George Marinos wishes to thank Douglas Roach for his contributions to this article. Roach is a manager in PricewaterhouseCoopers' Data Management Group focused primarily on helping enterprises manage the complexity of real-time data environments.

Before the meeting started, the data management team looked a bit uncomfortable. It wasn't every day that they were asked to sit down with senior business-unit executives - in this case, the COOs of the organization's top three performing divisions.

Actually, these meetings should have been routine - especially because business benchmarks always represent the yardstick required to measure data management's value. As a global conglomerate, however, this company had something even more important on the line.

In fact, its entire cross-sector strategy depended in great measure on how well it translated market-leading knowledge in one of its industry positions into competitive advantages driving its performance in others.
Clearly that point wasn't lost on the global CIO. "After two years of investment, the first wave of results from the RFID applications in our consumer products and retail lines of business look promising," he began. "Particularly with respect to core supply chain functions such as our inventory tracking, distribution center operations and retail store management."

He went on. "But looking down the road, we believe that the benefits of RFID will also extend well beyond the supply chain to more strategic if less obvious applications."

He paused and looked across the table at two of the COOs. "For instance, the benefits may be seen in pharmaceutical applications such as anti-counterfeiting and product identifications to support drug recalls. Or healthcare applications in direct support of the digital hospital."

The CIO leaned forward. "Here's the question we need to answer: What are the specific opportunities and obstacles to our using what we already know about RFID in our retail and consumer packaged goods (CPG) channels to establish forward-facing market positions within the most attractive of these emerging applications?"
Two years ago, this boardroom conversation would have only been about data - pilot-testing the point-to-point integration of RFID data to a legacy application or internal data source.

Today the conversation is quite a bit more sophisticated. It's about RFID and the value of context. It's about that crucial bridge between data capture and reporting - complex event processing (CEP). And, it's about whether translating RFID data into knowledge might offer these frontline operations chiefs their first steps toward breakthroughs in real-time or, at least, near real-time operating intelligence.

Or not - as you might be tempted to point out. Especially if the RFID business case that your team is trying so hard to develop doesn't support it. Maybe you're acutely aware that RFID doesn't do a good job of pushing out short-term ROIs (if you're not, ask mandated suppliers). Or there's another reason: you're uncomfortably familiar with the risks and uncertainties that always seem to accompany the messy arrivals of these disruptive technologies.

An Expanding Market Impact

It's difficult to dismiss RFID's forward momentum. As the world's largest supply chain owners - such as Wal-Mart, Target, Tesco, Metro AG and the U.S. Department of Defense - mandate RFID compliance among their suppliers, lead participants in these supply chains are beginning to integrate RFID across their own operations and supply base.

Today RFID represents the fastest growing segment of the wireless infrastructure and application services market. And industry insiders are talking openly about a pretty torrid RFID market growth rate - the North American market for RFID technology and services alone is expected to grow at a 40 percent CAGR from 2003 to 2008.
Even some of the skeptics concede that RFID has the potential to help form the "Internet of Things" - a smart network of items helping companies respond in real time to changes in supply and demand.

There's a hook here though: cashing in on the promise of RFID depends on something even more important than how ready you are to manage this deluge of data. Reaping the rewards of RFID also fundamentally depends on how well you can build context into this data. Here's why.

Putting RFID in Context

An RFID system can be designed to provide information on the location and the identity of an item in a manufacturing line, an asset, property or person. It can also - with the recent introduction of biosensor chips - even convey information about the physical condition of materials in transit.

To begin, one must understand the core components of an RFID solution. These include, 1) tags (mobile transponders that store data, silicone providers, tag converters, label converters and packaging with integrated tags); 2) printers/readers (including RFID reader, printers, encoders and label applicators); 3) software (including software applications, middleware and Web services supporting RFID solutions); and 4) integration services (including all services related to designing, developing, integrating and deploying RFID).

What allows an RFID application to shift from being location-based to identity-based is savant middleware - the sophisticated software that introduces context to the simple data that RFID readers are designed to capture.
It's a critical solution component that functions essentially as an automated "traffic cop," performing key roles such as managing tag-based and reader-based events, accepting and forwarding interrupt commands from an application, and flagging users when exceptions warrant intervention.

As depicted in Figure 1, it is this combination of data and context - this complex event processing - that translates data into information and eventually creates real-time intelligence.

Figure 1: The Role of Complex Event Processing in RFID

Think about the technical implications. Unlike bar code technology for which information is scanned only when someone puts a printed label in front of a reader, RFID is always "turned on."

Clearly, you must be able to continuously filter, mine and manage these huge unending streams of data. Some RFID experts have pointed out that, depending on the application, a single warehouse can generate multiple terabytes per day.

However, simple data is only useful to a point. It requires context to become information. That means you will need to 1) augment this data with meaning, 2) support it with relevant location- and time-specific detail from internal applications, third-party systems and Web services, and 3) tie this information to specific business processes.

Finally, in order to translate this into actionable intelligence, this information must be aggregated, summarized into meaningful reports and distributed to various touchpoints along the information supply chain (ISC) - managers and key constituents who are in the best position to analyze, interpret and take action in real time.

PML: The Language of Context

Process aside for the moment, it takes the right protocol to accomplish this. This protocol - at least with respect to RFID - is known as Physical Markup Language or PML. Simply explained, it's XML for RFID. In short, PML is designed as a collection of common, standardized vocabularies that represent and distribute information related to RFID-enabled physical objects and savant communications.

In particular, PML helps ensure that savant middleware performs filtering, aggregation and counting of data tags - some of the most crucial steps necessary to reduce the volume of data before it gets forwarded to enterprise applications.

The Barriers to Mass Adoption

Even within industries best positioned to capture any new efficiencies, planners can't count on clear visibility into just how this new technology will play out - if it will at all. Some analysts, in fact, view RFID merely as a "disruptive technology in waiting," because the great majority of current deployment practices are focused on retail compliance.

This uncertainty may go a long way toward explaining the caution that describes companies in RFID's "tactical" camp. These are organizations that, for better or worse, remain largely hesitant and reactive in response to mandates, choosing to limit their project objectives to tactical goals such as baseline compliance or minimizing total cost of ownership (TCO).

It's the "slap and ship" approach to RFID: complying with a mandate merely as a cost of business rather than as a strategic initiative, and its proponents tend to point to similar lists of concerns - in effect, the "street level" barriers to mass adoption. These include:

  • Standardization. Presently, global standards have not yet been established for either markup language or radio frequency spectrum utilization.
  • The Elusive $0.05 Tag. Today, the cost of a tag has fallen to less than 20 cents in quantities of a million, while readers cost as much as $1,000 each. Though technical and manufacturing advancements continue to reduce costs, an RF tag's integrated circuit will probably never be as cost effective as a bar code.
  • Technical Hurdles. Stubborn technological barriers remain. For example, passive tags are often unreadable when used on liquids (beverages) or metal (foil wrappers), and interference with other wireless devices can affect readers' ability to read tags.
  • Integration. RFID isn't a standalone technology, and determining just how it should be integrated across a particular business process can be a complex task. As one component of a supply chain system, for example, RFID may have to be integrated with systems such as electronic catalogs, GPS vehicle location devices, master supply chain scheduling, activity hub supply chain monitoring, supply chain visibility dashboards and electronic communication platforms such as AS2 or XML.

Actionable Considerations

In addition to the challenges with respect to cost, performance and standards, RFID planners would do well to take into account the following:

  • Understand the physics of RFID before jumping in. The technology is undergoing rapid changes, and some applications just won't work using "standard approaches."
  • Get a 360-degree view of RFID's expected impact across the enterprise - Particularly in terms of how RFID will affect your people, strategic business processes and existing technologies. Recognize that you won't realize the benefits of RFID if RFID cannot be interpreted correctly.
  • Plan based on the fact that RFID isn't simply about collecting data. It's about leveraging that data for real-time knowledge. Expect, at many stages, to be swamped with a deluge of irrelevant data, and plan to have the capabilities necessary to manage that. Take an extended look, if necessary, at how similar challenges have been addressed in other sectors - such as trading system development among investment banks and network management applications in the telecommunications industry.
  • Recognize that the middleware and complex event processing markets are still developing. One of the core premises behind any RFID strategy is that key facts (such as a rapidly closing window of opportunity or a critical business process mistake) are most valuable when they're accessible on a shared basis in a timely (as in "real time") manner. Another core premise? That the software enabling this is robust, reliable and well proven. The former premise is valid. The latter is not.
  • Acknowledge that RFID systems have a unique set of data integration issues that are different than enterprise application integration (EAI). Why? Most traditional applications and database systems are not designed for the capture, aggregation and distribution of real-time data or interfacing with third-party service providers.
  • Mind the security and privacy issues. Be particularly aware of RFID's security and privacy considerations - not just as they pertain to the exchange or transfer of data to external parties, but also as they relate to unauthorized exposure, usage or compromise.
  • Consider stepping into RFID through an operational industry supply chain network program. Consortiums can drive down the price of RFID efforts while shortening the learning curve.

One last point. Remember that by itself, a strategy-driven approach to RFID isn't enough, anymore than good RFID system integration capabilities will be effective in the absence of clear goals.
The most successful RFID implementations emerge when the approach is engaged from a wide-angle enterprise management perspective, one that's focused on process first and technology second. And one that is supported by the right mix of expertise in IT security and data management, business process efficiency and enterprise performance improvement.


Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access