Proceedings of the 2002 Winter Simulation Conference

E. Ycesan, C.-H. Chen, J. L. Snowdon, and J. M. Charnes, eds.

THE EXTEND SIMULATION ENVIRONMENT

David Krahl
Imagine That, Inc.
6830 Via Del Oro
Suite 230
San Jose, CA 95119, U.S.A.

scheduling, resource assignment, and even how subtle timing issues are resolved. These tools reduce the amount of
time required to gain confidence in the model.
When the model is completed, an interface is often
created to make the model usable and immediately recognizable by someone unfamiliar with the construction of the
model. Extend includes drag-and-drop tools for user interface creation as well as a variety of methods for communicating with other programs such as Microsoft Excel and
Access. This type of interface creation generally requires
no programming and can be done in very little time.
Finally, should the system prove particularly demanding to model, Extend includes a full featured programming
language, ModL. Anyone familiar with the C programming language will be instantly familiar with ModLs syntax and capabilities. However, ModL is easier to use in a
simulation application than a general purpose language and
more powerful than the scripting languages typically included with other simulation software.
These factors combine to make an efficient tool for
modeling large and small systems alike.

ABSTRACT
The Extend simulation environment provides the tools for
all levels of modelers to efficiently create accurate and
credible models. Extends modern, advanced design and
rich feature set reduce the amount of time developing,
validating, verifying, and analyzing simulation models.
Model builders can use Extends pre-built modeling components to quickly build and analyze systems with little or
no programming. Simulation tool developers can use Extends built-in, compiled language, ModL, to develop new
reusable modeling components. All of this is done within a
single, self-contained software program that does not require external interfaces, compilers, or code generators.
1

INTRODUCTION

Simulation software needs to be capable of easily adapting

to the type of problem at hand. With some modeling tools,
simulation modelers can find themselves adding model
constructs such as overlapping resources or synthetic
workstations to capture the system behavior in the simulation model. The extra modeling time required to determine
and build these components into the model slows the
model development process. Extends design greatly reduces the need for artificial constructs. The wide variety of
modeling components and the way in which they can be
combined together allows modelers to capture the elements
of the real-world system directly.
Once a model is completed, the modeler must first verify the model to ensure that it is operating as he or she believes that it should and then validate the model to ensure
that it is a reasonable facsimile of the real system. Extends
model transparency allows the modeler to easily see how
the model is operating. This transparency includes interactive model execution, modeling components that display
historical information about the model behavior and the interaction between other components, and an interactive debugger which, along with open source, allows the modeler
to see every detail of the model operation including event

EXTEND PRODUCTS

The Extend family of products is designed to meet the needs

of the entire enterprise. Table 1 illustrates the range of Extend based products sold directly by Imagine That! In addition to these, third-party developers have created their own
vertical market modules in diverse areas such as chemical
processing, supply chain, pulp and paper manufacturing, and
others. All products based on Extend include:

205

Drag and drop modeling.

A full suite of interprocess communication tools
for communicating with other applications
Hierarchical modeling capabilities
Evolutionary optimization
A complete development environment for building custom components.

Krahl
Table 1: The Extend Product Family
Extend Product

Description

Typical use

Extend CP

Drag and drop simulation for continuous

modeling
Advanced discrete event modeling capabilities added to the continuous modeling of Extend CP
Adds an integrated database and high speed
systems modeling to Extend OR

Continuous modeling of scientific and engineering

systems
Manufacturing, healthcare, communications, service industries, transportation, logistics, and business processes
Complex systems where it is useful to separate the
model data from the structure or high speed / high
volume processes
Organizations which need to model complex processes and build high quality animations

Extend OR
Extend Industry
Extend Suite

Adds Proof Animation and Stat::Fit for distribution fitting to the Extend Industry package

Extend models are constructed with library-based

iconic blocks. Each block describes a calculation or a step
in a process. Block dialogs are the mechanism for entering
model data and reporting block results. Blocks reside in libraries. Each library represents a grouping of blocks with
similar characteristics such as Discrete Event, Plotter,
Electronics, or Business Process Reengineering. Blocks are
placed on the model worksheet by dragging them from the
library window onto the worksheet. The flow is then established between the blocks. Figure 2 illustrates the overall
structure of an Extend model.
There are two types of logical flows between the Extend blocks. The first type of flow is that of items, which
represent the objects that move through the system. Items
can have attributes and priorities associated with them. Examples of items include parts, patients, or a packet of information. The second type of logical flow is values,
which will change over time during the simulation run.
Values represent a single number. Examples of values include the number of items in queue, the result of a random
sample, and the level of fluid in a tank.
Each block has connectors that are the interface points
of the block. Figure 1 shows the connector symbols for the
item and value connectors.

EXTEND INNOVATION

Extend was the first simulation program to successfully

bring simulation to the desktop. Originally released in 1988,
Extend brought capabilities to analysts that were previously
available only on mainframe computers. The vision which
lead to the first graphical user-interface based simulation environment continues today. In the process of developing and
enhancing Extend, Imagine That! has scored a number of
firsts in the simulation industry. Table 2 illustrates a few
of the pioneering features in Extend.
Table 2: Firsts for the Extend Simulation Software
Year
1988
1988
1988
1992
1992
1995
1998
2001
2001
2001
2001
2002
4

Innovation
First template-based (library) simulation system
First open source modeling components
First simulation software designed for a GUI
First hierarchical modeling environment
First message based discrete event architecture
First Windows/Macintosh simulation system
First DDE scriptable simulation environment
First open source optimizer
First drag and drop ActiveX/COM support
First integrated support for Proof Animation
First integrated support for network communication
First integrated support for Internet data transfer

THE EXTEND MODELING ENVIRONMENT

Before looking into how Extend can be used to build models, it is helpful to understand the Extend modeling environment (Imagine That, Inc. 2001).

Item Input

Item Output

Value Input

Value Output

Figure 1: Value and Item Connector

206

Krahl

Model Worksheet

Libraries of blocks
Dialog
Help
Icon
Initial data
Behavior (code)

Block dialogs
Input data
Simulation results

ActiveX, COM, DLL...

Input data
Simulation results
Interface
Behavior (code)

Figure 2: Extend Modeling Structure

this is a Queue, FIFO block that holds the cars until requested by the next block. The wash bay is represented by
the Activity Delay block with a limited capacity of one
processing unit. The delay for the activity is specified by
an Input Random Number block (connected to the D or
delay connector). Each time a car arrives to the activity, a
new value is sampled from the Input Random Number
block. The last block in the model is an Exit block that removes the cars from the system.

Connections are lines used to specify the logical flow

from one block to another. Double lines represent item
connections and single lines represent value connections.
The concept of value connections in addition to item connections is unique to Extend. Other contemporary simulation applications require that a function be written whenever a simulation input is based on a value from another
point in the model. In Extend, this type of logic is performed without programming of any type. More importantly, the logic of the model is visible to anyone examining the model structure. To simplify the appearance of the
model, the connections can be hidden. Figure 2 illustrates
the relationship between the libraries, blocks, worksheet
and any external programs (such as an ActiveX object, Excel, or a DLL) which may be linked to Extend. It also
shows the visual nature of an Extend model. Note that the
Input Random Number blocks can be clearly recognized as
providing the delay (D) for the activities.
5

Figure 3: A Single Server Single Queue Model

Combining basic elements like this allows the Extend
modeler to build complex systems quickly and accurately.
The overall structure of the model segment is easily determined by glancing at the model.

SINGLE SERVER, SINGLE QUEUE EXAMPLE

The following example is of a single server, single queue

system. For the purpose of illustration, the model of a car
wash will be used. This car wash will include one wash
bay and one waiting line. The model for this car wash is
shown in Figure 3.
The block on the far left is a Generator block that periodically creates items (in this case dirty cars). Following

5.1 Graphical Output

A Discrete Event Plotter graphically displays model metrics (values). In this example (Figure 4), the Plotter will
graph the contents of the Queue (the number of dirty cars

207

Krahl
waiting in line) over time. Here the length connector (L) on
the Queue FIFO is connected to an input on the Plotter.

Simulation results may be stored in a table, plotted,

cloned to a different area of the worksheet, exported to another program such as a spreadsheet or database, displayed
in an animation, or even used to control some aspect of the
outside world through external device drivers.
5.3 Connectivity with Other Applications
The term interprocess communication (IPC) describes the
act of two applications communicating and sharing data
with one another. This feature allows the integration of external data and applications into and out of Extend models.
Automatic communication between Extend and other applications can take one of five forms:

Figure 4: Discrete Event Plotter Added to Model

Figure 5 illustrates a sample plot from this model.
Note the data for the plotter is displayed below the plot.
This information is saved with the model and does not disappear when the model is saved and then reloaded.

Paste-Link where the information is automatically updated between Extend and Excel
Blocks that utilize the IPC functions to communicate directly with other applications
ODBC (Open DataBase Connectivity)
Embedded ActiveX or OLE (Object Linking and
Embedding) objects
DLL (Dynamic-Link Library).

Figure 7 illustrates an embedded Excel spreadsheet

used as a reporting mechanism for a model. The information in the spreadsheet is linked to the model with Extends Paste-Link functionality.

Figure 5: Plot of Queue Length

5.2 Model Results
During and after the simulation run, the results of the simulation are reported within the blocks, displayed on plotters,
sent to reports, and exported to other applications. Doubleclicking on each block reveals the information collected
from the simulation run. For example, double-clicking on
the Queue, FIFO block opens a dialog showing the information found in Figure 6. Each block reports its own statistical information.

Figure 7: ActiveX Excel Spreadsheet Embedded in Extend

The popularity of interfacing models with other applications (especially Microsoft Excel) makes these features
powerful tools for model developers. Extend modelers do
not have to program in an external language to communicate with other applications. Instead, information can be
transferred using standard modeling components. As user
interface development is a large portion of the time required to build a model, particularly when the model is de-

Figure 6: Dialog of Queue FIFO

The Plotter block shows the number of items stored in
the Queue, FIFO over time in both graphical and tabular
format.

208

Krahl
Finally, the Statistics library helps users to collect and
analyze output data. Blocks from the Statistics library
automatically gather data from the specific blocks and calculate confidence intervals.
Extends block architecture aids the modeler in developing custom statistical calculations. Virtually any statistic
can be calculated by combining the appropriate blocks.
This makes it easier to develop custom reports displaying
familiar statistics to the model end-user.

signed to be used by non-modelers, this can represent a

significant time savings.
5.4 Integrated Database
The Extend+Industry package contains an integrated relational database. This database provides a complete data
management system for model input and output. The database is built directly into the model to house product data,
process information, and experiment with scenarios.
By separating your data from the model, the database
enables fast scenario implementation, flexible analysis and
improved project management.

5.6 Optimization
Extends Evolutionary Optimizer employs powerful enhanced evolutionary algorithms to determine the best possible model configuration. Using a drag and drop interface,
performance metrics and parameters that can be varied are
entered into the Optimizer block. These parameters are
used in an equation that defines the objective function.
When the model is run, the Optimizer block generates alternatives and locates the statistically best configuration.
Unlike external optimizers, Extends optimization is well
integrated into the program. For example, when the optimization process is complete, model parameters are automatically set to the optimal configuration. In addition, because the optimizer has been implemented in a block, the
source code is available for examination and modification.

Configure tables for experiments and reports

Use database-aware blocks to build powerful
model constructs
Assign strings to items using database-aware
attributes
Leverage dates, times and other data formats such
as currency.

The Industry database is relational and parent-child relationships can be used to better organize the information
in the model. For example, each entry in a table of part
types can reference its own unique routing table. This is an
extremely powerful feature for organizing the information
used in complex simulation models. This allows the modeler to easily modify model parameters in a central location
without having to change various values that are distributed throughout the model.

CUSTOMIZING EXTEND

The above discussion illustrates the highly graphical and

interactive nature of Extend. However, Extend can also
take the shape of the modeled system. Interfaces, components, and graphics can be created which tailor the model
to a specific application area.
The most visible aspect of a custom model is the user
interface. By modifying an existing interface or creating a
new one, the simulation modeler is able to create a model
which can be exercised by someone more familiar with the
system than with the simulation tool. This means that models can be built that fit naturally into the conceptual framework of the person using the model. The following sections
will describe some of the tools provided in Extend that facilitate customization.

5.5 Data Analysis

Extend offers a number of methods for analyzing both input and output data. These range from internal analysis features to built-in interfaces with other applications.
An interface to distribution-fitting programs is provided to aid users in selecting the appropriate statistical
distributions based on empirical data collected in the field.
In addition, sensitivity analysis can be performed to determine how sensitive a system is to changes in specific input parameters. For example: to determine how sensitive the
car wash is to changes in the inter-arrival time of dirty cars,
sensitivity analysis can be performed on the inter-arrival
mean parameter of the Generator block. By selecting the inter-arrival time dialog item and choosing Sensitize Parameter from the Edit menu, the change in the parameter value
from one run to the next is defined. Simulation parameters
such as the number of runs and simulation end time can be
specified in the Simulation Setup dialog under the Run
menu. By cycling through different inter-arrival times for the
dirty cars and comparing the results from the different runs,
an understanding of how sensitive the car wash is to the arrival rate of dirty cars can be obtained.

6.1 Animation
Animation is a powerful presentation and debugging tool
that can greatly increase model clarity. In Extend, animation icons moving from block to block represent the flow
of items through the system. Users can choose from a
number of icons provided with Extend, create their own in
an external drawing package, or import them.
Animation is automatically a part of every Extend
model. A default animation is displayed when Show
Animation is turned on. Animation features can be added

209

Krahl
to a model in the form of different animation pictures that
represent different types of items, displaying values, levels,
color changes, or even sounds in response to simulation
events. In addition, custom animation can be added to display pictures and text, level indicators, and pixel maps.
For more sophisticated animation, Extend Suite includes Wolverine Softwares animation package, Proof
Animation. Activities, Resources, Generators, and Exit
blocks each have specific functionality to send information
to the Proof animation during simulation execution (Wolverine Software Corporation 1995). Additional animation features in Proof can be accessed in Extend through the Proof
library of blocks. This allows Extend modelers to easily utilize the industrys most sophisticated animation package.
6.2 Hierarchical Modeling
Figure 8: Call Center Model with Hierarchical Blocks

In the past, there have been at least two definitions of hierarchy in simulation modeling. The first definition, coined by
Imagine That, Inc (Imagine That, 1992), describes the
grouping or aggregation of system components (blocks) into
a single object. The second definition, first referred to later
in 1992, (Pegden and Davis 1992) describes a modeling system in which new primitive modeling constructs can be created from existing primitives provided by the simulation system. It appears that, as other simulation software packages
have matured and added features, other simulation companies have recognized Imagine Thats leadership in this area
and adopted Extends definition (Bapat and Swets 2000).
Extend provides unlimited layers of hierarchy, created
using a simple menu command. Hierarchy allows models
to be subdivided into logical components or sub-models,
represented by a single descriptive icon. Double-clicking
on the hierarchical block opens a new window displaying
the sub-model. This greatly simplifies the representation of
a model and allows the user to hide and show model details
as appropriate for the target audience.
Even a medium-sized call center model can become
difficult to maintain if all of the modeling components
must be at the same level. Extends hierarchy allows the
modeler to decompose the model into smaller, more manageable segments. Additionally, new model segments can
be added by dropping in a new hierarchical block. Figure 8
illustrates the use of hierarchy to organize a model where
each icon encapsulates a separate model segment.
By selecting a group of blocks and choosing Make Selection Hierarchical from the Model menu, a section of the
model can be encapsulated within a hierarchical block. Extends hierarchy fully encapsulates the enclosed block and
does not require the renaming of variables and connections. All of the connection names within the hierarchical
block are local to that block. This allows multiple instances
of identical hierarchical blocks in the same model (Pidd
and Castro 1998). The hierarchical blocks can be copied
within a model or saved to a library to be used again in

other models. The icon for the hierarchical block can be

modified by using the built-in icon editor or by importing
an existing picture. While the representation of the model
is more intuitive and simple than a non-hierarchical model,
all of the detail of the model can still be accessed by double-clicking on any of the hierarchical blocks to display the
underlying sub-model.
By utilizing hierarchy, modelers are able to rapidly and
accurately create reusable model segments. This speeds
model development by allowing the modeler to develop reuse portions of the model. For example, a call center may
have a number of similar groups of agents (differing in
number of agents and call time distribution). One hierarchical block can be built and, using cloning, a user interface and
report are created. This block can then be replicated multiple
times. The only changes that need to be made are easily accessible in the hierarchical blocks user interface.
6.3 Dialog Cloning and the Notebook
As noted earlier, input and output parameters associated
with the model can be found in the dialogs of the appropriate blocks. While this provides an intuitive association between system metrics and the constructs used to model
them, it can make searching for specific data cumbersome.
This is especially true when working with large models
containing many layers of hierarchy. An effective way of
dealing with this is to use the Extend notebook and cloning
feature. With the notebook, a single custom interface can
be created that consolidates critical parameters, results, and
model control to a central location.
The notebook is a separate window associated with
each model. Initially, the notebook is a blank worksheet to
which text, pictures, and clones can be added. Clones are
direct links to dialog parameters and are created by selecting the Cloning Tool from the tool bar and using it to drag

210

Krahl
The ModL development environment with its interface
for editing the dialog, help, icons, connectors, and code, is
illustrated in Figure 10. Other tools include block performance profiling, include files, and an interactive debugger.

a dialog parameter from a block dialog to the notebook or

model worksheet. Once a clone is created, any changes to
the clone are immediately reflected in the block and viceversa. Therefore, it is no longer necessary to access the
blocks dialog to change an input parameter or view updated results. Creative use of the notebook can result in a
simple yet effective interface for a large, complex model.
As an illustration of how the notebook can be used to consolidate important parameters into one location, Figure 9
shows the notebook for the Call Center model.

Figure 10: ModL Block Development Environment

The advantages of a development environment such as
this are obvious. Model builders are able to easily and reliably create new or modified modeling constructs for demanding modeling situations or new applications. The significance of a powerful programming language such as
ModL should not be underestimated. Traditional simulation languages or scripting environments typically lack
full sets of language features such as flexible condition
statements (many are limited to a single condition at a
time), user defined data structures, and user interface development tools. Because ModL has all of these features
(and more), there is rarely a need for a modeler to resort to
an external language such as C++ or Visual Basic. With
Extend and ModL, only one language and interface needs
to be learned and since ModL is based on the C language,
its learning curve is typically short. With less time learning
and switching between languages, model developers are
able to develop more sophisticated models in less time.
This level of extensibility has prompted many users to
develop libraries of custom blocks for specific industries.
Users and third-party developers have created libraries for
modeling many systems including high-speed production
systems, chemical processes, silicon wafer fabrication,
pulp and paper mills, environmental processes, and radio
and microwave communication systems. Some blocks
coded by customers can be found on the company web site
(www.imaginethatinc.com).

Figure 9: Notebook for Call Center Model

Cloning and the notebook are another example of the
tools available in Extend which facilitate model development. Without this feature set, the modeler would be required to learn and utilize an external language to develop
a similar user interface.
6.4 Block Development
The block development environment is one of Extends
most powerful features. While the majority of Extend users
find the pre-built constructs sufficient for their needs, the
block development environment provides a way for users
to expand their modeling capabilities to perform unusual or
highly specialized tasks. It typically takes only minutes for
someone with any programming experience to learn the
basics of building modeling components in Extend.
Extends open source architecture allows access to the
structure of most blocks that are shipped with Extend. By
opening the structure, the icon, dialog, help text, and programming code of the block can be edited. The interface
and functionality of any block can be modified or a new
block created from scratch.
ModL is the powerful and flexible language used to
define the behavior of each block. This language provides
high-level functions and features while having a familiar
look and feel for users with experience programming in C.
In addition, external XCMDs and DLLs can be called from
within ModL, giving the option of programming in any
language which supports this feature (such as C or Pascal).

6.5 Scripting
Scripting is a feature that allows models to be created
and/or modified through a suite of ModL functions. With
this functionality, users can create objects that automati-

211

Krahl
cally build and modify models. With scripting, users can
develop their own model building wizards or selfmodifying models. Without having to rely on generalpurpose wizards provided by the software vendor, users
can develop wizards specific to their needs and can have
complete control over the level of detail and accuracy resulting from automated model building.
Coupled with Extends ability to communicate with
other applications using interprocess communication
(IPC), scripting provides an easy way to allow other applications to control every aspect of Extend, including
building the model, importing/exporting data, and running the simulation.
7

AHEAD OF SCHEDULE
AND UNDER BUDGET

Simulation modelers and consultants consistently report

that they are able to build models in Extend more rapidly
than with other simulation tools. One organization, which
uses multiple simulation products, reports that the projects
which use Extend typically require less time than budgeted. A large aerospace company completely scrapped a
network model done in another simulation software product because the single person who understood the product
left the company. This model was replaced with one done
in Extend. Because Extend was easier to learn and use, an
entire group of engineers is now able to use the model.
Many consultants who use Extend, use it exclusively
because it provides the modeling power that they need
and they are able to efficiently model any size system.
James Dailey & Associates (www.james-dailey
.com) notes that typical invoice values for completed
models are less than $10,000 and generate a more than 10
fold return on investment.
Another example is the satellite and ground station
portion of the model in section 9.1 which was built in a
single day. This was designed as a separate, stand alone
model and then incorporated into a much larger model of
the overall system.
8

SAMPLE APPLICATION

Since Extend is a general purpose simulation program, it

has been used in many types of simulation projects. Areas
where Extend has been successfully applied include manufacturing, service industries, business process reengineering, communications, logistics, healthcare, control systems,
science, environmental studies, and high speed processing.
The sample application here is of a communication system
between a group of weather satellites and ground control.
Other sample applications including medical laboratory automation, supply chain management, pulp and paper
processing, and high speed manufacturing are available in
earlier versions of this paper (Krahl 2001).

WHAT MAKES EXTEND UNIQUE

Extend provides features and capabilities not found in

other simulation software. This allows the modeler to concentrate on the modeling process and quickly produce a
model that is easy to manipulate and communicate to others. These features include:

Reuseability: Extend blocks (modeling components and hierarchical sub-models) can be saved
in libraries, reused in other simulations, and even
distributed to other modelers. This feature increases productivity and consistency of design.
Scalability: Because of Extends unlimited hierarchical structure, it is used to produce enterprisewide models with hundreds of thousands of blocks.
Visual transparency: Extends block icons are
designed specifically to convey the structure and
behavior of the model at a glance.
Connectivity: Extend supports the COM model
(ActiveX/OLE) and ODBC. Unlike other simulation tools, these technologies have been implemented in Extend as modeling components so that
interapplication communication is a drag and drop
operation, with no programming necessary.
Extendability (open source): The blocks that
come with Extend are developed using Extends
compiled language and integrated development
environment. They are open source to allow
modification and enhancement. This speeds the
evolution of better modeling techniques, as the
user can improve components and develop new
proprietary components.
Third Party Support: Because of its integrated
development environment, Extend has proven to
be the simulation engine of choice for more third
party applications than any other simulation tool.

9.1 Weather Satellite System Simulation

The model shown in Figure 11 simulates the end-to-end
performances of the TRW polar-orbiting weather satellite
system NPOESS. It models more than 3 terabytes of
weather sensor data collected daily by multiple satellites,
transmitted to ground stations at scheduled contacts, and
processed into science products at government ground facilities. The Space segment model plots the satellites
ground tracks and animates the active contacts between

Interactivity: Even during a model run, Extend

parameters and model logic can be changed on
the fly. Extends point and click interactivity
translates into faster answers and quicker, easier
restating of problems.

212

Krahl
ence, ed. J.J. Swain, D. Goldsman, R. C. Crain, J. R.
Wilson, 390-399. IEEE, Piscataway, NJ
Pidd, M. and R. Bayer Castro. 1998. Hierarchical modeling
in discrete simulation. In Proceedings of the 1998
Winter Simulation Conference Proceedings, ed. D. J
Medeiros, E. F. Johnson, J. S. Carson, M. S. Manivannan, 383-389. IEEE, Piscataway, NJ
Wolverine Software Corporation. 1995. Using Proof Animation. Annandale, VA

satellites and the ground. Ground tracks and orbital data

are imported from Satellite Tool Kit (STK) for accurate
timing of sensor data collection. The data processing facility in the upper right corner models the generation of science products executed on multiple, parallel processors. It
models the latency of generating weather data products
based upon weather and terrain content in the satellite
viewed scenes.

AUTHOR BIOGRAPHY
DAVID KRAHL is Vice President of Technical Sales
with Imagine That, Inc. He received a MS in Project and
Systems Management in 1996 from Golden Gate University and a BS in Industrial Engineering from the Rochester
Institute of Technology in 1986. Mr. Krahl has worked extensively with a range of simulation programs including
Extend, SLAM II, TESS, Factor, AIM, GPSS, SIMAN,
XCELL+ and MAP/1. He is actively involved in the simulation community and is an adjunct faculty member at
Golden Gate University. His email address is
<[email protected]> and the Imagine
That Inc. site is <www.imaginethatinc.com>.
Figure 11: TRW Polar-Orbiting Satellite Model.
10 SUMMARY
Is Extend the one-click answer to all of the worlds simulation needs? Of course not, but its intuitive interface, rich
set of modeling components, extensive authoring and development environment, and more advanced simulation
technology make it a better solution for simulation engineers who need to efficiently utilize their modeling time.
REFERENCES
Bapat, Vivek and Nancy Swets. 2000. The Arena product
family: Enterprise Modeling Solutions. In Proceedings
of the 2000 Winter Simulation Conference Proceedings, ed. J. A. Joines, R. R. Barton, K. Kang, and P.
A. Fishwick, 163-169. IEEE, Piscataway, NJ
Imagine That, Inc. 1992. Extend Software Manual. San
Jose, CA.
Imagine That, Inc. 2001. Extend Users Guide. San Jose,
CA.
Krahl, Dave. 2001. The Extend simulation environment. In
Proceedings of the 2001 Winter Simulation Conference, ed. B. A. Peters, J. S. Smith, D. J. Medeiros, and
M. W. Rohrer, 217-225. IEEE, Piscataway, NJ.
Pegden, C. Dennis and Deborah C. Davis. 1992. Arena: a
SIMAN/Cinema-based hierarchical modeling system.
In Proceedings of the 1992 Winter Simulation Confer-

213