Apparatus and method for determining the estimated number of removal attempts for successful automatic removal of a part cut from a sheet metal from the sheet metal
By determining the estimated number of ejection attempts using part geometry and tool parameters, and leveraging AI, the method optimizes the ejection process, reducing failures and enhancing manufacturing efficiency.
Patent Information
- Application Number
- JP2024518445
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2021-09-23
- Filing Date
- 2022-09-07
- Publication Date
- 2024-11-13
- Estimated Expiration
- 2042-09-07
Smart Images

Figure 2024541801000001_ABST
Abstract
Description
[Technical field]
[0001] The invention relates to a method for determining the estimated number of removal attempts for successful automatic removal of a part cut from a sheet metal from said sheet metal.
[0002] Furthermore, the invention relates to a device for carrying out the method according to the invention.
[0003] From DE 10 200 03 133 A1 a method is known in which a part of a workpiece is removed from a remaining workpiece by multiple removal attempts.
[0004] From US Pat. No. 5,399,543 a method for generating a take-out forecast is known. [Prior art documents] [Patent documents]
[0005] [Patent Document 1] German Patent Application Publication No. 102018215738A1 [Patent Document 2] German Patent Application Publication No. 102018208126A1 Summary of the Invention [Problem to be solved by the invention]
[0006] The object of the present invention is to provide a method which allows a better prediction of the unloading process, which in turn allows a better control of the manufacturing process of which it is a part. [Means for solving the problem]
[0007] The problem is solved by a method for determining an estimated number of removal attempts for a successful automatic removal of a part cut from a sheet metal, the part being removed from the sheet metal by an automatic removal tool, during which multiple removal attempts may be made, the method comprising the following steps: a. reading the geometry of a part; b. reading parameters of an automatic extraction tool; c. determining an estimated number of ejection attempts based on the part geometry and the automated ejection tool parameters. The geometry of the part may for example originate from design data relating to the cutting process of the part. Such design data may for example originate from a CAD program. The design data is usually available in a so-called Manufacturing Execution System (MES) and can be retrieved from the MES. Alternatively, the data can also be determined from a camera image of the cut sheet metal using image processing.
[0008] The ejection tool may, for example, include a number of ejection elements and holding elements. The parameters of the ejection tool may, for example, be the type of the ejection tool, the number and / or arrangement of the ejection elements, the number and / or arrangement of the holding elements, the ejection speed, the ejection acceleration, the wear value, or the presence of hold-downs. The simultaneous use of multiple parameters may improve the method. The estimated number of ejection attempts is preferably determined from experience with parts of the same or similar shape and with ejection tools having the same or similar parameters.
[0009] Preferably, in the determining step, the possible number of times of the removal attempts is determined using a related probability value. By determining the possible number of times of the removal attempts using a related probability value, the certainty of the determination can be evaluated. If the possible number of removal attempts is small due to a high probability value, the certainty of the determination is high. If the possible number of removal attempts is large due to a similar probability value, the certainty of the determination is low.
[0010] Preferably, the step of determining an estimated number of retrieval attempts includes determining a probability of failure of automatic retrieval, where failure of automatic retrieval may be assumed, for example, if a predefined maximum number of retrieval attempts is exceeded.
[0011] Preferably, the time for successful automatic removal is determined from the estimated number of removal attempts. By determining the estimated time for automatic removal of the parts, the estimated duration for sorting the sheets and / or the working duration of the automatic removal tool can be determined, which can significantly improve production planning and / or operation planning.
[0012] Preferably, an estimated number of pick attempts is determined for the multiple parts. Most preferably, a pick order is determined for the multiple parts based on an estimated number of repeated attempts.
[0013] Preferably, the pick order includes parts with a lower estimated number of iterations before parts with a higher estimated number of iterations. Typically, the higher the estimated number of iterations, the higher the risk of automatic pick failure. The proposed pick order ensures that parts with a lower probability of failure are picked first.
[0014] Preferably, the parts are cut from the sheet metal by a manufacturing machine, whereby at least one parameter of the manufacturing machine is read and taken into account in the step of determining the estimated number of removal attempts. The manufacturing parameters are the type of material, the thickness of the material, the type of manufacturing machine, the cutting parameters or the wear values. The cutting parameters are for example the cutting width, the feed rate, the nozzle type of the laser cutting head or the distance between the nozzle of the laser cutting head and the sheet metal. The manufacturing parameters primarily determine the estimated number of removal attempts. By taking into account one or more manufacturing parameters, the determination of the estimated number of removal attempts can be improved.
[0015] If a manufacturing machine is used in different production shifts, and different reaction times of the machine operator of the manufacturing machine are known for the different production shifts, parts with a high estimated number of removal attempts are integrated into a production shift that is assigned a short reaction time of the machine operator. By such integration of parts into production shifts, the production process and thus the production planning can be improved, since a short reaction time of the machine operator is guaranteed for parts with a high estimated number of removal attempts and therefore a high probability of automatic removal failure.
[0016] Preferably, an artificial intelligence, in particular a trained neural network, is used in the step of determining the estimated number of removal attempts. The artificial intelligence can learn to determine the estimated number of removal attempts based on past removal attempts. Furthermore, the artificial intelligence can be further trained with each removed part, which allows the determination of the estimated number of removal attempts to be improved.
[0017] The invention also includes a computer program product for carrying out, by a computer, the method according to the invention.
[0018] The invention also includes an apparatus comprising a computing unit and a memory, wherein a computer program having instructions for carrying out the method according to the invention is stored in the memory, and wherein the computing unit is provided and configured to carry out the method.
[0019] The following description of preferred embodiments, together with the drawings, serves to explain the invention in more detail. [Brief description of the drawings]
[0020] [Figure 1] FIG. 2 is a process flow diagram of the method according to the invention. [Diagram 2] FIG. 1 is a schematic diagram of a machine for laser processing having an extraction tool. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0021] In all the embodiments, identical or functionally equivalent elements are provided with the same reference numbers.
[0022] 1 shows an exemplary process sequence of the method according to the invention. In a first step 1, the necessary data is read in. The first step 1 is divided into reading in the geometry of the part 1a and reading in one or more parameters of the extraction tool 1b. The data for the two sub-steps can be read from one or more memories. If the method is only carried out for a given extraction tool, the parameters of the extraction tool can also be predefined. In the present embodiment, additional manufacturing parameters are also read in 1c. As manufacturing parameters, the type of material (in the present embodiment, special steel), the material thickness (in the present embodiment, 1 mm) and the cutting width (in the present embodiment, 1 mm) are read in. Preferably, the manufacturing parameters are read in together with the geometry of the part.
[0023] In a second step 2, an estimated number of required removal attempts is determined. In this embodiment, this is done using artificial intelligence. In this embodiment, the artificial intelligence is formed as a neural network. The neural network is trained using a large number of removal attempts for different parts. The artificial intelligence can be trained for a specific removal tool and / or manufacturing machine. In this case, the parameters for the removal tool and / or manufacturing machine are inherent in the artificial intelligence.
[0024] The first two steps are performed for multiple parts. The manufacturing parameters valid for multiple parts are preferably loaded only once. For example, the material type and thickness of the sheet metal from which the multiple parts are cut may be the same for all parts, and the shape and cutting width may differ for the parts.
[0025] In a third step 3, the result of the determination of the estimated number of necessary pick attempts is used for production planning. In this embodiment, the result is used in the first variant 3a for determining the pick sequence for a large number of parts. Since the risk of failure of the automatic pick is higher the more the estimated number of repeated attempts, parts with a smaller estimated number of repeated attempts are then included before parts with a larger estimated number of repeated attempts. Correspondingly, parts with a lower failure probability are selected first. This reduces the possibility of interrupting the automatic sorting at the start of the sorting. In a second variant 3b, the result is used for planning the production shifts. In this case, it is advantageous if different reaction times of the machine operators of the production machines are known for different production shifts. Parts with a larger estimated number of pick attempts are then included in production shifts to which a short reaction time of the machine operator is assigned. This makes it possible to limit interruptions of the sorting process due to picking failures to a short time.
[0026] Fig. 2 shows an exemplary structure of a device for implementing the method according to the invention. In a memory 10, data on the part geometry 11, the removal tool parameters 12 and the manufacturing parameters 13 are stored. A computing unit 20 loads the data 11, 12, 13 from the memory 10. From the part geometry 11, the removal tool parameters 12 and the manufacturing parameters 13, the computing unit 20 determines an estimated number of removal attempts 22. To do so, the computing unit makes use of an artificial intelligence 21. In the present embodiment, the artificial intelligence is in the form of a trained neural network. This neural network has been trained with a large number of removal attempts for parts of a large number of different geometries. The artificial intelligence determines a number of possible removal attempts with associated probability values. Furthermore, the artificial intelligence determines a failure probability 23 of the automatic removal. From the estimated number of removal attempts 22, the computing unit 20 determines an estimated time 24 for a successful automatic removal.
[0027] The computing unit 20 reads from the memory 10 information about the number of production shifts, in particular about the reaction times of the machine operators in the production shifts. Together with the estimated number of pick attempts 22 and the probability of failure of automatic pick 23, the computing unit 20 places the parts into production shifts, with parts with a high estimated number of pick attempts 22 being placed into production shifts that are assigned short reaction times of the machine operators.
[0028] For parts cut from sheet metal, the computing unit 20 plans a sorting sequence based on the estimated number of pick attempts 22 and the probability of automatic pick failure 23, with parts with a lower estimated number of pick attempts 22 being placed before parts with a higher estimated number of pick attempts 22.
[0029] In this embodiment, computing unit 20 is part of a manufacturing execution system (MES) 25. MES 25 controls manufacturing machine 30. Manufacturing machine 30 includes an extraction tool 33.
[0030] A manufacturing machine 30 cuts a part 32 from a sheet metal 31. The part 32 is then automatically removed from the sheet metal 31 by means of an removal tool 33. An exemplary manufacturing machine 30 with an removal tool 33 is known from the above-mentioned US Pat. No. 5,399,623, the disclosure of which is incorporated herein in its entirety. The number of removal attempts required during automatic removal is stored and used together with the part geometry and manufacturing parameters for improving the artificial intelligence. [Explanation of symbols]
[0031] 1a. Loading part geometry 1b Read the parameters of the automatic extraction tool 1c Loading manufacturing parameters 2. Determining the Estimated Number of Retrieval Attempts 3a Determining the order of removal 3b. Planning production shifts 10. Memory 11 Part Shape 12 Extraction tool parameters 13 Manufacturing parameters 20 Calculation Units 21 Artificial Intelligence 22 Number of retrieval attempts 23 Probability of Failure 24 Take-out time 25 MES 30 Manufacturing machinery 31 Metal thin plate 32 parts 33 Extraction tool
Claims
1. 1. A method for determining an estimated number of removal attempts (22) for successful automatic removal of a part (32) cut from a sheet metal (31) from said sheet metal (31), comprising: The part (32) is removed from the sheet metal (31) by an automatic removal tool (33); During the automated ejection of the part (32), multiple ejection attempts may be performed; The method comprises the steps of: 1a. A step (1a) of reading the shape of the part (32); 1b. reading parameters of the automatic extraction tool (33); 2. determining (2) an estimated number of ejection attempts (22) based on the geometry of the part (32) and the parameters of the automated ejection tool (33).
2. 2. The method of claim 1, wherein in said determining step (2), a plurality of possible numbers (22) of retrieval attempts are determined with associated probability values.
3. 3. A method according to claim 1 or 2, characterized in that in the step (2) of determining the estimated number of retrieval attempts (22), a probability (23) of failure of the automatic retrieval is determined.
4. Method according to any one of claims 1 to 3, characterised in that the time (24) for successful automatic ejection is determined from the estimated number (22) of ejection attempts.
5. The method according to any one of the preceding claims, characterized in that the estimated number of removal attempts (22) is determined for a number of parts (32).
6. 6. The method of claim 5, further comprising determining (3a) a pick sequence for said number of parts (32) based on said estimated number of pick attempts (22).
7. 7. The method of claim 6, characterized in that when determining (3a) the pick order, parts (32) with a lower estimated number of pick attempts (22) are installed before parts (32) with a higher estimated number of pick attempts (22).
8. said part (32) being cut from sheet metal (31) by a manufacturing machine (30); Reading (1c) at least one manufacturing parameter (13); Method according to any one of the preceding claims, characterized in that said production parameters (13) are taken into account in the step (2) of determining said estimated number of removal attempts (22).
9. 9. The method according to claim 8, characterized in that the manufacturing parameters (13) are the type of material, the thickness of the material, the type of manufacturing machine, cutting parameters or wear values.
10. Using said manufacturing machine (30) in different manufacturing shifts; different reaction times of machine operators of said manufacturing machines for different production shifts are known; 10. The method according to claim 8 or 9, characterized in that parts (32) with a high estimated number of removal attempts (22) are grouped (3b) for production shifts allocated short machine operator reaction times.
11. Method according to any one of the preceding claims, characterized in that an artificial intelligence (21), in particular a trained neural network, is used in the step (2) of determining the estimated number of retrieval attempts (22).
12. A computer program product for carrying out, by a computer, the method according to any one of claims 1 to 11.
13. An apparatus comprising a computing unit (20) and a memory (10), A computer program having instructions for carrying out the method according to any one of claims 1 to 11 is stored in said memory (10), The computing unit (20) is provided and configured to perform the method.
Citation Information
Patent Citations
Workpiece attitude detecting device, workpiece attitude detecting method, picking system, and picking method
JP2007245283A
Workpiece takeout device
JP2013052490A
Bin picking performance evaluation device and bin picking performance evaluation method
JP2014240110A
Method for handling a workpiece with a removal tool and machine for carrying out this method - Patent Application 20070122997
JP2021524385A
Component suction-attachment rate calculation system for component mounter
WO2016162974A1
