To achieve the goals of the circular economy, significant improvements in non-hazardous solid waste processing/treatment must be made. Large deficits in the digitalization of mechanical waste treatment plants (smart waste factory) offer great potential for improvement. Real-time material flow monitoring is carried out in very few plants, thus wasting considerable potential for improving plant performance.

This article describes results from the authors’ own practical analyses using sensor-based technologies for monitoring material flows, an on-site investigation in a large waste treatment plant and also in a pilot-scale plant (Technical Line 4.0) using mixed commercial waste (MCW) from Austria. The obtained data shows that the quantitative monitoring of volume and mass flow (via laser triangulation as well as near-infrared (NIR) based monitoring) and material composition (NIR sensor) is possible. The observed fluctuations were categorised in short-, mid- and long-term fluctuations and were led back to their causes, i.e. discontinuous feeding process, material and machine-specific characteristics. Using the quotient of the 90% (Q90) and 10% (Q10) quantiles of time-resolved volume-flow data for the assessment of fluctuations, for the considered time-intervals, resulted in Q90 / Q10 ratios between 3.39 and 4.58. If short-term fluctuations (within the observed time-intervals) are related to the 29.6 s moving average, deviations between 1.8% and 6.8% result. To verify the relevance of such fluctuations, sensor-based sorting (SBS) experiments were conducted, revealing a reduced product purity of 6% due to short-term fluctuations in the feed of the SBS-Machine using light packaging waste (LPW).

为了实现循环经济的目标，必须对无害固体废物的处理/处理进行重大改进。机械废物处理厂（智能废物工厂）的数字化方面存在巨大缺陷，这有很大的改进潜力。实时物料流监控仅在极少数的工厂中进行，因此浪费了改善工厂性能的巨大潜力。

本文介绍了作者自己的实际分析结果，这些分析使用基于传感器的技术监控物料流，在大型废物处理厂以及使用混合商业废物的中试规模工厂（技术路线4.0）进行了现场调查（ MCW），来自奥地利。获得的数据表明，可以对体积和质量流量（通过基于激光三角测量以及基于近红外（NIR）的监视）和材料成分（NIR传感器）进行定量监视。观察到的波动分为短期，中期和长期波动，并归结为它们的原因，即进料过程不连续，物料和机器特定的特性。使用时间分辨的体积流量数据的90％（Q90）和10％（Q10）分位数的商来评估波动，Q90 / Q10比率在3.39和4.58之间。如果短期波动（在观察到的时间间隔内）与29.6 s的移动平均值相关，则会导致1.8％和6.8％之间的偏差。为了验证这种波动的相关性，进行了基于传感器的分选（SBS）实验，结果表明，使用轻包装废料（LPW）的SBS机器进料中的短期波动会导致产品纯度降低6％。