Information-Sensitive Inventory Management when Records are Inaccurate
<-- Return to the list
Date: 04-28-2009
Start Time:
1:00pm
End Time: 2:00pm
Speaker: Adam Mersereau, University of North Carolina
Location: Uris 333
ABSTRACT
The vast majority of inventory management research assumes that the inventory manager knows with certainty his inventory position. Recent empirical research, however, calls this assumption into question and reveals the reality of inventory management in practice: inventory records do not necessarily match what is on the shelf.
We study an inventory system with imperfect inventory records and unobserved lost sales. We propose tracking inventory using a Bayesian Inventory Record (BIR), a probability distribution that evolves over time to reflect the inventory manager's beliefs about the true inventory level, given replenishment and sales observations. A numerical study shows that simple replenishment policies based on the BIR recoup much of the cost of record inaccuracy. A BIR-based audit policy significantly outperforms the "zero-balance walk'' audit policy commonly employed by retailers.
We then investigate optimal BIR-based replenishment, formulating the problem as a partially observed Markov decision process (POMDP). Analysis of one- and two-period versions of the problem suggest that an optimal forward-looking policy may stock less than a myopic policy because of an inventory carryover effect and because of an "information effect:'' smaller stocking levels have an advantageous impact on the shape of future BIRs. Numerical experiments over longer horizons corroborate these findings and explore the tradeoff between policy complexity and quality.
Portions of this talk represent joint work with Nicole DeHoratius and Linus Schrage.
BIO
Adam J. Mersereau is Assistant Professor of Operations, Technology, and Innovation Management at the Kenan-Flagler Business School at the University of North Carolina, where his research interests are in information-sensitive decision making in operational and marketing contexts. Before joining UNC Kenan-Flagler, he served on the faculty of the University of Chicago Graduate School of Business for four years. He received his PhD from MIT's Operations Research Center in 2003.