You are on page 1of 6

Introduction to Simulation

In the past three modules, we discussed optimization and how to apply it to various supply chain cases. However, sometimes there
are complex problems that cannot be simplified and solved by regular optimization techniques, especially if risk is present.
Simulation is a powerful and expressive analytics tool that can solve these more complex problems.

Generally speaking, a simulation is an imitation of reality within a computer environment or the computerized representation of a
given real-world situation. In order to simulate, a model of the real-world scenario must be developed. Such a model is designed
to represent the key characteristics, behaviors, and functions of the particular physical or abstract world, system, or process.
Simulation models are created using software tools designed to capture and represent the system components and their
interrelationships and calculate/record their behavioral outcomes over time.

Simulations are used for predicting both the effect of changes to existing systems and the projected performance of an imaginary,
futuristic, or planned new system. Simulations are frequently used in the assessment of alternative designs, testing and validation
of operations, and calculation of risk propensity of current and future systems.

Types of Simulation Models


Simulations can be categorized based on several dimensions. The most common categorizations of simulations are based on:
1.Whether or not they include time in their representation of the underlying system (i.e., static versus dynamic simulation).
2.Whether they handle the probabilistic nature of the system variables (i.e., deterministic versus stochastic simulation).
3.Whether they perceive and represent the underlying phenomenon as a continuous system (i.e., discrete versus continuous
simulation).
In this module, we will focus on stochastic simulations called Monte Carlo simulations.
Monte Carlo Simulation
A Monte Carlo simulation is a mathematical technique used to estimate all possible outcomes under uncertainty. These
outcomes can then be used to help make the least risky decision in uncertain situations. A Monte Carlo refers to any
simulation, manual or computer-based, that utilizes random numbers to represent one or more of the variables in the
simulation model. As the name suggests, this simulation is named after the city of Monte Carlo in Monaco. In addition to its
natural beauty, Monte Carlo is famous for its casinos, chance gaming, and gambling. Although Monte Carlo simulation and
stochastic simulation are two synonymous terms, because of the name appeal, this type of simulation is commonly called
Monte Carlo—though stochastic simulation is a more descriptive and technical term for it.

The Monte Carlo simulation process involves generating random (or chance) variables to represent and illustrate random
collective behavior. Figure 1 below illustrates a simple depiction of developing a Monte Carlo simulation model.

A Monte Carlo simulation is an ideal tool to gain insight into expected outcomes with different risk factors and
assumptions. Because of its close representation of reality, the Monte Carlo simulation is quite popular in finance, options
pricing, portfolio management, and financial planning. All of these sectors involve risks, and a probability distribution as
done in a Monte Carlo simulation can represent those risks well and aid in finding an optimal solution to business
problems.
Monte Carlo Examples in the Supply Chain

Since a Monte Carlo simulation estimates all possible outcomes under uncertainty, its usage extends to supply chain and
operations management. Here are some examples:

•Make vs. Buy Decision: Organizations typically decide whether to make a product in-house or through a supplier. Assuming we
do not need to worry about intellectual property rights (usually a significant part of the decision process excluded to simplify our
discussion here), we need to worry about the cost of making it in-house vs. outsourced. We know that cost and demand can vary
over time and will have variations. A Monte Carlo simulation is a great way to simulate a range of cost and demand based on
probability distributions. Once we can simulate this with all the possible outcomes, we are in a better position to make a sound
decision.

•Determine Safety Stock Value: In Course 2, we learned how to compute safety stock by considering service level. The
assumption behind that approach is that demand follows a normal distribution. With this assumption, we can integrate a Monte
Carlo simulation into our analysis by building a model to simulate the variability and a cyclic production schedule.

•Determine the Warehouse Location: Warehouse location is typically determined by analyzing demand and supply origins. The
objective is to pick a location that best serves the customers while minimizing cost. Since variability exists in both demand and
supply, a Monte Carlo simulation is ideally suited to model the variability through probability distributions. The outcome, coupled
with a network analysis method (e.g., center of gravity), can significantly improve the chance of choosing a location that meets the
objective (minimizing cost while providing excellent service).
Benefits of Using Monte Carlo Simulation

A Monte Carlo simulation is a beneficial tool to aid decision-makers in making informed decisions. Benefits can vary depending
on the industry and the context. Here are some expected benefits:

•Repeatable: Once set up, the parameters of a simulation model can be easily updated and the simulation re-ran to obtain the
most current results.

•Scalable: The model’s input, criteria, and constraints can be added or removed. This means that the model is flexible and can
adapt to changes in business requirements. It also means that one can conduct a what-if analysis by adjusting the input, criteria,
and constraints.

•Make Informed Choices Over Taking Chances: The core of the Monte Carlo simulation is the probabilistic distribution of
events. This distribution is simulated many times to account for all the possible combinations of outcomes, which include
potential risks. As such, outcomes can be closer to reality and thereby provide decision-makers with more informed choices.
https://www.youtube.com/watch?v=emtaBhH7kf0

The formula LOOKUP(RAND(),$T$3:$T$8,R3:R8) in Excel performs a random lookup based on the generated random
number. Here's an explanation:

RAND(): Generates a random decimal number between 0 and 1.


$T$3:$T$8: Represents the range of probabilities in column T.
R3:R8: Represents the corresponding values in column R.
The purpose of this formula is to use the randomly generated number to look up the corresponding value in column R
based on the probability distribution provided in column T.

Here's how it works:

The RAND() function generates a random number between 0 and 1.


The LOOKUP() function searches for this random number in the range of probabilities in column T.
Once it finds the appropriate range where the random number falls, it returns the corresponding value from column R.
This is commonly used in scenarios where you want to simulate a random event based on given probabilities. In your case,
it seems to be used to simulate a demand scenario with associated probabilities.

You might also like