Method and device for monitoring a milling process

Abstract

A method for monitoring a milling process of a circuit board, the method comprising the following steps: (a) detecting (S1) a rotational speed of a milling head (2) of a milling machine (1) and at least one other operating parameter of the milling machine (1) during the milling process; wherein the other operating parameter is a power supply current for operating the milling machine, and (b) evaluating (S2) the detected rotational speed and the detected operating parameter by means of a trained learning algorithm to identify anomalies during the milling process.

Description

Technical Field

[0001] The invention relates to a method and a device for monitoring a milling process, preferably with the aid of artificial intelligence. Background Art

[0002] Patent application US 2016 / 091393 A1 describes a method or system for prediction and planning in a production plant. In order to detect tool wear, the performance of the spindle is also monitored. In order to compensate for tool wear, the performance of the spindle is increased and thus the rotation speed of the spindle is increased. In this case, the long-term trend of the spindle performance is inferred to predict the remaining service life.

[0003] Patent application US 2007 / 088550 A1 describes a method for predictive maintenance of a machine. For this purpose, data related to the operation of the machine, ie, for example, vibrations, speed or current, are detected. So-called signatures are generated from the data in order to predict faults.

[0004] Patent application US 2004 / 179915 A1 describes a machine tool. In order to analyze the behavior of the machine tool, a signal known as a signature variable is detected there and used as an input variable of an artificial neural network.

[0005] US Pat. No. 8,781,982 B1 describes a method and a device for determining the remaining service life. For this purpose, operating data are fed into an artificial neural network. The operating data can be vibration data, acoustic data and acceleration data.

[0006] Milling is usually a cutting method for manufacturing workpieces with geometrically determined shapes. During milling, the milling tool or milling head rotates at high speed around its own axis, while the milling tool itself passes over the contour to be manufactured at the workpiece, or the workpiece moves accordingly relative to the milling head to remove material. Milling machines can be used to mill different workpieces, in particular, also for milling circuit boards (Print Circuit Boards). For example, the dividing line between the conductor line surfaces of the circuit board can be established by milling. During the milling process, a large amount of dust can be formed, so that the milling spindle can get stuck after a certain running time. In conventional milling machines, such faults or failures are detected only after the occurrence of corresponding faults, such as blocking the milling spindle. The failure of the milling machine during the milling process can lead to a significant delay in the processing technology of the object to be manufactured. In addition, the workpiece that has not been manufactured in the manufacturing step is often presented as a defective product and cannot be further processed. During the running manufacturing process, the stoppage of the milling machine requires maintenance personnel or service technicians to be put into operation without delay to implement maintenance measures or repair measures. In many cases, such service technicians do not arrive immediately but must come on request. This can lead to the temporary stoppage of entire production lines within a factory. Summary of the invention

[0007] Therefore, the object of the present invention is to provide a method and a device for monitoring a milling process, which method and a device predict a possible failure of the milling machine concerned at an early stage, so that countermeasures can be taken in good time.

[0008] According to the invention, this object is achieved according to a first aspect of the invention by a method for monitoring a milling process having the features specified in the invention.

[0009] The invention therefore provides a method for monitoring a milling process, the method comprising the following steps:

[0010] detecting a rotational speed of a milling head of the milling machine and at least one further operating parameter of the milling machine during the milling process; and

[0011] During the milling process, the detected rotational speed and the detected operating parameters are evaluated by a trained learning algorithm for the purpose of identifying anomalies.

[0012] By means of the method according to the invention, the productivity of the milling machine can be significantly increased. Undesired downtimes of the milling machine are prevented. Thus, failures of the milling machine during the milling process can be avoided. Furthermore, the quality of the products or intermediate products produced in the milling process can be improved. Monitoring, in particular technical testing, of the function of the machine tool, in particular the milling machine, is also necessary in order to detect states that are not conducive to safety. Furthermore, monitoring can be used to assess the remaining service life of the milling machine. Thus, by means of such monitoring, unplanned downtimes can be avoided and a longer operation of the milling machine can be achieved.

[0013] In one possible embodiment of the method according to the invention, the further operating parameter comprises the supply current for operating the milling machine.

[0014] This operating parameter offers the advantage that it can be detected relatively easily and with little technical effort.

[0015] In another possible embodiment of the method according to the invention, after the anomaly has been detected, a message is generated for an operator of the milling machine and displayed in real time on a display unit of the milling machine.

[0016] In another possible embodiment of the method according to the invention, maintenance measures are automatically initiated after the anomaly has been detected.

[0017] This further minimizes operating downtimes.

[0018] In a further possible embodiment of the method according to the invention, after the detection of an anomaly during the milling process, a failure probability and / or a predicted downtime period of the milling machine is calculated.

[0019] This makes it possible to optimize the production planning of the manufacturing process.

[0020] In a further possible embodiment of the method according to the invention, after the anomaly has been detected during the milling process, at least one possible cause is determined for the occurrence of the anomaly.

[0021] This allows maintenance measures or repair measures to be carried out in a targeted manner, so that downtimes or downtimes of the milling machine are reduced.

[0022] In a further possible embodiment of the method according to the invention, the trained learning algorithm has a trained neural network.

[0023] In a further possible embodiment of the method according to the invention, the learning algorithm is trained in a training phase before the start of the production milling process based on historical training data regarding the rotational speed and at least one further operating parameter of the milling machine.

[0024] In a further possible embodiment of the method according to the invention, the learning algorithm is downloaded as an application from a database via a network into a computing unit of the milling machine and is trained in a training phase for a production milling process.

[0025] In a further possible embodiment of the method according to the invention, the learning algorithm is selected as a function of the type of milling machine and / or the type of workpiece to be milled in the milling process.

[0026] In a further possible embodiment of the method according to the invention, the workpiece to be milled comprises a circuit board.

[0027] In another possible embodiment of the method according to the invention, the rotational speed and at least one further operating parameter of the milling head of the milling machine are detected and the corresponding data are supplied at a data rate of approximately 1 / second to a trained learning algorithm executed in a computing unit of the milling machine.

[0028] In a further possible embodiment of the method according to the invention, anomalies are detected in real time during the production milling process by means of a trained learning algorithm.

[0029] According to a further aspect, the invention provides a milling machine for milling a workpiece having the features specified in the invention.

[0030] The invention thus provides a milling machine for milling a workpiece, the milling machine having

[0031] - a rotatable milling head for machining a workpiece in a milling process,

[0032] - an electric motor for driving the milling head, and

[0033] An artificial intelligence module is provided which, during the milling process, detects anomalies in the milling process based on the detected rotational speed of the milling head and at least one further operating parameter of the milling machine, in particular the supply current of the electric motor.

[0034] In a further possible embodiment of the milling machine according to the invention, the artificial intelligence module executes a trained learning algorithm during the milling process.

[0035] In a further possible embodiment of the milling machine according to the invention, the artificial intelligence module of the milling machine is designed to calculate a failure probability and / or a predicted downtime period of the milling machine if an anomaly is detected during the milling process.

[0036] In another possible embodiment of the milling machine according to the invention, the artificial intelligence module of the milling machine is further designed to determine at least one possible cause for an anomaly that occurs during the milling process. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Possible embodiments of the method according to the invention and of the device according to the invention for monitoring a milling process are explained in more detail below with reference to the attached drawings.

[0038] The accompanying drawings show:

[0039] Figure 1 A flow chart is shown for illustrating a possible specific embodiment of the method according to the invention for monitoring a milling process;

[0040] Figure 2 A block diagram is shown for representing an embodiment of a milling machine according to the present invention;

[0041] Figure 3 A block diagram is shown for illustrating a further possible embodiment of the milling machine according to the invention. DETAILED DESCRIPTION

[0042] As according to Figure 1 It can be seen from the flow chart of FIG. 1 that the method according to the invention for monitoring a milling process has two main steps in the embodiment shown.

[0043] In a first step S1, the rotational speed or the rotational speed of the milling head 2 of the milling machine 1 is detected. In addition, in step S1, at least one further operating parameter of the milling machine 1 is detected during the milling process. The further operating parameter has, for example, a supply current for operating the milling machine 1. The supply current is used, for example, to power an electric motor 3 driving the milling head 2 of the milling machine 1.

[0044] In a further step S2 , the detected rotational speed R and the detected operating parameters, in particular the supply current I, are evaluated by a trained learning algorithm in order to detect anomalies A during the milling process.

[0045] In another possible step ( Figure 1 In the embodiment of the present invention, after the anomaly is identified, a prompt is generated for the operator of the milling machine 1. For example, the identified anomaly can be displayed in real time on a display unit of the milling machine 1. In addition, in another step, maintenance measures can be automatically taken to eliminate the possible causes of the identified anomaly A. After the anomaly is identified during the milling process in step S2, in a feasible embodiment, the failure probability or the predicted downtime period of the relevant milling machine 1 is calculated. The feasible maintenance measures can coordinate the calculated failure probability and / or the predicted failure duration. For example, the milling machine 1 with a high failure probability is maintained or repaired with high urgency or priority. In addition, based on the calculated predicted downtime period, it is possible to control which maintenance measures are performed at which milling machines 1 in which time period. In another feasible embodiment of the method according to the present invention, after the anomaly is identified during the milling process, at least one possible cause is derived for the anomaly A that occurred, such as wear or blocking of the drive spindle. In a feasible embodiment, the trained learning algorithm for identifying anomalies has a trained neural network NN. The learning algorithm, in particular the neural network NN, is preferably trained in a training phase before the production milling process begins based on historical training data about the rotational speed R and at least one further operating parameter P of the milling machine 1. In one possible embodiment, the trained neural network NN comprises an RNN (recurrent neural network).

[0046] In one possible embodiment, a learning algorithm for detecting anomalies can be downloaded as an application (app) from a database via a network to a computing unit of the milling machine 1 and trained in a training phase for a subsequent production milling process. In one possible embodiment variant, the learning algorithm used can be selected as a function of the type of milling machine 1 concerned and / or the type of workpiece W to be milled in the milling process. In one possible embodiment, the workpiece W to be milled by the milling machine 1 comprises a circuit board.

[0047] In one possible embodiment, the rotational speed R and further operating parameters, in particular the supply current I, are detected in step S1 based on corresponding data, which are supplied at a data rate of approximately 1 / second to a trained learning algorithm implemented in a computing unit of the milling machine 1. In step S2, anomalies are determined or detected during the production milling process, i.e. during the production process, preferably in real time, by means of a trained learning algorithm, in particular a trained neural network NN.

[0048] In particular, one or more (short-term) current peaks of the supply current detected, for example for about 5 to 10 seconds, are identified as an anomaly. Here, the sudden rise of the supply current is characterized by a subsequently slowly falling edge. For example, such an anomaly is caused by milling chips entering the milling spindle. There, the milling chips cause an increase in friction. In order to keep the rotational speed or rotational speed of the milling head constant, the supply current is increased (sudden rise) by readjusting the supply voltage. If the milling chips are ground by the rotational movement of the milling spindle, the supply current will be normalized again (falling edge). This can be achieved so that the rotational speed of the milling head or milling spindle remains constant. In addition, in order to identify anomalies, the sampling rate of the supply current can be appropriately selected so that current peaks or anomalies are not missed due to undersampling of the supply current. In addition, a sufficiently low sampling rate of the supply current can be selected at the same time so that an excessive amount of data is not accumulated.

[0049] Finally, a quality indicator can be formed which indicates whether a detected measuring point of the supply current is no longer within the normal range for the milling machine. For example, the higher the value, the more likely the measuring point is outside the normal range. In this case, the quality indicator can be formed as a function of the rotational speed and the supply current. Based on the number of anomalies that occur, the remaining service life or failure of the milling machine can then be determined or predicted.

[0050] According to one embodiment, the milling process is therefore monitored, for example during the formation of milling chips by the milling of circuit boards, in particular in order to determine or predict a failure or a remaining service life of the milling machine.

[0051] In another embodiment, the supply current is monitored in the following manner to determine whether a base value of the supply current deviates briefly or momentarily and whether the supply current returns to the base value again. This monitoring can be performed by a trained learning algorithm. Additionally or alternatively, it is feasible to provide a threshold value by which such a current peak value of the supply current is identified. After identifying such a current peak value of the supply current, it is possible to monitor whether the supply current drops below the threshold value again, in particular to the base value of the supply current. The base value can be the value of the supply current, which constitutes the current peak value. The base value of the supply current can be predetermined by the desired rotational speed of the milling machine, more precisely the spindle of the milling machine, or the rotational speed of the milling machine can be predetermined by the base value of the supply current. In addition, the frequency of such current peaks can be detected and the failure or remaining service life of the milling machine can be determined based on this, which can be performed, for example, by quality indicators.

[0052] In order to ensure the identification of current peaks, sufficient sampling of the supply current is required. For example, the supply current can be sampled so that current peaks caused by milling chips entering the milling spindle can be detected. For example, the supply current can be sampled approximately once per second or more frequently than once per second.

[0053] In another embodiment, it is monitored whether a change in the supply current, in particular the current peak value, is caused when the rotational speed is (essentially) constant or unchanged. Here, for example, such a change in the supply current is caused by grinding milling chips. Therefore, in particular when the rotational speed of the milling head or the milling spindle is constant, the supply current can be monitored for the current peak value formed by grinding milling chips. This current peak value can have a typical variation curve. For example, this above-mentioned typical variation curve of the current peak value can be monitored and / or detected by a trained learning algorithm. For this purpose, the learning algorithm can be trained using historical data. Subsequently, the current peak value can be identified as an anomaly by the trained learning algorithm. In particular, the above-mentioned trained learning algorithm can distinguish the current peak value caused by grinding milling chips from current peak values ​​from other sources.

[0054] Furthermore, the workpiece to be processed can be a workpiece made of a homogeneous material or a section of a workpiece processed by a milling machine, which section is made of a homogeneous material. This makes it possible to keep the rotational speed of the milling head constant. Otherwise, in the case of inhomogeneous workpieces, this would result in a change in the rotational speed or the rotational speed would have to be adapted to the section of the workpiece that has just been processed. It can therefore be provided that a plurality of sections are processed at different predetermined rated rotational speeds. In this case, the respective rated rotational speed can be used as a base value to identify current peaks.

[0055] Instead of or together with the supply current, the supply voltage can also be detected and evaluated. Alternatively, other variables derived from the supply current and / or the supply voltage can also be used to monitor the milling process or to predict the remaining service life or failure probability of the milling machine.

[0056] The printed circuit board consists of an electrically insulating material with electrically conductive connections (conductor tracks) attached thereto. Fiber-reinforced plastics are common as insulating materials, and in cheaper devices, cardboard is often used. The conductor tracks consist primarily of a thin copper layer. The printed circuit board can also be structured by milling the copper layer ("insulation milling", see the solder grid printed circuit board in the figure below). Such printed circuit boards do not consist of conductor tracks, but of surfaces that are separated from one another by milling tracks.

[0057] In milling technology, the dividing lines between the conductor surfaces are created with the aid of a milling machine, for example a milling cutter. In this case, all copper remains in place (island method). Wet chemical steps and photolithography steps are eliminated. CNC programs can be generated with the aid of special CAD software, so that circuit boards can be produced quickly.

[0058] Figure 2 A block diagram is shown for illustrating an embodiment of a milling machine 1 according to the invention. Figure 2The milling machine 1 shown is used for milling a workpiece W. The workpiece W is, for example, a circuit board. The milling machine 1 has at least one rotatable milling head 2 for machining the workpiece W in a milling process. The milling head 2 is driven by an electric motor 3 of the milling machine 1. The electric motor 3 is supplied with a supply current. The supply current can include one or more current phases (L1, L2, L3). The milling machine 1 according to the invention has an artificial intelligence module 4, such as Figure 2 As shown. The artificial intelligence module 4 identifies an anomaly occurring during the milling process based on the detected rotational speed R of the milling head 2 and at least one further operating parameter P of the milling machine 1 during the milling process. The artificial intelligence module 4 can evaluate a plurality of operating parameters P to identify the anomaly A. In one possible embodiment, the artificial intelligence module 4 implements a trained learning algorithm during the milling process.

[0059] In a feasible embodiment, the artificial intelligence module 4 implements a trained neural network NN with multiple layers. The trained neural network NN has an input layer, one or more intermediate layers (hidden layers) and an output layer. Neuron nodes with corresponding activation functions are in each layer. Neuron nodes in different layers are linked to each other, wherein the links are weighted.

[0060] The weighting factor w is set in the training phase. The artificial intelligence module 4 or the neural network can be continuously trained in the running milling process. In a feasible embodiment, the training of the artificial intelligence module 4 is carried out in an unsupervised learning process. The learning process is preferably carried out based on historical training data. The historical training data includes, for example, historical power supply data and historical rotation speed data of the milling head 2. The historical training data is located in a data memory. In a feasible embodiment, the training data is loaded from the database of the operator of the milling machine and used for training purposes.

[0061] In a possible embodiment, the learning algorithm used, for example the neural network NN used, can be selected according to the type of milling machine 1 and / or the type of workpiece W to be milled in the milling process. In a possible embodiment, the learning algorithm can also be downloaded from a database to the artificial intelligence module 4 of the milling machine 1. For example, training data and a suitable learning algorithm can be downloaded from a cloud platform 6 of a milling machine manufacturer or a milling machine operator via a data network. For example, a learning algorithm suitable for monitoring a milling process at a circuit board is downloaded. In a possible embodiment, a vector is attached to the input layer of the neural network, which vector includes different operating parameters P of the milling machine 1. The state vector z For example, there are two operating parameters P, in particular the rotational speed R of the milling head 2 and the current I taken by the electric motor 3. Depending on the application, the state vector φt is attached to the input layer of the neural network NN.z Other operating parameters P can be included. For example, each phase L of a multi-phase supply current can be represented as an operating parameter P. The artificial intelligence module 4 can output the calculated failure probability of the milling machine 1 as an output value, in particular for a predicted downtime period. In addition, the identified anomaly A can be output to the user of the milling machine 1 in real time via a user interface. In a feasible embodiment, the milling process itself can be controlled by a CNC program, which is implemented by a controller or microprocessor of the milling machine 1 according to the present invention. During the milling process, the feed movement of the milling head 2 is perpendicular to or inclined to the axis of rotation of the milling head 2. If the workpiece W is a circuit board, the milling machine 1 generates a trajectory in a manner controlled by software or a CNC program, which circumvents the desired conductor line. For example, the conductor line to be manufactured is insulated from the remaining conductive material, such as copper. This is also called insulation milling. In order to mill the insulating channel, a smaller conical milling head can be used, which cuts off the conductive material, in particular copper, along the calculated motion trajectory. The removed material can lead to a blockage or impairment of the movement capability of the milling head 2, wherein such abnormalities or anomalies can be identified or detected early by means of the method according to the invention. In a possible embodiment, corresponding evaluation numbers can also be calculated for different anomalies A, which indicate the extent of the identified anomaly or the failure probability derived therefrom. In a possible embodiment, such anomaly characteristic values ​​can be output to the operator of the milling machine 1 via a display unit in real time. With the method according to the invention, possible anomalies can be identified around the clock during the operation of the milling process and possible causes of failure can be derived. As a result, repair measures or maintenance measures can be taken or implemented in a timely manner at the relevant milling machine 1, so that downtimes in the milling process are largely avoided. The method according to the invention can be used for a large number of milling machines used simultaneously in a production line.

[0062] exist Figure 2In the embodiment shown, the artificial intelligence module 4 is implemented locally in the computing unit of the milling machine 1. In an alternative embodiment, the artificial intelligence module 4 can also be implemented on a remote server of the cloud platform 6, which communicates bidirectionally with the control device of the milling machine 1 via a data connection. In a possible embodiment, the technical data connection includes a local network or an intranet. Alternatively, the data connection can also be a global network or the Internet. The method for monitoring the milling process according to the invention is a computer-implemented method, which is stored on a computer-readable medium in a possible embodiment. The computer-readable medium includes, for example, a data memory of the milling machine 1. In addition, the computer-readable medium can also have a portable data carrier, such as a CD-ROM or a USB stick. Depending on the application, in a possible embodiment, different learning algorithms can be used to identify anomalies during the milling process. For example, a support vector machine can be provided for calculating quality indicators. In a possible embodiment, additional sensors are provided around the milling head 2, which provide additional data to the artificial intelligence module 4. For example, an image sensor can use image data as a state vector z The input operating parameter P is provided to the artificial intelligence module 4. As a result, the reliability or accuracy of the method according to the invention can be further improved.

[0063] Figure 3 Another embodiment of the device according to the present invention for monitoring a milling process is schematically shown. In the illustrated embodiment, the milling machine 1 provides the operating parameters detected via the interface to a cloud platform 6 via a data network or cloud 5, in which an artificial intelligence module 4 is integrated. Abnormalities or prompts can be output to the operator of the milling machine 1 in real time via a display device 7. In a possible embodiment, the display device 7 is also connected to the data network or cloud 5. In a possible embodiment variant, the display device 7 is a portable device, such as a user's smartphone. With the aid of the method according to the present invention, deviations in the feature space of the operating parameters are detected at an early stage. In addition, with the aid of the data model M, the cause of the abnormality or failure of the milling machine 1 can be automatically derived. The data model M preferably describes the structural construction of the milling machine 1, for example as a body. In Figure 3In the illustrated embodiment, two characteristics or operating parameters P, namely the supply current I and the rotational speed R, are used to train the learning algorithm or the neural network. In a possible embodiment, the trained learning algorithm requires a relatively small storage space of less than 10 MB. This allows the learning algorithm to be implemented by the cloud platform 6 on the automation device. The method according to the invention can be expanded in a simple manner for a large number of milling machines 1. In a possible embodiment, the artificial intelligence module 4 provides an evaluation number for the probability of an identified anomaly or possible failure. In a possible embodiment of the method according to the invention, the operating parameter P is evaluated with the aid of a trained learning algorithm on a server of the cloud platform 6 based on operating data, which is transmitted from one or more milling machines to the server of the cloud platform 6 via the network 5. In a possible embodiment, the learning algorithm can be downloaded from a database as an application and implemented by the artificial intelligence module 4 after training. In a possible embodiment, the artificial intelligence module 4 can be set locally on the milling machine 1, such as Figure 2 As shown, or remotely implemented on a server of the cloud platform 6, such as Figure 3 shown.

Claims

1. A method for monitoring a milling process of a circuit board, the method comprising the following steps: a) detecting the rotational speed of a milling head (2) of a milling machine (1) and at least one further operating parameter of the milling machine (1) during the milling process; wherein: The further operating parameter is a supply current for operating the milling machine, wherein a current peak value of the supply current is detected while the rotational speed remains constant, The sampling of the supply current is performed in such a way that the current peak value caused by the milling chips entering the milling spindle is detected. b) evaluating said detected rotational speed and detected operating parameters by means of a trained learning algorithm to identify anomalies during said milling process, Therein, a current peak of the supply current caused by grinding milling chips is detected, the current peak having a characteristic curve that is recognized by the trained learning algorithm, wherein the learning algorithm is trained in a training phase before the start of the production milling process based on historical training data about the rotational speed and at least one of the further operating parameters of the milling machine.

2. The method according to claim 1, wherein: After the abnormality is detected, a prompt is generated for the operator of the milling machine (1) and the prompt is displayed in real time on a display unit of the milling machine (1) and / or maintenance measures are automatically taken.

3. The method according to claim 1 or 2, wherein: After an anomaly is detected during the milling process, a failure probability and / or a predicted downtime period of the milling machine (1) is calculated.

4. The method according to claim 1 or 2, wherein: After an anomaly has been detected during the milling process, at least one possible cause for the occurrence of the anomaly is determined.

5. The method according to claim 1 or 2, wherein: A trained learning algorithm has a trained neural network.

6. The method according to claim 1 or 2, wherein: The learning algorithm is downloaded as an application from a database via a network into a computing unit of the milling machine (1), and in a training phase the learning algorithm is trained for a production milling process.

7. The method according to claim 1, wherein: Using the rotational speed and the further operating parameter, a quality indicator is formed by means of the trained learning algorithm, and the quality indicator is used to determine one or more anomalies by means of the trained learning algorithm.

8. The method according to claim 1 or 2, wherein: The rotational speed and at least one further operating parameter of the milling head (2) of the milling machine (1) are detected and the corresponding data are supplied at a data rate of 1 / second to a trained learning algorithm implemented in a computing unit of the milling machine (1).

9. The method according to claim 1 or 2, wherein: The anomalies are derived in real time during the production milling process by means of a trained learning algorithm.

10. A milling machine (1) for milling a workpiece (W), wherein: The workpiece is a circuit board, and the milling machine has - a rotatable milling head (2) for machining the workpiece (W) in a milling process, - an electric motor (3) for driving the milling head (2), and - having an artificial intelligence module (4) which, during the milling process, identifies an anomaly in the milling process based on the detected rotational speed of the milling head (2) and at least one further operating parameter of the milling machine (1), wherein the further operating parameter is the supply current of the electric motor (3), The artificial intelligence module is further used to detect the current peak value of the power supply current when the rotation speed remains unchanged. The sampling of the supply current is performed in such a way that the current peak value caused by the milling chips entering the milling spindle is detected. Therein, the artificial intelligence module is used to detect current peaks of the supply current which are caused by grinding milling chips and have a characteristic curve, and to identify the current peaks with the aid of a trained learning algorithm, wherein the learning algorithm is trained in a training phase before the start of the production milling process based on historical training data about the rotational speed and at least one of the further operating parameters of the milling machine.

11. The milling machine according to claim 10, wherein: The artificial intelligence module (4) implements a trained learning algorithm during the milling process.

12. The milling machine according to claim 10 or 11, wherein: The artificial intelligence module (4) of the milling machine (1) is designed to, when an anomaly is detected during the milling process, calculate a failure probability and / or a predicted downtime period of the milling machine (1) and / or to derive at least one possible cause for the occurrence of the anomaly during the milling process.

13. The milling machine according to claim 10 or 11, wherein: The artificial intelligence module is used to form a quality indicator using the rotational speed and the further operating parameter, and the quality indicator is used to determine one or more anomalies, also using the trained learning algorithm.

Images