RFID technology has advanced a lot in recent years. In order for enterprises to be competitive, an increasing number have integrated the technology into their operations. The prime challenge for RFID middleware is not to shield the bottom device-level heterogeneity but to accelerate the integration with legacy system and application in a complex scenario. In this paper, we introduce a distributed work flow based RFID middleware deployed in cluster environment. It is modeling-based and service & component oriented. The architecture focuses on the top-level application: the business integration of RFID system. It combines using work flow engine with ESB (Enterprise Service Bus) to separate the execution logic and execution entity. The mechanism makes it easier to distribute in a cluster scenario. It facilitates system's maintenance by conforming to JMX (Java Management Extension). RFID events as engine triggers, first of all, mapped to units of work flow by the RFID adapter. As a result, business integration people can concentrate business process design leaving all other things to our system. At the end part of this paper, we provide solutions for one scenario in BOLUO Logistics Company using this system and evaluate the performance.
[1]
Jorge González Fernández,et al.
Transf-ID: Automatic ID and Data Capture for Rail Freight Asset Management
,
2009,
IEEE Internet Computing.
[2]
Wilhelm Schäfer,et al.
Concepts and implementation of a rule-based process engine/sup +/
,
1992,
International Conference on Software Engineering.
[3]
Dong Wang,et al.
Design of RFID Middleware Based on Complex Event Processing
,
2006,
2006 IEEE Conference on Cybernetics and Intelligent Systems.
[4]
Keunhyuk Yeom,et al.
Contextual Events Framework in RFID System
,
2006,
Third International Conference on Information Technology: New Generations (ITNG'06).
[5]
Wil M. P. van der Aalst,et al.
Verification of Workflow Nets
,
1997,
ICATPN.