Complex, Intelligent Landbases: Essential for Enterprise Geo-Spatial Solutions
M.J. (Marv) Everett Geomatics Services Supervisor Survey & Photogrammetry Department - BC Hydro 6911 Southpoint Drive, 5’hfloor, Burnaby B. C., Canada, V3N 4X8
Introduction
“Enterprise GIS (EGIS).” How many times in the last few years have you heard that “so and so” company is going to expand their AM/FM system to an Enterprise GIS? How many utility companies are currently engaged in studies to identifi the costs and benefits of transforming their current AM/FM system to an Enterprise GIS? How many computer hardware and sofiware vendors are currently touting their products as the “ideal solution for your EGIS needs?” How many AM/FM/GIS consultants are offering their expertise to help you effect the transition from your “application challenged” AM/FM system to a fully functional Enterprise GIS system with all the attendant tangible, intangible and strategic benefits? The current interest in EGIS solutions is very high, and rightfully so. There are significant potential advantages to integrating all of a corporation’s gee-spatial data analysis and management systems into one homogeneous and application rich facility that is readily accessible to all levels of stakeholders. The fundamental factor that is going to facilitate this is not just new hardware or software or common data formats, or any of the other issues that are typically thought (or advertised) to be the key to your EGIS solution. Certainly these will help, but generally speaking these issues have already been (or are in the process of being) well addressed through conformity to different hardware and software standards and the availability of a number of data conversionhranslation software products. The two absolutely fimdamental issues that must be addressed before an effective Enterprise GIS can be implemented are:
The Enterprise Landbase Concept The foundation of any successful Enterprise GIS is an Enterprise landbase (ELB) - a single digital landbase that contains all of the gee-spatial features and topology necessary to satisfy the fill spectrum of Enterprise AM/FM/GIS applications. The concept is simple. However, a needs analysis will quickly reveal the broad spectrum of gee-spatial data and data relationships (topology) that must be modeled and constructed in order to satis~ the full spectrum of EGIS applications. All of the necessary gee-spatial data richness cannot be secured from a single digital map source. There is no choice therefore but to search for other data sources (both internal and external) and look for ways to integrate these disparate sources into one homogeneous landbase. Also, vector data alone will not satisfy all corporate stakeholder expectations and application needs. Rasters in the form of digital orthophotos and satellite imagery have become extremely popular for a number of utility GIS applications and are therefore an essential component of any ELB. Finally,an ELB comprised of 2-D data only will severely constrain the functionality and utility of an Enterprise GIS. It is essential therefore that ELB planners and developers accept the fact that the world truly is a 3-D place and design their landbases accordingly. A comprehensive ELB is the essential foundation of any Enterprise GIS. It is the component that will consume up to 50°/0 of EGIS implementation budgets to develop and therefore deserves considerable planning and attention. It is also the component that determines what applications can be run and therefore what benefits your EGIS can achieve. Finally, it is the component that, with an on-going maintenance program, will stand the test of time. Hardware and software will be replaced in 3 to 6 year cycles as technology advancements offer more effective solutions. However, a well-constructed ELB is a corporate asset that will provide benefits for decades. Geo-Spatial Data Standards It is widely accepted that the three primary keys to successful real estate development are location, location, and location. Similarly, the three primary keys to successful Enterprise landbase development are standards, standards, and standards. The implementation of an effective ELB requires the establishment of a set of corporate gee-spatial data standards that will satis~ the fill spectrum of identified EGIS applications. Data standards will determine how successfi.dly you are able to import spatial data from other sources (both internal and external) into your ELB and utilize them to your advantage. They will also determine how successfidly you are able to export your EGIS data to other agencies (local governments, provincial/state governments, etc.) to foster good working relationships and potentially reduce EGIS operational costs. Gee-spatial data standards can be broken into three primary components:
Not too many years ago, the only barrier to gee-spatial data sharing was considered to be data format. “If only we could establish a common data format, the world of GIS would be a much happier place.” Although the universal data format never materialized, some excellent format translation software was developed which now makes it easy to move gee-spatial data to and from numerous data formats. However, we have found through experience that this is only a partial solution. Almost without exception, we find that when we import spatial data into our landbase there are extensive spatial disparities - some of them subtle, some of them extreme, but almost certainly none of them are consistent in either magnitude or direction. In response to this reality, most GIS vendors were quick to provide “warping” or “rubber sheeting” fi.mctions that would “normalize” these disparities and deliver a spatially transformed composite. Unfortunately, these “transformations” were usually done without the benefit of an absolute spatial control framework. Furthermore, the objective was usually only to achieve good facilities data correlation. Therefore, the results rarely yielded any true spatial improvement in the underlying landbase. Consequently, most landbases remain as “spatially challenged” islands with few opportunities for either “inter” or “intra” Enterprise data sharing. When designing an ELB, it is imperative that suitable resources be allocated to establishing appropriate spatial accuracy standards - both absolute and relative. Furthermore, it is important to ensure that these accuracy standards are achievable - both from the cost/budget perspective and from the technology/delivery perspective. A critical point to appreciate is that not all land features or objects in an ELB need to enjoy the same level of spatial accuracy. Mixed spatial accuracy is highly desirable from a cost point of view. Why build an ELB to +/- 0.25 metre accuracy throughout if this accuracy is only required for 5’XOof the landbase area? Mixed spatial accuracy offers the added advantages of minimizing the size of your ELB and maximizing EGIS performance. Furthermore, it is highly unlikely that the fill spectrum of requisite land data will be available for an entire service area at a single accuracy standard. This being the case, mixed spatial accuracy in an ELB is essential. Consider a case where the established minimum accuracy standard for an ELB is +/- 10 metres (absolute and relative). Digital topographic mapping compiled to a 10 metre spatial standard is available for the entire service area and is utilized as the minimum base for the ELB. In addition, digital mapping of a high voltage electrical transmission line corridor which traverses the service area has been compiled to a +/- 0.25 metre accuracy standard to satis~ critical encroachment and conductor clearance analysis applications. This mapping is integrated into the base mapping and adds significant utility to the ELB. The +/- 10 metre base mapping alone could not support the encroachment and clearance applications. On the other hand, the +/- 0.25 metre corridor mapping is very right-of-way specific and can’t support an application that requires downstream tracing of fish bearing streams to identify where herbicides applied to vegetation on the transmission line could migrate. Together, these mixed accuracy gee-spatial data satisfy multiple user needs. A mixed accuracy ELB offers clear advantages to an Enterprise GIS. In order to be completely effective, EGIS users must be able to readily determine the spatial accuracy of the features or objects that are analyzed for specific applications and queries. It is essential therefore that comprehensive meta data be maintained as a component of the ELB and that EGIS users have ready access to this critical information. For EGIS platforms utilizing relational database technology, a separate meta database would likely be established. For object based EGIS platforms, meta data would be a component of ELB object definitions. Over the past several years a great deal of effort by numerous organizations and agencies has gone into the development of geomatics standards. Before reinventing the wheel, it is highly recommended to undertake a thorough search of established geomatics standards that will satisfy your corporate requirements. Generally speaking, it is too much to expect to find a standard that provides a perfect fit, especially with large scale features. However, there is almost certainly a provincial, state, or federal geomatics standard that will provide an excellent framework for your corporate standard, which can be enhanced to include your specific unique requirements. This approach offers two distinct advantages:
The land and facilities features of the ELB needed to satisfy the broader application requirements of your Enterprise GIS will vary in type and quantity for each individual system. However, one thing is absolutely certain - the new ELB will be at least an order of magnitude larger in numbers of features and topological complexity than any legacy AM/FM system. The old parcel-based AM/FM land model that was used primarily as a facilities backdrop will be replaced with an ELB comprised of a rich complement of cartographic objects including a detailed cadastre, planimetry, hypsography, hydrology, cultural features, natural resources, rasters, and facilities. The challenge will be to find economic sources for these data that also conform to the spatial standards that you have established. Assuming that you have developed suitable data richness, the controlling factor for EGIS applications will be the level of topology that is built into the ELB. Topology is the definition of feature or object relationships. Without topology, ELB data has no “intelligence” and is nothing more than a digital representation of a hard copy map. In the legacy AM/FM world, the emphasis was usually on the facilities model and a great deal of effort was put into building the requisite “network topology” that would facilitate circuit traces and other electrical connectivity analyses. Relatively little effort was put into building an intelligent landbase. In the EGIS world, the facilities model will continue to be very important. However, a very high percentage of new applications will depend on the ELB for data. Without well-defined topology the ELB will constrain new applications and the potential benefits of the Enterprise GIS will be severely limited. The Data Model All of the ELB data and their attendant relationships, behaviors and qualities are ultimately defined in the data model. This is the interface between the complex spatial data of the ELB and the various specialized application programs that will make queries, undertake complex analyses, and deliver the answers which provide the benefits of an Enterprise GIS. Since the data model controls the success or failure of any GIS, it is imperative that the EGIS development team ensure that adequate resources and skills are allocated to this development task. This is an activity that requires a resource team with a very clear understanding of both current and fhture stakeholder application needs. If this knowledge isn’t readily available it must be secured through user interviews and other information gathering processes. The data model for virtually any EGIS is very complex. It is highly unlikely that even the most talented, enthusiastic development team is going to get it 10O% Acorrect the first time (or even the second or third). Furthermore, over time parts of the data model are likely going to have to be changed for any number of reasons. It is imperative therefore that the selected EGIS platform provides good tools for developing and maintaining the EGIS data model with minimal effort and with minimal impact on daily EGIS operations. How many dimensions are enough? Almost without exception, legacy AM/FM systems provide 2-D solutions only. A survey of legacy AM/FM system users would almost certainly reveal similar answers. “We don’t need 3-D, we are only interested in electrical connectivity and the relative XY position of our facilities to others.”” Our Trouble Call Management application doesn’t need 3-D.” There are several reasons for this:
It is important to understand what is meant by 3-D data. Some people think that if you drape your 2-D land and facilities model over a DEM (digital elevation model), you will have a 3-D model. In fact, what you will have is simply a 3-D rendering of your 2-D land and facilities base which you can view from some perspective to allow better visualization of an area of interest and facilitate determination of ground surface elevations at selected XY locations. Certainly this provides considerable advantages and the writer would be the first to argue that the DEM adds important hypsographic information and should be a standard component of any ELB. However, if your land and facilities data is only 2-D, you cannot, for instance, determine the elevation difference between a geographic surface position and the transmission line conductor that passes directly overhead. How close is the conductor to the ground? What is the EMF rating under the conductor at that geographic location? What is the level of public safety at this point? What potential vegetation management problems prevail here? What multiple use can be planned for this area? By adding the third dimension to your ELB data, you not only add considerable utility to your EGIS, but also add significantly to the value of the benefits that your EGIS can deliver. Time is another dimension worthy of serious consideration. It also has the added advantage of being quite inexpensive to include. As with other meta data, time can be provided in different ways depending on the EGIS platform. Ready access to temporal information can add significant value and fimctionality. A simple example is being able to determine that some specific data that you are looking at in your EGIS is ten years old. This knowledge could influence how you use that information. A more complex example would be to reconstruct a component of an electric transmission system and adjacent lands as they were five years ago and compare them to how they are today. This would provide an extremely powerful and effective method for monitoring changes and how these changes are influencing your transmission line management methods and costs. New Skills for new Challenges For most corporations, the Enterprise GIS concept with its hardware and software challenges, requisite gee-spatial data complexities and sophisticated application development requirements is being driven by a number of factors. Internal factors include such issues as changing corporate policies, downsizing/rightsizing, knowledge attrition due to early retirement incentives, etc. External factors include electrical deregulation, changes to provincial/state and federal regulations, regulatory body policy changes, etc. Regardless of the specific driver(s), one thing is clear - Enterprise GIS systems are far more complex than earlier AM/FM systems and therefore require broader planning, greater stakeholder involvement, and a broader skill set to ensure an efficient and effective implementation. Geomatics is the vocation of compiling, manipulating, and managing gee-spatial data. It includes such disciplines as survey, photograrnmetry, remote sensing, computer science, and cartography. Since much of the success of an effective EGIS implementation depends upon a well designed, feature rich Enterprise landbase, these are essential skills for the development of an effective ELB and should be employed extensively by EGIS developers. Although today there are no real technical barriers, the task of implementing an effective EGIS can be quite daunting. There will certainly be significant organizational and logistical problems to overcome. Furthermore, there will be many technical issues, which if not addressed by people with the appropriate skills, will put your EGIS project at considerable risk. As noted earlier, the development of an effective ELB to support the full spectrum of EGIS applications can represent up to 50°/0 of the cost of an EGIS project. It is also the component which has the most influence on how well the finished EGIS will perform and the value of the benefits that it will deliver to the corporation and its stakeholders. It is clear that the employment of experienced geomatics resources to assist with the development of appropriate gee-spatial standards and the design and implementation of the ELB will ensure satisfactory project deliverables and significantly reduce unnecessary risk. The Roll of the Vendors To a large degree, the rate at which comprehensive EGIS implementations evolve, is controlled by the vendor community. Until recently (the past 18-24 months), neither the hardware vendors nor the software vendors had a sufficiently rich set of products to facilitate the development and implementation of comprehensive EGIS solutions for large, geographically dispersed utility organizations. Today the basic computer hardware, operating system, database technology, and communication tools are sufficiently well developed and mature to support effective Enterprise GIS solutions. Much of this is the direct result of the excellent cooperation between the vendor community and different standards organizations such as OSF, OGIS and IEEE, to mention a few. Both the vendor community and the client community clearly benefit from this and it is reasonable to expect that this positive trend will continue. In the opinion of the writer, the largest outstanding area of technical development to be addressed is the responsibility of the GIS vendors. The true test of any EGIS implementation will be how easily and intuitively stakeholders from all levels of a corporation, with all levels of computer skills, are able to access information and achieve benefits. Like it or not, Windows 95/NT has raised the level of user interface expectations. Almost over night the average corporate computer user has come to expect this environment as a standard and is quick to discredit (and reluctant to use) any software product that doesn’t conform. In addition, there area number of other less visible but critical issues that the GIS vendors must address. Because an EGIS environment has the potential for a much broader spectrum of applications, it is imperative that the vendors provide powerful tools for application development and maintenance. Underlying this is the need for more sophisticated CASE tools that will facilitate easy development and maintenance of very complex data models. Full cartographic quality graphics are required to represent complex topographic features using symbols and line types that are consistent with established geomatics standards. Support for easy import and export of data from different sources and formats (including rasters) is critical. Also, sophisticated tools for spatial transformations between different map projections and datums, and support for rigorous, non-linear adjustments of spatial disparities are essential. Finally, a single vendor EGIS solution will require a platform that provides equally rich functionality for both network and spatial applications. Summary To twist an old adage a bit, “behind every successful EGIS there is a good ELB.” This will never happen by accident. It requires good planning and the employment of appropriate specialized skills and services. This investment is essential. Without a comprehensive ELB, an Enterprise GIS will be highly constrained and will not yield maximum benefits. Conversely, a well constructed, data rich, topologically complete ELB will provide the requisite foundation for an Enterprise GIS. This will satisfy the full spectrum of corporate application requirements and yield maximum tangible, intangible and strategic benefits. | ||
|
|