Power Industry Data Modeling tool selection implementation case sharing

Due to the need to record increasingly complex data warehouse implementations, IT staff in the U.S. Omaha Public Power Division had to purchase a new set of graphical data modeling tools, but it was almost impossible to obtain approval from the parent company for the project.

The utilities were responsible for providing electricity to most of the Nebraska State, and they decided to purchase Er/studio XE data modeling software from Embarcadero Company. Company Business Intelligence (BI) and Data Warehouse architect Bill Harrison said he managed to get approval for the purchase and, by ignoring the technical aspects, saved the bottom line and explained the effect of action.

"It's really hard, it's hard to get approval," Harrison said, "So I decided to sell it in a marketing way instead of using a technical approach." ”

After deciding to put aside the High-tech talk, Harris came to Omaha Power Company to answer questions from senior managers. These questions include: Why do we need this tool? What does it do for the company? What is the return on investment? How long does it take to realize value? What happens if we don't act?

Data Model--shows the relationship between the data source and database tables, as well as the association with the Data Warehouse, where part of the Omaha Power is recorded in the form of a Microsoft Excel spreadsheet. But these documents are hard for the general business workers to understand. Harrison efforts to make the company understand that this situation will eventually lead to costly mistakes.

Harrison explains that the power company's corporate customers are increasingly generating bi reports by themselves without the help of IT staff, requiring user-friendly data modeling documents to help them get the right information, get the right results, and make the right decisions.

"A lot of people have tried to push this as a technical project, and I said, ' That's not going to work, we need to be market-oriented." "he said.

Data modeling software quickly produces results

Most of Omaha's IT applications are focused on Oracle PeopleSoft financials and General Ledger (Accounting and general ledger management systems) and a powerful information system that handles more than 300,000 business and home user data. Harrison says that all of the information that these applications are constantly producing will eventually be purged, weighed, and then imported into a large data warehouse for BI and analytics reports. The company uses the IBM Cognos BI reporting software.

"The customer information system has 15000 tables and views," Harrison said, "How do you deal with these?"

Omaha Power used Embarcadero Company's software for only a few months, but Harrison said he was delighted with the results. The organization has started using Er/studio to generate some much-needed data warehouse documents, and Harrison is expected to use the tool to create new data models soon as other applications are coming online.

Harrison revealed that Er/studio has two ways of displaying data models. The first is the standard report format, which basically uses text to describe the relationship between the data sources. Users can also use the entity diagram (ERD) feature, which shows rich and colorful graphical data sources, and how they relate to each other.

"For example, the ERD shows the architecture of our budget data mart." It shows the table structure and how they relate to each other, "Harrison said." You can browse directly from the workspace, or you can print it out. ”

Omaha also utilizes the Er/studio "reverse engineering" function to extract the structure of its data warehouse. Harrison and his team create a child model of the data mart so that it is easier for the user to understand. Other data modeling tools on the market, including CA Erwin, Oracle SQL Developer Data Modeler, and Quest's Toad model.

Standing on the edge of the "big data" explosion.

One of the great challenges that Omaha power companies are experiencing today is the unprecedented rise in data growth, the so-called "big data" explosion. With the scale and speed of data volume growth, it is becoming more important than ever to correctly record the data model and the overall data Warehouse architecture.

"Big changes are taking place, and we're facing big data challenges, this is just the beginning," Harrison said: "We are talking about hundreds of TB or even petabytes of data, as we introduce the smart grid and some other technology, which will force us to analyze the huge data before we have never seen the amount of data. ”

Smart grids and similar technical schemes allow tools to track high-precision information about consumers, such as how much electricity is consumed by coffee pots or other appliances at a given time.

"We need new tools and techniques to retrieve such data, because first, it is unstructured; second, the volume of data is huge," Harrison said. "You may not have recorded every project yet, but at least one data model should describe the overall architecture." ”