WO2025132960A1

WO2025132960A1 - Systems and methods for determining configuration for sorting systems

Info

Publication number: WO2025132960A1
Application number: PCT/EP2024/087662
Authority: WO
Inventors: Dirk BALTHASAR; Felix Alexander FLEMMING; Marcus Sebastian BAULIG; Jan Jelmer SCHOLTEN; Ritesh Tiwari
Original assignee: Tomra Sorting GmbH
Current assignee: Tomra Sorting GmbH
Priority date: 2023-12-20
Filing date: 2024-12-19
Publication date: 2025-06-26
Anticipated expiration: 2026-06-20

Abstract

Systems and methods for determining configurations for one or more sorting systems for sorting of objects based on simulation are disclosed. Preliminary data specifying properties of a received plurality of objects are received. Sorting of the plurality of objects in one or more sorting systems is then iteratively simulated based on configurations for the one or more sorting systems, on a respective simulation model, and on the received preliminary data. The configurations are updated between each iteration based on one or more target functions. Determined configurations according to which the plurality of objects are to be sorted by the one or more sorting system are then received from the simulation.

Description

_{SYSTEMS AND METHODS FOR DETERMINING CONFIGURATION FOR} _{SORTING SYSTEMS} Technical Field T_{he present disclosure relates to systems and methods for determining} _{configurations for one or more sorting systems for sorting of objects.} Background Throughout a wide range of industries identification, detection, classification and sorting of various objects are frequently required and desired. In its simplest form, manual identification of objects by a person may be employed to advantage when a limited number of objects are to be identified, sorted and classified. The person in question may then, based on his/her knowledge identify and classify the objects concerned. This type of manual identification is however monotonous and prone to errors. Also, the experience level of the operator will significantly influence the results of the operation performed by the operator. Moreover, manual identification of the above kind suffers from low identification speeds. In industry, identification, sorting and classification of a plurality of _{objects (bulk objects) is therefore often performed by sorting systems where} _{the plurality of objects are supplied in form of a continuous object stream.} Such machines are generally faster than an operator and can operate for longer periods of time, hence offering an enhanced overall throughput. Machines of this kind are for instance used in agriculture for fruits and vegetables, and in recycling for identifying and sorting objects and materials that are to be recycled. F_{or a sorting system for sorting received unsorted plurality of objects in} a sorting assignment, the unsorted objects may be of any kind, such as plastic, metal, minerals, glass, wood, etc. The properties, such as material, color, and size, of a received plurality of objects may vary between sorting _{assignments. Furthermore, the sorting system may be configured in a large} _{number of ways by setting different parameters and optimization of sorting} _{may be performed for several sorting systems. Additionally, the requirements} _{on the sorted output from the sorting system, e.g. in relation to purity of the} _{output, the total amount of objects having a particular property in the output,} _{and timing, may also vary over time and between sorting assignments.} Hence, determining a configuration of the sorting system for sorting a received plurality of objects is a complex task and a preferred configuration may also vary over time for the same plurality of objects. Hence, enhanced systems and methods for determining configurations for one or more sorting systems for sorting of objects are desired.

It is a general object of the inventive concept to mitigate, alleviate or eliminate one or more of the above-identified deficiencies in the art and disadvantages singly or in any combination and at least partly solve the above-mentioned problems. Embodiments provide approaches for one or more of the following: -_{determining configurations for one or more sorting systems for sorting} _{of objects based on simulation.} _{- enabling improved sorting of objects.} _{- enabling sorting of objects with minimal waste.} _{- enabling a faster sorting of objects.} _{- enabling a more environmentally friendly sorting of objects.} _{- enabling sorting of objects with minimal human intervention.} _{- enabling a more energy efficient sorting of objects} The present inventive concept provides an overall improved system for _{enabling sorting received unsorted objects. The unsorted objects may be of} _{any kind, such as plastic, metal, minerals, glass, wood, etc. The system may} sort objects based on the identified properties with minimal manual _{intervention. The system may facilitate a more efficient sorting process and} thus save time and electrical power, and thereby having an overall smaller climate impact. The system may further account for local variations in the property- based distribution of objects, hence being flexible with respect to spatial or temporal variation of objects to be sorted. The present inventive concept is at least partly based on the realization by the inventor that optimal and/or desired sorting results may vary over time and in location. Furthermore, since there are typically a lead time from receiving a plurality of unsorted objects for sorting until the sorting can be performed, such variations may result in change of optimal and/or desired sorting result during that lead time. The term “processing unit” is to be interpreted broadly, and is, e.g., not limited to be a single processing unit in a local computer/server. Rather, by way of example, the “processing unit” may form part of a cloud based server, being one or more units having processing capabilities, wherein different units may be located or distributed differently, e.g., one unit may be present on a cloud based server while another unit may be locally present at a sorting _{system in relation to which the system is implemented.} _{The term “demand […] for objects having a specific property” in the} _{context of the appended claims may be understood as a demand of such} _{objects based on availability of objects having the specific property, based on} _{energy considerations, resource availability considerations, economic} _{grounds, or environmental considerations etc. Hence, the system may by way} _{of example account for a current change in demand of copper, a current} _{change in availability of transparent glass, or the like.} “Real-time data” may refer to data based on a current _{value/price/availability of a specific object type, or any other time-dependent} _{variable related to a specific object type. The real-time data may further} comprise a plurality of parameters such as a value/availability parameter of the specific object type and a present amount of the specific object type in a specific batch of objects to be sorted. Hence, a total value of the specific object type in the specific batch of objects may be estimated, and sorting may be performed based on this total value. _{The terms “sorting system” and “sorting plant” are used} _{interchangeably and mean a system for sorting including one or more sorters.} The term “configuration” in relation to a sorting system in the context of the appended claims relates to how the sorting system is configured and may for example mean how different parameters are set for the sorting system. A_{ccording to a first aspect, a system for determining configurations for} _{a first sorting system and a second sorting system for sorting of objects based} on simulation is provided. The system comprises a processing unit comprising circuitry configured to execute a data receiving function, a _{configuration simulation function, and a determining function. The data} receiving function is configured to receive preliminary data specifying properties of a received plurality of objects. The configuration simulation _{function is configured to iteratively simulate sorting the plurality of objects in} the first sorting system based on a first configuration for the first sorting system, on a first simulation model for the first sorting system, and on the received preliminary data, and simulate sorting an output from the first sorting system in the second sorting system based on a second configuration for the second sorting system, on a second simulation model for the second sorting system, and on the output from the first sorting system according to simulated sorting of the plurality of objects in the first sorting system. The first configuration and the second configuration are updated between each iteration based on one or more of a first target function for the first sorting system, a second target function for the second sorting system, and a third target function common to the first and second sorting systems. The _{determining function is configured to receive from the simulation configuration} _{function a determined first configuration according to which the plurality of} objects are to be sorted by the first sorting system and a determined second _{configuration according to which the output from the first sorting system} should be sorted by the second sorting system One or more of the first target function, the second target function, and _{the third target function depends on real-time data regarding demand in one} or more subsequent processes for objects having one or more of the properties of the received plurality of objects. The circuitry of the system according to the first aspect may further comprise a detector configured to detect properties of a subset of objects of _{the plurality of objects, thereby producing the preliminary data. The data} _{receiving function may the be further configured to receive the preliminary} data from the detector. T_{he circuitry of the system according to the first aspect may be further} _{configured to execute a feedback function configured to receive feedback} data from an actual physical sorting of objects based on the determined first configuration and the determined second configuration, and an adjustment _{function configured to adjust the first simulation model and the second} _{simulation model based on the feedback data.} _{According to a second aspect, a system for determining a configuration} _{for a sorting system for sorting of objects based on simulation is provided.} The system comprises a processing unit comprising circuitry configured to execute a data receiving function, a configuration simulation function, and a _{determining function. The data receiving function is configured to receive} preliminary data specifying properties of a received plurality of objects. The configuration simulation function is configured to iteratively simulate sorting _{the plurality of objects in the sorting system based on a configuration for the} sorting system, on a simulation model, and on the received preliminary data, and wherein the configuration is updated between each iteration based on a _{target function. The determining function is configured to receive from the} _{configuration simulation function a determined configuration according to} _{which the plurality of objects are to be sorted by the sorting system.} _{The target function may depend on real-time data regarding demand in} one or more subsequent processes for objects having one or more of the properties of the received plurality of objects. T_{he system according to the second aspect may further comprise a} detector configured to detect properties of a subset of objects of the plurality of objects, thereby producing the preliminary data. The data receiving function is then further configured to receive the preliminary data from the detector. T_{he circuitry of the system according to the second aspect may further} be configured to execute a feedback function and an adjustment function. The _{feedback function is configured to receive feedback data from an actual} _{sorting of objects based on the decided configuration, and the adjustment} function is configured to adjust the simulation model based on the feedback data. In the system according to the first aspect or according to the second _{aspect, the data receiving function may further be configured to receive} preliminary data for a further received plurality of objects. According to a third aspect, a computer implemented method for determining configurations for a first sorting system and a second sorting _{system for sorting of objects based on simulation. The method comprises} receiving preliminary data specifying properties of a received plurality of objects. The method further comprises iteratively simulating sorting the plurality of objects in the first sorting system based on a first configuration for the first sorting system, on a first simulation model for the first sorting system, and on the received preliminary data, and simulating sorting an output from the first sorting system in the second sorting system based on a second configuration for the second sorting system, on a second simulation model for the second sorting system, and on the output from the first sorting system according to simulated sorting of the plurality of objects in the first sorting system, and wherein the first configuration and the second configuration is updated between each iteration based on one or more of a first target function for the first sorting system, a second target function for the second sorting system, and a third target function common to the first and second sorting _{systems. The method further comprises receiving from the iterative simulation} _{a determined first configuration according to which the plurality of objects are} _{to be sorted by the first sorting system and a determined second configuration} _{according to which the output from the first sorting system should be sorted} by the second sorting system. The above-mentioned optional additional features of the system of the _{first aspect, when applicable, apply to the method according to the third} aspect as well. In order to avoid undue repetition, reference is made to the above. A_{ccording to a fourth aspect, a computer implemented method for} _{determining a configuration for a sorting system for sorting of objects is} _{provided. The method comprises receiving preliminary data specifying} properties of a received plurality of objects. The method further comprises iteratively simulating sorting the plurality of objects in a sorting system based on a configuration for the sorting system, on a simulation model, and on the received preliminary data, and wherein the configuration is updated between each iteration based on a target function. The method further comprises _{receiving from the iterative simulation a determined configuration according to} _{which the plurality of objects are to be sorted by the sorting system.} _{The above-mentioned optional additional features of the system} _{according to the second aspect, when applicable, apply to the method} _{according to the fourth aspect as well. In order to avoid undue repetition,} reference is made to the above. A_{ccording to a fifth aspect, a non-transitory computer-readable storage} medium having stored thereon program code portions for implementing the method according to the first aspect or the method according to the second _{aspect when executed on a device having processing capabilities.} The above-mentioned optional additional features of the method of the third aspect and of the method according to the forth aspect, when applicable, apply to the non-transitory computer readable storage medium according to the fifth aspect as well. In order to avoid undue repetition, reference is made to the above. T_{he feedback data may comprise real-time data of blockage in a flow} of material in the first sorting system, the second sorting system, between the first sorting system and the second sorting system, and/or after the second sorting system. _{The feedback data may comprise real-time data of blockage in a flow} _{of material in the first sorting system, and/or after the first sorting system.} The first target function, the second target function, and/or the third _{target function may depend on real-time data of blockage in a flow of material} in the first sorting system, the second sorting system, between the first sorting system and the second sorting system, and/or after the second sorting system. The term “blockage” in the context of the present disclosure refers to a condition where a flow of materials is obstructed, disrupted, or completely stopped within a specific part of the sorting system. Blockage can occur due to the accumulation of objects, improper alignment of materials, mechanical failure, or foreign objects getting stuck in conveyors, chutes, feeders, or other machinery. Blockages can result in downtime, reduced efficiency, or damage to the equipment, necessitating immediate identification and resolution to _{restore normal operations. Blockage can be complete (a material flow is} completely impacted by blockage) or partial (only a part of the material flow is impacted by blockage whereas the remainder of the material flow can _{proceed according to expectations).} _{Real-time data of blockage may comprise data of at least one blockage} occurring at any one of at least one given location. Real-time data of blockage _{may comprise data indicative of at least one future blockage at one of said at} _{least one given location. The at least one given location may be a monitored} location. “_{Monitored location” may refer to a location that is being monitored,} _{continuously or intermittently, by a surveillance camera or a sensor} _{arrangement of a sorting system.} The system may comprise at least one surveillance camera. The first sorting system and/or the second sorting system may comprise at least one sensor arrangement. Here, sensor arrangement may be an image sensor, such as an VIS, visible spectrum, camera, or any other type of sensor configured to provide sensor data which can be represented as image data. The at least one given location may also be a non-monitored location _{located somewhere between a first monitored location and a second} _{monitored location. Real-time data of blockage of the non-monitored location} _{may comprise information inferred by a comparison of sensor data of the first} _{monitored location and sensor data of the second monitored location.} In view of the above, the system may be configured to take blockage _{and/or risk of blockage into account when updating a configuration of the first} sorting system and/or a configuration of the second sorting system. The system may be configured to update the first configuration of the first sorting _{system to reduce blockage or to reduce a risk of blockage. The system may} be configured to update the second configuration of the second sorting _{system to reduce blockage or to reduce a risk of blockage. When blockage is} _{accounted for via the first target function, the second target function, and/or} _{the third target function, the system may determine a configuration for the first} _{sorting system and/or the second sorting system to reduce the risk of} blockage and consequently allow for a higher uptime, thus improving the efficiency of the first sorting system and/or the second sorting system. Risk of blockage may be based on at least one parameter, or any combination of a plurality of parameters, that is measured or at least derivable by at least one blockage sensor (e.g., a flow sensor or a proximity sensor) or by at least one virtual blockage sensor (e.g., a VIS-camera of a sorting _{system). Here, virtual blockage sensor which is configured to provide sensor} data for another purpose other than blockage detection, but said sensor data may still be used to infer blockage. T_{he at least one parameter or any one of the plurality of parameters} may be any one of material accumulation, time to recovery, throughput impact, sensor feedback tolerance, and system downtime threshold, each of which are described in the following. Threshold of material accumulation may indicate or specify a maximum permissible buildup of materials at specific points in the system, quantified in terms of weight, volume, or height, before the situation is classified as a blockage. Time to recovery may indicate or specify a time duration for which the system can endure an obstruction and resolve it autonomously—using mechanisms such as vibrations, reverse motion, or increased motor power— without requiring manual intervention. Throughput impact may indicate or specify a maximum allowable reduction in material throughput (e.g., as a percentage of nominal capacity) that can occur during a partial blockage while maintaining acceptable performance levels. Sensor feedback tolerance may indicate or specify an acceptable deviation in sensor readings (e.g., proximity, flow, or pressure sensors) that _{indicates a blockage without necessitating a system shutdown.} System downtime threshold may indicate or specify an acceptable frequency or duration of downtime caused by blockages within a defined _{operational period (e.g., "no more than 5 minutes of downtime per hour due to} blockages"). The system may be configured to determine at least a first blockage tolerance based at least on the first configuration of the first sorting system and/or at least a second blockage tolerance based at least on the second _{configuration of the second sorting system. Each sorting system may be} _{assigned at least one blockage tolerance. Each blockage tolerance may} specify or indicate a maximum acceptable risk of blockage at any given location within the sorting system. B_{lockage tolerance may be a fraction (or percentage) of the system’s} maximum allowable risk. The blockage tolerance may be based at least on _{the maximum risk and a safety margin subtracted from the maximum risk. As} _{a non-limiting example, if the maximum risk is 1.0 (100%) and the safety} margin is 0.2 (20%), the blockage tolerance is 0.8 (80%), meaning that the system can tolerate up to 80% of the maximum risk before considering it a blockage. The blockage tolerance, or the system’s maximum allowable risk, _{may be based at least on a nominal capacity at a given location within the} _{sorting system. The nominal capacity at a given location within the sorting} system may depend on the configuration determined for said sorting system. Thus, when updating the configuration for said sorting system, the blockage tolerance may be updated also. As mentioned, “Real-time data” may refer to data based on a current value/price/availability of a specific object type, or any other time-dependent _{variable related to a specific object type. However, “Real-time data” may,} _{alternatively or in combination, refer to visual data of at least one object or a} material flow. V_{isual data may include at least one image and/or at least one video} _{captured by means of said at least one sensor arrangement or said at least} _{one surveillance camera. The at least one sensor arrangement and/or the at} _{least one surveillance camera may be communicatively coupled with the} _{processing unit (e.g., the circuitry in particular) to provide said at least one} _{image and/or said at least one video.} The system may comprise a monitoring system. The monitoring system may comprise at least one monitor display. The at least one sensor arrangement and/or the at least one surveillance camera may be communicatively coupled with a monitoring system. Each monitor display may be configured to display in at least one window said at least one image or said at least one video. A single monitor display may be configured to display _{a plurality of images and/or a plurality of videos in respective windows. The} monitoring system may be arranged in a monitoring room. An operator stationed in the monitoring room may thus observe the at least monitor _{display to identify a blockage in said at least one image and/or said at least} one video. The at least one sensor arrangement and/or the at least one _{surveillance camera may be communicatively coupled with a deep learning} _{system. The deep learning system may comprise a neural network configured} to process said at least one image and/or said at least one video as input data and generate an output indicative of whether said at least one image and/or said at least one video depicts a blockage. The real-time data of blockage may comprise said output from said deep learning system. As mentioned, the deep learning system comprises a neural network. _{The neural network comprises an input layer configured to receive said at} least one image and/or said at least one video as input data. The neural network comprises one or more hidden layers, each hidden layer comprising a plurality of nodes configured to process the input data based on a set of weighted connections. The neural network comprises an output layer configured to provide the output based on how the one or more hidden layers processes the input data, which output is indicative of whether said at least one image and/or said at least video depicts a blockage. The neural network may be trained using a training dataset and a learning algorithm to perform a _{task of detecting whether said at least one image and/or video depicts a} _{blockage. The learning algorithm may be any applicable learning algorithm.} _{The training dataset may include at least one labeled image and/or at} _{least one labeled video. The labeled image and/or the labeled video may be} provided in various ways, e.g., by image and/or video being labeled in real- time (by an operator) or labeled retroactively. The monitoring system may comprise an interface allowing an operator to provide an input when he/she observes a blockage in at least one image and/or at least one video, thereby allowing the operator to label in real-time said at least one image and/or video as depicting a blockage. The monitoring system may be configured to interface with the deep learning system. The monitoring system may be configured to provide a blockage alert associated with at least one image and/or video if the deep learning system determines said at least one image and/or video depicts a _{blockage. This may advantageously alert the operator of a blockage in said} _{one or more flow of materials. In case of a false alert, the operator may} override said blockage alert. The monitoring system may be configured to re- label said at least one image and/or video whenever an operator overrides a blockage alert. T_{he monitoring system may be configured to store an anomaly} _{database and update said anomaly databased with entries of anomalies and} associated information (date, time, location, event, etc.). The anomaly database may be updated to store entries of blockages. An entry of blockage may comprise information about date, time, and location of a blockage, and also at least a reference to said at least one image and/or video in which a _{blockage was determined to be depicted. The entry may further comprise} information about whether a blockage alert was overridden. The entry may further comprise information about other real-time data. T_{he deep learning system may be configured to train the neural} network on a selection of training data, which selection of training data is _{associated with at least on age and/or a system configuration. Training data} _{associated with a certain age may be omitted from said selection of training} data. Training data associated with a certain system configuration may be omitted from said selection of training data. As a non-limiting example, training data associated with a certain image sensor may be omitted if said _{certain image sensor is replaced.} “Real-time data” may also comprise auxiliary data of the material flow at any of the at least one monitored locations. The system may be configured to use auxiliary data in combination with visual data to facilitate training of the deep learning system or facilitate detection of blockage. As a non-limiting example, auxiliary data may comprise material _{density. High-density materials can place a significant load on the system and} may cause clogging if the equipment is not designed to handle such materials. As a non-limiting example, auxiliary data may comprise object size and _{shape. Irregularly shaped or oversized objects can accumulate in certain} areas, leading to obstructions in the material flow. As a non-limiting example, auxiliary data may comprise moisture _{content. Materials with high moisture contents or sticky properties tend to} clump together, which increases the likelihood of blockages in the sorting plant. As a non-limiting example, auxiliary data may comprise throughput rate. Excessive input of materials into the system at a high rate can overwhelm the machinery, creating jams and disrupting operations. As a non-limiting example, auxiliary data may comprise conveyor _{speed. Incorrectly set conveyor speeds can result in material misalignment or} excessive buildup, which may eventually block the system. As a non-limiting example, auxiliary data may comprise information _{about component wear and tear. Worn-out or degraded components in the} machinery can lead to uneven material handling, which might contribute to clogs and system malfunctions. As a non-limiting example, auxiliary data may comprise information _{indicative of motor load of a conveyor motor. An unexpected or sudden} increase in the load on a conveyor motor may indicate that a blockage is present and restricting movement. If the system incorporates a pneumatic transport system, which it may, _{auxiliary data may comprise pressure levels of a pneumatic transport system.} In pneumatic transport systems, an increase in pressure levels can signal that airflow is obstructed due to a blockage. As a non-limiting example, auxiliary data may comprise information about component temperature. Unusual overheating of specific components in the system can suggest mechanical issues or blockages impeding normal operations. As a non-limiting example, auxiliary data may comprise information _{about vibration patterns. Irregular or abnormal vibration patterns detected in} machinery can indicate interruptions in material flow, often caused by blockages. As a non-limiting example, auxiliary data may comprise proximity _{sensor feedback. Data from proximity sensors may show material} accumulation at critical locations, such as choke points, which could suggest the presence of a blockage. As a non-limiting example, auxiliary data may comprise flow sensors _{data. Flow sensors provide feedback on the movement of materials and can} indicate interruptions or abnormalities in the material stream. As a non-limiting example, auxiliary data may comprise information _{about contaminants or foreign objects. The presence of non-sorted objects or} debris in the material stream can obstruct the flow, becoming lodged in machinery and causing system downtime. A further scope of applicability of the present invention will become apparent from the detailed description given below. However, it should be understood that the detailed description and specific examples, while indicating preferred embodiments of the invention, are given by way of illustration only, since various changes and modifications within the scope of the invention will become apparent to those skilled in the art from this detailed description. H_{ence, it is to be understood that this invention is not limited to the} _{particular component parts of the device described or acts of the methods} _{described as such device and method may vary. It is also to be understood} _{that the terminology used herein is for purpose of describing particular} _{embodiments only and is not intended to be limiting. It must be noted that, as} _{used in the specification and the appended claim, the articles "a," "an," "the,"} _{and "said" are intended to mean that there are one or more of the elements} _{unless the context clearly dictates otherwise. Thus, for example, reference to} _{"a unit" or "the unit" may include several devices, and the like. Furthermore,} _{the words "comprising", “including”, “containing” and similar wordings does} _{not exclude other elements or steps.} _{Brief Description of the Drawings} _{The above, as well as additional objects, features, and advantages of} _{the present invention, will be better understood through the following} _{illustrative and non-limiting detailed description of preferred embodiments,} _{with reference to the appended drawings, where the same reference} _{numerals will be used for similar elements.} _{Fig. 1 schematically shows a flowchart of a method for determining} _{configurations for one or more sorting systems for sorting of objects based on} simulation. _{Fig. 2 shows a block diagram of a system for determining} _{configurations for one or more sorting systems for sorting of objects based on} simulation. F_{ig. 3 illustrates surface coverage in relation to throughput.} _{Figs. 4a and 4b illustrate cascades of sorters.} _{Figs. 5a and 5b illustrate cascades of sorters including at least one} _{feedback loop.} _{Fig. 6 illustrates optimization of a single sorting system using} simulation. F_{igs. 7a and 7b illustrates optimization of two sorting systems using} simulation. F_{ig. 8 schematically illustrates a sorting plant comprising a deep} _{learning system trained to detect blockages.} Detailed Description T_{he present invention will now be described more fully hereinafter with} _{reference to the accompanying drawings, in which currently preferred} _{embodiments of the invention are shown. This invention may, however, be} _{embodied in many different forms and should not be construed as limited to} _{the embodiments set forth herein; rather, these embodiments are provided for} _{thoroughness and completeness, and to fully convey the scope of the} _{invention to the skilled person.} Figure 1 shows a flow chart in relation to embodiments of a method for determining configurations for one or more sorting systems for sorting of objects based on simulation. T_{he method 100 may be performed in relation to a single sorting} system or in relation to two or more sorting systems for sorting of objects based on simulation. Embodiments of the method in relation to a single sorting system is described in the following. T_{he method 100 comprises receiving 120 preliminary data specifying} properties of a received plurality of objects. _{The method 100 further comprises iteratively simulating S130, C135,} _{S140 sorting the received plurality of objects in the sorting system. This} _{comprises simulating S130 sorting the received plurality of objects in the} _{sorting system based on a configuration for the sorting system. Two} conditions are then checked C135, namely if 1) the result of the simulation converges and 2) if a maximum number of iterations have been performed. If none of the conditions are fulfilled, the configuration is updated S140 based _{on a target function and the method proceeds to simulating S130 sorting} based on the new configuration for the sorting system. If at least one of the _{two conditions is fulfilled, the method proceeds to receiving S150 from the} _{iterative simulation S130, C135, S140 a determined configuration according} _{to which the plurality of objects are to be sorted by the sorting system.} The target function may depend on real-time data regarding demand in one or more subsequent processes for objects having the one or more respective properties of the properties of the received plurality of objects. The configuration for the sorting system may for example relate to setting of a set of parameters of the sorting system. Examples of such parameters are given below in “Parameters that can be modified for different simulation results”. T_{he method 100 may further comprise detecting S110 properties of a} _{subset of objects of the plurality of objects by means of a detector thereby} producing the preliminary data. For example, a random subset of objects of _{the plurality object is selected such that the number of objects of the subset is} substantially smaller than the total amount of objects of the plurality of objects. The subset may for example be checked using a detector for _{detecting the properties. The detector may be of a spectroscopic or optical} type. Properties of the subset of the received plurality of objects may then be detected by spectroscopy such that the detector detects a material type of respective object by comparing/matching a detected spectrum to a database comprising a plurality of spectra and thus determining the material type of respective object. Spectral methods in connection hereto may be X-ray, Raman, or the like. For optical detectors, object detection/image recognition may be utilized to provide detection and determining properties or material types of the subset of objects. Based on the determined properties of the subset, the properties of the plurality of objects may then be estimated to produce the preliminary data. The preliminary data is then received S120 from the detector. Furthermore, actual physical sorting according to the determined _{configuration may be performed S160 in the sorting system. Feedback may} _{then be received S170 from the actual physical sorting by the sorting system} based on the decide configuration of the sorting system. The simulation _{model used in the iterative simulation S130, C135, S140 may then be} _{adjusted S180 based on the feedback data.} _{The sorting system may further comprise a plurality of sorters. The} configuration then comprises configuration for each of the plurality of sorters. _{The configuration used in the iterative simulation S130, C135, S140 then} _{comprises setting according to which respective configuration the plurality of} objects are to be sorted by a respective sorter of the plurality of sorters. Furthermore, the method may further determine a preferred order of sorting. By way of example, a first sorter may sort out polyethylene terephthalate, _{PET, from the objects, a second sorter may sort the PET based objects} according to their visual transparency, and a third sorter may sort the remaining (at least partially) transparent PET objects based on their color. Determining and performing sorting this way may facilitate a substantially _{optimized sorting, and may thus facilitate sorting accuracy, and save time and} power. Furthermore, preliminary data for a further received plurality of objects _{may be received S120. The method 100 may then further determine a} _{preferred order of sorting the received plurality of objects and the further} received plurality of objects. T_{he method 100 may be performed in relation to two or more sorting} systems. Embodiments of the method 100 in relation to two sorting systems is described in the following. _{The method 100 then comprises receiving S110 preliminary data} specifying properties of a received plurality of objects. The method 100 further _{comprises iteratively simulating S130, C135, S140 sorting the received} plurality of objects in the first sorting system and the output from the first sorting system in the second sorting system. This comprises simulating S130 sorting the plurality of objects in the first sorting system based on a first configuration for the first sorting system, on a first simulation model for the first sorting system, and on the received preliminary data, and to simulate sorting the output from the first sorting system in the second sorting system based on a second configuration for the second sorting system, on a second simulation model for the second sorting system, and on the output from the first sorting system according to simulated sorting of the plurality of objects in the first sorting system. Two conditions are then checked C135, namely if 1) _{the result of the simulation converges and 2) if a maximum number of} iterations have been performed. If none of the conditions are fulfilled, the first configuration and the second configuration are updated S140 based on one or more of a first target function for the first sorting system, a second target function for the second sorting system, and a third target function common to the first and second sorting systems and the method proceeds to simulating S130 sorting based on the new candidate configuration for the sorting _{system. If at least one of the two conditions is fulfilled, the method proceeds} _{to receiving S150 from the iteratively simulation S130, C135, S145 a} _{determined first configuration according to which the plurality of objects are to} _{be sorted by the first sorting system and a determined second configuration} _{according to which the output from the first sorting system should be sorted} by the second sorting system. One or more of the first target function, the second target function, and _{the third target function may depend on real-time data regarding demand in} one or more subsequent processes for objects having the one or more respective properties of the properties of the received plurality of objects. T_{he method 100 may further comprise receiving S170 feedback data} from an actual physical sorting of objects based on the on the determined first _{configuration and the determined second configuration, and adjusting S180} _{the first simulation model and the second simulation model based on the} feedback data. T_{he configuration may be updated S140 in relation to a number of} different parameters, e.g. the parameters identified hereinbelow in the section “Parameters that can be modified for different simulation results”. The convergence in relation to the target function or target functions is _{in relation to the resulting sorting results for the simulations after updating of} the configuration. It is to be noted that, in case of updating the simulation model in response to feedback from actual physical sorting by sorting system or sorting systems, there may also be a convergence in relation to the simulation model. F_{igure 2 shows a block diagram in relation to embodiments of a system} for determining configurations for one or more sorting systems for sorting of objects based on simulation. T_{he system 200 comprises a processing unit 205. The processing unit} is, e.g., not limited to be a single processing unit in a local computer/server. Rather, by way of example, the processing unit may form part of a cloud based server, being one or more units having processing capabilities, wherein different units may be located or distributed differently, e.g., one unit may be present on a cloud based server while another unit may be locally present in _{a sorting system in relation to which the system is implemented/placed. The} processing unit 205 comprises circuitry 210. The circuitry 210 is configured to carry out functions of the system 200. The circuitry 210 may include a _{processor or processors 212, such as for example one or more of a central} processing unit (CPU), graphical processing unit (GPU), tensor processing unit (TPU), microcontroller, or microprocessor. The processor 212 is configured to execute program code. The program code may for example be _{configured to carry out the functions of the system 200.} _{The system 200 may further comprise a memory 220. The memory} 220 may be one or more of a buffer, a flash memory, a hard drive, a removable media, a volatile memory, a non-volatile memory, a random access memory (RAM), or another suitable device. In a typical arrangement, the memory 220 may include a non-volatile memory for long term data storage and a volatile memory that functions as device memory for the circuitry 210. The memory 220 may exchange data with the processing unit 205/circuitry 210 over a data bus. Accompanying control lines and an address bus between the memory 220 and the processing unit 205/circuitry 210 also may be present. F_{unctions of the system 200 may be embodied in the form of} executable logic routines (e.g., lines of code, software programs, etc.) that are stored on a non-transitory computer readable medium (e.g., the memory 220) _{of the system 200 and are executed by the circuitry 210 (e.g., using the} _{processor or processors 212). Furthermore, the functions of the system 200} may be a stand-alone software application or form a part of a software _{application that carries out additional tasks related to the system 200. The} described functions may be considered a method that a processing unit, e.g., _{the processor or processors 212 of the circuitry 210 is configured to carry out.} Also, while the described functions may be implemented in software, such functionality may as well be carried out via dedicated hardware or firmware, or some combination of hardware, firmware and/or software. The circuitry 210 is configured to execute a data receiving function _{221, an configuration simulation function 222, a determining function 223, and} _{optionally a sorting function 224, a feedback function 225 and an adjustment} function 226. The system 200 may be adapted for determining a configuration for a single sorting system for sorting of objects based on simulation. The configuration of the data receiving function 221, the configuration simulation _{function 222, the determining function 223, and the optional sorting function} _{224, feedback function 225 and adjustment function 226 for a single sorting} system is described in the following. The data receiving function 221 is configured to receive preliminary data specifying properties of a received plurality of objects; The configuration simulation function 222 is configured to iteratively simulate sorting the plurality of objects in the sorting system based on a configuration for the sorting system, wherein the configuration is based on a simulation model, and on the received preliminary data, and wherein the configuration is updated between each iteration based on a target function; and T_{he determining function 223 is configured to receive from the} _{configuration simulation function a determined configuration according to} _{which the plurality of objects are to be sorted by the sorting system.} _{The target function may depend on real-time data regarding demand in} one or more subsequent processes for objects having the one or more respective properties of the properties of the received plurality of objects. T_{he system 200 may further comprise a detector 230 configured to} _{detect properties of a subset of objects of the plurality of objects, thereby} producing the preliminary data. The detector may be of a spectroscopic or optical type. Properties of the subset of the received plurality of objects may then be detected by spectroscopy such that the detector detects a material _{type of respective object by comparing/matching a detected spectrum to a} database comprising a plurality of spectra and thus determining the material type of respective object. Spectral methods in connection hereto may be X- ray, Raman, or the like. For optical detectors, object detection/image recognition may be utilized to provide detection and determining properties or material types of the subset of objects. Based on the determined properties of the subset, the properties of the plurality of objects may then be estimated to produce the preliminary data. The data receiving function is then further configured to receive the preliminary data from the detector. T_{he circuitry 210 may further be configured to execute sorting function} _{224, a feedback function 225, and an adjustment function 226. The sorting} function 224 is configured to perform actual physical sorting of the plurality of data in the sorting system according to the determined configuration received _{form the determining function 223. The feedback function 224 is configured to} _{receive feedback data from the actual physical sorting of objects based on the} _{determined configuration. The adjustment function 225 is configured to adjust} the simulation model based on the feedback data. I_{n embodiments, the system 200 comprises a plurality of sorters. The} configuration then comprises configuration for each of the plurality of sorters. The configuration used in the configuration simulation function S222 then _{comprises setting according to which respective configuration the plurality of} objects are to be sorted by a respective sorter of the plurality of sorters. Furthermore, the system 200 may determined which order the sorters should be used. By way of example, a first sorter may sort out polyethylene terephthalate, PET, from the objects, a second sorter may sort the PET based objects according to their visual transparency, and a third sorter may sort the remaining (at least partially) transparent PET objects based on their color. _{Determining and performing sorting this way may facilitate a substantially} optimized sorting, and may thus facilitate sorting accuracy, and save time and power. The data receiving function 221 may further be configured to receive preliminary data for a further received plurality of objects. The system 200 may then provide a preferable order of sorting the received plurality of objects and the further received plurality of objects based on properties of the objects. The system 200 may also be adapted for determining configurations for at least two sorting systems for sorting of objects based on simulation. The configuration of the data receiving function 221, the configuration simulation function 222, the decision function 223, and the optional feedback function 224 and adjustment function 225 for two sorting systems is described in the following. T_{he data receiving function 221 is configured to receive preliminary} data specifying properties of a received plurality of objects. The configuration _{simulation function 222 is configured to iteratively simulate sorting the plurality} of objects in the first sorting system based on a first configuration for the first sorting system, on a first simulation model for the first sorting system, and on the received preliminary data, and to simulate sorting an output from the first sorting system in the second sorting system based on a second configuration for the second sorting system, on a second simulation model for the second sorting system, and on the output from the first sorting system according to simulated sorting of the plurality of objects in the first sorting system. The first configuration and the second configuration are updated between each iteration based on one or more of a first target function for the first sorting system, a second target function for the second sorting system, and a third target function common to the first and second sorting systems. The _{determining function 223 is configured to receive from the configuration} _{simulation function a determined first configuration according to which the} _{plurality of objects are to be sorted by the first sorting system and a} _{determined second configuration according to which the output from the first} _{sorting system should be sorted by the second sorting system. One or more} of the first target function, the second target function, and the third target _{function may depend on real-time data regarding demand in one or more} subsequent processes for objects having one or more of the properties of the _{received plurality of objects. Objects may be decided to be sorted differently} _{among sorting systems located on different locations despite the objects to be} sorted have a similar distribution among the plants. Such differing decisions may be made according to a local demand, price, or availability on a specific plant or a region in connection to the specific plant. T_{he above-mentioned optional additional features of the method 100} described in relation to Figure 1, when applicable, apply also to the system _{100 described in relation to Figure 2.} Simulation of one sorter Simulating a sorting device requires a model of the input of the sorter _{and a sorting heuristics, that splits the input of the sorter into different} fractions. Typically, two fractions are used, but there are also sorters with more than two output fractions like double or multiple valve block systems or robotic sorters. The heuristic for sorting calculates from the input material the output material. We model the input of the sorter as a material vector ^, that contains in each dimension materials in metric tons per hour. The sorting model ^ transforms the input vector into the outputs, represented as a matrix ^. Each column of the matrix contains the output per valve block or fraction. The behavior of the sorter is furthermore relying on the capabilities of the sorter and the sorting configuration ^. ^_{(^, ^) = ^} The input material is divided completely into the different outputs, i.e., the sum of all elements in one line of the matrix is equal to the mass represented by the corresponding component in the input vector ^.

Furthermore, there is no material loss or material produced in a sorter, i.e., the sum of all elements in the matrix are equal to the sum of mass represented by the input vector ^.

If material loss in a sorter is scope of the simulation, it can be modeled as an additional column in the output Matrix ^. Modelling of the sorter configuration The sorter capabilities can be modeled by a probability that a material is ejected by the sorter. Typically, the probability of ejection depends on the throughput of the sorter, the ejection configuration set by the user and the position of the valve block. The first valve block of a system is typically more reliable than the second or consequent valve blocks. Additionally internal parameters like air pressure, image processing settings, classifier settings, ejection settings have a high impact on the result. Whenever a particle is ejected, there might be particles overlapping or _{in the neighborhood, that are ejected by mistake. The false reject rate is} dependent on the machine capabilities and ejection settings, the amount of ejections that are happening, and the throughput of the machine. With high throughput, the number of overlapping objects increases and therefore the false ejection rate. The modelling of the configuration Matrix of the sorter is possible in many ways. A basic model is shown here, that can be used as a starting point, and can be refined after mapping with actual machine data. T_{he quality of the sorting is dependent on the parameters that are} independent of the throughput and parameters depending on the throughput. Parameters, independent of throughput are the following: ^_{Ejected material definition. The ejected material is defined by a vector} ^, that contains a probability of 1 for each material that should be ejected. If for example the first and third material should be ejected at a 1 v_{alve block ^ is defined as following: ^ = ^} 0 ^ 1 ^_{False positive: Errors in sensing, classification and image processing.} For each material that should be ejected, there is a probability for misclassification with other materials. This leads to unintended ejections, independent of how much of the target material is contained in the sorting stream. In an ideal case, there are not errors. Assuming 1 1_{% loss of material 2, leads to a modified ejection vector ^′ = ^} 0.01 ^ 1 ^_{False negative: The classification and processing in a system is not} perfect, and also ejection systems operates with an efficiency below 100%. These errors need to be anticipated in the ejection. Assuming 0.98 2_{% error would lead to the ejection vector ^′′ = ^} 0.01 ^ 0.98 Predicted S_{orted = Yes Sorted = No} _{Sorted = Yes True Positive False Negative} Actual _{Sorted = No False Positive True Negative} Parameters dependent on the throughput: A sorting system can be operated at different throughputs. With increasing throughput, the area that is covered with objects is increasing. In the beginning the area increases linearly, and converges with increasing _{throughput to 100%. This can be modeled by ^(^) = 1 − ^^^^ . The slope of} the curve is controlled by the parameter ^, which is dependent on the object _{types and the mass density of the material. This is illustrated in Figure 3.} _{^ A low surface coverage ^ leads no additional sorting errors. With} increasing surface coverage, overlapping objects and fully covered objects might occur with higher probability. This leads to a decreased d_{etection rate. The effect can be modelled by a function ^(^) that can} _{be expressed e.g. by ^(^) = 1 − ^} The relationship could be expressed as following for the example 0_{.98 ^(^))} _{above ^′′′ = ^} 0.01 ^ 0_{.98 ^(^))} _{^ Another effect of higher surface coverage is, that the distance between} neighboring objects is shrinking, which may lead to unintended false ejection induced by air turbulence or, if unwanted objects are covered by ejected objects. This effect leads to an increased false ejection, that is dependent on the surface coverage and on the amount of material t_{hat is ejected at the stage. Let define the ejection activity ^ by the} _{ejected material divided by the total material: ^ =}

. The false ejections can be expressed as following for the example _{0.98 ^(^))} _{above ^′′′′ = ^} 0.01^^ ^ 0_{.98 ^(^))} All elements mentioned above lead to the following sorting model for the ejection: ^_{= ^^^(^) + (1 − ^)^^^} _{^ With ^ is a binary vector, containing a 1 for the ejected materials, 0 for} non ejected materials. ^_{^ is a vector, containing the misclassification probability without having} the selected material present in the input (e.g. all elements 0.01) ^_{^ is a vector, with the sorting efficiency under optimal conditions (e.g.} 0.97) ^_{^ is the covered area of the conveyor belt} _{^ ^(^) expresses the reduced detection due to system load} _{^ ^ is the fraction of ejected material at valve block} The simulation is defined as following for a double valve block system. The first column contains the material that is not sorted out with the first valve block, the second column contains the material that is sorted by second valve block and the third column contains the remaining material that is not sorted by any valve block: ^_{(^, ^) = (^^^ (^ − ^^^)^^ ^−^^^ − (^ − ^^^)^^ )} The first column of the matrix contains the material that is not sorted, sort of the first valve block, second column uses the rest material as input for the second valve block, and so on. The last column contains the remaining material. F_{or multiple valve blocks ^^, … , ^^ , the simulation can be written as (for} simplicity ^_^ is defined as the i-th column of the matrix): Parameters that can be modified for different simulation results ^_{Material configuration (i.e., switch on and off materials, i.e. set that} material should be sorted out or that the material is to be allowed to pass) ^_{Material Classifier aggressiveness (Switch between two classifiers, e.g.} pure PET and contaminated PET) ^_{Image processing (filter neighborhood) parameters like erosion or} dilatation of classified image (Pixel processing, selectivity filter) ^_{Object processing vs. pixel processing} _{^ Ejection parameters (air pressure, ejection strategy, parameters)} _{^ Conveyor belt speed (Belt coverage vs. particle movement)} _{^ Mechanical catcher hood settings, splitter} _{^ Mechanical preprocessing (Screens, ballistic separators, eddy} currents) ^_{Mechanics or devices for even distribution of the material on the belt} (plates, disk spreader) Mapping the sorter configuration to real machines With the simulation model in place, a sorter behavior can be modelled. To adjust and use the model for simulation of real sorters, the model parameters need to be restricted to the real sorters capabilities. X-Ray Machines for example, can eject other material types than Polymer sorters or camera systems with deep learning. Therefore, the ejected material definition needs to be limited to the sorter capabilities. Each sorter has also a specific maximum for the false positive and false negative rate. Both rates are related to each other and can be configured by processing parameters. With more aggressive sorter settings, the false positive rate increases, but the false negative rate decreases. Simulation of sorting plants without feedback loop The simulation of one sorter can be used to define the simulation of a sequence of sorters. To cascade multiple sorters, the topology of the sorter needs to be defined. S_{ince sorters may have multiple outputs, and one output is} represented by one column in the matrix, we define a selection of one column as a function ^_^, which selects the i-th column of the matrix. A cascade with three sorters where ^_^ is connected to the first output of and ^_^ is connected to the second output of

can be written as follows:

I_{n short form we can write it}

_{→ ^^, as shown in Figure} 4a. With this notation, we can write down arbitrary complex tree structures without loops, in closed formulas for each output and calculate the simulated result. In plant layouts of real sorting plants, the input of multiple sorters might be combined on a conveyor belt. This is equivalent of adding (+) all corresponding material vectors as shown in Figure 4b. ^_{^(^^(^^(^^, ^^)) + ^^(^3(^^, ^^)) , ^^)} By combining both, arbitrary forests with multiple input material vectors can be simulated. Simulation of sorting plants with feedback loop For graphs with at least one feedback loop, a closed formula for a general simulation is not trivial since loops cannot be unrolled. As example of a loop the output of ^_^ is input to the preceding node ^_^, as is shown in Figure 5a. To calculate simulation models with loops on a computer system, a stepwise solution can be applied to simulate the loops, whereas the input _{vector is kept constant during all simulation steps. The simulation stops} either, when the output matrix is converging or if the system does not converge after a predefined maximum of steps. To execute the simulation, a buffer for material vector from the last simulation step is required as additional input for nodes with loops, as is shown in Figure 5b. The simulation of one step of the example is calculated as follows: First step: v1=v2=v3=0; Iterate until matrix converges or maximum amount of steps are reached

For arbitrary cyclic graphs, the simulation can be performed accordingly with the same buffer scheme each loop that occurs. Evaluation of a target function for a sorting plant Evaluation based on internal state During the simulation, several heuristics like air pressure consumption, power consumption could be collected. These measurements could be used for evaluation of the sorting plant. Evaluation based on Input and Output In the simulation model, a sorting plant gets multiple material input vectors as input and produces material matrices as output. The matrices can be split into the column vectors. Therefore, a function that qualifies the operation of a sorting plant, can be written as: ^_{(^, ^)} _{Whereas ^ = {^^, ^^, … , ^^} represents the incoming material vectors} _{and ^ = {^^, ^^, … ,}

_{the produced material output vectors of the plant.} One example is a monetary target function:

Here the function $() represents the monetary value of the input and output fractions. It is clear, that the commodity price changes over time, which will lead to different results at different times. Other target functions: ^_{Sort to target specification. If a sorting plant produces fractions with a} quality that is better than required, blending in low value materials could increase the overall value proposition of the plant ^_{Sort with minimum overall ^^^ footprint. This includes efficiency in the} sorting plant as well as the ^^_^ footprint associated to the products, like incineration or reuse. ^_{Sort with maximum earnings by maximizing earnings on products and} minimizing operating expenses like electricity or compressed air or maintenance. Optimizing a sorting plant For a sorting plant simulation model, the configuration, which consists of the configurations of all sorters, can be changed. Consequently, the output of the plant changes in a simulation run with a changed configuration and therefore the result of a target function will change as well. The configuration of the plant contains several parameters which can be changed to improve the result of the target function. For the optimization of the configuration between simulation runs, several strategies can be applied. Examples are exhaustive search, gradient decent, grid search or combination hereof. The optimization is performed _{based on the changes of the target function value. See further Figure 6.} Multiple target functions might be applied and combined. Simulation of plant topologies with flexibility The plant topology of a sorting plant is typically fix. However, it is possible to change parts of the plant topology with e.g., reversable conveyor belts. This is possible automatically within operation. Alternative plant topologies can be considered in the simulation. Plant topology optimization with defined set of sorters As a special type of optimization, the topology of a plant can be optimized as well. This can support the comparison and optimization of different plant layouts. The evaluation of the best topology can be done, by generating all possible topologies for a set of sorters, simulating, and selecting the best topology for the intended use case and target function. Plant topology simulation When only the input and output of the plant and the target function is given, we are flexible to select the number of sorters and the topology. The goal is to find the minimum amount of sorting machines and the best plant layout, that is solving the specification in the most efficient way. In this case, a set of sorters can be gradually increased and optimized as described in “Plant topology optimization with defined set of sorters” until the target function converges. Optimizing multiple sorting plants Optimizing multiple sorting plants can be performed in the same way as optimization one plant. In the case that a sorting plant is relying on the input of another plant, output material streams can act as input for the next _{plant as is illustrated in Figure 7a.} Constraints like transport or storage prices could be modeled as well. In addition Target functions might be combined if different targets are of _{interest as is illustrated in Figure 7b.} Deep learning system for blockage detection F_{ig. 8 schematically illustrates a sorting plant comprising a deep} _{learning system trained to detect blockages. Blockages can result in} _{downtime, reduced efficiency, or damage to the equipment, necessitating} _{immediate identification and resolution to restore normal operations. Blockage} _{can be complete (a material flow is completely impacted by blockage) or} _{partial (only a part of the material flow is impacted by blockage whereas the} _{remainder of the material flow can proceed according to expectations).} _{The sorting plant (also referred to as “plant” or “system”) comprises a} _{monitoring system. As a non-limiting example, the monitoring system} _{comprises a plurality of video cameras (surveillance cameras) arranged to} _{monitor a respective monitored location through which the flow of material is} _{provided. The sorting plan also comprises virtual cameras. Here, virtual} _{cameras refer to the sensor arrangement(s) of the sorting systems and} _{analysis systems, which may output sensor data or analysis data that can be} _{interpreted as image data.} The sorting plant comprises an AI system (also referred to as a deep learning system). The deep learning system comprises a neural network. The neural network comprises an input layer configured to receive said at least one image and/or said at least one video as input data. The neural network comprises one or more hidden layers, each hidden layer comprising a plurality of nodes configured to process the input data based on a set of weighted connections. The neural network comprises an output layer configured to provide the output based on how the one or more hidden layers processes the input data, which output is indicative of whether said at least one image and/or said at least video depicts a blockage. The sorting plant comprises an AI training system. The AI training system is configured to train the neural network of the AI system using a training dataset and a learning algorithm to perform a task of detecting whether said at least one image and/or video depicts a blockage. The learning algorithm may be any applicable learning algorithm. The AI training system may be configured to implement full automatic labeling, wherein a labelling of said at least one image and/or video comprises labelling said at least one image and/or video as depicting a blockage based on feedback from at least one blockage sensor (e.g., flow sensor) and/or at least one virtual blockage sensor (e.g., any sensor from which real-time data of blockage may be derived). The AI training system may be configured to receive manual labeling via operator input. An operator stationed in a monitoring room where said at least one image and/or said at least video is displayed on at least one monitor display may thus observe the at least monitor display to identify a blockage in _{said at least one image and/or said at least one video. The operator may thus} provide input whether said at least one image and/or said at least one video depicts a blockage. The AI training system may comprise a training database comprising _{training data provided by full-automatic labelling and/or manual labeling. The} AI training system may be configured to train the AI system using training data selected from the training database to train the AI system to comprise at least one classifier of blockage. The AI training system may advantageously incorporate a dementia process, meaning that the selection of training data used is selected based at least on age. This allows omitting irrelevant training data to be used during the training of the AI system. T_{he untrained AI system may be configured with at least one classifier} to form an AI interference system. If properly trained, the AI interference system is configured to detect blockage in at least one image and/or at least one video captured by means of any surveillance cameras or virtual cameras. O_{ne or more classifiers may be transferred to other AI systems of other} sorting plants. Such a transfer may be performed via wireless connection, wired connection, via internet, via a shared server connection, and/or by _{transfer using a memory storage capable unit (e.g., USB, a computer, or the} like). The sorting plant may comprise a notification system. The notification system may be configured to provide a blockage alert if the AI interference system detects a blockage in at least one image and/or at least one video. The notification system may be configured to provide said blockage alert in a monitoring room in which the monitoring system is arranged. Thus, an operator of the monitoring system may thus receive said blockage alert _{(which may e.g., be a visual alert or a sound alert). Following a blockage} alert, the operator may inspect the at least one image and/or the at least one video to verify if said at least one image and/or said at least one video depicts a blockage. In case of a false blockage alert, the operator may provide feedback to the AI training system. The AI training system may update the training database with the at least one image and/or at least one video, now relabeled as not depicting a blockage. Thus, the AI training system may improve training of the AI (interference) system over time based on operator feedback. The first target function, the second target function, and/or the third _{target function may depend on real-time data of blockage in said one or more} flow of materials. The real-time data of blockage may comprise said output from the AI (interference) system. Consequently, the system may be configured to take blockage and/or risk of blockage into account when updating a configuration of the first sorting system and/or a configuration of the second sorting system. When blockage is accounted for via the first target function, the second target function, and/or the third target function, the system may determine a configuration for the first sorting system and/or the second sorting system to reduce the risk of blockage and consequently allow for a higher uptime, thus improving the efficiency of the first sorting system and/or the second sorting system. The person skilled in the art realizes that the present inventive concept _{by no means is limited to the preferred embodiments described above. On the} contrary, many modifications and variations are possible within the scope of the appended claims. Variations to the disclosed embodiments can be understood and effected by the skilled person in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims.

Claims

CLAIMS 1. A system for determining configurations for a first sorting system and a second sorting system for sorting of objects based on simulation, the system comprising: a processing unit comprising circuitry configured to execute: a data receiving function configured to receive preliminary data specifying properties of a received plurality of objects; a configuration simulation function configured to iteratively simulate sorting the plurality of objects in the first sorting system based on a first configuration for the first sorting system, on a first simulation model for the first sorting system, and on the received preliminary data, and simulate sorting an output from the first sorting system in the second sorting system based on a second configuration for the second sorting system, on a second simulation model for the second sorting system, and on the output from the first sorting system according to simulated sorting of the plurality of objects in the first sorting system, wherein the first configuration and the second configuration are updated between each iteration based on one or more of a first target f_{unction for the first sorting system, a second target function for the} second sorting system, and a third target function common to the first and second sorting systems; and a determining function configured to receive from the configuration simulation function a determined first configuration a_{ccording to which the plurality of objects are to be sorted by the first} _{sorting system and a determined second configuration according to} _{which the output from the first sorting system should be sorted by the} second sorting system. 2_{. The system according to claim 1, wherein one or more of the first} _{target function, the second target function, and the third function depends on} real-time data regarding demand in one or more subsequent processes for objects having one or more of the properties of the received plurality of objects. _{3. The system according to claim 1 or 2, wherein the circuitry is further} configured to execute: a feedback function configured to receive feedback data from an actual _{physical sorting of objects based on the determined first configuration and the} determined second configuration, and a_{n adjustment function configured to adjust the first simulation model} _{and the second simulation model based on the feedback data.} _{4. The system according to claim 3, wherein the feedback data comprises} _{real-time data of blockage in a flow of material in the first sorting system, the} second sorting system, between the first sorting system and the second sorting system, and/or after the second sorting system. _{5. A system for determining a configuration for a sorting system for sorting of} _{objects based on simulation, the system comprising:} a processing unit comprising circuitry configured to execute: a data receiving function configured to receive preliminary data specifying properties of a received plurality of objects; a configuration simulation function configured to iteratively s_{imulate sorting the plurality of objects in the sorting system based on} _{a configuration for the sorting system, on a simulation model, and on} _{the received preliminary data, wherein the configuration is updated} between each iteration based on a target function; and a_{determining function configured to receive from the} _{configuration simulation function a determined configuration according} _{to which the plurality of objects are to be sorted by the sorting system,}

_{6. The system of claim 5, wherein the target function depends on real-time} data regarding demand in one or more subsequent processes for objects _{having one or more of the properties of the received plurality of objects.} _{7. The system according to any one of claims 1-6, further comprising:} a detector configured to detect properties of a subset of objects of the plurality of objects, thereby producing the preliminary data, wherein the data receiving function is further configured to receive the preliminary data from the detector. _{8. The system according to any one of claims 5-7, wherein the circuitry is} _{further configured to execute:} _{a feedback function configured to receive feedback data from an actual} _{physical sorting of objects based on the decided configuration, and} _{an adjustment function configured to adjust the simulation model based on} the feedback data. 9. The system according to claim 8, wherein the feedback data comprises _{real-time data of blockage in a flow of material in the first sorting system,} _{and/or after the first sorting system.} 10. The system according to any one of claims 1-9, wherein the data receiving function is further configured to receive preliminary data for a further received plurality of objects. _{11. The system according to any one of claims 1-10, wherein the first target} function, the second target function, and/or the third target function depends _{on real-time data of blockage in a flow of material in the first sorting system,} the second sorting system, between the first sorting system and the second sorting system, and/or after the second sorting system.

_{12. A computer implemented method for determining configurations for a} first sorting system and a second sorting system for sorting of objects based on simulation, the method comprising: receiving preliminary data specifying properties of a received plurality of objects; i_{teratively simulating sorting the plurality of objects in a first sorting} system based on a first configuration for the first sorting system, on a first simulation model for the first sorting system, and on the received preliminary data, and simulating sorting an output from the first sorting system in a second sorting system based on a second configuration for the second sorting system, on a second simulation model for the second sorting system, and on the output from the first sorting system according to simulated sorting of the plurality of objects in the first sorting system, wherein the first configuration and the second configuration are updated between each iteration based on one or more of a first target function for the first sorting system, a second target function for the second sorting system, and a third target function common to the first and second sorting systems; and receiving from the iterative simulation a determined first configuration according to which the plurality of objects are to be sorted by the first sorting system and a determined second configuration according to which the output from the first sorting system should be sorted by the second sorting system 13. The method according to claim 12, wherein one or more of the first target function, the second target function, and the third target function depends on real-time data regarding demand in one or more subsequent processes for objects having one or more of the properties of the received plurality of objects. _{14. The method according to claim 12 or 13, further comprising:} receiving feedback data from an actual physical sorting of objects based on the determined first configuration and the determined second configuration; and _{adjusting the first simulation model and the second simulation model} based on the feedback data. 15. The method according to claim 14, wherein the feedback data comprises _{real-time data of blockage in a flow of material in the first sorting system, the} second sorting system, between the first sorting system and the second _{sorting system, and/or after the second sorting system.} _{16. A computer implemented method for determining a configuration for a} _{sorting system for sorting of objects based on simulation, the method} comprising: receiving preliminary data specifying properties of a received plurality of objects; iteratively simulating sorting the plurality of objects in a sorting system _{based on a configuration for the sorting system, on a simulation model, and} _{on the received preliminary data, wherein the configuration is updated} between each iteration based on a target function; and _{receiving from the iterative simulation a determined configuration according to} _{which the plurality of objects are to be sorted by the sorting system.} _{17. The method according to any one of claims 12-16, further comprising} _{detecting properties of a subset of objects of the plurality of objects by} means of a detector, thereby producing the preliminary data, w_{herein the preliminary data are received from the detector.} 18. A non-transitory computer-readable storage medium having stored thereon program code portions for implementing the method according to any _{one of claims 12-17 when executed on a device having processing} capabilities.