Jun 01, 2010RFID-tagging and tracking individual DVDs through the supply chain could improve in-store promotions, reduce out-of-stocks, and curb counterfeiting and theft. DVDs are commonly packed and shipped in cartons that hold 40 discs, and a pallet typically holds about 50 cartons. In 2009, EPCglobal asked the Auto-ID Lab at the University of Adelaide, in Australia, to investigate the feasibility of reading roughly 2,000 RFID-tagged DVDs stacked on a pallet.
|
|
The first step was to determine where to place the EPC RFID label on the DVD case. We achieved maximum readability by folding the label around the spine in the lower left corner. This keeps the label away from the disc, to avoid disturbing the reader field.
DVDs are always placed in the carton so that the spine of each case is perpendicular to the base of the carton. The cartons can be stacked on their sides or bases, so we explored both possibilities.
It became immediately clear that it's not possible for a fixed or handheld interrogator to be far from the stack and read the interior labels from a single position, because of interference from the DVDs and the labels on the front. Instead, it's necessary to read the labels from close up, front and back, scanning each surface horizontally and vertically. This can be achieved by having the pallet pass through a portal with moving reader antennas on both sides, or by having workers with handhelds read the cases from multiple positions.
|
|
We also tested the optimal way to stack cases to achieve high read rates. When 45 cartons (1,800 DVDs) were stacked on their sides—three high, three wide and five deep, with the carton openings facing outward—we achieved a read rate of 96.25 percent. Our results showed that side stacking is not expected to achieve 100 percent read rates, because the reader field has to pass through two rows of cartons to reach the middle labels.
When we stacked 48 cartons (1,920 DVDs) on their bases—four high, three wide and four deep—we achieved a read rate of 98.75 percent. These results were obtained with the spines of the cases in the front two rows facing forward, and those in the back two rows facing rear.
We attributed the failure to read all tags to the fact that some tags were weak responders. Even better results could be obtained by using tag quality control, or developing tags designed specifically for folding across the spines of DVD cases; the tuning of such labels could be adjusted to allow for the proximity of the folds.
The study shows the importance of a disciplined approach to stack structure, and in the comprehensive report we prepared for EPCglobal, we recommended use of the base-stacking method. Increasing stack height or width (but not depth) should make it possible to read more DVDs at once if necessary.
Peter Cole is the research director of the Auto-ID Lab at the University of Adelaide, in Australia. Zhonghao Hu is a research associate at the lab and a postgraduate student in the university's School of Electrical and Electronic Engineering. Illustration by John Hull.