Locating and tracking data center assets using active RFID tags and a mobile robot

Abstract

We describe an approach to completely automated asset tracking in data centers using a vision-based mobile robot in conjunction with active RFID tags. Typically, active RFID tags are tracked using fixed readers. The granularity with which one can localize tags is based on the number of these fixed readers, each of which is not inexpensive. In large data centers, the cost of such a solution can therefore become excessive. We describe the use of a mobile robot, utilizing a single RFID reader and a variety of algorithms for locating the plethora of RFID tags one can encounter in such data centers. These approaches are validated through experiments performed in a production industrial data center.