International Journal of Advanced Engineering, Management and Science (IJAEMS) [Vol-2, Issue-10, Oct- 2016]

Infogain Publication (Infogainpublication.com) ISSN : 2454-1311

Application of Discrete Event Simulation in

Industrial Sectors: A Case Study
Ndudim Ononiwu1, Mary Macharia2, Mohamed Al-Ali3
1
Department of Advanced Industrial and Manufacturing Systems, Kingston University, London
2
Department of Engineering Projects and Systems Management, Kingston University, London
3
Department of Engineering Projects and Systems Management, Kingston University, London

Abstract— Discrete Event Simulation (DES) has become among others. The second approach discrete event
a useful tool in the evaluation of changes that may bring simulation treats systems as a network of queues and
positivity to manufacturing and process organizations for activities whereby there is a change of state of the entity
both goods and services provision. The main focus of any at discrete points in time along the process line. DES has
business entails the reduction of cost and lead time while been applied extensively in the area of operation
increasing profits and this is why refining of production management and manufacturing.
processes is essential. This paper reports the application
of DES in two case studies. The case studies selected for
the implementation of Discrete Event Simulation are a
packaging company and a local mobile phone service
provider using the software FlexSim. The implementation
aims at showcasing the versatility and its ability to
provide the relevant data to make more informed decision Fig.1: Schematic representation of the DES process
while optimizing the entire processes involved in
production. DES works with the rule that an entity must pass through
Keywords — Implementation, Discrete Event a series of activities and in between an entity and an
Simulation (DES), FlexSim, production processes. activity, a queue exists where the entity waits until it is
ready to be processed. Fig 1 represents the DES process.
I. INTRODUCTION In the manufacturing industry, DES by virtue of its
The emerging trend in the manufacturing industry, which characteristics has been applied in different areas along
moves towards globalization and decentralization of product development including; inventory, production
managerial decision making processes require floor and supply chain management. Fig 2 below shows
instantaneous information exchange between different areas in the manufacturing industry simulation can be
points in the development of goods and services [7]. This applied for investigation purposes.
instantaneous information described can be obtained
through the process of simulation modelling. The
simulation of manufacturing process proves to extremely
simplified and cost effective. It is also capable of
providing the user with past, present and future data
(forecasting).
Simulation techniques have over time proved to be
powerful tools for evaluating different processes in key
areas of research and development. [5] argues that there
are two approaches to simulation modelling, system
dynamics and discrete event simulation (DES). The
Fig.2: The investigated domains of contemporary
system dynamic approach to simulation aims at modelling
manufacturing
processes that are continuous in nature. This type of
Source; [7]
simulation is mainly used in areas of fluid mechanics,
thermodynamics among other engineering scenarios. The
Discrete event simulation (DES) as a method of
application of system dynamics is not limited to
simulation has been present since the 1960s and overtime,
engineering applications, as use as been found in the area
it has evolved progressively for wider industrial
of economics, ecological systems, population modelling
www.ijaems.com Page | 1751
International Journal of Advanced Engineering, Management and Science (IJAEMS) [Vol-2, Issue-10, Oct- 2016]
Infogain Publication (Infogainpublication.com) ISSN : 2454-1311
applications. [1] states that till date, there have been four I. Using simulation to validate the existing plant
generations of software products being used for DES. layout
This evolution has been from the use of programming in II. Simulating the manufacturing process in order to
high level languages to interactive computer aided identify bottleneck activities and proffer the
software. The generation of DES software are outlined optimal solutions.
below in table 1; In order to achieve the set aim, [2] set out steps that were
Table.1: Generations of DES software followed sequencially. These steps were, model
development, model verification and model validation.
In the model development step, the factory’s current
layout was studied and the process times and probability
distribution of these ‘process times’ were evaluated.
Assumptions were made in the development of the DES
model. According to [2], these assumptions are;
I. Operators are always available during the two
shifts (1 shift =8 hours), an hour of lunch break
and fifteen minute of tea breaks will be provided.
II. There are no significant equipment or station
failures.
III. The production is continuous.
Source: [1] IV. Processes for 60 cells and 72 cell modules have
independent process time in pre-lamination
FlexSim is a 3-D DES computer software utilized for the stage.
simulation, validation, prediction and the visualization of V. Materials are always available at each assembly
systems in industrial sectors such as manufacturing, station.
healthcare, logistics and material handling among others. VI. Transfer times between stations are taken as
The described computer software has the capabilities of constant values
optimizing a planned and already existing process for the Furthermore, [2] explains that in data acqusition, four
purpose of lean production, reduction of production cost, modules were required to supply all the data required to
increasing revenue and productivity. perform the simulation. These modules are;
I. Capacity Input: Information provided are the
II. LITERATURE REVIEW number of machines in each process and the
Several works have been carried out over the years schedule cycle of workers to operate.
whereby DES has been applied to several other industries II. Product specific data: All the data required for
for the purpose of increase in productivity, cost reduction, processing each product in terms of time,
issues of health care among others. For the purpose of this production rate and batch size.
work, two sources of literature will be reviewed. These III. User specific data: The modeller can customize
sources of literature will be reviewed in terms of the the simulation by changing certain requirements
description, methodologies collection and interpretation as needed in the model.
process/results obtained. IV. Scheduling production plan data: This provided
A research with the title ‘Validating the Existing Solar data for the time taken by the plant to complete a
Cell Manufacturing Plant Layout and Pro-posing an demand.
Alternative Layout Using Simulation’ was carried out to
obtain the best layout in order to optimize the use of
resources in terms of operators and machines. This
optimization aimed at the reduction of production costs
while increasing productivity. The case study chosen for
this work was a solar photovoltaic (PV) module
manufacturing company located in southern India. In
order to achieve the set out aims, the DES tool ARENA©
was used to ‘examine the key performance variables’ by
using DES to obtain the best layout of operators and
equipment. [2] Fig.3: Simulation Model data flow
The objectives chosen for this works were; Source [2]
www.ijaems.com Page | 1752
International Journal of Advanced Engineering, Management and Science (IJAEMS) [Vol-2, Issue-10, Oct- 2016]
Infogain Publication (Infogainpublication.com) ISSN : 2454-1311
Model verification, the next step in the simulation process difficult for these orders to be fulfilled promptly, hence
was conducted to make sure that the model worked as it the end of the month syndrome.
was intended to. This step was described as an iterative The aim of the work as described by [3] was to suggest
activity performed at each step along the model alternatives in overcoming the experienced end of the
development stage. month syndrome after thorough analysis of the exsiting
Model validation describes the demonstration to confirm processes using modelling techinques. The objectives of
that the model developed is able to describe the system it undertaking the researcher were
was modelled after. [2] split this process into two stages. I. Modelling and Simulation of manufacturing
The face validation and statistical validation. line to analyse end of the month syndrome
The simulation was run for 400 hours during which the II. Reduce bottlenecks.
results represented in table 2 were obtained; III. Prevent underutilization of resources
Table.2: Results obtained after simulation run of 400 IV. Optimize system performance
hours V. Inclusion of new orders/customers
VI. Capacity improvement
The DES carried out by [3] was done using ARENA©.
Based on the data collected during the discussion with the
prodeuction heads, it was deduced that the ‘end of the
month syndrome’ was caused in the Yamaha gear shifter
manufacturing line (YMG Line). Further study by [3]
Sources; [2] reviewed that the contributing stages to the end of the
From the simulation, it was established that there was a month syndrome in the YMG line were rough honing,
bottleneck in the lamination stage and the assembly radius milling, pin machining, bending and pad grinding.
station consumed a large amount of time. The data collected which was used in the simulation
Based on the recommendations made by [2] with the modelling are;
objective optmizing the average resource utlization, work I. Machine wise data: The data under this category
in process, system through put and % reduction in idle includes; number of machines, standard cycle
cost. The results obtained after the simulation with the time, manpower, start up loss, end up loss, target
recommendations are shown in table 3 below, output, achieved output/shift, setting time,
Table.3: Results obtained after optimization rework and rejection.
II. Machine cycle time
III. Dispatch plan
IV. Delay timing in the processing line.
Based on the data obtained, a virtual model of the process
was built and fine-tuned until an ‘As-is’ condition was
achieved [3]. 15 days of production runs for the YMG-7
and YMG-8 was simulated and results as described in
Sources; [2]
table 4 below were obtained;
[3] describes ‘end of monthe syndrome’ in manufacturing
Table.4: Results obtained
settings as a situation whereby manufacturers experience
difficulties in achieving their full potential If products
exceeds 25% of the set monthly shipment plan in the last
week of the month or more than 33% of their quarterly
shipment plan in the last month of the quarter. This gave
rise to the research titled ‘Modelling and Simulation of
Manufacturing Process to Analyse End of Month
Syndrome’. The case study chosen is into the manufacture
of two wheeler parts and special purpose machines. It is a
tier 1 supplier to the seven customers which include;
Honda, Ducati, Bajaj, Piaggio, Yamaha, Motorai Miner,
Motto Guzzi. According to Kulkarni & Prashanth,
discussions with the production heads of the company
showed that it was overloaded with orders making it Source; [3]

www.ijaems.com Page | 1753

International Journal of Advanced Engineering, Management and Science (IJAEMS) [Vol-2, Issue-10, Oct- 2016]
Infogain Publication (Infogainpublication.com) ISSN : 2454-1311
Based the results obtained by [3] from the simulation The resources employed in the simulation any model
model, it was shown that the bottleneck was the radius- using FlexSim are categorised into fixed resources and
milling machine. They recommended that ‘by increasing task executers. Fixed resources as the name suggest are
radius milling machine capacity to 2 shifts and 50% of referred to as those resources which are stationery while
down time reduction on delay time in the line results into task executors are used to perform the functions of
achieving the production schedule target on date thus moving items to and from the point of production. Table
reducing the effects of end of month syndrome 5 and Table 6 below represent these resources;
substantially’. Table.5: Fixed resources in Case study 1

III. CASE STUDY IMPLEMENTATION

As described in the abstract, the DES tool was
implemented in to scenarios using the FlexSim 3D
simulation software. These case studies are described
below;
3.1. Case study 1; Simulation of a customer service
center in a mobile phone store
A customer service centre of a local mobile phone
company has customers arriving exponentially with an
average of 5 minutes. After entering the company, all
customers go directly to the ticket machine to take an Table.6: Task executers in Case study 1
ordered ticket. They take 1±0.5 minute uniformly to do
so. Then they go to the waiting area and sit down. The
customers are waiting to request 1 of 3 representatives.
Historical data shows that 40% of the customers want to
The completed representation of the defined problem
see the representative of Network Service, 40% want to
described in case study 1 is shown in figure 4.
see the representative of Repair Service, and 20% want to
see the representative of Billing Service. When the
desired representative is available, a staff will escort the
customer from the waiting area to the representative.
When seeing the representative of Network Service, the
visit lasts to be normally distributed with a mean of 7
minutes and a standard deviation of 2 minutes. When
seeing the representative of Repair Service, the visit takes
to be normally distributed with a mean of 10 minutes and
a standard deviation of 2 minutes. The visit of
representative of Billing Service lasts to be normally
distributed with a mean of 5 and a standard deviation of 1.
Some of the customers also see the Promotion Service Fig.4: Completed FlexSim model Case study 1
after they have finished their servicing with the 3.2. Case study 2; Simulation of a tin can production
representatives of Network Service, Repair Service or company
Billing Service. An Assistant Manager will escort such A tin can arrives a Queue every 14 seconds exponentially
customers to his room and introduces the promotion distributed and then routed to any one of 3 machines
package individually. The introduction takes to be where it is processed for 20 seconds (24 seconds for
normally distributed with a mean of 10 minutes and a rework). Machined parts are placed in a common queue
standard deviation of 3 minutes. Marketing Department and wait to be tested, 20% are found faulty and must be
found that 10% of the customers that enquire Network reprocessed. The test time is a constant 9 seconds. Parts
Service listen to the introduction of Promotion Service, passing test enter another Queue and wait to be packaged
20% that see the Repair Service listen to the introduction at an automatic packaging machine.
of Promotion Service, and 5% who have seen the The packaging machine accumulates 10 products into a
representative of Billing Service listen to the introduction box and then closes, seals, and labels the box in 57
of Promotion Service. After listening to the introduction seconds. The supply of boxes comes from a Queue fed by
of promotion with the Assistant Manager, the customers a box forming machine having a cycle time of normal
leave. (50,2) seconds. The box former jams regularly per
www.ijaems.com Page | 1754
International Journal of Advanced Engineering, Management and Science (IJAEMS) [Vol-2, Issue-10, Oct- 2016]
Infogain Publication (Infogainpublication.com) ISSN : 2454-1311
Weibull (151.1,50,24.9) and takes between 20 and 30 Table.8 Task executers for Case study 2
seconds to fix uniformly distributed. The Tester has an
output of 80% acceptable and 20% faulty by chance. As
performed in Case study 1 the fixed resources and task
executers were defined as shown in tables 7 and 8.
Table.7 Fixed resources for Case study 2

Figure 5 represents the model created based on the

available case study.

Fig.5: Completed FlexSim model Case study 2

IV. RESULTS AND DISCUSSION

4.1. Case study 1
The simulation model was run from 9:00am to 1:00pm (4 hours), which produced the results shown in tables 9 and 10.
Table.9: FlexSim summary report for case study 1

www.ijaems.com Page | 1755

International Journal of Advanced Engineering, Management and Science (IJAEMS) [Vol-2, Issue-10, Oct- 2016]
Infogain Publication (Infogainpublication.com) ISSN : 2454-1311

Table.10: FlexSim state report for Case study 1

Based on the report generated in tables 9 and 10, the following observations were made;
I. The percentage of the staff’s time spent in use as seen in the FlexSim state report is;
(100%-idle time); (100%-95.77%) = 4.23%
II. The percentage of the assistant manager’s time spent in use is; (100% - idle time);
(100% - 70.40) = 29.60%
III. From the FlexSim state report the utilization of the 3 representatives are; Network service representative 61.34%,
Repair service representative 76.93%, Billing service representative 26.44%.
IV. On the average, the waiting time for customers in the waiting area is 2.614777mins, 3mins approximately.
V. The most number of customers waiting to see the representative at one time as seen in the FlexSim report is 2
customers.

4.2. Case study 2

The simulation was run for 9 hours which represents the time for required for the completion of one shift. The results
obtained as shown in table 11.

www.ijaems.com Page | 1756

International Journal of Advanced Engineering, Management and Science (IJAEMS) [Vol-2, Issue-10, Oct- 2016]
Infogain Publication (Infogainpublication.com) ISSN : 2454-1311
Table.11: Summary report for Case study 2.

The reports obtained from the analysis of the data [2] S. V. Kulkarni and L. Gowda, “Validating the existing
provided in the generated report shows that the total solar cell manufacturing plant layout and propsing an
breakdown time for the packaging machine which was alternative layout using simulation,” Springer, 2012.
reported to be approximately one hour. This constitutes [3] S. V. Kulkarni and K. G. Prashanth, “Modeling and
11% of the total production time. The analysis further Simulation of Manufacturing processes to analyze end
shows that the box making machine has the highest of month syndrome,” Springer, 2012.
processing time of 7.6 hours.
[4] T. Sedding and G. Sun, “Application of discrete event
simulation to the activity based costing of
V. CONCLUSION
manufacturing systems,” Elsevier, 1999.
DES has overtime proved to be an important method of
solving a large variety of industry related problems [5] D. Mourtzis, M. Doukas and D. Bernidaki1,
spanning from production issues to managerial decision “Simulation in Manufacturing: Review and
making processes. The concept of DES has found Challenges,” 2014.
importance in other areas not limited to healthcare,
inventory control, and supply chain management among
others. As demonstrated in the sources of literature
reviewed as well as the selected case studies, employing
DES into the production processes will go a long way in
not only identifying defective areas along the production
process but also create an avenue to recommend
solutions.

REFERENCES
[1] E. Babulak and M. Wang, Discrete Event Simulation,
2010.

www.ijaems.com Page | 1757