smartFoodTechnologyOWL - TH OWL
smartFoodTechnologyOWL - futurefoodfactory

EP4b: PETauthent - Authentication of PET recyclate for food packaging using data-intensive sensors and machine learning methods

Explorative project

In the PETauthent research project, recyclate content and quality of granules and preforms as part of the production of beverage bottles made of polyethylene terephthalate (PET) will be investigated using chemical, spectroscopic, thermal and other analytical methods. By using machine learning (ML) or KI algorithms, models can be created that allow the quality determination. The aim is to develop a method that is as simple as possible, sufficiently reliable and inexpensive, and with which it is possible to differentiate between recyclate and virgin material and to determine the recyclate content.

PET bottles are important for the transport, storage and manageability of beverages. They are stable, have good barrier properties and can be collected and recycled by type through the deposit system. According to the legislator, the recyclate content of PET beverage bottles should increase to at least 25 % by 2025. This is not least an economic aspect, because currently the prices for recyclates are higher than those for new material.

In the previous project "Recyclate Transparency", the use of near-infrared spectroscopy (NIR spectroscopy) to distinguish virgin material from recyclate and to quantify the recyclate content in PET granules was investigated. This method can be used easily, quickly and also inline. However, this method has not been sufficiently investigated to provide proof of quality even for unknown samples. In "PETauthent" the applicability of NIR spectroscopy will be further investigated and the results will be additionally combined with other analytical methods.

Chemical, spectroscopic and thermal methods (e.g. electronic nose, DSC, UV-Vis-NIR) will be used to obtain an easily applicable and robust method for quality assessment of PET recyclates by means of data fusion and ML algorithms. This should make it possible to determine the content of recyclate as well as the quality of granules, preforms and bottles. Together with the previous project "Recyclate Transparency", "PETauthent" should contribute to more transparency among companies as well as consumers regarding the use of recyclates in packaging, so that the required higher recyclate rates can be safely realized.

 

Project leaders:  Prof. Jan Schneider, Prof. Miriam Pein-Hackelbusch

Project partners:

Right panel

Clicking outside this panel wont close it because clickClose = false