Data modeling for GIS

A pipeline data model is the basis for any pipeline GIS system. Robert Brook outlines the different options.

A pipeline data model provides the structure for data storage, maintenance, and use, and it also provides the environment for solutions, interoperability and enterprise integration. This makes it the key to any pipeline GIS system. This need has become predominant as pipeline operators’ regulatory and integrity requirements increase and they implement new technologies that make more data available. It has become vital that the data model suit the needs of an operator.

Models and templates
When implementing a data storage system, an operator needs to understand the similarities and differences between data templates and a data standard. A data standard defines the data: what should be recorded, how it should be recorded, and how it should be supported by a system in order to retain its full meaning. A standard is an established format.

A data template establishes a framework that can be used to collect and store information. It is usually created for a specific purpose but it can be used as a best practice to collect information beyond its original scope. While a template can be used rigidly to provide consistently, it is meant to provide flexible.

While each of these options can provide an operator with different opportunities, the selection of the appropriate type should be based on the organizations data requirements, work flows, processes and culture. Currently two models are at the forefront of the industry, the Pipeline Open Data Standard (PODS) and the ArcGIS Pipeline Data Model (APDM). The PODS model is relational, allowing vendors and software companies to provide different spatially enabled solutions. The APDM model is an ESRI Geodatabase template.

Similarities and the differences
APDM and PODS are not competing models; they simply offer operators options for meeting their data modeling needs. Both of the models have been specifically designed for the pipeline industry and each is appropriate for use in either a transmission or gathering environment. Both models can be utilized by a GIS system and each organization subscribes to the philosophy that an operator should be able to select from a group of appropriate vendors to provide them with a solution. In addition, they share a fundamental belief that any operator should be able to integrate products from more than one vendor into a unified solution for their organization.

While there are similarities between the two data models, there are also considerable differences. PODS is an open industry standard that defines in rich detail the features that describe a pipeline. The entire PODS table structure is intended to be implemented ‘as is’ and all the tables comprise the standard. Any operator who implements PODS must have a good understanding of their business needs; because the model is a completed standard, it could dictate how the business operates rather than allowing the business to dictate the model.

APDM is an ESRI template with a standard set of core table. The standard core must be implemented as described in order to promote interoperability between database implementations and in order for ESRI business partner software to function. The remainder of APDM model is a series of best practices that are optional. These best practices are essentially a set of rules that define how an additional feature class should be created, and are called the APDM abstract classes. APDM provides an operator with a lot of flexibility but it is open-ended. An operator needs to understand what they want to store and why. Without assistance it can be difficult for operators to clearly understand the models requirements, define their specific needs and implement the model.

While the PODS model attempts to describe in detail all the features that define the pipeline the APDM model seeks to describe the ‘response’ or ‘behavior’ of pipeline features as they are edited, or more importantly as the underlying centerline is edited. Both organizations provide support for their model.

Linear structure
A gathering system can be stored and managed from either model, but with PODS gathering system the pipeline must be represented by measured features. In APDM either measurements or a network could be used. In either case linear referencing and absolute co-ordinates are the primary means for locating features.

APDM is an ESRI Geodatabase, and therefore it uses ESRI’s linear referencing and topology data structures, and it is specifically designed for use with ESRI’s ArcGIS technology. Essentially it has GIS ‘built right into’ the model. Its implementations are relatively standard and since it is constructed for ESRI’s applications, deployments involve the same types of technology. PODS on the other hand is built to run in a relational database management system. Therefore, other than storing linear referencing and co-ordinate information in tables, there is no embedded GIS functionality within PODS. This establishes the openness of the PODS model as it gives pipeline operators the ability to choose their own GIS application. Thus PODS has a broad spectrum of implementation styles, uses multiple GIS platforms, and allows for various levels of GIS integration. This being said, many PODS users have also selected ESRI software.

ESRI has selected the service oriented architecture (SOA) model for enterprise integration as outlined by Microsoft and APDM follow suit. The SOA provides a modularity of pipeline business logic, which can be presented as a service for other enterprise clients to consume. These services are loosely coupled where the interface exists in the specific enterprise application and remains completely independent of the service layer. APDM does not support tight integration because it can be hazardous to the integrity of the geodatabase.

Where APDM has a single enterprise integration strategy, PODS provides multiple options. An operator can choose to use the SOA methodology and deploy services through the selected RDBMS client using that vendor’s technology and a spatialization tool. Alternately PODS users can select a more traditional model where information is tightly coupled using foreign keys or related records.

Operators who select PODS are generally companies who are looking for a standard model that is managed using relational technology. They feel that pipeline operators are performing the same business function and should do it in a like way. When an organization selects APDM, they have made a decision based on the desire to manage information spatially. In addition, they see the value in flexibility and through the use of a template they store information in a way that best suits their organization.

On the horizon
While today operators only have a few models to chose from there may be another option looming on the horizon. PODS and APDM are currently working jointly to develop a spatial implementation of PODS using ESRI technology. While PODS will continue to work with other spatially enabled software products, the intent of this project is to merge aspects of each model into a single model that calls upon their individual strengths, thereby creating the PODS ESRI Geodatabase. The concept of this model was proposed in early 2007, but it is still in its formative stages.

Robert Brook is ESRI’s Pipeline Industry Manager. With over 19 years of experience he has worked with many major national and international pipeline operators. He is currently Chair of ESRI’s Pipeline Industry Group, a member of the ArcGIS Pipeline Data Model steering committee, and Chair of the PODS New Construction Working Group. He can be contacted at the Environmental Systems Research Institute (ESRI), 380 New York Street, Redlands, California 92373 or at rbrook@esri.com.