Proceedings of the First Australian International Conference on Industrial Engineering and Operations

Management, Sydney, Australia, December 20-21, 2022

Towards Developing Real-Life Models: A Simulation

Modelling of an Ideal Vehicle Assembly Line Using Arena
Simulation
Gbeminiyi John Oyewole
Department of Industrial Engineering, Operations Management and Mechanical Engineering
Vaal University of Technology, Vanderbijlpark, South Africa
gbeminiyio@vut.ac.za

Tsosang Daniel Khitleli

Department of Industrial Engineering, Operations Management and Mechanical Engineering
Vaal University of Technology, Vanderbijlpark, South Africa
danielkhitleli@gmail.com

Abstract
We present an ideal model of a typical vehicle assembly plant by generalizing selected case studies from the literature.
The ideal model was extended to capture certain realities of material handling to obtain new scenarios. The Arena
simulation platform was used to model both the base and scenario models. Data and probability distribution were
obtained from specifications from the literature and random projections. Using the number of outputs from the vehicle
assembly plant, the base model and the scenarios were analyzed. Results of the sensitivity analysis of the scenario
models confirm that as real-world systems emanate from the ideal systems for example by capturing material handling
realities, the production or service outputs are constrained either positively or negatively. Thus, showing the need for
understanding different configuration settings of material handling equipment when scaling up production outputs is
desired. The simulation modelling and analysis employed are useful both as classroom illustrations on material
handling, aid for engineering education and other users in need of general understanding and applications of discrete
event simulation modelling.

Keywords
Generic vehicle assembly, Material handling modelling, Scenario analysis and arena simulation

1. Introduction
The automotive industry is one of the several manufacturing industries that contributes positively to the industrial and
economic base of countries. The automotive industry also largely impacts diverse sectors of a country's economy due
to its end products that facilitate the distribution and movement of goods and persons. Developed countries such as
Germany and Japan, are known to have several lines of vehicle assembly plants (Wirabhuana et al., 2008). On another
hand, developing countries such as Kenya, South Africa and Nigeria have seen the need to scale up their low-capacity
vehicle assembly plants due to the large volume of imported vehicles (Ikome et al., 2022, Gorham, 2022).

The vehicle assembly process could come with several variations depending on the type of vehicle being manufactured
such as trucks, cars, bus. However, the importance of a generic assembly procedure for different types of vehicles was
discussed by (Wang et al., 2011, Wy et al., 2011). According to Wang et al. (2011), Oumer et al. (2016) the generic
assembly starts from the arrival of different vehicle parts into the assembly facility. The parts are then distributed into
different sections of the assembly plants based on the manufacturing schedule. Typical sections/areas of the assembly
plant include the body shop, paint shop, assembly section, buffer area, quality inspection and motor pool section.
Important in the vehicle assembly process irrespective of the manufacturing plan is the need to ensure that the process
is efficient and effective, delivering the expected number of vehicles while ensuring resources are adequately utilized.
Therefore, productivity improvement through the optimization of several decisions and resources such as indicated by
Oyewole and Adetunji, (2020) and Oyewole, (2020) become critical. The simulation modelling technique has been
very useful as a productivity, analytical and improvement tool in a lot of manufacturing environments (Wy et al.,

© IEOM Society International

171
Proceedings of the First Australian International Conference on Industrial Engineering and Operations
Management, Sydney, Australia, December 20-21, 2022

2011, Dengiz et al., 2016). In the vehicle assembly plants several studies have been conducted using the concepts of
simulation modelling such as (Wang et al., 2011, Dewa and Chidzuu, 2013, Zhao and Li, 2015, Oumer et al., 2016,
Moon et al., 2021, Thanou and Matopoulos, 2021).

In this paper, we present a discrete-event simulation model of an ideal vehicle assembly line. This is done by
generalizing selected vehicle assembly models presented in the literature. We also extend the models by considering
some material handling realities such as the modelling of a type of conveyor within the assembly plant and the use of
resource-constrained material handling. Using the ideal or base model as a benchmark, initial parameters of the
extended models were obtained. Furthermore, sensitivity analysis was performed on the extended models to obtain
insights into the impact of selected variables on the type of material handling system introduced. We also indicate the
assumptions we have introduced in obtaining the basic and extended models. We specifically use the Arena simulation
modelling platform by Rockwell automation due to its convenience and vast use in the modelling of Discrete Event
Simulation (DES) studies useful for our generic vehicle assembly (Guiguet and Pons, 2022).

This study hopes to be useful to the general interest readers in engineering towards teaching the development of
complex models from basic models. In addition, this study could assist in understanding selected material handling
design specifications and data to collect for real experimentation. For example, during the covid-19 pandemic, site
visits to local plants were limited for real data collection. Therefore, understanding certain models were possible
through sensitivity analysis of appropriate projected data.

Section 2 presents a related study on the use of arena simulation software in vehicle assembly. In Section 3, we discuss
the model formulation and assumptions. The experimentation and analysis performed are presented in section 4. We
present the results obtained and provide some insights on the findings in section 5. This study ends in section 6 with
a conclusion and possible future directions.

2. Related Study
In this section, we present a selected review of studies that have considered a generic vehicle assembly plant, including
considerations for material handling and have used the Arena simulation platform in modelling. The importance of
using a generic model due to the time-consuming and error-prone nature of building a real manufacturing simulation
model was emphasized by Wy et al. (2011). The arena simulation software is built and updated with specific material
handling constructs or logic that enables the incorporation of material handling such as conveyances, and transporters
into the modelling operations of a typical manufacturing plant (Kelton et al., 2015, Wilson et al., 2022).

Wirabhuana et al. (2008) performed a scenario analysis of a general truck assembly problem using Arena and
considered performance measures such as outputs, cycle time, and line efficiency. Their focus was on improving
material handling through line balancing and facility re-layout. Using a case study of an automobile company, Dewa
and Chidzuu (2013) demonstrated optimizing a typical manual automobile assembly line using Arena. Their interest
was in improving production outputs through bottleneck management. Analysis was also presented to show the effects
of variables such as vehicle sequencing, batch sizes, and individual vehicle models. Wy et al. (2011) in their study on
material handling for cellular and conveyor assembly considered material handling logistics in assembly lines such as
parts feeding, cart circulation, and kitting of parts. They acknowledged the use of simulation tools such as Arena but
they used the Auto mod software as the simulation language. Wang et al. (2011) studied a general automotive
assembly system and developed a data-driven simulation methodology to model and conduct a what-if analysis of the
system using real-time online data. Arena simulation was used to model the material handling and assembly line
developed. However, they suggested optimization using certain material handling requirements such as driver capacity
as a future improvement to their work. A discrete event simulation of a door production line in an automotive assembly
plant making two types of doors using Arena simulation was studied by Zhao and Li (2015). Though material handling
was not discussed in their model, their interest was in investigating the key performance of production systems and
proposing a control policy for high throughput Soroush et al., (2014) used the Arena software for modelling and
analyzing the product's assembly process to minimize cycle time. They considered only forklifts as material handling
equipment and showed sensitivity analysis with different capacities of the forklifts to investigate operation, material
and waiting time.

© IEOM Society International

172
Proceedings of the First Australian International Conference on Industrial Engineering and Operations
Management, Sydney, Australia, December 20-21, 2022

Thanou and Matopoulos (2021) used the arena simulation to analyze the material flow and specifically the returns in
an automotive plant and to suggest areas of the plant that could result in efficiency gains. Furthermore, they noted
several studies that have considered material handling flow in the literature.

This study specifically contributes to the literature on material handling requirements of automobile plants by
highlighting possible relationships between selected material handling variables such as conveyor speed, length,
transfer resources and outputs from a generic vehicle assembly perspective. Therefore, a general contribution is made
to productivity improvement through the understanding of the impact of resource settings.

3. Model Formulation
3.1 System Description
We present below a basic description of the typical vehicle assembly plant from which the complex configurations in
existence could be obtained. In a typical vehicle assembly plant, different parts are shipped from different sources or
distribution warehouses and are either temporarily stored in an in-plant warehouse or moved into the plant on a just-
in-time basis. Following this is the separation of different parts based on the assembly schedule. Parts scheduled for
bodywork are assembled and enter the body shop. In the body shop, operations such as spot welding and assembly of
parts such as sheet metals are performed (Moon et al., 2021). After the bodywork, checks are conducted to ensure the
right in-process subassemblies from the body shop get into the next process called the paint shop. At the paint shop,
it is ensured that operations such as priming, cementing, sandpapering, chassis priming, and final painting are
conducted on the parts. Quality checks are then conducted on finished work from the paint shop including other parts
needed for final assembly. The finished or final assembled vehicles are sent to the motor pool having passed through
proper interior and exterior inspection.

3.1.1 Assumptions to obtain the ideal vehicle assembly model

To obtain the ideal or basic modelling configuration, the following assumptions were made. These assumptions are
often altered when real-life cases are modelled.
• Defective parts are disposed of immediately and need no rework.
• Buffer sections are not included.
• Material handling is not considered for the moving of parts within the system.
• A termination simulation model with a defined start and stop for a day (one shift) is modelled. (Usually
manufacturing environments run with a continuous shift and on a steady state basis).
Figure 1 below illustrates how parts move from the point of entering the assembly plant to exiting as an assembled
product.

Arrival Parts Quality Sub Quality Body

of parts sorting check Assembly check Shop

Quality Main Paint Quality

check Assembly Shop check

Finished motor
pool and check

Figure 1. Flow chart showing an ideal vehicle assembly process

3.3 Simulation model building

3.3.1 Simulation modelling for the base model
The logic and assumptions used in building the simulation model for the base problem are discussed.

Assumptions during model translation

© IEOM Society International

173
Proceedings of the First Australian International Conference on Industrial Engineering and Operations
Management, Sydney, Australia, December 20-21, 2022

• There could be several numbers of parts ( 𝑛𝑛-parts) referred to as entities required to be assembled and with
several processes and different quality checks. However, a 3-part assembly are utilized for illustration
purposes to be assembled without loss of generality.
• Instantaneous movement of parts in the system with no material handling.
• There are resources such as machines used to process the parts.
• Entities (parts) seize the resource for a time and releases the resource after the time duration.
• No scheduled breaks and failures for resources and resources operate on a fixed capacity.
• Queues are First in First out.

Figure 2 below presents a pictorial representation of the modelling logic used. For the general 𝑛𝑛 parts, the first quality
check sort the parts into the first (1 to 𝑘𝑘) parts, which are scheduled for other quality checks, bodywork and painting
while the other (𝑘𝑘+1 to 𝑛𝑛 ) parts go to the final assembly operation to produce the finished product (vehicle product)
𝑟𝑟 often less than the starting number of 𝑛𝑛 parts.

Sorting
1 to k
parts Process(es)/
n- parts Operation(s)

(k+1) to n
parts
Final operation 𝑟𝑟 < n

Figure 2. The basic logic used in building the model

We further present the actual arena logic modules used and reasons for choice in Table 1 below, while Figure 3
below presents the model logic flow chart in arena symbols.

Table 1. Arena modules selected to develop base simulation logic

Arena Module Reason(s) for selection

Create Arrival of parts into the system
Assign Assign the different part types to be assembled

Decision Sort the parts into the respective section of operations

Process For Quality checks, subassembly operations, and other operations such as
body shop, painting etc.

Batch Group parts based on part types

Dispose To end the simulation by disposing of the parts

© IEOM Society International

174
Proceedings of the First Australian International Conference on Industrial Engineering and Operations
Management, Sydney, Australia, December 20-21, 2022

Create Assign Decision Process Batch

Assemble
Arrival of Check1 Collate part Part 1 and 2 Section 1
Assign parts Split of parts
parts process 1and 2 to obtain Part
4
0
Ent it y. Type==Part 1

El s e
Ent it y. Type==Part 2
0 0 0

Check 2 Body work Section 2

process process

0 0

Check3 Paint shop Section 3

process process

0 0

Check 4 Collate Part

Main
End of Section 4
process 3 and 4
assembly
process
assembly
0
0 0
Dispose

0
Figure 3. Flow chart showing the model logic units for the basic/ideal model

3.3.2 Extensions to the ideal vehicle assembly process

Two model extensions were derived from the ideal model and termed Scenarios 1 and 2. These scenarios are
discussed below:
Scenario 1: Extension with transfer resource constraints. This incorporates a transfer resource which could be a forklift
for transferring in-process inventory between one major section and the other. A resource is stationed between one
section and another and moves forward and backwards between the sections.
Scenario 2: Extension with Conveyor modelling. A non-accumulating conveyor (Kelton et al., 2015), which stops
momentarily at the location of the entity is used as the material handling device for the movement of parts between
sections.

The two scenarios are illustrated in Figure 4 below. In this Figure, scenario 1 is captured with double arrow lines
(forward and backward movement of resource) in blue, while scenario 2 is captured with dashed thick red lines. It is
assumed that there are 𝑥𝑥 sections of the assembly plant consisting of different processes grouped.

Sorting
of parts 1 to k
parts Section(s) Section ( 𝑥𝑥 − 1)
n- parts (1 𝑡𝑡𝑡𝑡 𝑥𝑥 −2)

k+1 to n
parts
Section 𝑥𝑥 final
operation
𝑟𝑟

Transfer resource Conveyor segment

(Scenario 1) (Scenario 2)

Figure 4. The basic logic used in building the model scenario 1 and 2

The logic modules useful for modelling scenarios 1 and 2 and the reason for selection are further presented in Table
2. below. An illustration of the flowchart showing the extended simulation model for scenario 2 using the logic models
of Table 2. is presented in the Appendix.

© IEOM Society International

175
Proceedings of the First Australian International Conference on Industrial Engineering and Operations
Management, Sydney, Australia, December 20-21, 2022

Table 2. Arena modules for extended simulation logic

Arena Module Reason(s) for selection

Create Arrival of parts into the system
Assign Assign the different part types to be assembled

Decision Split the parts into the operations

Station Location of parts
Enter Access to material handling resource, present station of parts and routing
to the next station
Leave release of material handling resource, present station of parts and routing
to the next station
Process For quality checks, subassembly, operations, other operations such as
body shop, painting etc.

Batch Group split parts based on part types

Dispose To end the simulation by disposing of the parts

3.4 Performance measures for analysis

We utilized the number of outputs (final product or vehicle produced) to analyze the ideal and extended scenarios.
Other measures such as the number of parts waiting to be processed at the body shop and paint shop, and utilization
of resources at the Body shop, are also usable but due to the quick opportunity the number of outputs provides in
measuring the performance of the system, the number of outputs was focused on.

3.5 Model Verification and Validation

Verification ensures the simulation model work without errors and functions to the users’ intention. All the models
developed were checked for errors using the check model function, debug bar and animation. Using animation, we
stepped through the model in small spaces to visualize the movement of parts through various logic for the
appropriateness of our intentions.

One of the aims of validation is to check the performance of the developed system when compared to the real system.
Due to the goals of this study which give us the flexibility of either using, not entirely or not using real-life data at all,
validation to compare scenario outputs with real output was not done. However, we ensured the ideal models were
generated from the established operations in the literature, the effect of random data used in the model was taken into
consideration by replicating the experiment a few times and sensitivity analysis spanned across possible real values
of variables for material handling.

4. Data and Experimentation

4.1 Data generation
In this section, we present the type and values of data input for the base and extended problems used for the module
logic within Arena software. To obtain real data for the different sections, observations using time study, reasoning
from similar established processes in the literature, and expert opinions comprise some of the sources of data used.
However due to our objective of this paper and more emphasis on the simulation model formulation, some random
data values were utilized, and some basic assumptions were used to model the scenarios.

For the resources used in every process, we assume the "seize, delay and release" of all parts and have assumed the
triangular distribution as it is often used to model task activities. The actual probability distribution could follow any
of the known distributions such as gamma, uniform, Weibull, and exponential and could be determined with the Arena
input analyzer. The data input used is shown in Tables 3 and 4 below.

4.2 Parameter Fixing for the extended scenarios

© IEOM Society International

176
Proceedings of the First Australian International Conference on Industrial Engineering and Operations
Management, Sydney, Australia, December 20-21, 2022

Our interest was first to ensure a fair comparison between the base and extended scenarios. Since the extended
scenarios were modelled with material handling assumptions, our goal was to search out the parameters that could
still give the same number of outputs as the base model. We essentially wanted to determine the configuration settings
of the material handling of the extended scenarios that can still obtain the same number of outputs as the base scenario.
This guided us in decreasing or increasing resources of only the material handling as we ensured process/activity times
were kept constant for the base model and extended scenarios to achieve a fair comparison of the outputs.

Table 3. Data input type and value for the base model

Arena Module Key input type or Input value(s) Other Assumptions

Probability
distribution
Create: 2 hours Parts arrive in batches
Inter arrival Random (Expo) and slowly into the
Parts per arrival 10 parts/batch assembly process
Assign Discrete (DISC) DISC (0.2, 1, 0.3, Three main parts for
2,1.0,3) assembly
Decide n- way by condition 3- Parts index Parts are split based
on the type
Process Triangular (TRIA) 1-3: (5,6,7) mins Final inspection more
(Check 1 to 4) 4: (8,9,10) mins detailed other
inspections
Batch + Process Parts to group Parts 1 and 2 Parts grouped by
(subassembly ) attribute
Process: All triangular (TRIA) A: (10,12,15) secs all operational
subassembly B: (10,12,15) mins activities within this
Body shop C: (8,9,10) mins section
Paint shop D: (10,12,15) mins
Main assembly
Resource Fixed capacity 1 to 4: No failures and
(Check 1 to 4) Units to seize =1 (2 resources) Schedules
Resource: Fixed capacity A: No failures and
subassembly Units to seize =1 (2 resources) schedules
Body shop (B to D)
Paint shop (4 resources)
Main assembly

For the extended scenarios, the data inputs comprised additional modules including those listed in Table 3 above
were utilized. In Table 4 below, we present the additional modules utilized. Type 1 and 2 refer to the material
handling Arena modules for transfer resources and conveyors respectively. The set of letters and numbers coded in
Table 4 is defined, for example, A11 is coded as (first letter first digit second digit). The first letter represents the
sections of the process (Part release), the first digit represents the type of material handling (transfer resource, and
the second digit represents the logic module used (leave).

Table 4. Data input type and value for the extended model

New Arena Module Key input type or Input Value(s) Other Assumptions
Introduced probability distribution
Stations Parts Name of location -

Leave (type 1): A12, B12 to E12: loading and routing the
A12 parts release Constant delay time Move time= 10mins parts to the next station
Connect type: route Load time=5mins

© IEOM Society International

177
Proceedings of the First Australian International Conference on Industrial Engineering and Operations
Management, Sydney, Australia, December 20-21, 2022

B12: subassembly Station type: Station Transfer resource through a capacitated

section Transfer resource per Capacity = per transportation resource
C12: Body shop section section (Fixed capacity) section unloading time same as
D12: Paint shop section loading time
E12: Assembly section

Enter (type 1): B11 to E11: Move time= 10mins Different transfer resource
B11: subassembly unload time=5mins used between sections
section Unload time
C11: Body-shop section Transfer resource per Unloading time same as
D11: Paint-shop section section loading time
E11: Assembly section

Leave (type 2): A22, B22 to E22: For loading and moving
A22 parts release Access to conveyor the parts to the next station
B22: subassembly Constant delay time load time=5secs using a conveyor.
section Connect type: Conveyor load time estimated during
C22: Body-shop section Station type: Station parameter tunning
D22: Paint-shop section

Enter (type 2): B21 to E21: unload time=5secs For unloading of parts to
B21: subassembly Constant delay time discharge the conveyor.
section Exit from Conveyor Unload time estimated
C21: Body shop section during parameter tunning
D21: Paint shop section
E21: Assembly section

Conveyor One loop conveyor Velocity = Non accumulating

20ft/minute Assuming average length
Cell size=7ft of vehicle = 14 ft
Maximum cell
occupied = 2

Segment Part release to entering Conveyor segment= Segments connect each

of Assembly section 28ft per section section containing the
processes

4.3 Experimentation and Sensitivity Analysis

For this experimentation, we limit the number of replications to five (5). The simulation was conducted based on the
assumption of 1 shift (8 hours per day) for a week (5 days per week). Arena Version 16.1 student edition was used to
perform the simulation.

Sensitivity analyses were based on the extended scenarios to understand the effect of some changes in the material
handling model configuration and parameters used. We specifically were interested in observing how changes in the
parameters of the material handling variables of the extended scenarios can affect the performance of an ideal vehicle
assembly model. For example, how an increase in the speed of the conveyor in scenario 2 affects the outputs produced.
In addition, for scenario 1, it will be interesting to quantitatively experiment with the effect of capacity decrease or
increase of the transfer resource.

5. Results and discussions

5.1 Parameter fixing
The results of the parameter tunning discussed in section 4.2 are presented in table 5 below. The table shows the
material handling transfer resource and conveyor parameters obtained by ensuring the number of outputs is the
same.

© IEOM Society International

178
Proceedings of the First Australian International Conference on Industrial Engineering and Operations
Management, Sydney, Australia, December 20-21, 2022

Table 5. Parameter fixing results for fair model comparison

Performance measure Base model Scenario 1 Scenario 2

Number of outputs in 15 15 15
units (final parts)

Material handling No material handling. Four different transfer Conveyor segment=

resource parameters Flow depends on the resources per section 28ft per station
After tunning instantaneous flow With a fixed capacity Velocity = 20ft/minute
between connecting logic Cell size=7ft
modules Maximum cell
occupied = 2
Assuming average
length of vehicle = 14
feet

5.2 Sensitivity Analysis result

Figure 5 below shows the number of outputs obtained as the segment of the conveyor is increased based on three
different levels of velocity for the conveyor. The centroid was obtained using the popular k-means clustering algorithm
(Awad and Hamad, 2022).

120
Number of outputs( Vehicle)

100
20 ft/min
40 ft/min
80 65 ft/min
Centroid

60
28 42 49 56 70 77 84 98 112
Conveyor length per section (ft)

Figure 5. increasing conveyor segment at three levels of velocity

The results of Figure 5 suggest a downward and upward trend as the conveyor segments increased in length. However,
the centroid shows more of a downward trend before outputs finally increased with the long conveyor segment.
Looking at these results could initially indicate that the longer the conveyor, the longer the time it takes for parts to
travel through the system. However, an increased number of outputs with longer conveyor segments subject to
increasing velocity could indicate more space on a fast-moving conveyor.

Our second interest was to observe the effect of increasing the conveyor velocity at different conveyor segments. It is
assumed that the conveyor segments do not change per section. Figure 6 below illustrates the results obtained.

© IEOM Society International

179
Proceedings of the First Australian International Conference on Industrial Engineering and Operations
Management, Sydney, Australia, December 20-21, 2022

110

Number of outputs( Vehicle)

100

90
28ft
49 ft
80
70 ft

70 Centroid

60
20 25 30 40 50 65 75 85 100
Conveyor velocity(ft/min)

Figure 6. Increasing conveyor velocity at three different lengths of the conveyor segment

Similarly, increasing the conveyor velocity at a constant conveyor length could either result in increased or reduced
outputs. Looking at the centroid line, a gradually increasing number of outputs followed by a gradually decreasing
number of outputs even with increased conveyor velocity is shown. The results follow the logic that the increased
length of the conveyor increases more parts on it and increasing speed also increases the flow of the parts. However,
a high very high-speed conveyor might be very unstable for parts on it thus counteracting the gains of increased output.

Thirdly, our interest is to observe the impact of resource-constrained material handling on the number of outputs. This
was done under a constant move time and assuming a fixed capacity for the resource. Figure 7 below shows the effect
of increasing the transfer resource capacity under constant move time.

90
Number of outputs (Vehicle)

88

86

84

82

80

78
1 2 3 4 5 6 7 8 9
Resource transfer capacity(units)

Figure 7. Resource transfer capacity effect on outputs

Figure 7 above suggests that an increase in transfer resources initially increases outputs which agrees with the
reasoning of speeding operations or increasing flow through capacity increase. Furthermore, results from the table

© IEOM Society International

180
Proceedings of the First Australian International Conference on Industrial Engineering and Operations
Management, Sydney, Australia, December 20-21, 2022

also show the likelihood of outputs becoming insensitive to resource increase. This will be most likely when the
number of parts flowing through the system is much smaller than the transfer resource capacity.

6. Conclusion
In this study, a simulation model of a generic vehicle assembly plant was presented to illustrate the transitions of real
models from ideal models using material handling assumptions. The conveyor and resource-constrained capacity
scenarios were modelled into an ideal vehicle assembly model.

Sensitivity analysis showed that in the short run material handling devices have the likelihood of either increasing,
decreasing, or keeping constant the outputs of a typical operation or process such as vehicle assembly. Therefore,
showing the importance of different configuration settings and trading of variables to yield intended results for
production outputs.

Findings from this study are useful during material handling and design specification to observe the possible
relationships among certain variables of the system being designed. In addition, the discrete event simulation and
modelling logic employed in this study could be a useful aid in engineering education research to study the flow of
students under possible semester and assessment constraints.

This study could be improved with the use of more real-life data. In addition, several assumptions considered could
be removed to see how this will impact the results. For example, some plants operate continuously, and steady-state
modelling might be appropriate. Furthermore, resource schedules and failures could provide more practicality when
considering resource utilization. More performance objectives such as the throughput, utilization of resources, cycle
time, and average number in the queue for each process could also provide more insight into the material handling
process. Lastly, complex robotic features, automated conveyors, automatic storage, and retrieval systems are realities
to consider as the world moves in the fourth industrial revolution direction.

Appendix

B22
B21
Leave
Pa rt 1 a n d 2
Arri v a l o f
En te r
Ch e c k 1 Ba tc h 1 Su b a s s e m bSection
ly 1
Su b a s s e m b l y s ubas s em bly
p a rts As s i g n p a rts Sp l i t o f p a rts p ro c e s s s e c ti o n
s e c ti o n

0
Ent it y . Type==Par t 1 0 0
Ent it y . Type==Par t 2 0 C22
Els e A22 C21
Ch e c k 2 Section 2
En te r Bo d y Bo d y s h o p L e a v e Bo d y
p ro c e s s
Pa rts re l e a s e wo rk s e c ti o n wo rk s e c ti o n
to s ta ti o n 0 0
D21 D22
Pa rts re l e a s e Section 3
s ta ti o n En te r Leave
Ch e c k 3 Pa i n t s h o p
Pa i n ti n g wo rk Pa i n ti n g wo rk
p ro c e s s p ro c e s s
s e c ti o n s e c ti o n

E21 0 0

0
Ro u te to p a rt
3 s ta ti o n En te r
As s e m b l y Di s p o s e o f
s e c ti o n Ch e c k 4 As s e m b l y fi n a l p a rts Ro u te to e x i t M o to r p o o l p a rts
Ba tc h 2
p ro c e s s p ro c e s s s ta ti o n pool ex it

0 0 0
Pa rt 3 a rri v a l
Sta ti o n Section 4

Figure A1 Flow chart showing the model logic units for Conveyor modelling (Scenario 2)

References
Awad, F. H. and Hamad, M. M. Improved k-Means Clustering Algorithm for Big Data Based on Distributed
SmartphoneNeural Engine Processor. Electronics, 11, 883, 2022.

© IEOM Society International

181
Proceedings of the First Australian International Conference on Industrial Engineering and Operations
Management, Sydney, Australia, December 20-21, 2022

Dengiz, B., İç, Y. T. and Belgin, O. A meta-model based simulation optimization using hybrid simulation-analytical
modeling to increase the productivity in automotive industry. Mathematics and Computers in Simulation,
120, 120-128, 2016.
Dewa, M. and Chidzuu, L. Managing bottlenecks in manual automobile assembly systems using discrete event
simulation: case study. South African Journal of Industrial Engineering, 24, 155-166, 2013.
Gorham, R. Motorization Management for Development, 2022.
Guiguet, A. and Pons, D. A Framework for Interactive Development of Simulation Models with Strategical–Tactical–
Operational Layering Applied to the Logistics of Bulk Commodities. Modelling, 3, 272-299, 2022.
Ikome, J. M., Laseinde, O. T. and Katumba, M. G. K. The Future of the Automotive Manufacturing Industry in
Developing Nations: A Case Study of its Sustainability Based on South Africa’s Paradigm. Procedia
Computer Science, 200, 1165-1173, 2022.
Kelton, W. D., Sadowski, R. P. and Zupick, N. B. Simulation with Arena, 2015.
Moon, D. H., Kim, D. O. and Shin, Y. W. Automotive Body Shop Design Problems Using Meta-Models Considering
Product-Mix Change and Reconfiguration Strategy. Applied Sciences, 11, 2748, 2021.
Oumer, A., Atnaw, S. M., Cheng, J. K. and Singh, L. Improving energy efficiency for the vehicle assembly industry:
A discrete event simulation approach. IOP Conference Series: Materials Science and Engineering, IOP
Publishing, 012069, 2016.
Oyewole, G. Heuristic To Solve The Fixed Charge Solid Location And Transportation Problem With Truck-Load
Constraint. SAIIE31 Proceedings 5th – 7th October,2020 Muldersdrift, South Africa.
Oyewole, G. J. and Adetunji, O. Solving the facility location and fixed charge solid transportation problem. Journal
of Industrial & Management Optimization, 13, 0.,2020.
Soroush, H., Sajjadi, S. M. and Arabzad, S. M. Efficiency analysis and optimisation of a multi-product assembly line
using simulation. International Journal of Productivity and Quality Management, 13, 89-104, 2014.
Thanou, E. and Matopoulos, A. Improving efficiency of material flows in an automotive assembly plant: A case study.
CIRP Journal of Manufacturing Science and Technology, 35, 959-967, 2021.
Wang, J., Chang, Q., Xiao, G., Wang, N. and Li, S. Data driven production modeling and simulation of complex
automobile general assembly plant. Computers in industry, 62, 765-775, 2011.
Wilson, R., Mercier, P. H. and Navarra, A. Integrated Artificial Neural Network and Discrete Event Simulation
Framework for Regional Development of Refractory Gold Systems. Mining, 2, 123-154, 2022.
Wirabhuana, A., Haron, H. and Imtihan, M. R. Simulation and re-engineering of truck assembly line. 2008 Second
Asia International Conference on Modelling & Simulation (AMS), IEEE, 783-787, 2008.
Wy, J., Jeong, S., Kim, B.-I., Park, J., Shin, J., Yoon, H. and Lee, S. A data-driven generic simulation model for
logistics-embedded assembly manufacturing lines. Computers & Industrial Engineering, 60, 138-147, 2011.
Zhao, C. and Li, J. Production control of door manufacturing line at an automotive assembly Plant. IFAC-
PapersOnLine, 48, 1652-1657, 2015.

Biographies
Gbeminiyi John Oyewole: lectures at the Department of Industrial Engineering, Operations Management and
Mechanical Engineering, Vaal University of Technology, South Africa. He obtained his doctorate In Industrial
Engineering from the Industrial and Systems Engineering Department University of Pretoria, South Africa.
Furthermore, his post-doctorate research was done in the Engineering and Technology management department at the
University of Pretoria. He has published journal and conference papers. He has worked in the field of Logistics, and
transportation and as a supply chain officer at the Candel Agrochemical manufacturing Plant in Lagos Nigeria. His
research interests are in Facility location problems, Modelling and simulation, data analytics, applied optimization,
operations research, supply chain and engineering designs.

Khitleli Tsosang Daniel: He is currently working in a supply chain logistics company in Kwazulu Natal South Africa.
He started his higher education in 2014 at the Vaal University of Technology in Vanderbijlpark; where he obtained
his National Diploma in Industrial Engineering in 2017, Advanced Diploma in Industrial Engineering in 2020, Post
Graduate Diploma in 2021, and he is a Master of Engineering candidate in Vaal University of Technology. He did his
Work Integrated Learning at BMW Plant Rosslyn, and furthered his experimental learning at BMW Spartanburg in
South Carolina. He has worked on several modelling and continuous improvement projects while at the Vaal
University of Technology.

© IEOM Society International

182