Improving Dynamic Decision-Making Through RFID: A Partially Observable Markov Decision Process (POMDP) for RFID-Enhanced Warehouse Search Operations

Abstract

Misplaced items contribute significantly (2-10%) to the operational expense of a typical warehouse. In this work we develop a Partially Observable Markov Decision Process (POMDP) model for RFID directed search to detect misplaced items within a storage environment. A forklift operator (FLO) equipped with an RFID reader is assigned to search for a misplaced item in a warehouse. A FLO does not know the location of the tagged misplaced item and is guided by the imperfect variations in the strength of the signal received from the RFID tag (active or passive). The model considers five actions, five observations in scenarios with different RFID signal strength distributions namely, excellent, good and poor. An extensive simulation study has been conducted to evaluate the performance of RFID-driven POMDP search method. Specifically, the effects of signal strength distributions, initial beliefs at the start of the search and, the discount factor have been studied.