Assessment of sensor-based sorting performance for lightweight packaging waste through sensor-based material flow monitoring: Concept and preliminary results

Abstract

Background: Sensor-based sorting (SBS) units are a crucial part of lightweight packaging (LWP) waste sorting plants. Compared to sorting trials under technical lab conditions, significantly lower sorting performances are observed in many LWP sorting plants. One reason for this discrepancy is assumed to be the insufficient material flow representation under real sorting conditions. Aim: This paper aims to quantitatively determine how material flow presentation influences the performance of SBS units on the example of LWP waste. Method: In a case study, near-infrared (NIR) sensors were used to monitor the input, eject and drop fraction of an industrial-scale, NIR-based SBS unit at different throughputs. Result: Preliminary results show that higher occupation densities and insufficient material singling lead to significantly a lower sorting performance both in terms of purity of the eject fractions and yield of eject materials. Conclusion: The findings suggest that much of the discrepancy between theoretically possible and practically achieved LWP sorting performance can be explained by suboptimal material flow presentation. Optimized material flow presentation thus might offer considerable, but so far largely untapped, optimization potentials in LWP sorting.