Using the ModelBuilder of ArcGIS 9 for Landscape Modeling Jochen MANEGOLD, ESRI-Germany Geoprocessing in GIS A geographic information system (GIS) provides a framework to support planning tasks and decisions, to help managing the natural and man-made environment and the resources of the earth. With providing tools for all kind of geoprocessing, the GIS framework allows the user to define, manage and analyze all the information used to support planning or to make decisions. ESRI’s ARC/INFO was built upon such a framework, using coverages as an intelligent data model for storing geometry, topology and attributes, and offering all tools to capture, store, and manage as well as to analyze and present landscape data. With Arc Marco Language (AML) a scripting language can be used to store geoprocessing workflows and dialogs. ArcGIS 8 marked a new milestone in ESRI’s GIS client and server architecture. On the client side, completely new software architecture, was released, built on modern GIS and application framework components, including new applications like ArcMap and ArcCatalog. On the server side, the Geodatabase offers a new object-relational data model to store intelligent geo-objects with properties and behaviour in a Relational Database Management System (RDBMS) as well as “link” those objects and layers by flexible topological rules, thus extending the possibilities far beyond the coverage model. ArcGIS 9, a new geoprocessing framework which includes a model builder as a graphical environment to create diagrams of steps for complete geoprocessing tasks will be introduced to make geoprocessing much easier, more flexible and very user friendly. Geoprocessing Tools with ArcGIS 9 The main concept of geoprocessing is based on a concept of data transformation. A typical geoprocessing operation takes an input dataset, performs an operation on that input dataset and returns the result of the operation as an output dataset. ArcGIS 9 will have hundreds of these single geoprocessing tools for processing all types of ESRI compatible spatial and non-spatial data formats, namely coverages, shapefiles, personal and ArcSDE geodatabase featureclasses, with CAD and VPF files, as well as tables and text files. These tools can be grouped into different types of operations, such as data conversion, analysis or data management. For convenient search and retrieving, these groups of similar tools are organized in ToolBoxes and Toolsets, which can be stored in system files or database tables. These ToolBoxes are presented to the user as a list of his favourites in the ArcToolbox window for easy access in the ArcGIS applications like ArcMap and ArcCatalog. The number of potentially available tools varies, depending on the type of product license such as
30
Embed
Using the ModelBuilder of ArcGIS 9 for Landscape Modeling · Using the ModelBuilder of ArcGIS 9 for Landscape Modeling 5 A model contains information about data, how to process the
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Using the ModelBuilder of ArcGIS 9 for Landscape Modeling
Jochen MANEGOLD, ESRI-Germany
Geoprocessing in GIS
A geographic information system (GIS) provides a framework to support planning tasks
and decisions, to help managing the natural and man-made environment and the resources
of the earth. With providing tools for all kind of geoprocessing, the GIS framework allows
the user to define, manage and analyze all the information used to support planning or to
make decisions.
ESRI’s ARC/INFO was built upon such a framework, using coverages as an intelligent
data model for storing geometry, topology and attributes, and offering all tools to capture,
store, and manage as well as to analyze and present landscape data. With Arc Marco
Language (AML) a scripting language can be used to store geoprocessing workflows and
dialogs.
ArcGIS 8 marked a new milestone in ESRI’s GIS client and server architecture. On the
client side, completely new software architecture, was released, built on modern GIS and
application framework components, including new applications like ArcMap and
ArcCatalog. On the server side, the Geodatabase offers a new object-relational data model
to store intelligent geo-objects with properties and behaviour in a Relational Database
Management System (RDBMS) as well as “link” those objects and layers by flexible
topological rules, thus extending the possibilities far beyond the coverage model.
ArcGIS 9, a new geoprocessing framework which includes a model builder as a graphical
environment to create diagrams of steps for complete geoprocessing tasks will be
introduced to make geoprocessing much easier, more flexible and very user friendly.
Geoprocessing Tools with ArcGIS 9
The main concept of geoprocessing is based on a concept of data transformation. A typical
geoprocessing operation takes an input dataset, performs an operation on that input dataset
and returns the result of the operation as an output dataset. ArcGIS 9 will have hundreds of
these single geoprocessing tools for processing all types of ESRI compatible spatial and
non-spatial data formats, namely coverages, shapefiles, personal and ArcSDE geodatabase
featureclasses, with CAD and VPF files, as well as tables and text files. These tools can be
grouped into different types of operations, such as data conversion, analysis or data
management.
For convenient search and retrieving, these groups of similar tools are organized in
ToolBoxes and Toolsets, which can be stored in system files or database tables. These
ToolBoxes are presented to the user as a list of his favourites in the ArcToolbox window
for easy access in the ArcGIS applications like ArcMap and ArcCatalog. The number of
potentially available tools varies, depending on the type of product license such as
J. Manegold 2
ArcView, ArcEditor or ArcInfo and the types of ArcGIS extensions like 3D Analyst or
Spatial Analyst.
Fig. 1: ArcCatalog and the ArcToolbox window
Geoprocessing Methods in ArcGIS 9
There are four different ways to perform geoprocessing tasks in ArcGIS 9. Which method
you choose depends on which method is best suited to the particular task and your personal
preference. You can choose a dialog from the ToolBox window or use the command line
for input typing to perform a single geoprocessing task. With dialogs, the tools provide a
form, where you can specify the data and parameters for the geoprocessing operation.
Using the ModelBuilder of ArcGIS 9 for Landscape Modeling 3
Fig. 2: The command line window in ArcMap
The command line is similar to the Workstation ArcInfo command line, where you type in
the command including all input data and parameters. The command line in ArcGIS 9
prompts you with the usage of the specified command and helps you with intellisense and
dropdown lists.
For more complex tasks or workflows, involving multiple functions, choose a model tool or
create a new model and link single processes together. To perform the same function many
times on different datasets or with different parameters, you can use or create a tool derived
from a script.
Fig. 3: A geoprocessing script written in Python using the PythonWin IDE
J. Manegold 4
Scripting offers an efficient and effective way of managing geoprocessing needs. They
provide an environment, which makes it easy to process large volumes and quantities of
data. They are recyclable and easily modified. Any scripting language, which is COM
compliant, interacts well in the ArcGIS environment such as VBScript, JScript or Python.
Using ModelBuilder in ArcGIS 9
The ModelBuilder in ArcGIS 9 provides a graphical environment to create a diagram of
multiple steps to complete a complex geoprocessing task. The diagram you build represents
a model. Tools can be draggen from ToolBoxes in the ArcCatalog tree or from the
ArcToolbox window into the Model diagram to build the processes that make up the
model, then filled in the necessary input data and parameters for each tool and connect the
processes together. When the model is run, ModelBuilder processes the input data in the
order specified and creates output data. The model can be saved, modified and rerun.
Geoprocessing functions can be tied together in a model that keeps track of the datasets,
processes, parameters and assumptions that are used. This makes it easy to redo the exact
same procedure multiple times, or to alter data and parameters slightly.
Fig. 4: The ModelBuilder window with a typical model
Using the ModelBuilder of ArcGIS 9 for Landscape Modeling 5
A model contains information about data, how to process the data, and the sequence of
processing. The model can be displayed as a process flow called a model diagram or as a
tool in a toolbox. A tool exposes model parameters selected by the model author and is the
common way of executing a model. In the model diagram, there are symbols, called model
elements, which represent data and tools that operate on data. The model elements are
connected together into processes. A process is a set of input data, the tool that operates on
the input data, and the resulting output data. Connector arrows indicate the sequence of
processing. There will usually be several processes in a model, and they can be chained
together so that the derived data from one process may become the input data for another
process.
The model diagram provides a graphical way to present models to decision-makers and the
public. So the ModelBuilder can be used for cartographic modeling and analysis, such as
environmental or land use modeling.
Fig. 5: A conceptual model with different processes
Figure 5 is a conceptual overview of a model built from three processes. The processes are
built by connecting model elements representing data and functions. Each process has one
or more input datasets, a tool, and output data. Connecting the output from one process to
the input of another process chains the processes together.
In addition to the basic model elements such as input and output data, tools, and
connectors, there are text labels, which are graphical elements used to place explanatory
J. Manegold 6
text in a model. A label is not part of the processing sequence. Labels can be attached to
connectors or they can float freely in the model diagram.
Model elements have properties and connection rules that define how they behave. Data
elements have a property that describes the location of the data. Function elements have
properties that describe how to process data. For example, the Buffer function has a buffer
width property. Properties are editable through dialogs, wizards, and property sheets.
Connection rules describe what type of input each model element will accept and what type
of output it produces.
Sharing a model is as simple as sending the model file via email or publishing a ToolBox
on a server or with a web service. The receiving party simply saves the file anywhere on
their system. A user can then add the file to the ArcToolbox window for easy access. The
concept behind model sharing, or any ToolBox tool for that matter, is to maximize
resources and productivity by avoiding duplication of effort, and because it simply makes
more sense, other than from just a cost-effective point of view.
Summary
The evolution of geoprocessing has shown that there is no single user experience that
satisfies all user requirements and preferences. Early packages exposed geoprocessing
operations as a set of commands while more contemporary systems have a graphical user
interface (GUI) in which the user interacts and composes the operation to be executed. A
third method exposes the systems core functions in an application programmer interface
(API) that allows a software developer to invoke geoprocessing operations outside of the
GIS systems GUI. Each method has its advantages and disadvantages, but when all three
are supported in a GIS package, the user gets to choose which best suits his or her needs,
thereby offering the best possible user experience for the task at hand.
References
ESRI Inc. (2002): Geoprocessing in ArcGIS 9.0, White Paper
ESRI Inc. (2002): Creating a Model in ArcGIS 9.0, White Paper
From Professional to People’s Software –
Tracing the Development of 3D GIS Software at ESRI
Jinwi MA, ESRI Environmental Systems Research Institute, USA
1 Abstract
3D Analyst is ESRI’s commercial 3D GIS software that was initially released in 1998.
Over the years, it has significantly increased its popularity, gaining a wide range of support
from GIS professionals to casual desktop users. In its development path, it has
continuously evolved with the goal to meet the needs of user communities like that of
landscape architects. This paper analyzes 3D Analyst development history in various
aspects. Three main phases of 3D Analyst development are identified: the formative age as
an extension product of ArcView 3, the architecture shift implementation for
ArcInfo/ArcView 8, and the enhancement and new establishment at ArcGIS 9. 3D
visualization and surface analysis remain as the two fundamental pillars for 3D Analyst.
After the architecture shift to COM-based implementations at version 8, 3D Analyst has re-
directed its focus on realistic visualization and high performance over large datasets.
Combining the powers of 2D GIS and 3D CAD systems, it is expected to be an effective
tool for planning and design communities. This paper, as an attempt to promote
communication between the software developer and its user communities such as that of
landscape architects, provides an overall picture to the community about the development
of 3D Analyst.
2 Introduction
Founded thirty years ago, ESRI has developed a number of sophisticated professional GIS
software. The flagship product, ArcInfo, was the first vector-based overlay and
cartographic GIS solution released commercially. Over the years, ArcInfo has evolved
from an early monolithic Unix workstation program that was used by a relatively small
circle of GIS professionals into a component-based desktop software solutions employed
by both professionals and casual users. Along the way, ESRI has expanded the core
desktop product with many value-added extensions. 3D Analyst is one such extension.
From 3D Analyst’s first release in 1998 up to now, it has been evolved and matured in its
own right. For GIS applications in landscape architecture, one cannot ignore the role of 3D
Analyst as it is continuously catering to the needs of its users. To understand its current
status, it is important to trace its development path.
3 Pre 3D Analyst Era
There were a suite of ArcInfo workstation surface functions already in use well before the
3D Analyst product was first released. Those functions mostly deal with surface analysis
J. Ma 2
with Triangulated Irregular Networks (TIN). Based on these functions, some highly
professional surface analysis custom application modules were developed. They are
distinctively non-conventional from 2D cartographic point of view and visually rich and
attractive. It tended to portray the world intuitively in a non-abstract way. This was
already a breakthrough in a cartography-oriented GIS, but seemed lacking some key
component if it were geared toward a 3D analysis environment. That key component is
real-time interactivity.
In the mid-1990s, the success of desktop product ArcView 2/3, a product designed for light
GIS users to complement the heavy end of professional ArcInfo workstation GIS,
prompted the development of various extension products. The popularization of the
Windows NT operating system, the standardization of the industry-standard OpenGL API,
and the increasing performance and the decreasing cost of personal computers all added
more fuel to the flame. As a result, a new desktop product, called 3D Analyst, went under
development.
4 Early Stages with ArcView 3
In the Summer of 1998, after two to three years of internal research and development, the
first version of 3D Analyst was released. It was released as an extension product of the
core ArcView 3 desktop GIS soon after the release of another extension product, Spatial
Analyst. It was the first major desktop GIS application software released from ESRI that
had real-time interactive 3D visualization and surface analysis capabilities.
For the first time, the data display in a GIS by ESRI was not confined to a 2D environment
(with limited zooming and panning capabilities), but in a lively, dynamic, and interactive
one not necessarily oriented toward cartographic map production. In other words, the
exploration process in the 3D environment itself IS the communication media, i.e. to
communicate virtually, rather than through hardcopy. Viewing objects in 3D perspective is
important as landscape architects recognize that “the ability to support design creativity
might be enhanced if designs could easily be viewed and evaluated in 3-D during earlier
stages of the design process” (Tai, 2002).
As an extension of the GIS core software, 3D Analyst utilizes standard GIS data formats.
To obtain maximum usability, it also supports other non-proprietary standard data formats.
For surface data types, 3D Analyst uses TIN and Grid, which are ESRI proprietary, and can
also use standard surface data such as USGS DEM. For vector data types, it uses ESRI’s
ArcInfo coverage data as well as the non-proprietary, de facto industry standard
‘shapefile’, which was initially defined by ESRI but its format had been published. It can
also directly use standard CAD data such as DWG, DXF, and DGN files. Most standard
raster image formats are supported by 3D Analyst. For 3D output, it can export to standard
VRML 2.0 format, which can be viewed in VRML browsers.
Beside 3D visualization, surface analysis is another important aspect of a 3D GIS. With
3D Analyst, users can easily perform basic surface analysis tasks such as contour, slope,
From Professional to People’s Software – Tracing the Development of 3D GIS 3
aspect, hillshade, viewshed, cut/fill analyses, and area and volume calculation, all in an
intuitive 3D environment. More visualization and analysis functionalities would be added
to the product. At this time, however, there was an industry-wide movement and a major
architecture shift was in the making.
5 3D Analyst with ArcView 8/ArcInfo 8
Even though 3D Analyst for ArcView 3 supports multiple platforms and operating systems,
it was built on an aging architecture lacking extensibility and scalability. At this time,
Microsoft has released its Component Object Model (COM) after several years of matured
implementation with its own software products. Considering this fact, and the majority of
Windows users, ESRI decided to embark on an architecture shift to the Windows-specific
COM technology, on which ArcInfo 8 and its various extension products would be built.
Substantial resources were spent on the architecture shift throughout the ESRI product line,
and for 3D Analyst, there was a completely new application under consideration, named
ArcScene.
The first release of the COM-based 3D Analyst was in the Summer of 2001 at
ArcInfo/ArcView version 8.1 (with ArcScene as its main application program) with its
main goal of a stable transition from the old architecture. Yet, along the way many new
features were added to make it quite enhanced from the old 3D Analyst for ArcView 3.
Among them the most prominent ones are some of the navigation tools such as the fly tool
(fly through the scene using the mouse alone as a control) and the gesture tool (rotate/spin
at either direction and with adjustable speed by mouse gestures). At the rendering side,
new features included interactive light source positioning, layer drawing priority setting,
picture fill symbols for textures with transparency, and front and back face culling. 3D
perspective view had been the default setting for the scene viewer and users can opt for
orthographic view (2D map like), if needed. For image displays, higher default texture
resolution had been achieved through internal texture tiling. For export, in addition to
VRML 2.0, GeoVRML became another supported format. Additional raster formats (such
as PNG format) were supported for scene viewer snapshot. A new, extensible ActiveX
scene viewer control was also developed so that a new, navigable 3D viewer could be
embedded into other standard Microsoft applications such as Word, Excel, and Power
Point etc. Good news for third party developers, Avenue was replaced by Visual Basic or
Visual Basic for Applications as the standard customization language. Finally, by building
software based on COM, the problems associated with incompatible software versions
went away.
The subsequent release of 3D Analyst was in early April 2002 with ArcInfo 8.2. The most
prominent feature with this release was the addition of the built-in animation capability (see
Ma and Bayarri, 2001). After an animation is created, it can be exported to standard AVI
animation files for sharing with users who do not have the 3D software.
J. Ma 4
6 3D Analyst under ArcGIS 9
If the main goal of the 3D Analyst released at ArcInfo 8.1 was to migrate from the old
architecture, then the coming new release of 3D Analyst at ArcGIS 9 (Summer 2003) will
be a quantum leap in its own right. A brand new application, called ‘ArcGlobe’, brings
significant enhancement in the visualization and the performance aspects of the product.
For 3D applications, one cannot ignore the quest for realism. Even though this is almost
the first thing to have in many other 3D software, it was not in the product function list of
3D Analyst, until ArcGIS 9. There could be many factors in this delay, such as the fact
that 3D Analyst was essentially derived from a traditionally 2D cartography oriented GIS
setting, the gap between CAD and GIS worlds (see following analysis), high cost or low
performance of graphics hardware, and lack of efficient data structure/algorithms and so
on. Now we have special software dedicated to 3D applications, most of GIS data can be
managed in standard formats and stored in the GeoDatabase, graphics hardware
performance and cost ratios are historically low, and software researchers have designed
optimal algorithms using more effective data structures. The time was ripe and super-
realistic real-time navigation can be achieved, and it is being realized in 3D Analyst for
ArcGIS 9.
Unlike ArcScene, which is based on a conventional Cartesian coordinate system, ArcGlobe
is globe-centric. It employs a hierarchical structure to store data with varying levels of
detail (LOD). The hierarchy of data, representing different levels of resolutions, are either
pre-processed or processed on the fly and stored on local or remote cache. The appropriate
resolutions of data are retrieved from cache to memory for viewing based on the extent of
the viewer frustum cast on the terrain, thus the larger the scale (meaning the closer the
observer to the terrain), the smaller the extent and the more detail of the data – the memory
consumed keeps relatively constant (see Crawford et al, 2003). This is truly an elegant
solution for managing large amount of geographic data in a 3D viewing environment that
requires real-time navigation because the interaction performance will not suffer even if the
amount of raw data is extremely large (gigabytes of data is very common). Since it is a
globe-centric application, meaning the model is really a globe, there will be no ‘edge of the
world’ effect, as can be seen in some ArcScene applications. All GIS data with valid
spatial reference can be loaded into the application without the need for special treatment at
the user’s end. Since it uses a 3D globe as its core and not two dimensional, like ArcMap,
users need not worry about conventional cartographic projection issues. Considering
traditional GIS users are more or less cartographers, the elimination of map projection
requirement in ArcGlobe is significant. The implicit requirement for cartographic
knowledge on 3D GIS users is relaxed.
Plans for future ArcGlobe development also include the transition from a desktop product
to a web-based product. With data shared across the Internet, globe viewer would be more
like a web browser rather than a desktop software. We expect to see a lot more non-
professional users attracted to 3D Analyst via ArcGlobe.
From Professional to People’s Software – Tracing the Development of 3D GIS 5
The introduction of 3D symbology at ArcGIS version 9.0 makes it possible to render
realistic objects that can match those shown in popular computer gaming software. This is
realized through direct importing of some popular 3D data formats, namely 3D Studio,
OpenFlight, and VRML, into 3D symbol libraries. These 3D symbols do not only take 3D
model’s accurate geometries, they also have the original textures carried over. Before
version 9.0, 3D objects in scene look like a working model in an architecture lab for their
monotone color and lack of textures (unless textures are added by customization). Now it
can simulate the real world with rich textures. With realism handled at both the global
macro level and 3D symbols at the local micro level, a full range of 3D realism is achieved.
The built environment (or hardscape) is relatively easy to simulate. The real challenge,
however, is to simulate the natural landscape (or softscape), especially trees and shrubs.
There are various levels of vegetation models (see Ervin and Hasbrouck, 2001), but as a
first attempt to simulate realistic trees, the surface billboard implementation is adopted for
its visual effectiveness with a simple geometry that provides for efficient rendering. To
simulate natural environment effectively is always a challenge and much research needs to
be undertaken in this area.
7 3D Analyst for Landscape Architects
Landscape architects were among the earliest users of GIS (see Hanna, 1999). ESRI’s
president Jack Dangermond (himself a landscape architect) referred to landscape architects
as “geographic designers”, or “people who approach spatial problem solving holistically”
(ibid). There are many stimulating threads on the Landscape Architecture Electronic
Forum (maintained by Prof. James Palmer of Syracuse University). One of the interesting
topics is about CAD vs GIS. It is natural and efficient to use GIS as an aid to help with
landscape planning, especially natural resource planning, but is GIS suitable also for
landscape design considering the current situation of CAD domination in the field? The
emergence 3D GIS software like 3D Analyst, comfortably assures a positive answer.
However it would be beneficial for both GIS software vendors and the landscape
architecture community to understand what has been achieved in the development of 3D
GIS towards landscape architecture applications, and what has not. This will guide future
development.
CAD software (especially AutoCAD) is a mature tool in the design field and its user base is
well-established. Yet the total number of landscape architects (most of them using a CAD
package as the design aid) remains small, compared to that of architects and civil engineers.
In this smaller community, however, more and more landscape architecture firms require
CAD experience with new hires (Tai, 2002). On the other hand, GIS is still a relatively
new field and its impact is still growing. A GIS is first about geography, i.e. sensitive to
locations, and second it is an ‘information system’. A GIS user benefits from the system
by being able to efficiently retrieve and effectively display location sensitive information
about a place. Is this what a landscape architect wants? Not always but sometimes. It is a
full-time job to be a GIS professional. Is it required that landscape architects should also
be GIS professionals? Not exactly but some knowledge would definitely help. Therefore,
J. Ma 6
an ideal modern landscape architect would need to possess knowledge and expertise not
only in CAD but also in GIS, at least partially.
According to the Fact Sheet about Landscape Architecture (see ASLA), there are basically
two kinds of activities in the profession: planning related and design related. The new 3D
Analyst, combining the best functionalities of both worlds (CAD, with 3D capability, and
conventional GIS, i.e. 2D GIS), stands as a strong contender as an effective tool for
landscape architects. It can handle landscape planning at smaller scale covering large areas
as well as landscape design at a larger scale covering small areas. Global or regional
landscape planning is about large scale design solutions that concern extensive areas of
land (ibid). This is where ArcGlobe, the new application of 3D Analyst, shines. Local area
large scale designs focus on physical dimensions and psychological impact or expressions,
and that is where 3D Analyst’s CAD-like symbology functionalities come into play. The
new 3D Analyst is not only good at 3D functions at the local scale (as a 3D CAD system
can do), it also excels in 3D operations at the global scale. In this sense, 3D Analyst
appears to have bridged the gap between planning and design fields; it is a GIS but can also
be utilized as a CAD system.
8 Summary & Outlook
Looking at the development path of the 3D Analyst, we can identify roughly three phases.
The first is the formative and exploration stage, when an extension product of ArcView 3
was produced. At this stage, the basic 3D GIS environment and functionalities were
established. The second is the transition stage, as 3D Analyst remained as an extension of
ArcView 8 or ArcInfo 8, and a new application program ArcScene was created. At this
stage, the new architecture shift was completed and some new features were added. The
third is the expansion and self-identity stage and it is where we are standing right now. On
one hand ArcScene is improving, and on the other, a brand new program, ArcGlobe, is
taking its new shape as a new 3D GIS application that is uniquely different than its
cartography counterpart, ArcMap. Through these three major stages, 3D Analyst has re-
created itself by completing an architecture shift, greatly enhancing its visualization and
significantly boosting its performance. It is not difficult to foresee its potential to emerge
from the shadow of being an extension product of ArcMap, a professional cartography-
oriented GIS software, to become a self-contained product on a par with or even surpass
ArcMap in its influence and to become a more popular, rather than professional software.
We are now at a critical point in the development of 3D Analyst as the quest for realistic
visualization and high performance takes higher priority. Another important feature with
increasing demand is 3D data interactive editing. Since the data to be edited could be the
same as those used by ArcMap, the proper execution and implementation of such an editor
needs to be resolved. Moreover, true 3D volumetric model may become the next task to
tackle. Dealing with temporal data is perhaps the most challenging job of all because to
date, all existing GIS data are virtually time ‘dumb’, meaning the temporal information, if
any, are still relegated to feature attribute status; they should be elevated to the same level
as the fundamental geometry of the data. The good news is that research has been initiated
From Professional to People’s Software – Tracing the Development of 3D GIS 7
on integrating the temporal data oriented product, Tracking Analyst, with 3D Analyst.
These two products, with much commonality especially for their animation features, were
separately developed, until today. As 3D Analyst is being developed, advanced users
would keep posing more challenging requests, and ESRI would need to carefully evaluate
various user demands/requests for future development.
Of course, the biggest help for the 3D Analyst’s development will have to come from its
user community like that of landscape architects. There are various efforts from both
inside and outside of ESRI trying to make GIS tools more effective and user friendly for
landscape architects, and 3D Analyst is proudly one of such endeavors.
9 Acknowledgements
The author appreciates the help from 3D Product Manager Clayton Crawford in reviewing
the original draft and providing constructive comments and suggestions. The author is also
grateful for the reviewing and comments from Paul Yoshitomi of ESRI’s
Osterloh, Horst, Straßenplannung mit Klothoiden, Bauverlag GmbH 1991
Interdisciplinary Cooperation in the Development of
Customer-Oriented Brand Architecture
Johannes DROSDOL, Joachim KIEFERLE, Andreas WIERSE and Uwe WÖSSNER
Abstract
At DaimlerChrysler innovative technologies such as Mixed Reality (MR) and interdisciplinary cooperation technologies are not just applied to vehicle design engineering, they are also tested in marketing to ensure consistent and continued development for customer-oriented brand architecture. Under the draft title of ‘BrandStudio’, all brand experts will initially be provided with a laboratory where realistic simulations of architectural elements, exhibition concepts and theatrical presentation processes can be generated, new potential approaches can be discussed and prepared without the constraints of a hierarchy and customer-interaction processes can be researched and scientifically evaluated.
This concept was applied during the development of the Group’s new BrandCenter and used as the basis for the ‘MB Center Milan’ pilot project.
The knowledge gained from this project is described below.
1 The Task
A brand is the discriminator that forges the purchasing decision.
Customers no longer purchase the pure product – they purchase a brand promise and a brand style – a range of items that offer personal distinction and image.
Purchasing preference is based on trust
Customer perception relies on all instruments of communication, from conventional advertising at trade fairs, exhibitions, the Internet, right through to brand architecture, working in synergy to create an intrinsically coherent brand world.
Architecture can’t not communicate
The architectural image of a brand is part of its brand world. It is this image that communicates with the (potential) customer. The selective development of a brand architecture contributes towards images being implanted in customers’ minds and therefore to the brand character itself.
J. Drosdol, J. Kieferle, A. Wierse and U. Wössner 2
Brand architecture development is a never-ending task for
interdisciplinary experts.
To keep step with the spirit of the age and exceed customer expectations at any time, architectural development must continue to evolve. The Group’s products are sold world-
wide. Brand spirit and local involvement should emanate from every point of sale. To round out the list of requirements, the architecture should also consider the various aspects of multiple brand management.
The function of DaimlerChrysler’s Architect-Center is the continued development of the customer-oriented brand architecture.
The sum total of the issues described here reflect the complexity of this function and demands an effective collaboration between many experts – brand managers, architects, artists, media designers, suppliers, distribution network planners, communication strategists, dramatists and many others – right from the initial planning stages.
2 The Initial Solution
The initial solution for this function is a method developed to both facilitate and foster a trusting collaboration between experts from different disciplines. The fundamental concept is to replicate the company’s knowledge of the brand, architecture and presentation etc. using network communication, thereby activating the implied knowledge of the Group. The draft title for this human-machine interface for the development of brand worlds is ‘BrandStudio’. It ideally combines the confidence-building face-to-face communication with the virtual and digital worlds of simulations. The immersive design environment enables all participants, from the uninitiated to the expert, to find their solution faster as they understand more clearly and earlier what their conversation partner is actually thinking.
Moderated meetings in this studio serve to improve the quality of the architectural design process, because all issues ‘are on the table’ quickly and clearly.
Development of Customer-Oriented Brand Architecture 3
Each of the participating experts is given the opportunity to be the co-author of the future. Each expert with his specialist knowledge is involved and heard. This is how the implied knowledge of the Group is activated.
Crucial elements of this working environment are:
• consistent visualization of all designs
• intermediary support for the collaboration
• complex process simulation, also with future users.
Digital mock-ups of the proposed architecture projects reduce costs through being able to make an early, holistic and interactive assessment of the building design and all its variants in their real scale. Expensive follow-up work on the real building is largely eliminated.
3 Project History
Initially, existing architectural designs were visualized in Mixed Reality (MR) in order to test the potentials of the VR laboratory. The dialogue process was then incorporated as a structured procedure and a communication between experts was established.
These activities were implemented in the DaimlerChrysler Virtual Reality Center (VRC) in Sindelfingen. The VR environment (3 CAVEs and more than 15 different powerwalls, high-performance computers and PC clusters) already in use for efficient vehicle design engineering was for the first time applied to building development.
Various VR software products were tested for their applicability to architecture development and the working methods of the various brand experts were also observed, controlled and optimized.
This highlighted a number of interesting facts.
COVISE proved its flexibility as the VR software extremely suitable for being adopted into the VR environment unfamiliar to architects and brand creators. Operation through mike and keyboard, movement modes even on steps and location-related simulated noises were of prime importance, all this, of course, was reliant upon good team interaction and an approach backed by equal understanding and absolute commitment.
Other important observations: by presenting designs in any scale, even to a model scale of 1:100, many newcomers got the hang of the new medium faster. For important discussions and the issue regarding the best arrangement of architectural elements, the 1:1 presentation
J. Drosdol, J. Kieferle, A. Wierse and U. Wössner 4
in the confined CAVE was preferred even over the generous powerwall (7.5 x 2.5 m). The interactions and real-time simulations were crucial.
The activities focused on 3 essential aspects:
• A new concept for the Mercedes-Benz BrandCenter was developed as a result of the distribution process restructuring. Then came the prominent BrandGallery, an eccentric frustum that houses changing brand exhibitions and will soon be on show in some important European capitals.
• Work on the ‘MB Center Milan’ project, which also represents the prototype for the new BrandCenter. Priority was given to refining concrete and creative issues in this project. Also the experimental illustration of potential planning and design alternatives, color variations and much more.
• The software required to depict the MB Center in Mixed Reality was further adapted to meet architectural demands. Many functions have emerged through observing the actions of all participants and as a result of requirements from working with architects and exhibition designers.
The following illustration contains several examples that are possible in a VR environment with COVISE and that represent added value during the development of brand architecture:
COVISE and Architectural Visualization
COVISE was developed initially at the High Performance Computing Center at the University of Stuttgart. Even back in the 90s, architectural and planning applications were utilizing the Virtual Reality functionality of COVISE; the collaboration with University institutes guaranteed that the needs of these applications would be considered during development of the software framework.
The advantage of COVISE lies in the flexibility of its modular approach. It is not only possible to visualize geometric data (such as CAD or landscape data), the visualization of physical properties can also be performed simultaneously. This allows the user to go beyond the simple visualization of “visible” entities, such as the photo-like rendering of buildings. COVISE enables standard visible geometries to be combined and what are normally invisible entities to be displayed: the flow of the air around a building, the flow of water in a river, the temperature in an air-conditioned environment.
Development of Customer-Oriented Brand Architecture 5
Although it is possible to display high-quality, photo-like scenarios in COVISE, its focus during the development has been on providing a useful tool for the day-to-day tasks of the architects.
Exhibition Generator
The use of an “exhibition generator” allows any number of elements to be interactively and freely integrated into the scene. It is possible to create, vary and assess exhibitions from dramaturgy to arranging individual elements – right through to the ultimate decision. And all that can be done in
just one meeting within the interdisciplinary team.
The result of the discussion can be visually reported and means that the imminent decision can be taken from a mature perspective.
Models and Sectional Planes
The representation of a complex design to any scale (from 1:500 to 1:1), variable sectional planes through these virtual models, the online separation of individual components and entire assemblies, the visualization of the air flow through or in the building and the representation of the temperature flow creates new opportunities for architects, energy experts, brand managers and exhibition designers alike.
Any necessary changes can be reviewed and initial solutions discussed and implemented within the team.
Augmented Tool for Input∗
The elements – in the picture 3 vehicles and a plasma screen – of a still fictitious exhibition can be positioned using an additional intuitive input tool. At the same time, all participants in the CAVE can visually assess this position change immediately, they can move between the vehicles and the plasma screen and assess the influence of the anticipating audience much more
quickly and clearly, thereby adopting the design of such an exhibition more quickly and with a higher quality.
The figure below shows the model vehicles and corresponding markers loaded into the virtual scene by video camera.
The combination of virtual scene and reality (of the model) affords the observer an intuitive interaction, especially with complex scenes.
∗ This Augmented input device is a research project with the University of Stuttgart
J. Drosdol, J. Kieferle, A. Wierse and U. Wössner 6
4 What are the Highlights?
The ‘BrandStudio’ - a meeting
point for experts and decision
makers.
This design immediately generates ‘spatial’ objects which can be portrayed and observed in human proportions; to a ratio of 1:1 and in real time.
This means that building owners and architects have before their eyes the designed building in its real size and perspective and can even walk inside the building (CAVE). The figures merge to form a continuous 3-D impression, whereby the user is no longer just the ‘external’ observer, he becomes a part of the virtual environment.
All those involved in a design process – brand managers, brand strategists, sales planners, designers, architects and building owners and decision makers can walk through the virtual BrandCenter, interact with the model, move components such as walls, staircases, projection elements and much more, vary light sources and therefore, in this simulated reality, hold a clear dialogue regarding requirements and their realization.
Creating real added value in the design process:
• Decisions are taken from a mature perspective • Variants can be discussed in the VR meeting in a timely fashion • Real discussions that result in new knowledge even among experts are possible
5 The Future
Look the customer in the eye and know the market.
In the ‘BrandStudio’, we will be offering architects and planners another unique opportunity.
We plan to guide test customers through the virtual car showroom. The reactions of these customers will reveal whether the planned room dramaturgy and the presentation concept are appropriate. Applying special interview techniques, indicators will be determined that allow conclusions to be drawn on the purchasing decision process. The brand researchers have what is referred to as a ‘showroom clinic’ for this purpose. Because only those who really understand their customers can make customer-oriented presentations. That’s why it’s important to be able to test and simulate the selection and purchasing process under laboratory conditions.