A new hyperspectral imaging based device for quality control in plastic recycling

Abstract

The quality control of contamination level in the recycled plastics stream has been identified as an important key factor for increasing the value of the recycled material by both plastic recycling and compounder industries. Existing quality control methods for the detection of both plastics and non-plastics contaminants in the plastic waste streams at different stages of the industrial process (e.g. feed, intermediate and final products) are currently based on the manual collection from the stream of a sample and on the subsequent off-line laboratory analyses. The results of such analyses are usually available after some hours, or sometimes even some days, after the material has been processed. The laboratory analyses are time-consuming and expensive (both in terms of equipment cost and their maintenance and of labour cost).Therefore, a fast on-line assessment to monitor the plastic waste feed streams and to characterize the composition of the different plastic products, is fundamental to increase the value of secondary plastics. The paper is finalized to describe and evaluate the development of an HSI-based device and of the related software architectures and processing algorithms for quality assessment of plastics in recycling plants, with particular reference to polyolefins (PO). NIR-HSI sensing devices coupled with multivariate data analysis methods was demonstrated as an objective, rapid and non-destructive technique that can be used for on-line quality and process control in the recycling process of POs. In particular, the adoption of the previous mentioned HD&SW integrated architectures can provide a solution to one of the major problems of the recycling industry, which is the lack of an accurate quality certification of materials obtained by recycling processes. These results could therefore assist in developing strategies to certify the composition of recycled PO products

Classification of Common Household Plastic Wastes Combining Multiple Methods Based on Near-Infrared Spectroscopy

“…To categorize different polymers, data analysis methods and classification algorithms are needed to analyze the NIR spectral data. Some multivariate statistical methods are often adopted, like principal component analysis (PCA), , linear discriminant analysis (LDA), classification and regression trees (CART), partial least-squares discriminant analysis (PLS-DA), − spectral angle mapper (SAM), etc.…”

A Review of Sorting and Separating Technologies Suitable for Compostable and Biodegradable Plastic Packaging

“…Serranti et al ( 2019) also used the combination of chemometrics and SWIR hyperspectral imaging in the range between 1,000 and 2,500 nm to identify the polymer particles such as PE, PP and PS (size <5 mm) from marine environment. Bonifazi et al (2013) have applied hyperspectral imaging to enhance the efficiency of polyolefin recycling system while Moroni et al (2015) used hyperspectral imaging to separate PET and PVC. The selected wavelength range in this experiment is visible (400-1,000 nm) and SWIR (900-1,700 nm).…”

Separation of virgin plastic polymers and post-consumer mixed plastic waste by sinking-flotation technique

“…Similarly, Serranti et al (2015) through the magnetic density separation technique managed to recover more than 94% of polyolefins. Bonifazi et al (2013) using hyperspectral imaging technology (HSI) obtained PE and PP recoveries higher than 96%.…”