LSI says supply chain complexity isn't always worth it
By Malcolm Wheatley, senior contributing editor (malcolm_wheatley@compuserve.com) -- Manufacturing Business Technology, 11/1/2007 12:00:00 AM
As companies globalize, their supply chains gain complexity—especially when globalization warrants greater reliance on manufacturing outsourcing partners. It's a problem that's more than familiar to Ashok Santhanam, chief executive of Bristlecone, a technology and business process consulting services provider.
Conventional enterprise applications aren't best-positioned to minimize that complexity, Santhanam asserts. “Transactional tools deal with the supply network that they've been given,” he says. “Their task is managing at the micro level. What they don't do is shape the decision about what that network should look like, and how much complexity is permissible.”
While specialist modeling tools exist—among them offerings from i2Technologies and Oracle—the ultimate decision, says Santhanam, is a strategic one, ideally made at an appropriate level within the organization.
|
|
"With a global supply chain, the decision about how much complexity to live with is essentially a policy one, supplemented by spreadsheet-type analysis." —Ashok Santhanam, chief executive, Bristlecone |
“With a global supply chain, the decision about how much complexity to live with is essentially a policy one, supplemented by spreadsheet-type analysis. It's a trade-off between economies of scale and lower costs on the one hand, and shrinking the distance between the demand point and the supply point on the other,” he says.
Once that trade-off has been made, Bristlecone works with clients such as Milpitas, Calif.-based semiconductor manufacturer LSI Corp. and Toronto-based contract electronics manufacturer Celestica to “operationalize” the resulting supply network, often using SAP's advanced planning and optimization module (APO) as the transaction and optimization engine.
At LSI, says Diana White, senior manager of supply chain solutions, the challenge is optimizing a fab-less network, where end-customer demands can be met by capacity at any one of a number of semiconductor fabrication plants and their respective supply chains.
“When fabs aren't highly loaded, and people aren't fighting for every wafer, it isn't an issue. But as spare capacity shrinks, we need to be able to plan and source across multiple paths, and update those plans more frequently,” explains White.
While still relying on SAP's APO—again fine-tuned with Bristlecone's help—Celestica is pursuing a different strategy. “Increasingly, we were dealing with a more complex network, with more nodes, and greater distances between those nodes,” says Harvinder Sembhi, VP of supply chain strategy and planning. “We faced a choice: We could continue to model that complexity, or simplify it.”
Simplification won out. Today, notes Sembhi, a growing proportion of Celestica's manufacturing takes place in eight mega-plants.
“Our objective is fewer, but larger, plants, with suppliers co-located—or located very close by—making just-in-time deliveries as required,” he says. “It's a strategy we think has put us two to three years ahead of the competition.”
