Managing Short Life-Cycle Technology Products for Agere Systems

Over the past decade, the high-tech industry has been experiencing an unprecedented acceleration of technology innovations and rapid product introduction cycles. At the same time, the industry has been moving rapidly away from regional vertical integration toward a supply chain structure that is fundamentally global. In this paper, we describe a critical operational problem that arises in this environment. The research stems from a project started in the fall of 2002 partnering Agere Systems and the Center for Value Chain Research at Lehigh University. The main focus of the project is to study means of characterizing demands for short-lifecycle technology products. Agere is particularly interested in using demand characterization tools for capacity planning and capacity negotiation with their global supply partners. The shortened technological lifecycle poses major challenges to the time series forecasting methods fundamental to all commercial demand planning systems. As part of this project, we propose a new demand characterization method based on the notion of demand “leading indicators.” Given a cluster of products, our method identifies one or more products that provide advanced indication of demand behaviors for the rest of the cluster. Based on extensive analysis of a data set from Agere that covers twenty-six months from December 2001 to January 2004 and includes some 3,500 semiconductor products, we have discovered that we can consistently find leading indicators that predict the cluster demand pattern two to eight months ahead of time with correlation values ranging from 0.51 to 0.95. These findings have significant implications to capacity management in an increasingly global high-tech supply chain. We discuss opportunities for applying the leading indicator approach to various planning functions.