Plastics are ubiquitous in our lives. Whether in the form of kitchen items, toys, technical items or in vehicle components. Due to their wide variety (thermoplastics, thermosets and elastomers), many requirements for a product can be fulfilled through the use of plastics. The trade of plastics is flourishing worldwide.


When receiving a plastic delivery for subsequent processing or for direct sale, however, many companies are faced with the great challenge of determining whether the goods delivered are actually the materials ordered. In addition to the visual test, the breakage or fingernail test as well as the burning and odor test have established themselves as common practices for identity verification. All tests are based on purely subjective perceptions and in the best case can only be carried out by very experienced employees. The burning and odor test also represent an incalculable health risk for employees.


With the InProSens sensor systems you can safely check in a few seconds and at any location whether the plastic is your preferred material. In addition, it can also be determined whether additives have been added to the plastic. The plastic is not damaged by the sensor measurement, since only a spectrum has to be recorded with a sensor system.


A measured spectrum can be used to predict which material the plastic is made of. For this purpose, a prediction model is used, in which various calibration measurements were integrated as a data basis.


In the graphic with the calibration data you can see that each plastic material has an individual spectrum. The point cloud also shows how big the differences are between the individual spectra. The closer two materials are to each other in the point cloud, the more similar the materials are.

If a measured spectrum corresponds to one of these calibration spectra, it can be predicted with a certain probability which material the present one is.


The measurements shown in the graphics are only an extract from the possible measurable plastics. Upon request, the model can be customized with additional materials. 


Further information on the prediction model is available in the PDF file that can be downloaded here.

Adobe Acrobat Dokument 279.1 KB