Machine and method for laser processing

The machine and method automate quality control in laser processing by using a neural network to analyze cutting images, ensuring accurate scrap detection and detachment, addressing incomplete cuts and collision risks.

WO2026074455A1PCT designated stage Publication Date: 2026-04-09ADIGE SPA
View PDF 4 Cites 0 Cited by

Patent Information

Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Filing Date
2025-10-01
Publication Date
2026-04-09

AI Technical Summary

Technical Problem

Existing laser processing machines lack automation for quality control, particularly in detecting and ensuring the detachment of scrap material defined by closed cut geometries, which can lead to incomplete cuts and potential collisions.

Method used

A machine and method incorporating a monitoring device with an image acquisition system and an analyzing unit using a trained neural network to analyze laser cutting images, determining the presence of material in closed cut geometries and ensuring accurate scrap detachment.

Benefits of technology

Enables automated quality control of laser-cut workpieces by detecting and ensuring the detachment of scrap material, reducing the need for manual verification and minimizing risks of incomplete cuts or collisions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000030_0000
    Figure 00000030_0000
  • Figure 00000031_0000
    Figure 00000031_0000
  • Figure 00000031_0001
    Figure 00000031_0001
Patent Text Reader

Abstract

A machine for the laser processing (1) is described, configured to cut a workpiece (2), comprising a laser device (4) configured to direct a laser beam (5) onto the workpiece (2); a movement unit (6) for inducing a relative movement between the workpiece (2) and the laser beam (5); a control unit (7) configured to control in a coordinated manner the laser device (4) and the movement unit (6) to cut the workpiece (2) according to a predetermined cutting pattern; a monitoring device (10) configured to acquire, in use, a plurality of images (11) of at least a portion of the workpiece (2); and an analyzing unit (17) connected to the monitoring device (10) and configured to analyze each image (11) to attribute to each of one or more regions present in the image (11) a respective characteristic parameter. The analyzing unit (17) is configured to attribute to each region (19) an expected characteristic parameter dependent on the predetermined cutting pattern and to compare for each region (19) whether the characteristic parameter corresponds to the expected characteristic parameter.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] MACHINE AND METHOD FOR LASER PROCESSING

[0002] Cross-Reference to Related Applications

[0003] This Patent Application claims priority from Italian Patent Application No . 102024000021858 filed on October 2 , 2024 , the entire disclosure of which is incorporated herein by reference .

[0004] Technical Field

[0005] The present invention relates to a machine for the laser processing of metal workpieces selected from the group of metal sheets and / or elongated metal elements . In particular, the present invention relates to a machine for the laser processing which enables an automation of the quality control of the cut workpiece , capable of detecting the presence of material delimited by a closed cut geometry .

[0006] The present invention also relates to a method for laser processing of metal workpieces selected from the group of metal sheets and / or elongated metal elements . In particular, the present invention relates to a method for laser processing of metal sheets and / or elongated metal elements which enables an automation of the quality control of the cut workpiece , capable of detecting the presence of material delimited by a closed cut geometry .

[0007] Background

[0008] Machines for laser processing of metal workpieces to produce cut workpieces are well known . For example , a typical machine for the laser processing comprises a laser device adapted to direct a laser beam onto a workpiece , a movement unit for inducing a relative movement between the laser beam and the workpiece and a control unit configured to control the laser device and the handling unit to cut the workpiece according to a predetermined cutting pattern . By workpiece is meant any planar, deformed, tubular or more generally elongated element ( therefore also including L, C, U, H, I - shaped, etc . , beams ) , in which laser processing, which also includes laser cutting, is to be carried out , in order to obtain a cut workpiece .

[0009] The predetermined cutting pattern associates a respective shape to be cut with each of a plurality of zones of the workpiece , so as to obtain a cut workpiece having a plurality of zones , each provided with a respective cut shape . In other words , the predetermined cutting pattern defines the desired cut workpiece .

[0010] Furthermore , each shape to be cut is expressed in terms of a cutting profile and / or in a speci fic shape of a portion of the workpiece devoid of any material .

[0011] In use , the laser beam removes material from the workpiece according to the respective shapes to be cut as defined in the respective predetermined cutting pattern . For example , not all parts are to be removed by the laser beam but only those parts that , after the cut geometry is closed, cause the detachment of a respective piece of scrap, scrap being understood as the metal part contained within a completely cut closed geometry : for example , a circular geometry results in a piece of scrap material of a circular shape , or a geometry that follows a complete section of a tube results in a piece of scrap of the shape of the tube section itsel f .

[0012] At the end of a laser cutting along a closed geometry, the relative piece of scrap usually detaches and falls .

[0013] In the case of an incomplete cut , the piece of scrap might not fall but remain in contact with the workpiece , creating the need of a reworking of the cut workpiece .

[0014] Or a piece of scrap, even i f the closed geometry is completely cut , might remain in place and not separate from the workpiece , with the possibility of tilting with respect to the cutting surface , thus generating a potential risk of collision with the laser device .

[0015] To prevent cut workpieces from having undetached or tilted pieces of scrap, a veri fication is required, usually visual , by a technical operator, who determines the potential presence of undetached or tilted pieces of scrap .

[0016] Machines for laser processing and the respective methods for laser processing make it possible to obtain very satis factory results .

[0017] However, in the field there is a need for a further improvement of the machines for laser processing and / or of the methods for laser processing, in particular to veri fy that the cutting process is completed in the desired manner, and more speci fically that the piece of scrap defined by a laser cutting along a closed geometry is subsequently detached and / or does not obstruct subsequent operations .

[0018] Summary

[0019] The obj ect of the present invention is to provide an improved machine for laser processing, in particular a machine for laser processing which enables an automation of the quality control of the cut workpiece .

[0020] More particularly, the obj ect of the present invention is to provide a machine for laser processing capable of detecting the presence of material in the zone delimited by a closed cut geometry .

[0021] Furthermore , the obj ect of the present invention is to provide an improved method for laser processing, in particular a method which enables an automation of the quality control of the cut workpiece .

[0022] More particularly, the obj ect of the present invention is to provide a method capable of detecting the presence of material in the zone delimited by a closed cut geometry .

[0023] The aforesaid obj ects are achieved by the present invention, as it relates to a machine for laser processing and a method for laser processing as defined in the respective independent claims . Alternative preferred embodiments are protected in the respective dependent claims .

[0024] Brief Description of the Figures

[0025] For a better understanding of the present invention, a preferred embodiment thereof is described hereinafter, purely by way of a non-limiting example and with reference to the accompanying drawings , in which :

[0026] - Figure 1 illustrates in a schematic and partial manner a machine for laser processing according to the present invention;

[0027] - Figures 2a, 3a, 4a and 5a schematically illustrate respective images acquired by a monitoring device of the machine for laser processing of Figure 1 during a laser cutting method;

[0028] - Figures 2b, 3b, 4b and 5b schematically illustrate the respective image processing of the images of figures 2a, 3a, 4a and 5a ; and

[0029] Figures 6a and 6b schematically illustrate the operation of a neural network trained for image processing .

[0030] Description of Embodiments

[0031] In Figure 1 , 1 generically indicates , as a whole, a machine for laser processing configured to perform a laser cutting of a metal workpiece 2 to obtain a cut workpiece .

[0032] By workpiece is meant any planar, deformed, tubular or more generally elongated element ( therefore also including L, C, U, H, I -shaped, etc . , beams ) , in which laser processing, which also includes laser cutting, is to be carried out , in order to obtain a cut workpiece .

[0033] More speci fically, workpiece 2 can be selected from the group of a planar element ( also deformed) , for example a planar sheet or a deformed sheet , or an elongated element , for example a tube or a bar .

[0034] Advantageously, workpiece 2 can be of and / or comprise a metallic material .

[0035] Preferably, workpiece 2 is selected from the group of metal sheets and elongated metal elements , such as tubes or bars .

[0036] In particular, machine for laser processing 1 can be configured to perform a laser cutting, in particular of portions of workpiece 2 with thicknesses of less than ca . 20 mm for tubular elements and ca . 60 mm for planar sheets .

[0037] In more detail , machine for laser processing 1 comprises :

[0038] - a laser device 4 configured to direct a laser beam 5 onto workpiece 2 ;

[0039] - a movement unit 6 for inducing a relative movement between workpiece 2 and laser beam 5 ; and

[0040] - a control unit 7 configured to control the operation of machine for laser processing 1 itsel f .

[0041] In more detail , control unit 7 can be configured to control laser device 4 and movement unit 6 to cut workpiece

[0042] 2 according to a predetermined cutting pattern .

[0043] In more detail , the predetermined cutting pattern can associate a respective shape to be cut with each of a plurality of zones 8 of workpiece 2 , and in particular to obtain a cut workpiece having a plurality of zones 8 , each provided with a respective cut shape .

[0044] In other words , the predetermined cutting pattern describes how the plurality of zones 8 of workpiece 2 are to be cut to obtain the desired cut workpiece .

[0045] In more detail , each cutting shape defines a cutting profile ( i . e . a cutting outline ) .

[0046] In particular, each cutting profile defines a cutting edge .

[0047] Furthermore , at least some cutting profiles define a closed geometry which, in turn, defines a closed cut geometry following the laser cutting process .

[0048] In particular, laser device 4 can be configured to cut each zone 8 along the respective cutting profile by means of laser beam 5 .

[0049] In particular, each cutting profile defines the part of the respective zone 8 that is to be removed by laser beam 5 .

[0050] In further detail , each shape to be cut also defines a speci fic shape of a portion of workpiece 2 devoid of any material , for example a hole .

[0051] In particular, each portion devoid of any material is delimited by the respective cutting profile , in particular by the respective cutting edge .

[0052] In further detail , the cutting profile delimits the part of the respective zone 8 that is removed during the execution of the laser cutting which in turn can define a portion of workpiece 2 that is detached from the rest of workpiece 2 . This portion is a piece of scrap of the respective zone 8 that must separate from the rest of workpiece 2 , thus leaving the respective portion devoid of any material .

[0053] Following the execution of the laser cutting of a respective zone 8 , in the case of a correct execution, a cut shape is present, characteri zed by the respective cutting profile and the presence of the respective portion devoid of any material ( i . e . , the piece of scrap is not present in the portion devoid of any material ) .

[0054] Furthermore , following the execution of the laser cutting of a respective zone 8 , in the case of a correct execution, a closed cut geometry is present which delimits an area devoid of any material .

[0055] Furthermore , control unit 7 can be configured to control laser device 4 so that laser beam 5 cuts , in use , workpiece 2 at the respective zone 8 according to the cutting profile . Following the execution of the laser cutting at the respective zone 8 , a piece of scrap is present which, in the case of a correct cut , should detach from the rest of workpiece 2 and should leave free the portion devoid of any material .

[0056] For example , a shape to be cut could be a rectangular area, a circular area, an area of generic shape or a trim cut that defines the start and end of the workpiece ( see also Figures 2a, 3a, 4a and 5a ) . Obviously, the shapes can also be more complex . In the case of a rectangular area, the respective cutting profile is defined by two pairs of parallel lines perpendicular to each other . After the execution of the laser cutting, at the respective zone 8 a rectangularly shaped piece of scrap detaches , leaving free a rectangularly shaped portion devoid of any material .

[0057] Possibly, the predetermined cutting pattern can be saved in control unit 7 in the form of a two-dimensional and / or three-dimensional drawing . For example , the two- dimensional and / or three-dimensional drawing can represent the desired cut workpiece which is nothing other than an overlay of workpiece 2 and zones 8 with the respective shapes to be cut .

[0058] More speci fically, the two-dimensional and / or three- dimensional drawing can be based on software for Computer- Aided Design, CAD and / or software for Computer-Aided Manufacturing, CAM .

[0059] Control unit 7 can comprise a memory .

[0060] Preferably but not necessarily, the memory can be configured to contain a plurality of cutting patterns , in particular each associated with a respective type of workpiece 2 .

[0061] In more detail , laser device 4 can comprise an emission source of laser beam 5 operatively connected to control unit 7 and configured to emit laser beam 5 .

[0062] Furthermore , laser device 4 can comprise an optical assembly configured to define the optical path of laser beam 5 .

[0063] Advantageously, the optical assembly can be disposed in a housing 9 of laser device 4 .

[0064] Movement unit 6 can be configured to move housing 9 and / or components of the optical assembly, in order to move laser beam 5 relative to workpiece 2 .

[0065] Alternatively or in addition, movement unit 6 can be configured to modi fy, in use , a respective position of workpiece 2 relative to a reference point and / or to angularly move workpiece 2 around an axis of rotation A.

[0066] In particular, the movement relative to the reference point can be a movement along a linear axis or the movement can be described by linear movements , each along a respective axis di f ferent from the others . That is , the movement relative to the reference point is a movement in Cartesian space and the reference point can be defined as the zero point .

[0067] More speci fically, control unit 7 can be configured to control movement unit 6 so as to coordinate the movement of laser beam 5 and / or of workpiece 2 to perform the laser cutting according to the cutting pattern; for example , control unit 7 controls laser beam 5 and workpiece 2 by means of movement unit 6 to cut the respective cutting profiles which then determine the cut shapes .

[0068] Preferably, machine for laser processing 1 can also comprise :

[0069] - a device for generating a pressuri zed gas , for example nitrogen, oxygen or compressed air, operatively connected to control unit 7 and configured to direct a j et of gas onto the cutting zone in order to remove molten material from the laser cutting process from workpiece 2 itsel f .

[0070] Machine for laser processing 1 can also comprise a suction unit configured to remove fumes and / or by-products of the laser cutting and / or small-si zed pieces of scrap .

[0071] Control unit 7 can also be configured to control process parameters of machine for laser processing 1 , for example an intensity of laser beam 5 and / or a frequency and / or a duty cycle of the pulsed mode of laser beam 5 and / or a focal position of laser beam 5 and / or a diameter of laser beam 5 and / or a determined speed of the relative movement between laser beam 5 and workpiece 2 and / or the gas j et and / or a gas pressure of the gas j et and / or a position of a noz zle configured to emit a gas j et .

[0072] With particular reference to Figure 1 , machine for laser processing 1 also comprises a monitoring device 10 configured to acquire a plurality of images 11 of at least a portion of workpiece 2 during the execution of the laser cutting .

[0073] In particular, monitoring device 10 can be configured to acquire a sequence of images 11 in which each image 11 corresponds to a di f ferent point in time with respect to the other images .

[0074] In the illustrated, non-limiting example , during the execution of the laser cutting monitoring device 10 is fixed in space .

[0075] In consideration of the movement of workpiece 2 relative to the reference point and / or of the angular movement of workpiece 2 , it is to be noted that during the acquisition of images 11 , the relative position of workpiece 2 in images 11 also changes .

[0076] Alternatively, monitoring device 10 can be movable in space , and in particular it can move during the execution of the laser cutting . For example, monitoring device 10 can be integral with housing 9 .

[0077] In more detail , monitoring device 10 can comprise an image acquisition device 15 , for example a camera, a photo camera or a video camera, for example of the CCD or CMOS type , configured to acquire images 11 .

[0078] More speci fically, acquisition device 15 can be configured to acquire images 11 continuously, in such a way as to obtain a time sequence of images 11 .

[0079] Monitoring device 10 can also comprise more than one acquisition device 15 suitably spaced from each other .

[0080] Preferably but not necessarily, monitoring device 10 can also comprise an illumination source 16 configured to illuminate workpiece 2 .

[0081] In particular, illumination source 16 can be configured to emit light in a range of wavelengths that does not include the process light , such as a blue illumination light in the case where the process laser has a wavelength in the infrared range , to reduce the disturbance introduced by the process light .

[0082] Furthermore , illumination source 16 can also comprise a filter .

[0083] Furthermore , illumination source 16 can be positioned to avoid the formation of unwanted shadows on the portion of workpiece 2 .

[0084] Machine for laser processing 1 can also comprise a protective housing 18 delimiting a working space in which workpiece 2 is cut . In particular, the protective housing is configured to ensure that during the cutting process the laser radiation cannot escape outs ide of protective housing 18 .

[0085] Preferably but not necessarily, monitoring device 10 , more speci fically acquisition device 15 , and even more speci fically illumination source 16 , can be disposed in the working space .

[0086] Machine for laser processing 1 also comprises an analyzing unit 17 operatively connected to monitoring device 10 and configured to analyze each image 11 to attribute a respective characteristic parameter to each of one or more regions 19 present in image 11 .

[0087] Analyzing unit 17 is also configured to attribute to each region 19 an expected characteristic parameter dependent on the predetermined cutting pattern and to compare for each region 19 whether the characteristic parameter corresponds to the expected characteristic parameter .

[0088] In particular, each region 19 of each image 11 is associated with a respective zone 8 of workpiece 2 .

[0089] In more detail , analyzing unit 17 is configured to compare each image 11 with a respective expected image , which expected image depends on the predetermined cutting pattern and which expected image provides for each region 19 the respective expected characteristic parameter .

[0090] Advantageously, analyzing unit 17 is configured to determine , based on the comparison of each image 11 with the respective expected image , whether for each region 19 the characteristic parameter corresponds to the expected characteristic parameter .

[0091] Preferably but not necessarily, the expected image depends on the respective cutting pattern and / or the respective two-dimensional and / or three-dimensional drawing and / or is obtained from the respective two-dimensional and / or three-dimensional drawing .

[0092] Furthermore , each expected image can depend on the position in space and / or the orientation of workpiece 2 associated with the respective image 11 .

[0093] In addition or alternatively, each expected image can depend on the respective relative position of workpiece 2 relative to the reference point and / or on the angular position of workpiece 2 relative to the axis of rotation A and / or on the relative position of workpiece 2 relative to monitoring device 10 , in particular the respective acquisition device 15 .

[0094] More speci fically, each expected image may have been determined (previously) as a function of the position in space and / or the orientation of workpiece 2 associated with the respective image 11 .

[0095] Even more speci f ically, each expected image may have been determined (previously) as a function of the relative position of workpiece 2 at the respective point in time of the cutting process and relative to the reference point and / or of the angular position of workpiece 2 relative to the axis of rotation A and / or of the relative position of workpiece 2 relative to monitoring device 10 , in particular the respective acquisition device 15 .

[0096] In particular, the expected images can be saved in the memory .

[0097] Alternatively, analyzing unit 17 can be configured to calculate , during the execution of the laser cutting and for each image 11 , the respective expected image .

[0098] In the speci fic illustrated case , the position of monitoring device 10 is fixed and the representation in image 11 of the various zones 8 depends on the position in space and the orientation of workpiece 2 . For example , each image 11 represents portions of workpiece 2 depending on the position in space and the orientation of workpiece 2 .

[0099] In the examples of Figures 2a, 3a, 4a and 5a, for reasons of simplicity, we have chosen to indicate the progress of the laser cutting process along the linear movement only, highlighting only di f ferent pos itions of workpiece 2 along axis X .

[0100] Furthermore , we have chosen the example of a workpiece 2 with a rectangular cross-section to make the example more understandable . However, the concepts described in the present application also apply to workpieces 2 with di f ferent cross-sections such as circular, rhombic cross-sections , open sections such as H, L, I , T , C beams , flat bars , etc .

[0101] More speci fically, analyzing unit 17 is configured to analyze for each zone 8 a plurality of images 11 which contain one or more respective regions 19 , and in such a way as to evaluate in each of the images 11 whether, for each region 19 , the respective characteristic parameter corresponds to the respective expected characteristic parameter .

[0102] It is to be noted that in the case of multiple regions 19 present in an image 11 , each region 19 is associated with a di f ferent respective zone 8 .

[0103] Advantageously, analyzing unit 17 is configured to determine whether the respective cut shape of each zone 8 corresponds to the respective shape to be cut depending on the comparison between one or more respective characteristic parameters and the respective expected characteristic parameters .

[0104] Preferably but not necessarily, analyzing unit 17 is configured to determine the correspondence between the shape to be cut and the respective cut shape of each zone 8 i f the respective characteristic parameter corresponds to the expected characteristic parameter in the respective regions 19 of more than one ( analyzed) image 11 , typically at least three successive images 11 .

[0105] In further detail , the number of images 11 used for each zone 8 is determined by whether the analysis provided for a number of images 11 is suf ficient to determine whether the respective cut shape corresponds ( or not ) to the respective shape to be cut . I f the number of images 11 analyzed up to a certain point has not yet permitted an assured determination (within a tolerance defined by the system) , the analysis is continued on one or more further images 11 . The analysis for a speci fic zone 8 stops only when there is an assured determination that the cut shape corresponds to the shape to be cut . In the speci fic case , analyzing unit 17 waits until the respective characteristic parameter of the respective region 19 associated with zone 8 corresponds to the expected characteristic parameter in three images 11 .

[0106] It is to be noted that as long as the analysis returns a result of non-correspondence between the cut shape and the shape to be cut , the analysis itsel f is not interrupted for a speci fic zone 8 , as it may achieve the shape to be cut after a certain time .

[0107] In particular, until analyzing unit 17 determines the correspondence between the characteristic parameter and the expected characteristic parameter of the respective regions 19 associated with zone 8 in at least three images 11 , analyzing unit 17 continues the analysis of regions 19 related to zone 8 .

[0108] It is to be noted that a piece of scrap might only fall after a certain time , for example after a rotation of the workpiece , and therefore the analysis continues until the assured determination of the falling of the piece of scrap .

[0109] It is also to be noted that the time elapsed between each image 11 of the respective sequence of images 11 relating to each zone 8 does not necessarily have to be constant . For example , due to an angular movement of workpiece 2 around the axis of rotation A, it may result that in a series of images 11 a speci fic zone 8 is not reproduced, but only after further angular movements it is visible again and therefore reproduced in images 11 . Furthermore , it is common that images 11 can be used for analysis related to more than one zone 8 .

[0110] It is to be noted that the order of completion o f the analysis of the zones 8 need not necessarily correspond to the order of execution of the respective cut ; that is to say, it may happen that analyzing unit 17 determines for a first zone 8 that the respective cut shape corresponds to the respective shape to be cut after analyzing unit 17 determines that the respective cut shape of a second zone 8 corresponds to the shape to be cut , even i f the first zone 8 was cut before the second zone 8 .

[0111] In particular, the order of completion of the analysis of the various zones 8 depends on the information that analyzing unit 17 can extract from the acquired images 11 .

[0112] Advantageously, analyzing unit 17 can be configured to attribute to each region 19 of each image 11 a first characteristic parameter or a second characteristic parameter di f ferent from the first characteristic parameter .

[0113] In particular, analyzing unit 17 is configured to selectively determine for each region 19 and in each image 11 whether the respective region 19 has material in the respective portion of workpiece 2 or whether the respective region 19 is devoid of any material .

[0114] Furthermore , analyzing unit 17 is also configured to attribute the first characteristic parameter based on the determination that the respective region 19 has material and to attribute the second characteristic parameter based on the determination that the respective region 19 is devoid of any material . In other words , the first quality parameter indicates the presence of material and the second quality parameter indicates the absence of material . In further detail, analyzing unit 17 can be configured to analyze each image 11 to attribute to each region 19 the respective characteristic parameter by using an artificial intelligence algorithm. More specifically, the artificial intelligence algorithm uses a trained neural network to classify regions 19 in images 11 into a first class (e.g. presence of material) or into a second class (e.g. absence of material) .

[0115] According to some preferred embodiments, the first class corresponds to the first characteristic parameter and the second class corresponds to the second characteristic parameter .

[0116] In more detail, the trained neural network is a convolutional neural network. In particular, the trained neural network can use a model of the YOLO type.

[0117] Preferably, the trained neural network can be configured to segment images 11.

[0118] Preferably, the trained neural network is configured to attribute to each object identified in each image 11 a respective attributed class. In the specific case, the trained neural network is configured to attribute a first attributed class, a second attributed class and a third attributed class. In the specific case, the first attributed class corresponds to a surface (presence of material) of workpiece 2 identified in the respective image 11, the second attributed class corresponds to a part devoid of any material (i.e. a hole) and the third attributed class corresponds to a background identified in image 11.

[0119] Furthermore, analyzing unit 17 is configured to compare the analysis of the trained neural network with the expected result as defined by the predetermined cutting pattern and to attribute the respective characteristic parameter.

[0120] The trained neural network has been trained by means of a training dataset obtained from laser cutting processes of workpieces 2 of various types (such as different shapes, cross-sections, materials, dimensions, etc.) .

[0121] During the laser cutting process of such workpieces 2, images were acquired and / or obtained. In such images, labels were applied that correspond to the first attributed class, the second attributed class and the third attributed class.

[0122] The labelled images thus obtained define the training dataset and have furthermore been subdivided into training images, test images and validation images.

[0123] Figures 6a and 6b schematically illustrate the operation of the trained neural network. The image as schematically represented in Figure 6a is analyzed by the trained neural network. The trained neural network attributes the three associated classes as indicated in Figure 6b. In particular, the dark grey color indicates the first associated class, i.e. a surface of workpiece 2, the white color indicates the second attributed class, i.e. a part devoid of any material, and the light grey color indicates the third attributed class, i.e. the background identified in image 11.

[0124] Advantageously, analyzing unit 17 is configured, in particular also by using the artificial intelligence algorithm, to require for each zone 8 that the respective associated regions 19 be classified in the second class in a plurality of images 11, in particular in at least three images 11.

[0125] It is to be noted that in a simplified embodiment analyzing unit 17 can use a single determined characteristic parameter relating to the absence of material : for each zone 18 , the single determined characteristic parameter is associated with the respective regions 19 only when analyzing unit 17 detects the absence of material in a plurality of images 11 , in particular at least three images 11 .

[0126] Preferably but not necessarily, machine for laser processing 1 can comprise a display unit configured to display images 11 and / or workpiece 2 with an indication of the zones 8 for which the respective cut shape corresponds to the respective shape to be cut ( obtaining respective processed images 11 ' ) , and in particular in a di f ferent way for those for which the respective cut shape does not correspond to the respective shape to be cut .

[0127] In the case of Figures 2a, 3a, 4a and 5a and 2b, 3b, 4b and 5b, the display of the advancement of workpiece 2 , i . e . a change in the positioning relative to the reference point , is schematically indicated with reference to images 11 . In the speci fic example , the advancement is along a translation axis X around which the axis of rotation A rotates . For reasons of simplicity, we have omitted images 11 in which workpiece 2 is oriented in an angular position di f ferent from that visible in the attached figures .

[0128] Then, in Figures 2b, 3b, 4b and 5b, processed images 11 ' are illustrated . In particular, in Figures 2b, 3b, 4b and 5b, regions 19 for which the cut shape corresponds to the shape to be cut ( fallen piece of scrap ) are indicated with a hatching, and those for which the cut shape does not correspond to the shape to be cut (no certainty of the fallen piece of scrap ) are indicated only in white . In the speci fic case , the information contained in Figures 2b, 3b, 4b and 5b indicates with the hatching the zones 8 for which the respective piece of scrap has fallen . Figures 2b, 3b, 4b and 5b illustrate processed images 11 ' , which di f fer from the respective images 11 by the presence of the eventual hatching . In this way, the operator can obtain the information visually .

[0129] It can also be seen from figures 2b, 3b, 4b and 5b that the information is updated by analyzing unit 17 as the overall process progresses .

[0130] In use , a method for laser processing is performed, in particular by machine for laser processing 1 , to cut workpiece 2 .

[0131] The method comprises the steps of : a ) directing laser beam 5 onto workpiece 2 ; b ) inducing a relative movement between workpiece 2 and laser beam 5 ; c ) controlling, in particular in a coordinated manner, laser beam 5 and the movement of workpiece 2 according to the respective predetermined cutting pattern that associates with each zone 8 the respective shape to be cut , so as to obtain the cut workpiece having the zones 8 , each provided with the respective cut shape ; d) acquiring a plurality of images 11 of at least a portion of workpiece 2 ; e ) analyzing each image 11 , during which step of analyzing a respective characteristic parameter is attributed to each of one or more regions 19 present in the image 11 ; f ) attributing to each region 19 an expected characteristic parameter dependent on the predetermined cutting pattern; and g) comparing for each region 19 whether the characteristic parameter corresponds to the expected characteristic parameter .

[0132] In more detail , during the step of directing a ) , laser device 4 directs laser beam 5 onto workpiece 2 .

[0133] Advantageously, during the step of inducing b ) , workpiece 2 is moved, in particular by movement unit 6 , relative to the reference point in rotation with respect to the axis of rotation A and / or linearly along axis X , in order to obtain a relative movement between workpiece 2 and laser beam 5 .

[0134] In more detail , during the step of controlling c ) , control unit 7 controls , in particular in a coordinated manner, laser device 4 and movement unit 6 to perform the cutting of zones 8 according to the respective cutting pattern .

[0135] In further detail , during the step of acquiring, images 11 are acquired by means of acquisition device 15 .

[0136] Preferably, during the step of acquiring, illumination source 16 illuminates workpiece 2 .

[0137] In further detai l , during the step of comparing g) , each image 11 is compared with a respective expected image , which expected image depends on the predetermined cutting pattern and which expected image provides for each region 19 the respective expected characteristic parameter .

[0138] Furthermore , during the step of comparing g) , it is determined, based on the comparison of the respective image 11 with the respective expected image , whether for each region 19 present in image 11 the characteristic parameter corresponds to the expected characteristic parameter .

[0139] In particular, the step of analyzing e ) , the step of attributing f ) and the step of comparing g) are performed by analyzing unit 17 .

[0140] Preferably but not necessarily, during the step of analyzing e ) , a plurality of images 11 containing one or more respective regions 19 are analyzed for each zone 8 , and in such a way as to evaluate in each of the images 11 whether, for each region 19 (present in image 11 ) , the respective characteristic parameter corresponds to the respective expected characteristic parameter .

[0141] Preferably but not necessarily, during the step of comparing g) , it is determined that the respective cut shape of each zone 8 corresponds to the respective shape to be cut depending on the comparison between one or more respective characteristic parameters and the respective expected characteristic parameters .

[0142] In particular, during the step of comparing g) , it is determined that the respective cut shape of each zone 8 corresponds to the respective shape to be cut i f the respective characteristic parameter corresponds to the expected characteristic parameter in more than one , in particular in at least three , analyzed images 11 . In this way, the reliability of the obtained result is increased .

[0143] In further detail , during the step of analyzing e ) , the first characteristic parameter ( e . g . presence of material ) or a second characteristic parameter ( e . g . absence of material ) is attributed to each region of each image 11 .

[0144] In more detail , during the step of analyzing e ) , it is selectively determined for each region 19 and in each image 11 whether the respective region 19 has material or whether the respective region 19 is devoid of any material .

[0145] Furthermore , during the step of analyzing e ) , the f irst characteristic parameter is attributed based on the determination that the respective region 19 has material , and the second characteristic parameter is attributed based on the determination that the respective region 19 is devoid of any material .

[0146] Preferably but not necessarily, the method for laser processing also comprises a step of displaying h) , during which the respective image 11 is displayed with an indication of the regions 19 for which the characteristic parameter corresponds to the expected characteristic parameter .

[0147] From an examination of the characteristics of machine for laser processing 1 and of the method according to the present invention, the advantages that it makes it possible to obtain are evident .

[0148] In particular, following the execution of the process for laser processing of workpiece 2 , all the information related to the successful cut or whether inaccuracies may be present in one or more zones 8 is already available . This facilitates the quality control steps of the various cut workpieces .

[0149] Advantageously, information is obtained as to whether the respective pieces of scrap have fallen or not .

[0150] It is finally clear that modifications and variations can be made to machine for laser processing 1 and to the method described and illustrated herein which do not depart from the scope of protection defined by the claims .

[0151] For example , in the speci fic case the quality control performed concerns the presence ( or otherwise ) of pieces of scrap . Alternatively or in addition, it could also be analyzed whether the cutting profiles obtained following the cut correspond to the expected cutting profiles . Alternatively or in addition, dimensional analyses could also be performed .

Claims

CLAIMS1. A machine for the laser processing of metal workpieces (2) selected from the group of metal sheets and / or elongated metal elements to produce cut workpieces, comprising :- a laser device (4) configured to direct a laser beam (5) onto the workpiece (2) ;- a movement unit (6) for inducing a relative movement between the workpiece (2) and the laser beam (5) ; a control unit (7) configured to control in a coordinated manner the laser device (4) and the movement unit (6) to cut the workpiece (2) according to a predetermined cutting pattern that associates with each of a plurality of zones (8) of the workpiece (2) a respective shape to be cut, so as to obtain a cut workpiece having a plurality of zones (8) , each provided with a respective cut shape ;- a monitoring device (10) configured to acquire, in use, a plurality of images (11) of at least a portion of the workpiece (2) ; and- an analyzing unit (17) connected to the monitoring device (10) and configured to analyze each image (11) to attribute to each of one or more regions present in the image (11) a respective characteristic parameter; wherein each region (19) is associated with a respective zone ( 8 ) ; wherein the analyzing unit (17) is also configured to attribute to each region (19) an expected characteristic parameter dependent on the predetermined cutting pattern and to compare for each region (19) whether the characteristic parameter corresponds to the expected characteristicparameter .

2. Machine according to claim 1, wherein the analyzing unit (17) is configured to compare each image (11) with a respective expected image, which expected image depends on the predetermined cutting pattern and which expected image provides for each region (19) the respective expected characteristic parameter; wherein the analyzing unit (17) is configured to determine, based on the comparison of each image (11) with the respective expected image, whether for each region (19) the characteristic parameter corresponds to the expected characteristic parameter.

3. Machine according to claim 1 or 2, wherein the analyzing unit (17) is configured to analyze for each zone (8) a plurality of images (11) which contain one or more respective regions (19) , and in such a way as to evaluate in each of the images (11) whether, for each region (19) , the respective characteristic parameter corresponds to the respective expected characteristic parameter; wherein the analyzing unit (17) is configured to determine whether the respective cut shape of each zone (8) corresponds to the respective shape to be cut if the respective characteristic parameter corresponds to the expected characteristic parameter in the respective regions (19) of more than one analyzed image (11) .

4. Machine according to claim 3, wherein each shape to be cut is expressed in terms of a cutting profile and / or in a specific shape of a portion of the workpiece (2) devoid of any material .

5. Machine according to any one of the preceding claims, wherein the analyzing unit (17) is configured to attributeto each region (19) of each image (11) a first characteristic parameter or a second characteristic parameter different from the first characteristic parameter.

6. Machine according to claim 5, wherein the analyzing unit (17) is configured to selectively determine for each region (19) and in each image (11) whether the respective region (19) has material in the respective portion of the workpiece (2) or whether the respective region (19) is devoid of any material; wherein the analyzing unit (17) is also configured to attribute the first characteristic parameter based on the determination that the respective region (19) has material and to attribute the second characteristic parameter based on the determination that the respective region (19) is devoid of any material; or wherein the analyzing unit (17) is configured to attribute to each region (19) and in each image (11) a determined characteristic parameter based on the determination that the respective region (19) is devoid of any material .

7. Machine according to any one of the preceding claims, wherein the analyzing unit (17) is configured to analyze each image (11) to attribute to each region (19) the respective characteristic parameter by using a trained neural network.

8. Machine according to any one of the preceding claims, wherein the movement unit (6) is configured to modify, in use, a respective relative position of the workpiece (2) relative to a reference point and / or to angularly move the workpiece (2) around an axis of rotation (A) to control a relative angular position; wherein each image (11) depends on the relative positionand / or the relative angular position of the workpiece (2) .

9. Machine according to any one of the preceding claims, wherein the monitoring device (10) comprises an image acquisition device (15) and an illumination source (16) .

10. Machine according to any one of the preceding claims, further comprising a display unit configured to display an image (11) of the workpiece (2) and / or the workpiece (2) with an indication of the regions for which the characteristic parameter corresponds to the expected characteristic parameter.

11. Method for the laser processing of metal workpieces (2) selected from the group of metal sheets and / or elongated metal elements to produce cut workpieces, comprising the steps of:- directing a laser beam (5) onto the workpiece (2) ;- inducing a relative movement between the workpiece (2) and the laser beam (5) ;- controlling in a coordinated manner the laser beam (5) and the movement of the workpiece (2) depending on a predetermined cutting pattern that associates with each of a plurality of zones (8) of the workpiece (2) a respective shape to be cut, so as to obtain a cut workpiece having a plurality of zones (8) , each provided with a respective cut shape ;- acquiring a plurality of images (11) of at least a portion of the workpiece (2) ; and analyzing each image (11) , during which step of analyzing a respective characteristic parameter is attributed to each of one or more regions present in the image (11) ; attributing to each region (19) an expectedcharacteristic parameter dependent on the predetermined cutting pattern; and comparing for each region (19) whether the characteristic parameter corresponds to the expected characteristic parameter; wherein each region (19) is associated with a respective zone ( 8 ) .

12. Method according to claim 11, wherein during the step of comparing, each image (11) is compared with a respective expected image, which expected image depends on the predetermined cutting pattern and which expected image provides for each region (19) the respective expected characteristic parameter; wherein during the step of comparing it is determined, based on the comparison of the image (11) with the respective expected image, whether for each region (19) the characteristic parameter corresponds to the expected characteristic parameter.

13. Method according to claim 11 or 12, wherein the predetermined cutting pattern associates with each of a plurality of zones (8) of the workpiece (2) a respective shape to be cut, so as to obtain a cut workpiece having a plurality of zones (8) , each provided with a respective cut shape ; wherein during the step of analyzing, a plurality of images (11) containing one or more respective regions (19) are analyzed for each zone (8) , and in such a way so as to evaluate in each of the images (11) whether, for each region (19) , the respective characteristic parameter corresponds to the respective expected characteristic parameter; wherein during the step of comparing, it is determinedthat the respective cut shape of each zone (8) corresponds to the respective shape to be cut if the respective characteristic parameter corresponds to the expected characteristic parameter of more than one analyzed image (11) •14. Method according to any one of claims 11 to 13, wherein during the step of analyzing, a first characteristic parameter or a second characteristic parameter different from the first characteristic parameter is attributed to each region (19) of each image (11) .

15. Method according to claim 14, wherein during the step of analyzing, it is selectively determined for each region (19) and in each image (11) whether the respective region (19) has material of the respective portion of the workpiece (2) or whether the respective region (19) is devoid of any material; wherein during the step of analyzing the first characteristic parameter is attributed based on the determination that the respective region (19) has material and the second characteristic parameter is attributed based on the determination that the respective region (19) is devoid of any material; or wherein during the step of analyzing a determined characteristic parameter is attributed to each region (19) and in each image (11) based on the determination that the respective region (19) is devoid of any material.

Citation Information

Patent Citations

  • Laser processing machine and procedure

    IT202400021858A1

  • Methods for providing panel planning geometry data and laser flatbed machines for cutting out workpieces

    US20210308787A1

  • Edge Detection for Computer Numerically Controlled Fabrication

    US20230029940A1

  • Method and apparatus for automated quality control for cutting machines of flexible material parts

    US20230330780A1