Analytics in Manufacturing: The Collaboration EnablerThe quest for profitability has businesses in manufacturing industries on the hunt for analytics that serve collaborative execution models. Will BI providers step up?Or will solutions come first from ERP?
by Naeem Hashmi Continued from Page 2 PolyVista can help identify root causes of product and component failures or supply-chain network problems rapidly. Eventually, we will see such technologies embedded in other applications through APIs. This is how truly adaptable and intelligent manufacturing applications will emerge to drive collaborative processes. Database-centric data mining. Data mining engines are finding their way into the major database management systems, such as Oracle9i, IBM DB2 Universal Database, and Microsoft SQL Server. This presents good news for applications that, as part of their business processes, need to provide data mining from within the database engine upon which they are running. This approach reduces dramatically the time it currently takes to move data structures from the database into a form that the data mining tools can use. It also opens the way for new kinds of smart, database-centric applications. The drawback is that your application is tied to a specific database technology. In large organizations and especially in manufacturing industries, the array of database engines in use makes it hard to justify one single, database-centric, data mining platform. However, with the spreading adoption of XML-based languages such as the Predictive Model Markup Language (PMML), you will be able to exchange statistical and data mining models between PMML-compliant applications. PMML will eliminate many barriers. Now it is up to the BI vendors to provide solutions. Data mining from the ERP giants. With Oracle, PeopleSoft, and J.D. Edwards embroiled in merger and acquisition controversy, it's difficult to predict their ultimate product strategies. SAP, however, is marching ahead with its next-generation, integrated, collaborative, and adaptive applications for vertical industries and functions, including manufacturing. SAP is embedding data mining and real-time analytics within its technology framework without limiting which database, operating system, or hardware platform customers can use. Companies can use the SAP technology alone or distributed among collaborative partners, all of whom may share the same analytic application. Presently, SAP supports only a handful of algorithms, which focus on common customer relationship management (CRM) analytic needs. However, as SAP finds demand for additional algorithms, it will add them. Once the merger confusion is resolved, Oracle and PeopleSoft should jump back into advancing their approaches. Agility EnablerInformation is the primary enabling resource for agile manufacturing. How effectively you manage information determines the success of your business. However, what's required is an integrated environment within your business as well as service-level integration with collaborating partners — without question, a complex and demanding challenge. In 2002, I published a technical-reference architecture for service-oriented collaborative applications: the Extraprise Information Factory (EIF). I outlined four distinct integration models: tightly coupled, loosely coupled, on demand, and file-based data integration, as shown in Figure 2. Manufacturers at Stage 0 and Stage 1 processes should choose tightly coupled integration for activities such as statistical quality process control monitoring. Product activity monitoring analytics, however, should use loosely coupled integration through XML data exchanges. On-demand integration should be the chosen form for ad hoc requests to access and consolidate information. And finally, file-based integration should be the choice for moving large amounts of data, which is the case with product catalogs or large, integrated databases. It's quite interesting to observe how SAP is shaping up its core technology platform, NetWeaver, to support robust, collaborative business applications as outlined in the EIF. NetWeaver is not limited to typical J2EE-powered Web application servers; it includes SAP Business Warehouse (BW), Knowledge Management (KM), Process Integration (Exchange Infrastructure), and other components. SAP's direction sheds light on opportunities ahead for BI providers to develop a new breed of service-based, event-driven applications. The paradigm for future business applications is shifting away from standalone query, reporting, and "report mining" to exception and event notification that requires analytics to be tightly coupled with business processes. Will BI providers make the leap? Naeem Hashmi [nhashmi@infoframeworks.com] is CTO of Information Frameworks. An author and expert in emerging e-business intelligence and portal technologies, he has been a catalyst in defining the technical architecture for several products, including SAP Business Information Warehouse. RESOURCESRelated Article at IntelligentEnterprise.com: "BI for Sale" IntelligentERP, Oct. 20, 2000
|
 |
Most Popular This Week
IE Weekly Newsletter
Subscribe to the newsletter
| |
| |||||||||||||||||||||||||||||||
