PPWR: How machine vision supports companies with the new packaging requirements

Since 12 August 2026, the new EU Packaging and Packaging Waste Regulation (PPWR, Regulation (EU) 2025/40) has applied across the EU. In Germany, the Packaging Law Implementation Act (VerpackDG) replaces the previous Packaging Act. Companies must meet stricter requirements for recyclability, material restrictions and packaging labelling. Machine vision, meaning the use of cameras and software for the automated inspection of products, offers methods that can be integrated into existing production and recycling processes.

Process graphic: how machine vision makes PPWR-relevant packaging characteristics verifiable before placing on the market and during recycling.Process graphic: how machine vision makes PPWR-relevant packaging characteristics verifiable before placing on the market and during recycling.

What changes with the PPWR

The PPWR tightens requirements for packaging design, material restrictions and labelling:

 

  • PFAS in food packaging: maximum 25 ppb per individual substance, 250 ppb in total, threshold of 50 ppm for fluorine content
  • Heavy metals (lead, cadmium, mercury, hexavalent chromium): combined maximum of 100 mg/kg
  • Packaging must be demonstrably designed for recyclability
  • From 2028: recycling rate of 75% for plastic waste in the dual systems
  • Violations of the VerpackDG: fines of up to €200,000

These requirements affect production, quality control and recycling alike, exactly where automated inspection comes in.

PPWR in context

Germany officially reports a recovery rate of around 71% for plastic packaging (reporting year 2024), well above the statutory minimum of 63% and a clear increase from 42% in 2018. However, this figure only shows how much packaging material enters sorting, not how much of it actually ends up as recyclate in new packaging. This is precisely where the PPWR comes in: it requires recyclable design and greater traceability along the entire chain, from material selection through to reuse.

Where machine vision supports implementation

Material identification

 

  • NIR and hyperspectral cameras distinguish types of plastic that look identical to the human eye. Foreign objects such as paper, glass or metal can also be highlighted far more reliably with this technology than with standard cameras.

Recyclability checks

 

  • Systems detect material combinations that are difficult to separate, for example when a label or coating made from a second material is firmly bonded to the packaging, as well as interfering colours and unsuitable closures.

Labelling checks

 

  • Barcodes, recycling symbols, deposit markings and mandatory text are checked for presence, position and legibility using optical character recognition (OCR).

Packaging component recognition

 

  • Lids, labels, films, trays and secondary packaging can be automatically captured and classified.

Quality control in production

 

  • Damaged, incorrectly sealed, incorrectly printed or overfilled packaging is detected and removed from the line.

Sorting in recycling

 

  • Machine vision can classify individual packaging by material, colour, shape and level of contamination, so that a sorting plant can separate it correctly.

Reduction of packaging material

 

  • Dimensions, empty space and material thickness can be measured to avoid over-packaging.

Documentation and traceability

 

  • Inspection results can be stored and linked to batches, suppliers or packaging types.

When is a standard camera enough?

For labelling checks, a standard camera with text recognition is often sufficient. Material identification is more complex: two packages can look optically identical and still be made from different plastics.

One example: black-coloured packaging can contain industrial carbon black as a pigment. Carbon black absorbs the light that NIR sensors normally use to distinguish plastic types, which means dark packaging remains invisible to many sorting plants and ends up incinerated rather than recycled. Hyperspectral imaging or AI models, meaning software that learns from many example images to separate even hard-to-distinguish materials, help here.

Which sensors and software make sense in each case is tested by STEMMER IMAGING's Technical Competence Centre using real packaging samples.

Frequently Asked Questions

When does the new Packaging Regulation apply?

The PPWR (Regulation (EU) 2025/40) has applied directly in all EU member states since 12 August 2026. In Germany, the VerpackDG enters into force at the same time and replaces the previous Packaging Act.

Can machine vision alone ensure PPWR compliance?

Machine vision covers the automated inspection of material, labelling and packaging quality. Questions relating to manufacturer definitions, recyclate quotas or legal conformity assessment fall outside the system and require additional organisational and legal steps.

Which sensors are suitable for reliable material identification?

Standard cameras with text recognition (OCR) are sufficient for simple labelling checks. Reliable material identification, for example on dark-coloured or multi-layer packaging, additionally requires NIR or hyperspectral cameras along with software-based evaluation.

Does the PPWR only affect food packaging?

No, the regulation applies to packaging of all kinds. Certain limits, such as those for PFAS, are set specifically for packaging in contact with food, while requirements for recyclability and labelling apply across all industries.

Next step

Contact STEMMER IMAGING's Technical Competence Centre for an assessment of your own packaging: