Article ID: | iaor200627 |
Country: | United States |
Volume: | 35 |
Issue: | 1 |
Start Page Number: | 76 |
End Page Number: | 87 |
Publication Date: | Jan 2005 |
Journal: | Interfaces |
Authors: | Tayur Sridhar, Liu Yong, Troyer Loren, Smith James, Marshall Sean, Yaniv Elan, Barkman Martin, Kaya Alev |
Keywords: | production |
In 2001, Deere's Commercial and Consumer Equipment (C&CE) Division, with growing sales of $3 billion, set out to improve its on-time delivery from plants to dealers and to reduce its inventory while maintaining customer service levels. C&CE used state-of-the-art inventory optimization techniques embedded in SmartOps' multistage inventory planning and optimization (MIPO) product to set trustworthy weekly inventory targets. C&CE used these targets, together with appropriate dealer incentives, to transform to a pull system and exceed its goals. With 2,500 dealers, 100 product families, and a 26-week planning horizon, Deere's application of multiechelon inventory optimization may be the largest example of applied stochastic inventory theory in practice in a multiagent environment. With the enterprise-wide system integration, Deere improved its factories' on-time shipments from 63 percent to 92 percent, while maintaining customer service levels at 90 percent. Between 2001 and 2003, Deere reduced or avoided inventory by $890 million, improving annual shareholder value added (SVA) by $107 million. By the end of 2004, the C&CE Division will exceed its goal in $1 billion of inventory reduction or avoidance, a year ahead of schedule.