Mar 02, 2020Auburn University's RFID Lab has completed a proof-of-concept (POC) that it says proves blockchain can bring visibility to RFID-based data across a supply chain. The Chain Integration Project (CHIP) Initiative took place throughout the past year, tracking goods for five participating companies: three brands and two retailers. The result was that the blockchain network enabled these businesses to automatically share data regarding the flow of goods through the shared supply chain.
Based on those results, says Justin Patton, the RFID Lab's director, the use of blockchain with RFID technology could bring an end to claims related to missing goods in the supply chain, while improving supply chain efficiency. It could also create an accessible history of every RFID-tagged item from manufacturer to store. The lab has released a white paper describing the results.
Participants in the proof-of-concept were brands Nike, PVH Corp. and Herman Kay, as well as retailers Kohl's and Macy's. The companies collected and shared item-level data streams from RFID tag reads throughout 2019. They used software and hardware support from Avery Dennison, IBM, Mojix and SML.
The RFID Lab's POC follows Project Zipper, its 12-month pilot created to test the value RFID brings to the retail supply chain (see Project Zipper Finds Order Accuracy Jumps to Nearly 100 Percent With RFID). That makes the CHIP POC the second in a three-step process that Patton calls "crawl, walk, run," in which RFID management of supply chains could improve efficiency and eliminate waste. The third step, beginning now, will pilot blockchain data in real-world supply chain environments.
For the past 15 years or so, RFID technology use has grown in retail, as has its related benefits. Item-level tracking is taking place at many stores to provide in-store availability, and to enable omnichannel sales. Where RFID has had less traction, Patton explains, is in supply chain management, and that's partially due to the challenge of data management and data sharing between partners. Suppliers are not pre-sharing data for the RFID tags, he adds, so as an industry, "We're putting the tags on in the factories, and we scan them at the end of the supply chain, in the store, but we're not using them on the many points of exchange throughout the supply chain."
As a result, Patton says, although stores may know a product's location and status, they do not know its history. That lack of visibility in the supply chain introduces such problems as lost goods, along with claims related to ordered merchandise that go missing between factory and store. Similarly, he adds, advanced shipping notification accuracy is often poor, and the technology for data exchange is out of date or insufficient for current supply chains and technologies.
"That shared data-exchange medium is a huge need," Patton states. "We felt like blockchain was a good potential fit" to solve the problem." Mojix and Microsoft suggested the testing of blockchain technology to the RFID Lab in 2016, he recalls. The Auburn Blockchain Working Group then launched in February 2018 with brands, retailers and logistics providers. The group first conducted a survey to learn the priorities for those in the retail industry, which it found to be supply chain visibility, sustainability, consumer engagement and product authentication.
The working group then launched the CHIP initiative. Members organized the proof-of-concept, with PHV Corp. and Kohl's pairing to provide a single supply chain, and Herman Kay and Macy's providing another—both tracking goods through traditional wholesale channels. Nike provided the third channel throughout its closed loop supply chain, with tags being tracked at the point of manufacture through its own distribution center. PVH and Kohl's began capturing RFID tag read data at two DCs (one for each company). PVH provided two outbound read points at a single DC, while Kohl's contributed inbound data from its own DC.
Mojix contributed the technology to capture and manage the data streams. Herman Kay provided data from an outbound RFID read point at its DC, while Macy's provided inbound receiving data from a single DC for products destined for six stores. Herman Kay employed SML's solution to capture and manage RFID data, and Macy's partnered with Avery Dennison at the distribution center and Tyco at the store. Nike captured RFID data at the point of RFID tag encoding at the factory, then again at a distribution center, and leveraged its own technology team to manage the RFID data.
To standardize the data being captured from the readers in all three channels, the CHIP POC participants used Electronic Product Cole Information Services (EPCIS), which provides the format for the data across all participants in the blockchain. The companies employed Linux's Hyperledge Fabric blockchain framework on which the EPCIS-compliant data was uploaded.
Kohl's, Herman Kay and Nike sent their RFID data to a Web application known as the Translator Tool, developed by the CHIP team, which identified key data and transformed it into EPCIS-compliant data. PVH Corp and Macy's generated their own EPCIS-compliant data using their own teams or solution providers, in order to enable that process, then forwarded the information directly to the blockchain.
The CHIP team utilized three layers in the blockchain technology stack. The first primary interface and ingestion point was the CHIP Translator Tool application. The second was the Blockchain Client, also developed by CHIP, which connected the Web application to the third layer, the Blockchain Platform. Each channel has its own independent side chain within the platform, exclusive to that channel's participants.
To use the system, a company could make a query, such as an order of goods. This would be captured by the Blockchain Platform's Ordering Service, which organizes orders and distributes them to the appropriate channels, where they can be viewed by members of that channel. In the meantime, each read event generated data that was then forwarded to the platform, also through the Ordering Service. Thus, each time a tagged product left the brand's distribution center, that data was shared with the retailer. When the retailer's DC received the information, it was then shared back to the brand and could also be shared down the channel to the retailer's stores.
Once all participants had paired their data streams into a single blockchain solution, the CHIP team activated the blockchain process and data began flowing through the network. Some participants already had fully deployed RFID technology at their facilities, while others had established new systems for the POC. For that reason, Patton says, the read data was not intended to analyze the accuracy of the supply chains, but rather the data that could be captured. The POC also aimed to track the lifecycle of individual items so that users could view an item's history by simply reading its RFID tag and accessing its history in the blockchain network.
The first product to pass through the entire supply chain was a coat—a Michael Kors parka from Herman Kay. That was a big moment, Patton says, that he likens to the first scanning of a barcode in 1974 on a pack of Wrigley's chewing gum. The parka's date and location of departure at Herman Kay's distribution center, its arrival at Macy's warehouse and its arrival at the store were all visible in the blockchain, and were also available to the channel partners. "That first time we saw data transfer automated was amazing," he states. "It proved that the data was taking on a life of its own."
According to Patton, the POC's results, which the lab describes in its white paper, demonstrate that the blockchain network provided automated data exchange for RFID-tagged items throughout the supply chain. The next step for the RFID Lab is to conduct pilots with the existing participants, as well as other brand and retailer members of the CHIP team. In the case of a pilot, he says, the data will be compared against supply chain channels not using blockchain to view the impact such a system may actually have on claims, or other business value it might provide.
"A shared data-exchange medium is a huge need," Patton says. From here, he adds, companies need to consider how to ensure the right data is being fed into the blockchain in the first place. Because there will be exponentially more data, Patton notes, companies must examine the data at its origins. "The first thing we learned is this puts a spotlight focus on our ability to properly encode each item. We're really good at capturing data at the store." Where development is still needed is in the capture of data at warehouses, Patton says. "We're not as good at it, because not as many have been doing it."
"The takeaway from the POC is, it works," Patton says. He sees a scenario approaching soon in which "We will be living in the future with serialized items through the supply chain and no one in the middle who owns all the data." But good data, he notes, is just the beginning. "It takes a bit of work to get there, but that future everyone has been imagining and wondering how to get there—we're here. It's achievable and this is the way to get there."
The lab's next goal is to have data from the pilot, which is starting now, by the end of this year. Not all participants have yet been identified and named, Patton reports. "There's a huge amount of interest from new folks" who would like to participate, he says.