No other trend in technology evokes more discussions, questions and general confusion than radio frequency identification (RFID). While RFID is the first technology adoption driven by corporate mandates, many companies are discovering value in the detailed information that RFID delivers. Much like the efficiencies gained in the last 30 years with the adoption of bar coding, RFID is delivering competitive advantage for organizations willing to work through the technical challenges and move forward to leverage the data for fundamental business model change.

The early adopters who focused, albeit out of necessity, on the physical elements of RFID were deploying the technology without considering the business objectives of the enterprise. Today, organizations are focusing on RFID-generated data and the problems it can solve across an entire enterprise.

What is changing about RFID is the approach to return on investment (ROI) realization. Today, many RFID early adopters are asking, "How can we speed up the ROI capture and increase the adoption of the technology?" The answer is to flip the model around (see figure 1) and focus on defining enterprise-wide business challenges, evaluating whether RFID can address them, defining the metrics needed for success, and then sourcing the hardware to capture the data to address the problem.

A new point of diminishing returns
Let's face it, today we are hitting a point of diminishing returns when it comes to the data we have to analyze. Best-in-class organizations have consolidated their enterprise data into vast data warehouses and are running at near-peak efficiency. The next revolution for business will involve unleashing a wave of serialized event-level data from RFID (see figure 2).

RFID offers two key ROI tracks
There are two ways to capitalize on RFID data. The first involves the speed in which it is captured. Many of the analytics and business challenges RFID addresses are only solved if the read data is acted upon in near real time. The second way maximizes the advantage of unique serialization to capture and act upon detailed data.

enlarge To be successful, RFID projects must move from the current model of "hardware first" to a more classic model of process change driving the hardware and software requirements.

enlarge When RFID is utilized, analysis of the enterprise data gains additional value.

1) A case for speed: RFID offers a greater view of a business' supply chain (from raw materials to a customer's purchase). With the ability to actively monitor, analyze and act on events as they happen, RFID users can move from problem correction to problem avoidance. Consider in this example how, through RFID, a food retailer can avoid taking possession of a defective product.

The setup: The ice cream maker places active RFID tags on the shipping pallets when the product is packaged and tracks the temperature every five minutes.

The situation: The manufacturer receives notification from several retailers whose customers are complaining that a particular brand of ice cream had thawed and was refrozen at some point in the supply chain. This caused the product to separate, dramatically altering the taste. Retail sales of this brand of ice cream are dropping and customer satisfaction is at risk.

The solution: Through RFID, the manufacturer discovers that the spoilage occurred on the store dock; the product was left too long in the sun before it was moved from the dock to the freezer. The manufacturer, retailer and logistics provider all share RFID data. The manufacturer can set an alert (for example, via Teradata Supply Chain Intelligence) to monitor the movement of the ice cream in near real time as it travels from manufacturing to the various distribution centers (DCs). The manufacturer can identify where the temperature violation occurs and inform the retailer so that the bad product doesn't reach the consumer.

Likewise, if a retailer realizes through RFID that spoilage is occurring earlier on in the supply chain, the retailer can inform its DC not to accept receipt of any future bad products. By stopping the inbound damaged product from reaching the shelf, the manufacturer and retailer act to protect their sales and customer satisfaction. This same application of RFID technology can be applied to many perishable or sensitive products. Conversely, other challenges are best addressed with extensive analytics when an enterprise view and adequate history are available.

2) A case for deep analytics: Strategic decisions cannot be made in isolation. RFID data should be reviewed across the enterprise to ensure the decision will yield the best result—in many instances, this is the same when using RFID data to solve challenges within an organization.

The setup: A semi-annual category review between a retailer and a category captain; specifically, a manufacturer who agrees to work closely with the retailer to help drive the overall growth of a category, not just his or her own.

The situation: For more than 20 years, the retail and consumer product goods industries have practiced various forms of category management to increase share of market basket for all suppliers in the category. A market basket is the level of detailed data captured at the point of sale (POS) for each shopping event.

While this was the first time collaboration was seen as a way to drive business, the lack of accurate, timely and agreed-upon data nearly killed the inherent benefits to both parties.

The solution: Category management involves managing four key aspects of the business: assortment, promotions, pricing and shelf space. As each area is analyzed, more accurate decisions can be made using RFID event-level data:

Assortment: Decisions are made by the sales volume of a product as well as a unique selling position—event-level detail on product movement over the last year helps reduce the chance that a product is cut from the assortment because of an isolated incident in supplying the retailer.

Promotions: Many of the promoted items are also new items introduced into the marketplace. The one-year failure rate for new items exceeds 60%. By using RFID event data, the best stores, best times, best locations within the store and even the best crew can be determined to help ensure the success of a new product.

Base pricing: Detailed RFID event data for inbound product movement allows optimal prices to be set based on inbound product availability and margin requirements from the retailer.

Shelf space: As item-level tagging proliferates, shelf sets (or planograms) can optimize product affinities and enhance the shopping experience by providing additional information to the consumer at time of purchase. Detailed shopping flow patterns can also be analyzed (by tracking shopping carts with active RFID tags) to ensure categories are effectively located throughout the store. For example, does locating the beer category near salty snacks drive best-customers' purchase behaviors or make it easier for cherry pickers to get in and out without buying additional items?

By using RFID event-level data in the category review process, the retailer and the suppliers can optimize the store experience for the customer, driving up sales, margins and customer satisfaction.

In the end
How these two ROI tracks are optimized also depends on the willingness of various points in a supply chain to share information. The availability of RFID is beginning to blur responsibilities among manufacturers, logistics providers and vendors as each link in the supply chain becomes more transparent and takes on a greater level of ownership and responsibility for the final product.

This transparency will be driven by open standards for data exchange, detailed analytics throughout a product's life and the ability to act rapidly as problems arise during the movement of goods. Much like the illogic of isolating RFID from the rest of the enterprise, it will be impossible to fully realize optimal benefits if data is not shared among trading partners. And while exchanging data occurs today in many forms, the growth and demand for RFID data will drive the appeal for data sharing to higher levels.

Data sharing is undergoing an evolution in the retail and consumer-goods supply chain. Today, one of the biggest assumptions is that RFID data will be shared among the makers, movers and sellers throughout the supply chain.

As global data sharing standards are defined, the question is, who will pay for these services? Will the EPCIS (see bottom bar) be a central holding location for information, or will it serve only as a traffic cop for requested information?

History points away from an altruistic, global network to one or more networks driven by relationships and fees paid by each partner for access to other partners' data. As data keeps growing in importance as a competitive advantage, the trend will be to have very selective sharing for limited mutual benefit. The successful will use information to drive out competitors by more effectively managing business processes. Future business alliances will be heavily influenced by the level of sophistication in data acquisition by all parties.

Technologies don't evolve overnight, and anyone who remembers the slow adoption of the Uniform Product Code (UPC) will appreciate the historical precedent set that RFID will follow. It was more than 10 years after the standards adoption in the early 1970s before the vast majority of products contained UPCs and retailers had scanners to capture them.

Many of the things we take for granted today would not be possible without the data generated by the UPC. RFID and the data it creates should experience a similar growth curve and, in the end, will become an indispensable part of doing business—for everyone. T

Stephen Brobst, CTO for Teradata, is responsible for the research of strategic new technologies and products.

Richard Beaver, director of Global RFID Strategy at Teradata, is responsible for optimizing RFID data integration into the larger enterprise.

Illustration by Jeffrey Brown

Understanding the physical layer of RFID enlarge Radio frequency identification (RFID) is a technology that uses radio waves to transmit information over distance. The physical layer of RFID involves several components. These components contain and convey the serialized information necessary to feed meaningful information to the data warehouse. RFID tags or transponders—There are two main types of tags, passive and active: Passive tags require no internal power source, can be very small and have an unlimited life span. Electrical current induced in the antennae by the incoming radio frequency signal provides just enough power for the integrated circuit within the tag to power up and transmit a response. Active tags have an internal power source. They may have longer range and larger memories than passive tags, as well as the ability to store additional information sent by the transceiver. At present, the smallest active tags are about the size of a coin. Active tags frequently are the size of a deck of playing cards or larger. Many active tags have practical ranges of tens of meters and a battery life of up to five years. They can also capture environmental information like temperature, humidity and pressure. These tags are typically used today on high-value merchandise or reusable assets. Readers—A tag is useless if it cannot be read. An RFID reader is a device that is used to interrogate an RFID tag. By sending out a radio signal at a specific frequency, the antennae can read the information in the tag and then send the data to the reader. Operational server—The operational server is located with the middleware and local apps—it is the physical device holding the software. This device has many names (edge server, premise server) but, basically, it is a small computer server in the operational environment (store, distribution center or assembly plant) that controls all the readers and collects the data they gather. Middleware—The final piece of the physical puzzle is the middleware. This is software that controls the described devices. The middleware also sends captured data from the many readers back into various systems (enterprise resource planning, operational or enterprise data warehouse (EDW)) where robust analytics can be conducted. —S.B. and R.B.

Enabling better customer service with RFID One of the most hotly discussed and least understood developments to come out of RFID is the concept of the Electronic Product Code Information Service (EPCIS). The EPCIS is a specification for a standard interface for accessing EPC-related information. Because an EPC gives individual objects a unique serial number, each one can be tracked independently, and fine-grained real-time information can be collected, stored and acted upon. Electronic Product Code Information Services are a way for supply chain partners to share and exchange information efficiently. Because a standard interface allows trading partners to use the same functions or methods for querying data across the supply chain, less time is spent on integrating with partners, even if they store information in different types of underlying databases. This service will be a key component for the sharing of data between retail trading partners in the future as RFID proliferates across an enterprise. enlarge This diagram represents the retailer's use of EPCIS in a Teradata environment. Each retailer shares data from its data warehouse using protocols established by the EPCIS. These protocols control security and access rights to various data and other items. In this example, a customer returns a damaged product. The retailer can scan the RFID tag, which contains the unique EPC or product serial number, and then query the EPCIS (and other systems) to: > Verify that the product was sold by this retailer and not stolen along the supply chain > Verify warranty coverage or extended warranty purchased > Seek a pedigree of the points along the supply chain the product traveled > Notify the manufacturer of the damage and instantly process a return > Identify whether additional inventory is available to replace the product > Add the customer experience to its customer relationship management (CRM) records for future follow-up Clearly, the RFID serial number allows a new level of customer service, helps identify where failures occur in the supply chain and detects one type of pervasive fraud. In this example, understand that while RFID tag data was used to kick off the process, the retailer had to interact with data from other systems, including inventory, customer and supplier. Some data would flow through the EPCIS, but portions of it would be accessed via the retailer's internal enterprise data warehouse (EDW) and operational systems. —S.B. and R.B.

� Teradata Magazine-September 2006