6.

0 SIMULATION SOFTWARE AND APPLICATION

6.1 Categories o f M o d e l i n g a n d S i m u l a t i o n S oftware:

Modeling and simulation can be performed using three categories of software:
spreadsheets, programming languages and dedicated software packages.

Spreadsheets
Spreadsheets like Excel have some limited capabilities for simulation modeling:
 The time-slice model can be developed using the basic capabilities of a spreadsheet.
 The random numbers can be obtained using the ‘‘RAND’’ function. Samples can be
drawn from empirical distributions using the ‘‘IF’’ function.
 One can also use the lookup function (‘‘VLOOKUP’’ or ‘‘HLOOKUP’’).
 Some functions for sampling from statistical distributions are provided by Excel,
such as the normal and gamma distributions.
Above the limitations of the spreadsheets it is necessary to use some programming capabilities
within the spreadsheet, such as the macros or Visual Basic for Applications in Excel. Developing
a model animation using a spreadsheet is challenging.

Programming languages
General purpose programming languages such as Visual Basic, C++, Ruby, Python, Perl and Java
can be used to develop modeling and simulation software. The languages give the modeler a lot
of flexibility. However, it can be time consuming, since the modeler needs to develop the
simulation capabilities from scratch. Scripting languages such as Ruby and Java are useful for
developing simulations that run across the World Wide Web.

Dedicated simulation software packages

Several dedicated simulation software packages are now available. Two broad classes of
specialist simulation packages are as follows [Law and Kelton (2000)]:
i. General purpose simulation packages.
These are intended for use on a wide range of applications, albeit they might have special features
for some applications.
ii. Application orientated simulation packages.
These are focused on specific applications, for instance, medical, production scheduling or call
centers. These are more focused packages and they tend to be easier to use, mostly only requiring
the entry of relevant data, but they are however, much narrower in the range of application areas.
Many of them are Visual Interactive Modeling Systems (VIMS) [Pidd, 1998].

VIMS enable a simulation to be built as well as run in a visual and interactive manner. The
software provides a predefined set of simulation objects. The modeler selects the required
objects and defines the logic of the model through a series of menus. The visual display is
also developed through a set of menus. As a result, the modeler requires little in the way of
programming skills.
 6.2 Selection of Simulation Software
When selecting simulation software, the process can be viewed as a task that is performed in a
series of steps. The main steps include: establishing the modeling requirements; surveying and
short- listing the software; establishing the evaluation criteria; evaluating the software in
relation to the criteria; selecting the software. This process is likely to be fairly linear with
some possible iteration, between steps 3 to 5.

Step 1: Establish the modeling requirements

This step involves: identifying the nature of the systems to be modeled; determining the number
of software users and uses; determining the nature of the domain- whether general or
specific; determining the modeling approach whether the software is for fast track modeling or
complex/detailed modeling; determining any constraints that exist within the organization
such as the availability of finance, the level of software and modeling skills, and the hardware
and software policy of the organization.

Step 2: Survey and shortlist the software

Create a short list of software packages that meet the organization’s requirements. This list
should have five or less packages for consideration for a more detailed evaluation.

Step 3: Establish evaluation criteria

Establish some criteria for evaluating the software. Some Criteria for Simulation software
selection include the following:

 Hardware/software requirements: Hardware platform required; Operating system

required; Software protection (hardware security device?); Availability of network
licenses; Features for use on the world wide web.
 Model coding and testing: Ease of model development; Can a model be built and run in
small steps?; Availability of debugging aids (e.g. syntax checking, consistency checking,
trace); Maximum model size; Maximum dimensions of objects (e.g. arrays); Features
for documenting a model; Availability of help facility; Availability of software wizard.
 Visual features: Is the display concurrent with the run, or is it a playback feature?; Speed
with which display can be developed; Can user icons be drawn?; Availability of icon
libraries; Ability to pan and zoom; Ability to locate objects on the display; Smoothness of
animation; Availability of 3D animation.
 Input data and analysis features: Distribution fitting; Ability to sample from empirical
distributions; Which statistical distributions are available?; Ability to import data from
other software
 Reporting and output analysis features: Availability of standard reports for model objects;
Availability of graphical reporting; Ability to develop customized reports; Ability to
export results to other software; Statistical analysis of results.
  Experimentation: Probable run-speed; Run control (step, animated, batch); Interactive
capability; Number of random number streams available; Control of random number
streams; Ability to perform multiple replications; Facilities for organizing batches of runs-
Provision of advice on warm-up, run-length and multiple replications; Availability of
an optimizer; Ability to distribute runs across networked computers.
 Support: Availability of a help desk; Availability of consultancy support; Type of training
given; Frequency of software upgrades; What is in the next upgrade? Foreign language
versions and support; Quality of documentation.
 Vendor Background: Size of vendor’s organization; How long has the package been
available?; Have similar applications been modeled with the package?; the number of
users (in industry/private sector); Geographic usage of the package; Availability of
literature on the package and package use;
 Cost: Purchase price; Maintenance fee; Cost of support; Cost of training; Time to learn
the software; Availability of lower cost run-only licence.

Step 4: Evaluate the software in relation to the criteria

Evaluate each short-listed package according to see how they meet the selection criteria. The
following may be useful in the evaluation:
Discussion with the software vendor. Get information from technical staff rather than sales staff.
Be skeptical of the claims of software vendors, since they are trying to sell you their software.
Software demonstrations. Let the vendor demonstrate the capabilities of the software, but be wary
of demonstrations that only solve a test problem since they may show the cleverness of the
software, but not necessarily show how well the software suits the intended application.
Demonstration models. Let the software vendor develop a demonstration model of a simplified
version of the intended application. If possible, get them to develop the model in front of you.
Discussion with the users of the software. Visit and speak with other users of the software and
where possible try to locate them yourself, rather than rely upon the vendor to nominate a
reference site.
Obtaining a free evaluation copy of the software. Free evaluation copies of the software is a
useful idea if there is sufficient time to learn and use the package during the evaluation.
Software documentation. Get the copies of the documentation.
Literature. Obtain some information from the software reviews.
Expert opinion. Obtain the views of independent simulation experts, but bear in mind that they
may be subject to some biases. Also find some objective measures to evaluate the criteria.

Step 5: Software selection.

Devise some means of obtaining an overall score. It is useful to weigh the criteria according to
their importance. An overall score could then be calculated for each package as follows:
For Si = WjEji where: Si overall score for package i; Wj importance weight
criterion j; Eji evaluated score for criterion j for package i
Demonstrations:
Find some given simulations packages that work and give a demonstration.

6.3 Simulation Software

There are many simulation software available such as
1. SIMSCRIPT III
2. GPSS/H
4. Witness
5. Micro Saint Sharp
6. SIMPROCESS
7. ProModel
8. Arena
9. @RISK
10. MATLAB
They have some features in common. They provide the following features
i. Model development
ii. Random variate generation
iii. Collection and tabulation of simulation results
iv. Time keeping
v. Animation
The following simulation software are the most commonly used in the market
i. SIMPROCESS
ii. EXTENDSIM
iii. @RISK
iv. ARENA

6.2.1. SIMPROCESS
SIMPROCESS® is a hierarchical modeling tool that combines process mapping, discrete-event
simulation, and Activity-Based Costing (ABC) in a single easy to use interface.
The SIMPROCESS simulation engine provides the foundation for web-based Decision Support
Systems to move Modeling and Simulation to the management desktop.

SIMPROCESS provides ready-made building blocks for constructing dynamic business process
models, while the underlying expression language allows experienced programmers to add more
complicated business logic. SIMPROCESS is designed for organizations that wish to mitigate the
risk associated with implementing dramatic process changes. The tool allows users to quickly and
easily analyze various “what-if” scenarios, and by utilizing Java and XML technologies it provides
the necessary power and flexibility to meet these organizational needs.

Features
o Hierarchical Process Mapping
o Object-oriented Modeling
o Activity-Based Costing
o Process Animation
o Standard and Custom Reporting and Graphs
o Standard and Custom Real-Time Plots
o Model Customization through Attributes and Expression Language
o Distribution Library and Distribution Fitting
o File, Spreadsheet, and Database Interfaces
o Import and Export XPDL
o Export to PDF, RTF, PPT, and HTML

6.3.2. EXTENDSIM
ExtendSim (formerly known as Extend) is a simulation program for modeling discrete event,
continuous, agent-based, and discrete rate processes. There are four ExtendSim packages:
i. CP for continuous processes;
ii. OR (operations research) which adds discrete event;
iii. AT (advanced technology) which adds discrete rate, a number of advanced modeling
features,
iv. Stat-Fit for statistical distribution fitting; and Suite which adds 3D animation

ExtendSim is used for modeling manufacturing, healthcare, supply chain, communications,

defense, environmental, agricultural, biological, energy, reliability, service, information flow, and
recreational systems. Sample applications include resource optimization for food logistics, six
sigma process improvements for a hospital emergency department, communication systems, and
manufacturing facility design
6.3.3. @RISK
@RISK (pronounced “at risk”) is software developed by Palisade. It performs risk analysis using
Monte Carlo simulation to show you many possible outcomes in your spreadsheet model—and
tells you how likely they are to occur. It mathematically and objectively computes and tracks many
different possible future scenarios, then tells you the probabilities and risks associated with each
different one. This means you can judge which risks to-take and which ones to avoid, allowing for
the best decision making under uncertainty.

@RISK also helps you plan the best risk management strategies through the integration of
RISKOptimizer, which combines Monte Carlo simulation with the latest solving technology to
optimize any spreadsheet with uncertain values. Using genetic algorithms or OptQuest, along with
@RISK functions, RISKOptimizer can determine the best allocation of resources, the optimal asset
allocation, the most efficient schedule, and much more.

6.4.4 ARENA
Arena is a discrete event simulation and automation software developed by Systems Modeling
and acquired by Rockwell Automation in 2000 It uses the SIMAN processor and simulation
language. As of June 2014, it is in version 14.7, providing significant enhancements in
optimization and animation. It has been suggested that Arena may join other Rockwell software
packages under the "FactoryTalk" brand

In Arena, the user builds an experiment model by placing modules (boxes of different shapes) that
represent processes or logic. Connector lines are used to join these modules together and to specify
the flow of entities. While modules have specific actions relative to entities, flow, and timing, the
precise representation of each module and entity relative to real-life objects is subject to the
modeler. Statistical data, such as cycle time and WIP (work in process) levels, can be recorded and
outputted as reports.

Arena can be integrated with Microsoft technologies. It includes Visual Basic for Applications so
models can be further automated if specific algorithms are needed. It also supports importing
Microsoft Visio flowcharts, as well as reading from or outputting to Excel spreadsheets and Access
databases. Hosting ActiveX controls is also supported.
Arena is used by many large companies engaged in simulating business processes. Some of these
firms include
i. General Motors
ii. UPS
iii. IBM
iv. Nike
v. Xerox
vi. Lufthansa
vii. Ford Motor Company and others.

It has been noted that creating a simulation can require more time at the beginning of a project, but
quicker installations and product optimizations can reduce overall project time. Arena can simulate
diverse operation types, including call centers, for optimizing the use of agents and phone lines,
the size and routing of pancake stacks in a food processing facility, and the design of a gold mine.