Val Steven Schmidt
Advanced Visual Systems A/S, Blokken 15, DK-3460 Birker�d, Denmark, tel (+45) 45 99 95 99 fax (+45) 45 99 95 00, email info@avs.com

Abstract

Telecommunication companies that exploit their business data effectively can gain important insight into markets, and offer new products and services to address the needs of demanding customers. Analysts in the telecommunications industry, like those in other industries, are faced with the challenging task of extracting and analyzing complex information from large databases. Knowledge workers and application developers can find help in the combination of spatial data management and data visualization tools.
This paper examines an issue faced by most businesses: how to turn data into information, and make this information accessible to persons who rely on it. Data visualization tools, used in combination with the Esri Spatial Database Engine, provide revolutionary technology to address the challenge of discovering and exploiting information. Applications using data visualization tools from Esri business partner Advanced Visual Systems are discussed, along with solution case stories taken from the telecommunications industry. For innovative companies, powerful visual applications are improving their decision-making capabilities by performing spatial and multivariate business data analysis and providing rapid access to comprehensible information.

DATA EVERYWHERE? INFORMATION SOMEWHERE

Telecommunication companies face the challenge of managing and exploiting massive amounts of customer and network operations data that they accumulate at an ever-increasing rate. Data drives business decisions; it enables telecommunication companies to offer continuous improvement in cell coverage, optimally route calls, provide responsive and personal customer care services. Companies that succeed in turning data into information and products can gain important business advantage in an intensely competitive industry.
According to industry surveys, the majority of business data in any industry is spatially related. This is particularly true for the telecommunications industry, where most data can be associated with a telephone number, an address, or the location of a network node or receiver. This data typically resides in relational databases, making it difficult to fully exploit the spatial nature of this information. Extracting data based on location - termed a spatial query - is an intuitive way to access data and perform spatial investigations. Tools such as the Esri Spatial Database Engine are revolutionizing the way spatial data is managed by providing a fast and accurate way of accessing this type of data. Data visualization applications accessing data through SDE are enabling companies to turn their complex spatial and business data into comprehensible visual information.

TOOLS FOR MANAGING AND EXPLOITING DATA

The Esri Spatial Database Engine (SDE) is an application data server that sits on top of a relational database. SDE spatially enables relational databases and allows GIS data to be associated with business data in the corporate database. Furthermore, SDE provides data querying techniques based on spatial, as opposed to relational, data dependencies. The open architecture of SDE provides an API that enables any application to interface to it, and thereby use SDE as a source of data.
AVS/Express, from Advanced Visual Systems, is a multi-platform development tool for building visualization applications and components. Included with AVS/Express are libraries of modules that provide advanced 2D and 3D data visualization, as well as an interface to the full functionality of the Spatial Database Engine. With AVS/Express, spatial data analysis applications can be extended to include multidimensional views of data. Applications built using AVS/Express can access the functionality of SDE, and offer both rapid access to data and the supporting visualization methods for turning this data into comprehensible information.

VISUAL ANALYSIS OF SPATIAL DATA

Data that is associated with addresses, telephone numbers, sales territories, or points of sale can be geographically associated. This spatial information provides the reference through which related data, such as customer profiles, or market potential for a territory, can be analyzed. Spatial data has been traditionally stored in geographical information systems, with related attribute data stored in relational databases. Using relational methods to simulate spatial query operations is, at best, quite difficult. Products like the Spatial Database Engine spatially enable a relational database and provide the mechanism for storing the vector information defining geographical boundaries, and for associating existing business data to distinct geographical coordinates. Data in a spatially enabled database can be extracted using spatial queries, that is, retrieved based on geographical location. A spatial query might begin with a question like "What are the total number of subscribers within a 5 kilometer radius from this location?" Spatial queries of this type are an effective way to extract data from a large data store. The extracted data can then be the target for further, spatial or non-spatial, business analysis.
Spatial analysis of complex business data can contribute to the process of information discovery by enabling the viewer to see and analyze data with respect to its geographical relationships. Data visualization techniques enable spatial analysis to also be applied to multivariate business data. The application shown in Figure 1 uses spatial information in two ways: to both view and access information. Visual queries greatly simplify the interface to information and are used in this example to select information. The initial display provides the visual interface for navigating through the database, in addition to providing information for each postal code region. Users can click on any region to extract and analyze additional data associated the selected area. In Figure 1 average income, shown as color, is associated with each postal code region in the London area. A second variable, containing demographic data from the census bureau, is represented by the height of each region.

Figure 1
Business data visually analyzed based of geographical location

The spatial component of data can provide an important visual reference for interactive operations such as selecting areas of interest and drilling down. This type of interactive probing is often done in the early stages of the analysis process. A visualization tool can provide a spatial view of the data, and enable the user to drill down into further views of more abstract business information not necessarily associated with a map.
The application shown in Figure 1 serves as the visual initial interface to a data query operation. Clicking on a region selects data for that region. In Figure 2, demographic data for a region is extracted and visually represented as a 3D-scatter chart. The chart organizes the data into consumer profile groups. Groups with similar characteristics are clustered, indicating attribute association. Groups with unique characteristics appear as outliers. In the scatter chart, clustered and outlier groups are easy to distinguish, yielding information on both groups which is pertinent to the market analysis.

Figure 2
Customer profiles represented by a 3D model

The combination of spatial and multidimensional analysis techniques shown by Figure 1 and Figure 2 provides users with an effective way to search for and identify important information and monitor critical business operations. Geographical information, such as that found in spatially enabled databases and geographic information systems, can provide important positional reference points for data analysis and presentation. Visualization tools enable GIS-type information to be combined with multivariate information from business data repositories, and provide views and insight into data not obtainable from a purely geographical perspective. Spatial analysis combined with multidimensional data analysis in the same application is a powerful means of discovering information.

DATA VISUALIZATION IN THE TELECOMMUNICATIONS INDUSTRY

Telecommunications companies have traditionally been quick to recognize the potential of new technology and use it to improve their business processes. The following case stories examine how data visualization has been applied by leading telecommunications companies to solve data analysis issues associated with large repositories of spatial and business data.

BRITISH TELECOM

In the British Telecommunications network there are six thousand switches and twenty-five million customer lines generating megabytes of network status and control data every minute. Overlaid on the physical network is an expanding range of voice, data, and video services, each with their own data and management requirements. Dr. Graham Walker, Technical Group Leader, Systems Research at BT describes the main driver for rapid growth in visualization activity as "the overwhelming volume of data that routinely confronts both researchers and managers."
Dr. Walker’s group investigated how visualization could be used by the sales, finance, and marketing groups to analyze the results of special promotions, price changes, and the quality of service provided to their customers. According to Dr. Walker, "Visualization provides valuable information on patterns and trends, and can realize insights and perspectives that are not immediately available using conventional data analysis and display techniques."
Applications of visualization to non-spatial data, such as financial indicators or traffic statistics, are of great interest to BT. Data visualization provides the users of financial and statistical applications with the ability to interactively explore data, and to spot trends, cycles, and patterns that would otherwise be difficult, if not impossible, to identify. For example, in Figure 3 multivariate data is visually analyzed to determine quality of service levels.

Figure 3
Multidimensional view of call service statistics

3D views, which introduce the visual element of depth, provide an expanded viewing space within which data can be organized and displayed. The 3D viewing space, termed a workbox, has 6 walls (planes) onto which information can be projected. Additional planes can be defined within the interior space of the workbox. A considerable amount of information can be placed in a three dimensional view while still maintaining clarity in the display.

Consider the graph shown in Figure 4. Produced by a customer care application, this image summarizes the overall performance of a call center. The date and time axes of the middle graph show patterns in incoming call traffic during the day. Color is used on this graph to indicate wait time. The graph clearly shows that as call traffic increases, so does the wait time for a certain percentage of callers. Aggregated information is displayed on two of the back walls. The graph to the right of the display contains additional information on response performance per operator station.

Figure 4
Visual analysis of customer call center performance

The complete display provides a visual summary of related customer service factors and is an effective overview of the performance of the call center. Informative, comprehensible, and visually appealing displays of complex information, as exemplified by the image in Figure 4, enable companies like British Telecommunications to exploit large amounts of complex data and fine-tune critical business processes.

VODAFONE

Figure 5
2D overview of national coverage levels

The tools and services provided by Advanced Visual Systems have enabled Vodafone, one of the UK’s largest mobile telecommunications companies, to implement a system for monitoring cell coverage and network performance. Vodafone has been working closely with Advanced Visual Systems to develop Vodafone Information System Analyzer (VISA), an intuitive application which the engineering department is using extensively to monitor cell and network performance. Information that was once difficult to extract from multiple Oracle databases and required considerable SQL experience is now accessible through visual queries. For workers who rely on information, complex patterns in the data are immediately recognized and comprehended through the use of data visualization.

The application shown in Figure 5 runs on UNIX and Windows platforms and provides a variety of views into the data, including vector and raster map information, 2D charts, and data tables. Users can interact with the display to select areas of interest and drill on more detailed information, such as the precise cell coverage in a dense urban area, as shown as a 3D view in Figure 6.

Figure 6
Detailed 3D view of local coverage

Dr. Barry Skuse, Chief Engineer – Performance Management, is excited by the functionality provided by their visual solution: "we are very pleased with Advanced Visual Systems and the way they have worked with us. We look forward to continued successes and further applications…the benefits have already been seen by other departments within Vodafone." With improved access to information, and visual representations of data that hasten comprehension, Vodafone has been able to improve product quality and service to customers of their mobile telephone network.

DEUTSCHE TELEKOM

Deutsche Telekom is the largest telecommunications service provider in Europe. Its position and reputation ensure that it will play a leading role in key technology developments into the next century, not least the global shift to the information-driven society and the worldwide liberalization of telecommunications.
In the ongoing quest for efficiency and cost-effective coverage, Deutsche Telekom continuously assesses all its existing and projected transmitters. In order to fine tune coverage footprints and avoid interference with other transmitters, data on factors such as land usage, road maps, population density, and simulations are all included in the analysis process. 3D views, similar to the image shown in Figure 7, provide the first approximation of the effect of variations in the surrounding terrain.

Figure 7
3D view of signal propagation

To assist planning engineers in providing optimal coverage, proposed transmitter sites are marked in Deutsche Telekom’s GIS database and a wave propagation study, using Deutsche Telekom’s AVS/Express-based application RUVIS, is performed for the selected areas. The radiation propagation modeling tool shown in Figure 8, with its embedded data visualization capabilities, is being used at Deutsche Telekom to plan broadcasting coverage for the eastern part of Germany, as well as other countries worldwide.

Figure 8
Radiation propagation modeling tool

With the emergence of new technologies, such as digital broadcasting, Deutsche Telekom is confident that the modeling system will enable it to engineer the best possible solutions and meet the demands of all its customers. According to Visualization Engineer W. Schmeing, "Using AVS/Express we can easily increase or modify the functionality of applications by just changing or adding a few objects." Visualization helps Deutsche Telekom technical specialists explore "what-if" scenarios and quickly arrive at valuable information, thus ensuring services that offer the greatest possible coverage at the lowest cost.

MULTIDIMENSIONAL ANALYSIS USING GIS AND DATA VISUALIZATION TOOLS

Geographical Information System tools can be used with spatially enabled databases to perform data queries and some level of spatial data analysis. Visualization tools differ from GIS tools in their support of multidimensional data structures. With visualization tools and data structures, complex business data can be modeled and visually represented. The resulting display can include the GIS data for reference, but is not limited to a purely geographic view.
The spatial component of data can provide an important visual reference for selecting areas of interest and drilling down. These interactive methods are often used in the early stages of the analysis process. A visualization tool can provide a spatial view of the data, and enable the user to drill down into further views of more abstract business information not necessarily associated with a map.
Applications incorporating AVS visualization can access multidimensional data from SDE and relational databases, and thus extend spatial analysis to a greater range of problem domains. AVS/Express offers 3D functionality beyond the surface modeling provided by the ArcView 3D Analyst. In addition to its visualization techniques, AVS/Express provides a true 3D data model that can be applied to a variety of multidimensional problems. The fundamental difference between AVS/Express and ArcView 3D analyst is the data model. 3D Analyst supports triangulated irregular networks (TINs) to model variations in a surface. AVS/Express uses a more generic 3D data structure called a field. The field data structure supports a number of mesh types to represent data from a broad range of application domains, such imaging, engineering, defense, finance, and telecommunications. Meshes can have many different forms, from the simplest uniform mesh having equidistant nodes and rectilinear meshes where node spacing is defined along the edges. A structured mesh has each node location separately defined but the node connectivity can be assumed. With an unstructured mesh, the connectivity information must be supplied along with the node location. At each node an unlimited number of attributes values can be present. With a visualization data model, visual data analysis, including a spatial data component, can be applied to a variety of multidimensional problems.
The analysis of complex data is facilitated through familiar and innovative multivariate information displays. These visual data representations are even more powerful when they can be interactively used to uncover additional sources of information. Advanced Visual Systems’ products support a two-way drawing pipeline that allows all graphics objects to serve as hotspots, or links, to further processing. Visual objects can be probed to reveal exact data values, or picked to drill down on finer grained views of the data. Interesting areas of the display can be selected and zoomed in on, or flown through for closer investigation.
Information is often discovered in haphazard ways. Interactively enabled displays promote information discovery by enabling the viewer to pursue data relationships through multiple or modified views of the data. Each new view is invoked by intuitive use of the mouse. The power of many information discovery applications comes not from the use of sophisticated displays, but from the sophisticated use of display interactively to promote the data analysis process. In this respect, the highly interactive nature of visualization tools offers a distinct functional advantage, even for data with a relatively low number of attributes.
The combination of spatial and multidimensional analysis provides an effective way to search for and identify important information and monitor critical business operations. Spatial information, such as that found in spatially enabled databases and geographic information systems, can provide important geographical reference points for data analysis and presentation. Visualization tools enable GIS-type information to be combined with multivariate information from business data repositories, and provide views and insight into data not obtainable from a purely geographical perspective. Spatial analysis combined with multidimensional data analysis in the same application is a powerful means of discovering information.

DIFFERENT APPROACHES TO APPLICATION DEPLOYMENT

Data visualization can greatly enrich the functionality of business applications such as Executive Information Systems, Decision Support Systems, and Customer Care Systems. The usability and acceptance of these applications on the desktop can vary depending on how the applications are designed and how the visual operations are implemented. Visual tools embedded in desktop applications are intended to help the user absorb information in larger quantities, and in less time. The mind is quick to identify visual patterns and shapes. However, as more information is represented, patterns in a scene can become so numerous and abstract that the ability of the viewer to comprehend what is being seen can be diminished rather than enhanced. One of the major challenges of visual techniques is to find interfaces and display formats that maximize information content in applications without introducing corresponding levels of application complexity. Visual tools, in addition to representing information, can act to simplify an application’s user interface, and by so doing, contribute to the application’s acceptance on the business desktop.
AVS/Express shields the user from data access commands by enabling data retrieval operations to be easily implemented as interaction with the display. AVS/Express provides modules that support the full functionality of SDE from within its object-oriented, visual programming environment. Connectivity and transaction processing with SDE and the underlying SQL database can be fully implemented from within AVS/Express. The API to SDE is encapsulated in AVS/Express modules, enabling the connection to the database and data access operations such as query and update to be easily added to an application by inserting these modules into the visual application network. The AVS/Express network provides a visual representation of application logic.
In Figure 9, a visual application network is shown, including modules to access SDE data. Modules are used to dynamically connect to SDE and add data querying to the application. In this example, the data is extracted with a visual query to SDE and visualized using a viewer module. The full functionality of SDE is accessible, but the API to SDE is hidden from the user. An application built this way can be run, or modified, without compiling and linking.

Figure 9
AVS/Express modules provide dynamic connection to SDE

The AVS/Express development system allows applications to be rapidly prototyped. In addition to the database connectivity and the application code, AVS/Express can also be used to develop cross-platform user interfaces. For example, menus and forms used for user input are easily constructed using the AVS/Express GUI kit, and provide can another interface for accessing data and controlling the application. When run under Windows, the AVS/Express-built application interface is instantiated as a Windows GUI; under UNIX the same application would have a Motif GUI.
From the AVS/Express environment, developers can also encapsulate existing legacy code, such as analysis algorithms written in FORTRAN, C, or C++. Once tested, the completed application can be output as a binary executable, or as C++ or ActiveX components. Deployment can be in the form of a new application, an application component, or a web-based client-server implementation.
Powerful, database-enabled applications can be quickly deployed using AVS/Express. Because the sophisticated data model provided by AVS/Express can represent all types of, and dimensions of, data, the visualization tool can easily operate on data from SDE and various other sources, including relational databases, data warehouses, dynamic data feeds, and files output from other applications. Visualization tools are therefore often used to integrate diverse data sources and analytical processes in a single application.
The power of a desktop information discovery application is increased when the user can access numerous types and sources of data, and perform various analytical operations on this data from within a single application environment. The application stories presented in this paper were tailor made, integrating in-house developed algorithms and analysis code with transparent access to diverse data sources. In each case, the result is powerful, focused functionality combined with ease of use that helped ensure the success of these applications, and the new techniques that they provided, among their respective users.

CONCLUSION

The Esri Spatial Database Engine extends the relational database with spatial techniques, allowing data to be accessed with intuitive and accurate spatial queries. When combined with the AVS/Express application development system, companies with large repositories of geographical and business data can rapidly develop and deploy applications that visually query and drill down on data, and turn complex multivariate relationships into easy to comprehend visual displays.
"AVS is a forward-thinking innovator of world-class three-dimensional visualization and data products," says Jack Dangermond, president, Esri. "By taking advantage of AVS's three-dimensional visualization technology expertise, we can offer through SDE the most advanced solution available for storing and viewing two-dimensional and three-dimensional data in a fast, efficient manner. Businesses can gain a competitive advantage by harnessing the power of our combined technologies."

ABOUT ADVANCED VISUAL SYSTEMS

Since 1992, Advanced Visual Systems has helped hundreds of global organizations turn data into visual insight through innovative visual information solutions. Advanced Visual Systems, an Oracle business partner, specializes in developing custom applications, and application components, for discovering and visualizing information stored in relational, multidimensional, and spatially enabled databases. In the face of today’s massive data explosion, products and services from Advanced Visual Systems are enabling organizations to turn their data in information, improve information access, and use information to make better business decisions.
Professional Services consultants at Advanced Visual Systems have expertise in a range of industries and data processes. They understand the application of visualization technology as well as problem areas inherent in specific industries. With this exceptional team of visual application development experts, and proven, multi-platform software tools, Advanced Visual Systems offers the fastest route to deploying powerful business information solutions. For more information about Advanced Visual Systems and its products and services, visit our website at http://www.avs.com, or contact Advanced Visual Systems’ U.S. headquarters by telephone at 781-890-4300 or by email at info@avs.com.