DE102021110536A1 - Method for monitoring a conveyor system with conveyor elements, computer program and electronically readable data carrier - Google Patents

Abstract

The invention relates to a method for monitoring a conveyor system (10) with conveyor elements (12) for receiving a component (14) which is moved by means of the respective conveyor element (12) along a conveyor section (16) to assembly points (18). the steps:- detecting at least one respective parameter value for each of the conveying elements (12) and a respective location at defined points in time within a time interval; (S1)- filing of the parameter values to form a data record; (S2)- forming a respective data group which includes the parameter values and locations for the respective conveying element (12); (S3) - filtering the data group according to at least one known deviation value; (S4)- merging the data groups to form the data set, taking into account the at least one deviation value; (S5) - training an anomaly model depending on the detected parameter values and locations; (S6)- detecting an anomaly of one of the parameter values using the anomaly model; (S7) and calculation of an anomaly frequency. (S8) The invention also relates to a computer program and a data carrier.

Description

The invention relates to a method for monitoring a conveyor system with conveyor elements according to patent claim 1. The invention also relates to a computer program according to patent claim 9 and an electronically readable data carrier according to patent claim 10.

Conveyor systems are often used in assembly or production. A conveyor system with a plurality of conveyor elements, which can each be designed as a vehicle carrier, is used, for example, in vehicle assembly, in particular in motor vehicle assembly. Thus, the conveyor system is a conveyor system that can include several vehicle carriers, which generally move in a circle on a fixed path. Thus, repetitive tasks are performed for assembly with the conveyor system. A failure of one of the vehicle carriers usually leads to a standstill of the entire system and is therefore associated with high costs.

This is how it shows EP 1 800 196 B1 a device for predicting the mean time span between two failures of a technical system.

A disadvantage of the prior art is that, for example, an actual, current state of the conveyor system is unknown and therefore there is no early warning system for irregularities. As a result, there is a possibility of unforeseen failures due to a technical fault. This is associated with high costs, for example due to standstills in assembly.

It is therefore the object of the present invention to provide a method, a computer program and a data carrier by means of which failure of a conveyor element of a conveyor system can be prevented in a particularly advantageous manner.

According to the invention, this object is achieved by the subject matter of the independent patent claims. Advantageous designs are the subject matter of the independent patent claims and the description.

The method according to the invention serves to monitor a conveyor system with conveyor elements. In this case, the respective conveying element is designed to receive a respective component, for example a body part in motor vehicle production. The component is moved by means of the respective conveyor element along a conveyor section of the conveyor system, for example a circular path, for assembly steps or production steps to assembly points which are arranged along the conveyor section.

• In einem ersten Schritt erfolgt ein Erfassen zumindest eines jeweiligen Parameterwerts für jedes der Förderelemente. Der wenigstens eine jeweilige Parameterwert charakterisiert beziehungsweise beschreibt einen Zustand des jeweiligen Förderelements. Ferner wird ein jeweiliger Standort für das jeweilige Förderelement erfasst. Dabei erfolgt das Erfassen des Parameterwerts und des Standorts zu definierten Zeitpunkten innerhalb eines Zeitintervalls.

In order to prevent a failure of the conveyor system in a particularly advantageous manner, the procedure comprises several steps:

• In a first step, at least one respective parameter value is recorded for each of the conveying elements. The at least one respective parameter value characterizes or describes a state of the respective conveying element. Furthermore, a respective location for the respective conveying element is recorded. The parameter value and the location are recorded at defined points in time within a time interval.

In a second step, the recorded parameter values are stored with a respective identification value, which is, for example, an assignment ID and/or identification number, and the locations are stored in a memory device to form a data record.

In a third step, a respective data group is formed, which includes the parameter values and locations for one of the identification numbers and thus for the respective conveying element.

In a fourth step, the data group is filtered according to at least one known deviation value.

In a fifth step, the data points are combined to form the data set, taking into account the at least one deviation value.

In a sixth step, an anomaly model for the conveyor elements to be monitored is trained using the recorded parameter values, in particular as a function of the respective location.

In a seventh step, an anomaly of one of the parameter values is detected using the anomaly model.

Finally, in an eighth step, the frequency of anomalies per conveying element and per defined time interval is calculated.

The conveying elements are advantageously vehicle carriers during motor vehicle assembly. In other words, the method uses a continuous comparison of newly recorded parameter values with parameter values that have already been stored, which characterize a state of the respective conveying element and thus, for example, detect built-in components of the conveying elements designed as vehicle carriers, possible to detect anomalies under the vehicle carriers at an early stage. As a result, for example, preventive measures such as repairs can be initiated.

The anomaly is detected or recognized by the anomaly model, which is recorded, for example, with the aid of machine learning in relation to selected parameter values, such as, for example, an active current of a motor of a conveying element. This is followed by, for example, a static evaluation of the occurrence of anomalies and a comparison of the frequency of anomalies among all conveyor elements. This comparison makes it possible, since the conveyor system includes a large number of technically identical, in particular identical, conveyor elements to integrate a large number of parameter values into the data record, as a result of which a large number of comparative data is available for anomaly detection and for training.

In other words, conspicuous conveying elements should be found within the totality of the conveying elements by the method. The conveying elements can be identified by the unique identification value, so each conveying element is unique in the set of conveying elements. In order to identify a conspicuous conveying element, its at least one parameter value is compared with other detected parameter values.

For the method, or also while the method is being used, it is decided which of the at least one parameter values or which type of parameter value is to be used. This is followed by the aggregation or recording of the parameter values. Recording at defined points in time within a time interval means that a parameter value is recorded every ten seconds within an hour, for example. The respective location of the conveyor element in the conveyor system is also recorded during the detection. The short time interval, ie that which lies between two defined points in time, can be selected in such a way that the parameter value is also always recorded at a defined location.

For example, the parameter values are also dependent on the location or the position of the respective conveyor element in the conveyor system. The data group thus includes all parameter values of one of the conveying elements at different locations, depending on the point in time during the parameter value acquisition that the conveying element is in which position or which location. Thus, a possible example data group or a possible example data set as input for training the anomaly model can comprise n samples by m parameter values. Deviations from a mean parameter value can now be detected in the anomaly model using, for example, an n by m matrix, as a result of which an anomaly can be recognized and an anomaly frequency can be calculated. For example, deviations from Gaussian distributions in standard deviations can be determined in order to uncover an anomaly.

The location can be permanently specified as a respective block point, with the conveying section of the conveyor system thus being divided into discrete block points at which, for example, a parameter value is recorded. Such a block point can also coincide with an assemblage point, for example. However, this depends on the position of the assembly point or the block point on the conveyor line.

The method according to the invention thus results in a number of advantages. On the one hand, targeted prevention of system failures is possible. Furthermore, it is possible to generate very meaningful data due to the large number of available comparison elements in the form of parameter values and/or conveying elements. Furthermore, the method can be applied to different systems, regardless of the number of conveyor elements and/or the type of conveyor system, for example, this can be a suspension system, a pusher plate system or some other type.

Furthermore, the method can be carried out particularly favorably, for example, since existing sensors can be used to record the parameter values. In addition or as an alternative, parameters from virtually any built-in components can be used, which can be read out in the respective component, for example. In this way, these parameter values can also be recorded for another task. Furthermore, a production-independent use is possible since parameter data or parameter values of different components of the conveying elements can be used.

Furthermore, advantageous use is also possible when commissioning new systems, since this enables a faster reduction in the errors that occur.

In an advantageous embodiment of the invention, the parameter values are stored with each associated identification value and locations into the storage device over a network. As a result, the conveyor system and the storage device can be designed separately from one another. In other words, the sensors designed to record the parameter values on the conveyor system or on the conveyor elements are connected to the storage device via a network, for example by means of an IP protocol such as the Internet, with the storage device thus being used as a cloud storage device or cloud Computer can be designed. This results in the advantage that the method can be used particularly flexibly. In addition, there is the advantage that only one storage device has to be kept available, for example for monitoring several conveyor systems.

In a further advantageous embodiment of the invention, the anomaly model is trained using machine learning or is carried out using machine learning. In this case, the machine learning is carried out, for example, by a learning, in particular self-learning, algorithm and/or by a neural network. Thereby the algorithm can be executed by the network. This results in the advantage that an anomaly model can be created in a particularly advantageous manner, which can, for example, react more flexibly to further parameter values in comparison to a model created using conventional algorithms.

In a further advantageous embodiment of the invention, an active current, a reading quality, a temperature, a message number, a speed and/or a position are recorded as the at least one parameter value. The active current is, for example, the current of a chassis, hoist and/or slewing gear of the conveying element and this is advantageously aggregated, for example, at a ten-second interval, with a maximum mean value and a variance being recorded or calculated. The reading quality is advantageously a barcode scanner reading quality, as a result of which identification numbers, for example, can be detected in a particularly simple manner. Furthermore, various vehicle data of the vehicle carrier or the respective conveyor element can be recorded. Furthermore, encoder differences can be determined or recorded. The speeds are rotational speeds, for example. Position values such as a lifting height, angle of rotation or the like can be recorded as the position. In addition, it is conceivable that, for example, software versions of control units of the respective conveying element are recorded. In addition, operating hours can be recorded, which provide an indication of the wear and tear of the corresponding parts. In addition, generation data of a safety program can be recorded. Additionally or alternatively, for example, a route or a distance counter in the form of a speedometer, for example, can be recorded. In addition, average cycle times can be recorded. The examples given here are by no means exhaustive. Advantageously, a plurality of the parameter values mentioned are recorded for each of the conveyor elements, as a result of which a standstill of the conveyor system can be prevented in a particularly advantageous manner.

Advantageously, with the method according to the invention, a conveyor hanger, a pallet truck and/or a pusher plate is monitored as the respective conveying element. In addition, they can be elements of a conveyor belt or the like. This results in the advantage that the method can be used for a large number of different conveying elements and is therefore particularly flexible.

In a further advantageous embodiment of the invention, in a further step of the method, the respective anomaly frequency for a respective parameter value of the respective conveying element is visualized, for example as a thermal image card or heat map, on an output device such as a screen. By outputting the frequency of anomalies on the output device, maintenance or monitoring of the system that may be necessary can be carried out in a particularly advantageous manner.

In a further advantageous embodiment of the invention, a repair value, a travel value and/or a location value is recorded as the at least one known deviation value. In other words, data that would disrupt or worsen the modeling of the anomaly model is removed from the respective data group for specific identification values, ie for specific conveying elements.

This can be data that was collected at a specific location on the conveyor line. This can be, for example, the location at which a component is picked up by the conveying element, which can lead to process-related anomalies that should therefore not be evaluated as anomalies in terms of failsafety. In addition, this can affect data that was collected shortly after a process disruption, during the start-up phase process-related anomalies also arise. Furthermore, the data can be discarded for periods in which Ver search and/or repairs take place. The anomaly model can be determined particularly precisely by sorting out the deviation values mentioned, ie the repair values for repairs or tests taking place, the start-up values for start-up phases and the location values and/or time values. This results in the advantage that failures of the conveyor system can be avoided in a particularly advantageous manner.

In a further advantageous embodiment of the invention, a repair instruction is issued when the calculated anomaly frequency reaches a threshold value. In other words, based on the anomaly frequency, from which, for example, an additional anomaly score can be formed or the anomaly frequency represents an anomaly score, an instruction for a repair is issued or a repair is initiated if it becomes apparent from the calculated anomaly model that a failure of a component of the Conveying element and thus the conveying element, for example, is imminent. This results in the advantage that the method contributes particularly advantageously to avoiding failures of the conveyor system.

Another aspect of the invention includes a computer program. The computer program can be loaded into a memory of a control device of a conveyor system, for example, and includes program means to carry out the steps of the method when the program is executed in the control device of the conveyor system.

Another aspect of the invention relates to an electronically readable data carrier. The electronically readable data carrier includes electronically readable control information stored thereon, which includes at least one computer program as just presented or is designed such that it can execute a method presented here when using the data carrier in a control device of a conveyor system.

The advantages of the method are to be regarded as advantages of the computer program and the electronically readable data carrier and vice versa.

Further features of the invention result from the claims, the figures and the description of the figures. The features and combinations of features mentioned above in the description and the features and combinations of features mentioned below in the description of the figures and/or shown alone in the figures can be used not only in the combination specified in each case, but also in other combinations or on their own.

• 1 in einer schematischen Draufsicht eine Förderanlage mit Förderelementen zur jeweiligen Aufnahme eines Bauteils und mit einer Förderstrecke;
• 2 ein schematisches Ablaufdiagramm eines Verfahrens zur Überwachung der Förderanlage gemäß 1;
• 3 in einer schematischen Perspektivansicht ein Förderelement der Förderanlage gemäß 1;
• 4 in einem Diagramm durch das Verfahren erfasste Parameterwerte in Abhängigkeit der Zeit;
• 5 in einem Diagramm eine durch das Verfahren ermittelte Anomaliehäufigkeit; und
• 6 ein weiteres schematisches Ablaufdiagramm einer alternativen Ausführungsform des Verfahrens.

The invention will now be explained in more detail using a preferred exemplary embodiment and with reference to the drawings. It shows:

• 1 in a schematic plan view, a conveyor system with conveyor elements for receiving a component and with a conveyor section;
• 2 a schematic flow chart of a method for monitoring the conveyor system according to FIG 1 ;
• 3 in a schematic perspective view a conveying element of the conveyor system according to 1 ;
• 4 parameter values recorded by the method in a diagram as a function of time;
• 5 in a diagram, an anomaly frequency determined by the method; and
• 6 another schematic flow chart of an alternative embodiment of the method.

1 shows a schematic top view of a conveyor system 10 with a large number of conveyor elements 12. The respective conveyor element 12, which, for example, serves as an in 3 conveyor hanger shown is designed to receive a component 14. This component 14 can be designed, for example, as part of a motor vehicle body, which is moved by means of the respective conveyor element 12 along a conveyor path 16 for assembly steps, for example for assembling the motor vehicle, to assembly points 18.

It is desirable for the conveyor system 10 to be able to be operated as fail-safe as possible, in order in particular to save particularly high costs during assembly or during the production of motor vehicles, for example. A procedure for this is presented below.

• In Schritt S1 erfolgt ein Erfassen zumindest eines jeweiligen Parameterwerts für jedes der Förderelemente 12, welcher einen Zustand des jeweiligen Förderelements 12 charakterisiert beziehungsweise beschreibt. Ferner erfolgt das Erfassen eines jeweiligen Standorts für das jeweilige Förderelement 12 zum Zeitpunkt des Erfassens. Das Erfassen erfolgt dabei zu definierten Zeitpunkten innerhalb eines Zeitintervalls. Dabei ist der Abstand zwischen zwei Zeitpunkten innerhalb des Zeitintervalls ein weiteres und kürzeres Zeitintervall.

So shows 2 in a schematic flow chart the steps S1 to S8 of the method for monitoring the conveyor system 10. The steps are:

• In step S1, at least one respective parameter value is detected for each of the conveying elements 12, which characterizes or describes a state of the respective conveying element 12. Furthermore, a respective location for the respective conveying element 12 is detected at the time of detection. The recording takes place at defined points in time within a time interval. The distance between two points in time within the time interval is a further and shorter time interval.

In step S2, the parameter values are stored with a respective identification number or an identification value, which makes the respective conveying element 12 identifiable, as well as the locations for forming a data set in a storage device.

In step S3, a respective data group is formed, which includes the parameter values and locations for one of the identification numbers and thus for the respective conveying element 12.

In a step S4, the respective data group is filtered according to at least one known deviation value.

In step S5, the respective data group is combined to form the data set, taking into account the at least one respective deviation value.

In step S6, an anomaly model for the conveying element 12 to be monitored is trained as a function of the detected parameter values and/or as a function of the location.

In step S7, an anomaly of one of the parameter values is detected using the anomaly model.

In step S8, the anomaly frequency is calculated per conveyor element 12 and per defined time interval.

In step S1, in which the parameter values characterizing the respective conveying elements 12 are related, the following can be carried out, for example: For example, the at least one defined parameter value can be obtained repeatedly every ten seconds over a period of one hour for each of the conveying elements 12 and are always the same stored or filed in the memory device for step S2. The storage device is advantageously connected to the conveyor system 10 via a network and can thus be designed, for example, as a storage device in a cloud environment.

After being stored in the storage device, the data are available for further analysis in the further steps of the method. Furthermore, the recording is carried out, for example, in such a way that for one of the locations, which can describe a partial area of the conveyor line 16 as a block point, all parameter values of the conveyor elements 12 are recorded in a time window, with the values to be recorded being recorded every ten seconds in an hour, for example time intervals results in a set of parameter values for the data set. The set of parameter values, which is formed from the sum of the parameter values recorded for the respective conveying element 12 at the respective point in time, can be large enough (for example for good statistics) to determine a particularly advantageous anomaly model in step S6.

The data record from the parameter values, identification values or numbers and locations stored in the storage device thus contains the information which is to be analyzed or monitored in order to increase the failsafety of the conveyor system 10 . For example, it contains current, temperature or barcode reading quality, as well as the unique ID of the individual conveyor elements 12 . Furthermore, the data record contains the respective location parameters and thus the location, so that a location-dependent anomaly model can be formed, which can be used particularly advantageously for monitoring the conveyor system 10, as a result of which the reliability can be particularly high.

In step S3, the data group is formed by dividing the stored data, which contains parameter values and locations, according to conveying elements 12 using the identification values. In this step S3, the data set is grouped according to the individual identification values of the conveying elements 12. Then, for each identification value, certain data sets are removed, if conspicuous, which would disturb or worsen the modeling.

This abnormality is determined in step S4 by filtering the respective data groups. The data sets that are removed are data that were collected at certain locations on the belt area, this can be, for example, the location at which a (e) vehicle (body) is recorded as component 14 and there is a process-related anomaly comes. Here one can say that the at least one deviation value is a location value.

Furthermore, the data collected shortly after a process disruption and generated during a start-up phase can also be process-related anomalies. In this case, the deviation value can thus include an approach value. Furthermore, time periods in which tests or repairs take place are also removed from the data, as a result of which the deviation value can be a so-called repair value.

In step S5, a data set is formed that filters all data groups with their corrected values, values meaning the parameter values and/or location values, which in particular can be correlated with one another.

An anomaly model, in particular a standard-dependent one, is then trained. In principle, the anomaly model is trained according to the following logic: In the selected time period or time interval, a separate anomaly model is trained for each location area. This takes into account the fact that the conveying elements 12 perform different movement sequences depending on their location. The individual anomaly models are trained with one or more parameters as input and model the normal state at the respective location area. The individual anomaly models of the respective locations (block locations) can be combined into a common, global anomaly model.

Existing methods and algorithms can be used to train the (respective and/or global) anomaly model. In this case, the analysis method is advantageously specially tailored to the respective parameter value and depends on its detection and, for example, the possible error (values).

In step S7, the anomaly detection is used, and the anomaly frequency per hanger or conveyor element 12 is then calculated in step 8. The trained anomaly model is applied to the data set and a so-called anomaly score is calculated, which in principle ranges from 0 to 1 or from 0 to 100 percent goes, calculated. The anomaly score expresses the ratio of the number of anomalies in the parameter values per conveying element 12 to the total number of parameter values determined for a period of time for each of the conveying elements 12 . A higher anomaly score can thus signal a greater deviation from the normal state of a respective conveying element 12 .

Furthermore, in an additional, in the 2 not shown step of the method, a visualization of the respective anomaly frequency, in particular as a heat map or the like, on an output device. The heat map is a clear way of visualizing the respective parameter values. In addition, in an alternative or additional further step, a repair instruction can be issued at a threshold value of the calculated anomaly frequency or the anomaly score. With the help of this instruction, the conveying element 12 that has the anomaly can be removed, for example, from the conveying system 10 and repaired, while the conveying system 10 can continue to be operated during the repair. This avoids the conveying element 12 stopping, for example, in the middle of the conveying section 16 at a point from which it cannot be easily removed.

As 3 shows, the respective conveyor element 12 can be designed as a conveyor hanger and the component 14 is, for example, a vehicle body. In addition or as an alternative, the conveying element 12 can also be designed, for example, as a pallet truck and/or as a push plate or the like. The conveyor element 12 can have a first motor 20 and a second motor 22, for example.

4 FIG. 1 shows a current recorded as a respective parameter value of the respective motor 20, 22 in the form of diagrams, which is plotted against time. The upper diagram shows in 4 Data for the engine 20 and the bottom graph Data for the engine 22. It can be seen that there is an anomaly for the engine 20 at the rear of the graph, marked by the dashed box.

In general, a conveyor system 10 presented here supplies a large number of data for a number of types of parameter values, which are (can be) stored in the storage device, which is designed as cloud storage, for example. For example, the available parameter values can include plant flows for the chassis, hoist or slewing gear of the respective conveying element. Furthermore, parameter values can include: barcode reading scanner reading qualities, temperatures, message numbers, various vehicle data, encoder differences, speeds/rotational speeds, position values, lifting heights, angle of rotation, software version of a control unit of the conveyor element 12, VU error codes, operating hours, generated safety programs, odometer and/or average cycle times .

In summary, the method can be carried out with the parameter values presented here in such a way that, for example, all conveyor hanger data, i.e. conveyor element data, is retrieved from the cloud memory, then the data is divided according to vehicle carrier, i.e. the respective conveyor elements 12. Filtering also takes place according to location, time and, if necessary, process disruptions and other parameters. The data is then combined and an anomaly model is trained for the parameters to be monitored, depending on the location. The anomaly detection is then applied and the anomaly is calculated frequency per hanger and per defined time interval.

5 shows a schematic diagram of the before-after comparison of an anomaly score for a vehicle carrier, with the vertical line in the rear third of the diagram marking the point in time from which a repair of a corresponding component that caused an anomaly was carried out on the corresponding conveying element 12 .

For example, conspicuous deviations from the normal values for one of the vehicle carriers 12 can be detected in a heat map (not shown). In the subsequent analysis of the vehicle carrier, a worn brake can be determined as the cause, for example. Without the method of monitoring, the worn brake could not be replaced preventively, but would have resulted in an unforeseeable failure of the entire conveyor system 10.

6 shows an embodiment of the method in a systematic overview. while showing 6 furthermore, represented by the curly brackets 22, an IT infrastructure possibly behind the method, with which the at least one respective parameter value is recorded. A cloud computer 26 can be located in the IT architecture 24 .

The respective conveying element 12, which in the left area of the 6 are shown may have a programmable logic controller 28 (PLC). In this way, parameters (values) are transmitted to the respective PLC 28 . Data pre-processing can take place in the PLC 28 . Data processing can take place at a central PLC of the conveyor system 10, represented by the three left-hand arrows in FIG 6 . This central PLC or head PLC 30 is connected to the cloud computer 26 . The head PLC 30 is advantageously designed as a control device. The data is thus processed in the IT architecture 24. Visualization and alarming or alerting can take place there. For example, a gateway 32 can be used to connect to an alarm system. The alarm system can be designed, for example, as a process control system and/or fault reporting system.

The method presented makes it possible to monitor the conveyor system during assembly in a particularly advantageous manner.

reference list

10

conveyor system

12

conveying element

14

component

16

conveyor line

18

assemblage point

20

engine

22

engine

24

IT architecture

26

cloud calculator

28

Programmable logic controller (PLC)

30

head PLC

32

Gateway

34

alarm system

S1

First step

S2

Second step

S3

Third step

S4

fourth step

S5

Fifth step

S6

Sixth step

S7

Seventh step

S8

Eighth step

QUOTES INCLUDED IN DESCRIPTION

This list of documents cited by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent Literature Cited

• EP 1800196 B1 [0003]

Claims (10)

Method for monitoring a conveyor system (10) with conveyor elements (12) for receiving a component (14), which is moved by means of the respective conveyor element (12) along a conveyor path (16) for assembly steps to assembly points (18), with the steps : - detecting at least one respective parameter value for each of the conveying elements (12), which characterizes a state of the respective conveying element (12), and a respective location for the respective conveying element (12) at defined points in time within a time interval; (S1) - Storing the parameter values, with a respective identification value and the locations in a memory device, to form a data set; (S2) - Forming a respective data group, which includes the parameter values and locations for each of the identification values and thus for the respective conveying element (12); (S3) - filtering the data group according to at least one known deviation value; (S4) - merging the data groups into the data record, taking into account the at least one deviation value; (S5) - training an anomaly model for the conveyor elements (12) to be monitored as a function of the detected parameter values and locations; (S6) - detecting an anomaly of one of the parameter values using the anomaly model; (S7) and - Calculation of an anomaly frequency per conveying element (12) and per defined time interval. (S8) procedure after claim 1 , characterized in that the depositing in the storage device takes place via a network and thus the conveyor system (10) and storage device are designed separately. procedure after claim 1 or 2 , characterized in that the anomaly model is trained by means of machine learning. Method according to one of the preceding claims, characterized in that an active current, a reading quality, a temperature, a message number, a speed and/or a position is recorded as the at least one parameter value. Method according to one of the preceding claims, characterized in that a conveyor hanger, a pallet truck and/or a pusher plate is monitored as the respective conveyor element (12). Method according to one of the preceding claims, characterized in that in a further step of the method, the respective anomaly frequency is visualized, in particular as a thermal image map, on an output device. Method according to one of the preceding claims, characterized in that a repair value, a start-up value and/or a location value is recorded as the at least one known deviation value. Method according to one of the preceding claims, characterized in that a repair instruction is issued if the calculated anomaly frequency reaches a threshold value. Computer program, which can be loaded directly into a memory of a control device of a conveyor system (10), with program means to carry out the steps of the method according to one of Claims 1 until 8th to be executed when the program is executed in the control device of the conveyor system (10). Electronically readable data carrier with electronically readable control information stored thereon, which at least one computer program claim 9 include and are designed such that they use the data carrier in a control device of a conveyor system (10) a method according to one of Claims 1 until 8th execute.

Images