Take, for example, Preetha Pulusani, vice president for Intergraph Corporation's Government & Transportation Systems Division.
"I am seeing a growing demand to get this technology out of the back room and make it available to a variety of users within the organization. The direction is toward easier interfaces for non-GIS users and integration into the wider enterprise-information system," Pulusani said. "Organizations are asking for scaleable software from point to departmental, to enterprise solutions, without having to retrain their workforce or throw away the system they already have. We have been seeing this trend in the market for the past four years. GIS is coming of age."
Joe Astroth
vice president, AutoDesk
Astroth: No longer as the center of the universe -- those days are long gone. That's the "Copernican view" of GIS, so to speak. Now, we're in a post-Copernican era, where GIS melts into the IS/IT infrastructure. That's not to say it loses its importance; it will just become more transparent. It's no longer necessary to have a stand-alone GIS on a desktop or laptop; the application, the heavyweight, can reside on a server. It can part be of an organization's Internet and intranet platform strategy, acting as a plug-in to a browser, extending its reach throughout the organization, even to palmtops in the field.
One example of this is Northwest Water, one of the largest water utilities in the U.K. They are using geospatial technology software to automate, integrate and visualize their customer-information and network-management
systems. At the command-and-control center, just outside of London, a 25-foot-by-10-foot screen
displays a base map of their area and the locations of field crews and incoming customer calls. Other layers include the organization's infrastructure. On the same screen, at the same scale, is a realtime radar weather map. Since the utility also deals with storm-water runoff, supervisors must anticipate problems that may result from heavy thunderstorms and get service personnel into the area ahead of time.
The whole system is managed by an Oracle database, running what I call "geospatial middleware." In addition to modeling and representing the network, the system provides the visual information needed to solve problems as they arise. Again, GIS is not any less important here, just less visible as an independent system. It enables the staff at Northwest Water to visualize and model infrastructure and operations in a way that was much more difficult to do using loosely connected, one-off proprietary systems.
GT: Are advances in technology the primary forces underlying these changes?
Astroth: I think it's really a change in attitude by the GIS vendors, unlocking the real power of GIS by giving up trying to own all the data in a proprietary GIS system. It's the view that says GIS can become part of every system out there, from the box that helps you find the best Chinese restaurant in your neighborhood to the utility that gets your lights back on faster because they can visualize the network and the location of their service personnel. It's GIS melting into the overall IT infrastructure, rather than the other way around. It's a shift in attitude.
GT: What will happen to corporations that have built their identity on GIS? What will happen to corporations strongly identified with GIS as a distinct technology, rather than as a transparent application in a larger infrastructure?
Astroth: The harsh reality is that they will either have to change and assume a smaller role, or fade away. This has happened before as technology advances. Railroads are a good example: They chose to define their role far too narrowly by saying they were in the railway business rather than in the transportation business. The rest is history.
Carl Reed
vice president, Intergraph Corporation
GT: What will be the long-term impact of the Internet on GIS applications?
Reed: The Internet is going to radically change the way people interact with spatial data. Traditionally, GIS required trained technicians. In three to five years, people will be able to sign up for geospatial-technology services from an Internet service provider, just as they do now for Internet access. Many users won't be GIS specialists. They won't care or need to know how the technology works, any more than we care how lines, satellites and switches route our telephone calls.
The Internet will have a similar effect on the geospatial process; users won't care how the geospatial information is managed. They won't really buy licenses for GIS software, they'll just sign up for a service and pay by transaction or by a monthly service fee. The whole geospatial stuff will be demystified, consumed by larger IT industries and become much more universal. GIS as a unique industry will disappear. A corollary to this is if GIS vendors don't recognize this, they could go out of business.
Of course, some people will still be interested in how these things optimize work flows, how the technology can be used more effectively in existing resources and manpower and how it can be used
by the community and education. The emphasis will perhaps shift away from technology, toward social factors associated with using geospatial data.
GT: What other technologies will be involved in this transformation?
Reed: One of the underlying developments is the ability to store spatial data in a relational database. Right now, Oracle's Spatial Cartridge is probably the leader in terms of what it can do -- manage data, provide security and back-up measures and transaction processing. Data is much more accessible and easier to use stored in a single, consistent repository rather than scattered on desktops around the organization. Let the database manage the data. The move toward storing spatial data in a relational database will be a major contributing factor to the wider
use of geospatial data.
Another part of the equation is bandwidth. The Internet can be fairly slow. Much of that has to do with limited bandwidth. This is why the communications industry is madly laying new fiber. We have the technology to use cable-TV infrastructure to carry Internet signals. Now it turns out there are unused bandwidths on power lines that can be used to carry information for the Internet. There is also Internet2, capable of hundreds of megabytes per second.
People will want more multimedia capabilities via the Internet. The same capabilities will be important for GIS applications; users will want to view digital orthophotos, run video clips or take
a virtual drive around the city to capture infrastructure information. A whole new generation of technology and infrastructure is being put in place for the Internet of the future, and bandwidth is very important for this.
GT: What new or expanded applications do you see for GIS?
Reed: It's becoming apparent, especially at high political levels, that the use of geospatial data is critical to more effective development and implementation of public policy. This is something Al Gore is very keen on. The whole concept of livable communities, digital government and the national spatial-data infrastructure are all coming together at a very high political level, not just in the United States but in other countries. We are using geospatial data to develop policies for environmental protection. For example, the Global Disaster Information Network is an international effort to develop partnerships, infrastructure and procedures for sharing global disaster information across international boundaries, and to more effectively deploy resources in response to disasters.
Many countries are collaborating in the development of this network. If there's a disaster in the world -- forest fires in Indonesia or South America, earthquakes in Iran, tidal waves in New Guinea, hurricanes and flooding in Honduras and Nicaragua -- there will be more effective utilization of geospatial data available in responding to that situation.
Jean-Baptiste Monnier
vice president, Geoengineering Business Group, Bentley Systems, Inc.
GT: What developments do you see changing the way we work with GIS?
Monnier: Two trends converging and merging: the Internet and storage. GIS geospatial information is not stored in proprietary formats anymore, nor in files. The enterprise stores all of its data, both raster and vector graphics, within IT warehouse systems. Users can now go directly to a piece of information, a street number, a specific manhole and 20 meters around it, 50 meters around a specific gas valve, service line or all the parcels within a mile of a particular road. It's no longer necessary to download an entire file to access a piece of information in it; the user can go right to it. We're seeing demand from every level of government to provide this type of storage capability.
Instead of storing spatial data in proprietary files, they are being stored in the new relational databases that handle business and tabular (aspatial) data as well as spatial data. Oracle Spatial database offers a set of four tables in the core database, along with extended query mechanisms and a graphical user interface that provides access to the data without the need of proprietary products to interpret that data. One of the tables is an index, which enables users to make spatial queries and pull information out of the database.
Users can zoom in to a very high level of detail, and the server will know where the browser is and provide a resampling at the desired resolution. While the user is looking at a particular area, the server anticipates a zoom or pan, and automatically fills the buffer with a few hundred meters around that area, expecting a zoom in or pan of that area. The relational database can anticipate the user's move. In much the same way, Bentley's Imager provides raster information from a central raster warehouse. All of this allows for the new, more centralized architecture.
With the infrastructure to provide raster and vector data to everybody in the enterprise, far more information can be placed inside a graphic than was possible in the past. We used to serve corporate databases with SQL (Structured Query Language) alpha-numeric queries. I think it's easier to serve a government database through graphics, accessed by zooming into the area of concern, bringing up different layers then clicking on information and getting additional attribute queries.
So the map becomes the browser, the natural interface to start a query from many different accesses. I think we're still in the infancy with this type of navigation because people haven't populated maps with the proper information. Maps are still in the GIS department, but with easy access and the enhanced SQL, we'll get there.
David Maguire
director of products, ESRI
GT: How do you see GIS developing over the next few years?
Maguire: The technology is polarizing into two systems: desktop and server or network-based GIS. Desktop systems are able to run many GIS functions and they can be self-contained or operate in a client-server fashion in conjunction with a
GIS application server or data server. The
network-based GIS is dominated by applications running across the Internet,
typically, with a thin client running inside an Internet browser on the client's desktop and the data and smart GIS functions running on a server.
The technologies driving these developments are clearly the Internet and open-network productivity using Internet standards and TCP/IP open-network standards. It is also the development of component-based GIS for breaking monolithic systems into small component building blocks. I think Java and Microsoft's COM (Component Object Model) are the emerging standards here. On the server side, it
is the ability to store and manage spatial data inside relational database-management systems, a technology all major GIS vendors have embraced.
Mobile computing is another trend. We're already seeing the evolution of palmtop GIS and other portable devices with embedded mapping. As small, lightweight devices become more powerful, the evolution of the Windows CE operating system and Java coupled with the splitting of GIS into small components makes it possible to run a basic mapping capability on everything from cellular phones to field-based data-entry devices and palmtops. I think mobile computing is a key trend. As the workforce becomes more mobile and scientists move from the desktop to the field with these devices, there will be an increasing demand for GIS.
GT: What is driving this demand?
Maguire: It's coming from several areas: utilities that want to do field-based data entry, and foresters and soil scientists who want to get out in the field and do real science out where the action is, not back in the classroom or the laboratory. There is also a demand from transit agencies for remote passenger-recorder systems to record the number of passengers boarding and alighting from buses and trains. Then there are business professionals who want maps of local areas, hotels and restaurants.
GT: Will GIS maintain its identity through all of this change?
Maguire: I believe so. GIS is significant enough as a technology and as a science. There is no doubt, however, that mapping technology will be integrated much more closely into mainstream IT. We see that now in enterprise resource planning. Many companies, including Microsoft, Visio and SAP, are now incorporating mapping capabilities into their standard software products. Distributive GIS is going to happen, initially as a parallel to desktop GIS, maybe more than that in the long term, and definitely more cost-effective.
John Haller
chief technology officer, MapInfo
GT: How will evolving technologies today affect the use of GIS?
Haller: We don't really call it GIS because the term says big, high-end applications, isolated systems sitting in back rooms, typical land-use and asset-management applications. Instead, we use "spatial applications."
Since spatial data can now be stored in name-brand relational databases such as Oracle Spatial [Cartridge], IBM's DB2 and Informix IUS, they all treat spatial data accessed and manipulated in much the same manner as they do numeric and string data. This is a huge step forward because these relational databases enable us to manage spatial data in standard ways without having to use proprietary formats.
Also, with the old hurdles -- proprietary formats, high cost and data availability -- disappearing, spatial-data
management is becoming mainstream. Open standards and relational databases provide access to nearly all applications. There is so much data available today that scarcity is hardly an issue. As for complexity, that will be handled on the server side via Internet or intranet.
GT: What will these changes mean for the industry?
Haller: Our company and other companies need to start changing direction and start offering products and technologies that will apply to this new market, to server-side spatial applications. The demand from all our big customers now is for Internet use of these applications. We are currently building Java-based services and Internet tools that will allow users to do map publishing, geocoding and different types of spatial analysis -- all on the Internet. Since these applications must be scaleable, this requires all new technology.
Robert Moses
president and CEO, PCI Geomatics
GT: What are the new developments in remote sensing, and what effect will they have on GIS applications?
Moses: We are going to have several panchromatic, one-meter resolution satellites that will also carry four-meter, multispectral resolution capabilities. These new data sources are going to change the perspective of how we do geospatial processing. In the next two to five years, GIS will become more image-processing oriented and calculations will be done on the imagery in the computer without first having to make raster-to-vector conversions. The process of setting up and creating information products will be faster and more cost-effective than it has been in the past.
GIS will change from being primarily a vector-processing-oriented methodology encapsulated in vector-processing software systems to a raster-processing methodology encapsulated by image-processing companies.
Next year, there will be hyperspectral satellites carrying 200 chromatic bands, enabling us to further automate classifications of natural resources. That will allow us to more effectively determine the health of trees and crops. With 200 chromatic bands, we will have access to much more information than is available to us today. All of these satellites with radiometric and geometric resolution are going to give a huge impetus to a different source of data than was available to us in the past.
GT: How are classifications automated?
Moses: By comparing spectral signatures captured by satellites with the signatures stored in databanks, we can automatically classify trees, crops and other plant life according to type, stage of growth, health and nature of disease. The process will definitely reduce the need for ground-truthing. First, however, we must build up a spectral-signature databank by photographing crops at various stages, much as the USGS did for hyperspectral rock types.
GT: What other applications do you see coming from high-resolution satellite imagery?
Moses: Today, geospatial data makes up about 75 percent of the world's commercial mapping, most at 1:25,000 - and 1:50,000 - scale resolution and well within the capabilities of one-meter satellite imagery. Products derived from this imagery will be used in engineering, mapping, resource evaluation, urban planning, natural-disaster mitigation and reaction, environmental monitoring and defense-intelligence mapping. Governments will be able to use satellite imagery to set up their entire national base maps, and do predictions and thematic and topographic mapping straight from a 3-D raster. One-meter satellite resolution will give us access to a variety of information that, in the past, required aerial technologies.
GT: What other applications do you see evolving from these developments?
Moses: Right now, maybe 15 percent of the GIS market uses satellite imagery. Of that group, about 90 percent use it to update various processes, such as backdrops or foredrops for vector databases. I think more than half of that market, especially in the Third World, will be using satellite imagery to set up their databases and do primary processing. Then, of course, they would do secondary vector processing as needed. The paradigm will definitely shift toward venue-data sources.
