How Digital Watermarks Work in Recycling Lines
A digital watermark is an optical pattern spread across the pack artwork (and in some cases moulds). It’s designed to be invisible to shoppers but readable by high-speed, high-resolution cameras and processors on sorting lines, which can resolve the mark to product attributes such as material type, SKU, and other manufacturer data.
This differs from a visible barcode/QR in a material recovery facility: orientation is random, packs are dirty, overlap happens, and packaging is frequently crushed. A visible code sits in one location; a watermark is repeated across the design, improving detection odds when only part of the pack face is visible.
Digital watermark
Detectability in sorting lines: Designed for high-speed optical detection on sorting lines.
Robustness to damage/soiling: Repeated across the surface; can still be detected when partially obscured.
Privacy/data link: Can resolve to cloud-linked product attributes; data governance required.
Implementation complexity: Requires prepress “enhancement” + print QC to preserve readability.
Visible barcode/QR
Detectability in sorting lines: Typically designed for point scans; sorting-line performance varies and is often not reported (unspecified).
Robustness to damage/soiling: Single-location code is easier to miss if scratched, folded, or covered (sorting-line evidence: unspecified).
Privacy/data link: Can link to web data; consumer scanning can create tracking/data-retention expectations (implementation-dependent).
Implementation complexity: Simple to place visually, but recycling-grade detection is not guaranteed (unspecified).
Evidence from trials and research
Published Phase II semi-industrial results for HolyGrail 2.0 cite ~99% detection, ~95% ejection, and ~93%–95% purity under typical industrial conditions (belt speed ~3 m/s; rigids ~2.5 tons/hour; flexibles ~0.5 tons/hour).
Phase III is the headline. The industrial trial at Hündgen Entsorgung in Swisttal, Germany recorded ~56,000 detections/day and 5.66 million detections of 5,949 SKUs over 100 days, with 87.9%–93.8% detection efficiency on real post-household rigid packaging waste.
Academic work is starting to treat watermark identification as one component of “hybrid sorting” systems: a 2026 open-access life cycle assessment of advanced sorting scenarios (including optional digital watermark identification) reported net greenhouse-gas reductions up to 416 kg CO₂-eq per Mg of lightweight packaging collected and recycled (system-level result, not watermarking alone).
Pilots that show how brands intend to use it
One track is “better-than-NIR alone” sorting by adding digital watermark detection to existing optical approaches. In the 2024 industrial trial, detection prototypes combined near-infrared (NIR) and digital watermark technology, and the organizers explicitly point to follow-on market demonstrations (including planned activity in Belgium and additional streams) to scale and unlock dedicated recycled PP food-grade materials (exact commercial tonnages and pricing are unspecified).
Another clear track is fit-for-use streams. Packaging Dive (2023) described a 2024 France market launch where watermarks are added to packaging designs and then detected by optical sorters; the watermark can be linked to a cloud repository of product attributes. The cited goal at a Veolia facility was to sort detergent and cosmetic streams to produce “fit-for-use” recycled HDPE grades, with participation from The Procter & Gamble Company, L'Oréal, and Henkel.
Why regulation and economics are converging now
Extended Producer Responsibility is turning recycling outcomes into a producer cost line. CalRecycle states that California’s SB 54 establishes a packaging EPR program and shifts end-of-life responsibility toward producers. A 2025 update notes seven U.S. states have enacted packaging-focused EPR laws, with compliance timelines varying and some financial obligations starting in 2026 (state-by-state details vary; some specifics are outside public summaries).
In the European Union, the Ecodesign for Sustainable Products Regulation entered into force in July 2024 and sets the framework for Digital Product Passports. GS1 cites ESPR Article 10 requiring a DPP to be connected via a data carrier to a persistent unique product identifier; whether watermarks will qualify on equal footing with QR codes will depend on delegated acts (unspecified).
What this means for printers
Digital watermarks are becoming a manufacturing requirement—if you can’t prove detectability, you’ll lose the job.
For printers, watermark readiness is not a design add-on; it’s a workflow capability: controlled prepress “enhancement,” print QC that preserves readability, and verification/inspection that proves compliance. You also inherit data responsibility: the mark must resolve to the right attributes as SKUs change.
The upside is new billable work (enhancement services, verification documentation, sustainability-linked programs); the downside is operational risk (rejects if your process drifts, mis-sort liability, and privacy/data governance expectations). Cost/ROI benchmarks are often not public (unspecified).
Actionable next steps:
Pilot one substrate family and test watermark robustness across real press + finishing variability.
Add verification capability and clear acceptance criteria before selling “watermark-enabled” packaging.
Define data mapping ownership and mis-sort liability in contracts (audit-trail requirements often unspecified).
Pair watermarks with retrofit inkjet/print-bar capability for visible recycling instructions or serialization when customers require both layers.
Ask your customers now what watermark-ready proof they will require—because the sorting line will.