Prompt word generation device, prompt word generation method, and program

Abstract

The prompt word generating device includes: an input information acquisition unit that acquires input information; an information search unit that searches for information of an industrial machine, i.e., industrial machine information, associated with the input information from a storage unit that stores industrial machine information; and a prompt word generation unit that generates a prompt word to be sent to the AI generating device based on a search result of the information search unit.

Description

Technical Field

[0001] This disclosure relates to a prompt word generation apparatus, a prompt word generation method, and a program. Background Technology

[0002] Japanese Patent Application Publication No. 2023-058269 disclosed an information processing system that uses a chatbot to answer customer inquiries. Summary of the Invention

[0003] The goal is to develop a technology that can better assist users with questions and concerns related to industrial machinery.

[0004] The purpose of this invention is to solve the above-mentioned problems.

[0005] The first aspect of this disclosure is a prompt word generation device, comprising: an input information acquisition unit that acquires input information; an information retrieval unit that retrieves information about industrial machinery associated with the input information, i.e., industrial machinery information, from a storage unit storing industrial machinery information; and a prompt word generation unit that generates a prompt word to be sent to a generation AI device based on the retrieval result of the information retrieval unit.

[0006] The second aspect of this disclosure is a prompt word generation method executed by a computer, the prompt word generation method comprising: an input information acquisition step, acquiring input information; an information retrieval step, retrieving information about industrial machinery associated with the input information, i.e., industrial machinery information, from a storage unit storing industrial machinery information; and a prompt word generation step, generating a prompt word to be sent to a generating AI device based on the retrieval result of the information retrieval step.

[0007] The third aspect of this disclosure is a program for causing the computer to execute the prompt word generation method of the second aspect. Attached Figure Description

[0008] Figure 1 This is a diagram illustrating the structure of a prompt generation system as described in one implementation.

[0009] Figure 2A This is a diagram illustrating a fixed question. Figure 2B This is a diagram illustrating a fixed question.

[0010] Figure 3A It is a diagram illustrating one of several prototypes. Figure 3B This is an example and Figure 3A Diagrams of different prototypes.

[0011] Figure 4A This is a diagram illustrating prompts generated by the prompt generation unit. Figure 4B This is an example and Figure 4A Images with different prompts.

[0012] Figure 5 This is an illustration of a display showing the prompt words and the response of the AI ​​device to the prompt words.

[0013] Figure 6 This is a flowchart of a prompt word generation method according to one embodiment.

[0014] Figure 7 This is a diagram showing multiple standardized questions displayed in the display section of Example 1.

[0015] Figure 8 This is a schematic diagram of the prompt generation device for variation example 2.

[0016] Figure 9 It is a time-series curve of the injection pressure change during the injection process of an injection molding machine.

[0017] Figure 10 This is a diagram showing the structure of a part of the prompt generation system in variation example 3.

[0018] Figure 11 This is a diagram representing multiple industrial machines in variation 4.

[0019] Figure 12 This is a diagram showing the management device and multiple industrial machines in variation 5. Detailed Implementation

[0020] As a solution to address user questions and concerns related to industrial machinery, the use of generative artificial intelligence models can be considered. For example, generative AI models could be used to generate responses to user inquiries related to industrial machinery.

[0021] However, to obtain appropriate responses from generative AI models, it is necessary to provide them with suitable prompts that correspond to the content of the question. Creating appropriate prompts can be challenging for users of industrial machinery. For example, sometimes users want to ask the generative AI model about the cause of a malfunction in their industrial machinery. In such cases, users are required to create prompts that appropriately teach the generative AI model information related to the malfunction, including the parameters set for the machinery and its operational performance. Selecting the necessary information for asking questions related to industrial machinery and creating appropriate prompts is a significant burden for users.

[0022] Based on the above preliminary description, one embodiment will be described below.

[0023] (One implementation method)

[0024] Figure 1 This is a configuration diagram of a prompt word generation system 10 according to one embodiment.

[0025] The prompt word generation system 10 generates prompt words (PRT) containing questions related to industrial machinery 100 (see also...). Figure 4A , Figure 4B The computer system is a generator (computer system). Industrial machinery 100 is, for example, an injection molding machine 102, but could also be a robot, machine tool, etc. Additionally, the prompt word PRT is an AI prompt word used to ask questions to the generator AI device 200. The generator AI device 200 is a computer (computer system) that responds to the prompt word PRT using a generative artificial intelligence model. This generative artificial intelligence model includes a text-generating artificial intelligence model that generates text based on the prompt word PRT. The text-generating artificial intelligence model used by the generator AI device 200 can also be a conversational generation model. A conversational generation model is a text-generating artificial intelligence model capable of generating text that mimics the style of human spoken language.

[0026] like Figure 1 As shown, the prompt word generation system 10 includes: an industrial machinery information storage unit (storage unit) 12, one or more user input terminals 14, and a prompt word generation device 16.

[0027] The industrial machinery information storage unit 12 is implemented using one or more storage media (memory) accessible by the arithmetic unit 26 (information retrieval unit 38) included in the prompt word generation device 16. A more detailed description of the prompt word generation device 16 will be provided later. Figure 1 The text refers to a control device 104, for example, that controls industrial machinery 100. The control device 104 is, for example, a computer such as a numerical control (CNC) device. The memory (not shown) included in the control device 104 can also realize the industrial machinery information storage unit 12. Furthermore, in... Figure 1 The computer, or management device 300, used to manage multiple industrial machines 100 is shown in the figure. The memory (not shown) included in the management device 300 can also implement the industrial machine information storage unit 12. Alternatively, the prompt generation system 10 can include a dedicated storage device (not shown, such as a storage server) for implementing the industrial machine information storage unit 12. At least a portion of the storage unit 28 of the prompt generation device 16, described later, can also implement the industrial machine information storage unit 12.

[0028] The industrial machinery information storage unit 12 stores industrial machinery information. Industrial machinery information is information associated with industrial machinery 100. Therefore, information associated with, for example, injection molding machine 102 is included in the industrial machinery information. Industrial machinery information includes at least one of setting information, operation history information, and specification information.

[0029] The setting information refers to the settings of the industrial machinery 100. The setting information includes operating conditions and parameters that can be specified by the user to operate the industrial machinery 100. More specifically, for example, the molding conditions set by the control device 104 of the injection molding machine 102 are included in the setting information. Molding conditions include, but are not limited to, the nozzle temperature of the injection molding machine 102, the injection speed of the raw material in the injection process, the holding time in the holding process, and the holding pressure in the holding process. Furthermore, the nozzle is located at the top end of the injection barrel 102a of the injection molding machine 102.

[0030] Operation history information represents the operation history of the industrial machinery 100. For example, when the industrial machinery 100 operates according to the aforementioned operation conditions and parameters, information representing various physical quantities detected by the various sensors equipped with the industrial machinery 100 is included in the operation history information. More specifically, for example, when the injection molding machine 102 performs an injection process, the maximum pressure of the raw material detected by a pressure sensor (not shown) equipped with the injection molding machine 102 is included in the operation history information. Furthermore, the history of changes to the aforementioned setting information (operation conditions, parameters, etc.) can also be included in the operation history information. The history of alarms output in cases of abnormalities occurring in the industrial machinery 100 can also be included in the operation history information.

[0031] Specification information refers to information indicating the specifications of the industrial machinery 100. For example, information indicating the model number, serial number, etc., of the industrial machinery 100 may also be included in the specification information. Additionally, information indicating the design specifications of the industrial machinery 100 may also be included in the specification information. In this case, the specification information may include, for example, information indicating the screw diameter of the screw provided by the injection molding machine 102, but is not limited to this. The contents of the instruction manual for the industrial machinery 100 may also be included in the specification information.

[0032] The industrial machinery information may also include peripheral equipment information. Peripheral equipment information is information related to the peripheral equipment of the industrial machinery 100. Examples of peripheral equipment include, for instance, a temperature controller (not shown) and a hot runner, which are provided with the mold 106 of the injection molding machine 102. In this case, the peripheral equipment information may include, for example, temperature information of the temperature controller, hot runner, etc., detected by a temperature sensor (not shown). The temperature information of the hot runner may also be detected by a controller, i.e., a hot runner controller, provided with the control circuit that controls the hot runner. However, the peripheral equipment information is not limited to this temperature information. Furthermore, the peripheral equipment of the injection molding machine 102 is not limited to the temperature controller and the hot runner. For example, a resin dryer may be included among the peripheral equipment of the injection molding machine 102. A resin dryer is a dryer used to pre-dry the raw material (resin) supplied to the barrel provided with the injection molding machine 102. In this case, the peripheral equipment information may include, for example, parameters indicating the target dryness of the resin by the resin dryer.

[0033] User input terminal 14 is a terminal device (computer) for receiving information input from a user into prompt generation system 10. The control device 104 of the aforementioned industrial machinery 100 can be used as user input terminal 14. A PC (Personal Computer) 400 or a smart device (tablet terminal, etc.) 402 can also be used as user input terminal 14. Figure 1 As shown, the user input terminal 14 includes a communication unit 18, a display unit 20, and an operation unit 22.

[0034] The communication unit 18 includes a communication module (not shown) connected to the network NW. The network NW includes, for example, a LAN (Local Area Network) constructed within a factory where the industrial machinery 100 is installed. The network NW may also include the Internet.

[0035] Display unit 20 includes an output display device (not shown) that can be visually recognized by a user. This output display device may include, for example, a display panel (not shown) and display elements that form a display image (screen). The display panel may be, for example, a liquid crystal panel, but is not limited thereto.

[0036] The operation unit 22 includes an input device (not shown). This input device includes an operation panel, a pointing device, a keyboard, a mouse, etc. The pointing device includes, for example, a touch panel disposed on the display unit 20 along with the display panel described above, but is not limited to this. Furthermore, the input device provided in the user input terminal 14 may also include a voice input device. This voice input device includes, for example, a microphone.

[0037] The prompt generation device 16 is a computer that generates prompt words (PRTs) based on the information input to the prompt generation system 10, i.e., the input information. For example... Figure 1As shown, the prompt word generation device 16 includes a communication unit 24, an arithmetic unit 26, and a storage unit 28.

[0038] The communication unit 24 includes a communication module (not shown) connected to the network NW. Thus, the communication unit 24 can communicate with the communication unit 18 (user input terminal 14) via the network NW.

[0039] The arithmetic unit 26 includes processing circuitry (not shown). This processing circuitry may include, for example, one or more processors such as CPUs (Central Processing Units) or GPUs (Graphics Processing Units). It may also include integrated circuits such as ASICs (Application Specific Integrated Circuits) or FPGAs (Field-Programmable Gate Arrays). Alternatively, it may include discrete components.

[0040] Storage unit 28 includes one or more memory units. These memory units may include non-volatile memory. Non-volatile memory is a storage medium that non-temporarily stores programs, tables, mappings, etc. Examples include ROM (Read Only Memory) and flash memory. Storage unit 28 (one or more memory units) may also include volatile memory. For example, RAM (Random Access Memory) may be included in volatile memory. At least a portion of storage unit 28 (non-volatile memory) may also be implemented using storage media such as USB (Universal Serial Bus) memory, memory cards, optical discs, and magnetic disks.

[0041] The arithmetic unit 26 includes a display processing unit 30, an input information acquisition unit 32, a keyword extraction unit 34, a search scope determination unit 36, an information retrieval unit 38, a prototype selection unit 40, a prompt word generation unit 42, an editing and receiving unit 44, and a prompt word sending unit 46. The display processing unit 30, input information acquisition unit 32, keyword extraction unit 34, search scope determination unit 36, information retrieval unit 38, prototype selection unit 40, prompt word generation unit 42, editing and receiving unit 44, and prompt word sending unit 46 are implemented by the arithmetic unit 26 (processor) executing a program stored in the storage unit 28 (memory). At least a portion of the display processing unit 30, input information acquisition unit 32, keyword extraction unit 34, search scope determination unit 36, information retrieval unit 38, prototype selection unit 40, prompt word generation unit 42, editing and receiving unit 44, and prompt word sending unit 46 can also be implemented using the aforementioned ASIC, FPGA, discrete devices, etc.

[0042] The display processing unit 30 executes a function to display one or more typed questions 48 in the display unit 20 (see also...). Figure 2A , Figure 2B The information representing the content of the specification question 48 can be pre-stored by the storage unit 28. The display processing unit 30 sends the information representing the specification question 48 to the user input terminal 14 via the communication unit 24. As a result, the user input terminal 14 can display the specification question 48 on the display unit 20. Furthermore, the specific content of the specification question 48 can be predetermined, for example, by the manufacturer of the industrial machinery 100.

[0043] Figure 2A This is a diagram illustrating the definite question 48 (481). Figure 2A Example 48 is a type-setting question related to injection molding machine 102.

[0044] The predefined question 48 includes user-specific questions and confirmations for determining the inquiry to be made to the AI ​​generation device 200. The user answers the predefined question 48 via the operation unit 22 of the user input terminal 14. For example... Figure 2A As shown, the display processing unit 30 may display option 50 in the display unit 20 along with the predefined question 48. In this case, the user can also answer the predefined question 48 by selecting option 50 displayed along with it. The display processing unit 30 may also display radio buttons, checkboxes, drop-down lists, etc., for selecting option 50 in the display unit 20.

[0045] Figure 2B This is a diagram illustrating the 48th type of question. Figure 2B The example in Figure 2A The fixed question 481 is different from multiple fixed questions 48 (fixed question 482, fixed question 483, fixed question 484).

[0046] The display processing unit 30 can also display one or more additional standardized questions 48 on the display unit 20 based on the user's answer to the standardized question 48. For example, in Figure 2B Example in the selection Figure 2A In the case of option 50, "Want to know the countermeasures for poor molding", three molding questions 48 can be displayed on the display unit 20. That is, in Figure 2BThe display processing unit 30 can also display, along with the type of molding defect (e.g., type 482), the molding defect period (e.g., period 106), and the molding conditions (e.g., type 484). For type 482, the user answers the type of molding defect that occurred in the injection molding machine 102. For type 483, the user can answer the number of molding cycles (cycle range) during which the molding defect occurred, or the time (period) during which the molding defect occurred. The display processing unit 30 can also display, along with type 483, one or more numerical input fields 51, drop-down lists, etc., for the user to answer the period. For type 483, the user can also answer the number of cycles most recently performed by the injection molding machine 102. For type 484, the user answers, for example, the type of raw material (resin, etc.) of the molded product, the type of the molded product (mold 106), and the type of screw provided by the injection molding machine 102.

[0047] The display processing unit 30 can also display the text input field 52 together with the question-and-answer session 48. Figure 2B In this case, the user can also answer the predefined question 48 by entering a (freehand) string (question) into the text input field 52. Furthermore, the question can also be created based on voice information input to the user input terminal 14 via the aforementioned voice input device.

[0048] Input Information Acquisition Unit 32 ( Figure 1 The input information acquisition unit 32 can acquire user response information from the user input terminal 14 via the communication unit 24. The input information may include user response information. User response information represents the user's answer to one or more predefined questions 48 prompted to the user by the display processing unit 30. Examples include the selection result of option 50, the query statement entered into the text input field 52, and voice information entered into the user input terminal 14 via the voice input device. The input information acquisition unit 32 can acquire user response information from the user input terminal 14 via the communication unit 24. When the voice information is included in the user response information, the input information acquisition unit 32 can also generate a query statement based on that voice information.

[0049] The keyword extraction unit 34 extracts predetermined keywords from the query statement contained in the input information (user response information). For example, a keyword list representing multiple keywords determined based on the content of the typed question 48 is pre-stored by the storage unit 28. The keyword extraction unit 34 extracts keywords contained in the query statement by comparing the keyword list with the query statement. More specifically, for example, keywords related to the user's response to the question. Figure 2BThe type selection question 484 inputs a query statement containing information such as the type of raw material for the molded article, the type of the molded article, and the type of screw provided by the injection molding machine 102. In this case, the keyword extraction unit 34 uses a keyword list to extract keywords indicating the type of raw material for the molded article, the type of the molded article, and the type of screw from the query statement.

[0050] The search scope determination unit 36 ​​determines the search scope of industrial machinery information based on the input information acquired by the input information acquisition unit 32. For example, as described above, the user can select an answer to the specific question 48 via option 50. Option 50 and the types of industrial machinery information associated with that option 50 can be pre-associated. For example, for… Figure 2B One of the options 50 for the shaping question 482, namely "burr," can be pre-associated with injection speed, holding pressure, etc., as industrial machinery information (setting information) related to burrs (flash), a type of molding defect. Therefore, when the user selects "burr" as the answer to the shaping question 482, the search scope determination unit 36 ​​can include setting information such as injection speed and holding pressure in the search scope. Furthermore, for example, regarding... Figure 2B One of the options 50 for the type approval question 482, namely "shortage," can be pre-associated with nozzle temperature, injection speed, etc., as industrial machinery information (setting information) related to one type of molding defect, namely shortage (insufficient material in the molded part). Therefore, when the user selects "shortage" as the answer to the type approval question 482, the search scope determination unit 36 ​​can include setting information such as nozzle temperature and injection speed in the search scope. Furthermore, the search scope determination unit 36 ​​can also determine the search scope based on user instructions. For example, the user can arbitrarily specify at least one of the following information via the operation unit 22: setting information, operation history information, and specification information. In this case, the search scope determination unit 36 ​​can also include at least one of the following information specified by the user: setting information, operation history information, and specification information, in the search scope.

[0051] The search scope determination unit 36 ​​can also determine the time range of the industrial machinery information to be searched based on the input information. For example, the search scope determination unit 36 ​​determines the time range of the industrial machinery information to be searched based on the user's input information. Figure 2B The retrieval range or period is determined by the answer to the typed question 483, and industrial machinery information corresponding to that retrieval range or period is included in the retrieval range. More specifically, the retrieval range determination unit 36 ​​includes, for example, setting information of the control device 104 set during that period, and operation history information indicating physical quantities detected during that period in the retrieval range.

[0052] The search scope determination unit 36 ​​can also determine the search scope of industrial machinery information based on the keywords extracted by the keyword extraction unit 34. For example, the search scope determination unit 36 ​​can also determine the search scope of industrial machinery information based on keywords indicating the type of raw material of the molded product, keywords indicating the type of the molded product, and keywords indicating the type of screw, etc. Keywords and industrial machinery information corresponding to those keywords can be associated in advance.

[0053] The information retrieval unit 38 retrieves industrial machinery information associated with the input information from the industrial machinery information storage unit 12. By performing the retrieval according to the retrieval scope determined by the retrieval scope determination unit 36, the information retrieval unit 38 can accurately retrieve industrial machinery information associated with the input information.

[0054] The prototype selection unit 40 automatically selects the prototype (prototype text) 54 of the prompt word PRT sent to the AI ​​generation device 200 from a plurality of pre-made prototypes 54. The prototype selection unit 40 selects the prototype 54 based on the input information.

[0055] Figure 3A This is a diagram illustrating one of the multiple prototypes 54 (prototype 541). Figure 3B It means and Figure 3A Diagrams of different prototypes 54 (prototype 542).

[0056] exist Figure 3A Examples in the middle are for Figure 2A If the answer (selection result) to the finalization question 481 is "I want to know the countermeasures for molding defects", then prototype 541 can be selected by the prototype selection unit 40. Additionally, in Figure 3B Examples in the middle are for Figure 2A If the answer (selection result) to the question 481 is "I want to troubleshoot the injection molding machine 102", then the prototype selection unit 40 can select a prototype 542. Among the multiple prototypes 54 (prototypes 541, 542), there may be one or more fixed statements 56 (fixed statements 561, 562, 563) predetermined for each prototype 54, and one or more blank spaces 58 (blank spaces 581, 582, 583). The fixed statements 56 are pre-stored by the storage unit 28. The blank spaces 58 correspond to industrial machinery information (search scope) retrieved by the information retrieval unit 38 and are pre-set in the middle of the fixed statements 56 or between multiple fixed statements 56. Furthermore, at least one of the multiple blank spaces 58 may also be an article 60 (see also...). Figure 4A , Figure 4B Marking the insertion point, etc.

[0057] Prompt word generation section 42 ( Figure 1The information retrieval unit 38 combines the search results of industrial machinery information with one or more fixed phrases 56 to generate a prompt word PRT to be sent to the AI ​​generation device 200. More specifically, the prompt word generation unit 42 generates the prompt word PRT to be sent to the AI ​​generation device 200 based on the search results of industrial machinery information from the information retrieval unit 38 and the selection results of prototype 54 by the prototype selection unit 40. In this case, the prompt word generation unit 42 can generate the prompt word PRT by editing the prototype 54 according to the search results of the information retrieval unit 38. For example, the prompt word generation unit 42 generates the prompt word PRT by inserting industrial machinery information retrieved by the information retrieval unit 38 into one or more empty fields 58.

[0058] Figure 4A This is a diagram representing the prompt PRT (prompt PRT1) generated by the prompt generation unit 42.

[0059] exist Figure 4A The example shows the user selecting "burrs" as the target. Figure 2B In the case of a response to the question 482, a prompt word PRT1 can be generated by the prompt word generation unit 42. The prompt word PRT1 contains a document 60 (document 60A) indicating that the molded article has burrs. Document 60A can be generated by referring to the user's... Figure 2B Insert the corresponding statement for the answer to question 482. Figure 3A The blank space 581 (fixed phrase 561) is used to create the prompt. Additionally, the prompt word PRT1 contains an article 60 (article 60B) representing industrial machinery information retrieved by the information retrieval unit 38. When the search scope determination unit 36 ​​determines the aforementioned time range based on the input information, the prompt word generation unit 42 includes the content of the action history information corresponding to that time range in the article 60B of the prompt word PRT.

[0060] Figure 4B It means and Figure 4A A diagram of different prompt words for PRT. In addition, Figure 4A The prompt words PRT and Figure 4B The prompt word PRT can use the same prototype 541 ( Figure 3A ) to generate.

[0061] exist Figure 4B The example shows a user selecting "Bill Default" as the target. Figure 2B In the case of a definitive question 482, a prompt word PRT2 can be generated by the prompt word generation unit 42. The prompt word PRT2 contains a document 60 (document 60C) indicating a defect in the molded article. Similar to document 60A, document 60C can be generated by communicating with the user regarding... Figure 2B Insert the corresponding statement for the answer to question 482. Figure 3AThe blank space 581 (fixed phrase 561) is used to create it. In addition, the prompt word PRT2 contains the article 60 (article 60D) that represents industrial machinery information retrieved by the information retrieval unit 38.

[0062] According to this embodiment, the retrieved industrial machinery information varies depending on the input information (user response information). Therefore, if compared... Figure 4A , Figure 4B As shown, the content of prompt word PRT1 and prompt word PRT2 generated using the same prototype 54 can be different.

[0063] Editorial Department 44 ( Figure 1 The operation unit 22 can handle editing of the prompt word PRT generated by the prompt word generation unit 42. In this case, the display processing unit 30 can also perform processing for displaying the prompt word PRT on the display unit 20. The user can edit the prompt word PRT via the operation unit 22.

[0064] The prompt word sending unit 46 sends the prompt word PRT to the AI ​​generation device 200 via the communication unit 24. The AI ​​generation device 200 then generates a response to the prompt word PRT. The prompt word sending unit 46 may send the prompt word PRT to the AI ​​generation device 200 based on a user's instruction, but it may also automatically send the generated prompt word PRT to the AI ​​generation device 200 (see also the modified example 10 described later).

[0065] The generative AI device 200 includes, for example, a conversational generation model. In this case, the generative AI device 200 can generate text (response statements) that mimic spoken language as a response to the prompt word PRT.

[0066] The response statement generated by the AI ​​generation device 200 is sent to the prompt word generation device 16 (communication unit 24) via the network NW. Upon receiving the response statement, the display processing unit 30 performs processing for displaying the response statement on the display unit 20. This allows the user to confirm the response statement.

[0067] Figure 5 This is a diagram illustrating the display unit 20 showing the prompt word PRT and the response of the AI ​​generation device 200 to the prompt word PRT.

[0068] exist Figure 5 The display section 20 shows multiple operation buttons 62 (edit acceptance button 621, send instruction button 622), a prompt display bar 64, and a response display bar 66. The multiple operation buttons 62 include an edit acceptance button 621 and a send instruction button 622.

[0069] When a user wants to edit the prompt word PRT, they can operate the edit acceptance button 621. By operating the edit acceptance button 621, the user can begin editing the prompt word PRT. When a user wants to send the prompt word PRT to the AI ​​generation device 200, they can operate the send instruction button 622. By operating the send instruction button 622, the user can send the prompt word PRT to the AI ​​generation device 200.

[0070] The prompt word PRT is displayed in the prompt word display bar 64. The user can edit the prompt word PRT displayed in the prompt word display bar 64. The response content (AI answer) of the AI ​​generation device 200 to the prompt word PRT is displayed in the answer display bar 66. Thus, the user can confirm the response content of the AI ​​generation device 200. Furthermore, by arranging the prompt word display bar 64 and the answer display bar 66, the user can easily compare the prompt word PRT and the AI ​​generation device 200's answer to the prompt word PRT.

[0071] Figure 6 This is a flowchart illustrating a prompt word generation method in one embodiment.

[0072] The above-mentioned prompt word generation device 16 can perform Figure 6 The method for generating prompt words is shown. Figure 6 The prompt word generation method shown can be executed by the arithmetic unit 26 (processor) executing a program stored in the storage unit 28 (memory). For example... Figure 6 As shown, the prompt word generation method includes an information input acceptance step S1, a keyword extraction step S2, a search scope determination step S3, an information retrieval step S4, a prototype selection step S5, a prompt word generation step S6, and an editing acceptance step S7.

[0073] The information input acceptance step S1 includes a typed question display processing step S1A and an input information acquisition step S1B. In the typed question display processing step S1A, the display processing unit 30 performs processing to display one or more typed questions 48 on the display unit 20 of the user input terminal 14. In the input information acquisition step S1B, the input information acquisition unit 32 acquires input information. The input information may include user response information representing the user's answer to one or more typed questions 48.

[0074] The question display processing step S1A and the input information acquisition step S1B can also be repeated as needed. For example, as described above, based on the user's answer to the question 48, one or more additional questions 48 can be displayed on the display unit 20. In this case, the question display processing step S1A and the input information acquisition step S1B are repeated as needed.

[0075] In keyword extraction step S2, keyword extraction unit 34 extracts keywords from the query statement contained in the input information. Furthermore, if the input information does not contain a query statement, keyword extraction step S2 can be skipped.

[0076] In the search scope determination step S3, the search scope determination unit 36 ​​determines the search scope based on the input information or keywords. In the information retrieval step S4, the information retrieval unit 38 retrieves industrial machinery information based on the determined search scope.

[0077] In the prototype selection step S5, the prototype selection unit 40 selects a prototype 54 of the prompt word PRT sent to the AI ​​generation device 200. The prototype selection unit 40 selects the prototype 54 corresponding to the input information from a plurality of prototypes 54. Furthermore, the prototype selection step S5 may be performed before the search scope determination step S3, or between the search scope determination step S3 and the information retrieval step S4. The prototype selection step S5 may also be performed in parallel with at least one of the search scope determination step S3 and the information retrieval step S4.

[0078] In the prompt word generation step S6, the prompt word generation unit 42 generates the prompt word PRT based on the search results of the information retrieval unit 38. The prompt word generation unit 42 can generate the prompt word PRT by editing the prototype 54 selected by the prototype selection unit 40 based on the search results of the information retrieval unit 38.

[0079] In the editing acceptance step S7, the editing acceptance unit 44 accepts the editing of the prompt word PRT. The user can edit the prompt word PRT via the operation unit 22. Alternatively, the user can choose not to edit the prompt word PRT.

[0080] so, Figure 6 The prompt word generation method has ended. As described above, the prompt word generation device 16 includes a prompt word sending unit 46. Therefore, the prompt word generation device 16 can send the generated prompt word PRT to the AI ​​generation device 200. Furthermore, the display processing unit 30 of the prompt word generation device 16 performs processing for displaying the response statement received from the AI ​​generation device 200 on the display unit 20. Therefore, the user can confirm the response statement.

[0081] The prompt word generation system 10 described above can, for example, achieve the effects described below.

[0082] The prompt word generation system 10 includes a prompt word generation device 16. The prompt word generation device 16 includes an information retrieval unit 38. The information retrieval unit 38 retrieves industrial machinery information associated with the input information. This reduces the workload for users who want to filter and select the most suitable industrial machinery information contained in the prompt word PRT.

[0083] Furthermore, the prompt generation device 16 includes a prompt generation unit 42. The prompt generation unit 42 generates prompts (PRTs) based on the search results from the information retrieval unit 38. This reduces the workload for users who want to create prompts (PRTs).

[0084] The industrial machinery information includes at least one of setting information, motion history information, and specification information. In this case, the prompt word generation unit 42 can generate a prompt word PRT that teaches at least one of the setting information, motion history information, and specification information to the generation AI device 200. By sending the prompt word PRT to the generation AI device 200, an accurate response to a user's question related to the industrial machinery 100 can be easily obtained from the generation AI device 200.

[0085] Information about peripheral equipment may also be included in the information about industrial machinery. In this case, the prompt word generation unit 42 can generate a prompt word PRT that teaches the peripheral equipment information to the generation AI device 200. By sending this prompt word PRT to the generation AI device 200, it is easier to obtain an accurate response from the generation AI device 200 to user questions related to the industrial machinery 100.

[0086] The prompt word generation device 16 includes a search scope determination unit 36. The search scope determination unit 36 ​​determines the search scope based on the input information. As described above, by performing a search based on the search scope, the information retrieval unit 38 can accurately retrieve industrial machinery information associated with the input information. As a result, the prompt word generation unit 42 can generate prompt words PRT that include, for example, industrial machinery information within a period (time range) determined based on the input information.

[0087] The prompt generation device 16 includes a display processing unit 30. The display processing unit 30 performs processing for displaying one or more predefined questions 48 on the display unit 20. The user can answer one or more predefined questions 48 displayed on the display unit 20 via the operation unit 22. Thus, user answer information indicating the user's answer to one or more predefined questions 48 can be acquired by the input information acquisition unit 32. Furthermore, the display unit 20 and the operation unit 22 can be installed in various computers such as the control device 104, PC 400, and smart device 402.

[0088] The answer format for at least one fixed question 48 can also be a selection format where the user chooses the answer from pre-determined options 50. Selection formats are preferred because they require less user input of text.

[0089] The answer to at least one fixed question 48 can also be a free-form statement in which the user inputs text into the text input field 52 via the operation unit 22. In this case, the question is created by the user inputting text via the operation unit 22. For this question, the prompt word generation device 16 is equipped with a keyword extraction unit 34. The keyword extraction unit 34 extracts predetermined keywords from the question. As a result, the search scope determination unit 36 ​​can determine the search scope based on the keywords extracted by the keyword extraction unit 34. Consequently, the information retrieval unit 38 can accurately retrieve industrial machinery information associated with the keywords extracted by the keyword extraction unit 34.

[0090] The prompt word generation device 16 includes a prototype selection unit 40. The prototype selection unit 40 selects from a plurality of prototypes 54 the prototype of the prompt word PRT sent to the AI ​​generation device 200. The prompt word generation unit 42 generates the prompt word PRT based on the prototype 54 selected by the prototype selection unit 40. This reduces the quality deviation of the prompt word PRT for each user. In other words, according to this embodiment, even if the user of the industrial machinery 100 is not accustomed to creating AI prompt words, they can easily create prompt word PRTs that readily receive an appropriate response from the AI ​​generation device 200.

[0091] The prompt generation device 16 includes an editing and receiving unit 44. The editing and receiving unit 44 accepts the editing of the prompt PRT. The user can edit the prompt PRT via the operation unit 22. This provides convenience for the user.

[0092] One embodiment can also be modified as described below. In the following modifications, descriptions that are repeated in the embodiment are omitted as appropriate. In addition, in the figures used in the following modifications, the same reference numerals are given to the same configurations as those described in the embodiment.

[0093] (Variation Example 1)

[0094] Figure 7 This is a diagram showing the multiple type questions 48 displayed in the display section 20 of Modified Example 1.

[0095] like Figure 7 As shown, multiple pre-defined questions 48 and the user's respective answer history 67 for each of the multiple pre-defined questions 48 can also be displayed sequentially in the display unit 20 in the form of a chatbot. In this case, the display processing unit 30 performs processing for sequentially displaying the multiple pre-defined questions 48 and the multiple answer history 67 in the display unit 20 according to the chatbot program.

[0096] (Variation Example 2)

[0097] Figure 8This is a configuration diagram of the prompt generation device 16 (prompt generation device 16B) of Modified Example 2.

[0098] The input information acquisition unit 32 of the prompt generation device 16B automatically acquires the motion history information stored in the industrial machinery information storage unit 12. That is, the input information acquisition unit 32 automatically acquires the motion history information as input information. The input information acquisition unit 32 repeatedly acquires the motion history information, for example, according to a predetermined period.

[0099] In addition, the prompt word generation device 16B also includes an anomaly determination unit 68. The anomaly determination unit 68, for example, is implemented by the arithmetic unit 26 executing a program stored in the storage unit 28, similar to the input information acquisition unit 32. The anomaly determination unit 68 determines whether an anomaly has occurred in the industrial machinery 100 based on the action history information acquired by the input information acquisition unit 32 as input information.

[0100] Figure 9 This is a graph showing the time-series variation of injection pressure during the injection process of injection molding machine 102. The vertical axis of the graph represents the magnitude of injection pressure (MPa: megapascals). The horizontal axis of the graph represents time (s: seconds).

[0101] For example, the anomaly detection unit 68 determines whether an anomaly has occurred in the injection molding machine 102 based on whether the peak pressure P1 of the injection molding machine 102 deviates from the specified allowable range RNG. The peak pressure P1 is the maximum value of the injection pressure during the injection process. The injection pressure can be detected by a pressure sensor such as a force sensor. This pressure sensor is installed at the rear end of injection components such as screws and plungers included in the injection molding machine 102. Figure 9 The peak pressure P1 indicated in the figure deviates from the specified allowable range RNG. In this case, the anomaly determination unit 68 determines that an anomaly has occurred in the injection molding machine 102.

[0102] Furthermore, the determination criteria of the anomaly determination unit 68 can be appropriately determined. For example, the anomaly determination unit 68 can determine that an anomaly has occurred in the injection molding machine 102 if a molding cycle in which the peak pressure P1 deviates from the prescribed allowable range RNG has been performed more than a predetermined number of times. Additionally, as mentioned above, the industrial machinery 100 is not limited to the injection molding machine 102. The determination criteria of the anomaly determination unit 68 can be appropriately determined according to the type of industrial machinery 100.

[0103] When the anomaly determination unit 68 determines that an anomaly has occurred in the industrial machinery 100, the search scope determination unit 36 ​​determines the search scope of the industrial machinery information based on the content of the anomaly (the determination criteria of the anomaly determination unit 68). In this case, the search scope determination unit 36 ​​may also determine the time range based on the input information (operation history information). The search scope determination unit 36 ​​may also exclude operation history information that has already been obtained as input information from the search scope.

[0104] Based on the search scope determined by the search scope determination unit 36, the information retrieval unit 38 retrieves information on industrial machinery. Additionally, the prompt word generation unit 42 generates a prompt word PRT based on the search results from the information retrieval unit 38. The prompt word generation unit 42 can also generate a prompt word PRT based on action history information obtained as input information. The generated prompt word PRT is sent to the AI ​​generation device 200 by the prompt word sending unit 46.

[0105] According to this modified example, the prompt word generation device 16B (abnormality determination unit 68) determines whether an abnormality has occurred in the industrial machinery 100. Furthermore, according to this modified example, if the abnormality determination unit 68 determines that an abnormality has occurred in the industrial machinery 100, the prompt word generation unit 42 can automatically generate the prompt word PRT. Thus, the prompt word generation device 16B can assist the user.

[0106] (Variation Example 3)

[0107] Figure 10 This is a diagram showing the configuration of the prompt generation system 10 (prompt generation system 10C) of variant example 3.

[0108] The prompt generation system 10C includes a monitoring device (a specified monitoring device) 108. The monitoring device 108 is a device for monitoring the status of the industrial machinery 100. For example, the monitoring device 108 is an alarm device that issues an alarm based on the output signal of a sensor used to detect abnormalities in the industrial machinery 100. The monitoring device 108 can also be a quality inspection device for checking the quality of processed products manufactured by a machine tool, molded products formed by an injection molding machine 102, etc. For example, the quality inspection device checks the appearance of the molded product based on a signal output from an image sensor (camera) that captures an image of the molded product.

[0109] Based on its monitoring results, the monitoring device 108 outputs a predetermined signal to the prompt word generation device 16. For example, the alarm device described above outputs a signal indicating the alarm content to the prompt word generation device 16. Additionally, the quality inspection device described above outputs a signal indicating the defect of the molded product to the prompt word generation device 16 when it determines that the molded product is defective.

[0110] The input information acquisition unit 32 acquires a predetermined signal output from the monitoring device 108 as input information. In this case, the search scope determination unit 36 ​​determines the search scope of the industrial machinery information based on the acquired predetermined signal. For example, the search scope determination unit 36 ​​determines the search scope based on the type of alarm, the reason for determining that the molded product is defective, etc.

[0111] Based on the search scope determined by the search scope determination unit 36, the information retrieval unit 38 retrieves information on industrial machinery. Additionally, the suggestion term generation unit 42 generates suggestion terms PRT based on the search results from the information retrieval unit 38. The generated suggestion terms PRT are then sent by the suggestion term sending unit 46 to the AI ​​generation device 200.

[0112] According to this modification, when a specified signal is output from a monitoring device 108, which is an external device to the prompt word generation device 16, the prompt word generation unit 42 can automatically generate the prompt word PRT. Thus, the prompt word generation device 16 can assist the user. For example, in the event of an alarm, a pre-determined alarm message can be displayed to the user. According to this modification, not only is the alarm message provided to the user, but suggestions from the AI ​​generation device 200 can also be provided to the user.

[0113] Furthermore, the aforementioned alarm can also be output by the control device 104 of the industrial machinery 100. In other words, the control device 104 can also function as the monitoring device 108.

[0114] (Variation Example 4)

[0115] Figure 11 This is a diagram showing multiple industrial machines 100 (injection molding machine 102) of variant example 4.

[0116] The prompt word generation device 16 can also be located within the control device 104. In this case, the prompt word generation device 16 and the control device 104 are implemented as a single computer. The prompt word generation device 16 and the control device 104 can share hardware resources such as the processor and memory. Furthermore, the prompt word PRT generated by the control device 104 (prompt word generation device 16) is sent to the AI ​​generation device 200.

[0117] exist Figure 11The text indicates that the prompt word generation device 16 is installed in one of the multiple control devices 104, namely the first control device 104A. The prompt word generation device 16 installed in the first control device 104A can also be communicatively connected to control devices 104 (control devices 104B, 104C, ...) different from the first control device 104A. Thus, the user can cause the prompt word generation device 16 to generate the prompt word PRT via a control device 104 different from the first control device 104A. In this case, the control device 104 different from the first control device 104A functions as a user input terminal 14.

[0118] Alternatively, a prompt word generation device 16 can be installed in each of the multiple control devices 104. Furthermore, the prompt word generation device 16 can also be installed in the PC 400, smart device 402, etc. (see also...) Figure 1 )middle.

[0119] (Variation Example 5)

[0120] Figure 12 This is a diagram showing the management device 300 and multiple industrial machines 100 (injection molding machine 102) of Modified Example 5.

[0121] The management device 300 is a computer used to manage multiple industrial machines 100 (see also one embodiment). The prompt word generation device 16 may also be provided in the management device 300. Thus, the prompt word generation device 16 and the management device 300 are implemented as a single computer. In this case, the prompt word generation device 16 and the management device 300 can share hardware resources such as processors and memory. Furthermore, the prompt word PRT generated by the management device 300 (prompt word generation device 16) is sent to the AI ​​generation device 200. In this case, each of the multiple control devices 104 functions as a user input terminal 14.

[0122] (Variation Example 6)

[0123] If the answer to the typed question 48 does not contain an interrogative statement, the keyword extraction unit 34 may also be omitted. For example, if the answer to all typed questions 48 is in the form of a choice, the keyword extraction unit 34 may also be omitted.

[0124] (Variation Example 7)

[0125] The industrial machinery 100 includes machines that repeatedly execute predetermined routines (cycles). For example, the injection molding machine 102 can repeatedly execute predetermined molding cycles (see also one embodiment). In addition, machine tools can repeatedly execute predetermined processing cycles. Robots can repeatedly execute predetermined work routines.

[0126] Based on the above, the information retrieval unit 38 can also retrieve industrial machinery information for a predetermined number of cycles. For example, if the input information includes information indicating a certain time (period), the information retrieval unit 38 can also retrieve industrial machinery information for a period tracing back a predetermined number of cycles from that time. The predetermined number of cycles can be a predetermined fixed value (constant), or it can be specified via the operation unit 22.

[0127] (Variation Example 8)

[0128] There are cases where the AI ​​generation device 200 can access the industrial machinery information storage unit 12 based on access path information. Access path information refers to information such as the storage location of industrial machinery information in the information communication network and the access path to that storage location. For example, network addresses and network paths are included in the access path information. More specifically, the access path information may include, for example, a URL (Uniform Resource Locator), but is not limited to this.

[0129] The prompt generation unit 42 may also include access path information in the prompt word PRT. In this case, at least a portion of the specific content of the industrial machinery information retrieved by the information retrieval unit 38 may not be recorded in the prompt word PRT.

[0130] According to this variation, the AI ​​generation device 200 can access the industrial machinery information storage unit 12 by following the access path information represented by the prompt word PRT, and can then refer to industrial machinery information. For example, the prompt word generation unit 42 can include a URL for accessing the operation history information representing the peak pressure P1 in the prompt word PRT instead of the specific value of the peak pressure P1 of the injection molding machine 102 (see Variation 2). Thus, the AI ​​generation device 200 can refer to the peak pressure P1 corresponding to that URL.

[0131] (Variation Example 9)

[0132] The generative artificial intelligence models that can be used by the generative AI device 200 are not limited to text-generating artificial intelligence models (dialogue-based generative models). For example, the generative AI device 200 can also use image-generating artificial intelligence models that generate images, video-generating artificial intelligence models that generate videos, speech-generating artificial intelligence models that generate speech, etc., to respond to the prompt word PRT.

[0133] (Variation Example 10)

[0134] The prompt generation device 16 (display processing unit 30) may also omit displaying the generated prompt PRT on the display unit 20. For example, the editing reception unit 44 may also be omitted from the prompt generation device 16. In this case, it is unnecessary to display the prompt PRT on the display unit 20 so that the user can edit it. In this case, the display processing unit 30 may also omit displaying the prompt PRT generated by the prompt generation unit 42 on the display unit 20. Furthermore, if the editing reception unit 44 is omitted from the prompt generation device 16, the prompt sending unit 46 may automatically send the prompt PRT generated by the prompt generation unit 42 to the AI ​​generation device 200.

[0135] (A combination of multiple variations)

[0136] The aforementioned variations can also be appropriately combined.

[0137] Based on the above implementation method and variations, it is possible to effectively assist users with problems, questions, etc. related to industrial machinery 100.

[0138] Regarding the above-described embodiments, the following notes are further disclosed.

[0139] (Postscript 1)

[0140] The prompt word generation device 16 disclosed herein includes: an input information acquisition unit 32 that acquires input information; an information retrieval unit 38 that retrieves information about industrial machinery 100 associated with the input information, i.e., industrial machinery information, from a storage unit 12 storing industrial machinery information; and a prompt word generation unit 42 that generates a prompt word PRT to be sent to the AI ​​generation device 200 based on the retrieval result of the information retrieval unit.

[0141] (Postscript 2)

[0142] According to the prompt word generation device described in Appendix 1, the industrial machinery information may also include at least one of the following: setting information indicating the setting of the industrial machinery, operation history information indicating the operation history of the industrial machinery, and specification information indicating the specifications of the industrial machinery.

[0143] (Note 3)

[0144] According to the prompt word generation device described in Appendix 2, the prompt word generation unit can also generate prompt words containing the industrial machinery information with a predetermined number of cycles or the industrial machinery information within a period determined by the input information.

[0145] (Note 4)

[0146] According to the prompt word generation device described in Appendix 2 or 3, the industrial machinery information includes the motion history information, the input information acquisition unit automatically acquires the motion history information stored by the storage unit as the input information, and also includes an anomaly determination unit 68. The anomaly determination unit 68 determines whether an anomaly has occurred in the industrial machinery based on the motion history information acquired by the input information acquisition unit. If the anomaly determination unit determines that an anomaly has occurred in the industrial machinery, the prompt word generation unit can also automatically generate the prompt word.

[0147] (Note 5)

[0148] According to any one of the appendices 1 to 4, the prompt word generation device may also include information related to peripheral equipment of the industrial machinery, i.e., peripheral equipment information, in the industrial machinery information.

[0149] (Note 6)

[0150] The prompt word generation apparatus according to any one of Appendices 1 to 5 further includes a display processing unit 30, which performs processing for displaying one or more predetermined typed questions 48 on a display unit 20 that can be visually recognized by a user, and the input information may also include the user's answers to one or more of the typed questions.

[0151] (Note 7)

[0152] The prompt word generation apparatus according to any one of Appendices 1 to 6 includes an inquiry statement generated by text input via the operation unit 22 in the input information, and also includes a keyword extraction unit 34, which extracts predetermined keywords from the inquiry statement, and the information retrieval unit can also retrieve the industrial machinery information based on the keywords extracted by the keyword extraction unit.

[0153] (Postscript 8)

[0154] According to any one of the appendices 1 to 7, when a specified signal is output from a specified monitoring device 108, the input information acquisition unit may also acquire the specified signal as the input information.

[0155] (Note 9)

[0156] The prompt word generation apparatus according to any one of Appendices 1 to 8 may further include an editing receiving unit 44, which accepts editing of the prompt words generated by the prompt word generation unit.

[0157] (Postscript 10)

[0158] The prompt word generation device according to any one of Appendix 1 to 9 may also be provided in the control device 104 that controls the industrial machinery.

[0159] (Postscript 11)

[0160] The prompt word generation device according to any one of Appendix 1 to 10 may also be provided in the management device 300 for managing multiple industrial machines.

[0161] (Postscript 12)

[0162] According to any one of the appendices 1 to 11, the industrial machinery may also be an injection molding machine 102.

[0163] (Postscript 13)

[0164] According to any one of the appendices 1 to 12, the prompt word generation device may also include a dialogue-based generation model.

[0165] (Postscript 14)

[0166] The prompt word generation method disclosed herein is executed by a computer, and the prompt word generation method includes: an input information acquisition step S1B, acquiring input information; an information retrieval step S4, retrieving information about industrial machinery 100 associated with the input information from a storage unit 12 storing industrial machinery information; and a prompt word generation step S6, generating a prompt word PRT to be sent to the AI ​​generation device 200 based on the retrieval result of the information retrieval step.

[0167] (Postscript 15)

[0168] The program disclosed herein is a program for causing the computer to execute the prompt word generation method described in Appendix 14.

[0169] Although this disclosure has been described in detail, it is not limited to the individual embodiments described above. Various additions, substitutions, modifications, and partial deletions can be made to these embodiments without departing from the spirit of this disclosure or the spirit derived from the claims and their equivalents. Furthermore, these embodiments can also be implemented in combination. For example, in the embodiments described above, the order of actions and the order of processes are shown as examples and are not limited to. The same applies to instances where numerical values ​​or formulas are used in the description of the embodiments described above.

[0170] Symbol Explanation

[0171] 12…Industrial Machinery Information Storage Department (Storage Department) 14…User Input Terminal

[0172] 16… Prompt word generation device 20… Display unit

[0173] 22…Operation Unit 30…Display Processing Unit

[0174] 32…Input Information Acquisition Department 34…Keyword Extraction Department

[0175] 38…Information Retrieval Department 40…Prototype Selection Department

[0176] 42… Prompt word generation department 44… Editing and processing department

[0177] 48…Forming Questions 54…Prototype

[0178] 68…Anomaly Detection Department 100…Industrial Machinery

[0179] 102…Injection molding machine 104…Control device

[0180] 108… Monitoring device (specified monitoring device) 200… AI generation device

[0181] 300… Management device PRT… prompt.

Claims

1. A prompt word generation device, characterized in that, have: The input information acquisition unit acquires input information. The information retrieval unit retrieves information about industrial machinery associated with the input information, i.e., industrial machinery information, from a storage unit containing industrial machinery information; and The prompt word generation unit generates prompt words to be sent to the AI ​​generation device based on the search results of the information retrieval unit.

2. The prompt word generation device according to claim 1, characterized in that, The industrial machinery information includes at least one of the following: setting information, operation history information, and specification information.

3. The prompt word generation device according to claim 2, characterized in that, The prompt word generation unit generates prompt words for the industrial machinery information containing a predetermined number of cycles or for the industrial machinery information within a period determined by the input information.

4. The prompt word generation device according to claim 2 or 3, characterized in that, The industrial machinery information includes the motion history information. The input information acquisition unit automatically acquires the action history information stored by the storage unit as the input information. The prompt word generation device also includes an anomaly determination unit, which determines whether an anomaly has occurred in the industrial machinery based on the action history information obtained by the input information acquisition unit. When the anomaly determination unit determines that an anomaly has occurred in the industrial machinery, the prompt word generation unit automatically generates the prompt word.

5. The prompt word generation device according to any one of claims 1 to 4, characterized in that, The industrial machinery information includes information related to the peripheral equipment of the industrial machinery, namely peripheral equipment information.

6. The prompt word generation device according to any one of claims 1 to 5, characterized in that, It also includes a display processing unit that performs processing for displaying one or more predetermined typed questions on a display unit that can be visually recognized by a user. The input information includes the user's answers to one or more of the typed questions.

7. The prompt word generation device according to any one of claims 1 to 6, characterized in that, The input information includes an inquiry statement created via text input through the operation unit. The prompt word generation device includes a keyword extraction unit, which extracts predetermined keywords from the query statement. The information retrieval unit retrieves the industrial machinery information based on the keywords extracted by the keyword extraction unit.

8. The prompt word generation device according to any one of claims 1 to 7, characterized in that, When a specified signal is output from a specified monitoring device, the input information acquisition unit acquires the specified signal as the input information.

9. The prompt word generation device according to any one of claims 1 to 8, characterized in that, It also includes an editing and receiving department, which accepts the editing of the prompts generated by the prompt generation department.

10. The prompt word generation apparatus according to any one of claims 1 to 9, characterized in that, It is installed in the control device that controls the industrial machinery.

11. The prompt word generation apparatus according to any one of claims 1 to 10, characterized in that, It is installed in a management device that manages multiple of the aforementioned industrial machines.

12. The prompt word generation apparatus according to any one of claims 1 to 11, characterized in that, The industrial machinery mentioned is an injection molding machine.

13. The prompt word generation apparatus according to any one of claims 1 to 12, characterized in that, The AI ​​generation device includes a conversational generation model.

14. A method for generating prompt words, which is executed by a computer. The prompt word generation method is characterized by having: Steps for obtaining input information: Obtain input information; The information retrieval step involves retrieving information about industrial machinery associated with the input information, i.e., industrial machinery information, from a storage unit containing industrial machinery information; and The prompt word generation step generates prompt words to be sent to the AI ​​generation device based on the retrieval results of the information retrieval step.

15. A program, characterized in that, Used to enable the computer to execute the prompt word generation method of claim 14.

Images