Data Modeling and Simulation

Introduction to Simulation Modeling

Dr. Iman Abu Hashish

Department of Data Science and Artificial Intelligence
e-mail: [email protected]
Phone Ext.: 5104
Office: SB-G04

Fall Semester
Academic Year 2024/2025
Table of Contents

1. Objectives

2. Introducing Simulation Models

3. Classifying Simulation Models

4. Approaching Simulation Based Problems

5. Optimizing the Checkout Process in a Supermarket

6. References

Introduction to Simulation Modeling I. Abu Hashish 2

Objectives
Objectives

• Introduce fundamentals concepts in simulation modeling, along

with related terminologies.
• Understand the decision-making workflow.
• Classify and differentiate between different types of simulation
models.
• Formulate a simulation-based problem.
• Illustrate explained concepts through a real-life use case.

Introduction to Simulation Modeling I. Abu Hashish 4

Introducing Simulation Models
What is Simulation?

Simulation refers to reproducing the behavior of a system.

• We can use a real model which reproduces reality using a

computer – a model of an airplane replaced in a wind tunnel to
run simulated tests.
• Many theoretical laws and rules were, and still being, developed
– but to which extent are these laws applicable?

A simulation model is a tool capable of processing information and

data and predicting the responses of a real system to certain inputs.

• With the goal of providing effective support for analysis,

performance evaluation, and decision-making process

Introduction to Simulation Modeling I. Abu Hashish 6

What is Simulation?

Simulation uses digital models developed to replicate the

characteristics of a system.

• This can be achieved through probability distributions to

randomly generate system events – statistical observations are
obtained from the simulated system.
• Many scenarios can be simulated by changing input parameters
– instead of working on real systems, thus, limiting cost and
reducing time.
• Changing input parameters reproduces different scenarios by
identifying the most convenient one from various points of view.

Introduction to Simulation Modeling I. Abu Hashish 7

Decision-Making Workflow

The decision-making process starts with identifying a problem up to

making a well-informed data-driven decision.

Introduction to Simulation Modeling I. Abu Hashish 8

Decision-Making Workflow

Model building is a two-way process – definition of conceptual

models and continuous interaction between the model and reality.

1. Identify the problem that requires a change.

2. Analyze the identified problem: a) relevant elements are chosen,
b) relationships are highlighted, and c) objectives to be achieved
are defined.
3. Build a formal model – which allows the simulation of the
system to understand its behavior and identify the decisions to
be made.

Introduction to Simulation Modeling I. Abu Hashish 9

Pros of Simulation Modeling

Simulation is a widely used tool in a variety of fields.

• It reproduces the behavior of a system in reference to situations

that cannot be directly experienced.
• It represents real systems, even complex ones, while also
considering the sources of uncertainty.
• It requires limited resources in terms of data.
• It allows experimentation in a limited time.
• The models that are obtained are easily interpretable.

Introduction to Simulation Modeling I. Abu Hashish 10

Cons of Simulation Modeling

Since it is a technique capable of reproducing complex scenarios, it

has some limitations.

• It provides indications of the behavior of the system but not

exact results.
• The analysis of the output of a simulation could be complex.
• The implementation of a simulation model could be laborious
• The results that are returned by the simulation depend on the
quality of the input data.
• The complexity of the simulation model depends on the
complexity of the system it intends to reproduce.

Introduction to Simulation Modeling I. Abu Hashish 11

Simulation Modeling Terminology

Such terms refer to elements that make up a model and those that
characterize a simulation process.

• System: set of elements that interact with each other.

• State variables: a system is described in each instant of time by
a set of variables.
• Event: Any instantaneous event that causes the value of at least
one of the status variables to change.
• Parameters: essential terms when building a model, that are
adjusted through the simulation process.

Introduction to Simulation Modeling I. Abu Hashish 12

Simulation Modeling Terminology

• Calibration: the process by which the parameters of the model

are adjusted.
• Accuracy: the degree of correspondence of the simulation result
that can be inferred from a series of calculated values with the
actual data.
• Sensitivity: indicates the degree to which the model’s outputs
are affected by changes in the selected input parameters.
• Validation: the process that verifies the accuracy of the
proposed model.

Introduction to Simulation Modeling I. Abu Hashish 13

Classifying Simulation Models
Static vs Dynamic Models

Static models are the representation of a system in an instant of

time – aka representative models in which the time variable plays no
role.

A customer segmentation system is static – analyzes relationships

between customer data at a specific point in time.

Introduction to Simulation Modeling I. Abu Hashish 15

Static vs Dynamic Models

Dynamic models describe the evolution of the system over time – in

the simplest case, the state of the system at time t is described by a
function x(t).

An inventory management system is dynamic – simulates how

inventory changes, say, on a daily basis.

Introduction to Simulation Modeling I. Abu Hashish 16

Deterministic vs Stochastic Models

A model is deterministic when its evolution, over time, is uniquely

determined by its initial conditions and characteristics.

• They do not consider random elements and the output is well

determined.
• Linear programming optimization models (maximum profit or
minimum cost), are deterministic – given the same inputs, it
always produces the same optimal solution without randomness
or variability.

Introduction to Simulation Modeling I. Abu Hashish 17

Deterministic vs Stochastic Models

Stochastic models, on the other hand, can be evolved by inserting

random elements into the evolution.

• Can be obtained by extracting them from statistical distributions.

• The queuing system simulation is stochastic (customer arrival,
serving time) – the system’s behavior involves randomness,
leading to different outcomes each time the simulation runs.

Introduction to Simulation Modeling I. Abu Hashish 18

Continuous vs Discrete Models

In discrete models, the system is described by an overlapping

sequence of physical operations interspersed with inactivity pauses.

• These operations begin and end in well-defined instances –

events.
• Task scheduling models are discrete – involves scheduling
specific activities that can only start or finish at defined time
points, with each task being a distinct entity.

Continuous models represent systems in which the state of the

variables changes continuously as a function of time.

• A car moving on a road represents a continuous system – speed

and position change continuously with respect to time.

Introduction to Simulation Modeling I. Abu Hashish 19

Approaching Simulation Based
Problems
The Simulation Process Workflow

The simulation process workflow can be comprised of the following:

1. Problem analysis – identify goal, components, performance

measures.
2. Data collection – collect relevant data.
3. Setting up the simulation model – most crucial step.
4. Simulation so tware selection – simulators, simulation
languages, programming languages.
5. Verification of so tware solution – aka debugging.
6. Validation of simulation model – implement performance
measures.
7. Simulation and analysis of results – interpret performance
measures results.

Introduction to Simulation Modeling I. Abu Hashish 21

Optimizing the Checkout Process
in a Supermarket
Problem Formulation

Supermarket X has multiple checkout lanes where customers arrive

at random intervals, receive service, then leave once a purchase is
complete.

• Objective is to minimize customer average waiting time, and

decide how many lanes needed at peak time.
• Events include customer arrival (arrival rate), service start
(service time), service completion (number of lanes).
• State variables can be number of customers in queue, number
of customers being served.
• Performance metrics can be average waiting time, length of
queues, and lane utilization.

Introduction to Simulation Modeling I. Abu Hashish 23

Optimizing the Checkout Process in a Supermarket

Assuming NUM_LANES=3, ARRIVAL_RATE=1.5, SERVICE_TIME=5, and

SIMULATION_TIME=60

Introduction to Simulation Modeling I. Abu Hashish 24

References
References

• Giuseppe Ciaburro. 2022. Hands-On Simulation Modeling with

Python. Packt Publishing.

Introduction to Simulation Modeling I. Abu Hashish 26