Masterpiece: GS1 Tag Data Standard 1.5

GS1 EPCglobal ratified and published the most recent version of the Tag Data Standard (TDS), 1.5.  I have always been a fan of TDS, but earlier versions served as much to expose embarassing disconnects in GS1 standards as they did to explain how to apply GS1 identifiers in an Radio Frequency IDentification (RFID) context.  Earlier versions of TDS tightly bound the concept of the Electronic Product Code (EPC) to RFID, and that’s just wrong.  “EPC”, even as defined in those earlier versions of TDS, is a way of uniquely identifying objects on a global basis.  It’s purely a globally unique identifier standard.  That’s a hugely important and relatively new concept.  RFID–a simple data carrier technology that’s been around for many years–is almost insignificant in comparison.

Even the name “Tag Data Standard” reflects this wrong-headed binding of EPC to RFID by seeming to place the RFID “Tag” at the center of the “standard”.  Yes, there is a need for a “Tag Data Standard”, to show how to encode an EPC into an RFID tag, but it is wrong to hide the definition of the EPC inside such a limited specification.  EPC is carrier technology independent.  RFID is only one of several carrier technologies that can carry EPC information.

This disconnect between the standard and reality has caused a lot of confusion, even within GS1 where people seem to fall into one of two catagories distinguished by either bar codes or RFID.  People from both worlds within GS1 have assured me that and EPC can only be encoded into an RFID tag.  They are wrong, and GS1 would be a better organization if everyone associated with it would study the EPC and think about what it means to GS1 and its users.  I think it would expand their minds.

But finally, in TDS 1.5 it is made clear…

“Nevertheless, it should always be remembered that the EPC and RFID are not at all synonymous: EPC is an identifier, and RFID is a data carrier. RFID tags contain other data besides EPC identifiers (and in some applications may not carry an EPC identifier at all), and the EPC identifier exists in non-RFID contexts (those non-RFID contexts including the URI form used within information systems, printed human-readable EPC URIs, and EPC identifiers derived from bar code data following the procedures in this standard).”

Now GS1 needs to update a lot of their documentation to reflect this reality.  Much of it continues to perpetuate the tight linkage of EPC and RFID.  See this FAQ for example.

This separation of EPC and RFID is pertinent in the pharmaceutical supply chain right now because many companies are planning to apply globally unique serial numbers to their products to comply with regulatory requirements in countries like Turkey, Brazil, Italy and in states like California.  At the same time most of them are also planning to apply globally unique serial numbers to their cases, pallets and totes.  The companies I am aware of are currently planning to use GS1 EPC’s for these serial numbers, but, as I predict in “RFID is DEAD, at Unit-Level in Pharma“, most will not use RFID to carry those numbers, but will use 2D barcodes instead.  A few will use RFID and 2D barcodes to carry a single EPC.  None of these uses would be possible if the EPC wasn’t fully distinct from the carrier technology.


Yes, TDS 1.5 is a masterpiece.  For the first time there are separate sections for the EPC and for RFID as a carrier technology of EPC’s, and it is written in a way that is very easy to understand.  This version also includes clear information about encoding GS1 element strings into EPC’s.  GS1 element strings are values that are typically encoded into barcodes.  But if it were up to me in the future, GS1 would entirely remove the specification of the EPC from TDS.  There would be an EPC Standard, a Tag Data Standard (let’s call it TDS 2.0), and a Barcode Standard.  Or perhaps they should keep all carrier technologies in a single standard.  I’d call it the “EPC Carrier Standard” and it would have an RFID section and a barcode section.  It would show how to encode any EPC into one carrier technology or the other and how to convert from one to the other without data loss or ambiguity.  TDS 1.5 is very close to this ideal.

It seems simple and the need is great because pharmaceutical companies are moving forward right now, but GS1 has been grappling with RFID and bar code interoperability for quite a few years.  TDS 1.5 goes a long way toward resolving these issues, but full resolution is still ahead of us.  Oh well, maybe TDS 1.6 will be an even better masterpiece.  Watch for it to arrive sometime next year.


The EPC standard, encased inside of the RFID-Tag Data Standard 1.5, is so simple that it’s not surprising that people miss it.  Section 4 of TDS 1.5 expresses it this way:

“The EPC is a universal identifier that provides a unique identity for any physical object. The EPC is designed to be unique across all physical objects in the world, over all time, and across all categories of physical objects. It is expressly intended for use by business applications that need to track all categories of physical objects, whatever they may be.”

The most likely systems to recognize and operate on EPC’s are those that are based on the GS1 Drug Pedigree Messaging Standard (DPMS, a.k.a., the GS1 Ratified Pedigree Standard), or the GS1 EPC Information Services (EPCIS) standard.  Both of these standards specifically work with EPC’s.  DPMS, of course, is specificly designed for compliance with U.S. electronic document-based pharmaceutical pedigree laws like California, Florida and the federal PDMA.  EPCIS is an interface standard for use in systems that focus on documenting object visibility in a supply chain.  The future GS1 Discovery Services standard will also operate on EPC’s.

I assume that GS1 has copyrighted the acronym “EPC”, but what’s interesting is that there is nothing special–nothing GS1-specific–about creating an object identifier that can be used in systems based on DPMS or EPCIS.  That’s because those two standards simply require the use of unique identifiers that follow a set of open W3C standards, specifically RFC2396 (see also RFC2141 and RFC3406).  GS1 refers to this as an identifier in “pure identity” format, which is also known as “URN” format (“URI format” to purists).

For example, a valid identifier for use in DPMS or EPCIS for an object serial number that I took from TDS 1.5 is:


GS1 was issued the namespace “epc” by the Internet Assigned Numbers Authority (IANA) and that gives them the authority to define the schema for all of the data elements to the right of the letters “epc:” in the identifier above.  TDS 1.5 specifies what each of those segments mean and their value ranges.

But DPMS and EPCIS would work just as well if the object’s unique serial number were defined with some other valid URN: namespace besides “epc:”.  Nothing stops organizations (like the Heath Industry Business Communications Council, HIBCC) from acquiring their own namespace from IANA and defining their own schema that would encode their own unique, non-GS1 serial numbers.  And those URN: formatted (URI formatted for the purists) unique identifiers would work just as well and would be fully interoperable with other “URN:EPC:” serial numbers.  Of course, just don’t call them EPC’s.

For many types of commercial objects, there are no viable alternatives to GS1 numbering so use of GS1 EPC’s are a given.  But whenever medical products are serialized at some time in the future, GS1 is only the latest entrant (though growing fast as I point out in “WAR: GS1 Vs. HIBCC“).  HIBCC was the traditional numbering not-for-profit organization for these type of devices and they recently introduced their own approach to unique identification based on their Labeler Identification Code (LIC) and an added serial number.  Like GS1, they hid it inside of their RFID specification.

With version 1.5, TDS has gone through an impressive rewrite.  It’s easy to read, especially the first 6 sections (23 pages).  I highly recommend it to anyone wishing to better understand the EPC standard at the conceptual level.