Abnormality detection for cutting machine with electrically driven spindle

Through the electrical power consumption analysis and Gaussian hybrid model, the problem of low spindle abnormality detection accuracy in the prior art is solved, high-precision abnormality detection and fault prediction of the spindle of the cutting machine is realized, and maintenance arrangements are optimized.

CN118927017APending Publication Date: 2024-11-12SIEMENS AG
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410574563.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2023-05-12
Filing Date
2024-05-10
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing cutting machine spindle abnormality detection method requires a lot of unavailable information, and it is difficult to accurately distinguish whether the spindle is cut or idling at a given time, resulting in low detection accuracy and inability to predict spindle failures and arrangements in time. maintain.

Method used

By determining the electric power consumption of the motor during the working cycle, dividing the time window, calculating the median and expected deviations of the electric power value, using the Gaussian hybrid model to distinguish the cutting and idle states, and detecting abnormalities based on the deviation relationship of the electric power value, predicting spindle faults in advance and Lifespan, optimize maintenance arrangements.

Benefits of technology

It improves the accuracy of spindle abnormality detection, can distinguish without understanding the cutting and idle states, predict faults in advance, extend the running time of the cutting machine, and optimize maintenance arrangements.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118927017A_ABST
    Figure CN118927017A_ABST
Patent Text Reader

Abstract

Abnormality detection for a cutting machine having an electrically driven spindle. The invention relates to a method (200) for anomaly detection in a cutting machine (105) having a spindle (115) driven by an electric motor (120). The proposed method (200) comprises the steps of: determining (215) an electrical power consumption of the electric motor (120) during a duty cycle (305) of the machine (105); dividing the duty cycle into time windows (335); determining (225), for each time window (335), whether the spindle (115) is cutting and / or idle based on the associated electrical power value; determining (245) a median deviation and an expected deviation of the electrical power values for the cleaving and idling operations over the sequence of time windows (335); and determining (250) an anomaly if the electrical power value during cutting or idle running exceeds a predetermined relationship with the corresponding median deviation value and the expected deviation value.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to a cutting machine having an electrically driven spindle. More particularly, the present invention relates to determining abnormalities of the spindle. Background Art

[0002] The cutting machine is adapted to cut the sheet material into smaller pieces. The cutting can be done using a circular saw or a milling cutter. The cutting element may also be referred to as a spindle because it rotates at a high speed around a predetermined axis. The spindle is driven by an electric motor which can be controlled to a predetermined speed.

[0003] During cutting, the spindle undergoes a certain amount of wear and tear, so that its life may be limited. After a certain time, the spindle may bend or break, making it unable to be used for cutting any further. The spindle may be a consumable part, and a new spindle may be inserted into the cutting machine. However, replacing the spindle may take some time. The cutting machine may be involved in an integrated production process, which may have to be stopped to replace the spindle.

[0004] Several methods have been proposed to estimate whether the spindle in a cutting machine needs to be replaced. Some known methods utilize machine learning techniques. Based on the operating data of the spindle, a pattern observable during normal operation can be learned. Deviations from that pattern can be considered to indicate an anomaly and a warning can occur. However, most of such patterns require information that may not be available, such as whether the spindle is cutting or spinning idly at a given time. Summary of the invention

[0005] The object of the present invention is to provide an improved technique for determining anomalies in a cutting machine with an electrically driven spindle. The present invention solves the given problem by means of the subject matter defined in the attached independent claims. The dependent claims describe preferred embodiments.

[0006] The proposed method relates to anomaly detection in a cutting machine having a spindle driven by an electric motor. The method comprises the following steps: determining the electric power consumption of the electric motor during a working cycle of the machine; dividing the working cycle into time windows; for each time window, determining whether the spindle is cutting and / or idling based on the associated electric power value; determining the median deviation and the expected deviation of the electric power values ​​for cutting and idling operations over a sequence of time windows; and determining an anomaly if the electric power value during cutting or idling exceeds a predetermined relationship with the corresponding median deviation value and the expected deviation value.

[0007] To get the expected deviation, the interquartile range of the measurements can be adjusted so that it corresponds to the standard deviation of the normal distribution. To do this, just multiply it by a factor of 1 / (2*0.6745).

[0008] Anomaly detection can be performed on minimal data extracted from the cutting machine. The method can be performed in particular without knowing whether the spindle is cutting or just spinning idly at a given time. This distinction can be made by the method, thereby detecting anomalies with improved accuracy. The anomaly can indicate an impending spindle failure. Therefore, the operating state of the spindle can be determined with improved accuracy.

[0009] The cutting machine may be operated in a pattern where some of the time the spindle is cutting and other times the spindle is just spinning and not cutting (which is referred to herein as idling or spinning idly). This pattern may be predetermined or programmed, and it may not be known and required as a parameter for the proposed method.

[0010] In one embodiment, the cutting machine may comprise a depaneling machine for PCBs. The machine may be adapted to cut sheet material into smaller pieces. Various staggered processing steps may be performed on the PCB so that the spindle will occasionally remain rotating without cutting until another step has been completed.

[0011] It is particularly preferred that the determination of whether the spindle is cutting or idling during the time window is performed using a first Gaussian mixture model. A Gaussian mixture model (GMM) can be used to determine whether the spindle is always cutting, always idling, or cutting during a portion of the window and idling during another portion within the current time window. The GMM can also be used to determine statistically relevant numbers for distinguishing measurements related to cutting and those related to idling separately.

[0012] It is particularly preferred to determine the rotational speed of the spindle together with the electrical power value.The mean deviation and the expected deviation of the electrical power value can then be determined for the adapted rotational speed.

[0013] It can help to consider that the wear on the spindle may be related to the speed of the spindle. In addition, the electrical power value may be related to the speed. The higher the speed, the higher the power required to operate the spindle. Therefore, the behavior of the spindle at different speeds may change. The speed of the motor and the electrical power consumption can be obtained from the controller of the cutting machine.

[0014] Particularly preferably, a second Gaussian mixture model is used to determine the different rotational speeds of the spindle. The second Gaussian mixture model can determine how many different speeds and which different speeds are used during the working cycle. On this basis, it can be determined which of the time windows are related to which determined rotational speeds.

[0015] The electric power values ​​of different time windows can be collected in a histogram, and then based on the histogram the median deviation and the expected deviation of the electric power values ​​are determined. One histogram can be related to the rotational speed of the spindle, and entries can be collected from all time windows in which a matching rotational speed of the spindle is used. The histogram will usually show two accumulations related to the electric power consumption during cutting and during idling. However, it is also possible to determine the GMM as well as the median deviation and the expected deviation (or interquartile range) directly from the measurements.

[0016] Anomaly determination may take into account the relative frequency of the cutting time and the idling time of the spindle. For example, if there are a large number of entries for cutting operations and a small number of entries for idling operations in a time window, anomalies may be determined based on the relative weights of the cutting entries and the idling entries. An outlier may satisfy a predetermined relationship between the deviation of an observed data point from a determined median and the deviation expected based on the weights.

[0017] Based on the occurrence of anomalies of the spindle, an impending spindle failure can be determined. In addition, the remaining spindle life can be determined based on the observed anomalies. Thus, the maintenance of the cutting machine, especially for the replacement of the spindle, can be planned in advance with improved accuracy. The cutting machine can remain in operation for a longer period of time and maintenance downtimes can be advantageously scheduled. The quality of a type of spindle can be determined based on the observed or estimated life.

[0018] In general, it is preferred that the time windows have at least similar lengths. A typical time window may have a length of about 30 seconds to two minutes. A working cycle of a machine may take up to half an hour, an hour or even longer. More precisely, during the fitting phase, when the first GMM is applied, the median length of several working cycles can be calculated. The time window length range can then be determined accordingly. Exemplary default length ranges may include 10 to 30 seconds, 20 to 50 seconds, and 30 to 120 seconds. It is also possible to configure this range manually. If the spindle usually runs in a cycle of 10 minutes, an interval of 30 to 120 seconds can be selected; if the spindle usually runs in a cycle of 15 seconds, an interval of 10 to 30 seconds may be more appropriate.

[0019] In another refinement of the method, the portion of the operating cycle in which the spindle accelerates from a standstill or decelerates to a standstill can be excluded from the determination time window. During such a time, the relationship between the electric power consumption of the motor and the rotational speed of the spindle can differ from the relationship that can be observed during normal operation.

[0020] During these ramp-up times or ramp-down times, a monotonic large increase or decrease in speed can be observed, respectively. After excluding these times, we can assume that the speed is more or less constant over each time window. The second Gaussian model then allows the determination of the prevailing speed in the time window. The assumption of relatively constant speed is usually very valid, because the spindle is programmed to run at a specific speed.

[0021] According to another aspect of the present invention, a computer program product comprises instructions, which can enable an electronic device and / or an electronic control system to perform the method described herein. To this end, the computer program product can be executed on the electronic device and / or the electronic control system.

[0022] According to yet another aspect of the present invention, there is provided a computer readable medium on which a computer program product as described herein is stored.

[0023] According to another aspect of the present invention, a monitoring device for a cutting machine having a spindle driven by an electric motor is provided. Therein, the monitoring device comprises: a first interface of the cutting machine for determining the electric power consumption of the electric motor; a second interface for outputting an indication; and a processing device. The processing device is adapted to: determine the electric power consumption of the electric motor during a working cycle of the machine; divide the working cycle into time windows; determine, for each time window, based on the associated electric power values, whether the spindle is cutting and / or idling; determine the median deviation and expected deviation of the electric power values ​​for cutting and idling operations over a sequence of time windows; and determine an anomaly if the electric power value during cutting or idling exceeds a predetermined relationship with the corresponding median deviation value and expected deviation value.

[0024] The device may be adapted to carry out, in whole or in part, the method as described herein. The processing means may include an electronic execution device such as a computer system. The execution device may also include a programmable microcomputer, a microcontroller or an integrated circuit. The features or advantages of the method may be applicable to the device, and vice versa.

[0025] The indication may foreshadow an impending spindle failure. The indication may be given to personnel associated with operating the cutting machine. The indication may also be given in electronic form and further evaluated by a monitoring device or a different device. The evaluation may in particular comprise determining the operating state of the spindle or estimating the remaining life.

[0026] Furthermore, it is preferred that the cutting machine comprises a panel splitter. The panel splitter may be adapted to cut the PCB sheet into smaller parts. The cutting machine may be part of an integrated production system for PCBs. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The above-mentioned characteristics, features and advantages of the present invention and the manner in which they are achieved will be made clearer and better understood from the following discussion with reference to exemplary embodiments illustrated in the accompanying drawings, in which:

[0028] Figure 1 A system is shown;

[0029] Figure 2 A flow chart illustrating an exemplary method;

[0030] Figure 3 shows exemplary electrical power and speed data for a spindle during a work cycle;

[0031] Figure 4 The second processing step is shown. DETAILED DESCRIPTION

[0032] Figure 1 A system 100 is shown, comprising: a cutter 105 having a spindle 115 driven by a motor 120, a controller 125, and a monitoring device 130. The cutter 105 is adapted to cut a larger PCB (printed circuit board) 110 into smaller parts. The cutter 105 may be part of a production facility for the PCB 110. Cutting may be done before or after electrical or electronic components have been placed and soldered to the PCB material 110.

[0033] PCB material 110 is usually in the form of a laminated sandwich structure of a conductive layer and an insulating layer. The conductive layer may include copper, and the insulating layer may include phenolic cotton paper or glass and epoxy resin. In addition, the PCB production line may include machines for optical coating, exposure, development, etching, cleaning, stacking, riveting or lamination.

[0034] Controller 125 is adapted to control the rotational speed of spindle 115 and may also control horizontal and / or vertical movement of spindle 115 relative to PCB sheet material 110. For cutting tasks, machine 105 may be placed in a directed contact with PCB 110 through a predetermined motion or motion pattern.

[0035] Cutting one PCB board 110 into many smaller pieces requires a series of movements that can be included in one work cycle. Between work cycles, the spindle 115 can be braked to a stop and possibly moved to a parked position when the cut pieces of PCB 110 can be removed from the work area and a new sheet of PCB 110 can be prepared. Before the cutting of the next sheet occurs, the spindle 115 can be accelerated to the operating speed.

[0036] The controller 125 is adapted to provide a first signal indicating the electrical power consumed by the motor 120, preferably indicating the current flowing through the motor 120. The voltage of the motor 120 can be kept constant, or a signal indicating the voltage can be provided so that the electrical power consumed can be determined as the product of the voltage and the current. The controller 125 can also provide a second signal indicating the rotational speed of the spindle 115.

[0037] During the operation of the cutter 105, the spindle 125 experiences a certain amount of wear. Therefore, the spindle 125 may show a decreased performance at a certain moment, or even deform or break, so that it can no longer be used for cutting. If in fact the spindle 125 breaks during operation, the PCB 110 or the cutter 105 may be damaged. Replacing the spindle 125 or repairing the cutter 105 may take some time, during which the cutter 105 and possibly other machines on the production line may have to stop. Such a machine may require a complicated, expensive or time-consuming deactivation / reactivation process.

[0038] Therefore, it is desirable to have an estimate of the health of the spindle 125 so that replacement of the spindle 125 can be prepared or scheduled in advance. The health can be determined based on the type of spindle 125 used and a measure of the physical stresses it has been subjected to to date.

[0039] The monitoring device 130 is adapted to monitor the operation of the spindle 115 and to generate a signal if an anomaly of the spindle 125 is detected. To this end, the device 130 analyses information made available by the controller 125. Assume that the controller 125 does not provide a signal indicating whether the spindle 115 is currently cutting or idling.

[0040] The monitoring device 130 comprises a first interface 135, a second interface 140 and a processing means 145. The processing means 145 may comprise electronic circuitry, and preferably comprises a programmable microcomputer or microcontroller.

[0041] The first interface 130 can be connected to the controller 125 to obtain at least a first signal indicating the electrical power used to rotate the spindle 125, and preferably also a second signal indicating the rotational speed of the spindle 115. The second interface 140 can be connected to a human interface, such as an optical, acoustic or tactile output device. The second interface 140 can also be connected to another system for processing abnormal signals. The processing can particularly include determining a health or operating state, or estimating the remaining life of the spindle 115.

[0042] The processing device 145 is adapted to determine anomalies based on information provided by the controller 125. To this end, it may be determined whether the spindle 115 is cutting or idling at a given time or during a given period of time. The average electrical power consumption for cutting and idling may be determined, and the current electrical power consumption may be compared to the average electrical power consumption. The anomaly may be determined based on a predetermined relationship between those values ​​and possibly one or more other values.

[0043] Figure 2 A flow chart of an exemplary method 200 for detecting anomalies of a spindle 125 in a cutting machine 105 is shown.

[0044] In step 205, the current rotational speed of the spindle 125 can be determined. This can be done based on a signal from the controller 125. A useful unit of rotational speed is revolutions per minute (RPM). If the spindle 125 runs at a fixed speed, step 205 can be omitted and the fixed speed can be considered as the basis for the following steps. The electrical power consumed by the motor 120 can also be determined in step 205. The current flowing through the motor 120 can be used as an indication of the electrical power. The motor 120 can be a DC, AC or multi-phase motor, and various control schemes can be applied to control its rotational speed.

[0045] In step 210, the time in which the rotation speed of the spindle 120 is ramped up from stop to operating speed or vice versa can be skipped. In one embodiment, the time in which the rotation speed reported by the controller 125 is below a predetermined limit can be skipped. In another embodiment, a fixed time after the motor 120 is started or before the motor 120 is shut down can be skipped. It is assumed that one working cycle of the cutting machine 105 is outlined by such acceleration time and deceleration time.

[0046] In step 215, the speed of the spindle 120 and the electrical power consumed by the motor 120 can be collected during the current working cycle. During a working cycle, the speed of the spindle 115 can be controlled to different values. Therefore, the speed and electrical power measurements are collected in a sequence of time windows that are generally of the same length. For each time window, the subsequent steps 220 to 230 can be performed once. For example, a time window can cover 30 seconds to 2 minutes, while a working cycle can take up to 45 minutes or more.

[0047] In step 220, speed and electrical power data are collected in the window under consideration. This may produce a two-dimensional series of data points for the time window, where each data point represents a rotational speed and an associated electrical power value.

[0048] When the spindle 115 is cutting through the PCB material 110, the motor 120 consumes more electrical power, and the consumed electrical power varies more than when the spindle is rotating at a comparable speed without cutting. Therefore, the relationship between the rotational speed and the electrical power drawn can provide an indication of the operation of the spindle 120 at a given time. However, the electrical power consumed can also depend on the rotational speed, because the motor 120 can consume more electrical power at higher speeds. An abnormality of the spindle 115 may be accompanied by an increased or varying electrical power consumption of the motor 120.

[0049] In step 225, a first Gaussian mixture model GMM1 may be applied to separate data points representing cutting from data points representing idling. The electrical power value of each data point may be input into a first histogram, and the rotational speed value may be input into a second histogram.

[0050] In step 230, the first GMM can determine the fit on the values ​​of each histogram. The median deviation and expected deviation for the speed and electric power during cutting and idling can be provided. For each time window, steps 220 to 230 can be repeated once. However, the histogram processing is optional and is used in this article to better explain the process. Direct determination can be used instead.

[0051] In step 240, a second Gaussian mixture model GMM2 may be applied to the data from all time windows to determine which different rotational speeds were used during the duty cycle. Then, for time windows with matching rotational speeds, the median deviation and expected deviation of the electrical power values ​​for both cutting and idling operations may be determined.

[0052] In step 250, an anomaly during the duty cycle may be determined if the electrical power value in any of the time windows exceeds a predetermined threshold. This threshold may be determined as the quotient between the observed deviation from the associated median and the expected deviation from the fit.

[0053] Optionally, in step 255 , the remaining life of the spindle 115 may be determined based on the detected anomalies, the estimated cutting stresses, and / or data regarding the type of spindle 115 being used.

[0054] In step 260, a signal may be issued indicating a detected anomaly of the spindle 115 and / or an impending end of life of the spindle 115. The signal may be used to schedule replacement of the spindle 115. In another embodiment, it may be determined what stresses a given type of spindle 115 can withstand before the probability of failure exceeds a predetermined value.

[0055] Figure 3Exemplary electrical power and speed data 300 is shown for a spindle 115 during a work cycle 305. The horizontal direction reflects time, and the vertical direction reflects rotational speed and electrical power, respectively. A first series 310 shows speed, and a second series 315 shows electrical power values.

[0056] At the beginning of the work cycle 305, the rotational speed of the spindle 115 is ramped up from zero to the working speed. This process is shown in detail 320 in an enlarged manner. Similarly, the spindle speed is shown to be ramped down from the working speed to zero at the end of the work cycle 305 in detail 325. In both cases, the relationship between the rotational speed and the consumed electrical power does not follow the same relationship during the cutting or idling of the spindle 115. Therefore, the electrical power and speed data from the stages shown in details 320 and 325 may be discarded. After this, a time period 330 remains, which can then be subdivided into time windows 335 of equal duration.

[0057] Figure 4 Shown from Figure 3 4. For each time window 335, the electrical power data may be collected in a first histogram 405 and the speed data may be collected in a second histogram 410. A first Gaussian mixture model GMM1 may be used to create a fit to the histogram data and determine the median deviation and the expected deviation.

[0058] A second Gaussian mixture model GMM2 can be applied to the available speed data to construct a time window group with matching speeds. Therefore, the electrical power information related to the matching speed can be summarized. By connecting the fitted data with the GMM parameters, the median deviation and the expected deviation can be determined for the speed during cutting and during idling, respectively. These parameters can be determined for each of the speeds encountered during the working cycle 305. Anomalies of the spindle 115 can be determined based on the calculated parameters.

[0059] Considering an exemplary time window 335 in which the main shaft 115 has been idling, an anomaly may be determined if the observed electrical power deviates from the determined median value of the electrical power during idling by more than the expected power deviation multiplied by some configurable threshold. One possible default threshold is 20, so if the actual deviation is 20 times the expected deviation, an anomaly may be determined.

[0060] However, the time window 335 may include portions where the spindle 115 is cutting and other portions where the spindle 115 is idling. To properly determine anomalies, the relative idling time may be considered when processing the electrical power values. Anomalies may be detected if:

[0061] (I*(Mi-Pi)+(1-I)*(Mc-Pc)>C*(I*Di+(1-I)*Dc)

[0062] (I*Mc–Pc)-((1-I)*Mi–Pi)>I*Dc+(1–I)*Di

[0063] in:

[0064] I Relative idling time

[0065] (1–I) Relative cutting time

[0066] Mc median electric power during cutting

[0067] Pc Observed electrical power during cutting

[0068] Mi median power at idling

[0069] Pi Observed electrical power at idle

[0070] Dc Expected electric power deviation during cutting

[0071] Di Expected electric power deviation at idling

[0072] N Speed

[0073] Even though the invention has been described and explained in detail above with reference to preferred embodiments, the invention is not to be construed as being limited to the examples given. Subject matter experts may come up with variations or alternative combinations of features given in different embodiments without departing from the scope of the invention.

[0074] Regardless of grammatical term usage, individuals with male, female, or other gender identities are included in the term.

[0075] Reference List

[0076] 100 Systems

[0077] 105 Cutting Machine

[0078] 110 Printed Circuit Board PCB

[0079] 115 Spindle

[0080] 120 Electric Motor

[0081] 125 Controller

[0082] 130 Monitoring equipment

[0083] 135 First Interface

[0084] 140 Second interface

[0085] 145 Processing Device

[0086] GMM1 The first Gaussian mixture model

[0087] GMM2 Second Gaussian Mixture Model

[0088] 200 Methods

[0089] 205 Determine current RPM

[0090] 210 Skip Ascent / Descend

[0091] 215 Collect data during the work cycle

[0092] 220 collects data in windows 1...N

[0093] 225GMM1: cutting; idling

[0094] 230 Estimated idling power, cutting power, speed

[0095] 240GMM2: Group operation speed

[0096] 245 Determine the cutting and idling power and the expected deviation for each speed in all time windows

[0097] 250 Confirm abnormality

[0098] 255 Determine remaining spindle life

[0099] 260 Signal

[0100] 300 data

[0101] 305 Working Cycle

[0102] 310 Speed

[0103] 315 Power

[0104] 320 Details of the ramp up

[0105] 325 Details of the slope descent

[0106] 330 paragraphs

[0107] 335 Time Window

[0108] 405 First Histogram

[0109] 410 Second Histogram

Claims

1. A method (200) for anomaly detection in a cutting machine (105) having a spindle (115) driven by a motor (120), the method comprising the following steps: - determining (205) the electrical power consumption of the electric motor (120) during a working cycle (305) of the machine; - dividing the working cycle (305) into time windows (335); - for each time window (335), determining (225) whether the spindle (115) is cutting and / or idling based on the associated electrical power value; - determining (245) a median deviation and an expected deviation of the electrical power values ​​for cutting and idling operations over a sequence of time windows; as well as - If the electric power value during cutting or idling exceeds a predetermined relation to the corresponding median deviation value and the expected deviation value, an anomaly is determined (250).

2. The method (200) of claim 1, the cutting machine (105) operating in a mode in which some of the time the spindle (115) is cutting and other times the spindle (115) is rotating idly therebetween.

3. The method (200) according to claim 1 or 2, wherein: A first Gaussian mixture model (GMM1) is used to perform a determination of whether the spindle (115) is cutting and / or idling during a time window.

4. The method (200) according to one of the preceding claims, wherein: The rotational speed of the spindle (115) is determined (205) together with the electrical power value; wherein for the matched rotational speeds, a median deviation and an expected deviation of the electrical power value are determined (245).

5. The method (200) according to claim 4, wherein: A second Gaussian mixture model (GMM2) is used to determine different rotational speeds of the spindle (115).

6. The method (200) according to one of the preceding claims, wherein: The anomaly determination (250) considers the relative frequencies of the cutting time and the idle time of the spindle (115).

7. The method (200) according to one of the preceding claims, wherein: An impending spindle (115) failure is determined (255) based on the occurrence of the spindle (115) anomaly.

8. The method (200) according to one of the preceding claims, wherein: The time windows (335) have at least similar lengths.

9. The method (200) according to one of the preceding claims, wherein: The portion (320, 325) of the work cycle (305) in which the spindle (115) accelerates or decelerates from a stop to a stop is excluded from the time window (335).

10. Computer program product comprising instructions which, when executed on an electronic device (145) and / or an electronic control system (130), cause the electronic device and / or the electronic control system to perform the method (200) according to one of claims 1 to 9.

11. Computer readable medium having stored thereon a computer program product according to claim 10.

12. A monitoring device (130) for a cutting machine (105) having a spindle (115) driven by an electric motor (120), wherein: The monitoring device comprises: - a first interface (135) to the cutting machine (105) for determining the electrical power consumption of the electric motor (120); - a second interface (140) for outputting an indication; - a processing device (145) adapted to: - determining the electrical power consumption of the electric motor (120) during a working cycle of the machine; - dividing the working cycle into time windows; - determining, for each time window, based on the associated electric power value, whether the spindle (115) is cutting and / or idling; - determining the median deviation and the expected deviation of the electrical power values ​​for cutting and idling operations over a sequence of time windows; and - If the electric power value during cutting or idling exceeds a predetermined relation to the corresponding median deviation value and the expected deviation value, an anomaly is determined.

13. The device (130) according to claim 12, wherein: The indication indicates an impending spindle (115) failure.

14. The device (130) according to one of claims 12 or 13, wherein: The cutting machine (105) comprises a panel separator.