IT in manufacturing is still struggling with data integration, despite the success of enterprise resource planning (ERP) systems in bringing together data and processes within others sectors such as retail and the public sector.
Common IT challenges in manufacturing - including reducing costs, increasing efficiency, improving data visibility and extending collaboration up and down the supply chain - all have data integration at their core.
Analysts agree that the big suppliers in ERP have not done particularly well when it comes to manufacturing. They have failed to deal with what Teresa Jones, senior research analyst at Butler Group, calls the "nitty gritty" level of manufacturing data needs.
Manufacturers have attempted to tap into information generated by the highly specialised systems used to control particular manufacturing processes. As a result, they have been left with ERP portfolios that are more diverse than those found in other types of organisations.
Many find themselves with an array of different ERP systems within the firm, according to Duncan Jones, senior analyst at Forrester Research. This has necessitated difficult, expensive and often inefficient integration projects.
IT departments in the manufacturing industry are now looking to generate more value from existing systems. To do that, they need to find a more sophisticated and cost-effective means of integration across systems in different parts of the manufacturing business, according to Andrew Hughes, research director of manufacturing at Gartner.
"The challenge is to get more accuracy by integrating manufacturing systems with ERP systems and other business processes. It is not a matter of integrating the manufacturing system just with the ERP system it is about making use of the information within the business processes and making use of information within the factory to drive profitability," Hughes said.
Manufacturing organisations are beginning to collaborate more with their customers as well as with their suppliers, said John Griffiths, managing consultant with PA Consulting's IT group. This means manufacturers need to find a way of easily collaborating across an increasingly large number of systems as they form alliances outside the organisation.
Manufacturers are tackling this in several ways. Many are simply upgrading their existing ERP systems. Some 49% of European manufacturing companies surveyed by Forrester are planning some sort of upgrade to their ERP systems, compared with 42% in retail, 35% in finance, and 25% in the public sector.
Another common strategy IT departments in manufacturing are using is to implement business intelligence software as a way of extracting more information from existing data systems and moving all the important business data into one place.
The Forrester survey also shows that 50% of European manufacturing companies are implementing business intelligence software for the first time or doing a major upgrade, compared with 45% in the public sector, 43% in finance, and 38% in media and entertainment.
Although described as "flavour of the year" by some analysts, a more long-sighted approach has been the move towards ERP systems that are either based on a service oriented architecture (SOA) or can be offered as a web service.
Few analysts see SOA as the complete solution to the integration challenges facing IT in manufacturing, but most see it as the correct approach.
Among those who see SOA as part of the solution is Nick Slowe, IT director of The Institution of Engineering and Technology. "However, there is still a need for solid integration competencies," he said.
The role of data synchronisation
According to Griffiths, "Even though SOA is a move in the right direction, it needs to be complemented by the application of the relevant supporting or enabling technologies." Such technologies include RFID, new algorithmic-based data-mining tools, or something else that achieves near real-time data synchronisation across enterprise information systems, he said.
At the very least, Hughes sees SOA as a way of reducing the prohibitive costs traditionally associated with integration. However, "there is still a large cost to overcome to convert from traditional business systems to an SOA-based system," he said. In the long run, SOA is a viable approach to enabling a more efficient means of integration than before.
He adds that it is for this reason that some manufacturers are already going down the SOA path, and others will follow as the business benefits begin to outweigh the cost of conversion.
Hughes predicts that all companies will eventually have to follow suit when SOA is all that is on offer from the business-software providers.
Gartner has seen an almost 100% move over to SOA by all the software suppliers in the manufacturing operations space, including control systems and ERP, said Hughes. This indicates that IT departments in manufacturing need to be open to the idea of SOA in the longer term.
Rather than something to be resisted, SOA may well be an important part of the solution to the integration woes of the manufacturing industry.
It may enable IT departments to finally achieve their goals of reducing costs, increasing efficiency, improving data visibility, and extending collaboration with the manufacturing enterprise and beyond.
Case study: Dow Chemicals
The Dow Chemical Company is among those manufacturing companies already following a service oriented approach in a bid to gain competitive advantage, by replacing its existing SAP R/2 system with a global SAP service oriented architecture (SOA).
Dow's SOA, due to go into production this year, is based on SAP's Netweaver and MySAP ERP platforms.
Melanie Kalmar, Dow's information systems programme director, said Netweaver components had matured to the point that the time was now right to launch an SOA project. Netweaver's support for non-SAP applications was also key to Dow's decision, she said.
Kalamar said Dow was impressed by the reduction in planning time achieved through a proof-of-concept document-management system.
The SOA project is aimed at enabling Dow to re-use data definitions and code, speed up application development, reduce system maintenance costs, and support business operations more quickly and responsively.
The first projects using SOA will cover purchasing, work and plant maintenance, pricing, and a message-optimisation platform for tracking correspondence and documents.