Applying Sequential Pattern Mining to Portable RFID System Data

Abstract

This paper presents how data mining can be applied to RFID proximity tracking data captured in a production setting. The WINEPI algorithm is explained and used for mining sequential patterns from transaction data produced by portable RF transceivers that can be attached, for example, to the personnel and machines of a production facility. The contribution of this paper is the additional mindset of how data can be produced for a data warehouse with dedicated sensors in order to prototype the data warehouse itself - and how we can use this created knowledge as a help when designing intelligent manufacturing systems.