As organizations plot their future use of enterprise resource planning (ERP) systems, knowledge management will increasingly pop up on the radar screen. Once relegated to the realm of obscure academics or touchy-feely corporate reformists, knowledge management--a sometimes fuzzy term that incorporates many business intelligence tools--is latching onto ERP systems and finding a measure of mainstream acceptance.
"Once organizations have their ERP systems up and running, they start thinking about what they are going to do with all the information being captured by the systems," observes Phillip Russom, director/data warehousing and business intelligence, Hurwitz Group, a research firm based in Framingham, Mass. This thinking inevitably follows the path of the information processing chain from query and reporting to online analytic processing (OLAP) to business intelligence to, ultimately, knowledge management.
|Knowledge management proponents see knowledge as the top of a hierarchy (pyramid) that begins with raw transactional data. At each level, the data undergoes transformation, augmentation, and abstraction until it becomes knowledge and, ultimately, wisdom.|
That was the case at Kennametal, a Latrobe, Penn., manufacturer of tools for the metalworking industry. Kennametal implemented all major modules of SAP R/3 and brought up its last major site in August 1998. "We were so focused on putting up SAP that we put reporting and decision support on the back burner," recalls Michael Taccino, manager/information resources. When the organization finally turned its attention to SAP reporting, it was shocked and disappointed. "We thought the reporting built into SAP would be adequate, but it wasn't," he reports. The company found itself, in effect, building its own SAP executive information system (EIS) using about 100 custom ABAP reports it wrote itself.
But even this solution quickly proved inadequate. "Users exhausted what they could do with the system in about two months," Taccino notes. The problem revolved around the flat data structures and the difficulty in joining different types of data from different SAP modules. The company found itself scrambling to build intermediate data structures to support each different type of analysis.
Finally, the company turned to SAS Institute, of Cary, N.C., for tools to create a flexible information warehouse that would support a range of knowledge management functions, such as predictive data mining, ad hoc reporting, and a robust EIS. The new SAS-based system rolls out for production usage late in 1999.
The Kennametal experience illustrates the frustration organizations face when they try to tap into in an ERP system for information and knowledge. SAP is good for accessing data used in daily operations, Taccino explains, but it is not good for long-term knowledge management or for the information companies need for long-term planning and decision making. "There is a tremendous wealth of knowledge in SAP; the challenge is getting it out and turning it into something you can use for decision making," he concludes.