MODELING AND SIMULATION

Ajai Jain
Professor, Department of Mechanical Engineering
National Institute of Technology, Kurukshetra, India
Email: ajayjainfme@nitkkr.ac.in

1. INTRODUCTION

A system is collection of entities (people, parts, messages, machines, servers, …) that

act and interact together toward some end (Schmidt and Taylor, 1970). It can be facility
or process, actual or planned such as manufacturing facility, bank/ insurance office,
transportation /logistics /distribution operation, hospital facilities (emergency room,
operating room, and admissions), computer network, criminal justice system, chemical
plant, fast-food restaurant. It is essential to work with system in order to study and
understand it so as to measure its performance, improve it, design /redesign as well as
control it. It is the best option if one can play with the actual system as one looks at the
right thing in actual. However, in general it is impossible to do so in reality with the
actual system as the system doesn’t exist (one is planning the new one) or it would be
disruptive, expensive or dangerous due top management restrictions. As one can’t play
with the actual system, one is not left with no choice other than to use scientific methods
to analyze the system for understanding, improvement, optimization and decision
making. For this, a model of the system is built.

Modeling is the process of producing a model. A model is a representation of the

construction and working of some system of interest. There are two types of models: (i)
Physical: Scale models, prototype plants (ii) Mathematical: Analytical queuing models,
linear programs, simulation. One purpose of a model is to enable the analyst to predict
the effect of changes to the system and to optimize the system. Thus, studying the
model instead of the real system is, usually, much easier, faster, cheaper, and safer.
Further, analyst can try wide-ranging ideas with the model and one can make one’s
mistakes on the computer where they don’t count, rather than for real where they do
count. Thus modeling assists us in decision making process about the system. Figure 1
shows the various ways to study the system.
A model should be a close approximation to the real system and incorporate most of its
salient features. Further, it should not be so complex that it is impossible to understand
and experiment with it. Generally, a model has set of assumptions/approximations
about how the system works. Thus, it becomes essential to validate the model i.e. it
should be able to mimic the real world system faithfully so that one get the same
conclusions from the model as one would from the system and contains the desired
level of details that are required to analyze it. An important issue in modeling is model
validity. Model validation techniques include simulating the model under known input
conditions and comparing model output with system output.

Fig. 1: Ways to study a system (Law, A.M.)

2. SIMULATION

If model structure is simple enough, one could use mathematical methods to get exact
information on questions of interest and is called analytical solution. But most complex
systems require models that are also complex. They must be studied via simulation to
evaluate model numerically and collect data to estimate model characteristics.

A simulation of a system is the operation of a model, which is a representation of that

system or methods and applications to imitate or mimic real systems. The model is
amenable to manipulation which would be impossible, too expensive, or too impractical
to perform on the system which it portrays. The operation of the model can be studied,
and, from this, properties concerning the behavior of the actual system or its sub
systems can be inferred. In its broadest sense, simulation is a tool to evaluate the
performance of a system, existing or proposed, under different configurations of interest
and over long periods of real time. Generally, simulation is done using computer.
Although simulation can be used to analyze any model (simple or complex) but the real
power of simulation is in studying complex model.

Simulation is used before an existing system is altered or a new system built: (i) To
reduce the chances of failure to meet specifications, (ii) To eliminate
unforeseen bottlenecks, (iii) To prevent under or over-utilization of resources, and (iv)
To optimize system performance. In a simulation study, human decision making is
required at all stages, namely, model development, experiment design, output analysis,
conclusion formulation, and making decisions to alter the system under study.
The only stage where human intervention is not required is the running of the
simulations, which most simulation software packages perform efficiently. For example
a manufacturing company considering extending its plant and wish to see how it works.
Thus, the company’s model is built and validated. Then simulate current and expanded
operations and one could investigate many other issues along the way, quickly and
cheaply.
Advantages of Simulation

Simulation is intuitively appealing to a client because it mimics what happens in a real

system. The output data from a simulation should directly correspond to the outputs that
could be recorded from the real system. In contrast to optimization models, simulation
models are “run” rather than solved. Given a particular set of input and model
characteristics, the model is run and the simulated behavior is observed. The various
advantages of simulation are as follows:

• Decision aid.

• Time stretching/contraction capability.

• Cause-effect relations

• Exploration of possibilities.

• Diagnosing of problems.

• Identification of constraints.

• Visualization of plans.

• Building consensus.

• Preparing for change.

• Cost effective investment.

• Training aid capability.

• Specification of requirements.

Disadvantages of Simulation

• Training required.

• Interpretation of results required.

• Time consuming/expensive.
Application Areas

 Designing and analyzing manufacturing systems/ Materials Handling system/

inventory system (ordering policies)
 Evaluating military weapons systems or their logistics requirements ( military)
 Evaluating designs for service organizations such as call centers, fast-food
restaurants, hospitals, and post offices (Public and Health Systems )
 Determining hardware and software requirements for a computer system
(Computer Systems Performance)
 Determining hardware requirements or protocols for communications networks
(Communications)
 Designing and operating transportation systems such as airports, freeways,
ports, and subways (transportation)
 Analyzing financial / economic systems/ ecological system ( Financial/ Natural
resource)
 Reengineering of business processes

When is simulation appropriate?

 Allows access to system internals that may otherwise not be observable.

 Informational, organizational, and environmental changes can be simulated, and

the effect of these changes on the model’s behavior can be observed.

 Observations based on simulations give great insight into the system behavior,
and it can be determined which variables are most important and how they
interact.

 Analytic solutions can be verified.

 Simulation allows to experiment with new designs or policies prior to

implementation.

 Can be used for training without the cost and disruption of on-the-job learning.
  The simulated system is so complex, that its interactions can be treated only
through simulation

When simulation is not appropriate?

 Would common sense suffice?

 Is there an analytical solution?

 Is it easier to perform direct measurements on a physical system?

 Is there a shortage of resources for implementing the simulation?

 Is there a shortage of time for getting the desired results?

 Is data lacking for modeling the system and beginning a simulation study?

 Is there enough time and personnel to verify and validate the model?

 Are the managers’ expectations unrealistic?

 Is the system too complex to be modeled?

3. CLASSIFICATION OF SIMULATION MODELS

There are various types of simulation model and they are described below.

(i) Stochastic simulation model

A simulation that contains random variables (variables containing probability
distributions) e.g. Inter-arrival time or service time of customers at a
restaurant or store, amount of time required to service a customer. Thus,
output is a random quantity and multiple runs are required to analyze output.

(ii) Deterministic simulation model

A simulation containing no random or probabilistic variables. All variables are
known with certainty, or in other words, no probability is associated with them
e.g. Simulation of a digital circuit, Simulation of a chemical reaction based on
differential equations. Here, output is deterministic for a given set of inputs.
 (iii) Static simulation model
It is a representation of a system at a particular point in time or one that may
be used to represent a system in which time simply plays no role i.e. a
simulation model where time is not a significant variable e.g. determine the
probability of a winning solitaire hand. It is important to mention that
combination of static and stochastic is known as Monte Carlo Simulation in
which statistical sampling is used to develop approximate solutions to
numerical problems.

(iv) Dynamic Simulation Model

A simulation model focusing on the evolution of the system under
investigation over time.

(v) Discrete Simulation Model

A simulation model one in which the state variables change only at discrete
or countable points in time e.g. number of customers in the bank

(vi) Continuous Simulation Model

A simulation model in which the state variables change continuously over
time. Thus, the system is s described by a set of differential equations e.g. an
airplane moving through the air, e.g. position and velocity

4. DISCRETE EVENT SIMULATION

Most operational models are dynamic, stochastic, and discrete and in general they are
called discrete-event simulation models. Discrete event simulation concerns the
modeling of a system as it evolves over time by a representation in which the state
variables instantaneously changes at separate point s in time i.e. system can change at
only a countable number of points in time. These points in time are the ones at which an
event occurs.
Simulation Terminology

(i) State of the system: It is the collection of variables necessary to describe

the status of the system at any given time e.g. length of the job queue, number of
customers waiting in a queue, status of machine (busy, idle or breakdown). One
can restart simulation from given state variables
(ii) Entity: It is an object of interest in the system i.e. customers in the study
of a bank, machine in job shop etc.
(iii) Attribute: It is a property of an entity e.g. balances in the accounts of
customers, speed, capacity, breakdown of machine etc.
(iv) Activity: It represents a time period of specified length i.e. making deposits
by customers in a bank, operation on jobs in machine shop, welding, stamping.
(v) Event: An instantaneous occurrence that may change the state of the
system e.g. arrival of a job on the machine, beginning of a new execution;
departure of a job from the machine, breakdown of the machine
(vi) Endogenous activities/Events: Activities/events occurring within a system. I.e.
completion of an operation in a machine shop.
(vii) Exogenous activities events: These are activities/events in the environment
that affect the system i.e. arrival of a customer in a bank.

5. SIMULATION OF A SINGLE SERVER QUEUING SYSTEM

The purpose of this example is to show how computer simulation is carried out.
Although if system is not complex and duration of simulation is small, manual simulation
can be carried out.

Consider a single server queuing system for which the interarrival times A 1, A2….. are
independent and identically distributed (IID) random variables. (Identically distributed
means that the interarrival times have the same probability distribution). A customer
who arrives and finds the server idle enters service immediately and the service times
S1, S2….. of the successive customers are IID random variables that are independent of
the interarrival times. Upon completing service for a customer, the server chooses a
customer from the queue (if any) in a First-in, First out (FIFO) manner.
The simulation will begin in the empty and idle state, i.e. no customers are present and
the server is idle. At time 0, we will begin waiting for the arrival of the first customer,
which will occur after the first interarrival time, A1, rather than at time 0. We wish to
simulate this system until a fixed number (n) of customers have completed their delays
in queue i.e. the simulation will stop when the nth customer enters service. It is
important to mention that the time the simulation ends is thus a random variable,
depending on the observed values for the interarrival and service-time random
variables.

To measure the performance, we will look at estimates of three quantities, i.e. expected
average delay in queue of the n customers completing their delays during the simulation
(d (n)), expected average number of customers in the queue (q (n)), and how busy the
server is (u (n)? d (n) will give information about system performance from customers
viewpoint, q (n) and u (n) will give information to the management.

The estimates of various parameters can be computed as follows:

The expected average delay in queue

n

D i
dˆ (n)  i 1
n

Di = delay in queue of ith customer

The expected average number of customers in queue


q(n)   ipi
i 0


qˆ (n)   ipˆ i
i 0

 iT  i
T(n)
Q(t)dt
qˆ (n)  i 0
 0
T ( n) T ( n)

Where pi = the expected proportion of the time that Q (t) is equal to i

Q (t) = number of customers in queue at time t

The expected utilization of the server

T (n)

uˆ (n) 

0
B(t )dt
T ( n)

1 if the server is busy at time t

B(t )  
0 if the server is idle at time t

Let the interarrival and service-time of customers are

A1 = 0.4, A2 = 1.2, A3 = 0.5, A4 = 1.7, A5 = 0.2, A6 = 1.6, A7 = 0.2, A8 = 1.4, A9 = 1.9, …

S1= 2.0, S2 = 0.7, S3 = 0.2, S4 = 1.1, S5 = 3.7, S6 = 0.6

Let us assume that we want to carry out simulation when sixth customer enters the
service. Thus simulation will end when the sixth customer leaves the queue and enters
service. The simulation is shown on slides. At this point of time, the values of
parameters of interest are as follows:

Estimate of d (6) = 5.7/6 = 0.95

Estimate of q (6) = 9.9/8.6 = 1.15

Estimate of u (6) = 7.7/8.6 = 0.90

6. SIMULATION VALIDATION
Use of a model or simulation is a surrogate for experimentation with an actual system
(existing or proposed), where experimentation with that system could be disruptive, not
cost effective, or infeasible. If the model or simulation is unable to provide valid
representations of the actual system, any conclusions derived from the model or
simulation are likely to be erroneous and may result in poor decisions being made.
Validation can be performed for all models and simulations, regardless of whether the
corresponding real-world system exists in some form or will be built in the future.
Validation should always be focused on the intended use.

Validation is the process of determining the degree to which a model or simulation is an

accurate representation of the real world from the perspective of the intended uses of
the model or simulation .The following are some general perspectives on validation:
(i) Conceptually, if a simulation is “valid”, then it can be used to make decisions about
the system similar to those that would be made if it were feasible and cost effective to
experiment with the system itself.
(ii) The ease or difficulty of the validation process depends on the complexity of the
system being modeled and on whether a version of the system currently exists
(iii) A simulation of a complex system can only approximate the actual system, no
matter how much time and money are spent on simulation construction. There is no
such thing as absolute simulation validity, nor is it even desired. Indeed, a model or
simulation is supposed to be an abstraction and simplification of reality. However, the
most valid simulation is not necessarily the most cost effective. For example, increasing
the validity of a simulation beyond a certain level might be quite expensive, since
extensive data collection may be required, but might not lead to significantly better
insight or decisions.
(iv) A simulation should always be developed for a particular set of objectives. Indeed, a
simulation that is valid for one set of objectives may not be for another set of objectives.
(v) The measures and acceptability criteria (e.g., measures of performance [MOPs])
used to validate a simulation should include those that the decision-maker will actually
use for evaluating system configurations.
(vi) Validation of a stand-alone simulation is a process that should be conducted in
coordination with the development or modification effort. It is not something to be
attempted after the simulation has already been developed (or modified) and then only if
there is time and money remaining.
(vii) A federation of models still has to be validated even if the models (federates) that
compose it are believed to be valid.
A model or simulation, its data, and its results have credibility if the decision-
maker and other key project personnel accept them as “correct”. Note that a credible
simulation is not necessarily valid, and vice-versa. The following factors help establish
credibility for a model or simulation:

Decision-maker’s Understanding and Agreement with the Simulation’s

Assumptions
Demonstration that the simulation has been validated and verified.

Decision-maker’s ownership of and involvement with the project.

Reputation of the simulation developers.

A model or simulation that is both valid and credible is more likely to be formally
accredited for use in a particular application.

7. MODELING AND SIMULATION IN MANUFACTURING SYSTEM

Manufacturing and material handling system provide one of the most important
applications of simulation. Simulation has been used successful as an aid in the design
of new production facilities, warehouses and distribution centers. It has also been used
to evaluate suggested improvement to existing system. Analysis using simulation have
found it valuable for evaluating the impact of capital investments in equipment and
physical facility and proposed changes to material handling and layout. Managers have
found simulation useful in providing a “test drive” before making capital investments.

Manufacturing and material handling simulation need to contain the proper level detail.
The major guidelines for capturing the correct level of detail are the objectives of the
study and the questions being asked. Level of detail is constrained by the availability of
input data and the knowledge of how system components work. For new non-existent
system, data availability may be limited and system knowledge may be based on
assumptions. Models of manufacturing system may have to take into account a number
of characteristics of the system such as

 Physical layout
 Equipments
(i) Capacities
(ii) Breakdowns (time to failure, time to repair, Resources needed for repair)
 Maintenance
(i) Preventive maintenance schedule
(ii) Time and resources required
(iii) Tooling and fixtures
 Storage
(i) Supplies
(ii) Spare parts
(iii) Work-in-process (WIP)
(iv) Final goods

Models of material handling system may have to contain some of the following type of
subsystems.

 Conveyors

(i) Accumulating /Non-accumulating

(ii) Indexing and other special purpose
(iii) Fixed windows or random spacing
(iv) Power and free.
 Transporters
(i) Unconstrained vehicles (i.e. manually guided fork trucks)
(ii) Guided vehicles (automated or operator-controlled, wire-guided chemical paths,
rail-guided)
(iii) Bridge cranes and other overhead lifts
 Storage System
(i) pallet storage
(ii) Small part storage (totes)
(iii) Automated storage and retrieval system (AS/RS)
Performance Measures

Those who purchase and use simulation software want to gain insight and
understanding into how a new or modified system will work and ask several questions
such as will it meet throughput expectations?, what happens to response time at peak
periods?, what are the staffing requirements?, what is the system capacity?, what
conditions and loads cause a system to reach its capacity? While simulations are
expected to provide numeric measures of performance, such as throughput under a
given set of conditions, the major assistance in the communication of models
assumptions, system operations and model results is of course a rigorous analysis. For
stochastic simulation models a proper statistical analysis, is of almost importance for the
simulation analyst to draw correct conclusion from simulation output. Some common
measures of system performance include:

 Throughput under average peak load

 Utilization of resources, labor and machines
 Bottlenecks and choke points
 Queuing at work locations
 Queuing and delays caused by material handling devices and systems
 WIP storage needs
 Staffing requirements
 Effectiveness of scheduling system
 Effectiveness of control system

Often, material handling is an important part of a manufacturing system and its

performance. Non-manufacturing material handling system includes warehouses,
distribution centers etc. There are a number of modeling issues especially important for
the advisement of accurate and valid simulation models of manufacturing and material
handling systems. Two of these issues are the proper modeling of downtimes and
whether to use actual system data for some inputs versus a statistical model of those
inputs.
8. LANGUAGES/SOFTWARES USED IN SIMULATION

Computer simulation languages generally facilitate the development and execution of

simulations of complex real-world systems. It can be classified into three broad
categories.

(i) Special-purpose simulation software

(ii) General-purpose programming languages
(iii) General-purpose simulation languages

(i) Special-purpose simulation software

Depending upon the application particular software can be selected which is

highly user-friendly. For example, simulation packages such as Enterprise Dynamics,
Pro Model, Quest, WITNESS, and Arena are useful for the simulation of manufacturing
and material handling systems.

(ii) General purpose programming languages

In some special cases of simulations where the simulation software could not
serve the purpose due to complexity nature of the problem, some general-purpose
programming languages such as JAVA, C/C++ can be used to simulate the system.
Object oriented programming (OOP) was developed specially for writing discrete-event
simulation models. The object-oriented extensions to C which become C++ were
motivated by the need to model a telephone system.

(iii) General-purpose simulation languages

SIMAN V, SIMSCRIPT11.5, SLAM11, GPSS/H, SLAMSYSTEM etc. are high-

level simulation programming languages which have constructs specially designed to
facilitate model building. These languages provide the simulation analyst with a choice
of orientation (process interaction or event scheduling) or a model using a mixture of the
two orientations. Unlike general purpose programming languages, these languages
provide management of the future event list and other sets, built-in random variate
generators, and built-in statistics gathering routines. Complex computations can be
accomplished in both implementations of GPSS and in SIMAN V, SIMSCRIPT11.5, and
SLAM11. These languages provide the capability to conduct continuous simulations.

9. SUSTAINABLE MANUFACTURING

Sustainable manufacturing refers to developing and practicing technologies to transform

materials with reduced energy consumption, reduced emission of greenhouse gases,
reduced generation of waste, and use of non-renewable or toxic materials. It includes
the integration of processes, decision-making and the environmental concerns of an
active industrial system to achieve economic growth, without destroying precious
resources or the environment. Sustainability applies to the entire life cycle of a product.
It involves selection of materials, extraction of the materials, manufacture of component
parts, assembly methods, retailing, product use, recycling, recovery, and disposal.
Some examples include:

 Develop products that last longer or have extended service lives

 Re-design or reformulate products to use easily recyclable materials
 Substitute toxic with non-toxic materials in manufacturing processes
 Improve manufacturing technology, equipment, materials, tools, and operations
to Minimize environmental impacts
 Adopt cleaner production technologies to reduce waste, better manage industrial
waste, and maximize recycling
 Develop technologies to reduce energy consumption, recover waste heat, and
use renewable energy sources, such as solar, wind, bio-mass, landfill, and
digester gases
 Develop reusable packaging products

Presently, modeling and simulation is used to help a company to meet design and
production objectives and it seldom addresses sustainable manufacturing aspects.
There is very little published research where manufacturing simulation includes
environmental concerns or parameters in the modeling process. The research
addresses a narrow aspect of manufacturing sustainability and is mainly conducted
using existing simulation tools or a minor enhancement of the modeling. Current
simulation products do not typically support the modeling of environmental concerns or
impacts; e.g., energy consumption or carbon footprint, waste/hazardous materials
disposal, and pollution. The types and quantities of manufacturing waste associated
with manufacturing processes are not often a simulation consideration. Simulations
typically do not model the by-products of manufacturing, generation of effluents, or their
disposal. Information on alternative manufacturing technologies, processes, and data is
not readily available to the simulation analyst to incorporate into models. Effects of good
housekeeping on the reduction of waste and pollutants are not supported. Similarly,
regional differences in environmental safety requirements are not represented in
simulation environments. Occupational safety and health considerations may not be
accounted for in process models. Recovery, recycling, and life cycle costs (LCC) of
materials are often not addressed in design and manufacturing simulations. Simulations
usually do not deal with the usage and disposal practices of product users after sale.
These issues are not modeled today because of the way the manufacturing simulation
systems were developed and evolved. Further the metrics for sustainable
manufacturing are different than traditional manufacturing and incorporation of these
sustainable aspects into the modeling and simulation tool requires additional
programming. A metric is a simulation output measure of a process, operation, or
system that is crucial to the evaluation of alternative solutions. In the past, metrics have
focused on enhancing product quality, improving production efficiency, and/or reducing
costs. In the future, additional metrics will be required to evaluate “what-if” propositions
for sustainability. Possible sustainability metrics will measure energy consumption,
types and quantities of material used, pollution, manufacturing waste and by-products,
recycling, product reuse, worker health/safety, and other effects on the environment and
the community. A table of the traditional metrics used today and new sustainability
metrics for the future is presented in Table 1 below.
Table 1: Traditional and sustainability metrics

Traditional Metrics Sustainability Metrics

Efficiency Energy use
Number of jobs produced per unit of time % of energy use that is renewable
job flow, cycle, or manufacturing lead time energy use per unit of product made
makespan of a set of jobs energy cost per product unit
job queuing time energy cost as % of total expenses
queue lengths plant heating and cooling energy
job transfer time efficiency measures
worker and equipment utilization
equipment downtime due to breakage, blockage or Pollution
starvation emissions per unit of product output
machine utilization effluents that are captured and treated
balance of equipment utilization greenhouse gases that are captured and
overall plant capacity utilization treated
time to market cost of fines and charges due to
job lateness pollution
number of jobs tardy carbon footprint of products and
proportion of jobs tardy processes
order lead times noise level measures
travel distance for products and components
inventory turns Material usage
work-in-process % recycled/recyclable materials used
scrap and rework measures % environmentally-friendly materials
used
Costs and returns output per unit of material used
cost of carrying raw materials and work-in-progress output per units of water used
inventories lifetime of materials used
machine cost per unit time environmentally-friendly packaging
material handling/transportation costs measures
labor cost
energy cost per unit time
total job completion cost
payback periods
return on investment in plant and equipment
paper process management measures
Waste
quantity of waste produced per unit
output
percentage of waste materials recovered
waste water recovery measure
costs to recover (and dispose or reuse)
discarded product, if regulations impose
so

Worker health and safety

work-related accidents and injuries
ergonomics issues consideration in
material
handling and processing
healthcare costs due to occupational
accidents
compensation costs due to work related
injuries and suffering
worker job repetitiveness, satisfaction
level, morale factors
lost work days due to injuries
lost production due to injuries
Community impact
land usage, green space, etc.
traffic impact on local roads
use of public transportation by
employees
rain water capture
REFERENCES:

1. Deogratias Kibira & Charles Mclean, “Modeling And Simulation For Sustainable
Manufacturing” NIST paper
2. Law, A. M., “ Simulation Modleing and Analysis, 4 e Tata McGraw Hill, 2008.
3. Banks, J., J. S. Carson, and B. L. Nelson. 1996. Discrete-Event System
Simulation. 2d ed. Upper Saddle River,New Jersey: Prentice-Hall.
4. Lecture notes