Statistical modelling of sensor-based sorting processes

Research output: Chapter in Book/Report/Conference proceedingConference contribution


According to technical literature and manufacturer specifications for sensor-based sorting (SBS) machinery the performance of such technologies is subject to the load put on a respective unit. This load can be defined by the throughput rate (either volumetric or mass specific), material properties (e.g. grain size distribution) and the composition of the input material for a sorting unit (e.g. share of material that is supposed to be ejected via air shocks).
During the processing of waste, the utilized machine is subject to particularly strong fluctuations of throughput rate and material composition. This observation can be traced back to a multitude of influencing factors as e.g. waste heterogeneity and seasonal or regional fluctuations in the waste composition.
Experiments with SBS machinery provide insight into the interdependencies of throughput rate and input composition on the sorting performance. For this purpose, material mixtures with certain compositions and grain size distributions were created from waste fractions and sorted at various throughput rates. To evaluate the sorting performance of the SBS unit in dependence of the applied load, four assessment factors concerning the output fractions were studied: yield, product purity, recovery and incorrect discharged share of reject particles.
This data should be aquired on three different sorting machines: one is a lab machine, one a chute sorting machine and one a band sorting machine for the same material. The material are four samples of solid recovered fuel, which were taken from the same processing plant at the same position in a two-week rhythm so that the variation of the material composition on one specific point in the plant can be representatively investigated.
With these four samples the same trials were performed on each sorting machine for four different throughput rates and each of the trials was repeated three times to know if repeatability for the sorting results according to the input composition can be determined or not. In sum 144 trials were conducted.
In the end the relation between throughput rate and input composition can be displayed for three sorting machines and mathematical functions for yield, product purity, recovery and incorrect discharged share of reject particles can be derived for all three machines. According to this result a statement can be made whether a model for sorting machine performance related to throughput rate and input composition needs to be set up:
• Once and it is general valid for all SBS machines,
• For each SBS machine construction type (chute or band sorter) or
• For each SBS machine
To control the throughput rate of SBS machines automatically, when the input composition is analysed before the SBS machine in a sorting plant.


Original languageEnglish
Title of host publicationSardina Symposium 2021 Conference Proceedings
ISBN (electronic)2282-0027
Publication statusPublished - 15 Oct 2021
EventSardinia Symposium 2021: 18th International Symposium on Waste Management and Sustainable Landfilling - Forte Village Resort, Santa Margherita di Pula, Italy
Duration: 11 Oct 202115 Oct 2021


ConferenceSardinia Symposium 2021
Abbreviated titleSardinia 2021
CitySanta Margherita di Pula