Sorting system controller
The FFT-based method for analyzing material data streams in sorting plants allows indirect monitoring of machine parameters, addressing accessibility issues and enhancing operational reliability by detecting malfunctions and optimizing settings.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2025-09-16
- Publication Date
- 2026-03-26
AI Technical Summary
Existing sorting and processing plants face challenges in monitoring machine conditions and detecting malfunctions due to limited accessibility and unreliability of machine data, leading to potential operational disruptions and increased wear without direct access to machine parameters.
A method utilizing Fast Fourier Transform (FFT) to analyze material data streams from machines, allowing indirect monitoring of machine parameters by evaluating fluctuations in material flow, enabling early detection of malfunctions and optimizing machine settings without direct access to machine data.
Enables early detection of machine malfunctions and optimization of settings, reducing operational risks and extending machine lifespan by identifying wear and defects through material flow analysis.
Smart Images

Figure EP2025076291_26032026_PF_FP_ABST
Abstract
Description
[0001] Patent application Applicant: Stadler Anlagenbau GmbH Status: 15.09.2025 Our reference: ST 5768-PC / CST - 1 -
[0002] SORTING PLANT CONTROL
[0003] Description
[0004] The present invention relates to the technical field of methods for sorting system control.
[0005] In modern sorting or processing plants, machines are often comprehensively parameterized and controlled, and their reliability and wear are evaluated by recording various parameters, such as current consumption, operating temperature, motor speed, or similar, directly in or on the machine or its components. These parameters allow the operator to draw conclusions about the condition of the machine and thus about the process.
[0006] A disadvantage here is that direct access to the machine data is always required, which can sometimes be difficult to implement due to issues such as the accessibility or reliability of this data. Furthermore, interfaces for the relevant parameters are not always provided, or such parameters cannot always be obtained in a meaningful way, as manufacturers of purchased machines, for example, do not make this data available.
[0007] In particular, the early detection of malfunctions caused by defects or wear is of high economic relevance, since a defect does not usually lead directly to the failure of the machine, but if a defect is not detected or detected early, spontaneous, often time-critical consequences occur, which can lead to a complete shutdown of operations.
[0008] For example, worn bearings can still be operated with increased engine power for a certain period of time, but a failure to detect this development quickly leads to increased engine wear, which can be avoided by replacing the bearings early.
[0009] The technical objective of the invention is therefore to improve the state of the art and propose alternatives. In particular, a method for monitoring sorting or processing plants, especially the machines operating therein, preferably during operation, is to be proposed, without requiring direct access to the machine data of the machine being monitored.
[0010] The technical problem is solved by an object with the technical features according to the independent claim. Advantageous embodiments are the subject of the dependent claims, the description, and the figures.
[0011] All features described in connection with the individual embodiments of the invention can be provided in different combinations in the object according to the invention in order to simultaneously achieve their advantageous effect.
[0012] The scope of protection of the present invention is defined by the patent claims and is not limited by the features explained in the description or illustrated in the figures.
[0013] For the purposes of the invention, the term "material" means that material which is to be sorted or processed by the respective sorting or processing plant or which, at least partially, has already been sorted and processed.
[0014] According to one aspect, the technical problem of the invention is solved by means of a method for identifying machine settings and / or control and / or regulation of a sorting or processing plant comprising at least one machine by monitoring the machine, comprising the method steps: a) recording material data of a material stream processed by the machine as a data stream; b) filtering and evaluating the data stream by means of a fast Fourier transform; c) inferring machine parameters of the machine by means of the evaluation of the
[0015] Fast Fourier transform.
[0016] For the purposes of the invention, material data can also include, for example, physical quantities. It can be provided, for instance, that the recording of material data and / or machine data is carried out using a belt scale and / or a volumetric flow sensor and / or an area flow sensor and / or near-infrared spectroscopy (NIR) and / or by measuring fluctuations in a material volumetric flow and / or by measuring fluctuations in material volumetric flow data and / or by measuring the material flow. [Protection application: Stadler Anlagenbau GmbH, Status: 15.09.2025, Our reference: ST 5768-PC / CST - 3 - and / or by measuring a material class fraction and / or by measuring current consumption.]
[0017] Furthermore, in the context of the invention, the identification of machine settings can, for example, be understood as the identification of error sources.
[0018] The Fast Fourier Transform (FFT) assumes that an oscillation can be represented as the sum of various other harmonic oscillations. These individual summands can be calculated and displayed, or filtered out. Since machines or their settings, unlike fluctuations in the material flow, primarily generate harmonic oscillations or fluctuations in the material flow, these can be quantified using FFT. This allows for advantageous inferences about the machine settings and its parameters, and machine data can be monitored solely based on material data.
[0019] A material stream consisting of waste or refuse generally exhibits inhomogeneities, for example, in its volumetric flow rate and composition. This typically complicates analysis. However, it can be assumed that no periodic oscillations arise from such a material stream. Therefore, FFT signals can be candidates for malfunctions in sorting or processing plants that originate within the respective machines themselves.
[0020] A further advantage is that if machines in the sorting or processing plant are purchased components with undefined technical characteristics or those that change over time, the system can be optimized based on these unknown characteristics. For example, the internal settings of a shredder, such as reversing intervals, can change due to overheating in its shafts. The resulting fluctuations in the material flow would negatively impact the sorting or processing of the plant, which can be detected and prevented early on with this system.
[0021] Furthermore, upcoming maintenance tasks can be identified early and thus planned and implemented to their advantage. Patent application Applicant: Stadler Anlagenbau GmbH Status: 15.09.2025 Our reference: ST 5768-PC / CST - 4 -
[0022] Unlike existing FFT-based analyses, this is an indirect method that analyzes not the machine itself, but its output or downstream secondary machines in order to draw conclusions about the machines installed upstream. It can provide a tool for cases where machine parameters are not directly recorded or accessible, for example, for safety reasons.
[0023] For example, fluctuations in volume flow and / or material flow and / or material class proportion, or similar effects, could then be generated by influencing the material flow, such as by controlling and / or regulating the machine parameters.
[0024] In a technically advantageous embodiment of the method, it is provided that the filtering and evaluation of the data stream by means of fast Fourier transform is carried out in successive time intervals, wherein at each time a data record and / or a Fourier transform data record is transferred to a memory and stored together with this time.
[0025] This allows the created data set and / or data sets to be used advantageously over a longer period of time, and further insights can be gained from them, for example through further analyses.
[0026] Furthermore, in a technically advantageous embodiment of the method, it is provided that the data records of the memory are compared over the time intervals, in particular analyzed for changes occurring in temporal sequence and / or recurring.
[0027] This can also contribute to improving the process.
[0028] Furthermore, in a technically advantageous embodiment of the method, it is provided that the filtering and evaluation of the data stream is carried out continuously and / or discontinuously using Fast Fourier Transform.
[0029] This allows for lossless recording of the data stream and further improves the process. Patent application Applicant: Stadler Anlagenbau GmbH Date: September 15, 2025 Our reference: ST 5768-PC / CST - 5 -
[0030] Furthermore, in a technically advantageous embodiment of the method, it is provided that a smoothing of the data stream, in particular to reduce short-term fluctuations, especially small particle fluctuations, in the data stream, is carried out, preferably before and / or after filtering and evaluating the data stream by means of a fast Fourier transform.
[0031] This allows the recorded data to be processed and thus optimized, which can have an additional positive effect on the process.
[0032] Furthermore, in a technically advantageous embodiment of the method, it is provided that a limitation, in particular to a value range, preferably a frequency range, and / or a clipping, in particular to time segments of the data stream, is carried out.
[0033] Advantageously, for example, start-up intervals and / or plant shutdowns can be excluded from the evaluation. Furthermore, downtimes can be cut out or such intervals ignored during the evaluation.
[0034] Additionally, this method can also be used to advantageously filter out frequencies at which vibrations intended by the machine are to be expected.
[0035] Furthermore, in a technically advantageous embodiment of the method, it is provided that process step a) is carried out at only one location of the sorting plant or processing plant, in particular at only one single data stream recording point of the sorting plant or processing plant.
[0036] Advantageously, the FFT can be performed continuously with targeted analysis of the frequency ranges of all machines preceding the data stream recording point, and the state or settings of all machines before the data stream recording point can be determined with a single data stream.
[0037] Furthermore, in a technically advantageous embodiment of the method, it is provided that, in the case of at least two process lines being processed by one and the same machine, or in the case of at least two process lines being processed by several identical machines, a comparison of the filtering and evaluation of the data stream depending on the respective process line, in particular of the patent application. Applicant: Stadler Anlagenbau GmbH Status: 15.09.2025 Our reference: ST 5768-PC / CST - 6 -
[0038] Frequency mapping of the Fast Fourier Transform is performed, preferably for assessing machine performance.
[0039] An ideal operating state can be advantageously determined in this way by, for example, the presence of identical or similar frequency patterns.
[0040] Furthermore, in a technically advantageous embodiment of the method, it is provided that additional data processing is carried out in the form of filtering and evaluation of the Fast Fourier Transform.
[0041] Advantageously, for example, the data obtained by filtering and evaluation can be processed as a function of frequency, whereby, for example, smoothing and / or aggregation of the oscillation amplitudes over the frequency and subsequent determination of inflection points in a cumulative curve are carried out.
[0042] Furthermore, in a technically advantageous embodiment, it is provided that the method further comprises the process step a1 ) : a1 ) Recording machine data of at least one secondary machine downstream of the machine of a downstream process as a data stream.
[0043] Furthermore, in a technically advantageous embodiment of the method, it is provided that the material flow is designed as a, in particular continuous, material flow of a sorting or processing plant for waste and / or transport of waste material.
[0044] In particular, continuous material flows may be considered as opposed to a sequence or series of parts. The material flow can be designed as an inhomogeneous flow and / or as a bulk material.
[0045] Furthermore, in a technically advantageous embodiment of the method, it is provided that the recording (2) of material data by means of detection means, in particular sensors, cameras, etc., takes place before a machine and after a machine and is compared with each other.
[0046] For example, it may be intended that material data is recorded in front of a machine, such as a sorting unit, by means of detection devices, in particular sensors, cameras, etc. Patent application Applicant: Stadler Anlagenbau GmbH Status: 15.09.2025 Our reference: ST 5768-PC / CST - 7 -
[0047] The process will be recorded, and then the following procedural steps will be carried out.
[0048] In particular, for example, a pre-shredder that reverses at a specific frequency might only show a peak in the machine parameters at that frequency, assuming no disturbance is present. A disturbance occurring, for example, at a drum screen, with a different frequency than that of the pre-shredder, could then be detected as a separate or shifted peak in the spectrum. This would allow for the identification of changes in the fluctuation or interference of an anomaly, enabling subsequent optimization or countermeasures.
[0049] Furthermore, in a technically advantageous embodiment of the method, it is provided that only material data of a material stream processed by the machine is recorded as a data stream, and / or machine data of at least one secondary machine downstream of the machine in a downstream process is recorded as a data stream.
[0050] Furthermore, in a technically advantageous embodiment of the method, it is provided that the method, in particular the sorting or processing plant, is preferably exclusively suitable for processing and / or preparing waste and / or valuable materials and / or recycled materials.
[0051] Furthermore, in a technically advantageous embodiment, the method further comprises a final process step z): z) controlling and / or regulating the sorting or processing plant based on the preceding process steps, in particular initiating measures to remedy or prevent a malfunction in or with the sorting or processing plant.
[0052] Exemplary embodiments of the invention are shown in the figures and are described in more detail below.
[0053] It shows
[0054] Fig. 1 shows a schematic representation of a process;
[0055] Fig. 2 shows another schematic representation of the process.
[0056] Figure 1 shows a schematic representation of a method for identifying machine settings and / or control and / or regulation of at least one patent application. Applicant: Stadler Anlagenbau GmbH. Date: 15.09.2025. Our reference: ST 5768-PC / CST - 8 -
[0057] A machine-comprehensive sorting or processing plant by means of machine monitoring, comprising the following process steps: a) recording material data of a material stream processed by the machine as a data stream, and / or recording machine data of at least one downstream secondary machine of a downstream process as a data stream; b) filtering and evaluating the data stream by means of a Fast Fourier Transform; c) inference of machine parameters of the machine by means of the evaluation of the Fast Fourier Transform.
[0058] This approach, unlike existing FFT-based analyses, allows for the creation of an indirect method that analyzes not the machine itself, but rather its output. This can prove advantageous, for example, in situations where machine parameters are not directly recorded.
[0059] Figure 2 shows a further schematic representation of the process. A material output 11 is generated by a first machine 5, which can be configured, for example, as a shredder, with initially unknown machine parameters 6, which can be configured, for example, as the shredder's reversing intervals. The material output 11 is analyzed on a section 12 by a sensor and / or a camera 10, and material data 7, which can be configured, for example, as volumetric flow rate data, are recorded. Additionally, current machine data from a second machine 8, which can be configured, for example, as a drum screen, can be recorded. Using FFT analysis of the material data 7 and / or machine data 8, machine parameters 6 can be deduced, since they exhibit the same harmonic frequencies generated by the first machine 5.
[0060] The material data 7 and the machine data of the second machine 8 are the data to be measured, while the machine parameters 6 are unknown data to be determined using the method. The FFT analysis of the material data 7, in particular the volumetric flow rate data, can, for example, show fluctuations with the same frequency as machine parameters of the first machine, especially the reversing intervals of the shredder. Thus, it can advantageously be concluded that the fluctuations originate from the first machine and that corresponding parameters should be changed. Alternatively, an FFT analysis of the machine data of the second machine, in particular the drum screen, could show that [This appears to be a separate, unrelated sentence fragment: "Applicant: Stadler Anlagenbau GmbH, Date: 15.09.2025, Our reference: ST 5768-PC / CST - 9 -"]
[0061] Fluctuations with the same harmonic oscillation as the machine parameters of the first machine, in particular as the reversing intervals of the shredder show, according to which a corresponding procedure could be taken.
[0062] Patent application
[0063] Applicant: Stadler Anlagenbau GmbH Date: September 15, 2025 Our reference: ST 5768-PC / CST
[0064] - 10 -
[0065] Reference symbol list
[0066] 1 Procedure
[0067] 2. Recording material or machine data 3. Filtering and evaluating using FFT
[0068] 4. Conclusions about machine parameters
[0069] 5 first machine or shredder
[0070] 6 machine parameters of the first machine or reversing intervals
[0071] 7. Material data or volume flow data 8. Machine data of the second machine
[0072] 9 second machine or drum sieve
[0073] 10 Sensor or camera
[0074] 11. Issue of materials
[0075] Section 12
Claims
Patent application Applicant: Stadler Anlagenbau GmbH Status: 15.09.2025 Our reference: ST 5768-PC / CST - 11 - Claims 1. Method (1) for identifying machine settings and / or control and / or regulation of a sorting or processing plant comprising at least one machine by monitoring the machine, comprising the process steps of: a) recording (2) material data of a material stream processed by the machine as a data stream, b) filtering and evaluating (3) the data stream by means of a fast Fourier transform; c) inferring (4) machine parameters of the machine by means of the evaluation of the fast Fourier transform.
2. Method (1) according to one of the preceding claims, characterized in that the filtering and evaluation of the data stream by means of fast Fourier transformation is carried out in successive time intervals, wherein at each time a data record and / or a Fourier transformation data record is transferred to a memory and stored together with this time.
3. Method (1) according to one of the preceding claims, characterized in that the data records of the memory are compared over the time intervals, in particular analyzed for changes occurring in temporal sequence and / or recurring.
4. Method (1) according to one of the preceding claims, characterized in that the filtering and output of the data stream is carried out continuously and / or discontinuously using Fast Fourier Transform.
5. Method (1) according to one of the preceding claims, characterized in that a smoothing of the data stream, in particular to reduce short-term fluctuations, Patent application Applicant: Stadler Anlagenbau GmbH Date: September 15, 2025 Our reference: ST 5768-PC / CST - 12 - especially small particle fluctuations in the data stream, preferably before and / or after filtering and evaluating the data stream using Fast Fourier Transform.
6. Method (1) according to one of the preceding claims, characterized in that a limitation, in particular to a value range, preferably a frequency range, and / or a clipping, in particular to time segments of the data stream, is carried out.
7. Method (1) according to one of the preceding claims, characterized in that the method step a) is performed at only one location of the sorting plant or Processing plant, especially at only a single data stream recording point, the sorting plant or processing plant.
8. Method (1) according to one of the preceding claims, characterized in that, in the case of at least two process lines being processed by one and the same machine, or in the case of at least two process lines being processed by several identical machines, a comparison of the filtering and evaluation of the data stream is carried out depending on the respective process line, in particular of frequency images of the Fast Fourier Transform, preferably for the purpose of assessing machine performance.
9. Method (1) according to one of the preceding claims, characterized in that additional data processing of filtering and evaluation of the Fast Fourier Transform is carried out. Patent application Applicant: Stadler Anlagenbau GmbH Status: 15.09.2025 Our reference: ST 5768-PC / CST - 13 - 10. Method (1) according to one of the preceding claims, further comprising the method step a1) a1) recording machine data of at least one secondary machine downstream of a downstream process as a data stream.
11. Method (1) according to one of the preceding claims, wherein the material flow is designed as, in particular, a continuous material flow of a sorting or processing plant for waste and / or transport of waste material.
12. Method according to one of the preceding claims, characterized in that the recording (2) of material data by means of recording means, in particular sensors, cameras, etc., is carried out before a machine and after a machine and is compared with each other.
13. Method according to one of the preceding claims, characterized in that only material data of a material stream processed by the machine are recorded as a data stream, and / or machine data of at least one secondary machine downstream of the machine of a downstream process are recorded as a data stream.
14. Method (1) according to one of the preceding claims, characterized in that the method, in particular the sorting or processing plant, is preferably exclusively suitable for processing and / or preparing waste and / or valuable materials and / or recycled materials.
15. Method (1) according to any of the preceding claims, further comprising a final process step z) Patent application Applicant: Stadler Anlagenbau GmbH Date: September 15, 2025 Our reference: ST 5768-PC / CST - 14 - z) Control and / or regulation of the sorting or processing plant based on the preceding process steps, in particular initiating measures to remedy or prevent a malfunction in or with the sorting or processing plant.
Citation Information
Patent Citations
Data pipeline for process control system anaytics
US20160098037A1
Cloud and facility-based machine learning for sorting facilities
WO2023121903A1