TY - CONF

T1 - Data Mining Within a Lightweight Packaging Waste Sorting Plant: Long-Term Insights into the Data of a Sensor-Based Sorter

AU - Kuhn, Nikolai Emanuel

AU - Schlögl, Sabine

AU - Koinig, Gerald

AU - Tischberger-Aldrian, Alexia

PY - 2024/2/13

Y1 - 2024/2/13

N2 - Sensor-based material flow monitoring has great potential in sorting plants ingeneral, an implementation based on the data of a sensor-based sorter (SBS)provides an economic advantage compared to the use of additional sensors. Inthis work, we provide insights into the data stream collected by a SBS positioned atthe beginning of the sorting cascade for 3D-shaped particles in a sorting plant forlightweight packaging waste in Austria. The data was analyzed in three regards: Afirst analysis identifies plant downtimes within a year and allocates them to eventssuch as weekends or public holidays. A second analysis presents the compositionof the lightweight packaging material over a year. Here, the mean of the dominatingfractions is 45.1% for PET, 16.1% for PE, and 15.4% for PP. Lastly, we examinethe seasonal variability of the material composition. Here, no major changes in thecomposition in terms of mean and standard deviation were observed. Only the shareof clear PET bottles increases in summer, whereas a lower share of blue PET bottlesat the same time compensates the overall share of PET bottles. The results of thisresearch indicate the variety of possible applications for monitoring based on SBSdata to improve the operational efficiency of the plant.

AB - Sensor-based material flow monitoring has great potential in sorting plants ingeneral, an implementation based on the data of a sensor-based sorter (SBS)provides an economic advantage compared to the use of additional sensors. Inthis work, we provide insights into the data stream collected by a SBS positioned atthe beginning of the sorting cascade for 3D-shaped particles in a sorting plant forlightweight packaging waste in Austria. The data was analyzed in three regards: Afirst analysis identifies plant downtimes within a year and allocates them to eventssuch as weekends or public holidays. A second analysis presents the compositionof the lightweight packaging material over a year. Here, the mean of the dominatingfractions is 45.1% for PET, 16.1% for PE, and 15.4% for PP. Lastly, we examinethe seasonal variability of the material composition. Here, no major changes in thecomposition in terms of mean and standard deviation were observed. Only the shareof clear PET bottles increases in summer, whereas a lower share of blue PET bottlesat the same time compensates the overall share of PET bottles. The results of thisresearch indicate the variety of possible applications for monitoring based on SBSdata to improve the operational efficiency of the plant.

M3 - Paper

SP - 11

EP - 22

T2 - Sensor-Based Sorting &amp; Control 2024<br/>

Y2 - 13 March 2024 through 14 March 2024

ER -