Data and how it’s organized can impact all aspects of an organization’s mandate. As such, a decision to adopt a new data model, regardless of commodity, can be daunting. There are many far-reaching consequences as a result of decisions made during a data modeling workshop. Within the utility industry, this could include things like regulatory compliance, resolution of service outages, service territory expansion, and threat mitigation.  The intent of this article is to provide a set of commodity independent, data modeling best practices.

Preparation

It’s vital to the process to have the tools and access necessary to effectively conduct a data modeling workshop. Prior to the workshop, the following items should be addressed:

• A decision on which target model to use as a baseline.
• A document of record should be generated from the chosen base target schema. This document should be in a format that is granular enough to detail the lowest level of object characteristic/behavior including classes, subtypes, fields, relationships, domains, default values, etc. that comprise the various elements of the database. The format should be such that it can be easily edited and distributed.
• A representation or copy of the legacy database/schema should be made available in a format that can be analyzed in advance. This will allow the modeling lead to identify potential issues that can be targeted for discussion.

Attendees

Data modeling is a tedious, in-the-weeds process where every element of the source and target databases must be evaluated. Depending on the commodity, this effort will take days or weeks to accomplish. As such, it’s ideal to limit attendance to key personnel to avoid distraction and extended unnecessary discussion. Key personnel include:

• Utility Side
  • Individuals familiar with the legacy system including its design, maintenance, and utilization
  • Individuals that have “final say” regarding schema changes
• Consultant Side
  • Utility GIS SME to lead discussion
  • Scribe to document model changes, action items, and key decisions

Environment

This refers to the space and tools utilized during the workshop. It’s imperative for a successful workshop to have the following in place:

• Ability to access the legacy system and display via projector.
• Ability to access a representation of the target system and display via projector. This should include the base target schema as well as sample data so behavior unique to the target system can be demonstrated.

A white board to illustrate examples and talk thru options. Photos of the whiteboard illustrations should be captured as a reference.

Workshop Kickoff

Conducting an effective data modeling workshop begins with educating attendees. After introductions, start with a comprehensive presentation detailing the following:

• Review workshop deliverables
  • The completed document of record
  • Accompanying document detailing key decisions and how they were reached
  • Action items document detailing outstanding issues and who is responsible for resolution;
• A demo of the target system highlighting functionality that represent the “hot button” features that led them to choose the solution
• An explanation of terminology unique to the target system; A single page handout is recommended for attendees to reference throughout the workshop
• An explanation of how configuration and properties drive behavior of target system
• An overview of the data dictionary for the target base data model; Distribute a copy to attendees
• Description of the workshop process and how the decisions will define configuration and properties of their new system

Conclusion

While many of these recommendations are optional, this is a proven set of best practices to employ during a data modeling workshop.  Effective data modeling is a tedious but essential process to ensure an organization is maximizing their GIS.