For everybody implementing — or thinking about implementing — radio frequency identification (RFID) technology, Tim Crane, principal with consulting firm Computer Sciences Corp.'s Supply Chain Solutions group, has this simple advice: Stay flexible.
Plan to have multiple technologies doing different things within your own operation, expect to replace equipment within a year because of technological advances, and keep in mind that the middleware you buy today to link your RFID tools to data management applications will be incorporated into those applications within three to five years, so plan on throwing that out as well.
Only 53 of the 137 suppliers who were selected or volunteered for Wal-Mart Stores Inc.'s RFID initiative met the January deadline for compliance. Nearly 100 said they would be compliant by the end of February. That may sound like limited success for the technology, but in the decades since it was developed, RFID has made quiet inroads in data collection and management. Its recent "celebrity technology" status may be affecting many attitudes and some critical judgment when it comes to applying the technology.
It's true that RFID is not as mature as some other data management methods currently used in warehouses and along the supply chain. That's a good reason to limit exposure to hardware, according to Crane. But on the flip side, he strongly urges developing a strategic vision of how the technology will be employed and targeting some applications with a short payback.
Before getting into the strategic vision aspect of the discussion, it might be helpful to understand some trends in data collection, and specifically RFID. There is a trend, Crane notes, towards adding active memory to tags to reduce dependence on network solutions. The advantage of writing more information to a tag and adding to that information as it moves through the supply chain means that more of the intelligence accompanies the product.
Here's how it works:
A fixed identifier on a pallet or case has to be associated with other data in a real-time database to provide anything more than basic information (e.g., "This carton contains blue, short-sleeved shirts").
A serialized ID can help link a carton to more information (e.g., "This carton containers blue shirts from lot number 2 that were part of shipment A to Acme Co.'s Poughkeepsie store").
RFID tags are serialized, so they start out being more specific. An active tag can have information added as the tag moves through the supply chain, so the same information that is in the database can be written to the tag — maybe not as elegantly yet, but that's coming. Still, actively associating a carton to a bill of lading or PRO number so that anyone along the supply chain can know instantly where that carton belongs means less dependence on a data network.
Picture a truckload of pallets where 10% of the cartons are the subject of a recall. Reading the carton IDs on each pallet as it is unloaded can tell you whether that individual pallet contains any of the recalled product. Those with no recalled product can continue moving directly into stock without delay, and the pallets with recalled product can be culled at the dock.
Business solutions are being lost in the need to be compliant, says Crane. But in looking for the benefit beyond compliance, he warns that this is a technology where one size does not fit all. Applications, environment and the product itself will determine the best fit.
Even in the high-profile areas that are getting current media attention, needs vary. Retail has issues with the cost of tags. The U.S. Department of Defense needs to communicate more complex data, and it has many more vendors as well as international operations to consider. Healthcare is currently focused on drug distribution, but it has many more areas where the technology can provide benefits.
One-way packaging in a retail setting has a different set of requirements and cost tolerances than a reusable tote in a closed-loop replenishment operation. Thinking strategically and focusing on what the "end state" will be will avoid potential redundant systems, says Crane.
He describes a pick-and-pack operation where items are picked to a tote that moves along a conveyor. The density of the product makes it difficult to read every tag in the tote at the end of the sequence. A series of strategically placed RFID readers could read the top layer of product and the tote ID each time it passes a reader, confirming accuracy of the pick-and-pack sequence while providing a final, complete list of the contents of each tote as it comes to the shipping dock.
Current technology can use triangulation to determine a tag's location. A product (or a person) can be tracked as it moves through a facility. If a high-value product is headed for open storage, the exception can be noted in real time and dealt with. If a lift truck or driver is headed into an area where neither is authorized, gates or doors can be programmed not to allow entry or a system on the lift truck could shut down the vehicle and send notification to the appropriate party.
Aircraft companies like Boeing Co. — an early adopter of RFID technology — face a significant challenge tracking specific parts that are assembled and installed on an aircraft. As the aircraft is built, an assembly might be removed or a different part installed. When a section is completed, RFID tags on the various parts can be read during a "walk through" and the specific parts and components that are actually installed become part of the assembly documentation.
In daily airline operations, the serialized RFID tags allow a luggage cartsized device to be rolled down the aisle of an aircraft required to carry life jackets and read the individual identification of each life jacket to ensure there is one under each seat. This has significantly reduced-turnaround time in ground service and helped improve on-time departures, according to Crane.
But if airlines have recognized some areas where they can apply the technology for benefit, they haven't always recognized when the same technology can be used for multiple purposes. Using a system to identify and track baggage that is loaded on an aircraft can save the airlines millions of dollars. That same system can track catering carts that provide-the food service items for the flight and the "flight bag" materials used by the aircraft crew. Yet, says Crane, he saw an airline looking at a system to track flight bags without recognizing similar systems were being developed in parallel.
This speaks to Crane's point of visualizing the end state early in the process of developing RFID applications. The airline in his example could have built redundant systems at a very high cost and seen little or no return on investment (ROI) from either application. But don't expect to accomplish everything at once. Visualizing the goal will help not only with the implementation of solutions but also with identifying those opportunities with short ROI.
The short ROI is important initially, according to Crane, because of the high degree of obsolescence and change that will occur in this area of technology in the near term. Since Wal-Mart began its initiative, for instance, the second-generation RFID tag protocol (Gen 2) has been accepted by EPCglobal Inc. and will be placed before the International Standards Organization (ISO) to move it towards becoming a recognized global standard.
Develop solutions that provide benefits but limit risk, advises Crane, especially exposure to hardware. Pilot projects are needed, and they can take four to six months. Roll out can take another six months. Very few companies have budgeted for assessments or pilots, so it is increasingly important to look for places to apply the technology in a self-funded process. But in doing that, keep in mind the hardware will likely be changed in a couple of years. You'll also be throwing out your middleware that links data collection to the execution systems and enterprise systems as those systems suppliers incorporate the functionality into the larger application.
Expect the hardware companies to keep pushing the capabilities of the technology and expect consolidation in the industry, suggests Crane.
Computer Sciences Corp.
International Organization of Standardization
U.S. Department of Defense
Wal-Mart Stores Inc.