Information provision method and information provision device

Abstract

This provides an information delivery method that enables the generation of appropriate answers to user questions. [Solution] The information provision method performed by the information provision device 3 includes a question reception process that receives question information D1 including plant status data, a prompt generation process that generates prompt information D4 that instructs the generative artificial intelligence model 20 to output answer information D5 while referring to technical information D2A to D2C based on the question information D1, and an answer provision process that provides the answer information D5 output from the generative artificial intelligence model 20 by inputting the prompt information D4 to the generative artificial intelligence model 20. The prompt generation process and the answer provision process are repeated multiple times. In the first prompt generation process, prompt information D4 is generated to refer to technical information D2A, and in the second and subsequent prompt generation processes, prompt information D4 is generated to refer to technical information D2B and D2C, respectively.

Description

Technical Field

[0001] The present invention relates to an information providing method and an information providing apparatus.

Background Art

[0002] In recent years, generative artificial intelligence models called generative AI and generative AI have been utilized for various purposes and applications. For example, Patent Document 1 discloses an information processing apparatus that selects a database corresponding to a question input by a user based on the content of the question, generates a prompt based on the selected database and the question, and generates an answer by inputting the prompt into a language model. In addition, since various knowledge and experiences are required for plant operation, Patent Document 2 discloses a system that enables an answer by a service provider to a question from a customer to be shared via a network.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the information processing apparatus disclosed in Patent Document 1, since a database is selected when generating a prompt, the knowledge stored in the database is reflected in the generation of the prompt but not in the generation of the answer. In addition, in the information processing apparatus disclosed in Patent Document 1, an answer is generated by inputting a prompt into a language model, but the prompt is not repeatedly input into the language model. Therefore, there were cases where an appropriate answer could not be obtained for a user's question.

[0005] Furthermore, in the system disclosed in Patent Document 2, since the service provider answers questions from customers, the service provider must have the knowledge and experience to answer the questions, and there were limitations to the time that the service provider could respond.

[0006] This invention has been made in view of the above-mentioned problems, and aims to provide an information provision method and an information provision device that enable the generation of appropriate answers to user questions. [Means for solving the problem]

[0007] To achieve the above objective, an information provision method according to one aspect of the present invention is: A method of providing information that contributes to the maintenance of a plant using a computer, A question reception process that receives question information including plant status data indicating the state of the plant, A prompt generation process that generates prompt information that instructs a generative artificial intelligence model to output answer information that contributes to the maintenance of the plant, while referring to technical information related to the plant, based on the aforementioned question information. The process involves inputting the prompt information into the generative artificial intelligence model and then providing the response information output from the generative artificial intelligence model. The prompt generation process and the response provision process are as follows: This is something that is repeated multiple times. The first prompt generation process is as follows: Based on the aforementioned question information, prompt information is generated that gives the instruction to output the aforementioned answer information while referring to the aforementioned technical information. The prompt generation process from the second time onward is performed as follows: Based on the question information and the answer information output from the generative artificial intelligence model in previous answer provision processes, prompt information is generated that instructs the model to output the answer information while referring to the technical information. [Effects of the Invention]

[0008] According to one aspect of the present invention, prompt information generated based on question information including plant status data is repeatedly input to a generative artificial intelligence model, and answer information generated by the generative artificial intelligence model by referring to technical information is repeatedly generated. Therefore, it is possible to generate appropriate answers to user questions.

[0009] Other issues, configurations, and effects will be clarified in the embodiments for carrying out the invention described later. [Brief explanation of the drawing]

[0010] [Figure 1] This is an overall diagram showing an example of Information Provision System 1. [Figure 2] This is a data structure diagram showing an example of a technical information management database 50. [Figure 3] This is a data structure diagram showing an example of a user management database 51. [Figure 4] This is a block diagram showing an example of an information provision device 3. [Figure 5] This is a functional diagram illustrating an example of the information provision device 3. [Figure 6] This is a hardware configuration diagram showing an example of a computer 900 that makes up each device. [Figure 7] This is a flowchart illustrating an example of the operation of Information Provision System 1. [Figure 8A] This is a screen configuration diagram showing an example of a display screen 11A for inputting question information D1. [Figure 8B] This is a screen configuration diagram showing an example of a display screen 11B that includes the first prompt information D4 and response information D5. [Figure 8C] This is a screen configuration diagram showing an example of a display screen 11C that includes the second prompt information D4 and response information D5. [Figure 8D] This is a screen configuration diagram showing an example of a display screen 11C that includes the third prompt information D4 and response information D5.

Best Mode for Carrying Out the Invention

[0011] Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings. Hereinafter, the scope necessary for explaining for achieving the object of the present invention will be schematically shown, and mainly the scope necessary for explaining the relevant part of the present invention will be described, and the parts where the explanation is omitted shall be based on known techniques.

[0012] (Configuration of Information Providing System 1) FIG. 1 is an overall view showing an example of the information providing system 1. The information providing system 1 functions as a system for providing information contributing to the maintenance of the plant 10. The plant 10 is, for example, any plant such as a natural gas plant, an oil refining plant, a chemical treatment plant, a power generation plant, an ironmaking plant, etc., and is not limited to these examples.

[0013] The plant 10 is composed of a plurality of components, and each component performs a predetermined process. In the plant 10, as a plurality of components, for example, equipment 100 for processing any fluid such as gas, liquid, and fluidized granular material, piping 101 that connects between each equipment 100 and serves as a fluid flow path, and instruments 1 02 for measuring various state quantities of the fluid and the equipment 100, and controllers 103 for controlling the fluid and the equipment 100 are installed.

[0014] The equipment 100 includes, for example, but is not limited to, tower tanks and heat exchangers for performing reactions, distillations, extractions, absorptions, washings, temperature adjustments, etc. The piping 101 is three-dimensionally arranged within the site of the plant 10. The instruments 102 include, for example, but are not limited to, temperature sensors, pressure sensors, flow sensors, water level sensors, current sensors, voltage sensors, weather sensors, vibration sensors, speed sensors, acceleration sensors, gas sensors, timers, etc. The controllers 103 include, for example, but are not limited to, pumps, turbines, compressors, motors, valves, controllers, etc.

[0015] The information provision system 1 comprises, as its main components, a generative AI device 2, an information provision device 3 that operates in conjunction with the generative AI device 2, a terminal device 4 used by the user, and a database device 5 where various types of information are stored. Each of the devices 2 to 5 is, for example, composed of a general-purpose or dedicated computer (see Figure 6 below) and connected to a wired or wireless network 6, enabling the mutual transmission and reception of various types of data. Note that the number of each of the devices 2 to 5 and the connection configuration of the network 6 are not limited to the example in Figure 1 and may be changed as appropriate.

[0016] The generative AI device 2 is composed of, for example, a server-type computer or a cloud-type computer. The generative AI device 2 includes, for example, a generative artificial intelligence model 20, also known as generative AI or generative AI. The generative artificial intelligence model 20 receives instructions from prompt information D4 and generates response information D5 in response to those instructions.

[0017] The generative artificial intelligence model 20 is a pre-trained model constructed using a vast training dataset and deep learning technology, and may include foundational models (e.g., large-scale language models) or multimodal models. The generative artificial intelligence model 20 may also be provided as a service by an external provider; in this case, the generative AI device 2 may be omitted from the components of the information provision system 1.

[0018] The information providing device 3 is composed of, for example, a server-type computer or a cloud-type computer. When the information providing device 3 receives question information D1 from the terminal device 4, it generates prompt information D4 for the generative artificial intelligence model 20 and transmits it to the generative AI device 2. Also, when the information providing device 3 receives answer information D5 from the generative AI device 2 by the generative artificial intelligence model 20, it provides the answer information D5 to the terminal device 4.

[0019] Terminal device 4 consists of, for example, a stationary computer or a portable computer. Terminal device 4 is used by various users such as operators, maintenance managers, plant managers, and plant business managers. Terminal device 4 has programs such as applications and browsers installed, accepts various input operations, and outputs various information via display screens and audio. Terminal device 4 displays a display screen for accepting input operations to input question information D1 and outputting answer information D5, and sends and receives various data such as question information D1 and answer information D5 with information provision device 3.

[0020] The database device 5 is composed of, for example, a server-type computer or a cloud-type computer. The database device 5 includes a technical information management database 50 that manages multiple types of technical information D2 related to each plant 10, and a user management database 51 that manages user attribute information D3 related to the attributes of each user. The technical information management database 50 and the user management database 51 are accessible from each device 2 to 4, and are referenced, for example, when the generative artificial intelligence model 20 generates response information D5. In addition, editing operations such as adding, deleting, and modifying are performed on the technical information management database 50 and the user management database 51 via the information provision device 3 and the terminal device 4.

[0021] Figure 2 is a data configuration diagram showing an example of a technical information management database 50. The technical information management database 50 is a database for storing multiple types of technical information D2.

[0022] The technical information management database 50 stores several types of technical information D2, including industrial standards information D2A relating to the industrial standards required for plant 10, maintenance knowledge information D2B relating to the maintenance knowledge of plant 10 held by the construction company of plant 10, and plant history information D2C relating to the maintenance history or operating history of plant 10 held by the operator of plant 10.

[0023] Industrial standards information D2A includes information on various industrial standards such as JIS standards and API standards. These industrial standards may relate to the raw materials, products, exhaust gases, energy, and safety of plant 10, or they may relate to the equipment 100, piping 101, instruments 102, controllers 103, etc. that make up plant 10.

[0024] Maintenance knowledge information D2B includes the time, cost, and personnel required for maintenance work, the parts replaced during maintenance, the tools used, the skill level and qualifications required for maintenance work, and the details of the maintenance work itself.

[0025] The Plant History Information D2C includes maintenance data for Plant 10, such as the time, cost, number of personnel, and maintenance locations when maintenance work was performed in the past; spare parts data for parts and tools; special hazard data for special or dangerous maintenance locations; occupational accident data for past occupational accidents; schedule management data for maintenance workers (including those from partner companies); and maintenance worker data for maintenance workers (including those from partner companies), including their skill levels, qualifications, and safety training.

[0026] Furthermore, the Plant History Information D2C includes the following as the operating history of Plant 10: process data (fluid temperature, pressure, flow rate, operating time, etc.), raw material data (type of raw material, specific gravity, viscosity, component ratio, etc.), product data (product quantity, product quality, etc.), equipment data (temperature, pressure, flow rate, water level, current, voltage, vibration, speed, acceleration, etc. for each piece of equipment 100), exhaust gas data (component amounts such as CO2, SOx, NOx, etc.), energy consumption data (consumption of steam, electricity, gas, etc.), safety-related data (abnormal data such as temperature and pressure, operation history of safety valves and emergency stop systems), meteorological data (temperature, humidity, wind speed, precipitation), seismic motion data, etc.

[0027] Figure 3 is a data configuration diagram showing an example of a user management database 51. The user management database 51 is a database for storing the username and user attribute information D3 for each user, based on the user ID assigned to each user.

[0028] User attribute information D3 includes user attributes such as the name of the organization to which the user belongs, the job title within the organization, the user's authority, and qualifications. In the example in Figure 3, the user attributes are illustrated as being classified by job titles such as operator, maintenance manager, plant manager, and plant business manager.

[0029] (Configuration of Information Provisioning Device 3) Figure 4 is a block diagram showing an example of the information providing device 3. Figure 5 is a functional diagram illustrating an example of the information providing device 3.

[0030] The information providing device 3 consists of a control unit 30 composed of a processor, a storage unit 31 composed of an HDD, SSD, memory, etc., a communication unit 32 which is a communication interface with the network 6, an input unit 33 composed of a keyboard, mouse, etc., and a display, etc. It comprises a display unit 34 composed of the above. Note that the input unit 33 and the display unit 34 may be omitted.

[0031] The memory unit 31 stores the prompt syntax data 310 and the information processing program 311, as well as the operating system, other programs, various types of data, etc.

[0032] The prompt syntax data 310 is used to generate prompt information D4 from question information D1, and includes multiple syntaxes. Each syntax specifies technical information D2 to be referenced when generating answer information D5, or specifies the task to be imposed on the generative artificial intelligence model 20. Furthermore, each syntax is prepared for each user attribute to whom the answer information D5 will be provided.

[0033] The control unit 30 functions as a question receiving unit 300, a user attribute acquisition unit 301, a prompt generation unit 302, and an answer provision unit 303 by executing an information processing program 311 stored in the storage unit 31. Each of the units 300 to 303 of the control unit 30 functions as a user interface with the user by transmitting display information to the terminal device 4 to display various display screens on the terminal device 4 and accepting various input operations through those display screens.

[0034] The question reception unit 300 performs question reception processing to receive question information D1, which includes plant status data indicating the status of plant 10.

[0035] Question information D1 includes plant status data along with text data containing sentences representing the user's question. The plant status data is various types of data used to elaborate on the content of the question, and may be recorded using numerical values, characters, symbols, etc., but may also be graphs, tables, still images, moving images, etc. In this case, the plant status data may be input as a file in which the plant status data is recorded in any data format, or as a file name of a reference to access the file in which the plant status data is recorded.

[0036] Examples of plant condition data include, but are not limited to, plant design data (process flow diagrams, operation plans, plot plan diagrams, P&ID diagrams, I / O (input / output) lists, wiring block diagrams, structural design drawings, etc.), process data, raw material data, product data, equipment data, exhaust gas data, energy consumption data, safety-related data, weather data, and seismic data.

[0037] The process data included in the plant status data is the same as the process data included in the plant history information D2C, but the time periods covered are different. For example, the process data included in the plant status data covers the current point in time for plant 10 (which may be the most recent point in time) or a first period including the current point in time (for example, the one month from the current point in time to one month prior), while the process data included in the plant history information D2C covers a past point in time for plant 10 or a second period including the past point in time (for example, the one year from one month ago to 13 months prior). By comparing the process data included in the plant status data with the process data included in the plant history information D2C, it becomes possible to diagnose abnormalities, failures, performance degradation, etc. (including the detection and prediction of early signs).

[0038] The user attribute acquisition unit 301 performs a user attribute acquisition process to acquire user attribute information D3, which concerns the attributes of the user to whom the answer information D5 will be provided. The user to whom the answer information D5 will be provided may be the same as the user who entered the question information D1, or it may be a different user from the user who entered the question information D1.

[0039] The prompt generation unit 302 performs prompt generation processing to generate prompt information D4, which instructs the generative artificial intelligence model 20 to output response information D5 that contributes to the maintenance of plant 10, based on question information D1 and user attribute information D3, while referring to technical information D2 related to plant 10. For example, the prompt generation unit 302 performs natural language processing on the sentence (text data) representing the question in question information D1, and generates prompt information D4 according to the type of plant status data and the user attributes indicated by user attribute information D3. At that time, the prompt generation unit 302 refers to the prompt syntax data 310 stored in the storage unit 31, selects a syntax suitable for the content of the question and the user's attributes, and generates prompt information D4 by applying the type of technical information D2, the content of the question, and the content of the plant status data to that syntax.

[0040] The response provision unit 303 performs a response provision process that provides response information D5 output from the generative artificial intelligence model 20 by inputting prompt information D4 to the generative artificial intelligence model 20.

[0041] The prompt generation unit 302 and the answer provision unit 303 repeat the prompt generation process and the answer provision process multiple times. The first prompt generation process generates prompt information D4 that instructs the system to output answer information D5 while referring to technical information D2, based on the question information D1 and user attribute information D3. Subsequent prompt generation processes generate prompt information D4 that instructs the system to output answer information D5 while referring to technical information D2, based on the question information D1, user attribute information D3, and the answer information D5 output from the generative artificial intelligence model 20 in previous answer provision processes.

[0042] If the technical information management database 50 stores multiple types of technical information D2 (D2A to D2C), the first prompt generation process generates prompt information D4 that instructs the system to output answer information D5 while referring to one or more types of technical information D2 from among the multiple types of technical information D2, based on the question information D1. Subsequent prompt generation processes generate prompt information D4 that instructs the system to output answer information D5 while referring to one or more types of technical information D2 from among the multiple types of technical information D2 that were not referred to in the previous answer provision process, based on the question information D1 and the answer information D5 output by the generative artificial intelligence model 20 in the previous answer provision process.

[0043] In the example shown in Figure 5, the prompt generation process and the response provision process are repeated three times, and prompt information D4 is generated to instruct the generative artificial intelligence model 20 to refer to industrial standards information D2A in the first prompt generation process, maintenance knowledge information D2B in the second prompt generation process, and plant history information D2C in the third prompt generation process.

[0044] (Hardware configuration of each device) Figure 6 is a hardware configuration diagram showing an example of the computer 900 that constitutes each device. Each device 2 to 5 in the information provision system 1 is composed of a general-purpose or dedicated computer 900.

[0045] As shown in Figure 6, the computer 900 comprises, as its main components, a bus 910, a processor 912, memory 914, an input device 916, an output device 917, a display device 918, a storage device 920, a communication interface unit 922, an external device interface unit 924, an I / O device interface unit 926, and a media input / output unit 928. Note that the above components may be omitted as appropriate depending on the intended use of the computer 900.

[0046] The processor 912 consists of one or more arithmetic processing units (CPU (Central Processing Unit), MPU (Micro-Processing Unit), DSP (Digital Signal Processor), GPU (Graphics Processing Unit), NPU (Neural Processing Unit), etc.) and operates as a control unit that oversees the entire computer 900. The memory 914 stores various data and programs 930 and consists of volatile memory (DRAM, SRAM, etc.) that functions as main memory, and non-volatile memory (ROM), flash memory, etc.

[0047] The input device 916 consists of, for example, a keyboard, mouse, numeric keypad, or electronic pen, and functions as an input unit. The output device 917 consists of, for example, a sound (voice) output device or a vibration device, and functions as an output unit. The display device 918 consists of, for example, a liquid crystal display, an organic EL display, electronic paper, or a projector, and functions as an output unit. The input device 916 and the display device 918 may be configured as an integrated unit, such as a touch panel display. The storage device 920 consists of, for example, an HDD or SSD, and functions as a storage unit. The storage device 920 stores various data necessary for the execution of the operating system and program 930.

[0048] The communication I / F unit 922 is connected by wire or wireless to a network 940 such as the Internet or an intranet (which may be the same as network 6 in Figure 1) and functions as a communication unit that sends and receives data with other computers according to a predetermined communication standard. The external device I / F unit 924 is connected by wire or wireless to external devices 950 such as cameras, printers, scanners, and reader / writers and functions as a communication unit that sends and receives data with external devices 950 according to a predetermined communication standard. The I / O device I / F unit 926 is connected to I / O devices 960 such as various sensors and actuators and functions as a communication unit that sends and receives various signals and data with the I / O devices 960, such as detection signals from sensors and control signals to actuators. The media input / output unit 928 consists of, for example, a drive device such as a DVD drive or CD drive, a memory card slot, and a USB connector, and reads and writes data to media (non-temporary storage media) 970 such as DVDs, CDs, memory cards, and USB memory.

[0049] In the computer 900 having the above configuration, the processor 912 calls and executes the program 930 stored in the storage device 920 in the memory 914, and controls various parts of the computer 900 via the bus 910. The program 930 may also be stored in the memory 914 instead of the storage device 920. The program 930 may be recorded on the media 970 in an installable or executable file format and provided to the computer 900 via the media input / output unit 928. The program 930 may also be provided to the computer 900 by downloading it via the network 940 through the communication interface unit 922. Furthermore, the computer 900 may implement various functions realized by the processor 912 executing the program 930 using hardware such as an FPGA (Field-Programmable Gate Array) or ASIC (Application Specific Integrated Circuit).

[0050] Computer 900 is an electronic device of any form, consisting of, for example, a stationary computer or a portable computer. Computer 900 may be a client computer, a server computer, a cloud computer, or an embedded computer such as a control panel or controller (including microcontrollers, programmable logic controllers, and sequencers).

[0051] (Operation of Information Provision System 1) The following describes a series of operations performed by the information provision system 1. This series of operations is performed through the cooperation of each part 300-303 of the information provision device 3 (each process of the information provision method executed by the information processing program 311), the generative AI device 2, the terminal device 4, and the database device 5.

[0052] Figure 7 is a flowchart showing an example of the operation of the information provision system 1. In the following explanation, it is assumed that the technical information management database 50 contains industrial standards information D2A, maintenance knowledge information D2B, and plant history information D2C.

[0053] First, in step S100, the terminal device 4 receives, for example, a question regarding the maintenance of the plant 10 and plant status data as user input operations on the display screen shown in Figure 8A, which will be described later, and transmits question information D1 based on those input operations to the information providing device 3.

[0054] Figure 8A is a screen configuration diagram showing an example of a display screen 11A for inputting question information D1. The display screen 11A includes a question input area 110 for inputting a question, a file attachment button 110A for attaching a file, and a question confirmation button 110B for sending the question entered in the question input area 110 to the information providing device 3.

[0055] When a user enters a question in the question input area 110 on the display screen 11A, attaches a file containing plant status data using the file attachment button 110A, and presses the question confirmation button 110B, the question information D1 is transmitted to the information providing device 3 based on the input operation on the display screen 11A.

[0056] Next, in step S110 (question reception processing), the question reception unit 300 of the information providing device 3 receives the question information D1 transmitted in step S100, thereby accepting the question information D1 which includes plant status data.

[0057] Next, in step S110 (user attribute acquisition process), the user attribute acquisition unit 301 acquires user attribute information D3 related to the attributes of the user to whom the answer information D5 will be provided. For example, if the question in question information D1 contains characters that can identify the attributes of the recipient user, such as "for maintenance managers" or "plant business management meeting," the user attribute acquisition unit 301 acquires user attribute information D3 based on the question information D1. If the question in question information D1 does not contain characters that can identify the attributes of the recipient user, the user attribute acquisition unit 301 considers the user using the terminal device 4 that sent the question information D1 as the recipient user, and for example, refers to the user management database 51 based on the user ID entered as the login name when displaying the display screen 11A, and acquires user attribute information D3 associated with the user ID.

[0058] Next, in step S200 (first prompt generation process), the prompt generation unit 302 generates the first prompt information D4, which instructs the generative artificial intelligence model 20 to output answer information D5 while referring to industrial standard information D2A, based on the question information D1 received in step S110 and the user attribute information D3 acquired in step S120.

[0059] Next, in step S210 (first response provision process), the response provision unit 303 inputs the first prompt information D4 generated in step S200 into the generative artificial intelligence model 20, thereby generating the first response information D5 as an output result from the generative artificial intelligence model 20.

[0060] Next, in step S220 (first response provision process), the response provision unit 303 sends display screen information for displaying the first response information D5 generated in step S210 to the terminal device 4, which is the source of the question information D1. At this time, the display screen information also includes the first prompt information D4 generated in step S200.

[0061] Next, in step S230, when the terminal device 4 receives the display screen information transmitted in step S220, it displays a display screen 11B that includes the first prompt information D4 and response information D5 based on that display screen information.

[0062] Figure 8B is a screen configuration diagram showing an example of a display screen 11B that includes the first prompt information D4 and answer information D5. The display screen 11B comprises a question display area 111 for displaying question information D1, a prompt display area 112A for displaying the first prompt information D4, and an answer display area 113A for displaying the first answer information D5.

[0063] In the example shown in Figure 8B, the first prompt information D4 instructs the generative artificial intelligence model 20 to extract equipment 100 whose inspection period is within 3 months when compared to industrial standard information D2A, from the equipment list input as plant status data via an attached file. As a result, the first response information D5 shows the output of a list of equipment that requires inspection within 3 months. The display screen 11B may also include buttons to proceed to the subsequent step S300, or buttons to modify the first prompt information D4 and return to step S210.

[0064] Next, in step S300 (second prompt generation process), the prompt generation unit 302 generates a second prompt information D4 that instructs the generative artificial intelligence model 20 to output the answer information D5 while referring to the maintenance knowledge information D2B, based on the question information D1 received in step S110, the user attribute information D3 acquired in step S120, and the first answer information D5 generated in step S200.

[0065] Next, in step S310 (second response provision process), the response provision unit 303 inputs the second prompt information D4 generated in step S300 into the generative artificial intelligence model 20, thereby generating the second response information D5 as an output result from the generative artificial intelligence model 20.

[0066] Next, in step S320 (second response provision process), the response provision unit 303 sends display screen information for displaying the second response information D5 generated in step S310 to the terminal device 4, which is the source of the question information D1. At this time, the display screen information also includes the second prompt information D4 generated in step S300.

[0067] Next, in step S330, when terminal device 4 receives the display screen information transmitted in step S320, it displays a display screen 11C containing the second prompt information D4 and response information D5 based on that display screen information.

[0068] Figure 8C is a screen configuration diagram showing an example of a display screen 11C that includes the second prompt information D4 and the answer information D5. The display screen 11C comprises a prompt display area 112B that displays the second prompt information D4 and an answer display area 113B that displays the second answer information D5.

[0069] In the example in Figure 8C, the second prompt information D4 is the work time and cost when maintenance work is performed on each of the 100 pieces of equipment included in the equipment list in the first response information D5. The diagram illustrates a case where the generative artificial intelligence model 20 is instructed to refer to and reflect the maintenance knowledge information D2B. As a result, the diagram illustrates a case where the second response information D5 outputs the work time and work cost for the maintenance work on each piece of equipment 100. The display screen 11C may also include buttons to proceed to the next step S400, and buttons to modify the prompt information D4 from the second time or earlier and return to steps S210 and S310.

[0070] Next, in step S400 (third prompt generation process), the prompt generation unit 302 generates a third prompt information D4 that instructs the generative artificial intelligence model 20 to output the answer information D5 while referring to the plant history information D2C, based on the question information D1 received in step S110, the user attribute information D3 acquired in step S120, and the second answer information D5 generated in step S300.

[0071] Next, in step S410 (third response provision process), the response provision unit 303 inputs the third prompt information D4 generated in step S400 into the generative artificial intelligence model 20, thereby generating the third response information D5 as an output result from the generative artificial intelligence model 20.

[0072] Next, in step S420 (third response provision process), the response provision unit 303 sends display screen information for displaying the third response information D5 generated in step S410 to the terminal device 4, which is the source of the question information D1. At this time, the display screen information also includes the third prompt information D4 generated in step S400.

[0073] Next, in step S430, when the terminal device 4 receives the display screen information transmitted in step S420, it displays a display screen 11D containing the third prompt information D4 and response information D5 based on that display screen information.

[0074] Figure 8D is a screen configuration diagram showing an example of a display screen 11D that includes the third prompt information D4 and answer information D5. The display screen 11D comprises a prompt display area 112C that displays the third prompt information D4 and an answer display area 113C that displays the third answer information D5.

[0075] In the example shown in Figure 8D, the third prompt information D4 instructs the generative artificial intelligence model 20 to refer to the plant history information D2C to plan a schedule (date, time, number of personnel, priority) for performing maintenance work on each piece of equipment 100 as described in the second response information D5, and to also consider the availability of spare parts and safety precautions. As a result, the third response information D5 shows the maintenance work schedule for each piece of equipment 100, the availability of spare parts, and precautions. The display screen 11D may also include buttons to modify the prompt information D4 from the third prompt onwards and return to steps S210, S310, and S410.

[0076] Through the above series of processes, the user is provided with answer information D5 for question information D1. The same series of processes will be performed again if the user enters a new question.

[0077] As described above, according to the information provision device 3 and information provision method of this embodiment, prompt information D4 generated based on question information D1 including plant status data is repeatedly input to the generative artificial intelligence model 20, and answer information D5 generated by the generative artificial intelligence model 20 by referring to technical information D2 is repeatedly generated. Therefore, it is possible to generate appropriate answers to user questions.

[0078] Furthermore, when the prompt generation process and the response provision process are repeated multiple times, the technical information D2 referenced by the generative artificial intelligence model 20 is switched each time, and the response information D5 is generated in stages. Therefore, each time the prompt generation process and the response provision process are repeated, new technical information D2 is referenced, and the response information D5 is made concrete in stages. This makes it possible to generate more appropriate response information D5 compared to referencing various technical information D2 and generating response information D5 in one go.

[0079] Furthermore, multiple types of technical information D2 are referenced each time the prompt generation process and the response provision process are repeated, in the order of industrial standards information D2A, maintenance knowledge information D2B, and plant history information D2C. As a result, the response information D5 is progressively refined, such as the first response information D5 when viewed in light of industrial standards, the second response information D5 which incorporates maintenance knowledge based on the first response information D5, and the third response information D5 which incorporates maintenance history or operation history based on the second response information D5. This effectively reflects three technical perspectives in the response information D5, enabling the generation of more appropriate response information D5 compared to referencing various technical information D2 and generating response information D5 in a single step.

[0080] Furthermore, prompt information D4, generated based on the user attributes indicated by user attribute information D3, is input to the generative artificial intelligence model 20, which then generates answer information D5. Therefore, even for the same question, the user attributes to whom the answer information D5 is provided will differ. For example, the answer required for a maintenance manager will differ from the answer required for a plant business manager. As a result, prompt information D4 is generated based on the user's attributes, and these differences in user attributes are reflected in prompt information D4, which in turn are reflected in answer information D5. This makes it possible to generate appropriate answer information D5 according to the user's attributes.

[0081] (Other embodiments) The present invention is not limited to the embodiments described above, and can be implemented with various modifications without departing from the spirit of the invention. All such modifications are included in the technical concept of the present invention.

[0082] In the above embodiment, the functions of each part of the information providing device 3 were described as being realized in a single device, but the functions of each part may be distributed among multiple devices to be realized in multiple devices. Alternatively, the control unit of the terminal device 4 or the control unit of the database device 5 may execute the information processing program 311 to function as the information providing device 3.

[0083] In the above embodiment, a case was described in which the database device 5 includes a technical information management database 50 and a user management database 51. However, some or all of the technical information management database 50 and the user management database 51 may be stored in an external device (or more) connected to the network 6 or in any storage medium. In addition, at least one of the technical information management database 50 and the user management database 51 may be stored in the storage unit 31 of the information providing device 3.

[0084] In the above embodiment, the case in which the information providing device 3 operates according to the flowchart shown in Figure 7 has been described. However, the execution order of each step may be changed as appropriate, some steps may be omitted, or other steps may be added. For example, when the prompt generation unit 302 generates prompt information D4 in the prompt generation process (steps S200, S300, S400), a step may be added in which the prompt information D4 is provided to the terminal device 4, and a user input operation to change the prompt information D4 is received via the display screen of the terminal device 4, and the prompt information D4 is changed.

[0085] In the above embodiment, the case in which the information providing device 3 includes a user attribute acquisition unit 301 and performs a user attribute acquisition process to acquire user attribute information D3 was described. However, the information providing device 3 does not need to include a user attribute acquisition unit 301, and the user attribute acquisition process may be omitted. In that case, the prompt generation unit 302 should perform a prompt generation process to generate prompt information D4 without relying on the user attribute information D3.

[0086] In the above embodiment, the prompt generation process by the prompt generation unit 302 and the response provision process by the response provision unit 303 are repeated three times, and the prompt information D4 and response information D5 are generated in such a way that they are referenced in the order of industrial standard information D2A, maintenance knowledge information D2B, and plant history information D2C. In contrast, the number of times the prompt generation process and the response provision process are repeated may be changed as appropriate, and may be two times or four or more times. In this case, the order and type of technical information D2 referenced may be changed as appropriate, and technical information D2 other than industrial standard information D2A, maintenance knowledge information D2B, and plant history information D2C may be referenced. Furthermore, the number of times the prompt generation process and the response provision process are repeated may be changed as appropriate, and the order and type of technical information D2 referenced may be changed as appropriate, depending on the user attributes indicated by the user attribute information D3.

[0087] In the above embodiment, the generative artificial intelligence model 20 generates response information D5 by referring to the technical information D2 indicated by the prompt information D4 from among the multiple types of technical information D2 stored in the technical information management database 50 when prompt information D4 is input. In contrast, the generative artificial intelligence model 20 may use a trained model in which technical information D2 is systematically incorporated by including multiple types of technical information D2 as part of the training dataset. In that case, the generative artificial intelligence model 20 can generate response information D5 by referring to the technical information D2 incorporated into the generative artificial intelligence model 20 based on the instructions of the prompt information D4. [Explanation of symbols]

[0088] 1... Information provision system, 2... Generative AI device, 3... Information provision device, 4... Terminal device, 5...Database device, 10...Plant, 20...Generative artificial intelligence model 30...Control unit, 31...Storage unit, 32...Communication unit, 33...Input unit, 34...Display unit, 50...Technical information management database, 51...User management database, 300...Question reception unit, 301...User attribute acquisition unit, 302...Prompt generation unit, 303...Answer provision unit, 310...Prompt syntax data, 311...Information processing program

Claims

1. A method of providing information that contributes to the maintenance of a plant using a computer, A question reception process that receives question information including plant status data indicating the state of the plant, A prompt generation process that generates prompt information that instructs a generative artificial intelligence model to output answer information that contributes to the maintenance of the plant, while referring to technical information related to the plant, based on the aforementioned question information. The process involves inputting the prompt information into the generative artificial intelligence model and then providing the response information output from the generative artificial intelligence model. The prompt generation process and the response provision process are as follows: This is something that is repeated multiple times. The first prompt generation process is as follows: Based on the aforementioned question information, prompt information is generated that gives the instruction to output the aforementioned answer information while referring to the aforementioned technical information. The prompt generation process from the second time onward is as follows: Based on the aforementioned question information and the aforementioned answer information output from the generative artificial intelligence model in the previous answer provision process, prompt information is generated that instructs the model to output the aforementioned answer information while referring to the aforementioned technical information. Information provision method.

2. The aforementioned technical information is, It consists of multiple types of technical information related to each of the aforementioned plants, The first prompt generation process is as follows: Based on the question information, prompt information is generated that gives the instruction to output the answer information while referring to one or more types of the technical information, The prompt generation process from the second time onward is as follows: Based on the question information and the answer information output from the generative artificial intelligence model in the previous answer provision process, prompt information is generated that instructs the model to output the answer information while referring to one or more types of technical information from among multiple types of technical information that were not referenced in the previous answer provision process. The method for providing information according to claim 1.

3. Multiple types of the aforementioned technical information are Industrial standards information relating to the industrial standards required for the aforementioned plant, Maintenance knowledge information relating to the maintenance knowledge of the plant, held by the construction company of the plant, This consists of plant history information relating to the maintenance history or operating history of the plant, held by the operator of the plant. The first prompt generation process is as follows: Based on the aforementioned question information, prompt information is generated that gives the instruction to output the aforementioned answer information while referring to the aforementioned industrial standard information. The prompt generation process from the second time onward is as follows: Based on the question information and the answer information output from the generative artificial intelligence model in the first answer provision process, prompt information is generated that instructs the model to output the answer information while referring to the maintenance knowledge information. The prompt generation process from the third time onward is as follows: Based on the aforementioned question information and the answer information output from the generative artificial intelligence model in the second answer provision process, the plant history information is referenced to output the answer information. Generate the prompt information that gives the instruction to exert force, The information provision method described in claim 2.

4. A user attribute acquisition process is performed to obtain user attribute information regarding the attributes of the user to whom the aforementioned response information will be provided. The first prompt generation process is as follows: Based on the question information and the user attribute information, prompt information is generated that gives the instruction to output the answer information while referring to the technical information. The prompt generation process from the second time onward is as follows: Based on the aforementioned question information, the user attribute information, and the answer information output from the generative artificial intelligence model in previous answer provision processes, the prompt information is generated that instructs the model to output the answer information while referring to the technical information. The method for providing information according to claim 1.

5. An information providing device that provides information useful for plant maintenance, A question receiving unit that performs question receiving processing to receive question information including plant status data indicating the state of the plant, A prompt generation unit performs a prompt generation process that generates prompt information that instructs a generative artificial intelligence model to output answer information that contributes to the maintenance of the plant, while referring to technical information related to the plant, based on the aforementioned question information. The system includes a response provision unit that performs response provision processing by inputting the prompt information into the generative artificial intelligence model and providing the response information output from the generative artificial intelligence model, The prompt generation unit and the response provision unit are: The prompt generation process and the response provision process are repeated multiple times. The first prompt generation process is as follows: Based on the aforementioned question information, prompt information is generated that gives the instruction to output the aforementioned answer information while referring to the aforementioned technical information. The prompt generation process from the second time onward is as follows: Based on the aforementioned question information and the aforementioned answer information output from the generative artificial intelligence model in the previous answer provision process, prompt information is generated that instructs the model to output the aforementioned answer information while referring to the aforementioned technical information. Information provision device.

Images