As I was driving home this evening, I could not believe how many houses where already lit up for the holidays!  I guess that brief glimpse of sunshine this past weekend inspired folks to check that task off the "honey do" list... I was sneaking a ride on my Harley. :) 

 Driving a car gives me some time to drift into thought.  One thought tonight was based on a conversation with my colleague, Javed Sikander on some of the more "fuzzy" aspects of RFID architectures.  As the CPG manufacturers and others begin to RFID-enable aspects of their ship-to processes and other case/pallet-level processes to comply with "requests" from their retail partners, there becomes real opportunities to address current, "as-is" processes to re-coup some of the investments in this infrastructure.  The concepts of pushing RFID tagging deeper in the manufacturing processes becomes appealing.  The challenges, however, among many others, really are architecting the "linkage" between the physical world of case/item movement and virtual world of information movement in the context of a complicated collaborative architecture and managing all that data...

There becomes a real need to architect several levels of abstraction as an RFID read event does not always (and in fact, almost never) directly feed LOB applications.   These events typically live at the “edge” of the enterprise.  There are logical layers including readers, "middleware", Enterprise Integration Services, enterprise apps.  Many read events may never "penetrate" layers above "middleware", while others are aggregated, transformed, and published as part of a larger process on a enterprise integration service provider and end up in the application of record.  Clearly SOA and interoperability will be key in any strategy. 

I would love to hear your feedback on the concept of “data proliferation” in an RFID enterprise!