Search

Sensor-based monitoring of construction and demolition waste (CDW) streams plays an important role in recycling (RC). Extracted knowledge about the composition of a material stream helps identifying RC paths, optimizing processing plants and form the basis for sorting. To enable economical use, it is necessary to ensure robust detection of individual objects even with high material throughput. Conventional algorithms struggle with resulting high occupancy densities and object overlap, making deep learning object detection methods more promising. In this study, different deep learning architectures for object detection (Region-based CNN/Region-based Convolutional Neural Network (Faster R-CNN), You only look once (YOLOv3), Single Shot MultiBox Detector (SSD)) are investigated with respect to their suitability for CDW characterization. A mixture of brick and sand-lime brick is considered as an exemplary waste stream. Particular attention is paid to detection performance with increasing occupancy density and particle overlap. A method for the generation of synthetic training images is presented, which avoids time-consuming manual labelling. By testing the models trained on synthetic data on real images, the success of the method is demonstrated. Requirements for synthetic training data composition, potential improvements and simplifications of different architecture approaches are discussed based on the characteristic of the detection task. In addition, the required inference time of the presented models is investigated to ensure their suitability for use under real-time conditions. [ABSTRACT FROM AUTHOR]

To achieve a highly efficient recycling process using a database, we developed a sensor-based sorting system that recognizes individual models of waste digital devices by the combination of convolutional neural network (CNN)-based image recognition and mass and 3D shape measurements. The system uses the 1024-dimensional vector of CNN output as an index to characterize the image data and the separately measured 8-dimensional vector consisting of mass and seven 3D feature values of the target device, and simultaneously performs 8-item classification and individual model identification. In the model identification test using 1958 models of undamaged waste digital devices (728 smartphones, 581 handy phones, 493 compact digital cameras, 47 video cameras, 44 portable game consoles, and 65 small tablet devices), the overall values of the Top 1 success rate (rate of correctly answered as 1st) and Top 3 success rate (rate of correctly answered within 3rd) were 90.3% and 94.9%, respectively. Excluding the video camera, which had difficulties in 3D shape measurement, the overall Top 1 success rate was 90.9% for 1911 models of the other five item categories. Compared with the results in a previous study, the identification algorithm was more effective in improving the performance of individual model identification. [ABSTRACT FROM AUTHOR]

Copyright of BHM Berg- und Hüttenmännische Monatshefte is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Österreichische Wasser- und Abfallwirtschaft is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Österreichische Wasser- und Abfallwirtschaft is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Pyrrolizidinalkaloide (PA) sind sekundäre Pflanzenstoffe und dienen in erster Linie der Pflanze zum Schutz gegen Fraßfeinde. Durch Mitbeerntung PA-haltiger Beikräuter im Pflanzenbau gelangen PAs als Verunreinigung in Lebensmittel (Salate, Kräuter, Tees) oder pflanzliche Arzneimittel. Aufgrund ihrer lebertoxischen und genotoxischen Wirkungsweise stellen sie eine potenzielle Gesundheitsgefahr dar. PA-Pflanzen sind besonders im Arznei- und Gewürzpflanzenanbau ein Problem, denn schon wenige Pflanzen pro Hektar genügen, um die Ernte für den Handel unbrauchbar zu machen. Eine Bekämpfung bzw. Reduktion der PA-Beikräuter bereits auf dem Feld ist zumeist nur mechanisch und mit sehr hohem Personalaufwand möglich. Dies ist für die Anbauer zunehmend nicht mehr ökonomisch darstellbar. Darüber hinaus ist importierte, getrocknete Ware oftmals höher als zulässig mit PAs belastet. Hier fehlt es aktuell an Möglichkeiten, PA-haltige Pflanzenteile aus der Ware zu entfernen. Hierzu sollte das über das Bundesministerium für Ernährung und Landwirtschaft (BMEL) und über die Fachagentur für Nachwachsende Rohstoffe e. V. (FNR) (FKZ 220132165) geförderte Verbundprojekt „Detektion und Entfernung von Pyrrolizidinalkaloid-haltigen Unkräutern aus Kulturpflanzen nach der Ernte - PA-NIRSort“ mögliche Lösungswege aufzeigen. Mit der Entwicklung einer automatisierten, der Ernte nachgelagerten Detektion und Abtrennung der kontaminierenden, toxischen PA-Beikräuter, kann eine effiziente Verbesserung der Qualitätskontrolle geschaffen werden. Dies sollte auf Basis von Hyperspektral-Nah-Infrarot-Spektroskopie (hyperspektral-NIRS), in Kombination mit einer Druckluftsortiereinheit entwickelt werden. Daher wurde ein Verfahren zum Nachweis, zur Klassifikation und zur physikalischen Ausschleusung von PA-haltigen Pflanzenteilen in einem Materialstrom von Arznei- und Gewürzpflanzen (Brennnessel, Melisse und Minze) entwickelt und umgesetzt. Die Detektion basiert auf bildgebenden hyperspektralen Sensoren im kurzwelligen Infrarotbereich. Die gewonnenen Hyperspektraldaten bilden die Grundlage für das Training eines KI-basierten Klassifikationsmodells. Die Klassifikation mittels NIR-Spektroskopie erreicht eine Erkennungsrate von über 90 %. Mit Hilfe dieses Detektionsmodells wurde eine sensorgestützte Sortieranlage entwickelt, mit dem Ziel PA-haltige Pflanzenteile physikalisch auszuschleusen und den pflanzlichen Materialstrom aufzureinigen. Dabei lag der Fokus auf der Optimierung des Materialtransports auf dem Förderband. Hierfür wurde ein aus Druckluft erzeugter Luftstrom zur Materialberuhigung eingesetzt. Somit wurde ein sensorgestütztes Sortierverfahren mittels pneumatischen Schnellschaltventilen erfolgreich umgesetzt.

At present, sensor-based sorting machines are usually not operated at the optimal operation point but are either overrun or underrun depending on the availability of waste streams. Mathematical approaches for predefined ideal mixtures can be found based on the input stream composition and the throughput rate. This scientific article compares whether and under what conditions these approaches can be applied to sensor-based sorting machines. Existing data for predefined ideal mixtures are compared with newly generated data of real waste on three sensor-based sorting setups in order to make significant statements. Five samples of 3D plastics at regular intervals were taken in a processing plant for refuse-derived fuels. With the comparison of all these results, four hypotheses were validated, related to whether the same mathematical approaches can be transferred from ideal mixtures to real waste and whether they can be transferred to sensor-based sorting machines individually or depending on the construction type. The developed mathematical approaches are regression models for finding the optimal operation point to achieve a specific sensor-based sorting result in terms of purity and recovery. For a plant operator, the main benefit of the findings of this scientific article is that purity could be increased by 20% without substantially adapting the sorting plant. [ABSTRACT FROM AUTHOR]

Copyright of Julius-Kühn-Archiv is the property of Julius Kuehn Institut and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Technisches Messen is the property of De Gruyter and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of BHM Berg- und Hüttenmännische Monatshefte is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

To improve the recycling process of waste from electrical products, we developed a sensor-based sorting system using convolutional neural network (CNN)-based image recognition. We sampled a total of 1571 models in three categories of waste digital devices (484 smartphones, 580 non-smart mobile phones, and 507 digital cameras) and examined a method of fusing information from 2D/3D appearance, volume, and mass into a single input image to VGG16 (a well-known CNN). The identification performance of the VGG16 was improved by pre-image-processing that placed the digital device at a fixed orientation and fixed location. In the three-category classification of 628 untrained digital devices, the accuracy was 97.7% for using the "4ch" image that results from inserting the 3D image fusing density information of the digital device into a fourth channel of the BGR color image. In the individual product model identification of trained 943 digital devices using the output vector from the fully connected layer of VGG16, the correctly identified rate was 90.5% for using the "BG + 3D" image that results from overwriting the third channel of the BGR color image with the 3D image fusing density information of the digital device under "fine-tuning" of the VGG16 pre-trained by the ImageNet dataset. [ABSTRACT FROM AUTHOR]

Explainable artificial intelligence (XAI) can make machine learning based systems more transparent. This additional transparency can enable the use of machine learning in many different domains. In our work, we show how XAI methods can be applied to an autoencoder for anomaly detection in a sensor-based sorting system. The setup of the sorting system consists of a vibrating feeder, a conveyor belt, a line-scan camera and an array of fast-switching pneumatic valves. It allows the separation of a material stream into two fractions, realizing a binary sorting task. The autoencoder tries to mimic the normal behavior of the nozzle array and thus can detect abnormal behavior. The XAI methods are used to explain the output of the autoencoder. As XAI methods global and local approaches are used, which means we receive explanations for both a single result and the whole autoencoder. Initial results for both approaches are shown, together with possible interpretations of these results. [ABSTRACT FROM AUTHOR]

A DEM‐CFD (discrete element method ‐ computational fluid dynamics) model of an optical belt sorter was extensively compared with experiments of a laboratory‐scale sorter to assess the model's accuracy. Brick and sand‐lime brick were considered as materials. First, the transport characteristics on the conveyor belt, involving mass flow, lateral particle distribution and proximity, were compared. Second, sorting results were benchmarked for varying mixture proportions at differing mass flows. It was found that the numerical model is able to reproduce the experimental results with high accuracy. [ABSTRACT FROM AUTHOR]

Although artificial neural networks are widely used in various fields, including mining and mineral processing, they can be problematic for appropriately choosing the model architecture and parameters. In this article, we describe a procedure for the optimization of the structure of a convolutional neural network model developed for the purposes of metallic ore pre-concentration. The developed model takes as an input two-band X-ray scans of ore grains, and for each scan two values corresponding to concentrations of zinc and lead are returned by the model. The whole process of sample preparation and data augmentation, optimization of the model hyperparameters and training of selected models is described. The ten best models were trained ten times each in order to select the best possible one. We were able to achieve a Pearson coefficient of R = 0.944 for the best model. The detailed results of this model are shown, and finally, its applicability and limitations in real-world scenarios are discussed. [ABSTRACT FROM AUTHOR]

At present, sensor-based sorting machines are usually not operated at the optimal operation point but are either overrun or underrun depending on the availability of waste streams. Mathematical approaches for predefined ideal mixtures can be found based on the input stream composition and the throughput rate. This scientific article compares whether and under what conditions these approaches can be applied to sensor-based sorting machines. Existing data for predefined ideal mixtures are compared with newly generated data of real waste on three sensor-based sorting setups in order to make significant statements. Five samples of 3D plastics at regular intervals were taken in a processing plant for refuse-derived fuels. With the comparison of all these results, four hypotheses were validated, related to whether the same mathematical approaches can be transferred from ideal mixtures to real waste and whether they can be transferred to sensor-based sorting machines individually or depending on the construction type. The developed mathematical approaches are regression models for finding the optimal operation point to achieve a specific sensor-based sorting result in terms of purity and recovery. For a plant operator, the main benefit of the findings of this scientific article is that purity could be increased by 20% without substantially adapting the sorting plant.

Sensor-based sorting techniques offer the potential to improve ore grades and reduce the amount of waste material processed. Previous studies show that sensor-based sorting can reduce energy, water and reagent consumption and fine waste production by discarding waste prior to further processing. In this literature review, recent investigations of sensor-based sorting and the fundamental mechanisms of the main sorting techniques are evaluated to inform optimal sensor selection. Additionally, the fusing of data from multiple sensing techniques to improve characterization of the sensed material and hence sorting capability is investigated. It was found that the key to effective implementation of sensor-based sorting is the selection of a sensing technique which can sense a characteristic capable of separating ore from waste with a sampling distribution sufficient for the considered sorting method. Classes of potential sensor fusion sorting applications in mineral processing are proposed and illustrated with example cases. It was also determined that the main holdup for implementing sensor fusion is a lack of correlatable data on the response of multiple sensing techniques for the same ore sample. A combined approach of experimental testing supplemented by simulations is proposed to provide data to enable the evaluation and development of sensor fusion techniques. [ABSTRACT FROM AUTHOR]

Small plastic packaging films make up a quarter of all packaging waste generated annually in Austria. As many plastic packaging films are multilayered to give barrier properties and strength, this fraction is considered hardly recyclable and recovered thermally. Besides, they can not be separated from recyclable monolayer films using near-infrared spectroscopy in material recovery facilities. In this paper, an experimental sensor-based sorting setup is used to demonstrate the effect of adapting a near-infrared sorting rig to enable measurement in transflection. This adaptation effectively circumvents problems caused by low material thickness and improves the sorting success when separating monolayer and multilayer film materials. Additionally, machine learning approaches are discussed to separate monolayer and multilayer materials without requiring the near-infrared sorter to explicitly learn the material fingerprint of each possible combination of layered materials. Last, a fast Fourier transform is shown to reduce destructive interference overlaying the spectral information. Through this, it is possible to automatically find the Fourier component at which to place the filter to regain the most spectral information possible. [ABSTRACT FROM AUTHOR]

This work covers a lifecycle assessment of monolayer and multilayer films to quantify the environmental impacts of changing the management of plastic film waste. This lifecycle assessment offers the possibility of quantifying the environmental impacts of processes along the lifecycle of monolayer and multilayer films and mapping deviating impacts due to changed process parameters. Based on the status quo, the changes in global warming potential and abiotic fossil resource depletion were calculated in different scenarios. The changes included collecting, sorting, and recycling mono- and multilayer films. The "Functional Unit" under consideration comprised 1000 kg of plastic film waste, generated as post-consumer waste in Austria and captured in the lightweight packaging collection system. The results showed the reduction of environmental impacts over product lifecycles by improving waste management and creating a circular economy. Recycling all plastic film reduced global warming potential by 90% and abiotic fossil resource consumption by 93%. The necessary optimisation steps to meet the politically required recycling rates by 2025 and 2030 could be estimated, and the caused environmental impacts are presented. This work shows the need for increased collection, recycling, and significant improvement in the sorting of films to minimise global warming potential and resource consumption. [ABSTRACT FROM AUTHOR]

Plastics are a ubiquitous material with good mechanical, chemical and thermal properties, and are used in all industrial sectors. Large quantities, widespread use, and insufficient management of plastic wastes lead to low recycling rates. The key challenge in recycling plastic waste is achieving a higher degree of homogeneity between the different polymer material streams. Modern waste sorting plants use automated sensor-based sorting systems capable to sort out commodity plastics, while many engineering plastics, such as polyoxymethylene (POM), will end up in mixed waste streams and are therefore not recycled. A novel approach to increasing recycling rates is tracer-based sorting (TBS), which uses a traceable plastic additive or marker that enables or enhances polymer type identification based on the tracer's unique fingerprint (e.g., fluorescence). With future TBS applications in mind, we have summarized the literature and assessed TBS techniques and spectroscopic detection methods. Furthermore, a comprehensive list of potential tracer substances suitable for thermoplastics was derived from the literature. We also derived a set of criteria to select the most promising tracer candidates (3 out of 80) based on their material properties, toxicity profiles, and detectability that could be applied to enable the circularity of, for example, POM or other thermoplastics. [ABSTRACT FROM AUTHOR]

• Systematic literature review of 198 peer-reviewed journal articles. • Proposal of unified sensor-based material flow characterization (SBMC) terminology. • Elaborate overview on machine learning (ML) implementation in SBMC. • Innovative ML algorithm comparison based on Elo ratings adapted from chess games. • Identification and discussion of ten future research potentials in SBMC. Digital technologies hold enormous potential for improving the performance of future-generation sorting and processing plants; however, this potential remains largely untapped. Improved sensor-based material flow characterization (SBMC) methods could enable new sensor applications such as adaptive plant control, improved sensor-based sorting (SBS), and more far-reaching data utilizations along the value chain. This review aims to expedite research on SBMC by (i) providing a comprehensive overview of existing SBMC publications, (ii) summarizing existing SBMC methods, and (iii) identifying future research potentials in SBMC. By conducting a systematic literature search covering the period 2000 – 2021, we identified 198 peer-reviewed journal articles on SBMC applications based on optical sensors and machine learning algorithms for dry-mechanical recycling of non-hazardous waste. The review shows that SBMC has received increasing attention in recent years, with more than half of the reviewed publications published between 2019 and 2021. While applications were initially focused solely on SBS, the last decade has seen a trend toward new applications, including sensor-based material flow monitoring, quality control, and process monitoring/control. However, SBMC at the material flow and process level remains largely unexplored, and significant potential exists in upscaling investigations from laboratory to plant scale. Future research will benefit from a broader application of deep learning methods, increased use of low-cost sensors and new sensor technologies, and the use of data streams from existing SBS equipment. These advancements could significantly improve the performance of future-generation sorting and processing plants, keep more materials in closed loops, and help paving the way towards circular economy. [ABSTRACT FROM AUTHOR]

Through textile recycling, resources can be saved, and landfilling can be avoided. However, determined input streams are needed for high-quality recycling processes. The aim of this study is to investigate the detection, prediction and sorting of polyester-containing textile waste using near-infrared spectroscopy. In experimental trials, polyester-containing textiles with different fibre blends are scanned, and the spectra are recorded. The results are evaluated in a linear regression model and transferred to an industrial-scale sorting plant. Due to the different peaks in the spectra of the investigated fibre types and a visible correlation of the absorption intensity between the blend ratios, the use of NIR spectroscopy is very promising for application in the sorting of polyester-containing textiles. [Display omitted] [ABSTRACT FROM AUTHOR]

• 2D Materials are commonly incinerated which does not increase the recycling quota. • Usage of reflective backgrounds increases spectral quality and detectability. • Transflection can be used to improve the detectability of 2D Materials. • NIR spectroscopy in transflection mode can be used to recycle 2D materials. Most two-dimensional plastic packaging materials are thermally recovered, which does not add to the recycling quota of 50 % required by EU legislation for all lightweight packaging until 2025. Furthermore, the separation processes for 2D materials cannot reach the same level of accuracy, which is possible in the sorting of rigid plastic packaging. This study proposes new adaptations to existing sorting aggregates to increase the near-infrared spectral quality of two-dimensional materials. It aims to improve the spectral quality, which was defined by the deviation of the spectra from a reference spectrum and the variability of the recorded spectra, which can be achieved by installing reflectors behind the material made up of copper or aluminium. This setup enables detection in transflection rather than reflection mode. The variability could be reduced by a factor of 6 through the use of a reflective background. Meanwhile, the spectral fidelity to the reference spectrum could be enhanced, in some cases decreasing the deviation from the reference spectrum by 30 %, which means enhancing a spectrum from unrecognisable to useable. Apart from using reflective materials, the effects of emitter intensity, material and thickness were evaluated. [ABSTRACT FROM AUTHOR]

In this study, effective solutions for polyethylene terephthalate (PET) recycling based on hyperspectral imaging (HSI) coupled with variable selection method, were developed and optimized. Hyperspectral images of post-consumer plastic flakes, composed by PET and small quantities of other polymers, considered as contaminants, were acquired in the short-wave infrared range (SWIR: 1000--2500 nm). Different combinations of preprocessing sets coupled with a variable selection method, called competitive adaptive reweighted sampling (CARS), were applied to reduce the number of spectral bands useful to detect the contaminants in the PET flow stream. Prediction models based on partial least squares-discriminant analysis (PLS-DA) for each preprocessing set, combined with CARS, were built and compared to evaluate their efficiency results. The best performance result was obtained by a PLS-DA model using multiplicative scatter correction + derivative + mean center preprocessing set and selecting only 14 wavelengths out of 240. Sensitivity and specificity values in calibration, cross-validation and prediction phases ranged from 0.986 to 0.998. HSI combined with CARS method can represent a valid tool for identification of plastic contaminants in a PET flakes stream increasing the processing speed as requested by sensor-based sorting devices working at industrial level. [ABSTRACT FROM AUTHOR]

Although artificial neural networks are widely used in various fields, including mining and mineral processing, they can be problematic for appropriately choosing the model architecture and parameters. In this article, we describe a procedure for the optimization of the structure of a convolutional neural network model developed for the purposes of metallic ore pre-concentration. The developed model takes as an input two-band X-ray scans of ore grains, and for each scan two values corresponding to concentrations of zinc and lead are returned by the model. The whole process of sample preparation and data augmentation, optimization of the model hyperparameters and training of selected models is described. The ten best models were trained ten times each in order to select the best possible one. We were able to achieve a Pearson coefficient of R = 0.944 for the best model. The detailed results of this model are shown, and finally, its applicability and limitations in real-world scenarios are discussed.

Sensor-based sorting techniques offer the potential to improve ore grades and reduce the amount of waste material processed. Previous studies show that sensor-based sorting can reduce energy, water and reagent consumption and fine waste production by discarding waste prior to further processing. In this literature review, recent investigations of sensor-based sorting and the fundamental mechanisms of the main sorting techniques are evaluated to inform optimal sensor selection. Additionally, the fusing of data from multiple sensing techniques to improve characterization of the sensed material and hence sorting capability is investigated. It was found that the key to effective implementation of sensor-based sorting is the selection of a sensing technique which can sense a characteristic capable of separating ore from waste with a sampling distribution sufficient for the considered sorting method. Classes of potential sensor fusion sorting applications in mineral processing are proposed and illustrated with example cases. It was also determined that the main holdup for implementing sensor fusion is a lack of correlatable data on the response of multiple sensing techniques for the same ore sample. A combined approach of experimental testing supplemented by simulations is proposed to provide data to enable the evaluation and development of sensor fusion techniques.

Small plastic packaging films make up a quarter of all packaging waste generated annually in Austria. As many plastic packaging films are multilayered to give barrier properties and strength, this fraction is considered hardly recyclable and recovered thermally. Besides, they can not be separated from recyclable monolayer films using near-infrared spectroscopy in material recovery facilities. In this paper, an experimental sensor-based sorting setup is used to demonstrate the effect of adapting a near-infrared sorting rig to enable measurement in transflection. This adaptation effectively circumvents problems caused by low material thickness and improves the sorting success when separating monolayer and multilayer film materials. Additionally, machine learning approaches are discussed to separate monolayer and multilayer materials without requiring the near-infrared sorter to explicitly learn the material fingerprint of each possible combination of layered materials. Last, a fast Fourier transform is shown to reduce destructive interference overlaying the spectral information. Through this, it is possible to automatically find the Fourier component at which to place the filter to regain the most spectral information possible.

This work covers a lifecycle assessment of monolayer and multilayer films to quantify the environmental impacts of changing the management of plastic film waste. This lifecycle assessment offers the possibility of quantifying the environmental impacts of processes along the lifecycle of monolayer and multilayer films and mapping deviating impacts due to changed process parameters. Based on the status quo, the changes in global warming potential and abiotic fossil resource depletion were calculated in different scenarios. The changes included collecting, sorting, and recycling mono- and multilayer films. The “Functional Unit” under consideration comprised 1000 kg of plastic film waste, generated as post-consumer waste in Austria and captured in the lightweight packaging collection system. The results showed the reduction of environmental impacts over product lifecycles by improving waste management and creating a circular economy. Recycling all plastic film reduced global warming potential by 90% and abiotic fossil resource consumption by 93%. The necessary optimisation steps to meet the politically required recycling rates by 2025 and 2030 could be estimated, and the caused environmental impacts are presented. This work shows the need for increased collection, recycling, and significant improvement in the sorting of films to minimise global warming potential and resource consumption.

Can random deposition create dense non-overlapping material feeding? The question is very fundamental for the research of particle packing, while the answer is of great importance for any industrial process that applies single object operation. To gain an insight into this issue, we studied the overlap problems of convex particles in the manner of uniformly random deposition. The overlap probability of two convex particles with arbitrary shapes and sizes is formulated, and the coverage fractions of free particles and sticking particles (particles of the bottom layer) are precisely predicted. Simulations with rectangular particles verified the theory. Surprisingly, free particles can only occupy less than 7.5% of the plane area, much smaller than what is intuitively expected. Sticking particles, however, can easily cover 19%, a factor of 2.5 times larger. The finding is of great value for applications that need to create dense non-overlapping feeding. An evenly and densely distributed particle feeding is crucial in determining the capacity of many industrial processes that require single object operation. In this paper, we studied the overlap probability of convex particles when randomly dropping on a plane, to reveal the feasibility in creating dense monolayer feeding. [Display omitted] • The overlap probability of two convex particles in random dropping is theoretically formulated. • Coverage fractions of free particles and sticking particles (bottom layer) are predicted. • Free particles in random feeding can only occupy a very small area, less than 7.5% of rectangular particles. • A way to create dense non-overlapping feeding is to remove the overlapping layers while leaving the bottom layer. [ABSTRACT FROM AUTHOR]

Plastics are a ubiquitous material with good mechanical, chemical and thermal properties, and are used in all industrial sectors. Large quantities, widespread use, and insufficient management of plastic wastes lead to low recycling rates. The key challenge in recycling plastic waste is achieving a higher degree of homogeneity between the different polymer material streams. Modern waste sorting plants use automated sensor-based sorting systems capable to sort out commodity plastics, while many engineering plastics, such as polyoxymethylene (POM), will end up in mixed waste streams and are therefore not recycled. A novel approach to increasing recycling rates is tracer-based sorting (TBS), which uses a traceable plastic additive or marker that enables or enhances polymer type identification based on the tracer’s unique fingerprint (e.g., fluorescence). With future TBS applications in mind, we have summarized the literature and assessed TBS techniques and spectroscopic detection methods. Furthermore, a comprehensive list of potential tracer substances suitable for thermoplastics was derived from the literature. We also derived a set of criteria to select the most promising tracer candidates (3 out of 80) based on their material properties, toxicity profiles, and detectability that could be applied to enable the circularity of, for example, POM or other thermoplastics.

Sensor-based sorting is a machine vision application that has found industrial application in various fields. An accept-or-reject task is executed by separating a material stream into two fractions. Current systems use line-scanning sensors, which is convenient as the material is perceived during transportation. However, line-scanning sensors yield a single observation of each object and no information about their movement. Due to a delay between localization and separation, assumptions regarding the location and point in time for separation need to be made based on the prior localization. Hence, it is necessary to ensure that all objects are transported at uniform velocities. This is often a complex and costly solution. In this article, we propose a new method for reliably separating particles at nonuniform velocities. The problem is transferred from a mechanical to an algorithmic level. Our novel advanced image processing approach includes equipping the sorter with an area-scan camera in combination with a real-time multiobject tracking system, which enables predictions of the location of individual objects for separation. For the experimental validation of our approach, we present a modular sorting system, which allows comparing sorting results using a line-scan and area-scan camera. Results show that our approach performs reliable separation and hence increases sorting efficiency. [ABSTRACT FROM AUTHOR]

• Influence of pre-screening on down-stream processing is investigated. • Pre-screening improves the quality of the output. • Drum sieving cleans the surface of objects and hence increases sorting yields. • Selective size reduction enhances the enrichment of valuable materials. • Proper pre-treatment can raise the sorting yield by 3–5%. The use of plastic waste as resource gains more and more attention. In this context, material recycling is especially focused on packaging plastics. Further waste streams that contain a significant amount of plastics are mixed commercial and municipal solid waste. To assess the potential of plastics for recycling and energy recovery from these material streams large-scale experiments were conducted. The potential of mechanical pre-processing with the aim of generating a 3D-plastics pre-concentrate was assessed. The focus of these investigations was put on the relevance of the screening stage and its influence on down-stream material processing via ballistic separation and sensor-based sorting. Results demonstrate not only that the screening of both waste streams leads to enrichment of plastics in coarse particle size ranges (especially >80 mm) and transfer of contaminants, organics and minerals to fine fractions (especially <10 mm), but also that sensor-based sorting performance can be significantly enhanced due to cleaning effects on plastics, induced by the material circulation and the resulting interparticle friction in a drum screen. On the downside, the material rotation in a drum screen leads to tail-formation that can create plant down-time through clogging as well as material losses and impairment of pre-concentrates. [ABSTRACT FROM AUTHOR]

Experiments with sensor-based sorting (SBS) machinery provide insight into the effect of throughput rate and input composition on the sorting performance. For this purpose, material mixtures with certain compositions and particle size distributions were created from waste fractions and sorted at various throughput rates. To evaluate the sorting performance of the SBS unit (using near infrared technology) in dependence of the applied load, four assessment factors concerning the output fractions were studied: yield, product purity, recovery/product quantity and incorrectly discharged share of reject particles. The influences on the assessment parameters of light twodimensional (2D) particles in the input of a sorting stage and failing air valves in an SBS unit were evaluated for various input compositions at different throughput rates. It was found that a share of approximately 5 wt% 2D particles in the input had a similar negative effect on the yield as the malfunction of 20% of all air valves in an SBS machine at high throughput rates. Additionally, the failure of the air valves reduced the product purity of the sorting stage at increased throughput rates. Furthermore, qualitative observations concerning systematic effects of prior studies could be confirmed. Resulting graphs for a specific input composition of an SBS unit at varying throughput rates could be used to adjust the throughput rate to meet the exact demands for a sorting stage. [ABSTRACT FROM AUTHOR]

Copyright of BHM Berg- und Hüttenmännische Monatshefte is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Utilizing parallel algorithms is an established way of increasing performance in systems that are bound to real-time restrictions. Sensor-based sorting is a machine vision application for which firm real-time requirements need to be respected in order to reliably remove potentially harmful entities from a material feed. Recently, employing a predictive tracking approach using multitarget tracking in order to decrease the error in the physical separation in optical sorting has been proposed. For implementations that use hard associations between measurements and tracks, a linear assignment problem has to be solved for each frame recorded by a camera. The auction algorithm can be utilized for this purpose, which also has the advantage of being well suited for parallel architectures. In this paper, an improved implementation of this algorithm for a graphics processing unit (GPU) is presented. The resulting algorithm is implemented in both an OpenCL and a CUDA based environment. By using an optimized data structure, the presented algorithm outperforms recently proposed implementations in terms of speed while retaining the quality of output of the algorithm. Furthermore, memory requirements are significantly decreased, which is important for embedded systems. Experimental results are provided for two different GPUs and six datasets. It is shown that the proposed approach is of particular interest for applications dealing with comparatively large problem sizes. [ABSTRACT FROM AUTHOR]

In this study, effective solutions for polyethylene terephthalate (PET) recycling based on hyperspectral imaging (HSI) coupled with variable selection method, were developed and optimized. Hyperspectral images of post-consumer plastic flakes, composed by PET and small quantities of other polymers, considered as contaminants, were acquired in the short-wave infrared range (SWIR: 1000–2500 nm). Different combinations of preprocessing sets coupled with a variable selection method, called competitive adaptive reweighted sampling (CARS), were applied to reduce the number of spectral bands useful to detect the contaminants in the PET flow stream. Prediction models based on partial least squares-discriminant analysis (PLS-DA) for each preprocessing set, combined with CARS, were built and compared to evaluate their efficiency results. The best performance result was obtained by a PLS-DA model using multiplicative scatter correction + derivative + mean center preprocessing set and selecting only 14 wavelengths out of 240. Sensitivity and specificity values in calibration, cross-validation and prediction phases ranged from 0.986 to 0.998. HSI combined with CARS method can represent a valid tool for identification of plastic contaminants in a PET flakes stream increasing the processing speed as requested by sensor-based sorting devices working at industrial level.

The paper presents a comparison of the concentration methods conventional jig, air jig, and sensor-based sorting to treat construction and demolition waste. All tests were made with concrete, brick, and gypsum particles and the tests aim to separate these materials into different size ranges, depending on the method. The equipment tested, conventional jig, air jig, and sensor-based sorting present good results to concentrate construction and demolition waste particles, with different concentrations and mass recoveries. The results show particularly good mass recoveries and particle concentration for conventional jig, especially for concrete and gypsum particles. Sensor-based sorting should preferably use concentration circuits for best results.

Assessing the success of sensor-based sorting in the raw materials industry currently requires time-consuming and expensive empirical test work. In this contribution we illustrate the prospects of successful sensor selection based on data acquired by scanning electron microscopy-based image analysis. Quantitative mineralogical and textural data from more than 100 thin sections were taken to capture mineralogical and textural variability of two different ore types from the Hämmerlein Sn–In–Zn deposit, Germany. Parameters such as mineral grain sizes distribution, modal mineralogy, mineral area and mineral density distribution were used to simulate the prospects of sensor-based sorting using different sensors. The results illustrate that the abundance of rock-forming chlorite and/or density anomalies may well be used as proxies for the abundance of cassiterite, the main ore mineral. This suggests that sorting of the Hämmerlein ore may well be achieved by either using a short-wavelength infrared detector — to quantify the abundance of chlorite — or a dual-energy X-ray transmission detector to determine the abundance of cassiterite. Empirical tests conducted using commercially available short-wave infrared and dual-energy X-ray transmission sensor systems are in excellent agreement with simulation-based predictions and confirm the potential of the novel approach introduced here. Image 1 • Novel approach for the selection of sensors suitable for sensor-based sorting. • Approach validated by empirical test work for Sn–Zn–In ores of the Hämmerlein deposit. • Improvement of resource and energy efficiency in the raw materials industry. • Avoids time-consuming and expensive test work. • Important tool for realizing predictive geometallurgical models. [ABSTRACT FROM AUTHOR]

Near-infrared (NIR) and short-wavelength infrared (SWIR) hyperspectral imagery can be used to detect certain alteration minerals. At epithermal deposits, the formation of alteration minerals is, in theory, related to the mineralisation of gold and silver. In order to provide foundations for developing sensor-based sorting applications at a mine that exploits such a deposit, it was investigated if NIR-SWIR hyperspectral imagery can be used to distinguish between ore and waste particles by characterising the alteration mineralogy. Maps were produced from the NIR-SWIR hyperspectral images of 827 drill core samples that show mineral occurrences, mineral absorption feature intensities and characteristics of the iron oxide mineralogy. Partial least squares discriminant analysis (PLS-DA) was applied to the information contained in these maps to investigate if this information can be used to discriminate between ore and waste. The results showed that NIR-SWIR hyperspectral imagery could be used to segment a population of waste samples by detecting occurrences of pyrophyllite, dickite and/or illite. This result can be explained by the fact that these minerals are commonly deposited further away from the ore-bearing epithermal veins, while the absence of SWIR-active minerals or detected occurrences of alunite are more closely associated with these structures. The ability to identify waste with NIR-SWIR spectral sensors means there is potential that sensor-based sorting can be used to remove this waste from mineral processing operations. Additional research is still required to assess the economic feasibility of such a sensor-based sorting application. [ABSTRACT FROM AUTHOR]

In the project 'NEW-MINE' the use of sensor-based sorting machinery in the field of 'landfill mining' is investigated. Defilements pose a particular challenge in the treatment and sorting of plastics contained in landfills. For this reason, the effects of various pollutants caused by the interactions in the landfill body or the mechanical treatment steps in landfill mining are examined. In the following elaboration, the focus is on the influences of surface moisture and surface roughness of plastics on sensor-based sorting by means of near-infrared technology. Near-infrared radiation (NIR) in a wavelength range of 990 nm to 1500 nm has been used for the detection and classification of plastic particles. The experiments demonstrate that increased surface roughness reduces signal noise and thereby improves the classification of both spectrally similar and transparent plastics, but reduces the yield of low-softening plastics because their sliding speed on a sensor-based chute sorter varies as a result of the heating of the chute. Surface moisture causes the absorption of radiation from 1115 nm (high density polyethylene [HDPE], linear low density polyethylene [LLDPE], polyethylen terephthalate [PET] and polyvinylchloride [PVC]) or from 1230 nm (low density polyethylene [LDPE], polypropylene [PP] and thermoplastic polyurethane [TPU]) up to at least 1680 nm, which causes amplification or attenuation of various extremes in the derivative. However, the influence of surface moisture on the yield of plastics is usually very low and depends on the spectral differences between the different plastics. [ABSTRACT FROM AUTHOR]

• Identification of a robust and efficient strategy based on HSI for classification of plastic waste. • Plastic waste hierarchical classification by short-wave infrared hyperspectral imaging (SWIR). • Discrimination of plastic flakes in automated and non-destructive way. • Chemometric strategies applied to hyperspectral images for plastic waste identification. • Wavelengths selection and reduction for plastic flakes classification. Plastic waste management represents a global challenge in the framework of sustainable production and consumption of resources. One of the most critical issues in plastic recycling is polymer separation, necessary to obtain high-quality secondary raw material flow streams. The aim of this work was to build a classification strategy, based on pushbroom hyperspectral imaging, able to recognize the most common polymers found in mixed plastic waste to be applied at recycling plant scale. After exploring polymer spectral differences by principal component analysis, a hierarchical partial least squares-discriminant analysis, based on the acquired full spectra, and a hierarchical interval partial least squares-discriminant analysis, based on selected variables, were tested and their performances were evaluated and compared. High quality classification results were obtained in both cases, demonstrating that the developed multi-class models can be utilized in a flexible way for quality control and/or for on-line sorting actions in recycling plants. [Display omitted] [ABSTRACT FROM AUTHOR]