EPC: The End of Bar Codes? (Part 1 of a series)

Read Part 2

Introduction

When it comes to understanding EPC, people generally fall into one of two categories: those who know nothing about it and those who misunderstand it. Of course, there's a third group and, thankfully, they're the ones directing its development.

What is EPC?

First, let's understand what it's not. It's not an early pregnancy check. It's not the immediate death knell for bar codes. It's not just U.P.C. on a chip. And it's not quite ready for prime time.

EPC stands for "Electronic Product Code", which might lead one to suspect it's an extension of the U.P.C. (which stands for "Universal Product Code").

EPC, however, is much more than an RFID-based product code that includes electronic article surveillance (EAS) capabilities. The implications -- and complexity -- of EPC go far beyond what's being done with today's bar code systems.

At the same time, however, EPC is not a radical new concept. Many of its core item identification functions could be accomplished with the newer types of bar codes (such as matrix and composite symbols).

What sets EPC apart from all the existing standards and applications is the systematic approach being developed by the Auto-ID Center at MIT.

Background

EPC was conceived as "an Internet of things" (as opposed to the person-to-person or computer-to-computer Internet). EPC would leverage the benefits of RFID's non-line-of-sight reading, large data capacity and anti-theft/anti-counterfeiting features. Additional sensors (such as temperature, time, or drop force) might even be used to write data to the tag itself.

Initially, the vision was that every item in the supply chain -- down to the individual point-of-sale item -- would be RFID-enabled to constantly report its location, condition and status to on-site information systems or to remote databases via the Internet. This concept requires every item to have a unique identification code. This concept is what many people understand about EPC -- even though the concept has since been significantly refined. And this is the major point of misunderstanding.

Some recent, and overly optimistic, news releases have suggested that consumer goods manufacturers will soon begin using EPC on low cost commodities. Critics have pointed out that putting RFID tags on every can of soup or box of cereal would be cost prohibitive, given current cost points for even million-tag orders. And they're right.

But they're wrong to believe that's what is currently being proposed.

At some point in the future we may indeed see tag prices fall to the point where it makes sense to tag each item to provide completely automated inventories and point-of-sale check-out. Until that time, however, what's being proposed for EPC is far more realistic.

While the focus is still on promoting the concept of a universal Internet of things, unique identification does not necessarily mean every item gets an individual serial number.

Current Thinking

As with any implementation of AIDC technology, the current strategy is to apply EPC where there's a strong business case.

Unique identification is envisioned for applications where individual item tracking returns real benefits. For example, identifying and tracking an individual box of cereal may have little value at this point whereas uniquely identifying and tracking a pallet or even a case of cereal may offer tangible benefits for shipping, receiving and inventory applications.

For many low-cost, mass-produced items, however, a lot/batch number or expiration date would be far more valuable that individual item identification. A White Paper authored by David L. Brock, published by the Auto-ID Center states, "Although the objects in a batch are indistinct, they may, however, be assigned a unique EPC number that designates the entire lot." Brock's paper recognized that for many items, such as common fasteners or low-cost electronic components, "...it is unnecessary and cost prohibitive to consider individually tagging." It is envisioned that containers or bins containing these items would be tagged.

In this regard, EPC is not very different from many current bar code applications where groups of items are identified and tracked according to a common feature such as their lot/batch number. The major difference may well be in the wording.

Whereas most application standards reference "a unique lot or batch number," which is the common language of manufacturing, the White Paper uses the phrase "a unique EPC number." While this phrasing is consistent with the rest of the paper, it is likely the cause of much of the confusion about EPC.

A "unique EPC number that designates the entire lot" is not so different from "a lot/batch number" except that the EPC would implicitly contain manufacturer and other identification codes.

In other words, an EPC lot number could always stand alone whereas the lot/batch number on a bar code label often can not. In many bar code applications, lot/batch data is presented in a separate symbol and must be associated with the manufacturer code in another symbol elsewhere on the label. One final point on the current situation: EPC doesn't exist yet.

Any suggestion that a company is going to implement EPC today demonstrates a lack of understanding of what the EPC system involves. Yes, some companies may begin (or have begun) using RFID tags with EAN.UCC data and formats. This, however, is not and implementation of EPC.

The Vision

Just as the U.P.C. is far more than just a bar code symbology (incorporating international vendor identification and a number of coding conventions), EPC is more than just data on an RFID tag.

The EPC concept relies on an integrated, systematic approach that includes data content, technical standards, coding standards, mark-up language and software systems. This integrated approach is where EPC is breaking new ground -- not so much in the concepts themselves but rather in their integration from the very beginning.

In the past, various AIDC application standards were developed separately from the technical standards. Systems were designed on an industry-by-industry (or even company-by-company) basis and were forced to evolve as the vision changed and uses grew.

EPC is designed to be launched much further along that evolutionary path -- having learned the lessons of all the previous AIDC applications -- with all the tools in place to manage its future growth and change.
Here's the key point: the components for EPC are not yet in place. Therefore, it can not yet be implemented.

Next month, we'll examine EPC's systematic approach and its various components.

Note: David Brock's White Paper on EPC can be found at:
http://www.autoidcenter.org/research/MIT-AUTOID-WH-011.pdf