Improving Inventory Management at Lucent Technologies
The Problem
Like many large manufacturers, Lucent Technologies was challenged by the complexity of its parts inventory management needs. Inaccurate forecasting of demand for particular products or components was causing production bottlenecks and expensive delays in getting products to customers. The electronics and telecommunications products company decided to ask an analytics team to build a more sophisticated and robust Inventory Requirements Planning (IRP) computer-based system to master the situation.
The Analytics Solution
The inventory management system developed by the analytics group had two unusual and highly beneficial features. First, it was constructed to allow periodic input from managers to analyze business trade-offs inherent in any inventory management system and make adjustments if needed. Second, the system made the entire process of setting inventory levels transparent to all affected managers and departments. The purpose was to prevent “black-box syndrome,” in which material planners either ignore system-generated inventory recommendations or respond in the opposite manner by using system-generated recommendations blindly, even in the face of circumstances clearly beyond the system’s ability to accommodate. The new IRP system enabled Lucent to take advantage of its staff’s experience and expertise to weed out bad data and override the system’s recommendations when appropriate.
The mechanics of the system itself are based on a trial-and-error (heuristic) approach that separates basic inventory requirements from “buffer” inventory. After observed variations in demand and supply have been tabulated for each time period, a statistical model then describes “net provisioning variability” (variations in the difference between demand and supply) over different supply intervals. The system uses this data to estimate inventory required to obtain desired service levels.
The Value
Benefits of the IRP tool at three Lucent facilities were significant. For example, at a facility that manufactures optical networking products, parts inventory was reduced by 40%, for an $11 million inventory cost savings, and order backlogs were cut by 68%. At another facility, where fiber optic products are made, the plant achieved a 98% material availability level, resulting in a 20% increase in service levels. And at the third facility, where the IRP tool was applied to all levels of the internal supply chain involving some 15,000 parts, $55 million in inventory savings were achieved.
