Radio Frequency Identification

How and when it will impact the gases and welding industry

By Scott Ernschwender

Technology and warehouse distribution followers cannot pick up a current magazine issue without reading about the hottest topic in supply chain management, Radio Frequency Identification (RFID). From mass merchandisers and grocery stores to aircraft companies, small parcel delivery carriers and pharmaceutical firms, RFID trials are underway, in search of the most efficient way to track product as it moves through the supply chain. Given the breadth and scope of the RFID trials and research, how and when will this sweeping technology change our lives in the gases and welding industry?

As with many new high investment technologies, our industry will initially be affected only at its edges, such as large welding equipment manufacturers who sell pallet load quantities to national retailers or the Department of Defense. As RFID technology matures, however, and as RFID tag prices drop, our industry will have an excellent opportunity to track and account for its largest asset, cylinders. If the cylinder population could be tagged for approximately $.05 to $.10 each and scanned by automated scanners located at loading docks and truck gates, then the dream of inexpensive and automated cylinder tracking will have arrived. But before solving industry issues years down the road, let's look at RFID as it exists today.

RFID Technology
The RFID concept has been around since World War II, although too expensive to implement in industry. In 1999, two individuals at MIT founded the Auto ID Center, which created and patented the RFID process discussed below. RFID begins with an RFID tag, consisting of two parts; an antenna that can send and receive data within a set radius, and a microchip consisting of 90 bits of data, including the manufacturer's ID item number plus a 40 bit serial number that uniquely identifies the product. These data are called the “electronic product code,” or EPC number.

Similar to bar coding, manufacturers begin the process by embedding the RFID tag within the individual product, case or pallet of goods. As the product passes through the manufacturer's dock doors, an overhead RFID reader passively captures and activates each RFID tag on every product and passes this information along to a local server, which in turn passes the data through the Internet. For each occurrence that one or more of the tagged items passes an RFID reader, the data are passed on. Therefore, there would be a log entry for each product in each step in the supply channel, including a record of where the product is currently residing.

Each information server (called a “Savant”) covering a physical plant, whether a manufacturing facility, distributor store or end-user location, is linked to the Internet. The EPC information is passed from the Savant over the Internet to an Object Name Service database, which looks up the product's Web address, similar to how www.gawda.org is passed to a Domain Name Service which locates the GAWDA Web server address. All of the logged information is sent to and stored at the home server database address, creating an historical profile of each product. This is the data that can be mined by supply chain partners to locate the product's whereabouts and movement history.

In addition to product movement, the Internet home database of product information will also contain potentially dozens of product specifications which could not fit onto the embedded microchip, attributes such as expiration date, temperature requirements, “do not mix with” information (for drugs), manufacturer lot number, etc. The combination of knowing a product's location along with its expiration date, for example, provides an incredible opportunity to ensure that expired product does not reach the public.

RFID for Cylinders and Hardgoods
In our industry, this same product home database scenario could be used to hold a cylinder's history. This would include testing and maintenance dates, previous gas containment and “do not fill” warnings based on past content. The history would also include a log entry each time the cylinder passed by an RFID reader, whether at the filling dock location, the truck gate entering or leaving the filling facility, and possibly the customer's gate. This data then could be mined by the distributor not only for safety status checks, but to analyze cylinder turnover as well as settle any ownership issues. Currently, the data are obtainable from cylinder tracking software packages; however, each cylinder must be manually and individually scanned with a handheld scanner at each access point from a short range distance. Using RFID tags and automated scanning equipment at a dock door or gate could save hundreds of scanning hours.

Coming Soon?
Since automated cylinder tracking seems to make sense, why is it not yet available? There are several reasons:

1. The cost of the tags and the scanners are much higher than this industry will support. Over the next two years, both tags and equipment prices should drop dramatically, hopefully to an industry affordable level.
2. Scanning accuracy currently averages about 70 percent, unacceptable for any industry. Companies are finding issues with RFID tags near metal, liquids and other material that appears to absorb the transmitting signal. Each testing company must uniquely find the right combination of scanners and tag placement to improve the accuracy to 99 percent. For RFID to become mainstream, scanning accuracy must be near 100 percent right off the shelf.
3. Multiple scanners in one plant operating simultaneously often receive interference from one another.
4. Software is still developing to integrate the process of moving the scanned data to the Savant, as well as to the user's database.
5. There are still multiple communication standards for transmitting scanned data.

Each of these issues must be resolved before customizable solutions will be brought forward to most industries, including our own.

In addition to cylinders, RFID could be used in our industry for hardgoods. While extensive testing is being conducted at Wal-Mart, Target, Albertson's and the Department of Defense, I don't see RFID tracking coming to our channel in the foreseeable future. The expense of RFID tags, scanners and the software integration required to automate the process currently averages over $1 million per project. While this cost will continuously decrease for future projects, I do not foresee any reasonable ROI in the near or intermediate term. This is not to say, however, that manufacturers who produce product for large retailers may not be asked to supply RFID tags on pallets within the next 24 months.

All in all, RFID is an exciting technology. The breadth of new uses for RFID tracking grows daily, from implants in humans for security purposes to labeling all prescription drugs by 2007. For our industry, the technology also holds great promise, although perhaps several years from now. Nevertheless, I encourage distributors to discuss RFID technology, particularly as it relates to cylinder tracking, with their software providers, as advancement in this area will most likely come from within our ranks.