Program creation device, program creation method, and computer-readable storage medium

Abstract

A program creation device according to the present disclosure receives a request for creating a processing program used in an operation test in a natural language, creates a prompt in which an instruction to create the processing program is added to the request, transmits the prompt to an interactive response device, receives an answer from the interactive response device, extracts the processing program from the answer, and displays the processing program.

Description

Programming device, programming method, and computer-readable storage medium

[0001] The present disclosure relates to a program creation device, a program creation method, and a computer-readable storage medium.

[0002] Conventionally, there exists a technology for generating programming code using artificial intelligence, for example, Patent Document 1.

[0003] JP 2024-030940 A

[0004] For machine tools, operational tests are conducted to actually drive the machine and adjust the operating performance of the motor. Operational tests also include tests to actually operate the machine tool with a machining program.

[0005] Creating a machining program for each operation test is cumbersome, and for inexperienced operators, creating a machining program can be an obstacle to the operation test.

[0006] It is desirable to be able to easily create machining programs.

[0007] A program creation device according to one aspect of the present disclosure includes a request acquisition unit that receives a request in natural language to create a machining program to be used in an operating test, a pre-processing unit that creates a prompt that adds instructions for creating the machining program to the request, a transmitting unit that transmits the prompt to an interactive response device, a receiving unit that receives a response from the interactive response device, a post-processing unit that extracts the machining program from the response, and a display unit that displays the machining program.

[0008] 10 is a block diagram of a program creation device. FIG. 10 is an example of a screen for accepting input of a user's request. FIG. 10 is a diagram showing an example of a machining program created by a generation AI. FIG. 10 is an example of a display screen for a machining program. FIG. 10 is a block diagram of a program creation device of a second embodiment. FIG. 10 is a diagram showing an example of a prompt to which machine information is added as a constraint. FIG. 10 is a flowchart explaining the operation of a program creation device. FIG. 10 is a block diagram of a program creation device of a third embodiment. FIG. 10 is a diagram showing an example of a prompt requesting an explanation of a machining program. FIG. 10 is a diagram showing an example of a display screen for an explanation of a machining program. FIG. 10 is a block diagram of a program creation device of a fourth embodiment. FIG. 10 is a block diagram of a program creation device of a fifth embodiment. FIG. 10 is a diagram showing an example of waveform data. FIG. 10 is a block diagram of a program creation device of a sixth embodiment. FIG. 10 is a diagram showing an example of a screen to which an objective related to an operation test is accepted in natural language. FIG. 10 is a schematic diagram explaining a machine learning model. FIG. 10 is a block diagram of a program creation device of a seventh embodiment. FIG. 10 is a diagram showing an example of adding a tool's movable range to a prompt. FIG. 10 is a hardware configuration diagram of a program creation device.

[0009] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In the following description, components having the same or similar functions will be denoted by the same reference numerals. Duplicate descriptions of those components may be omitted.

[0010] In this application, "based on XX" means "based on at least XX," and includes cases where it is based on other elements in addition to XX. Furthermore, "based on XX" is not limited to cases where XX is used directly, but also includes cases where it is based on XX that has been calculated or processed. "XX" is any element (for example, any information).

[0011] The program creation device 100 creates a machining program used to adjust a motor or the like. The program creation device 100 is implemented, for example, in a numerical control device connected to an interactive response device. The program creation device 100 can also be implemented on a computer such as a personal computer connected to the interactive response device, a cell computer, a fog computer, or a cloud server.

[0012] The program creation device 100 is connected to an interactive response device via a network. The interactive response device may be distributed. The interactive response device includes a generation AI. By learning from a large number of samples, the generation AI is able to output text with a high probability as a response to input data when the data is input.

[0013] The generation AI is assumed to have learned data related to the machining program sufficiently to generate the machining program. The learned data includes the machining program, a manual, etc.

[0014] 1 is a block diagram of a program creation device 100 according to a first embodiment. The program creation device 100 includes a request acquisition unit 1, a preprocessing unit 2, a transmission unit 3, a reception unit 4, a postprocessing unit 5, and a display unit 6.

[0015] The request acquisition unit 1 accepts user requests in natural language (text). The request acquisition unit 1 of the first embodiment particularly accepts tool trajectories in an operation test in natural language. FIG. 2 is an example of a screen for accepting input of requests. The screen of FIG. 2 is composed of a text input area for accepting tool trajectories in natural language and a program display area for displaying the created machining program. Command buttons between the text input area and the program display area accept instructions for creating a machining program.

[0016] The preprocessing unit 2 performs preprocessing for sending the text input to the request acquisition unit 1 to the interactive response device. The text sent to the interactive response device is called a prompt. A prompt is an instruction or question sent to the interactive response device. In order to allow the generation AI to accurately understand the user's intention and elicit the expected output, it is necessary to create an effective prompt. The preprocessing unit 2 adds commands / instructions, adds context / background / constraints, specifies the output format of the generation AI, and converts the prompt into a structure that is easy for the generation AI to understand, so that the generation AI can generate appropriate responses and results.

[0017] The pre-processing unit 2 of the first embodiment adds instructions to the generation AI. For example, the instruction is, "Please create a CNC machining program that draws the following trajectory." The prompt with the added instruction becomes, "Please create a CNC machining program that draws the following trajectory. A square with sides of 150 mm." Markings may also be added to clarify the instructions. For example, adding the marking "# command statement" changes the prompt to "# command statement Please create a CNC machining program that draws the following trajectory," making the content of the command clear. Adding the marking "# trajectory shape" changes the prompt to "# trajectory shape A square with sides of 150 mm," making the shape of the trajectory clear.

[0018] The transmission unit 3 transmits the prompt created by the preprocessing unit 2 to the interactive response device. The interactive response device creates a processing program instructed by the prompt. The interactive response device may be a general-purpose model on a public cloud, an on-premise specialized model, or an extended model obtained by performing additional learning on a general-purpose model.

[0019] The receiving unit 4 acquires the machining program from the interactive response device. The machining program created by the generation AI may contain unnecessary text, such as part of the sent prompt and an explanation of the response. The post-processing unit extracts only the machining program from the response of the generation AI. Figure 3 shows an example of a machining program created by the generation AI. The machining program includes an explanatory statement such as "The following G-code can be used.", a machining program explanatory statement such as "This program first moves to coordinates (0,0),...," and a note such as "You can create a CNC program using the above programs. Please feel free to ask any questions." These are unnecessary text that is not part of the machining program. In addition, explanatory statements are added to each block of the machining program. For example, the first line of the machining program, "G00 X0.Y0.;," is accompanied by the explanatory statement "Rapid move to the starting point." This explanatory statement is also unnecessary text that is not part of the machining program. Because the generation AI generates text probabilistically, the answers of the generation AI may contain unnecessary text.

[0020] The post-processing unit 5 extracts a processing program from the response of the generation AI. As an example of a method for extracting text, pattern matching or the like is used to extract only text that conforms to the grammar of the processing program. However, the method for extracting text is not particularly limited.

[0021] The display unit 6 displays the machining programs extracted by the post-processing unit 5. Fig. 4 shows an example of a display screen of the machining programs. The program display area of ​​the screen in Fig. 4 displays the machining programs "G00 X0.Y0.;", "G01 X150.Y0.F100.;", "G01 X150.Y150.;", "G01 X0.Y150.;", and "G01 X0.Y0.;".

[0022] 5 is a block diagram of a program creation device 100 according to a second embodiment. The program creation device 100 includes a request acquisition unit 1, a preprocessing unit 2, a transmission unit 3, a reception unit 4, a postprocessing unit 5, a display unit 6, and a machine information acquisition unit 7. In the explanations of the second to seventh embodiments, overlapping explanations will be omitted, and only differences will be explained.

[0023] The machine information acquisition unit 7 acquires machine information from the numerical control device. The machine information includes the manufacturer of the numerical control device (corresponding to the syntax of the machining program), the axis configuration of the machine tool, the axis movement range, motor specifications, etc.

[0024] The preprocessing unit 2 adds information necessary for creating a machining program to the prompt. The manufacturer and model of the numerical control device are related to the syntax of the machining program. Information on the axis configuration of the machine tool is also related to the machining program. The movement range of the axes also becomes a constraint on the machining program.

[0025] Figure 6 is an example of a prompt with machine information added as a constraint. In the example in Figure 6, information about the axis configuration, "In addition, only the following axis names may be used in the program: X-axis, Z-axis," and the constraint that only the specified axes may be used are added.

[0026] The post-processing unit 5 modifies the machining program by referring to the machine information of the numerical control device. For example, if the only axes present in the machine tool are the X-axis and the Z-axis, the axes of the machining program created by the generating AI are modified to the present axes (X-axis and Z-axis). Modifying the machining program in the post-processing unit 5 eliminates the need to send information such as machine information to an external interactive response device, which is safe. As will be described later, the interactive response device may add machine information to a prompt to create a prompt that is easy for the generating AI to understand.

[0027] The operation of the program creation device 100 will be described with reference to the flowchart in Fig. 7. The request acquisition unit 1 receives a tool trajectory shape in natural language (step S1). When a program creation command is issued, the machine information acquisition unit 7 acquires machine information from the numerical control device. The pre-processing unit 2 reads information required to create a machining program from the machine information (step S2). The information required to create a machining program is set as parameters of the numerical control device.

[0028] The pre-processing unit 2 creates a prompt by adding commands / instructions, context / background / constraints, output format, etc. (Step S3). An example of an instruction is "Please create a CNC machining program that draws the following trajectory." Constraints include the axis configuration, and background information includes the manufacturer of the numerical control device. The prompt may also be marked with "# command" or similar to make it easier for the generating AI to understand.

[0029] The transmitting unit 3 transmits the preprocessed prompt to the interactive response device (step S4). The generation AI of the interactive response device creates a processing program in accordance with the prompt. The receiving unit 4 receives the processing program created by the generation AI (step S5).

[0030] The post-processing unit 5 performs post-processing of the machining program created by the generation AI (step S6). The machining program created by the generation AI contains unnecessary text such as part of the sent prompt and explanation of the answer. The post-processing unit 5 extracts only the actual machining program part.

[0031] The post-processing unit 5 acquires machine information of the numerical control device as necessary and modifies the machining program by referring to the acquired machine information (step S7). The display unit 6 displays the post-processed machining program on the display unit 6 (step S8).

[0032] 8 is a block diagram of a program creation device 100 according to a third embodiment. The program creation device 100 includes a request acquisition unit 1, a preprocessing unit 2, a transmission unit 3, a reception unit 4, a postprocessing unit 5, a display unit 6, an explanation request preprocessing unit 8, an explanation request transmission unit 9, an explanation reception unit 10, and an explanation display unit 11.

[0033] The explanation request preprocessing unit 8 adds an instruction to request an explanation of the corrected machining program after post-processing, and creates a prompt. An example of the prompt is shown in Figure 9. The prompt in Figure 9 has an added command: "Please tell me the meaning of the following CNC machining program."

[0034] The explanation request transmitting unit 9 transmits the prompt created by the explanation request preprocessing unit 8 to the interactive response device. The generating AI of the interactive response device responds with an explanation of the machining program.

[0035] The commentary receiving unit 10 receives the commentary of the machining program. The commentary display unit 11 displays the received commentary. On the display screen of Fig. 10, the commentary of the machining program is displayed as "G00 X0.Z0.; This means that the coordinates of the X and Z axes are moved to 0 by fast forward. At this time, the tool moves without machining. ..."

[0036] 11 is a block diagram of a program creation device according to a fourth embodiment. The program creation device 100 includes a request acquisition unit 1, a preprocessing unit 2, a transmission unit 3, a reception unit 4, a postprocessing unit 5, a display unit 6, and a program transmission unit 12. The program creation device 100 according to the fourth embodiment is an information processing device connected to a numerical control device. Examples of information processing devices include a personal computer, a cell computer, a fog computer, and a cloud server.

[0037] The program transmission unit 12 transmits the created machining program to the numerical control device. The numerical control device displays the acquired machining program on a screen. The numerical control device accepts manual adjustment of the machining program.

[0038] 12 is a block diagram of a program creation device according to a fifth embodiment. The program creation device 100 includes a request acquisition unit 1, a preprocessing unit 2, a transmission unit 3, a reception unit 4, a postprocessing unit 5, a display unit 6, a program transmission unit 12, a waveform data acquisition unit 13, an adjustment unit 14, and a parameter transmission unit 15. The program creation device 100 according to the fifth embodiment is a numerical control device or an information processing device connected to a numerical control device. Examples of the information processing device include a personal computer, a cell computer, a fog computer, and a cloud server.

[0039] The program transmission unit 12 transmits the created machining program to the numerical control device. The numerical control device controls the machine tool in accordance with the machining program. The waveform data acquisition unit 13 acquires waveform data that indicates the state of the machine tool. The waveform data is a collection of data that indicates the state of the machine tool. The waveform data indicates changes in the state of the machine tool, the trajectory of the tool, etc.

[0040] The adjustment unit 14 adjusts the parameters of the numerical control device based on the waveform data. The following describes corner adjustment as an example. FIG. 13 shows waveform data showing the tool trajectory at a corner. Ideally, corners should be turned at right angles, but overshooting occurs. In this example, a parameter called the allowable speed difference is adjusted to keep the corner overshoot within an allowable range.

[0041] It is assumed that the adjustment unit 14 has acquired a target threshold value. The adjustment unit 14 acquires data indicating the state of the machine tool from the waveform data and adjusts the parameters so that the data falls within the threshold value range. In the example of FIG. 13 , the shape error (overshoot) at the corner is acquired and the allowable speed difference is adjusted. For example, in the upper diagram of FIG. 13 , the allowable speed difference is "1000.000 mm / min" and the overshoot is "maximum approximately 40 μm." If the overshoot exceeds the threshold value, the adjustment unit 14 changes the allowable speed difference (parameter). For example, if the allowable speed difference is adjusted to "400.000 mm / min," the overshoot becomes "maximum approximately 10 μm" (lower diagram of FIG. 13 ). When the overshoot falls within the threshold value range, the adjustment unit 14 ends the adjustment of the parameter.

[0042] The parameter transmission unit 15 transmits the adjusted parameters to the numerical control device, thereby automating the process from creating a machining program to adjusting the parameters.

[0043] 14 is a block diagram of a program creation device 100 according to a sixth embodiment. The program creation device 100 includes a request acquisition unit 1, a trajectory determination unit 16, a preprocessing unit 2, a transmission unit 3, a reception unit 4, a postprocessing unit 5, a display unit 6, a label acquisition unit 17, and a log collection unit 18.

[0044] The request acquisition unit 1 acquires the purpose of the driving test in natural language (text). In the screen of FIG. 15, the natural language "check corner shape error" is input as the request. Also, in FIG. 15, multiple check boxes are provided. The check boxes allow the user to select that the natural language (text) input as the request is "purpose of the driving test." Note that the check boxes are not a required component. Even without the check boxes, it is possible to determine whether the content of the request (natural language) is "purpose of the driving test" by using classification in natural language processing.

[0045] The trajectory determination unit 16 determines the shape of the tool trajectory from the purpose (natural language) of the driving test. The shape determination uses, for example, machine learning. FIG. 16 shows an example of a machine learning model. The model inputs natural language text and outputs the shape of the trajectory. The model uses, for example, text classification. In text classification, the text input in natural language is subjected to morphological analysis, vectorized, and classified into qualitative variables. Here, the qualitative variables are variables that categorize the purpose. The qualitative variables may be numbers, letters, or words. The trajectory determination unit 16 determines the shape of the trajectory from the categorized purpose. The shape of the trajectory in the driving test is related to the purpose of the driving test. Therefore, once the purpose is determined, the shape can be determined. In other words, even if various natural language inputs such as "a trajectory suitable for checking corner shape errors," "please create a machining program for a driving test of part corner shape errors," or "I want to check corner shape errors" are input, they are categorized into a set of similar purposes. Once the purpose is determined, the shape associated with the purpose can be determined. The model may be supervised or unsupervised learning. The model is assumed to have sufficiently learned a natural language dataset that includes at least sentences expressing the purpose. Furthermore, while the above-described model only inputs the purpose (natural language), machine information obtained from a numerical control device or a user's behavior history may also be input as features to classify the purpose. Furthermore, a prediction or discrimination model may be used instead of classification.

[0046] The preprocessing unit 2 creates a prompt based on the shape of the trajectory output by the model. For example, if the shape is classified as a square, a prompt such as "square with a side length of 150 mm" is created. The side length may be fixed, or may be calculated from the machine tool's machine information, as shown in the seventh embodiment.

[0047] The transmission unit 3 transmits the prompt created by the pre-processing unit 2 to the interactive response device. The reception unit 4 acquires the machining program created by the generation AI. The post-processing unit 5 extracts only the machining program from the response of the generation AI. The program transmission unit 12 transmits the machining program to the numerical control device. The label acquisition unit 17 accepts the user's evaluation of the machining program. The log collection unit 18 records the user's evaluation as a label, linking it with the purpose of the driving test (natural language), the machining program, and machine information. The data recorded in the log collection unit 18 can be used as useful feedback. For example, it can be used to understand user behavior, as data for model learning, and as evaluation data.

[0048] 17 is a block diagram of a program creation device 100 according to a seventh embodiment. The program creation device 100 includes a request acquisition unit 1, a preprocessing unit 2, a transmission unit 3, a reception unit 4, a postprocessing unit 5, a display unit 6, and a machine information acquisition unit 7.

[0049] In the seventh embodiment, the size of the shape is determined based on machine information. For example, as shown in Figure 18, when a command for "square of appropriate size" is input, the shape is specified but the size is not.

[0050] The machine information acquisition unit 7 acquires machine information from the numerical control device. The pre-processing unit 2 acquires the movable range (amount of movement) of the tool from the machine information. The pre-processing unit 2 adds the movable range to the prompt as a constraint. In the example of FIG. 18 , the movable range (amount of movement) of the tool is "X axis: 300 mm," "Y axis: 400 mm," and "Z axis: 300 mm." The pre-processing unit adds the text "In addition, the movable range (maximum amount of movement) of each axis of the machine tool is as follows..." as a constraint.

[0051] The movement range may be limited by the post-processing unit 5. In this case, the generation AI creates a machining program that moves the tool in a specified shape, and the post-processing unit 5 adjusts the machining program based on machine information such as the axis configuration and axis movement range. Correcting the machining program by the post-processing unit 5 eliminates the need to send information such as machine information to an external interactive response device, making it safer.

[0052] As described above, the program creation device 100 of the first embodiment creates a machining program to be used in an operational test for adjusting a motor from natural language commands. The pre-processing unit 2 creates effective prompts to elicit an expected response from the generation AI. The post-processing unit 5 extracts a machining program from the response of the generation AI. The display unit 6 displays the machining program extracted by the post-processing unit 5. The user can check the contents of the machining program to be used in the operational test. The created machining program may be stored in a memory or the like without being displayed on the display unit 6.

[0053] The program creation device 100 of the second embodiment includes a machine information acquisition unit 7 that acquires machine information of a numerical control device. A pre-processing unit 2 adds the machine information required for creating a machining program to a prompt. A post-processing unit 5 modifies the machining program by referring to the machine information. This allows a machining program suitable for the machine tool to be created.

[0054] The program creation device 100 of the third embodiment sends a prompt to the generation AI instructing it to provide an explanation of the post-processed machining program, and receives the explanation of the machining program as a response from the generation AI. The explanation display unit 11 displays the post-processed machining program and an explanation of the machining program. The explanation helps a user who is unfamiliar with machining programs to understand the machining program.

[0055] The program creation device 100 of the fourth embodiment transmits the created machining program to a numerical control device, where the received machining program can be manually corrected.

[0056] The program creation device 100 of the fifth embodiment acquires waveform data obtained in an operation test and adjusts parameters of a numerical control device, thereby automating processes from creation of a machining program to adjustment of parameters.

[0057] The program creation device 100 of the sixth embodiment acquires the shape of the tool trajectory by inputting the purpose of the driving test into a trained learning model that has undergone machine learning to output the tool trajectory shape for the purpose of the driving test, and transmits the acquired shape to the generation AI to receive the machining program. This allows even a user who is unfamiliar with driving tests to create a machining program simply by inputting the purpose of the driving test.

[0058] Furthermore, the program creation device 100 of the sixth embodiment can collect learning data, evaluation data, user behavior, and the like that record user evaluations of machining programs as logs, and use the collected data to improve the model.

[0059] The program creation device 100 of the seventh embodiment acquires machine information from a numerical control device and adds the tool movement range as a constraint to the prompt. This allows the user to automatically determine the size of the shape and create a machining program simply by inputting the shape of the trajectory and the purpose of the test.

[0060] The hardware configuration of a program creation device 100 to which the present disclosure is applied will be described below. Fig. 19 is a hardware configuration diagram of the program creation device 100. As shown in Fig. 19, the program creation device 100 includes a CPU 111 that controls the entire program creation device 100, a ROM 112 that records programs and data, and a RAM 113 for temporarily expanding data. The CPU 111 reads out a system program recorded in the ROM 112 via a bus and determines the consistency between a block and a tool.

[0061] The nonvolatile memory 114 is backed up by, for example, a battery (not shown), and the stored state is maintained even when the power to the program creation device 100 is turned off. The nonvolatile memory 114 stores various data such as programs read from the external device 120 via the interfaces 115, 118, and 119 and operation inputs input via the input device 20. The nonvolatile memory 114 may store programs and data for executing the program creation device 100 of this embodiment.

[0062] The interface 115 is an interface for connecting the program creation device 100 to an external device 120 such as an adapter. Programs, various parameters, etc. are loaded from the external device 120. The interface 118 is an interface for connecting the program creation device 100 to a display device 30 such as a liquid crystal display. The display device 30 displays various data loaded into memory, data obtained as a result of executing programs, etc. The interface 119 is an interface for connecting the program creation device 100 to an input device 20 such as a keyboard or pointing device. The input device 20 passes commands, data, etc. based on operations by an operator to the CPU 111 via the interface 119.

[0063] Although the present disclosure has been described in detail, the present disclosure is not limited to the individual embodiments described above. Various additions, substitutions, modifications, partial deletions, etc. are possible in these embodiments without departing from the gist of the present disclosure or the gist of the present disclosure derived from the claims and their equivalents. Furthermore, these embodiments can also be implemented in combination. For example, in the above-described embodiments, the order of each operation and the order of each process are shown as examples and are not limited to these.

[0064] The following are supplementary notes related to embodiments of the present disclosure. (Supplementary Note 1) A program creation device (100) includes a request acquisition unit (1) that receives a request for creating a machining program to be used in an operation test in natural language, a pre-processing unit (2) that creates a prompt by adding instructions for creating the machining program to the request, a transmission unit (3) that transmits the prompt to an interactive response device, a reception unit (4) that receives a response from the interactive response device, a post-processing unit (5) that extracts a machining program from the response, and a display unit (6) that displays the machining program. (Supplementary Note 2) The program creation device (100) includes a machine information acquisition unit (7) that acquires machine information from a numerical control device, and the pre-processing unit (2) adds information necessary for creating the machining program for the operation test, which is included in the machine information, to the prompt. (Supplementary Note 3) The program creation device (100) includes a machine information acquisition unit (7) that acquires machine information from a numerical control device, and the post-processing unit (5) modifies the machining program by referring to the machine information. (Supplementary Note 4) The machine information acquired by the machine information acquisition unit (7) is axis information. (Supplementary Note 5) The program creation device (100) includes an explanation request sending unit (9) that sends a prompt to an interactive response device, to which an instruction requesting an explanation of a machining program is added, an explanation receiving unit (10) that receives an explanation of the machining program from the interactive response device, and an explanation display unit (11) that displays the explanation of the machining program. (Supplementary Note 6) The program creation device (100) includes a program sending unit (12) that sends a machining program to a numerical control device, a waveform data acquiring unit (13) that acquires waveform data obtained by executing the machining program from the numerical control device, an adjustment unit (14) that adjusts parameters of the numerical control device based on the waveform data, and a parameter sending unit (15) that sends the adjusted parameters to the numerical control device.(Supplementary Note 7) The program creation device (100) includes a trajectory determination unit (16) that, when receiving a request in natural language to create a machining program for the driving test, determines the shape of a tool trajectory corresponding to the purpose by inputting the natural language into a trained model that has undergone machine learning to classify the purpose of the driving test from the natural language, and the pre-processing unit (2) creates a prompt that instructs the creation of a machining program that moves the tool to the shape output by the model. (Supplementary Note 8) The program creation device (100) includes a label acquisition unit (17) that receives a user's evaluation of the machining program, and a log collection unit (18) that links the user's evaluation as a label and records the purpose of the driving test, the machining program, and machine information. (Supplementary Note 9) The pre-processing unit (2) acquires the tool's range of motion from the machine information and adds the tool's range of motion to the prompt as a constraint. (Supplementary Note 10) A method for creating a machining program includes the steps of: accepting a request for creating a machining program to be used in a driving test in a natural language; creating a prompt by adding instructions for creating the machining program to the request; transmitting the prompt to an interactive response device; receiving a response from the interactive response device; extracting a machining program from the response; and displaying the machining program. (Supplementary Note 11) A computer-readable storage medium stores instructions for causing one or more processors to execute processes of accepting a request for creating a machining program to be used in a driving test in a natural language, creating a prompt by adding instructions for creating the machining program to the request, transmitting the prompt to an interactive response device, receiving a response from the interactive response device, extracting a machining program from the response, and displaying the machining program.

[0065] 100 Program creation device 1 Request acquisition unit 2 Pre-processing unit 3 Transmission unit 4 Reception unit 5 Post-processing unit 6 Display unit 7 Machine information acquisition unit 8 Explanation request pre-processing unit 9 Explanation request transmission unit 10 Explanation reception unit 11 Explanation display unit 12 Program transmission unit 13 Waveform data acquisition unit 14 Adjustment unit 15 Parameter transmission unit 16 Trajectory determination unit 17 Label acquisition unit 18 Log collection unit 111 CPU 112 ROM 113 RAM 114 Non-volatile memory

Claims

1. A program creation device comprising: a request acquisition unit that receives a request in natural language to create a machining program to be used in an operation test; a pre-processing unit that creates a prompt that adds instructions for creating the machining program to the request; a transmission unit that transmits the prompt to an interactive response device; a reception unit that receives a response from the interactive response device; a post-processing unit that extracts the machining program from the response; and a display unit that displays the machining program.

2. A program creation device as described in claim 1, further comprising a machine information acquisition unit that acquires machine information from a numerical control device, and wherein the pre-processing unit adds information contained in the machine information that is necessary for creating a machining program for the operation test to a prompt.

3. A program creation device according to claim 1, further comprising a machine information acquisition unit that acquires machine information from a numerical control device, wherein the post-processing unit modifies the machining program by referring to the machine information.

4. A program creation device according to claim 2 or 3, wherein the machine information acquired by the machine information acquisition unit is axis information.

5. A program creation device as described in claim 1, comprising: an explanation request sending unit that sends a prompt to an interactive response device with an instruction added requesting an explanation of a machining program; an explanation receiving unit that receives an explanation of the machining program from the interactive response device; and an explanation display unit that displays the explanation of the machining program.

6. A program creation device as described in claim 1, comprising: a program transmission unit that transmits a machining program to a numerical control device; a waveform data acquisition unit that acquires waveform data obtained by executing the machining program from the numerical control device; an adjustment unit that adjusts parameters of the numerical control device based on the waveform data; and a parameter transmission unit that transmits the adjusted parameters to the numerical control device.

7. A program creation device as described in claim 1, further comprising a trajectory determination unit that, when a request to create a machining program for the driving test is received in natural language, determines the shape of a tool trajectory corresponding to the purpose by inputting the natural language into a trained model that has undergone machine learning to classify the purpose of the driving test from the natural language, and the pre-processing unit creates a prompt that instructs the creation of a machining program that moves the tool to the shape output by the model.

8. A program creation device as described in claim 7, comprising: a label acquisition unit that receives a user's evaluation of the machining program; and a log collection unit that records the user's evaluation as a label, linking it to the purpose of the operation test, the machining program, and machine information.

9. The program creation device according to claim 1, wherein the preprocessing unit acquires a tool movement range from machine information and adds the tool movement range to the prompt as a constraint.

10. A method for creating a machining program, comprising: a step of receiving a request in natural language to create a machining program to be used in an operation test; a step of creating a prompt that adds machining program creation to the request; a step of sending the prompt to an interactive response device; a step of receiving a response from the interactive response device; a step of extracting the machining program from the response; and a step of displaying the machining program.

11. A computer-readable storage medium storing instructions to cause one or more processors to execute the following processes: accept a request in natural language to create a machining program to be used in an operation test; create a prompt that adds instructions for creating the machining program to the request; send the prompt to an interactive response device; receive a response from the interactive response device; extract a machining program from the response; and display the machining program.

Images