this distribution centre; moving principle of the system, it is important to
Secondly the paper gets the efficiency of key understand the output response with different inputs
equipment and workers by using Flexsim so that the and different simulation mechanism.
preliminary output is analysed; Finally the paper
finds out the bottleneck and idle resource and g) Outputs and analysis of the simulation results.
makes several adjustments to improve the
operation efficiency for this fruits and vegetables Specific steps are shown in figure 1.
cold-chain logistics distribution centre.

2. Modeling and simulation steps by using

Flexsim software

Flexsim software is a set of computer 3-D image

processing technology, simulation techniques,
artificial intelligence technology, and data handling
techniques. Flexsim is suited to production-
manufacturing, storage and delivery, transport
system and other fields [16] [17]. Flexsim provides
original data, entering modeling, operation model to
realize simulation experiment and optimizes the
system. Modeling simulation usually follows the
basic steps below:

a)_Survey the system, and then determine

simulated targets.

b)_Collect basic data of the system. The collection

of simulation data involves two aspects. One is that Figure 1. Flexsim modeling and simulation steps.
the data is collected according to the established
simulation goals to keep the system running 3. Modeling for the Operation Process of a
normally. The other is that the data involves starting Cold-Chain Distribution Centre
conditions and internal variable of the system.
3.1 The Operation Process of a Cold-Chain
c) Build the system model. Discrete events should Distribution Centre
define the entities of the system accurately by using
flow chart or network diagram. Flow chart includes Cold-chain logistics distribution centre usually has
three parts: arriving model of temporary entities, the functions of gathering and distributing, linking,
queuing discipline, model of service. low-temperature transport, refrigeration storage,
sorting, loading and unloading, low temperature
d)_Build the simulation model. The process includes processing and logistics information handing etc.
confirming the storage form of model and data, The internal operation process of cold-chain
choosing software development platform and logistics distribution centre has six basic steps,
program design language according to mathematical including ordering, receiving goods, processing,
model and features of the system. storage, sorting and distribution etc. [18] [19].
Various problems may occur during these
e) Model validation. Determine whether simulation processes. The problems which involve the relevant
model and computer language can show every part context of working environments are complex [20].

JournalofAppliedResearchandTechnology 271
 AFlexsimͲbasedOptimizationfortheOperationProcessofColdͲChainLogisticsDistributionCentre, X.Zhuetal./270Ͳ278

Based on the above analysis and considering the 3.3 Data Collection
reality, a fruits and vegetables cold-chain distribution
centre has resources including technical cold By collecting data, the paper uses variable t (day)
storeroom, sorting and packaging equipment, (t•0) representing the refreshing time of the fruits
workers, information service etc. and vegetables. The paper classifies the fruits and
vegetables into three types according to variable t.
This distribution centre is divided into five areas: The kinds of fruits and vegetables for simulation
receiving area, circulating and processing area, include short refreshing time of the goods
tallying and sorting area, storage area, distribution (refreshing time t<3), refrigerated goods (3”t”7) and
and delivery area. The cold-chain distribution centre frozen goods (3”t”7). The arrival interval of the fruits
includes specific areas below: and vegetables follows Poisson distribution (10,1)
(min). The goods arrive 4000 per batch and the
a) Receiving area includes unloading, checking and average weight is 1000 kg per batch. The
storing goods. Fruits and vegetables remain a short distribution centre arranges the machine, equipment
time in consideration of shelf life in this area. This and staff according to different process. Receiving
area involves the elements including checking area is equipped with conveyors, workers, checking
equipment, conveyor and workers. equipment and the time dealing with the fruits and
vegetables per kg follows Uniform distribution(2,5)
b) Circulating and processing area mainly deals with (min); The time that circulating and processing area
fruits and vegetables from receiving area (washing, deals with the fruits and vegetables follows Uniform
partition, packaging etc.). In this area the resources distribution(3,5) (min); Tallying and sorting area is
contain processing equipment, workers etc. equipped with sorting machine and workers, the
time sorting fruits and vegetables per kg follows
c) Tallying and sorting area mainly sorts and packs Uniform distribution (1,3) (min); The working time of
the fruits and vegetables. This area involves the Distribution and delivery area follows Uniform
elements including sorting area, conveying distribution(5,10) (min).
equipment, automatic sorting machine, workers etc.
3.4 Modeling for the Operation Process of a fruits and
d) Storage area contains refrigerated storage and vegetables Cold-Chain Logistics Distribution Centre
freezing storage. The resources that this area
needed are goods shelf, forklifts, workers, freezing According to the analysis of the operation process
equipment etc. for cold-chain distribution centre, this paper uses
flow chart to build the system model. The model is
e) Distribution and delivery area: workers should also the simulation module for the operation
send fruits and vegetables according to customers’ process of this Cold-Chain Logistics Distribution
demand, reasonable route, and time restriction. The Centre. Compared with actual process of this cold-
resources of this area involve vehicles, workers, chain distribution centre, the simulation module is
conveying, equipment etc. As figure 2 shows, the reasonable. Figure 3 shows the simulation model
specific operation procedures are as follows for operation process of the system

3.2 Simulation Targets 4..Simulation and results analysis for the

Operation Process of a Cold-Chain Logistics
As cold-chain goods have features of perishability, Distribution Centre based on Flexsim
short refrigeration cycle etc. This paper sets two
simulation targets. One is transport turnover 4.1 Operation Parameter and Elements of Simulation
volumes of fruits and vegetables. The other is
operation capability indicator of the distribution Set simulation time: the unit time of simulation is 1
centre (efficiency of workers, machines) in order to min. Determine the simulation working hours as
find out the bottleneck of this operation system and 8/day (480 min), and get the results of simulation.
improve operation process balance of the fruits and Table 1 below shows the entity elements of
vegetables cold-chain distribution centre. simulation model.

272 Vol.12,April2014
 AFlexsimͲbasedOptimizationfortheOperationProcessofColdͲChainLogisticsDistributionCentre,X.Zhuetal./270Ͳ278

Figure 2. The operation process of a fruits and vegetables distribution centre.

3. The weight of fruits and vegetables is equal

without distinction.

4. The refreshing time is not the same for different

fruits and vegetables and the maximum refreshing
time is 7 days.

5. Simulation module includes receiving area,

circulating and processing area, tallying and
sorting area, storage area, distribution and delivery
area. These areas are consists of specific
operation process.

4.3 Output and Analysis of the Results

By using the Flexsim software and running the

simulation model, the paper get the results of
working state statistics of machines (Table 2),
working state statistics of conveyors (Table 3) and
working state statistics of workers (Table 4).

Table 2 shows the busy rate of Processor 13

(Processor of tallying and sorting area) is 39.94%.
So the use efficiency of Processor 13 is low.

Figure 3. Simulation model for operation process Table 3 shows the conveying rate of Conveyor
of a fruits and vegetables distribution centre. 19, Conveyor 22, Conveyor 25 is respectively
68.1%, 84% and 28.6%. However, the conveying
rate of conveyor 25 (Sorting-delivering conveyor)
4.2 Set the Assumptions is only 28.60%, and the block rate of conveyor
25 is 38.60%. It represents that staff or
Simulation model is abstraction from real system to equipment of conveyor 25 didn’t deal with the
mathematical model. So the paper sets several goods in time, in other words, staff or equipment
assumptions by surveying this distribution centre. of this area is not enough.
1. Fruits and vegetables are sent to distribution Table 4 shows the idle rate of Transporter 20
centre by fixed batch and the arriving time of fruits (Transporter in distribution centre), Operator
and vegetables each day follows Poisson 29(Operator in storage area), Operator 28
distribution. (Operator in distribution and delivery area) is
respectively 59.1%, 96.3%, 56.9%. It indicates that
2. The fruits and vegetables arrive continuously the system has bottleneck from tallying and sorting
during simulation. area, storage area to distribution area.

JournalofAppliedResearchandTechnology 273
 AFlexsimͲbasedOptimizationfortheOperationProcessofColdͲChainLogisticsDistributionCentre, X.Zhuetal./270Ͳ278

Name of elements Type Remark

Item 1 Item Goods of short refreshing time (t<3)
Item 2 Item Refrigerated goods(3”t”7)
Item 3 Item Frozen goods(3”t”7)
Source 1 Producer Produce Item 1
Source 2 Producer Produce Item 2
Source 3 Producer Produce Item 3
Transporter 22 Transporter Transporter of receiving area
Transporter 20 Transporter Transporter of distribution centre
ASRSvehicle 22 Vehicle Vehicle of storage area
Operator 24 Operator Operator of sorting are
Operator 29 Operator Operator of storage area
Operator 28 Operator Operator of distribution and delivery area
Processor 19 Processor Processor of receiving area
Processor 9 Processor Processor of circulating and processing area
Processor 13 Processor Processor of tallying and sorting area
Processor 24 processor Processor of distribution and delivery area
Conveyor 19 Conveyor Receiving-processing conveyor
Conveyor 22 Conveyor Processing-sorting conveyor
Conveyor 25 Conveyor Sorting-delivering conveyor
Sink 5 Sink place of receipt

Table 1. Entity Elements of Simulation Model.

Object Busy% Idle% Block%

Processor 19 69.87 30.13 0.00
Processor 9 79.44 20.56 0.00
Processor 13 39.94 60.06 0.00
Processor 24 61.95 16.90 0.00

Table 2. Working State Statistics of Machines.

Object Conveying% Empty% Block%

Conveyor 19 68.10 13.20 18.70
Conveyor 22 84.00 10.00 0.00
Conveyor 25 28.60 32.80 38.60

Table 3. Working State Statistics of Conveyors.

274 Vol.12,April2014
 AFlexsimͲbasedOptimizationfortheOperationProcessofColdͲChainLogisticsDistributionCentre,X.Zhuetal./270Ͳ278

The bottleneck is restriction of conveying vehicle

Travel Travel
Object
empty%
Idle%
loaded%
which can be solved by increasing the number of
vehicles.
Transporter 22 23.20 21.20 33.40
Transporter 20 17.00 59.10 18.40 2. The work efficiency of equipment in distribution
ASRSvehicle and delivery area is very low. The reason is that
0.00 9.90 82.60
22 the machine is idle for a long time to wait for
Operator 24 11.00 7.80 37.50 conveying workers which can be solved by
Operator 29 0.90 96.30 2.00 increasing a staging areas.
Operator 28 17.70 56.90 17.60
3. The idle rate of Operator 29 (Operator of
Table 4. Working State Statistics of Workers. storage area) is high, and a forklift can replace
operator 29 to solve this problem.

Object Input Output Turnover rate 5.2 Specific Improved Measures and Optimized
Item 1 47 32 68.10 Simulation Modeling Graph
Item 2 29 3 10.35
1. Adjust the number of machine and workers.
Item 3 23 3 13.04
Increase a vehicle ASRSvehicle 29 in storage area
Total 99 38 38.38 and increase. Processor 30 in distribution and
delivery area, and then cut Operator 29.
Table 5. Statistics of Elements Input and Output.
2. Change storage rule of goods shelf in frozen
The total quantities of Item 1, Item 2, and Item 3 storage area. When storing goods storage area,
are absorbed by Sink 5. Table 5 shows the take the distance into consideration first instead of
turnover rate of system. randomly putting goods.

Table 5 shows that the turnover rate of Item 1 is 3. Increase a staging area in front of distribution
68.10%; the turnover rate of Item 2 is 10.35%; the and delivery area. Optimized simulation modeling
turnover rate of Item 3 is 13.04%. graph is shown below (figure 4).

The total turnover rate is 38.38%. It means the 5.3 Contrastive Analysis
turnover rate of goods is extremely low and the
operation process of distribution centre has Table 6 below shows the contrast of two simulation
problems of delay and blocking. results.After the system is optimized, the turnover
rate per day of all kinds of fruits and vegetables
From the above analysis of simulation results, it surpasses 87%. The turnover rate of Item 1
shows that the utilization rate of equipment and (refreshing time t<3) reaches to 91.43% which is
workers etc. is low. The paper takes several improved by 23.33%; the turnover rate of Item 2
measures to improve operation efficiency of this (refrigerated goods 3”t”7) reaches to 88.24%; the
cold chain distribution centre. turnover rate of Item 3 (frozen goods 3”t”7)
reaches to 87.5%.
5. Improvement for the Operation Process of a
Cold-Chain Logistics Distribution Centre At the same time, the utilization rate of equipment in
the distribution centre is improved extremely. The
5.1 Problem Analysis using rate of worker 3 (operator 28) is improved by
31.5%; the using rate of conveyor 19 (receiving-
1. The backlog of goods from tallying and sorting processing conveyor) is improved by 16.70%; the
area to frozen storage area leads to low flow rate using rate of processor 29 in distribution and
of Item 2, Item 3. delivery area is improved by 12.96%.

JournalofAppliedResearchandTechnology 275
 AFlexsimͲbasedOptimizationfortheOperationProcessofColdͲChainLogisticsDistributionCentre, X.Zhuetal./270Ͳ278

Figure 4. The Optimized Simulation Model.

Evaluation
Name of entity Before optimization % After optimization % Contrast %
index
Item 1 68.10 91.43 23.33
Turnover rate Item 2 10.35 88.24 77.89
Item 3 13.04 87.50 74.46
Processor 19 69.87 73.45 3.58
Processor 9 79.44 81.08 1.64
Use efficiency
Processor 13 39.94 43.49 5.55
of processors
Processor 29 61.95 74.91 12.96
Processor 30 — 67.88 —
Use rate of Transporter 22 78.80 78.80 0.00
transporter Transporter 20 40.90 40.90 0.00
Use rate of ASRSvehicle 22 90.10 80.20 -9.90
ASRSvehicle ASRSvehicle 29 — 78.60 —
Operator 24 92.20 91.10 -1.10
Use rate of
Operator 29 3.70 — —
operators
Operator 28 43.10 74.60 31.50
Conveyor 19 68.10 84.80 16.70
Use rate of
Conveyor 22 84.00 86.99 2.99
conveyors
Conveyor 25 28.60 27.40 -1.20

Table 6. Contrast of two Simulation Results.

276 Vol.12,April2014
 AFlexsimͲbasedOptimizationfortheOperationProcessofColdͲChainLogisticsDistributionCentre,X.Zhuetal./270Ͳ278

6. Conclusion [7] Y.Y. Zhao, “Optimization research of food cold chain

logistics soft-windows distribution mode,” Anhui
On the basis of research status for cold-chain Agricultural Science, Vol.37, no. 17, pp. 8235-8238,
logistics distribution centre, this paper analyzes a 2009.
fruits and vegetables Cold-Chain Logistics [8] C.D.Tarantilis and C.T.Kiranoudis, “A meta-heuristic
Distribution Centre and collects the fundamental algorithm for the efficient distribution of perishable foods,”
data, then builds the simulation model for the Journal of Food Engineering, Vol. 50, pp.1-9, 2001.
operation process of the Cold-Chain Logistics
Distribution Centre. By using Flexsim software, this [9] Petersen and Charles, “Improving order-picking
paper obtains the basic data of equipment and performance through the implementation of class-based
workers and finds out the bottleneck and idle storage,” International Journal of Physical Distribution &
resources. Finally the paper provides specific Logistics Management, Vol. 34, no. 7, pp. 524-544, 2004.
improved measures to make the turnover rate of
[10] R.Ye et al., “Researching on benefit assignment of
cold chain goods and the using rates of equipment logistics system in cold chain for food,” 2011 2nd IEEE
and workers increase greatly. International Conference on Emergency Management
and Management Sciences, pp.513-515, 2013.
Acknowledgements
[11] S.J.Mason et al., “A Simulation Based Approach for
This work is partially supported by a Research Fund for Dock Allocation in a Food Distribution Centre,”
the Doctoral Program of Higher Education of China with Proceedings of the 40th 2008 Winter Simulation
Grant No. of 20120009110009, a major program of the Conference, pp. 2750-2755, 2008.
National Social Science Fund of China 13&ZD026, and
an EU 7th Framework Programme fund with contract [12] Hassini.E and Vickson.R.GA, “two-carousel storage
number EU IRSES 612546 ENRICH. location problem,” New Jersey: Institute of Electrical and
Electronics, pp.1177-1184, 2000.
References
[13] L.M. Liu, “Clustering Techniques for Stock Location
[1] X.L. Ning, “Chinese food cold chain logistics mode,” and Order-Picking in a Distribution Centre,” Computers
China Logistics & Purchasing, vol.2, pp. 68-69, 2010. & Operations Research, no. 26, pp. 989-1002, 1999.

[2] H.G. Lan and Y.B. Tian, “Analysis of the demand [14] Y.D. Zhao, “The picking process simulation analysis
status and forecast of food cold chain in Beijing,” Journal of a distribution centre,” Information Technology, no. 31,
of Industrial Engineering and Management, Vol. 6, no. 1, pp. 36-39, 2007.
pp. 346-366,2012.
[15] Fathollah, M et al., “Developing a Conceptual
[3] T.Y. Zheng and Michael Wang, “Application of Quality Framework for Simulation Analysis in a Supply Chain
Cost and Quality Loss Function in Food Supply Chain Based on Common Platform (SCBCP),” Journal Of
Systems Modeling,” Journal of System and Management Applied Research And Technology, Vol. 7, no. 1, pp.
Sciences, Vol. 2, no. 1, pp. 19-30, 2012. 163-184, 2009.

[4] Q.G. Wu, “The current situation and the counter [16]_L. Gang, “An Shengmin, Liao Junjun.The
measures of China’s cold chain logistics development,” Optimization and Simulation of Automobile Spread
China Business and Market, no. 2, pp. 24-28, 2011. Production Line,” Industrial Engineering, Vol. 11, no. 6,
pp. 71-76, 2009.
[5]_Cedillo-Campos._M and Sanchez-Ramirez.C,
“Dynamic Self-Assessment of Supply Chains [17] J.Q. et al., “The magnesium-smelting production line
Performance: an Emerging Market Approach,” Journal simulation and optimization based on Flexsim,” Industrial
of Applied Research and Technology, Vol. 11, pp. 338- Engineering, Vol.12, no. 2, pp.78-81, 2009.
347, 2013.
[18] Y.Bai et al., “How to Locate the Cold Chain Logistics
[6] Z.G. Liu and K. Wang, “Application of joint distribution Distribution Centre Based on the Gray Comprehensive
strategy in the cold chain logistics,” Logistics Evaluation,” 2010 International Conference on E-Product
Technology., Vol. 30, no. 10, pp. 1-3,2007. E-Service and E-Entertainment, pp.1-4, 2010.

JournalofAppliedResearchandTechnology 277
 AFlexsimͲbasedOptimizationfortheOperationProcessofColdͲChainLogisticsDistributionCentre, X.Zhuetal./270Ͳ278

[19] Y.D.Huang et al., “The simulation optimization of

goods loading model in distribution centre,” Logistics for
Sustained Economic Development Infrastructure,
Information, Integration Proceedings of the 2010
International Conference of Logistics Engineering and
Management, no. 387, pp. 2429-2436, 2010.

[20] Ke. CK. “Research on Optimized Problem-solving

Solutions: Selection of the Production Process,”. Journal
of Applied Research and Technology, Vol. 11, pp. 523-
532, 2013.






















278 Vol.12,April2014