Data capturing device and machine tool

Abstract

Provided are a data capturing device capable of capturing user-stored information into a database used for creating a machining program, and a machine tool.　A data capturing device according to the present disclosure captures data into a database used for creating a machining program for controlling a machine tool when machining a workpiece. The data capturing device executes: a digitization process for acquiring target data that does not correspond to a data input rule for a database, and digitizing the acquired target data; a stored information acquisition process for acquiring user-stored information pertaining to a machine tool; and a registration process for classifying, using AI technology, information about a numerical value digitized by the digitization process and the stored information acquired by the stored information acquisition process, and registering the classified information as data in the database.

Description

Data acquisition device and machine tool

[0001] The present disclosure relates to a technique for importing data into a database used for creating a machining program.

[0002] Patent Document 1 below describes a processing device that reads image information (image data) of drawings on paper media or non-standard documents, extracts text included in the read image information, and registers it in a database server. The processing device of Patent Document 1 analyzes the morphemes in the extracted text to extract words, and extracts keywords from the extracted word group by statistical weighting processing. The processing device classifies the drawings or documents read according to the extracted keywords.

[0003] Japanese Patent Application Laid-Open No. 2019-003472

[0004] When creating a machining program such as an NC program from a database, it is necessary to register information such as a work drawing and a machining proposal form in the database. However, at the machining site of a work using a machine tool, there is information that remains only in the memory of a user such as an operator. Therefore, when creating a machining program, a problem arises in that there is a shortage of necessary past information.

[0005] The present disclosure has been made in view of the above problems, and an object thereof is to provide a data acquisition device and a machine tool capable of importing user's memory information into a database used for creating a machining program.

[0006] To solve the above problems, this specification discloses a data acquisition device for acquiring data into a database used to create a machining program for controlling a machine tool when machining a workpiece, the data acquisition device performing: acquiring target data that does not correspond to the input rules for the database data, a digitization process that digitizes the acquired target data, a storage information acquisition process that acquires user storage information related to the machine tool, and a registration process that classifies the numerical information digitized by the digitization process and the storage information acquired by the storage information acquisition process using AI technology, and registers the classified information as data in the database. It should be noted that the contents of this disclosure are not limited to implementation as a data acquisition device, but are also extremely useful when implemented as a machine tool equipped with a data acquisition device.

[0007] According to the data acquisition device and machine tool disclosed herein, numerical information obtained by quantifying target data that does not conform to input rules, as well as user-remembered information, can be classified and registered in a database using AI technology. Not only numerical information, but also remembered information can be classified and registered in the database used to create machining programs. Information such as handwritten notes that are difficult to decipher and the know-how of experts can be accumulated in the database, making it possible to build a database in which even engineers with little knowledge or experience can create machining programs with the same level of expertise as experts.

[0008] A block diagram of a machine tool according to the first embodiment. A diagram showing the flow of data acquisition processing. A diagram showing a machining proposal table. A diagram showing a machining proposal table. A diagram showing a machining proposal table. A diagram showing a data acquisition system including a data acquisition device according to the second embodiment.

[0009] (First Embodiment) Hereinafter, a first embodiment, which is an embodiment of a machine tool equipped with a data acquisition device of the present disclosure, will be described with reference to the drawings. Figure 1 shows a block diagram of the machine tool 10 of this embodiment. As shown in Figure 1, the machine tool 10 is, for example, a turret-type lathe and includes a work spindle device 11, a turret device 12, an operation panel 13, an external IF (abbreviation for interface) 14, and a control device 15.

[0010] The workpiece spindle device 11 is equipped with a gripping mechanism for holding a workpiece, such as multiple chuck jaws. The workpiece spindle device 11 grips the workpiece with the gripping mechanism and rotates the workpiece around the workpiece spindle (Z-axis). The workpiece spindle device 11 is equipped with a servo motor, encoder, etc. for rotating the workpiece, and its rotational movement is controlled by the control device 15. Note that the gripping mechanism is not limited to multiple chuck jaws, but may be other mechanisms capable of gripping a workpiece, such as a collet chuck.

[0011] The turret device 12 includes, for example, a turret (tool post) capable of mounting multiple tools, and a servo motor for rotating the turret. The turret device 12 performs machining on a workpiece held by the workpiece spindle device 11 using cutting tools (such as cutting tools or rotary tools) mounted on the turret. Based on the control of the control device 15, the turret device 12 rotates the turret and changes the cutting tools (cutting tools indexed to the working position) used for machining the workpiece. The machine tool 10 also includes a sliding device for moving the turret device 12 in the X-axis and Y-axis directions, and controls the sliding device to change the position of the turret device 12 (the cutting tools indexed to the working position).

[0012] Furthermore, the control panel 13 is equipped with, for example, a touch panel and multiple operation switches, and performs functions such as displaying information about the machine tool 10 and receiving operation instructions based on the control of the control device 15. In this embodiment, the control panel 13 is also capable of receiving instructions for acquiring target data and creating an NC program 27, which will be described later. The control device 15 comprises a numerical control device 21 and a PLC 22. The numerical control device 21 comprises a CPU 23 and a storage device 24. The storage device 24 is equipped with, for example, RAM, ROM, flash memory, HDD, etc. Note that the configuration of the storage device 24 is not limited to the above configuration, and may be configured with an SSD instead of an HDD, or with an external storage device such as a USB memory, or with a storage medium such as a DVD-RAM, or a combination of these.

[0013] Furthermore, the external IF 14 has a communication interface, such as a USB interface or a LAN interface. The control device 15 is connected to the external IF 14 and can communicate with an external device via the external IF 14. This external device is, for example, an input / output device such as a scanner or a multifunction printer. The control device 15 can receive image data of workpiece drawings and processing notes from the external device via the external IF 14. The external device can also be a processing device such as a server, management PC, or tablet. The control device 15 can acquire image data from the processing device and execute processing based on instructions from the processing device via the external IF 14. For this reason, the machine tool 10 may receive instructions to execute a scan or instructions to register to the database 29, which will be described later, from the external processing device. Alternatively, the external device may be another machine tool. In addition to the communication interface, the external IF 14 also has an audio input interface such as a microphone. The control device 15 can acquire user audio data via the external IF 14. Furthermore, the control device 15 may acquire audio data from a management PC or tablet connected to the external IF 14. In addition, the control device 15 can input and output data via a chat function using a processing device (tablet, etc.) connected via the control panel 13 or the external IF 14.

[0014] Furthermore, the machine tool 10 includes a control device 15 and a plurality of drive circuits 25 that connect each of the above-mentioned devices (work spindle device 11, turret device 12, and control panel 13). The numerical control device 21 can control each device via the drive circuits 25 by executing the NC program 27 stored in the storage device 24 with the CPU 23. The drive circuits 25 are, for example, driver circuits (servo amplifiers).

[0015] The PLC 22 is a Programmable Logic Controller. The PLC 22, for example, executes ladder programs and processes various signals sequentially using ladder circuits. These various signals include, for example, output signals that drive various elements such as lamps provided by the machine tool 10, and input signals received from elements such as limit switches. The PLC 22 is connected to the numerical control device 21 via the communication bus 26 and performs signal input and output with the numerical control device 21.

[0016] The storage device 24 also stores the acquisition program 28, the database 29, and the creation program 30. The acquisition program 28 includes a scan program and an assistant program. The scan program includes, for example, a driver program for a scanner, and is a program that controls the scanner to generate image data from work drawings, etc. The assistant program includes, for example, a program for performing voice recognition and chat functions. The control device 15 acquires various data (such as target data described later) for registration in the database 29 by executing the scan program and assistant program of the acquisition program 28 on the CPU 23. The acquisition program 28 also includes an AI program. The control device 15 classifies the acquired data (numerical information that digitizes the target data, stored information, information on components included in the imaging data, etc.) by executing the AI ​​program of the acquisition program 28 on the CPU 23 and registers it in the database 29.

[0017] Furthermore, the database 29 stores information necessary for creating the NC program 27, such as shape information 31, machining theory information 32, machining knowledge information 33, and machine tool information 34. Shape information 31 is, for example, information related to the shape of the workpiece. Machining theory information 32 is information of pre-set values, such as theoretical values ​​of machining conditions. Machining knowledge information 33 is information that constitutes know-how at the machining site, such as a machining proposal table showing the machining process, a change history of the machining proposal table, and information on machining failures. Machine tool information 34 is information about machine tool 10 or other machine tools. The creation program 30 is a program for creating the NC program 27 from the machining proposal table. Details of the import program 28, database 29, and creation program 30 will be described later.

[0018] The configuration of the machine tool 10 shown in Figure 1 is just one example. For example, the machine tool 10 may be a lathe equipped with multiple sets of workpiece spindle devices 11 and turret devices 12, or it may be a so-called multi-tasking machine equipped with a tool spindle device in addition to the turret device 12. Therefore, the machine tool disclosed in this disclosure is not limited to a lathe, but can be a machine tool with various configurations such as a machining center, milling machine, or drilling machine. The machine tool disclosed in this disclosure can be equipped with various devices whose operation can be controlled by a machining program. Also, Figure 1 shows only a part of the devices that the machine tool 10 is equipped with. The machine tool 10 may also be configured to include a loader for transporting workpieces, a turning device for inverting workpieces, a measuring device for inspecting workpieces after machining, etc.

[0019] Furthermore, the machine tool 10 may be equipped with a wireless communication interface as an external IF 14. Alternatively, the external IF 14 is not limited to a communication interface, but may also be a device capable of detecting the shape of the workpiece, such as a camera or millimeter-wave radar. Therefore, in addition to a communication interface for communicating with external devices, various other devices can be used as the external IF.

[0020] Furthermore, the database 29 may be configured to omit at least one of the following: shape information 31, machining theory information 32, machining knowledge information 33, and machine tool information 34. Also, the machine tool 10 may be configured to omit the creation program 30. Furthermore, the machine tool 10 may be configured to include a user interface other than the control panel 13, such as a portable user interface like a teaching pendant.

[0021] (Regarding data acquisition) Next, the data acquisition process by the machine tool 10 with the above configuration will be explained. Figure 2 shows the flow of the data acquisition process. As shown in Figure 2, the control device 15 of this embodiment executes a first acquisition process (step 1) to acquire first target data D1 that does not correspond to the data input rules of the database 29, and a second acquisition process (step 2) to acquire second target data D2 such as user memory information, by executing an acquisition program 28 with the CPU 23. The control device 15 also executes a registration process (step 3) by executing the acquisition program 28 to classify the information acquired in the first and second acquisition processes using AI technology and register the classified information as data in the database 29. For example, the control device 15 starts the processes of steps 1 to 3 based on a predetermined operation input to the operation panel 13.

[0022] Furthermore, the data that does not comply with the input rules in this disclosure refers to data that is difficult to import into the database in its original state, such as image data of paper drawings, image data of handwritten notes, drawing data in a specific data format, and photographs. For this reason, unsupported data is data that requires a determination of its correspondence with the type, format, and values ​​of data to be stored in the database, and in its original state, it is necessary to determine how to import it into the database. AI technology, on the other hand, is a technology that enables computers to perform tasks close to human intelligence, and can be used by selecting and combining various technologies such as machine learning, deep learning, Bayesian statistics, and genetic algorithms.

[0023] In the following description, the control device 15 that executes the acquisition program 28 on the CPU 23 may be simply referred to as the control device 15. The control device 15 executes the first acquisition process in step 1 (hereinafter simply referred to as S). The targets of the first target data D1 include, for example, work drawings, manufacturing drawings, machining proposal tables, equipment drawings, tooling diagrams, diagrams showing tool arrangements, diagrams showing tooling, diagrams showing machine tools, specifications, etc. These targets may also be in paper format, such as printed materials or handwritten notes. As described above, the control device 15 can be connected to a scanner via an external IF 14, and can acquire image data (such as drawing data) obtained by scanning paper drawings, etc., as the first target data D1.

[0024] Furthermore, the first target data D1 may be electronic data such as 2D or 3D models of the object described above. Specifically, the first target data D1 can be, for example, drawing data created with 2D or 3D CAD, data with the extension pdf, or 2D graphic data in dxf format. The first target data D1 can be electronic data representing a workpiece, cutting tool, machine tool, etc.

[0025] Furthermore, the first target data D1 may be image data (photographic data) of an object. Therefore, the first target data D1 may be image data of a workpiece, cutting tool, machine tool, etc. Also, the first target data D1 may be a video. Specifically, it may be a video of the machining operation of a machine tool during processing.

[0026] Furthermore, the first target data D1 is not limited to the names and types described above. The first and second target data D1 and D2 of this disclosure can be used as appropriate, as long as they are data that can be registered in the database 29 and used to create the NC program 27. For example, a work drawing / manufacturing drawing is a drawing that includes the shape, dimensions, center line, leader line, scale of the drawing, creation date, creator's name, work name, etc., of the work to be processed, as shown in Figure 2. For this reason, any drawing that has such information printed or written on it may be treated as a work drawing / manufacturing drawing even if the name is different.

[0027] Similarly, information relating to the cutting tools used to process the workpiece, such as diagrams of the cutting tools, numerical values ​​indicating the shape of the cutting tools (nose radius, cutting angle, drill diameter, etc.), and information indicating the processing conditions of the cutting tools (workpiece material, feed rate, rotational speed, etc.), may be treated as tooling diagrams. Likewise, information indicating the functions and capabilities of a machine tool may be treated as the first target data D1 of the machine tool. Furthermore, specifications relating to the processing of the workpiece may also be treated as the first target data D1.

[0028] In the following explanation, to avoid complexity, we will describe a first acquisition process that involves scanning a paper drawing to capture the first target data D1 as an example. The control device 15 acquires image data of a paper work drawing as the first target data D1 from the scanner, for example, via an external IF 14. At this time, the control device 15 may perform correction processing on the first target data D1. Specifically, in addition to the machining shapes of the side and front, the work drawing may also show other lines and characters such as dimension lines, center lines, and leader lines. On the other hand, generally, outline lines showing the machining shape are shown with thicker lines than other lines. Preferably, the outer diameter line is shown with the thickest line among the multiple lines included in the drawing. Therefore, the control device 15 may detect the thickest line among the multiple lines included in the image data of the drawing as the outline line showing the machining shape of the work. Furthermore, if the control device 15 misrecognizes the shape or size due to variations in the spacing of dimension lines, etc., caused by printing accuracy, wrinkles, or distortions in the printed material, it may perform correction processing such as color correction or edge detection. The control device 15 may also perform detection of additional information such as dimension values ​​and dimension lines within the drawing. In addition, as will be described later, if a detection error occurs, the control device 15 may perform a query in the second acquisition process regarding the missing data and prompt the user to input dimension values ​​or correct dimension values. In this way, the control device 15 detects the processed shape and dimension values ​​of the workpiece from the workpiece drawing. Furthermore, if dimension values ​​are not indicated, the control device 15 may also detect the actual size of the workpiece (dimensional values, actual shape size) from the scale information of the drawing and the size of the workpiece on the drawing.

[0029] In the first acquisition process of S1, the control device 15 digitizes the first target data D1 described above. Specifically, the control device 15 detects information such as the workpiece name and machining dimensions (cutting diameter) as digitized information. As a technology to detect information such as the workpiece name and machining dimensions from such workpiece drawings, OCR technology or image processing technology can be employed. The control device 15 may also use AI technology (such as machine learning) to determine the machining content (such as outer diameter machining or inner diameter machining) for each machining position in the workpiece drawing and detect it as digitized information. The control device 15 can also similarly detect information from photographs and CAD data other than drawings. Furthermore, the control device 15 may also detect information from drawings and photographs of cutting tools and machine tools, not just workpieces.

[0030] The control device 15 uses AI technology to classify the numerical data and detected text information obtained through the first acquisition process, and then executes a registration process to register the classified information in the database 29 (S3). The database 29 stores shape information 31, machining theory information 32, machining knowledge information 33, and machine tool information 34. Some of this information may overlap, and all of the information may be of different types.

[0031] For example, the shape information 31 includes image data of the workpiece drawing, image data of handwritten notes, and 2D or 3D CAD data (drawing data). The shape information 31 also includes the machining dimensions (cutting diameter) of each machined part of the workpiece, the material of the base material, the workpiece name, the name of the customer who requested the workpiece machining, and the name of the person in charge of the drawing.

[0032] For example, the control device 15 executes the AI ​​program of the acquisition program 28 and performs machine learning on the data type and classification method of the shape information 31 described above. This machine learning is, for example, supervised learning in which the learning data is taught with correct answers provided. The control device 15 may also perform machine learning after the operation of the machine tool 10 has started. Based on the rules and characteristics of the information learned by machine learning, the control device 15 classifies the information detected from the image data acquired by the scanner described above and registers it in the database 29. For example, if the control device 15 determines through AI that a numerical value converted from a work drawing is written near a dimension line and is a dimension value, it classifies it as a dimension value and registers it in the database 29. The control device 15 also determines the type of machining (such as outer diameter machining or inner diameter machining) of the machined part indicated by the dimension value based on the relationship between the position where the dimension value is written, the position indicated by the dimension line, and the shape of the work piece. Furthermore, if the control device 15 determines through AI that a numerical value converted from a work drawing is text indicating the material of the work piece (such as aluminum or iron), it classifies it as a material and registers it in the database 29. The control device 15 detects information such as dimensional values, machining details, material, or drawing name and workpiece name, and associates this information to register it as shape information 31 data.

[0033] Furthermore, the machining theory information 32 includes, for example, information such as cutting tool identification information, machining conditions for each cutting tool (spinning speed, feed rate, etc.), nose radius, cutting angle, cutting edge angle, and, in the case of rotary tools, drill diameter. The values ​​for these machining conditions, such as spinning speed and feed rate, are theoretical values ​​set by the cutting tool manufacturer for each workpiece material. The machining theory information 32 also includes sample programs for the NC program 27, and a list of G-codes and M-codes. In addition, the machining theory information 32 includes information indicating the type of coolant (water-based or oil-based, etc.) and information on the performance of the coolant.

[0034] Furthermore, the machine tool information 34 includes information such as the machine's functions and specifications for machine tool 10 or other machine tools. Specifically, the machine tool information 34 includes machine information necessary for creating the NC program 27, such as the type of cutting tool attached to the turret and the holder number indicating the mounting position of the cutting tool. The machine tool information 34 also includes information necessary for determining machining conditions, such as the maximum rotational speed of the workpiece spindle unit 11. In addition, the machine tool information 34 includes information necessary for determining the time required for indexing, positioning, and retracting the cutting tool, such as the turret's rotational speed and the turret's movement speed. Information such as the loader's movement speed may also be included in the machine tool information 34.

[0035] Furthermore, the control device 15 acquires imaging data of the machine tool 10 in the first acquisition process of S1, for example. The control device 15 extracts and classifies information about the components contained in the imaging data acquired using AI technology, and registers the classified information as data in the database 29. For example, the control device 15 detects the shape, type, position, etc., of the components in the imaging data using image processing with AI technology. The control device 15 determines which information group (such as machine tool information 34) in the database 29 the detected component information should be registered in, and registers it in the determined information group.

[0036] Specifically, for example, the control device 15 extracts information such as the mounting position (holder number) of the cutting tool on the turret, the type of cutting tool mounted, and the blade protrusion (coordinates of the cutting edge) from the image data of the turret, and registers it as machine tool information 34. This allows the control device 15 to detect and register information from the photograph indicating what kind of cutting tool is mounted on the turret, in which holder number, and in what condition. In this case, the information on the cutting tool and the holder number are examples of the information on components included in the image data in this disclosure. The control device 15 may also extract information on the type of coolant (oil-based or water-based) from the image data of the coolant used in the machine tool 10 and register it in the machine tool information 34. The control device 15 may also associate the registered machine tool information 34 data with the coolant data in the machining theory information 32. This allows the control device 15 to register which machine tool uses which type of coolant. Alternatively, for any machine tool, it is possible to register information from the image data as to whether or not it is a machine that performs dry machining without using coolant. Furthermore, the control device 15 may extract and register information about the type of chuck (jaws, collet chuck, etc.) from imaging data obtained by imaging the chuck of the workpiece spindle device 11.

[0037] Furthermore, the machining knowledge information 33 stores, for example, information from the machining proposal table, a change history of the machining proposal table, and information on machining conditions when machining fails. The machining proposal table is data that includes each machining process of the workpiece (machining conditions, etc.) and information about the workpiece (material, etc.). The control device 15 can create an NC program 27 from this machining proposal table by executing the program 30 created by the CPU 23. The machining proposal table may also include information on the cycle time related to the machining of the workpiece (cycle time for each machining process, overall cycle time, loader cycle time, etc.) and information on the machine configuration of the machine tool 10. For this machining proposal table, for example, a table in the data format disclosed in International Publication WO2024 / 095400, International Publication WO2024 / 095401, International Publication WO2024 / 095402, and International Publication WO2024 / 224551 can be adopted.

[0038] Figures 3 to 5 show an example of a machining proposal table 41. Figure 3 shows the data that constitutes the basic information of the machining proposal table 41. The machining proposal table 41 includes a figure 43 showing the shape of the workpiece, information 44 indicating the workpiece name and material, and information 45 about the machine tools and chucks used in each machining process. Figure 4 shows the left side of the detailed information table for each machining process included in the machining proposal table 41, and Figure 5 shows the right side. As shown in Figure 4, the machining proposal table 41 includes, for example, information on the process identification, the cutting tool to be used identification, the required precision (tolerance), the shape of the cutting tool to be used, and the machining content.

[0039] Furthermore, as shown in Figure 5, the machining proposal table 41 contains machining settings such as cutting diameter and rotational speed, and time settings such as cutting time and positioning time for each machining process. The control device 15 uses AI technology to classify the information acquired and digitized in the first acquisition process and automatically sets it in each item of the machining proposal table 41. For example, the control device 15 sets the machining conditions and machining sequence for each machining process based on the shape and dimensions of the workpiece read by scanning, the machining theory information 32 in the database 29, the machining knowledge information 33 (past machining proposal tables 41), and the machine tool information 34. The control device 15 sets the digitized information in each item of the set machining process. This allows the machining proposal table 41 to be automatically generated based on the information read from the workpiece drawing, etc. For example, the control device 15 uses generation AI, etc., to input image data of the workpiece drawing, classify the detected information, and generate the machining proposal table 41, executing machine learning. The control device 15 learns in advance the correspondence between the classified information and each item in the machining suggestion table 41, and learns which item in the machining suggestion table 41 should be entered when certain information is detected and classified. Preferably, the control device 15 can automatically create the machining suggestion table 41 and, consequently, the NC program 27 from the workpiece drawing.

[0040] Accordingly, the database 29 of this embodiment contains data for a machining proposal table 41 used for machining a workpiece, in which machining conditions are set for each step of the workpiece machining process (Figures 3 to 5). The control device 15 learns the correspondence between the classified information and the data for each item in the machining proposal table 41 in advance using AI technology, and in the registration process of S3, it registers the information classified according to the pre-learned rules into the machining knowledge information 33 of the database 29.

[0041] According to this, the control device 15 can pre-learn the correspondence between the classified information and the data for each item in the machining proposal table 41 using AI technology, and register the information classified according to the pre-learned rules as data for the appropriate items in the machining proposal table 41 in the database 29. As a result, the information quantified in the first acquisition process in S1 and the stored information acquired in the second acquisition process in S2 (described later) can be incorporated as data for the machining proposal table 41. The control device 15 can create an NC program 27 by coding each machining process in the machining proposal table 41.

[0042] For example, the control device 15 may pre-learn the types of machining operations, such as internal diameter machining, external diameter machining, and end face machining, and their relationship to the machining position and shape in the workpiece drawing. The control device 15 may, according to the pre-learned rules, import the machining position indicated by a dimension value in the workpiece drawing as the cutting diameter for internal diameter machining if it is an internal diameter machining position. Furthermore, the control device 15 does not have to register the classified information in the machining proposal table 41. For example, the control device 15 may only perform a simple classification of the detected information. The control device 15 may learn rules that associate the names of the items to be classified with the detection conditions for those items. The control device 15 may then extract information from the workpiece drawing for each detection condition and register the extracted information as the value of the item associated with that detection condition.

[0043] Furthermore, as shown in Figure 5, the machining proposal table 41 includes fields for inputting whether the set machining settings are theoretical or actual values ​​(selection fields in Figure 5) and for inputting the reason (reason fields in Figure 5). Theoretical values ​​are, for example, theoretical values ​​recommended by the cutting tool manufacturer, and are set according to the material of the workpiece, the type of cutting tool, etc. Specifically, theoretical values ​​are set for machining settings such as cutting speed depending on the material, such as carbon steel, chromium-molybdenum steel, aluminum alloy, and ductile cast iron, and the type of cutting tool, such as cutting tools, drills, and end mills. Actual values ​​are, for example, values ​​changed after trial cutting, or values ​​adjusted based on the experience and past results of skilled users. Theoretical values ​​can be used as is as machining settings, or adjustments may be necessary. For this reason, the user may change the machining settings in the second acquisition process of S2, which will be described later. Alternatively, the user may create the NC program 27, perform trial cutting, and then modify the NC program 27 by changing the machining settings. For example, a user may change the machining settings by operating the control panel 13. This change history is also stored in the database 29 as machining knowledge information 33 and learned using AI technology. As a result, data can be imported into the database 29 with greater accuracy. For example, when registering new information where similar machining content or machining conditions have been registered in the past, if there is a history of changing from theoretical values ​​to actual values ​​in the past, the machining settings can be automatically changed or a warning can be issued.

[0044] Also, as shown in FIGS. 1 and 2, failure information for failed machining is registered in the machining knowledge information 33. Specifically, the failure information indicates, for example, for a machine tool, the material of the workpiece, the cutting tool, the machining content, the machining setting values (such as the cutting diameter), etc., what kind of machining failures occur when machining is executed in what kind of combination. Machining failure includes, for example, the occurrence of so-called chatter, the generation of burrs, the winding of chips, or tool breakage. By leaving the information of the setting values set in the machining process when such failures occur in the database 29, if there is information on past failures, similar to the above-described performance values, etc., changes in machining setting values or warnings can be executed.

[0045] Incidentally, in the above example, the machining proposal table 41 has been described. Similarly, the shape information 31, the machining theory information 32, and the machine tool information 34 can also be registered in the database 29 by executing the first acquisition process (S1) and the registration process (S3). For example, the control device 15 may determine and classify the image data of the workpiece drawing as information related to the shape of the workpiece by AI technology and register it in the shape information 31. Also, the control device 15 may determine and classify the information detected from the image data of the cutting tool manual as information on the machining conditions of the cutting tool by AI technology and register it in the machining theory information 32. Further, the control device 15 may determine and classify the information detected from the image data of the machine tool manual as information such as the specifications of the machine tool by AI technology and register it in the machine tool information 34.

[0046] Preferably, as described above, the control device 15 can use AI technology to generate a machining proposal table 41 from image data such as workpiece drawings, and register the data. However, in the machining area of ​​the machine tool 10, there are notes that can only be deciphered by the person who wrote them, i.e., know-how-related documents. Specifically, when machining a workpiece using the machine tool 10, there are notes (change history, etc.) of when machining conditions etc. have been changed based on knowledge and experience to improve machining accuracy. In addition, there is also memory information that is not recorded as notes, etc., but remains in the user's memory. For example, there is a memory of changing machining settings because machining could not be done well with theoretical values, or a memory of a machining failure. Such information may not be converted into data and may remain as notes or memories of the user, for example, an expert. Therefore, in the second acquisition process of S2, the control device 15 can acquire information that is difficult to register in the database 29 in the first acquisition process of S1 as second target data D2, and classify and register it using AI technology.

[0047] In S2, the control device 15 acquires second target data D2 from the user using voice recognition and chat functions. This allows information that remains in a person's memory or information in handwritten notes that only the person can decipher to be acquired as second target data D2 via voice or chat. Memory information and the like can be efficiently captured.

[0048] By executing the assistant program of the capture program 28 with the CPU 23, the control device 15 can acquire the voice data of the user via the microphone of the external IF 14. The control device 15 executes voice recognition processing (S2) on the voice data input via the external IF 14 using the assistant program and the AI program, and executes registration processing (S3) such as classification using the AI technology on the recognized voice data. For example, the control device 15 receives the items to be input in the processing proposal form 41 by voice input via the external IF 14. When the control device 15 receives an input instruction for a specific item, it receives voice input of the value of that item. The user, for example, vocally instructs the item to be input for each item shown in FIGS. 3 to 5, and then vocally inputs the value to be input. The user, for example, makes a voice input such as "Please set the cutting diameter 1 value for the processing step 30 to 102". The user registers the content of the memory and their own memo (second target data D2) by voice data. The control device 15 detects the item instructed by the voice chat function or the like using the AI technology, and classifies the information by registering the value received by voice in the detected item. Alternatively, the control device 15 may receive and register the reason for the value change and the processing failure condition by voice data. Further, when the data of the work drawing detected, classified, and registered through the process of S1 is incorrect, the control device 15 may receive a correction by voice. Therefore, the second target data D2 is not limited to the stored information and the memo information. Further, the control device 15 may perform machine learning on the voice characteristics of each user.

[0049] Furthermore, in S2, the control device 15 may accept input of the second target data D2 by a chat function instead of, or in addition to, the voice input function described above. The control device 15 executes the assistant program of the acquisition program 28 on the CPU 23 to perform data input and output using the chat function with the operation panel 13 or a processing device (tablet, etc.) connected via the external IF 14. In this case as well, the control device 15 accepts input, classifies the accepted information, and registers it using the chat function with AI technology. For example, the control device 15 uses the chat function to ask questions such as "What data would you like to register?" via the touch panel of the operation panel 13. The user registers data about stored information and the contents of their own notes while interactively answering the questions displayed by the chat function. The control device 15 uses AI technology to judge and decide on the content of the questions and answers, determines the registration destination of the input second target data D2, and performs registration to the database 29.

[0050] Furthermore, the control device 15 may be configured to acquire stored information from the user in S2 using at least one of the voice recognition function and the chat function. Also, the method for inputting the second target data D2 is not limited to the functions described above; for example, it may be a method of inputting by handwriting or a method of determining and inputting input content from eye movements.

[0051] Furthermore, if the types of data required to be registered in the database 29 for creating the NC program 27 increase, and the information obtained from processing S1 and S2 is insufficient to obtain the necessary data, the control device 15 will request the registration of the missing data. This means that if data that was initially thought to be unnecessary for creating the NC program 27 becomes necessary for creating the NC program 27 later on and is no longer present in the database, the control device 15 can request the registration of that data.

[0052] The control device 15 requests input of the missing data, for example, through voice recognition or chat functions. For example, suppose the technology for creating the NC program 27 is updated, and information on the type of coolant (such as water-based or oil-based) is required for its creation. The administrator of the machine tool 10 adds a coolant type item to the machining proposal sheet 41, for example. When the control device 15 detects that an item has been added to the machining proposal sheet 41, it determines that coolant type information is needed to create the NC program 27 from the machining proposal sheet 41. When registering a work drawing, or when creating an NC program 27 corresponding to that work drawing, the control device 15 notifies the user that coolant type information is missing and requests its registration. For example, when executing the registration process (S3) or receiving an instruction to create an NC program 27, the control device 15 makes an inquiry via voice or chat, such as, "Is the coolant used in machining this workpiece water-based or oil-based?" The user confirms the requested data and inputs it via voice or other means. This allows the type of coolant to be taken into account when creating the NC program 27, enabling the creation of an NC program 27 that is in line with the latest technology.

[0053] Furthermore, the control device 15 may request missing data for previously registered data, not just when new data is being registered. The information related to the machine tool 10 is diverse, and some information may be good to keep, but at the time it may not be clear which information should be kept. Later, as technology advances and other factors arise, the information may only remain in memory or as notes. For example, if a new item is added to the machining proposal sheet 41, the control device 15 may accept input of that new item for the machining proposal sheet 41 that has already been registered.

[0054] Furthermore, when the control device 15 receives an instruction from the user to execute the created program 30, it creates an NC program 27 based on the data in the database 29 that was registered and constructed by the registration process (S3) described above. For example, the control device 15 receives a selection of a machining proposal table 41 registered in the machining knowledge information 33, codes each machining process in the selected machining proposal table 41, and creates an NC program 27. This makes it possible to create an NC program 27 that machines the workpiece indicated by the acquired workpiece diagram, etc.

[0055] Incidentally, the correspondence between the terms used in the first embodiment and the terms used in the claims will be explained below. In the first embodiment described above, the control device 15 is an example of a data acquisition device of the present disclosure. The first target data D1 is an example of target data. The second target data D2 is an example of stored information. The NC program 27 is an example of a machining program. The first acquisition process in S1 is an example of a digitization process. The second acquisition process in S2 is an example of a stored information acquisition process.

[0056] As described above, the first embodiment provides the following effects. The control device 15, which is one aspect of the present invention, acquires first target data D1 that does not correspond to the input rules for the data in the database 29, and digitizes the acquired first target data D1 (S1, an example of the digitization process in this disclosure). The control device 15 also acquires user memory information (second target data D2) related to the machine tool 10 (S2, an example of the memory information acquisition process in this disclosure). The control device 15 uses AI technology to classify the numerical information digitized in S1 and the memory information acquired in S2, and registers the classified information as data in the database 29 (S3, an example of the registration process in this disclosure).

[0057] According to this, the control device 15 can use AI technology to classify and register in the database 29 numerical information obtained by quantifying the first target data D1 that does not correspond to the input rules, and user memory information, etc. (second target data D2). Not only numerical information but also memory information can be classified and registered in the database 29 used to create the NC program 27. Information from handwritten notes that are difficult to decipher and the know-how of experts can be accumulated in the database 29, making it possible to construct a database 29 that enables even engineers with little knowledge or experience to create NC programs 27 with the same level of expertise as experts. It should be noted that the memory information in this disclosure is not limited to information that remains only in a person's memory, but also includes information from handwritten notes that are difficult for anyone other than the creator to decipher, that is, information that can be determined by comparing handwritten notes with memories, etc., as part of the concept of memory information.

[0058] (Second Embodiment) Next, a second embodiment of the present disclosure will be described. In the first embodiment described above, the case in which the control device 15 of the machine tool 10 is used as the data acquisition device of the present disclosure was described, but the present disclosure is not limited to this. The data acquisition device of the present disclosure may be a device other than a machine tool. Figure 6 shows a data acquisition system 51 including a data acquisition device 52 according to the second embodiment. In the following description of the second embodiment, the same reference numerals will be used for the same components as in the first embodiment described above, and their descriptions will be omitted as appropriate.

[0059] As shown in Figure 6, the data acquisition system 51 comprises a data acquisition device 52 and client terminals 54 installed in each production plant 53. The data acquisition device 52 is, for example, a server connected to a WAN 55 such as the Internet, and includes a CPU 57, a user interface 58 such as a mouse and keyboard, a network interface 59 connected to the WAN 55, and a storage device 60.

[0060] The storage device 60 stores the NC program 27, the acquisition program 28, the database 29, and the creation program 30, similar to the storage device 24 in the first embodiment. The data acquisition device 52 provides, for example, the functions performed by the machine tool 10 in the first embodiment as a service over the network. The client terminal 54 is, for example, a PC that manages the machine tool installed in the production factory 53. Alternatively, the client terminal 54 may be a terminal capable of running CAD or CAM (Computer-Aided Manufacturing) applications.

[0061] Even with this configuration, similar to the first embodiment described above, it is possible to extract information from work drawings, classify the extracted information, register the classified information, and register stored information using voice recognition and chat functions. The storage device 60 stores, for example, an OS 61 that functions as a web server. The data acquisition device 52 accepts access to the web page from the client terminal 54 by executing the OS 61 on the CPU 57. The data acquisition device 52 performs the first and second acquisition processes (S1, S2) and registration process (S3) shown in Figure 2 with the client terminal 54 via the WAN 55. As a result, a user operating the client terminal 54 in the production plant 53 can register work drawings, notes, stored information, etc., in the database 29 of the data acquisition device 52 via the WAN 55. In addition, the user can create an NC program 27 for processing the work in the work drawing by having the data acquisition device 52 execute the creation program 30.

[0062] Furthermore, the information of the machining proposal table 41 used to create the NC program 27 for each machine tool installed in the production plant 53 may be stored as information of the machining proposal table 41 included in the machining knowledge information 33 of the database 29. Also, information such as specifications corresponding to each machine tool installed in the production plant 53 may be stored as machine tool information 34 of the database 29. This makes it possible to import data and create NC programs 27 according to each type of machine tool, even if different types of machine tools are installed in each of the production plants 53.

[0063] Furthermore, the contents of this disclosure are not limited to the embodiments described above, and can be implemented in various forms with various modifications and improvements based on the knowledge of those skilled in the art. For example, the configuration of the machine tool 10 in the first embodiment is just one example. The machine tool 10 is not limited to a turret-type lathe, but may be a machining center, for example. Also, the machining program of this disclosure is not limited to the NC program 27, but may be any other machining program capable of controlling the machine tool 10. In addition, the control device 15 may not be configured to learn in advance the correspondence between the classified information and the data of each item in the machining suggestion table 41 using AI technology. Also, the control device 15 may not be configured to require registration of missing data. Also, the control device 15 may not be configured to extract and classify information of components included in the imaging data using AI technology.

[0064] Furthermore, the contents of this disclosure are not limited to the dependencies described in the claims. For example, this specification also discloses a technical concept in which "the data acquisition device described in claim 1 or claim 2" in claim 4 is changed to "the data acquisition device described in any one of claims 1 to 3". Also, for example, this specification also discloses a technical concept in which "the data acquisition device described in claim 1 or claim 2" in claim 5 is changed to "the data acquisition device described in any one of claims 1 to 4". Also, for example, this specification also discloses a technical concept in which "equipped with the data acquisition device described in claim 1 or claim 2" in claim 6 is changed to "equipped with the data acquisition device described in any one of claims 1 to 5".

[0065] 10 Machine tool, 15 Control device (data acquisition device), 27 NC program (machining program), 29 Database, 41 Machining suggestion sheet, 52 Data acquisition device, D1 First target data (target data), D2 Second target data (storage information).

Claims

1. A data acquisition device for importing data into a database used to create a machining program for controlling a machine tool when machining a workpiece, the device performing: a data acquisition process that acquires target data that does not correspond to the data input rules of the database and converts the acquired target data into numerical data; a memory information acquisition process that acquires user memory information related to the machine tool; and a registration process that uses AI technology to classify the numerical information converted into numerical data by the data acquisition process and the memory information acquired by the memory information acquisition process, and registers the classified information as data in the database.

2. The data acquisition device according to claim 1, wherein the database contains data of a processing suggestion table used for processing a workpiece, the processing conditions for each step of processing the workpiece are set in the processing suggestion table, the data acquisition device learns in advance the correspondence between classified information and the data of each item in the processing suggestion table using AI technology, and in the registration process registers the classified information in the database according to the rules learned in advance.

3. The data acquisition device according to claim 1 or 2, wherein, in the memory information acquisition process, the memory information is acquired from the user using at least one of the voice recognition function and the chat function.

4. The data acquisition device according to claim 1 or 2, which, when the types of data required to be registered in the database for creating the processing program increase, and the numerical information obtained by the numerical processing is insufficient to obtain the necessary data, requests the registration of the missing data.

5. The data acquisition device according to claim 1 or 2, comprising acquiring imaging data of the machine tool, extracting and classifying information of components contained in the acquired imaging data using AI technology, and registering the classified information as data in the database.

6. A machine tool comprising a data acquisition device according to claim 1 or claim 2, which creates the machining program based on data registered in the database.

Images