GISdevelopment.net ---> GITA 1997 ---> Business Evolution & Platform Migration

Migrating to an enterprise GIS


Ted MacKinnon
BC Hydro 6911 Southpoint Drive, loth floor Bumaby, B.C. V3N 4X8

Robert Wong
Westech Information Systems Inc. 6911
Southpoint Drive, 10th floor Bumaby, B.C. V3N 4X8


Abstract
After spending over $30 million dollars on developing an AM/FM system for electric distribution, BC Hydro is looking into expanding their mature production system to an Enterprise GIS. The new Enterprise GIS would address corporate and different Strategic Business Units (SBU) business requirements. The proposed system has to deal with different hardware/sofhvare platforms, as well as different land base, spatial data standards and system interfaces. This paper presents the approach that BC Hydro followed to migrate to an Enterprise GIS. At time of writing, the Enterprise GIS Prototype is in the early development stage. A data translator has been set up and data from a selected geographic area is being tested. The results to March, 1997 from the Prototype Project will be presented at the AIWFM cofierence.

Background
British Columbia (B.C.) is the third largest province in Canada. It is located on the Pacific west coast, north of Washington state. It covers a territory of 948,900 square kilometres. By comparison, this is 2.5 times as large as Japan or 4 times as large as Great Britain. In rough figures the area of B.C. consists of 60?4.forested lan~ 357. barren alpine tundra, snowfields and glaciers, 4% lakes, rivers and streams and 0.594.urban and farm land. B.C. has a population of 3.85 million. BC Hydro is the 3rd largest electric utility in Canada. It is responsible for the generation, transmission and distribution of electricity in B.C. It serves over 1.4 million customers in the province. In the early 1980’s BC Hydro had a need for faster updates of their many paper map products. This led to the exploration of various Computer Assisted Mapping systems. After examining several systems that were available in the marketplace, IBM’s Geographic Facilities Information System (GFIS) was selected to provide automatic mapping/facilities management of BC Hydro’s electric distribution system. The initial prototype was one municipality (Bumaby) and it operated on a single workstation. BC Hydro now has a mature production AIWFM system that supports data conversion, quality assurance/quality control, mapping, facility planning, work order design and integration with their Pole Management System, and Work Management System. The AM/FM system is used by more than 350 service pkumers/engineers/work order staff. There are over 250 workstations throughout the province connected with 19.2 Kb and 56Kb lines, as well as 10 workstations connected to a T 1 Metropolitan Area Network.

In 1995 BC Hydro was re-organized to better position itself for a de-regulated marketplace. This brought Transmission and Distribution (T&D) together into one Business Unit increasing the need to have a common GIS that would meet T&D needs, at least. The 51,500 km of distribution primary is in the GFIS database but a GIS is needed to help manage the 18,000 km of transmission line right of way. The GFIS land base was only entered for areas in which there was distribution plant and there is only about a 15% overlap of land coverage between transmission and distribution lines. This will require considerably more land coverage to include transmission.

The Survey and Photogammetry Department has had projects to prepare large scale mapping of the provincial transmission line right of ways and this data also needs to be incorporated into the Enterprise GIS.

Needs/Opportunities
Although the existing AM/FM system has the fimctionality to support the operational needs of electric distribution, there are areas where improvements are needed. These include:
  1. Does Not Address Requirements of Other Business Units
    There are many groups within BC Hydro that require a spatial component to the data they already have, or they need access to the transrnission/distribution data. The current GFIS supports many of the needs of electric distribution but does not support those of other Business Units. Although there is a growing amount of transmissiorddistribution digital data, there is no support system that provides user access or analysis capability for this information.

  2. Old Technology
    The life cycle of the GFIS product line is coming to an end. It is no longer being enhanced and product support by the vendor is expected to diminish in the fhture. BC Hydro must plan for the eventuality of no GFIS product support.

  3. PerformancdOperating Cost
    The GFIS system is costly to operate and maintain. The product has little built in fi.mctionality. Large programming effort is required to modifi or extend system functionality. The system response time to users is slow. Although part of the problem is due to communication line capacity, the other factor is inefficient handling of retrievalhefresh data.

  4. Infl-”ble Data Model
    The GFIS product has several major deficiencies, including the following:

    • limited 3-D capabilities (Z coordinates can only be stored as an attribute).
    • limited polygon definition, maintenance, and analysis capabilities.
    • no distributed data base capabilities.
    • data model changes often require modifications to application programs and require data base conversions.
    • all data components must be input through a spatial relationship.
Solution
Given that the vendor is not going to replace or significantly enhance the GFIS products, BC Hydro wanted to protect their current investment by migrating the existing system functionality and data to a more responsive, fictional and cost effective environment. In addition, BC Hydro wanted to take a more enterprise approach to GIS. Instead of a replacement platform for their electric distribution system, the requirement is for a platform that meets many internal business needs (e.g. Distribution, Transmission, Properties, Environment, Aboriginal, Taxation, Power Supply, Customer Services, etc.) and enables easier ties to external agencies. An enterprise GIS study was commissioned by the office of the Chief Information Officer (CIO) in 1994. The study concluded that an enterprise-wide approach will be cost effective. The alternative is the ad hoc growth of applications and data bases with minimal opportunities for cost-sharing, standardization, and long-term business effectiveness.

Obstacles/Risks
There are several major obstacles/risks with the enterprise GIS approach. These are associated with:
  1. Data Accuracy
    The GFIS data was digitized from different sources that have different levels of accuracy. As a result, the absolute accuracy of the data is a concern. Relative accuracy of the data is adequate to meet the needs of electric distribution. However, higher levels of absolute accuracy are required to support the needs of Transmission and to utilize a common landbase.

  2. Multiple Landbase
    Many separate land tiormation sources are maintained by different groups within BC Hydro (Properties, Distribution, Transmission, Environment, Vegetation, Taxation, Aboriginal Relations, etc.) in many formats. While these groups have different data requirements, they operate in the same geographic area and share many of the spatial data needs. They need an easy and intuitive method to access the data. The land data model developed must be capable of meeting all these needs.

    A major concern is whether the different information sources can be assimilated using a common landbase. If there are significant positional discrepancies between the land sources, the different land sources may have to be kept in separate layers or the facilities/features data may have to be re-positioned with reference to the common landbase.

  3. Multiple Standards
    The data from different Business Units have different standards. Incorporating the different sources in an enterprise GIS will require data standards to be adopted (e.g. field names, sizes, date sequence, etc.). Getting agreement from the various Business Units, on the data standards maybe difficult and time consuming.

  4. Legacy Data Sources
    Moving the data to an enterprise GIS may eliminate some of the legacy data sources. However, some of the legacy systems are not well doeurnented. In the migration process, obsolete fimctions and data may be perpetuated or required fimetions and data maybe unintentionally dropped.

  5. cost
    The eat and benefits for the enterprise GIS approach are based on certain assumptions. There is uncertainty about the reality of the costs and the benefits. Before proceeding with a fill scale GIS project, a method is required to confirm feasibility, costs and benefits.
Approach
Before diving into an enterprise wide GIS that requires large expenditures, BC Hydro is proceeding with a prototype on a contemporary GIS platform. The advantage of this approach is that the key assumptions can be tested and verified with minimized risk and lower cost. The scope of the prototype is to develop several applications to determine how well they meet corporate and SBU business requirements. An evaluation will be made on benefits, costs (development and operating), and system fimetionality. The GIS prototype will be evaluated on the following areas:
  1. Common Landbase
    The use of a common Iandbase is key to many of the Enterprise GIS applications. The proposed system will need to show that it can assimilate gee-referenced data with other data sources into a single landbase and make the data available to satisfi a variety of applications.

  2. GFIS Data Migration
    The existing GFIS land and electric distribution data will have to be migrated to the new platform. A small geographic area with representative facilities/features will be used in the prototype. The proposed system will need to demonstrate that data migration can be accomplished effectively and efficiently.

  3. Transmission Line Right-of-Way Applications
    Besides the fimctionality to support electric distribution, the proposed system will need to show applications that support managing right-of-way issues. These will include property rights and asset management and utilization of complex polygons in managing vegetation, environmental and habitat programs.

  4. Out-of-Box Functionality
    The contemporary GIS platform will need to show that more built-in fimctionality is available than in the current GFIS soRware. The prototype will evaluate system functionality related to enhanced display and reporting features, data version capability, network analysis features, polygon analysis features, data model flexibility, CAD features and ease of use.

  5. DevelopmentiOperating Costs
    The proposed system will need to show that it has lower development, maintenance, and operating costs than the current GFIS software. Costs will be collected and compared with those associated with the existing distribution system. The review process will also re-examine and verifi previously estimated effort, benefits and cost of migrating to a new platform.

  6. System Interfaces
    The technical issues and benefits of integrating GIS with other BC Hydro information systems and external systems will be examined, Also, the technical issues and benefits of utilizing portable GIS viewing systems in the field will be evaluated,

  7. Implementation Plan
    Demos of the GIS prototype modules will be provided to various departments across the company to assist them in better quanti~g the benefits. Based on the associated costs and benefits, a prioritized implementation plan will be developed based on the best business cases. This will incorporate an RFP and evaluation for the sofhvare and hardware requirements.
Results
Budget approval for the Enterprise GIS Prototype Project was received in October, 1996. The prototype software was obtained in November. Key user staff members have attended training on the vendor products, Additional workstations were in place during December. The sofhvare and hardware was obtained on a lease arrangement. The GIS prototype modules are scheduled to be completed by March 1997.

The project plan for the GIS prototype will include the following tasks:
  1. Design a suitable land data model
  2. Establish gee-spatial data standards.
  3. Build the modules to demonstrate the required system functionality.
  4. Measure the costs in the process. Identi@ the direct measurable benefits (e.g. development and maintenance savings, hardware usage savings).
  5. Demonstrate the fimctionality of the prototype modules to various user groups so they can accurately access their potential benefits.
  6. Assemble a project completion report.
  7. Based on a positive result, develop an Enterprise GIS business case and an implementation plan.
At time of writing, the Enterprise GIS Prototype is in the early development stage. The underground electric distribution data model and land data model are being finalized. A data translator has been setup and data from a selected geographic area is being tested. The results to-date from the Enterprise GIS Prototype will be presented at the AM/FM conference.

© GISdevelopment.net. All rights reserved.