Analyzing Risk Dependencies on RFID-driven Global Logistics Processes

In this paper, we try to formally analyze risk dependencies on a global logistics process that is made up of a set of RFID-driven biz-steps2. In enacting a global logistics process, it is very important to control and trace each bizstep's execution as well as to visualize its status. Moreover, for the sake of improving QoS (the quality of services), it ought to be crucial for the system to provide an autonomous error-detection functionality on its running exceptions and very safe self-recovery mechanisms, as well, for the exceptional and risky situations. In resolving the QoS issue, the essential technologies might be RFID and BPM/workflow technologies; workflow is for representing a global logistics process in a natural, and RFID is for implementing each biz-step's application program tagging the logistics information. Also, the essential theory of the self-recovery mechanisms must be on the roll-back mechanism that determines up to where the roll-back operations have to be applied among the previous performed biz-steps on the corresponding global logistics process. As an essential theory of implementing a reasonable roll-back mechanism, the paper particularly formalizes a risk dependency analysis algorithm that is able to produce a set of risk dependency knowledge existing among biz-steps of a RFID-driven global logistics process, and that is eventually used for realizing the self-recovery mechanisms.