King Saud University

College of Computer and Information Sciences

Information Systems Department

IS 481: Business Process Management

1st Semester 1445 H
Course Project

Quantitative Process Analysis:

Process Simulation

Prepared by:

SN Student Name ID
#1 Alhanouf Alsagir 442201424
#2 Nouf Alsaleh 442201372
#3 Aljohara Alharbi 442202612
#4 Haya Alsultan 441204325

Supervised by:
Dr. Ashraf Youssef
 Abstract
The group's report focuses on the importance of using business process simulation as a strong
tool for studying and improving how organizations work. This report gives an overview of
business process simulation and discusses its benefits, along with looking at earlier studies
and how it's used in real life.

The report is split into four main parts: simulation anatomy, simulation input, simulation
output, and simulation tools. Each part talks about these topics and includes theories,
methods, and best practices that are important.

In the simulation anatomy part, the report looks at the main parts of business process
simulation, like the process model, simulation engine, and parameters. In the simulation input
part, it talks about the different types of important information needed for business process
simulation, like diagrams, performance numbers, and resource use data.

The simulation output part looks at all the different results that come from using business
process simulation, such as performance numbers, key points, and suggestions for making
things better. The simulation tools part talks about the software used to make models, find
problems, test different situations, and make decisions using data. It also talks about problems
and difficulties related to these tools.

Overall, business process simulation is seen as a really important way for organizations to
understand and improve how they work. Learning from simulation results helps make better
decisions, improves how well things work, saves money, and boosts how well the
organization does.

II
 Acknowledgements

We begin by acknowledging and expressing our gratitude to Allah for granting us the
capability to diligently pursue our research study.

Additionally, we extend our sincere appreciation and thanks to Dr. Ashraf Youssef for
generously sharing his extensive knowledge, invaluable assistance, and guidance throughout
the process, enabling us to successfully complete this research.

III
 TABLE OF CONTENTS
ABSTRACT ................................................................................................................................................................................II
ACKNOWLEDGEMENTS ...................................................................................................................................................... III
TABLE OF CONTENTS ......................................................................................................................................................... IV
LIST OF FIGURES ................................................................................................................................................................... V
LIST OF TABLES .................................................................................................................................................................... VI
1. INTRODUCTION .............................................................................................................................................................. 1
2. LITERATURE REVIEW ................................................................................................................................................... 2
3. PROCESS SIMULATION ................................................................................................................................................. 3
3.1 PROCESS SIMULATION ANATOMY .............................................................................................................................. 3
3.2 SIMULATOR’S LOGIC .................................................................................................................................................. 4
3.3 PROCESS SIMULATION INPUT ..................................................................................................................................... 5
3.4 PROCESS SIMULATION OUTPUT.................................................................................................................................. 6
4. SIMULATION TOOLS ..................................................................................................................................................... 8
5. CONCLUSION ................................................................................................................................................................ 10
REFERENCES.......................................................................................................................................................................... 11

iv
 LIST OF FIGURES

Figure 1. Simulator’s Logic and Activities .............................................................................................................. 4

Figure 2. Process Simulation Input and Output........................................................................................................ 7

v
 LIST OF TABLES

Table 1. Process Simulation Techniques. ................................................................................................................ 8

vi
 1. Introduction
In business process management, checking a process’s performance is crucial during process
monitoring. Process performance measures are utilized to assess the effectiveness of a process,
considering four crucial factors: time, cost, quality, and flexibility. These measures empower
process owners to make informed decisions about whether immediate enhancements are needed or
not. Nevertheless, ensuring the success of a proposed process model during the process analysis
phase is extremely challenging. This is due to the fact that certain business processes may perform
exceptionally well under normal conditions but struggle when faced with unexpected situations.
However, leveraging process simulation allows decision makers to select the most suitable process
model, by creating fictional scenarios and providing them with insightful measures of the
fictionalized scenarios, based on three important success criteria which are:
1. Cycle times.
2. Average waiting times.
3. Average resource utilization.
The aforementioned criteria allow the process owner to conduct a fair evaluation of the process's
performance and determine the most suitable approach. Moreover, this enables decision makers to
make well-informed choices by relying on quantitative statistics. [1]

1
2. Literature Review

The utilization of quantitative process analysis, specifically through process simulation, is a

valuable approach for understanding and improving complex business processes. Process
simulation involves the construction of models that simulate the behavior and performance of
business processes. It allows organizations to create hypothetical instances of a process, simulate
their execution step-by-step, and capture important data such as cycle times, resource utilization,
and waiting times [1]. By employing process simulation, organizations can gain insights into the
factors influencing process performance and evaluate different scenarios to identify areas for
improvement.

One of the primary benefits of business process simulation is its ability to provide insights into the
impact of various factors on process performance. Simulation models enable the testing of different
scenarios, allowing organizations to identify areas for enhancement. For example, simulation can
be used to assess the effects of changes in staffing levels or resource allocation on process
performance [2]. Another advantage is the ability to make informed decisions regarding resource
allocation and project prioritization, as simulation provides insights into how different scenarios
might unfold. Various techniques for simulating business processes, such as system dynamics
modeling, agent-based modeling, and discrete-event simulation, are available, each with its own
strengths and weaknesses [3].

Ongoing research focuses on developing simulation techniques capable of handling more complex
processes. Agent-based modeling, for instance, simulates independent agents within a system rather
than treating them as homogeneous entities, potentially enhancing the modeling of complex
processes [2].

Numerous studies demonstrate the effectiveness of business process simulation in improving

processes. For example, simulation can optimize staffing levels in call centers while maintaining
low waiting times. By determining the optimal number of agents required to handle incoming calls
and minimize average waiting time, organizations can improve efficiency [1]. Simulation is also
valuable in analyzing the impact of process changes before implementation. By testing different
scenarios, organizations can identify the most effective approach in terms of cost reduction or
performance improvement [1].

Simulation has been applied in specific industries, including healthcare, manufacturing, and
logistics. In healthcare, simulation optimizes patient flow through hospitals and emergency
departments. In manufacturing, it enhances production line efficiency and reduces downtime. In
logistics, it improves supply chain operations [2].

Despite the benefits, challenges exist in the use of business process simulation. Accurate data
inputs are crucial for reliable simulations, requiring precise information about an organization's
processes and operations. Additionally, building accurate models that reflect all aspects of an
organization's operations can be complex [3].

2
3. Process Simulation

3.1 Process Simulation Anatomy

Simulation is a methodology of conducting experimental simulation, studying potential improvements,

and validating the proposed enhancements. Although that simulation is relatively new, it is
considered to be one of the standard analysis techniques and the most essential skill that a process
analyst must possess. A simulator can generate an infinite number of possible experimental scenarios
with changing environmental settings using intelligent processing algorithms. The main purpose of
simulation is only to search for possible improvements and validate them, other than that, leveraging
such a feature won’t be as beneficial as we might believe [3][4].
In order to gain a better understanding, we must shed light on the urging motives of using simulators.
The motives as listed below [5]:

1- Training and Skill Development:

Safety Training: Simulators are widely used in industries such as aviation, healthcare, and
manufacturing to train personnel in a safe environment without real-world risks.
Skill Enhancement: Provides a controlled and repeatable environment for individuals to practice and
enhance their skills, whether it's flying an aircraft, performing surgery, or operating complex
machinery.

2- Cost Savings:
Reduced Resource Costs: Simulators allow users to practice without the need for expensive
resources, fuel, or materials. This can significantly reduce costs associated with traditional training
methods.
Avoidance of Real-World Risks: In industries like aviation and military, mistakes can be costly both
in terms of human lives and equipment. Simulators offer a way to train without the risk of real-world
consequences.

3- Research and Development:

Prototyping and Testing: Simulators are used to test and refine prototypes in a virtual environment
before investing resources in physical prototypes. This is common in industries such as automotive
and aerospace.
Scenario Analysis: Researchers and engineers use simulators to model and analyze various scenarios
to understand potential outcomes and optimize designs or systems.
The motives mentioned above, indicate the immediate need for employing simulators in the process
analysis phase. Yet, there are critical issues related to unrealistic human forces modeling during process
simulation. The following list illustrates the issues in-depth [2]:
• Safety Concerns: Incorrect modeling of human forces can lead to inaccurate predictions of
ergonomic issues, potentially resulting in safety hazards for workers. For example, if the
simulation does not accurately represent the forces required for a task, it may underestimate
the risk of musculoskeletal disorders.
• Inefficient Workflows: Unrealistic human force modeling may result in inefficient workflows.
If the simulated forces do not match the actual forces exerted by workers, the simulation may
suggest suboptimal layouts, equipment placements, or process designs.
• Productivity and Performance Impact: Incorrect force modeling can impact worker
performance and productivity predictions. If the simulated forces do not align with the actual
physical demands of a task, the simulation may fail to identify potential bottlenecks, delays, or
inefficiencies in the process.
• Complexity of modeling business priorities: the prioritization of processes is critical, in order
3
 to focus more on the important ones sequentially, this concern arose from dealing with multiple
processes at the same time.
• Processes change based on context: simulators accommodate processes in a steady state only.
This isn’t true, because processes are continuously changed and affected by external factors
which implies the instability of processes on long term

3.2 Simulator’s Logic

To tell apart simulation techniques from tools, it's important to understand how simulators work using
the logic in the model. Now, let's explain the basic idea of how the simulation model works. It's
important to note that all processes can be repeatedly executed until a specific goal is achieved. As
depicted in Figure no. 1 below, the model's logic encompasses eight activities connected in sequence
[4].

Figure 1. Simulator’s Logic and Activities

1. Modeling Objectives Definition: is the activity in which the business process practitioners
determine a set of desired outputs and information that must be extracted from the model.
2. Modeling Scope Definition: is the activity of identifying the business process scope and its
boundaries. The scope of a process should be based on:
a. The essentiality of the process, and whether it is a high priority or not.
b. The inefficiency of a process which requires redesign.
c. The proportionality of a process to be simulated.
d. The captured resources. Such as, time, specialized workforce, and technical equipment.
3. Data collection and Analysis: this activity concerns collecting related data of the pre-specified
process and analyzing it through statistical measures.
4. Model Development: is the activity of which the simulation model is iteratively and
incrementally constructed using a simulation software suite.
5. Model Testing: is the activity of defect testing and validating the model outputs. This activity
could be intersected with the earlier activity “Model Development”, conducting tests during
model implementation.
6. Model Experimentation: is the activity of the model that generates a wide range of alternatives
of the given process. Note that the random error rates must be low, in order to produce
reasonable improvements based on valid assumptions.
7. Output Analysis: is the activity of analyzing the results captured from the previous activity
“Model Experimentation”. A set of statistical measurements must be calculated, to conduct a
comparison between the generated alternatives.
8. Feedback: this activity concerns observing the calculated outputs, and defining a set of
recommended actions, regardless of process change or enhancement

4
 3.3 Process Simulation Input

Creating a process model begins with mapping the business process. Most process modeling tools
provide a drag-and-drop kind of GUI with process documentation. Then you discover processes
where subprocesses and activities are identified. Description of passing entities process and link
process using connectors facilitate the definition of workflow. Finally, you define the resources and
assign them to the activities where they are used. Before simulating a model, you must choose
interesting performance measures. For example, you can care about throughput and cycle time
reporting for entities, operational costs for processes, and use reports for resources. When you run a
simulation, the process simulation tool will automatically check the model and start the simulation
clock progress. During the simulation you will see an animation of flow that helps you visualize the
process in motion. You can also ask Business Process Simulator (BPS) to generate charts in real-
time, allowing you to visualize key performance measures, in the process simulation [6].

From the above description of how a simulator works, we can see that the following information that
must be specified for each task in the process model to simulate it:
• Probability distribution for the processing time of each task.
• Other performance attributes for the task such as cost and added value created by the task.
• The resource group is responsible for performing the task.
For example, in the loan application process, there are three groups of resources: Claims Processors,
Clerks, and Managers. For each resource group, we need to specify its size (e.g., number of claims
handlers or the number of employees) and optionally their cost per unit of time (e.g.: hourly cost of a
complaint handler). If we specify the cost per unit time for each resource pool, the simulation calculates
the average labor cost for each instance in addition to calculating the cycle time and waiting time.
Common probability distributions for task duration in the context of process simulation include:
• Fixed: This is the case where the task processing time is the same for all executions of this
task.
• Exponential distribution: can be applied when task processing time is usually around a certain
average value, but sometimes significantly longer.
• Normal distribution: this distribution is used when the task's processing time is around a certain
mean and the deviation around this value is symmetric, meaning the actual processing time
may be higher or lower than the average. average with the same probability.
To execute the simulation effectively, the analyst needs to provide additional definitions. For task
duration, when opting for an exponential distribution, specifying the mean is sufficient. Conversely,
when utilizing a normal distribution, both the mean and standard deviation must be defined. These
values are typically derived from informed estimations, often gathered through interviews with
involved parties. Additionally, for each thread departing from a decision gate, the analyst must assign
a branch probability. These probabilities can be determined through stakeholder interviews, process
observations, or the analysis of relevant information system logs. Finally, to initiate the simulation,
the analyst must further articulate the following:
• Mean time between arrivals and its associated probability distribution. A very common time
distribution between arrivals is the exponential distribution, and this is often the default
distribution supported by business process simulators. However, the time between arrivals may
follow a different distribution, such as a normal distribution. By providing a sample of arrival
times over a given period to a statistical tool, we can determine which distribution best fits the
data. Some simulators provide a module to select the distribution of time between arrivals and
to calculate the average time between arrivals from a data sample.

5
 • The simulation starts date and time (e.g., "8:00 am 11/11/2017"). One of the following:
a. The simulation's end date and time. If this option is selected, the simulation will stop
generating more process instances after the simulation clock to the end time.
b. Real-time duration of the simulation (e.g., 7 days, 14 days). In this way, the end time of
the simulation can be inferred by adding this interval to the start time.
c. The number of process instances needed to simulate (e.g., 1000). If this option is selected,
the emulator will spawn process instances based on the incoming rate until the required
number of process instances is reached. At this point, the simulation stops. Some
emulators will not stop immediately but will allow active process instances to terminate
before stopping the simulation [1].
Upon completion of the simulation, presenting and analyzing the performance measures of care is
essential. Effectively drawing valuable and precise conclusions from simulation results involves
comprehensive statistical analysis of both input and output data. Addressing issues such as distribution
is crucial for accurately representing variables like uptime or data volume. Collecting relevant data
points is imperative for establishing a statistically valid model. Questions regarding the duration of the
simulation process and the number of iterations become pivotal in ensuring valid and accurate results.
While the power of process simulation lies in its capability to incorporate randomness into the model, it
doesn't inherently provide the optimal solution. To identify the best solution, it is imperative to define
various scenarios and simulate them. Experimental design proves valuable in this context, facilitating
the exploration and identification of optimal solutions. By employing experimental plans, one can
systematically compare the performance metrics of the current state with those of potential future state
alternatives [6].

3.4 Process Simulation Output

Business process simulation is a powerful tool that enables organizations to model and analyze their
business processes in a virtual environment. By simulating different scenarios and adjusting
parameters, organizations can gain valuable insights and outputs about their process performance,
helping them identify areas for improvement and make informed decisions to enhance efficiency and
reduce costs [7][6].

Business process simulation outputs provide quantitative, time-based, and cost-related data regarding
how processes are carried out and resources are utilized. Additionally, it can generate a range of
performance measures, which are the basic performance measures, including cycle time, entity count,
resource utilization, and activity cost [8]. The first performance measure is cycle time, which can be
defined as the total time an entity spends traversing a process, including value-added process time,
waiting time, movement time, etc. The calculation of the minimum, average, and maximum cycle times
based on all running process instances, which is considered to be the most vital and fundamental output
of a business process simulation. For the second performance measure, entity count measure, it
includes the total number of processed entities plus the entities that are still in-process. The third
performance measure is resource utilization, which is the percentage of time that a resource spends in
each state, because throughout a simulation, entity states can change from busy to idle or from
unavailable to reserved, for example. Resource utilization offers useful indicators for measuring and
analyzing under-utilization or over-utilization of resources. The last performance measure is activity
cost, process model defines resources by the number of units that are available that includes usage
costs, setup costs, and fixed costs, while an activity is defined by the resources that are needed to
complete it and the duration of its completion. Following the process simulation, the amount of time
each entity spends in an activity and the amount of time each resource is assigned to that activity are
calculated, which provides us with the cost calculations we need to measure and analyze the activity

6
 cost. These are the general business process simulation outputs [6]
Business process simulation can generate several additional possible outputs beside the basic
performance measures, as illustrated in Figure 2. First is the process log, which is a detailed process
log that contains the data obtained from the simulation process and can be analyzed with a process
mining framework. Second is a set of benchmarks with some diagrams. Benchmark is a popular
technique that a business can use to compare its performance with other best-in-class competitors.
Additionally, Key performance indicators (KPIs) encompass not only the basic performance metrics
mentioned above, but also advanced measurements that provide a comprehensive assessment of
process performance [12].

Figure 2. Process Simulation Input and Output.

The output results from process simulation serve as a foundation for recommending changes or
improvements to the business processes. It helps organizations understand the underlying issues and
provides a basis for informed decision-making. However, the effectiveness of simulation outputs
depends on their applicability to real-world processes and the accuracy of the simulation model
[7][6].

Business process simulation offers numerous advantages. It allows organizations to predict process
performance using quantitative measures, providing a basis for informed decision-making and process
optimization [8][12]. Simulations are flexible and can be used to investigate complex situations,
enabling organizations to explore various scenarios and evaluate the impact of different process
parameters. Simulations foster creativity and simplify the understanding of intricate concepts by
providing visual representations and interactive models [5]. However, it is important to consider the
limitations of simulation studies. They can be time-consuming, requiring extensive simulation runs to
obtain accurate results. The interpretation and reliability of simulation outcomes can be challenging, as
they do not provide definitive proof and may not capture all real-world scenarios [5].

7
 4. Simulation Tools
Business process simulation tools are software applications enabling businesses to model their
processes and simulate them virtually [6].
These tools leverage mathematical algorithms and statistical models to replicate real-world scenarios,
offering insights for process enhancements. With these tools, businesses pinpoint bottlenecks, explore
diverse scenarios, and make data-driven decisions for process improvements [9].

The advantages of employing simulation tools in business processes are multifaceted. They enable
testing scenarios and informed decision-making prior to real-life implementation, leading to cost
reduction, risk mitigation, and performance optimization. By simulating processes, businesses uncover
inefficiencies, bottlenecks, and areas ripe for enhancement. Experimenting with strategies and variables
sheds light on their impact on process outcomes. Moreover, simulation aids in employee training,
preparing them for unforeseen situations or emergencies [10]. Enhanced process efficiency through
simulation can elevate customer satisfaction levels by providing better service [9].

A simulator, also known as a simulation model, is crucial for conducting simulation experiments and
comes in two distinct types [5]:
1. Simulation language: Utilizing a programming language compliant with simulation standards.
2. Simulation package: A tool within the business process modeling and notation software suite,
streamlining model specification selection to expedite model construction.

Furthermore, four commonly known techniques exist for process simulation, detailed for better
comprehension in the table below [3]:

Table 1. Process Simulation Techniques.

Technique Description Focus Examples

Discrete-Event Modeling and simulating processes Process event Arena, Simul8,
Simulation (DES) involving discrete events like sequencing [10]. ProModel, FlexSim
Tools manufacturing, logistics, and supply [10].
chain operations. These tools
employ queuing theory, handling
entities (e.g., customers or products)
moving sequentially through
systems [3].
System Dynamics These software tools are employed System changes AnyLogic, Vensim
Simulation (SD) to simulate and model processes analysis [10]. [10].
Tools System characterized by continuous
Dynamics (SD) variables, such as population growth
or environmental changes. System
Dynamics (SD) tools operate on the
principle of feedback loops, wherein
alterations in one variable can exert
influence on other variables within
the system [3].
Monte Carlo Uses random sampling to model Probabilistic outcomes, Finance, engineering
Simulation uncertainty in a system[13] risk assessment[13] designs[13]

8
 Agent-Based P portrays system participants as Process entities [11]. NetLogo, Repast,
Simulation (ABS) active components within a business Repast, GAMA
process, adhering to predetermined [11].
rules. These participants, also
referred to as agents, encompass
various entities such as departments,
institutions, deliverables, and the
workforce [6].

All simulation tools offer businesses a cost-effective means to replicate real-world scenarios and base
operational decisions on data. As previously noted, simulating processes aids in pinpointing
bottlenecks, inefficiencies, and areas ripe for improvement. Additionally, businesses can experiment
with diverse strategies and variables to gauge their impact on process outcomes [9].
However, utilizing simulation tools for business processes presents certain limitations and challenges.
Firstly, these tools demand substantial data for creating accurate models. The reliability of simulation
models hinges on the accuracy and relevance of the data used, emphasizing the need for precise and
updated information. Moreover, interpreting simulation models might pose complexity, particularly for
non-technical users, necessitating specialized skills or additional training [6]. Additionally, the cost
associated with acquiring and maintaining simulation tools might render them inaccessible to some
businesses. Building and running simulations can also be time-intensive, potentially limiting their
applicability in fast-paced business environments. Furthermore, resistance to change among employees
might impede the adoption of suggested changes stemming from simulation results, hindering the
efficacy of these tools in driving process enhancements [9].
Nevertheless, business process simulation tools empower businesses to make informed decisions that
yield better outcomes. When effectively harnessed, these tools enable businesses to optimize processes
for maximum efficiency and profitability [10].

9
5. Conclusion
Business processes, known for their complexity and constant evolution, exceed the capabilities of
flowcharting and spreadsheet analysis in conducting thorough assessments. Simulation emerges as
the most powerful and feasible method for evaluating business process performance. Furthermore,
simulation accommodates diverse activity durations, complex resource connections, and other
multifaceted elements influencing performance over time. Developing business process models
directly from process modeling tools and real data through simulation offers significant benefits to
organizations. However, advancements are still required in integrating simulation effectively within
this domain, as its application continues to advance and mature.

10
References
[1] D. JAN. MARLON. LA ROSA, MARCELLO. MENDLING, Fundamentals of Business Process
Management. Berlin, Germany: SPRINGER, 2018.

[2] W. M. van der Aalst, “Business process simulation revisited,” Lecture Notes in Business
Information Processing, pp. 1–14.

[3] G. Hook, "Business Process Modeling and simulation," Proceedings of the 2011 Winter
Simulation Conference (WSC), Phoenix, AZ, USA, 2011, pp. 773-778.

[4] V. Hlupic and G. J. Vreede, “Business process modeling using discrete-event simulation: Current
opportunities and future challenges,” International Journal of Simulation and Process Modelling,
vol. 1, no. 1/2, p. 72, 2005.

[5] W. M. van der Aalst, “Business process simulation survival guide,” Handbook on Business
Process Management 1, pp. 337–370, 2014.

[6] K. Tumay, “Business process simulation,” Proceedings of the 28th conference on Winter
simulation - WSC ’96, 1996.

[7] R. J. Paul, G. M. Giaglis, and Vlatka Hlupic, “Simulation of Business Processes.”, Aug. 1999, pp.
1551–1576.

[8] G. Vaglini and M. Giovanni, “An Interval-Valued Approach to Business Process Simulation
Based on Genetic Algorithms and the BPMN.” International Journal of Computer Science and
Applications, May, 2014.

[9] V. Bosilj-Vuksic, V. Ceric, and V. Hlupic, “Criteria for the evaluation of Business Process
Simulation Tools,” Interdisciplinary Journal of Information, Knowledge, and Management, 2007.

[10] Aalst, W.M.P. Voorhoeve, M, “Business Process Simulation, Lecture notes 2II75,” Technical
University Eindhoven, Netherlands, 2008.

[11] M. Halaška and R. Šperka, “Is there a need for agent-based modeling and simulation in Business
Process Management?,” Organizacija, vol. 51, no. 4, pp. 255–269, 2018. doi:10.2478/orga-2018-
0019.

[12] J. Banks, J. S. Carson, B. L. Nelson, and D. M. Nicol, "Discrete-Event System Simulation," 5th ed.,
Pearson, 2009.

[13] Rubinstein, R. Y., & Kroese, D. P. (2016). Simulation and the Monte Carlo method. John Wiley
& Sons.

11