What successes have you been achieving so far?
Schneider: To date, the first prototype, which was developed by Pellenc ST and Digimarc and which combines the digital watermarks technology and NIR/VIS infrared for sorting of packaging waste, achieved a >95 percent ejection rate. This prototype was installed in the Amager Resource Centre (ARC) in Copenhagen for the semi-industrial test phase. For four months, trials and demonstrations with around 125.000 pieces of packaging representing up to 260 different stock-keeping units (SKUs), all prepared by HolyGrail 2.0 members, were held in Copenhagen. Engineers tested for several parameters including the speed and accuracy of the system, to ensure its ability to withstand the pressures of full-scale industrial operations. Beginning of February 2022, the prototype was shipped back to the headquarters of Pellenc ST in France for the final tests.
How are the digital watermarks applied to the products?
Schneider: Digital watermarks can either be printed on the packaging or embossed in molds. Digital watermarks for print consist of repeated tiles, whose pieces can be combined together to recover a barcode and which are applied by the encoder in a mosaic manner. For enhancing the artwork with digital watermarks, existing pixels are used, hence, this process does not require special inks or printing processes. On the other hand, digital watermarks for molds are the result of micro-topological variations in the substrate of the mold, creating signal tiles.
How does the technology for reading them work and what technical requirements do the recycling companies have to bring along?
Schneider: Digital watermarks on packaging would be detected through an add-on module that would be mounted onto existing detection sorting units at sorting and recycling facilities, and would be used in combination with existing NIR detection. The aim is that once the coded packaging has entered into a waste sorting or recycling facility, the digital watermark can be detected and decoded by a high resolution camera of this add-on module on the sorting line, which then – based on the transferred attributes (for example food versus non-food) – is able to sort the packaging in corresponding streams.