See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/354499701

Unity3D-based Simulation for Operations Management Teaching

Conference Paper · January 2021

DOI: 10.46354/i3m.2021.mas.017

CITATIONS READS
3 169

2 authors, including:

Javier Pernas
Universidade da Coruña
16 PUBLICATIONS 11 CITATIONS

SEE PROFILE

All content following this page was uploaded by Javier Pernas on 28 November 2022.

The user has requested enhancement of the downloaded file.

 20th International Conference on Modelling and Applied Simulation
18th International Multidisciplinary Modeling & Simulation Multiconference

ISSN 2724-0037 ISBN 978-88-85741-61-4 © 2021 The Authors.

doi: 10.46354/i3m.2021.mas.017

Unity3D-based Simulation for Operations Management

Teaching
Javier Pernas-Álvarez1,* and Diego Crespo-Pereira1
1
University of A Coruña, Campus de Esteiro s/n, Ferrol, 15403, Spain
*Corresponding author. Email address: [email protected]

Abstract
The use of digital applications as support tools for teaching has opened a new world of possibilities when it comes to applying
and consolidating theoretical concepts. Within the field of operations management, game-based learning is still nowadays an
innovative approach pending further exploitation. Digitalisation and advances in 3D graphics remove constraints inherent to real
world by immersing user in fictional yet realistic simulations. In this paper we present a Unity3D-based 3D simulation of a
manufacturing process to be used as case study for the application of operations management concepts and modelling and
simulation techniques. Acting as operations manager, the user is responsible for handling material supply, manufacturing orders
and product delivery to ultimately increase total profit. As proposed teaching activity, the user must analyse the process by means
of a simulation model through which optimise the decision strategy. Finally, she must prove the validity of the devised strategy
by implementing it in a final gameplay.

Keywords: discrete-event simulation; education: optimisation; serious game; operations management

that we, as modellers, must overcome. The only

1. Introduction approach into this challenge is the firm acquisition of
good modelling and simulation (M&S) skills along with
Within operations management field, discrete-event solid operations management (OM) concepts.
simulation (DES) is long known for its unquestionably
advantages over other types of simulation techniques In parallel to this, several teaching techniques have
when it comes to model a complex system and been used for teaching OM and understanding the
understanding the specifics for further optimisation suitability of M&S as to the target case. From the
(Maidstone, 2012; Robinson, 2005). Recent classical lectures and test-based assessment, to
advancements of commercial packages respond to a project-based and assignment-based learning, and
growing use of DES in the industry as a decision- lastly to game-based learning or teaching based on
support and planning tool. This in turn is due to the serious games (Delago et al., 2016; Zhao et al., 2019).
current enhancement of complementary technologies Although the discussion on the most suitable approach
ranging for Internet of Things (IoT), Big Data to Cloud for theoretical knowledge is on the table (Girard et al.,
Computing, all of them encompassed by the so-called 2013), what is quite clear is that a combination of them
Industry 4.0 (Liao et al., 2017; Rekapalli et al., 2016). As provides a more comprehensive training. In other
one of its pillars, DES has become the key to the digital words, as stated in (Costantino et al., 2012), game-
twin paradigm when applied to manufacturing or based teaching provide the necessary link between OM
warehousing systems. Nevertheless, an effective and notions and their use and impact on the reality,
comprehensive DES model brings in several hurdles something oftentimes lacked by traditional methods.

© 2021 The Authors. This article is an open access article distributed under the terms and conditions of the Creative
Commons Attribution (CC BY-NC-ND) license (https://creativecommons.org/licenses/by-nc-nd/4.0/).

136
 Pernas-Álvarez and Crespo-Pereira | 137

If the game is to be digital, as the case presented manufacturing, and delivering decisions to face
here, it must be noted the infinite possibilities demand and process variability, being her performance
educational games prompt regardless of real-world measured by means of a profit function. Therefore, user
constraints. If we wanted to devise a traditional game must use M&S to study the system an adopt a proper
(often table-top, card or role playing) which simulates strategy that optimises the total profit.
the actual system, we would be conditioned by the
The overall intention of the game is to provide
impossibility of students for visiting or controlling the
students with a realistic simulation that acts as the real
real system. In that case, and depending on our aim, we
system to be optimised. Thereby, hard skills such as OM
could even argue that the game loses its sense, as
concepts, M&S techniques and experimentation mixed
players may not feel sufficiently engaged. However,
with soft skills like teamwork, problem-solving and
simulation-based digital games can replicate the real
critical thinking are trained all together.
system on a computer and immerse the user by
inducing her to think that the world she has in front of
is a faithful copy of the real system (Pourabdollahian et 2. 3D Simulation Game
al., 2012). When it comes to applying and consolidating
OM concepts, the user can perfectly see their 2.1. General Description
importance and effects since she is allowed to make
The 3D manufacturing game presented here is a DES
changes and take decisions otherwise unthinkable onto
model developed in the game engine Unity3D. This
the real twin.
model is devoted to representing the real
In this sense, varied examples can be found in the manufacturing plant where the user performs the role
literature regarding the use of simulations and games of the operations manager and thus must take a set of
as a teaching-support tool. (Aqlan & Walters, 2017) predefined decisions. User can switch between a fixed
present a series of hands-on experiments devoted to third-person perspective that facilitates an overall
teaching lean concepts to undergraduate students in view of the process (Figure 1). and a first-person
collaboration with local industries. Despite the absence perspective (FPP) by means of which she can move
of digitalisation, it is an interesting example of how around the manufacturing shop (Figure 2). Although it
games can be used to reinforce OM basics such as batch cannot interfere with the process physics, the FPP
and queue production and lean flow production. yields a more immersive experience.
(Delago et al., 2016) initially developed a lean-board
game which afterwards was levelled-up by means of
commercial M&S package FlexSim. Out of this model
they define simulation cases that gradually grow in
complexity as to the user skills. They remark how the
digitalisation of the game make it gain realism and
upgrade the potential to teach lean manufacturing
concepts.
(Zhao et al., 2019) focus on craft production to
deliver a virtual reality-based game where students
must design and produce LEGO cars attending to
costumer’s requirements at the same time they
optimise total production cost. They conduct a pilot
Figure 1. Overall view of the model and initial menu.
study to prove the efficiency of the game for improving
student’s problem-solving skills. It represents another
interesting instance of the use of games in OM
education.
Finally, and closer to our case, (Costantino et al.,
2012) and (Crespo-Pereira et al., 2014) present both
digital simulation-based games where users must deal
with decisions relative to system’s functioning
strategy. In the first, user must select capacity and
maintenance strategy of different resources whereas in
(Crespo-Pereira et al., 2014) the user must simulate the
system in question to formulate solutions that optimise
the score function.
In this line, we present a simulation-based game Figure 2. First-person perspective.
consisted of a 3D Unity-based simulation of a
manufacturing process where user plays the role of
operations manager. She must take smart supplying,
 138 | 20th International Conference on Modelling and Applied Simulation, MAS 2021

2.2. Rules
Provider A Provider B
Once the game is initialised, the user must decide as
operations manager the capacity of inventory buffers,
the processing capacity of the assembler, the
Queue Queue processing capacity of all the workstations, and the size
A B
of the stock buffer. These decisions have a direct effect
on the initial cost, so that at the beginning the profit
Conveyor Conveyor
will be always negative (Figure 2). The game also asks
for user’s name which is essential for the game to start.
Assembler In this way, we can track user identity and compare her
performance to that of the rest of students.

Conveyor Afterwards, it is user’s turn to decide when to

request to manufacturers A and B and how many items
are to be ordered. Although she can see on her own how
Processor 1
many raw items are already stored, a control menu
(Figure 2) displays the exact number to provide more
Conveyor
accurate supervision.
The second decision revolves around when and how
Processor 1 many products must be manufactured. The menu
displays both fabrication pending and customer
pending products. The first figure increases when user
Conveyor
sends fabrication orders, but parts are not yet available
or cannot enter the process. Once parts are in process,
Processor 1
the figure decreases. As for the second, it increases
every time customer demands a new product, and it is
reduced when user ships a truck.
Conveyor Conveyor Customer
Stock
Buffer This latter is precisely the last decision user must
take. Every shipment has an associated cost and
Figure 3. Process Diagram. revenue based on the product price, and same is applied
to unsatisfied customers along with final stock buffer.
Therefore, user must study trade-offs between the
The simulation model implemented in Unity is delivery cost and the opportunity cost plus the buffer
depicted in Figure 3. It consists of a set of workstations cost of not shipping in time.
connected by conveyors, two inventory buffers and
final stock buffer. Conceptual game-objects represent Maximum simulation time is set so profit function is
parts to be manufactured. The first workstation is an always assessed over the same time span. Simulation
assembly station that joins together items from two time can be previously readjusted to the available
providers (provider A and B). The rest of workstations session duration by the activity supervisor.
represent single activities with the same processing
capacity. 2.3. Goal

As for activity times, different distributions are used The goal of the student is to maximise the score
as shown in Table 1. Customer demand and supply function which is modelled by means of a profit
times are modelled as random processes by means of function. This function considers variable costs (VC) a
uniform distributions with time means set differently function of inventory units (BA, BB) and stock units (BS),
to each process. Customer demand can be ordered (O) and purchased (P) units, pending orders
automatically generated or read from a file. Assembling (PO) and manufacturing orders (MO) and shipments
and processing activities are modelled as Poisson (S). Inventory and pending order costs are per unit of
process with exponential distribution of task times. time. Fixed costs (FC) (2) involve initial inventories
Lastly, conveyor times are constant and set equal to all capacities (BCA, BCB) and stock buffer capacity (BCS),
conveyors. assembler capacity (AC), and workstations processing
capacity (WSC). The only revenue (REV) comes from
Table 1. Model Elements and processes’ time models.
the final product sale price (P) and the number of
Element Time Modelling shipped units (3). Costs and sale price have been
Provider Uniform Distribution balanced to determine the difficulty of the game as
Assembler Poisson Process moderate. Profit function is represented in (4).
Processor Poisson Process
Conveyor Constant 𝑉𝐶 (€) = (𝐵! + 𝐵" + 𝐵# ) ∗ 𝐶"$% + 𝑂 ∗ 𝐶&' + 𝑃 ∗ 𝐶($' (1)
Customer Uniform Distribution + 𝑃𝑂 ∗ 𝐶(& + 𝑀𝑂 ∗ 𝐶)& + 𝑆 ∗ 𝐶#*+(
Where CBUF stands for cost per buffered unit, COR for
 Pernas-Álvarez and Crespo-Pereira | 139

cost per ordered unit, CPUR for cost per purchased unit,
CPO for cost per pending order, CMO for cost per
manufacturing order and CSHIP for cost per shipped unit.
𝐹𝐶 (€) = (𝐵𝐶! + 𝐵𝐶" + 𝐵𝐶# ) ∗ 𝐶𝐶"$% + 𝐴𝐶 ∗ 𝐶𝐶!## (2)
+ 𝑊𝑆𝐶 ∗ 3 ∗ 𝐶𝐶,#
Where CCBUF stands for cost per capacity unit of
buffer, CCASS for cost per unit of processing capacity of
the assembler, and CCWS per unit of processing capacity
of the workstations.
𝑅𝐸𝑉 (€) = 𝑃 ∗ 𝑄 (3)
Figure 4. Proposed simulation model developed in FlexSim.
𝑃𝑅𝑂𝐹𝐼𝑇 (€) = 𝑅𝐸𝑉 − 𝐹𝐶 − 𝑉𝐶 (4)
Finally, users must check their strategy in the
2.4. Audience simulation model and perform a final gameplay. Result
The target audience of the game is composed of is registered in an encrypted file whose encoding is only
known by the professor. In this way, assessment is
undergraduate and graduate students. The game has
already been used as an education tool within the performed in a way that prevents data tampering.
course “Simulation and Optimization” of the 3rd
academic year of the Bachelor’s Degree in Industrial Start
technology at University of A Coruña. The results have
been satisfactory, having shown the students higher
engagement. They claimed they saw the game as an aid
to deal with simulation project from scratch and Session 1
transferred results to real decisions. • Game Presentation
• First Gameplay
• Rules explanation: Profit Function
3. Proposed Teaching Program and Maximum Simulation Time
The development of the teaching activity must be at
least distributed between four different sessions as
illustrated in Figure 5. Session 2
In the first session, users are presented the • Conceptual Model Development
simulation game which is provided as a standalone • Record task times and obtain
distributions
application. Time is given to them for practicing and
becoming familiar with the simulation. Afterwards,
users are explained the basic rules and the evaluation of
the profit function to maximise. They are also told the Session 3
breakdown of the different costs involved and the sale
• Simulation model development by
price. means of commercial software
During the second session, users must develop a • Validation and verification of the
model
conceptual model of the shop and identify key • Definition of decision strategy
parameters to measure. They must run the simulation
iteratively and record the different processing times,
including conveyors’ one. They must calculate the
Session 4
minimum sample size for an adequate confidence
interval and obtain approximate distribution functions • Final gameplay based on planned
adjusted to recorded sample times. decision strategy

In the third session, users must carry out a

simulation model with a commercial package. An
example developed with FlexSim simulation software
is provided in Figure 4. Having finished this step users End
must study the sensitivity of the profit function as to Figure 5. Proposed schedule for the teaching activity.
the different parameters and undertake experiments to
define an optimised decision strategy.
4. Conclusions
In this paper we present a simulation-based 3D game
for operations management teaching and propose a set
of guiding sessions to undertake the activity. The game
was developed by means of Unity3D and in it, assuming
 140 | 20th International Conference on Modelling and Applied Simulation, MAS 2021

the role of operations manager, the user must take a Maidstone, R. (2012). Discrete Event Simulation,
series of decisions that have diverse impacts on a profit System Dynamics and Agent Based Simulation:
function, which is to be maximised. Thus, the user must Discussion and Comparison. In researchgate.net.
identify key factors that fully describe the behaviour of https://www.researchgate.net/publication/30615
the system and develop a simulation model in a 7680
commercial M&S software. Having performed this step,
she must extract a suitable decision strategy that Pourabdollahian, B., Taisch, M., & Kerga, E. (2012).
maximises the profit function. Finally, the user must Serious games in manufacturing education:
prove the validity of the strategy through a complete Evaluation of. Procedia Computer Science, 15,
run of the game. 256–265.
https://doi.org/10.1016/j.procs.2012.10.077
As main insights out of this development, we aim at
providing a teaching case that demonstrates once more Rekapalli, P. V., Martinez, J. C., Back, M., Kimber, D.,
the validity of this tools as methodology for teaching Rieffel, E., Dunnigan, A., Liew, B., Gattepally, S.,
OM concepts together with M&S skills. We also Foote, J., Shingu, J., Vaughan, J., Kusiak, A.,
emphasise the potential of game engines as a Wang, S., Wan, J., Li, D., Zhang, C., Qi, Q., Tao, F.,
framework to develop simulation-based games, since Robinson, S., … Brooks, J. R. (2016). Adding
they greatly facilitate the development phase and
immersive virtual reality to a science lab
widen future research paths with, for example, further
simulation causes more presence but less
integration with virtual reality.
learning. Proceedings - Winter Simulation
Conference, 56(1), 40–63.
References https://doi.org/10.1061/(asce)co.1943-
Aqlan, F., & Walters, E. G. (2017). Teaching Lean 7862.0000270
principles through simulation games. ASEE Robinson, S. (2005). Discrete-event simulation: From
Annual Conference and Exposition, Conference the pioneers to the present, what next? Journal of
Proceedings, 2017-June. the Operational Research Society, 56(6), 619–629.
https://doi.org/10.18260/1-2--28921 https://doi.org/10.1057/palgrave.jors.2601864
Costantino, F., Di Gravio, G., Shaban, A., & Tronci, M. Zhao, R., Aqlan, F., Elliott, L. J., & Lum, H. C. (2019).
(2012). A simulation based game approach for Developing a virtual reality game for
teaching operations management topics. manufacturing education. ACM International
Proceedings - Winter Simulation Conference. Conference Proceeding Series, 3–6.
https://doi.org/10.1109/WSC.2012.6465028 https://doi.org/10.1145/3337722.3341831
Crespo-Pereira, D., Lamas-Rodriguez, A., & Rios-
Prado, R. (2014). A modified beer game for
simulation and optimization teaching. 26th
European Modeling and Simulation Symposium,
EMSS 2014, c, 498–503.
Delago, L. C., Machado, M. E. F. H. S., De Brito, F. O.,
Landgraf, G. C., De Andrade Schroeder, M., &
Torezzan, C. (2016). Learning lean philosophy
through 3D game-based simulation. Proceedings
- Winter Simulation Conference, 3385–3392.
https://doi.org/10.1109/WSC.2016.7822369
Girard, C., Ecalle, J., & Magnan, A. (2013). Serious
games as new educational tools: How effective
are they? A meta-analysis of recent studies.
Journal of Computer Assisted Learning, 29(3), 207–
219. https://doi.org/10.1111/j.1365-
2729.2012.00489.x
Liao, Y., Deschamps, F., Loures, E. de F. R., & Ramos, L.
F. P. (2017). Past, present and future of Industry
4.0 - a systematic literature review and research
agenda proposal. International Journal of
Production Research, 55(12), 3609–3629.
https://doi.org/10.1080/00207543.2017.1308576

View publication stats