RFID data cleaning tips
RFID (Radio Frequency ID tags) are again in the news. The FDA has recently scolded the pharmaceutical industry for delays in implementing RFID track-and-trace technologies. Pressure is also growing for using RFID tags in consumer products, clothing, and industrial equipment for asset management. RFID tags significantly increase the volume of shop-floor-to-ERP communication, but RFID experience...
RFID (Radio Frequency ID tags) are again in the news. The FDA has recently scolded the pharmaceutical industry for delays in implementing RFID track-and-trace technologies. Pressure is also growing for using RFID tags in consumer products, clothing, and industrial equipment for asset management. RFID tags significantly increase the volume of shop-floor-to-ERP communication, but RFID experience from Pfizer and other early adaptors reveal additional technological complications.
The biggest complication is the need for exceptionally clean data transfer from the tags to the reader. Integration of product level RFID tags into the supply chain often requires that received data has 6 sigma quality (3.4 defects per million), or even higher in the case of controlled substances. Achieving this level of accuracy requires that RFID data is cleaned before entry into the supply chain. Working against this accuracy requirement are at least three main factors: angles, interfering environmental materials, and phantom tags.
RFID tags include small antennas that pick up a signal from a reader and gather enough energy to send a response. In order to achieve this, the RFID antenna cannot be perpendicular to the reader. A stack of RFID tags perpendicular to the reader cannot be read and are a source of errors. But changing the angle by even 30% allows the tags to be read. Three redundant readers can ensure complete coverage, but most applications use only a single reader and rely on the tags to be correctly positioned.
Materials surrounding the tags will change the tags' ability to be read by interfering with the signal. In particular, many liquids or metals that are close to the tag antenna will change the frequency the tag is sensitive to. For example, a typical RFID tag on a bottle of mouthwash cannot be read no matter how close it is to the reader's antenna. Even worse, the material interference can come from production equipment.
Phantom tags are another problem with clean data gathering. These are RFID tag reads that are not associated with the product but come from unidentified sources in the area. They might come from tags on operator's clothes, production equipment tagged for asset management, or even other products from nearby pallets or containers. With the growing use of RFID tags in all industries the problem of phantom tags will only get worse.
Cleansing problem data
These problems mean that there is a requirement to cleanse RFID reader data. The best practice is to clean and verify the data as close as possible to the source, while the product is still readily available and operators are nearby to take corrective action. This implies that there is a layer of real-time software between the readers and the supply chain ERP software.
The data cleaning software may stand alone or be part of a SCADA system, but it must know what it expects to read. For example, the data cleaning software should know how many tags should be on a pallet or carton. If fewer tags are seen than are expected, then an alarm can be raised to operators to manually verify the actual inventory. If some of the tags are outside of the expected range of IDs, then the tags could be phantom tags, or could flag incorrectly packed product for rechecking. Because the number of tagged items per pallet and the expected range of tag IDs may vary per pallet or carton, the data cleaning software may have to communicate to a recipe system. The data cleaning software should also trend error rates in order to detect changes in the environment that can impact tag readability and help identify phantoms. For example, a new production line added next to an existing line may cause enough radio interference to increase errors. Moreover, some types of material stored nearby, even if not obvious, may increase errors.
Dennis Brandl, firstname.lastname@example.org , is the president of BR&L Consulting, Cary, N.C., which is focused on manufacturing IT solutions.
|Search the online Automation Integrator Guide|
Case Study Database
Get more exposure for your case study by uploading it to the Control Engineering case study database, where end-users can identify relevant solutions and explore what the experts are doing to effectively implement a variety of technology and productivity related projects.
These case studies provide examples of how knowledgeable solution providers have used technology, processes and people to create effective and successful implementations in real-world situations. Case studies can be completed by filling out a simple online form where you can outline the project title, abstract, and full story in 1500 words or less; upload photos, videos and a logo.
Click here to visit the Case Study Database and upload your case study.