The test targeted two areas of the supply chain: The partners applied battery-assisted passive (BAP) RFID tags with built-in temperature sensors to 60 cases loaded on a single, bar-coded pallet. Each case was tagged because the cases were bound for different customers via the national distribution center. The pallet was tracked with the existing system as it moved through the supply chain. The partners also placed the same type of RFID tags with temperature sensors inside 40 freezer cabinets at select stores.
During the initial
phase of the test, the tagged cases of ice cream moved from the factory to Unilever's national distribution center and on to the retailers' distribution centers, then to about 40 retail shops. The 60 tags were interrogated twice in the course of the test—when they were activated, or switched on, at the ice-cream factory, and when the tags were returned to Unilever by participating stores. The purpose of the reads was to collect sensor data recorded every 10 minutes, and to make sure the sensors were working properly in a deep-frozen environment.
The temperature-sensing, battery-assisted RFID tags, known as MTsens, are made by Montalbano Technology. Operating at 13.56 MHz and complying with the
ISO 15693 standard, the tags acquire temperature at a programmable rate, storing it on internal
non-volatile memory. (The tags use a battery only to acquire and store the temperature data, and data is transferred only when the tag is interrogated.)
Proximity RFID interrogators, provided by Montalbano Technology, were used during the proof-of-concept project to program, initialize, activate and deactivate tags. They were also used to
read thermal history and search for temperature recordings made at different times.
"Throughout the test, there were no breaks in temperature data, despite the severe environment," says Daniele Grosso, general manager of Montalbano Technology. "We were able to work in -30 Celsius temperatures for the whole battery duration, and each tag acquired more than 400,000 time-temperature readings. In addition, MTsens is very accurate in measuring time with a 0.006 percent accuracy."
Although Unilever Hungary was positive about the hardware and software test and says tag prices are decreasing steadily, thus reducing the barriers for larger tests, the company will still take years before implementing a tracking system able to provide real-time information on storage conditions. That, Buzinkay explains, is because tagging several million cartons of ice cream every year would be too expensive at this point. Furthermore, he adds, the tags should be read at the end of the supply chain so the company can determine what happened to the product the consumer finally receives. Since Unilever Hungary has 18,000 points of sale, Buzinkay notes, the tag interrogation step would require significant investment.