RFID Demands a Measured Response

By Oliver Hedgepeth

As RFID adoption grows, we'll need new metrics to understand and fully realize the technology's benefits.

  • TAGS
image_pdfimage_print

A bar code can add value to a product when recorded as a product count, usually as a single custody event, at a single location, isolated in time. An RFID tag can add value to that product by describing its chain of custody from birth up to the current point in time, and maybe even forward to the next product disposition event. With bar codes, you think about what to count; with RFID, you think about how much more to count. And for any technology, the first step in this thinking process of defining a new metric, or measure of performance, is to understand clearly the functions the technology performs.

A bar code label may display only a product’s invoice number, whereas an RFID tag can tell the story of where that product was born, where it has been and where it is going, and it may also contain an alert notice of when to drop the product’s retail price. During the RFID Academic Convocation hosted by MIT on Jan. 23-24, 2006, the need to start analyzing the assumptions about current and new metrics was discussed, if not all too briefly, by a few attendees in response to repeated questioning.




Some attendees indicated RFID would change nothing about the units currently used to measure supply chain performance, such as weeks or months of inventory held in a warehouse, the number of days customers must wait for product to arrive, or the days or weeks required to process product requisitions. Others described RFID as just the latest technology that would fade from memory in three or four years.

In the 1960s, similar statements had been made about the impact of computer usage, at a time when decision-making started to shift from using a computer as a high-speed calculator to using it as a tool that could model manufacturing processes, examine complex weapon system metrics, and give recommendations on which truck or weapon to purchase. What changed from the 1960s to the 1980s were some of the old metrical units of weeks and months. The previous units defined a perceived need for computers to simply summarize logistics supplies in a warehouse. Today, however, units such as nanoseconds are the movers of global wealth.

The metaphors and paradigms describing the changes technology brings to our data and information-demanding society seem to shift first, before the metrics and their units shift, when a new technology enters the business stream. Do you remember the record player? It’s been replaced by the compact disc (CD) player. Revolutions per minute (RPM) is a term the music industry once used while producing analog vinyl records; recordings, now digital, are measured in kilobits per second. Just as the automobile used to be called the horseless carriage, so is RFID now being hailed as wireless bar codes.

Presently, we measure inventory in units of months or days, not hours, minutes or seconds. RFID usage could change that, depending on how retail stores use the technology. The stage is set for RFID to shift some metrics, refining the units of measurement as companies modify some of their operations to add RFID, replace bar codes, redefine the workspace and workers’ functions, and update their computer middleware and product communications protocols.

At the convocation, Simon Langford, Wal-Mart Stores executive VP and CIO, said RFID’s current benefits are just the “tip of the iceberg.” The history of computer technology’s insertion into logistics and supply chain activities seems to create such icebergs of benefits, as we have seen from the 1960s to the present. And with the global complexity of Internet-based data transfers available to today’s businesses, the logistics threads—the critical arcs and nodes in the supply chains—may produce a new set and quantity of data, as well as benefits. How to manage and measure the use of the emergent RFID databases is still an unknown. Given that bar codes speak a dozen times in a product’s lifetime, while RFID codes speak 200 to 1,000 times a second, the thread between measuring RFID supply chain performance and achieving supply chain management benefits may become a study into complex-systems behavior.

A few standard metrics for bar codes do exist, such as product number and retail price. There may possibly be hundreds of metrics, meanwhile, for an RFID tag’s encoded data. These include a product’s date of manufacture, its supply chain entry and exit points, the last location it was processed and its on-sale and expiration dates.

So, how do we find these metrics? We start with a systems approach, looking at the RFID functional and technical parameters, the goals and objectives of their use. The cybernetic landscape—that is, the communication links and threads along the supply chain network arcs and nodes for the flow of goods, information and cost—should help us examine the span of a product’s supply chain, from its strategic production-planning system and its production to the dock doors in a distribution center, and on to the point of sale where a customer purchases it.

Metrics for RFID, then—much like transportation—are set to be a derived demand. Transportation engineers discuss how people’s desires for goods and services actually create a demand for transportation, and how communications is a replacement for transportation. We see today that the desires of people, or of organizations such as Wal-Mart or the U.S. Defense Department, for goods tracked and traced in real-time is creating a demand for new metrics, new units of measurement, new understanding of the concept of supply chains, and new concepts of outsourcing logistics services. As such, RFID is creating a new demand, and metrics is a key part to understanding and responding to that demand. Communications technology such as the cell phone is essential to transportation and logistics services, and possibly as a new partner with RFID. New laws on the use of the cell phone are emerging—and the freedom of using your cell phone has new boundaries on usage. What will those boundaries be for RFID five years from now?

One search for metrics is to examine the use of RFID at the intersection of its technology, the analysis of RFID functions and its actual application. But how?

The first step, based on RAND’s Assumption-Based Planning (ABP) method for analyzing strategic operational military plans, is to examine a metric’s assumptions. One assumption is that inventory levels will still be counted on the current monthly or weekly basis. Will this remain the metric and unit you use when you have second-by-second visibility of 120,000 retail items in one store—and more than 5,000 retail stores?

The second step is to identify the vulnerabilities of those assumptions, while the third is to identify any warning signs signaling that the assumption or one of its vulnerabilities is becoming invalid. Some version of ABP has proven to work in areas of great uncertainty, such as going to war, drilling for oil or developing a global supply chain.

In a warehouse, you might employ a worker using a hand-held bar code scanner to search for bar codes on multiple boxes stacked on a pallet. If each box had an RFID tag on it, then an RFID interrogator placed above the dock door could search your entire inventory and update the records accordingly. Your employee’s time would be saved; the accuracy of the data would be preserved; and shipping dates would be more accurate. Now, your employee could move on to some other operational function.

We already have a controversy over what seems like simple metrics when processing products as they move in and out of a warehouse or a distribution center. We have differences of definition over such simple terms as “shipping date.” The units of this metric are the day, month and year, and, sometimes, the time—but determining when and where this event takes place is still not performed in a standard way in the logistics industry.

The process of analyzing where to place a reader might create new metrics. If the reader is on a dock door, on a forklift, in a handheld or on a wearable attachment, does that mean we will have new metrics? Certainly, it would be possible to have new metrics for tracking goods as they move through a distribution center.

One new idea for using RFID technology for the milk industry comes from Matanuska Maid Dairy in Anchorage, Alaska. Joe VanTreeck, the company’s president and CEO, says he would like to have RFID tags identify temperature problems in order to prevent premature spoilage, thereby becoming a tattletale for temperature abuse. Similar ideas are emerging all across industry and the military. One possible metrics decision may involve how often that tattletale wakes up and checks the milk supply. Should this occur daily? Hourly? And, who receives this data—the consumer, the retail store’s manager or the milk company?

Metrics—whether quantitative, such as the monthly inventory count of tires, or qualitative, such as customer confidence—arise with the emergence of a new technology. Without questioning metrics, however, business-decision makers’ expectations of RFID’s benefits may lead to new problems instead of to new opportunities for improving visibility into our complex, global supply chain network.


Oliver Hedgepeth is the chairman of the logistics department at the University of Alaska Anchorage (UAA). He was also the first director of the Artificial Intelligence Center for Army Logistics.