LaBarge, Inc. installed its Enterprise Resource Planning and Manufacturing Resource Planning software in 1992. The mainframe/terminal-based application had served the company well.
“I couldn’t find anything that system didn’t do that the business needed,” says George Hayward, Director of Information Systems at LaBarge, a St. Louis-based manufacturer of high-performance electronics and electro-mechanical equipment. “It captured the manufacturing data, did the math we needed, told us what we needed to buy. But it was cumbersome getting data into and out of it.”
To get any reports out of ERP, users typically would have to put in a request to IT, wait overnight for the batch to run, and then pore through a few hundred pages of 132-column printouts to find the needed information. If the data wasn’t there, they would have to wait another day to receive a new report. To speed that process and improve decision making, but without ripping out what was already in place, Hayward decided to install WebFOCUS Business Intelligence software as a new user interface into the enterprise systems. WebFOCUS is the creation of Information Builders, Inc. of New York City.
“Implementing a new ERP system is expensive, risky, and sometimes you have to shut down parts of the business for a while,” Hayward explains. “Since we didn’t really need new guts, I just needed a better way to help people get the data they wanted.”
LaBarge uses custom designed ERP and MRP software from Western Data Systems (now JDS Software). The system collects data from LaBarge’s six ISO 9001:2000 certified manufacturing plants to provide unified reporting. From 1991 to 2004, the company ran the systems on hosted mainframe.
It required a series of actions to make the data more readily accessible to users. One was to bring the ERP/MRP in house. Since the mainframe it was designed to run on was no longer manufactured, LaBarge purchased an X235 series emulated mainframe from T3 Technologies, Inc. of Tampa, Florida. The server uses runs 32-bit UNIX. On top of the UNIX, it runs FLEX-ES mainframe simulation software. IBM’s OS/390 then runs on top of the FLEX-ES.
“We were sharing CPU availability with other customers and now run our own dedicated 33 MIPS (million instructions per second) machine,” says LaBarge ERP Systems Manager Dennis M. Carter.
The next element was the BI software. Among other reasons, LaBarge selected WebFOCUS because of its ability to access the Virtual Storage Access Method (VSAM) files used by IBM mainframes. WebFOCUS resides on two Dell blade servers running Windows 2003.
The BI software could directly pull the data from the VSAM files when executing a query. Instead, however, LaBarge prefers to pull it into SQL databases that mirror the data in the VSAM files. This provides a structure that is easier for end users to understand. For example, in the VSAM files, a purchase order may consist of data from a header file, a purchase order detail file and a customer master file. In the SQL databases, users can see the data expressed as purchase orders. Users are able to see the structure of the data and how things relate to each other. They can also dump the data into an Excel file for further analysis.
“We spent a lot of time before writing a report figuring out the relationships and what we wanted the interface to look like,” Hayward explains. “Once we had a good list of rules and a map in place, then it was fairly simple to write the reports.”
LaBarge did have Information Builders consultants come in to help with the initial report writing and to train the IT staff. The reports were formatted so they were similar to the paper reports employees were used to seeing. A new front end was also added to the reports so that the users could, for example, limit a report to a certain subset of customers. The system went live in about two months. About 120 users in the areas of finance, manufacturing, engineering, purchasing, receiving and customer service currently access the system directly rather than expecting IT to run reports for them.
“Now that they are directly interfacing with the data themselves, they understand the relationships of the data better and execute multilevel queries against multiple data sources in ways they probably wouldn’t have thought of in the past,” says Hayward. “They are thinking of the data as a whole, are asking better questions and getting the answers themselves.”
And this, he says, leads to better decision-making.
This article was first published on EnterpriseITPlanet.com.