Automatic setting device and automatic setting method

Abstract

This invention provides an automated setting device and method that automatically configures application settings without human intervention. [Solution] The automatic setting device 2 includes a log acquisition unit 232 that acquires application logs, an extraction unit 233 that extracts operation information related to the logs based on operation procedure data, an operation control unit 234 that operates the application settings based on the operation information, and a determination unit 236 that determines whether the logs are normal or not by repeatedly checking the log acquisition unit 232, the extraction unit 233, and the operation control unit 234.

Description

Technical Field

[0001] The present disclosure relates to an automatic setting device and an automatic setting method for automatically operating application settings.

Background Art

[0002] Conventionally, an information processing system for a specific industry such as BIM (Building Information Modeling) that can construct a design model used in the design of structures such as buildings and bridges is known (see, for example, Patent Document 1).

[0003] The information processing system described in Patent Document 1 receives code and setting data to embody a semiconductor device, classifies the operation characteristics and design history documents of design rules of the embodied semiconductor device into predetermined classes, and proposes a code modification policy and a setting data modification policy based on the classified classes.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] The information processing system described in Patent Document 1 has a function of proposing a code modification policy and a setting data modification policy, but is not a technology for automatically operating application settings. Like the information processing system described in Patent Document 1, current AI agents can modify code and setting data, but do not consistently execute error log analysis and automatic reprocessing loops.

[0006] Therefore, an automatic setting device and an automatic setting method for automatically operating application settings without human intervention are desired. [Means for solving the problem]

[0007] The characteristic configuration of the automatic setting device according to this disclosure is that it comprises a log acquisition unit that acquires application logs, an extraction unit that extracts operation information related to the logs based on operation procedure data, an operation control unit that operates the application settings based on the operation information, and a determination unit that repeatedly performs the log acquisition unit, the extraction unit, and the operation control unit to determine whether the logs are normal or not. Furthermore, the automatic setting method that can be executed by a computer includes a log acquisition process that acquires application logs, an extraction process that extracts operation information related to the logs based on operation procedure data, an operation control process that operates the application settings based on the operation information, and a determination process that repeatedly performs the log acquisition process, the extraction process, and the operation control process to determine whether the logs are normal or not.

[0008] The automatic setting device and automatic setting method described herein include a determination unit (determination process) that repeatedly performs log acquisition (log acquisition processing), extraction (extraction processing), and operation control (operation control processing) to determine whether the log is normal or not. This makes it possible for even inexperienced users to use specialized applications, and since the process is completed within the application, there is no need for human intervention. Furthermore, if the determination unit (determination process) determines that the log is normal, the system can automatically proceed to the next operation, thereby reducing the time required to resolve errors, reducing mistakes, and standardizing operations.

[0009] Thus, this is an automated setting device and method that automatically operates the application settings without human intervention. [Brief explanation of the drawing]

[0010] [Figure 1] This is a block diagram showing the configuration of an automatic setting device. [Figure 2] This is a flowchart showing the automatic setup method. [Figure 3]This is an example of a screen displaying drawing data errors using an automatic setting device. [Figure 4] This is an example of an explanatory screen after the AI ​​has been driven by the automated setting device. [Figure 5] This is an example of a screen illustrating the extraction process performed by an automated setting device. [Figure 6] This is an example of an explanatory screen showing that the extraction process has been completed by the automatic setting device. [Figure 7] This is an example of an explanation screen showing settings changed by an automatic setting device. [Figure 8] This is an example of a screen where settings have been automatically changed by an automatic configuration device. [Modes for carrying out the invention]

[0011] The embodiments of the automatic setting device and automatic setting method relating to this disclosure will be described below with reference to the drawings. In this embodiment, an example will be described in which the settings of a BIM (Building Information Modeling) that includes drawing data 33a consisting of a piping and instrumentation diagram 1 (P&ID) are automatically changed (operated) using a user device 3 on which the automatic setting device 2 is installed. However, the embodiments are not limited to the following, and various modifications are possible without departing from the gist thereof.

[0012] As shown in Figure 1, the BIM system 100 includes an automatic setting device 2, a user device 3, and a generating AI (Artificial Intelligence) 5. The automatic setting device 2 is a server owned by a vendor that provides an application used by users to automatically change various settings such as digital data in the piping and instrumentation diagram 1 (an example of operation) via SaaS (Software as a Service), on-premise, or a hybrid of these. The user device 3 is a terminal on which the automatic setting device 2 is installed, and is owned by the user who designs, constructs, and manages the plant, etc. Although one terminal is shown for convenience, different users who perform design, construction, or management may own multiple terminals. These terminals consist of desktop PCs, laptops, tablets, smartphones, etc.

[0013] The following describes an example of manipulating application settings, specifically the automatic modification of various settings such as the digital data in Piping Instrumentation Diagram 1, but is not limited to this example. For instance, instead of changing parameters as part of executing a specific process or command, a command to resolve a specific error (e.g., clearing the cache) could be directly executed. Alternatively, a rollback could be performed, automatically undoing the operation or series of operations immediately preceding the error and restoring the system to its normal state before the error occurred. Furthermore, if the system alone cannot resolve the issue, an effective workaround (e.g., opening a specific dialog box and highlighting recommended items) could be automatically implemented. The system could also communicate with an external database or other software to check for updates and consistency of information.

[0014] The automatic setting device 2, the user device 3, and the generating AI 5 are connected to each other via the network 4 so that they can communicate with one another. The automatic setting device 2 includes a processor 21, a communication interface 22, a memory 23, an input / output interface 24, and a display 25. Although the automatic setting device 2 has other functional components, only the functional components relevant to this embodiment are described.

[0015] Memory 23 stores the automatic configuration program 23a. The processor 21 is the central part of the computer and receives instructions and performs calculations and data processing. The processor 21 includes a CPU (Central Processing Unit), GPU (Graphics Processing Unit), NPU (Neural Network Processing Unit), or other hardware for executing the automatic configuration program 23a stored in memory 23. In other words, the processor 21 has arithmetic processing circuits, input ports, and output ports. The processor 21 may be an ASIC, FPGA, or SoC, and is not particularly limited.

[0016] Communication IF22 is a communication interface that has the function of transmitting signals output from the processor 21 to the user device 3 via the network 4, and the function of sending signals received from the user device 3 via the network 4 to the processor 21. The processor 21 can transmit and receive signals to and from the memory 23 and the communication IF22. Input / Output IF24 functions as an interface with input devices (e.g., keyboard, mouse, touch panel, touchpad, etc.) and output devices (e.g., display, speaker, etc.).

[0017] Memory 23 stores data within the computer and consists of main memory and secondary memory. Main memory is a storage area for temporarily saving programs and data and is composed of RAM, etc. Secondary memory is a non-temporary storage area that permanently stores programs and data and is composed of HDD, SSD, etc. Note that secondary memory may also be external hardware such as a cloud server or rental server.

[0018] The automatic setting program 23a is stored in the secondary memory (storage unit 23A), and is add-on software that is read into the main memory and executed by the processor 21. The automatic setting program 23a is software that is transmitted to the user device 3 via the network 4 and can be executed on the user device 3. Note that the automatic setting program 23a may be add-on software that adds functions to a creation support tool (e.g., AutoCAD P&ID) for the piping and instrumentation diagram 1, or may be provided as API (Application Programming Interface) software.

[0019] The automatic setting program 23a is software for integrally managing digital information related to piping and instrumentation in a plant. The automatic setting program 23a can automatically operate the settings of BIM (Building Information Modeling) (an example of an application, hereinafter referred to as "BIM application") that includes two-dimensional or three-dimensional digital models (drawing data 33a). In the present embodiment, as an example of BIM, it is composed of CAD software capable of an automatic piping function and high-precision layout creation in plant design. Note that BIM may be composed of other programs such as REVIT (registered trademark), Archicad (registered trademark), and GLООBE (registered trademark).

[0020] Log database DBa is associated with the drawing data 33a. Thereby, the drawing data 33a can be utilized in all phases of plant design, construction, operation, and maintenance. Note that the log database DBa may be stored only in the memory 33 of the user device 3 without being stored in the memory 23 of the automatic setting device 2, or may be stored in both the memory 23 of the automatic setting device 2 and the memory 33 of the user device 3.

[0021] The display 25 is provided in the automatic setting device 2 and is a display device that displays text, images, videos, and other visual information. This display 25 can display information by executing the automatic setting program 23a on the processor 21. The display 25 is composed of an LCD display, LED display, OLED display, plasma display, etc. However, the display 25 is not particularly limited as long as it is hardware capable of executing the automatic setting program 23a on the processor 21 and displaying the drawing data 33a.

[0022] User device 3 comprises a processor 31, a communication interface 32, a memory 33, an input / output interface 34, a display 35, and an operation unit 36. The processor 31, communication interface 32, memory 33, input / output interface 34, and display 35 have a basic configuration similar to the processor 21, communication interface 22, memory 23, input / output interface 24, and display 25 of the automatic setting device 2, so a brief explanation is omitted. Although user device 3 has other functional units, only the functional units relevant to this embodiment are described.

[0023] Memory 33 stores digitized drawing data 33a of the piping and instrumentation diagram 1. The drawing data 33a is digital data generated by automatic routing logic, or digital data generated by scanning the piping and instrumentation diagram 1. This drawing data 33a includes vector information in which shapes, lines, characters, etc., are represented by mathematical formulas (coordinates and vectors). The drawing data 33a is associated with the log database DBa. The operation unit 36 ​​is constructed from at least one element from among touch switches, keyboards, mice, scanners, voice input speakers, etc. When a user operates the operation unit 36, a signal corresponding to that operation is input to the processor 31.

[0024] (Detailed configuration of the automatic setting device) Figure 1 shows a block diagram of an automatic setting device 2 for automatically changing the settings of a BIM application, including digital data of a piping and instrumentation diagram 1 (an example of operation). The automatic setting device 2 can exchange information with a generating AI 5. The generating AI 5 includes large language models (LLMs) or other language models (e.g., VLMs: Vision-Language Models), and is composed of, for example, ChatGPT, Gemini, Claude, etc. The memory 23 of the automatic setting device 2 comprises a storage unit 23A and an automatic setting program 23a. In this embodiment, the automatic setting device 2 analyzes operation procedure data and logs containing character information using a large language model.

[0025] The memory unit 23A functions as a secondary memory to the memory 23 described above. The memory unit 23A stores the log database DBa, the operation procedure database DBb, and the trained model 23b. The memory unit 23A also stores block patterns, various objects, and tools, but a detailed explanation is omitted.

[0026] The log database DBa is a database that stores logs of the BIM application. This log database DBa stores the generated scripts and the automatic setting history from the operation control unit 234 (described later) associated with the operation information for logs that the determination unit 236 (described later) determined to be normal. The logs in this embodiment consist of access logs, error logs, and operation logs, etc. The access log contains history information of accesses to the BIM application. Error logs include tag errors such as "duplicate equipment numbers or instrumentation tags" or "undefined tags being used"; connection errors such as "piping lines are not properly connected (broken in the middle and not connected to equipment)" or "instrumentation signal lines are not connected to the target equipment"; specification errors such as "valve specifications (diameter, pressure, material) do not match piping specifications" or "equipment inlet / outlet conditions (temperature, pressure) do not match"; symbol errors such as "using prohibited symbols" or "symbols that do not conform to standards (ISO / JIS / ISA)"; loop errors such as "duplicate or missing instrumentation loop numbers" or "control loop configuration incomplete (e.g., AI input exists but output destination is not set)"; and database linkage errors such as "P&ID information does not match equipment list or I / O list".

[0027] The operation log is, for example, operation history information executed via the automatic setting program 23a or the operation unit 36 ​​of the user device 3 for multiple components that form start and end points extracted from vector information contained in the drawing data 33a. The multiple components include a string such as an ID, the coordinates of the rectangular area surrounding this string and its type (Line number, equipment, nozzle, drawing number, connector number, battery limit, etc.). The multiple components include, for example, structured data (CSV, etc.) that lists the Line number, From (starting point number), and To (ending point number) associated with the drawing number of the drawing data 33a for each of the multiple components contained in the drawing data 33a. As described above, since the drawing data 33a in this embodiment contains character information including a string such as an ID as a component of the log database DBa, the character information of the Line number, the character information of the From (starting point number), and the character information of the To (ending point number) are interrelated.

[0028] The Operation Procedure Database DBb is a digital data (natural language, audio, images, video, etc.) that systematically compiles specialized manuals such as the creation, interpretation, and operation rules of piping and instrumentation diagrams 1 used in plant and factory design, as well as company-specific operation rules and the background and constraints of the work. This digital data includes basic information such as "the purpose and role of P&ID (how to use it in design, construction, operation, and maintenance)" and "relationship with other drawings (piping diagrams, instrumentation loop diagrams, I / O lists, etc.)", "drawing format (frame, title field, revision management method)", "symbols and symbols (JIS, ISO, ISA, company-specific rules)", "notation methods for equipment, valves, and instrumentation equipment", "equipment number (P-101: pump, T-202: tank, etc.)", and "instrumentation loop number (PT-101: pressure gauge, FV-202: The topics covered are complex and diverse, including tag numbering rules such as "flow control valves, etc." and "line numbering system (including fluid, system, bore diameter, material, etc.)", control and instrumentation methods such as "line numbering system (including fluid, system, bore diameter, material, etc.)", "how to write control loops (PID control, sequence control, etc.)", and "how to record alarms and interlocks", check reviews such as "consistency checks (matching P&ID with equipment list and I / O list)", "how to check error logs" and "how to proceed with reviews (basic design review, detailed design review)", and case studies such as "P&ID samples using actual plants as examples" and "common mistakes and how to correct them".

[0029] The trained model 23b is generated by the learning unit 237 (described later) inputting operation information for logs determined to be normal by the judgment unit 236 included in the log database DBa to the generating AI 5, and then performing deep learning using the generated script and the automatic setting history of the operation control unit 234 (described later) as training data. In other words, the trained model 23b becomes a generating AI 5 customized for each plant. In this embodiment, the trained model 23b is trained by the learning unit 237 without user intervention by referring to operation information for normal logs included in the log database DBa in real time or at predetermined intervals. As a result, it is possible to naturally improve the inference accuracy of multiple components that form the start and end points output by the generating AI 5 via the trained model 23b, and the user can experience the generating AI 5 without any discomfort.

[0030] The automatic setting program 23a comprises a setting unit 231, a log acquisition unit 232, an extraction unit 233, an operation control unit 234, a display control unit 235, a determination unit 236, a learning unit 237, and an output control unit 238. The automatic setting program 23a is software that causes a computer to execute these functional units. The automatic setting program 23a does not have to have all of these functional units; it may have only some of them, and other parts may be linked to external software via API.

[0031] The configuration unit 231 sets up an AI-driven screen 35d (see "AI Route Assistant" in Figure 4) that allows users to view a partially enlarged view of the drawing data 33a consisting of the piping and instrumentation diagram 1, etc., created by the automatic routing logic ("Route Assist" in Figure 3), and error logs related to this partially enlarged view. Specifically, when drawing data 33a consisting of the piping and instrumentation diagram 1, etc., is created by the automatic routing logic using "Route Assist" in Figure 3, selecting "Piping" in Figure 3 as the design mode and checking "Error Report" will launch "AI Route Assistant" in Figure 4. In the example shown in Figure 4, multiple error logs are displayed in "AI Route Assistant," and the generated AI 5 is driven from the request selected by the user. However, the generated AI 5 may be automatically driven when there are error logs without displaying "AI Route Assistant." This AI-driven screen 35d may be automatically displayed on the browser screen shown on the display 35 of the user device 3, or a display button for the AI-driven screen 35d may be provided in the browser. In other words, the setting unit 231 sets the AI-driven screen 35d using the generated AI 5 as add-in software.

[0032] As shown in Figure 3, the setting unit 231 has an error log display area 35a on the right side of the drawing, a drawing display area 35b in the center of the drawing, and a filter area 35c on the left side of the drawing. Also, as shown in Figure 4, the setting unit 231 has an AI drive screen 35d between the error log display area 35a and the drawing display area 35b. On this AI drive screen 35d, the user can input instructions ("Request") for the generating AI 5, including natural language, images, or files, and the generating AI 5 can be driven by clicking the arrow in the lower right corner. Input of instructions ("Request") for the generating AI 5 is performed via an operation unit 36 ​​consisting of at least one element from among a touch switch, keyboard, mouse, scanner, voice input speaker, etc. Here, natural language includes words such as keywords, groups of words, phrases consisting of words and their suffixes, and sentences consisting of combinations of phrases. Images may be still images or videos. Files include project data and Excel data, etc.

[0033] The natural language, images, or files entered into the instructions ("Request") for the AI5 generating the AI-driven screen 35d may include instructions for modifying the script. Instructions for modifying the script are natural language, such as "Please correct the bend in the pipe," which prompts the AI5 to generate program code. Based on this natural language, the AI5 generates a modification script.

[0034] The log acquisition unit 232 acquires logs of the BIM application, including drawing data 33a containing vector information and text information, stored in the BIM system 100 (user device 3). Specifically, the log acquisition unit 232 acquires access logs, error logs, and operation logs of the BIM application, including the drawing data 33a. Note that the drawing data 33a is not limited to being stored in the memory 33 of the user device 3, but may also be stored outside the system, such as on a cloud server, rental server, or other storage medium. In this case, the log acquisition unit 232 will acquire the drawing data 33a containing vector information and text information stored in the external storage device of the BIM system 100 via the network 4 or via the input / output IF 34.

[0035] The extraction unit 233 inputs operation procedure data stored in the operation procedure database DBb to the generation AI 5, which then extracts operation information related to the log. For example, the operation information shown in Figure 6 is the operation procedure of the BIM application analyzed by the generation AI 5 to resolve the error log. At this time, the BIM application operation information is output by executing a script generated by the generation AI 5 based on the content of a natural language, image, or file (for example, "Fix Axis Alignment mismatched") input from the operation unit 36 ​​of the user device 3 to the generation AI 5 on the AI ​​drive screen 35d. The natural language, image, or file input to the instruction ("Request") on the AI ​​drive screen 35d is sent to the automatic setting device 2, and the generation AI 5 generates a script based on the input content of the generation AI field IP. The "instruction to the generation AI 5 on the AI ​​drive screen 35d" includes using the content of the error log as is and processing the content of the error log to make it suitable for script generation.

[0036] The extraction unit 233 executes the script generated by the generation AI 5 to extract multiple candidates as operation information. For example, as shown in Figure 4, by selecting and inputting "Fix Axis Alignment mismatched" as an instruction ("Request") to the generation AI 5, as shown in Figure 5, the script generated by the generation AI 5 is executed, and as shown in Figure 6, multiple solutions (such as enabling "Allow Detach" or enabling "Allow Non-Orthogonal Pipes") are extracted. This script is program code (e.g., Python, Java, C#, etc.) that extracts operation information.

[0037] The extraction unit 233 may execute a correction script generated by the generating AI 5 based on the correction content entered in the instruction ("Request") from the operation unit 36 ​​of the user device 3 to the generating AI 5. This correction script is program code that performs correction, deletion, and addition of operation information. For example, if "Fix Axis Alignment mismatched" is entered but the solution is irrelevant, the generating AI 5 will generate a correction script based on the correction instruction entered in the instruction ("Request") to the generating AI 5, such as "Are there any other solutions?", and construct the corrected operation information by executing this correction script.

[0038] The operation control unit 234 operates the settings of the BIM application, including the drawing data 33a, based on the operation information extracted by the extraction unit 233. For example, as shown in Figure 6, the operation control unit 234 operates the settings of the BIM application, including the drawing data 33a, based on one candidate selected by checking "Allow Detaching" in the "Route Assist Condition" column from among the multiple operation information extracted by the extraction unit 233. As a result, as shown in Figure 7, the error log "Fix Axis Alignment mismatched" shown in Figure 3 is resolved, and only multiple warning logs "Warnings" remain. In this way, the operation control unit 234 may operate the settings based on one candidate selected from among the multiple operation information candidates extracted by the extraction unit 233. Since the operation control unit 234 selects one candidate from multiple candidates and operates the settings, the log acquisition unit 232 can acquire logs corresponding to the operation information selected by a highly skilled professional.

[0039] In this embodiment, the operation control unit 234 may, from among the multiple operation information extracted by the extraction unit 233, have the generating AI 5 calculate an expected value according to the log type, and change the setting (operate) based on the candidate with the highest expected value. For example, as shown in Figure 3, when an error log occurs, the extraction unit 233 and the operation control unit 234 are activated, and the generating AI 5 changes the setting (operates) based on the candidate with the highest expected value among the multiple operation information, and the error log is automatically resolved as shown in Figure 8. The expected value is calculated by the generating AI 5 using a statistical method to determine the probability of resolving the error log. By comparing these expected values, the candidate with the highest expected value is the operation information that enables "Allow Detach," which is labeled "Recommended," as shown in Figure 6. In this way, by changing the setting (operating) based on the candidate with the highest expected value calculated according to various log types, the log is optimized, and even inexperienced users can use the BIM application without feeling any hassle.

[0040] The display control unit 235 displays the settings changed (operated) by the operation control unit 234 on the screen (display 35) (see Figure 7 or Figure 8). Specifically, as shown in Figure 7, the display control unit 235 may display the drawing data 33a of the settings changed (operated) by the operation control unit 234 along with the multiple operation information extracted by the extraction unit 233, and the error log ("Errors"), or as shown in Figure 8, the log acquisition unit 232, the extraction unit 233, and the operation control unit 234 may be invisible to the generated AI5, which may then automatically loop through them, and the drawing data 33a of the settings changed by the operation control unit 234 and the error log ("Errors") may be displayed as the final setting change result, following Figure 3.

[0041] The display control unit 235 may highlight the components that have been modified (operated) by the operation control unit 234. For example, it may highlight the portion where the misalignment of the pipe axis has been corrected, such as when the solid line entering the pipe shown in Figure 3 disappears into the pipe shown in Figure 8. Highlighting the components that have been modified (operated) by the operation control unit 234 in this way makes it easier for the user to confirm the changes.

[0042] Although not shown in the diagram, the display control unit 235 may display the script generated by the generation AI 5 based on the content of the natural language, image, or file entered in the generation AI field IP. For example, it may visualize the flow of "instruction" → "generated script" → "preview". In this case, the display control unit 235 may also display an explanation of the script generated by the generation AI 5. The display control unit 235 may also provide a dashboard-like screen that displays a summary of the script or only the important parameters in an easy-to-read format.

[0043] The display control unit 235 may display the modified script that has been changed (operated) based on the modification instructions entered into the generated AI field IP via the operation unit 36 ​​of the user device 3. This modified script may display a comment field for each line, or it may have a screen that allows comparison of the original and modified versions. The display control unit 235 may also display a video on the display 35 that shows the procedure for modifying (operating) the drawing data 33a. This makes it possible to use it as a learning tool for operating the BIM system 100, further enhancing user convenience.

[0044] The determination unit 236, using the generating AI 5, repeatedly checks the log acquisition unit 232, the extraction unit 233, and the operation control unit 234 to determine whether the log is normal or not. If the log is an error log, the determination unit 236 determines that it is normal when the error log is resolved. If the determination unit 236 resolves all error logs and determines that the log is normal, the user can proceed to the next task. Thus, the system is equipped with a determination unit 236 that repeatedly checks the log acquisition unit 232, the extraction unit 233, and the operation control unit 234 to determine whether the log is normal or not. In other words, the system autonomously completes a series of cycles, from error log analysis to the generation of correction proposals, automatic application of settings, automatic reprocessing, and result verification, without human judgment. This makes it possible for even inexperienced users to use specialized BIM applications, and since it is completed within the BIM application, there is no need for human intervention. In addition, since it is possible to automatically proceed to the next operation if the determination unit 236 determines that it is normal, it is possible to reduce the time spent on error resolution, reduce errors, and standardize operations. Thus, the automatic setting device 2 automatically operates the BIM application settings without human intervention.

[0045] Furthermore, the determination unit 236 may calculate the effect verification result of clearing the error log. When this effect verification result satisfies predetermined conditions, the operation control unit 234 (operation control processing) may determine that the log based on the settings operated is normal and terminate the process. The predetermined conditions include when the error log is cleared and only warning logs remain, or when the numerical value of the effect verification result, when expressed numerically, is 80% or higher. The determination unit 236 may also determine the version of the script stored in the log database DBa. By determining the script version, the determination unit 236 can track which settings were used at which stage in the drawing data 33a.

[0046] The determination unit 236 may determine the access rights related to the construction of the drawing data 33a. The determination unit 236 assigns different permissions to each user and restricts the range of operations that can be performed. This makes it possible to manage access rights according to the difficulty and importance of the modification work on the drawing data 33a, for example, by allowing only specific users to modify the drawing data 33a.

[0047] The learning unit 237 inputs operation information for logs determined to be normal by the judgment unit 236 into the generating AI 5 (large-scale language model) for training. The learning unit 237 inputs operation information for logs determined to be normal by the judgment unit 236, which are contained in the log database DBa, into the generating AI 5, and uses the generated script and the automatic setting history of the operation control unit 234 as training data to perform deep learning and generate a trained model 23b. This trained model 23b may be a locally generated AI or a cloud-generated AI. In this way, training a large-scale language model based on operation information for logs determined to be normal by the judgment unit 236 improves model performance and enhances usability.

[0048] As described above, the log database DBa stores the generated scripts and the automatic setting history in the operation control unit 234 associated with the operation information for logs that the determination unit 236 has determined to be normal. The operation information for logs that the determination unit 236 has determined to be normal is the most accurate information contained in the operation procedure database DBb, and the scripts based on this accurate information are used as training data to perform deep learning and generate a trained model 23b. Furthermore, by inputting the operation information for logs that the determination unit 236 has determined to be normal into this trained model 23b, the validity of the parameters and structured data in the accumulated drawing data 33a can be checked or evaluated. In addition, the trained model 23b can learn from the log database DBa related to the accumulated drawing data 33a, refer to and automatically present best practices for each piece of drawing data 33a, and optimize the design using the previous operation information.

[0049] The output control unit 238 can output a list of components included in each drawing data 33a, organized by start and end point. The data output format is not limited to CSV files; it supports various formats such as Excel files and JSON (JavaScript Object Notation). Outputting such a list allows for importing the drawing data 33a along with the list into other systems, making it highly versatile.

[0050] (Automatic setup method) Figure 2 shows a flowchart of the automatic setting method executed by the automatic setting device 2 in the BIM system 100 where the drawing data 33a is stored. The automatic setting method shown in Figure 2 is only a representative example, and other flowcharts for executing each function of the automatic setting device 2 in the above-described embodiment are omitted.

[0051] As shown in Figure 2, the automated configuration method executable by the computer includes a log acquisition process (#23) for acquiring logs of the BIM application, an extraction process (#26) for extracting operation information related to the logs by inputting operation procedure data into the generating AI 5, an operation control process (#27) for operating the BIM application settings based on the operation information, a display control process (#29) for displaying the settings operated by the operation control process on the screen, and a determination process (#28) for determining whether the logs are normal or not by repeating the log acquisition process (#23), the extraction process (#26), and the operation control process (#27). Alternatively, the extraction process (#26) may extract multiple candidates as operation information, and then the operation control process (#27) may operate the settings based on one of the selected candidates, or the generating AI 5 may calculate an expected value according to the type of log and operate the settings based on the candidate with the highest expected value. The judgment process calculates the effect verification result by resolving the error log, and when the effect verification result satisfies the predetermined conditions, it may determine that the log with the settings operated by the operation control process is normal and terminate the process (#28Yes). Furthermore, the display control process (#29) may highlight the components operated by the operation control process (#26).

[0052] The automatic setting method in this embodiment may include a learning process (#30) that trains a large-scale language model based on logs determined to be normal in the judgment process (#28). The judgment process may also calculate the effect verification result of resolving error logs, and based on this calculated effect verification result, it is possible to determine whether to continue repeating the log acquisition process, extraction process and operation control process to determine the degree to which the error logs have been resolved. Furthermore, the display control process may further display the effect verification result on the screen, and by displaying the effect verification result on the screen, the user can intuitively confirm whether the error logs have been resolved.

[0053] First, the user inputs drawing data 33a, which contains the arrangement of multiple equipment parts from the piping and instrumentation diagram 1, into the BIM application on the user device 3, where the automatic setting program 23a of the automatic setting device 2 is installed (#21). Then, the automatic routing logic ("Route Assist" in Figure 3) of the BIM application is executed to perform automatic piping, connecting each part (#22). Next, the log acquisition unit 232 acquires a log of the drawing data 33a, which includes vector information and text information, stored in the user device 3 (#23, log acquisition process). For example, as shown in Figure 3, "Axis Alignment mismatched" and "Angular Alignment mismatched" are acquired as error logs ("Errors").

[0054] Next, the setting unit 231 sets up an AI-driven screen 35d (see "AI Route Assistant" in Figure 4) that allows the user to view a partially enlarged view of the drawing data 33a, which consists of the piping and instrumentation diagram 1, etc., created by the automatic routing logic ("Route Assist" in Figure 3), and the error log related to this partially enlarged view (#24). Note that this AI-driven screen 35d (see "AI Route Assistant" in Figure 4) may be set by the user clicking the AI-driven button (not shown), or it may be set automatically, or the error log may be automatically cleared without being set.

[0055] Next, as shown in Figure 3, if there is an error log ("Errors") and it is necessary to change (operate) the settings of the BIM application (#25 Yes), the extraction unit 233 inputs the operation procedure data stored in the operation procedure database DBb to the generation AI 5, so that the generation AI 5 extracts operation information related to the log (#26, extraction process). For example, as shown in Figure 4, by selecting and inputting "Fix Axis Alignment mismatched" as an instruction ("Request") to the generation AI 5, the generation AI 5 executes the script it generates, as shown in Figure 5, and as shown in Figure 6, multiple solutions (such as enabling "Allow Detach" and enabling "Allow Non-Orthogonal Pipes") are extracted.

[0056] Next, the operation control unit 234 modifies (operates) the settings of the BIM application including the drawing data 33a based on the operation information extracted by the extraction unit 233 (#27, operation control processing). For example, as shown in Figure 6, the settings of the BIM application including the drawing data 33a are modified (operated) based on one candidate selected by checking "Allow Detaching" in the "Route Assist Condition" column from among the multiple operation information extracted by the extraction unit 233. As a result, as shown in Figure 7, the error log "Fix Axis Alignment mismatched" shown in Figure 3 is resolved, and only multiple warning logs "Warnings" remain.

[0057] Next, the determination unit 236, using the generating AI 5, repeatedly performs log acquisition, extraction, and operation control processing to determine whether the log is normal or not (#28, determination process). The determination unit 236 determines that the log is normal when the error log has been resolved (#28 Yes). Then, the display control unit 235 displays the settings changed (operated) by the operation control unit 234 on the screen (display 35) as shown in Figure 7 or Figure 8 (#29, display control process). In this way, if the determination unit 236 resolves all error logs and determines that the log is normal, the user can proceed to the next task.

[0058] Furthermore, the learning unit 237 inputs operation information for logs that have been determined to be normal by the judgment unit 236 into the generating AI 5 (large-scale language model) for training (#30, training process). The learning unit 237 inputs operation information for logs that have been determined to be normal by the judgment unit 236, which are included in the log database DBa, into the generating AI 5, and uses the generated script and the automatic setting history of the operation control unit 234 as training data to perform deep learning, thereby generating a trained model 23b.

[0059] In the embodiment described above, the following configuration can be envisioned. (1) The automatic setting device 2 includes a log acquisition unit 232 that acquires application logs, an extraction unit 233 that extracts operation information related to the logs based on operation procedure data, an operation control unit 234 that operates the application settings based on the operation information, and a determination unit 236 that determines whether the logs are normal or not by repeatedly performing the log acquisition unit 232, the extraction unit 233, and the operation control unit 234. Furthermore, an automatic setting method that can be executed by a computer includes a log acquisition process (#23 in Figure 2) that acquires application logs, an extraction process (#26 in Figure 2) that extracts operation information related to the logs based on operation procedure data, an operation control process (#27 in Figure 2) that operates the application settings based on the operation information, and a determination process (#28 in Figure 2) that determines whether the logs are normal or not by repeatedly performing the log acquisition process (#23 in Figure 2), the extraction process (#26 in Figure 2), and the operation control process (#27 in Figure 2).

[0060] The automatic setting device 2 and automatic setting method according to this embodiment include a determination unit 236 (determination process) that repeatedly performs log acquisition processing 232 (log acquisition processing), extraction processing 233 (extraction processing), and operation control unit 234 (operation control processing) to determine whether the log is normal or not. This makes it possible for even inexperienced users to use specialized applications, and since the process is completed within the application, there is no need for human intervention. Furthermore, if the determination unit 236 (determination process) determines that the log is normal, the system can automatically proceed to the next operation, thereby reducing the time required to resolve errors, reducing mistakes, and standardizing operations. In this way, the automatic setting device 2 and automatic setting method automatically operate the application settings without human intervention.

[0061] (2) In the automatic setting device 2 (automatic setting method) of (1), it is preferable that the determination unit 236 (determination process) determines that the system is normal when the error log is cleared as a log.

[0062] As in this embodiment, if the system is deemed normal when the error log among various logs is resolved, even inexperienced users can use the application without any issues.

[0063] (3) In the automatic setting device 2 (automatic setting method) of (1) or (2), the extraction unit 233 (extraction process) preferably extracts multiple candidates as operation information.

[0064] As in this embodiment, by extracting multiple candidates as operational information, it is possible to understand the process of optimizing the settings.

[0065] In the automatic setting device 2 (automatic setting method) of (4)(3), it is preferable that the operation control unit 234 (operation control processing) operates the setting based on one candidate selected from among multiple candidates.

[0066] As in this embodiment, by selecting one candidate from multiple candidates and manipulating the settings, it is possible to obtain logs selected by highly skilled experts.

[0067] In the automatic setting device 2 (automatic setting method) of (5)(3), it is preferable for the operation control unit 234 (operation control processing) to calculate an expected value according to the type of log and operate the setting based on the candidate with the highest expected value of 1.

[0068] As in this embodiment, by manipulating the settings based on the candidate with the highest expected value calculated according to various log types, log optimization is achieved, allowing even inexperienced users to use the application without any hassle.

[0069] (6) In any one of the automatic setting devices 2 (automatic setting method) of (1) to (5), it is preferable to further include a learning unit 237 (learning process) which inputs and learns operation information for logs that have been determined to be normal by the determination unit 236 (determination process).

[0070] As in this embodiment, by inputting operation information for logs that have been determined to be normal and allowing the system to learn from it, it becomes possible to reduce the number of error logs themselves and improve the performance of the application.

[0071] In the automatic setting device 2 (automatic setting method) of (7)(2), the determination unit 236 (determination process) preferably calculates the effect verification result by resolving the error log.

[0072] As in this embodiment, the determination unit 236 (determination process) can calculate the degree to which the error log has been resolved based on the calculated effect verification results and decide whether to continue repeating the log acquisition unit 232 (log acquisition process), the extraction unit 233 (extraction process), and the operation control unit 234 (operation control process).

[0073] In the automatic setting device 2 (automatic setting method) of (8)(7), it is preferable that the determination unit 236 (determination process) determines that the log resulting from the settings operated by the operation control unit 234 (operation control process) is normal when the effect verification result satisfies predetermined conditions, and terminates the process.

[0074] As in this embodiment, if the processing of the determination unit 236 (determination process) is terminated when the effect verification results satisfy predetermined conditions, the error log is automatically cleared, allowing the user to proceed to the next task, thus improving usability.

[0075] (9) In any one of the automatic setting devices 2 (automatic setting method) of (1) to (8), the application is preferably a construction-related vertical SaaS.

[0076] As in this embodiment, if the application is a construction-related vertical SaaS, the operation difficulty is particularly high, making the automatic setting device 2 (automatic setting method) useful.

[0077] [Other embodiments] (a) In the above-described embodiment, the automatic setting device 2 was described as an add-in software for a BIM application installed on the user device 3 of the plant. However, the automatic setting device 2 may also be an add-in for structural design software of a CAD (Computer-Aided Design) system used for BIM (Building Information Modeling) (such as AutoCAD or REVIT®), and can be used in various systems capable of creating construction drawings. (b) The automatic setting device 2 in the above-described embodiment is not limited to construction-related vertical SaaS, but can be applied to vertical SaaS in other industries. (c) In the embodiment described above, the determination unit 236 repeatedly checked the log acquisition unit 232, the extraction unit 233, and the operation control unit 234 to determine whether the log was normal or not, but the display control unit 235 may also repeatedly check the log. (d) Some of the steps in the above-described embodiments may be omitted, or they may be combined as appropriate to implement the functionality. [Industrial applicability]

[0078] This disclosure can be used in automatic configuration devices and automatic configuration methods that automatically operate application settings. [Explanation of symbols]

[0079] 2: Automatic setting device, 232: Log acquisition unit, 233: Extraction unit, 234: Operation control unit, 236: Judgment unit, 237: Learning unit

Claims

1. A log acquisition unit that acquires logs including operation logs of a BIM application, An extraction unit that extracts operation information related to the operation log by inputting operation procedure data stored in the operation procedure database into a generating AI, An operation control unit that operates the settings of the BIM application based on the aforementioned operation information, The system includes a determination unit that sequentially executes the log acquisition unit, the extraction unit, and the operation control unit, and determines whether the log is normal or not. The operation control unit sequentially changes the settings based on the operation procedure included in the operation information, and when the determination unit determines that the error log has been cleared and the system is functioning correctly, it confirms the settings. An automatic setting device that continuously executes the log acquisition unit, the extraction unit, the operation control unit, and the determination unit.

2. The automatic setting device according to claim 1, wherein the extraction unit extracts a plurality of candidates as the operation information.

3. The automatic setting device according to claim 2, wherein the operation control unit operates the setting based on one of the candidates selected from a plurality of candidates.

4. The automatic setting device according to claim 2, wherein the operation control unit calculates an expected value according to the type of operation log and operates the setting based on the candidate with the highest expected value (1).

5. The automatic setting device according to claim 1, further comprising a learning unit that inputs and learns the operation information for the operation log that has been determined to be normal by the determination unit.

6. The automatic setting device according to claim 1, wherein the determination unit calculates the effect verification result obtained by resolving the error log.

7. The automatic setting device according to any one of claims 1 to 6, wherein the BIM application is a construction-related vertical SaaS.

8. A log acquisition process that acquires a log including an operation log of a BIM application, The process involves inputting operation procedure data stored in the operation procedure database into a generating AI to extract operation information related to the operation log, and An operation control process that controls the settings of the BIM application based on the aforementioned operation information, The log acquisition process, the extraction process, and the operation control process are executed in order, and a determination process is performed to determine whether the log is normal or not. The operation control process sequentially changes the settings based on the operation procedure included in the operation information, and when the determination process determines that the error log has been cleared and the situation is normal, it confirms the settings. An automated setting method that can be executed by a computer, which continuously performs the log acquisition process, the extraction process, the operation control process, and the determination process.

Images