WO2025009185A1 - Architectural drawing creation support system - Google Patents

Abstract

Provided is an architectural drawing creation support system suitable for making an editable editing element in an architectural drawing compatible with the specifications of a finishing table.　A drawing creation support device 100 acquires finishing table data and editing elements, estimates matching elements from the acquired finishing table data and editing elements using a trained model trained on the basis of information about finishing elements appearing in a finishing table and information about compatible elements compatible with the specifications of the finishing elements and appearing in an architectural drawing, and, when the acquired editing elements and the estimated compatible elements do not match, indicates that defects exist in the editing elements.

Description

Architectural drawing support system

The present invention relates to a system that supports the creation of architectural drawings, and in particular to an architectural drawing creation support system that is suitable for adapting editable elements in architectural drawings to the specifications of a finishing table.

Conventionally, one known technology that uses AI (Artificial Intelligence) to assist in the creation of architectural drawings is the technology described in Patent Document 1.

The technology described in Patent Document 1 includes a table storage unit that extracts legal judgment elements related to numerical values from laws and regulations and stores a legal judgment table composed of the legal judgment elements for each law; an assignment element extraction unit that uses AI to extract assignment elements containing numerical values that can be substituted for the legal judgment elements from the initial layout data; an appropriateness judgment unit that assigns the assignment elements extracted by the assignment element extraction unit to the legal judgment table stored in the table storage unit and judges it to be appropriate if it satisfies the conditions in the legal judgment table and judges it to be inappropriate if it does not; and an judgment result creation unit that records the judgment results made by the appropriateness judgment unit and creates a floor plan judgment document including the judgment results.

JP 2022-100026 A

When designing a building, plans and detailed floor plans are created based on the finish table, but the dimensions and other specifications listed on the plans must match those listed on the finish table. Also, the tags listed on the detailed floor plans must match those listed on the finish table. Since the designer performs these tasks while checking the finish table, mistakes are likely to occur.

However, while the technology described in Patent Document 1 can determine whether editable elements in architectural drawings comply with architectural laws, it cannot make the editable elements conform to the specifications of a finishing table.

The present invention was made with a focus on these unresolved issues in the conventional technology, and aims to provide an architectural drawing creation support system that is suitable for adapting editable elements in an architectural drawing to the specifications of a finishing table.

[Invention 1] In order to achieve the above object, the architectural drawing creation support system of invention 1 comprises an element information acquisition means for acquiring finishing element information on finishing elements described in a finishing schedule and editing element information on editable editing elements in an architectural drawing, and an estimation means for estimating the matching element information from the finishing element information and editing element information acquired by the element information acquisition means, using a trained model that has been trained based on information on the finishing elements described in the finishing schedule and matching element information on matching elements described in the architectural drawing that match the specifications of the finishing elements.

Here, the finishing element information can be configured, for example, as information for identifying the finishing element (for example, a name, a number, an ID, a code, link information such as a URL). Also, the finishing element information can be configured, for example, as letters, numbers, figures, codes, symbols, images, sounds, and other information. Also, the finishing element information can be configured as keywords relating to the finishing element (for example, one or more keywords indicating part of the name of the finishing element). The same applies to other element information below.

The element information acquisition means may, for example, input element information from an input device or the like, acquire or receive element information from an external terminal or the like, read element information from a storage device or storage medium or the like, or generate or calculate element information by information processing or the like. Therefore, acquisition includes at least input, acquisition, reception, reading (including search), generation and calculation. The same concept of acquisition applies below.

Furthermore, this system may be realized as a single device, terminal, or other equipment, or may be realized as a network system in which multiple devices, terminals, or other equipment are communicatively connected. In the latter case, each component may belong to any one of the multiple devices, etc., as long as they are communicatively connected to each other. The same applies below to the architectural drawing creation support systems of inventions 4, 7, and 9.

[Invention 2] The architectural drawing creation support system of Invention 2 further includes an output means for outputting information about edited elements related to edited element information acquired by the element information acquisition means based on the matching element information estimated by the estimation means in the architectural drawing creation support system of Invention 1.

[Invention 3] Furthermore, the architectural drawing creation support system of invention 3 is an architectural drawing creation support system of either invention 1 or 2, in which the trained model has been trained based on information on the finishing elements described in the finishing table and on matching element information on matching elements and their contents described in the architectural drawing that match the specifications of the finishing elements, and is equipped with a modification means for modifying an edited element related to the edited element information acquired by the element information acquisition means based on the matching element information estimated by the estimation means.

[Invention 4] The architectural drawing creation support system of Invention 4 further includes an element information acquisition means for acquiring changed element information relating to an edit element that has been changed among editable edit elements in an architectural drawing, and an estimation means for estimating the related element information from the changed element information acquired by the element information acquisition means, using a trained model that has been trained based on information relating to the editable edit elements in the architectural drawing and related element information relating to other edit elements that should be changed or have been changed in conjunction with the change to the edit element.

[Invention 5] The architectural drawing creation support system of invention 5 further includes an output means for outputting information about other edit elements related to the related element information based on the related element information estimated by the estimation means in the architectural drawing creation support system of invention 4.

[Invention 6] Furthermore, the architectural drawing creation support system of invention 6 is the architectural drawing creation support system of either invention 4 or 5, in which the trained model has been trained based on information on editable editing elements in architectural drawings, and related element information on other editing elements that should be changed or have been changed in conjunction with a change to the edit element and their contents, and is equipped with a modification means for modifying other editing elements related to the related element information based on the related element information estimated by the estimation means.

[Invention 7] The architectural drawing creation support system of invention 7 further comprises an element information acquisition means for acquiring changed element information on edit elements that have been changed among editable edit elements in an architectural drawing, and an estimation means for estimating the judgment necessity information from the changed element information acquired by the element information acquisition means, using a trained model that has been trained based on information on the editable edit elements in the architectural drawing and judgment necessity information on whether a human judgment is required to change other edit elements that should be changed or have been changed due to the change in the edit element.

[Invention 8] The architectural drawing creation support system of invention 8 further includes an output means for outputting information on whether or not a change to another editing element related to the judgment necessity information is required, based on the judgment necessity information estimated by the estimation means, in the architectural drawing creation support system of invention 7.

[Invention 9] The architectural drawing creation support system of invention 9 further comprises an element information acquisition means for acquiring changed element information relating to an edit element that has been changed among editable edit elements in an architectural drawing, and an estimation means for estimating the related element information and the judgment necessity information from the changed element information acquired by the element information acquisition means, using a trained model that has been trained based on information on the editable edit elements in the architectural drawing, related element information relating to other edit elements that should be changed or have been changed due to the change in the edit element and their contents, and judgment necessity information relating to whether or not a human judgment is required to change the other edit elements.

[Invention 10] Furthermore, the architectural drawing creation support system of invention 10 is the architectural drawing creation support system of invention 9, and when it is determined that a human judgment is required to change another edit element related to the related element information based on the judgment necessity information estimated by the estimation means, the system includes an output means for outputting information about the other edit element based on the related element information estimated by the estimation means.

[Invention 11] Furthermore, the architectural drawing creation support system of invention 11 is the architectural drawing creation support system of either invention 9 or 10, and when it is determined that no human judgment is required to change other edit elements related to the related element information based on the judgment necessity information estimated by the estimation means, the architectural drawing creation support system of invention 11 is provided with a modification means for modifying the other edit elements based on the related element information estimated by the estimation means.

As described above, the architectural drawing creation support system of Invention 1 estimates compatible element information regarding compatible elements that correspond to the finishing elements listed in the finishing table, so it is expected that the edited elements will be adapted to the specifications of the finishing table.

Furthermore, according to the architectural drawing creation support system of Invention 3, compatible elements corresponding to the finishing elements described in the finishing table and compatible element information related to the contents thereof are estimated, so it is expected that the edited elements and their contents will be adapted to the specifications of the finishing table.

Furthermore, according to the architectural drawing creation support system of Invention 4, related element information regarding other edit elements related to the edit element that has been changed is estimated, so that other edit elements that should be changed can also be changed at the same time.

Furthermore, according to the architectural drawing creation support system of Invention 6, related element information regarding other edit elements related to the edit element that has been changed and their contents is estimated, so that other edit elements that should be changed and their contents can also be changed at the same time.

Furthermore, according to the architectural drawing creation support system of Invention 7, judgment necessity information is estimated as to whether or not a human judgment is required to change other edit elements related to the edit element that has been changed, so it is possible to know whether or not a human should change the other edit elements that need to be changed.

Furthermore, according to the architectural drawing creation support system of Invention 9, related element information on other edit elements related to the edit element that has been changed and their contents, as well as judgment necessity information on whether or not a human judgment is required to change the other edit elements, are estimated, so that it is possible to grasp whether or not a human should change the other edit elements and their contents that need to be changed.

FIG. 1 is a diagram illustrating a hardware configuration of a drawing creation support device 100. FIG. 13 is a diagram showing the structure of finishing list data. FIG. 1 is a diagram showing the structure of CAD data of a plan view. FIG. 13 is a diagram showing the structure of CAD data of a detailed plan view. Wall type diagram. 13 is a flowchart showing a trained model generation process. FIG. 1 is a block diagram showing the process of generating and using a trained model. 13 is a flowchart showing a matching element estimation process. This is a plan view before changing the editing element. FIG. 11 is a plan view after changing the editing element. FIG. 13 is a plan view detail before changing the editing element. FIG. 13 is a plan view detail after changing the editing element. FIG. 1 is a block diagram showing the process of generating and using a trained model. 13 is a flowchart showing a related element estimation process. FIG. 1 is a block diagram showing the process of generating and using a trained model. 13 is a flowchart showing a related element estimation process.

First Embodiment
A first embodiment of the present invention will now be described with reference to Figures 1 to 12.

First, the configuration of this embodiment will be described.
FIG. 1 is a diagram showing a hardware configuration of a drawing creation support device 100. As shown in FIG.

As shown in FIG. 1, the drawing creation support device 100 is composed of a CPU (Central Processing Unit) 30 that controls calculations and the entire system based on a control program, a ROM (Read Only Memory) 32 in which the control program etc. of the CPU 30 is stored in advance in a specified area, a RAM (Random Access Memory) 34 for storing data read from the ROM 32 etc. and calculation results required in the calculation process of the CPU 30, and an I/F (InterFace) 38 that mediates the input and output of data to external devices. These are connected to each other and capable of sending and receiving data by a bus 39, which is a signal line for transferring data.

Connected to the I/F 38 are external devices, such as an input device 40 consisting of a keyboard, mouse, etc. that can input data as a human interface, a storage device 42 that stores data, tables, etc. as files, and a display device 44 that displays a screen based on an image signal.

The storage device 42 has installed therein CAD (Computer Aided Design) software and BIM (Building Information Modeling) software (hereinafter collectively referred to as "CAD software"). CAD software is software that assists in the creation of drawings in response to user operations. When a request is made to start the CAD software, the CPU 30 starts a CAD program stored in a specified area of the ROM 32 and executes processing according to the program. A designer can start the CAD software to create plan drawings, detailed floor plans, and other architectural drawings.

Next, the data structure of the storage device 42 will be described.
FIG. 2 is a diagram showing the structure of finishing list data.

As shown in FIG. 2, the storage device 42 stores finishing table data relating to finishing tables.
The finishing table is a table that summarizes the finishes of each part of a building. There is an exterior finishing table that shows the finishes of the roof and exterior walls, and an interior finishing table that shows the finishes of the walls, floors, ceilings, etc. of each room. In the finishing table data, one row is registered for each finishing element. In the third row in Figure 2, for the finishing element "internal corridor", "tile carpet (A), border" is registered as the floor finish, "vinyl cloth (A)" is registered as the wall finish, and "vinyl cloth (A)" is registered as the ceiling finish. This indicates that the floor of the internal corridor is finished with "tile carpet (A), border", and the walls and ceiling of the internal corridor are finished with "vinyl cloth (A)". In addition, in the ninth row in Figure 2, "2450" is registered as the ceiling height for the finishing element "windbreak room". This indicates that the ceiling height of the windbreak room is finished to 2450 [mm].

FIG. 3 is a diagram showing the structure of CAD data of a plan view.
As shown in FIG. 3, the storage device 42 stores CAD data and BIM data of plan views (hereinafter collectively referred to as "CAD data").

A rectangular plan is a detailed drawing of the cross section of a building. CAD data for a rectangular plan is created by a designer using CAD software. A designer creates a rectangular plan in the CAD software by adding, editing, or deleting editable elements in the rectangular plan. In the example of Figure 3, the edit elements of the floor, wall, and ceiling of the area labeled "internal corridor" are arranged, as are the edit elements of the floor, wall, and ceiling of the area labeled "windbreak room." These correspond to the finishing elements in the third and ninth rows of the finishing table data in Figure 2. Corresponding finishing elements and editing elements may be associated by setting the same ID within the finishing table data and the CAD data.

FIG. 4 is a diagram showing the structure of CAD data of a detailed plan view.
FIG. 5 is a diagram showing wall types.

The storage device 42 stores CAD data of detailed plan views, as shown in FIG.
A detailed plan drawing is a drawing that draws the plan of a building in detail. CAD data of the detailed plan drawing is created by a designer using CAD software. The designer creates the detailed plan drawing by adding, editing, or deleting editable editing elements in the detailed plan drawing in the CAD software. In the detailed plan drawing, a tag is placed for the editing element of the wall so that the wall type can be identified. The tag is a three- or four-digit alphanumeric character surrounded by a square or ellipse, and the designer sets the value by referring to the wall type diagram in FIG. 5. In the example of FIG. 4, the tag "L4DA" is placed in the area displayed as "LD", and the tag "L4JA" is placed in the area displayed as "Bathroom/Dressing Room". These correspond to the finishing elements in the 13th and 17th rows of the finishing table data in FIG. 2. The corresponding finishing elements and editing elements may be associated with each other by setting the same ID in the finishing table data and the CAD data.

The storage device 42 stores CAD data of other architectural drawings, and data of trained models. The trained models are trained based on information about the finishing elements listed in the finishing schedule and information about compatible elements listed in the architectural drawings that fit the specifications of those finishing elements. The finishing schedule data and CAD data for plan drawings, detailed floor plans, and other architectural drawings are used for AI training, so the storage device 42 stores a large amount of data that was created in the past. This CAD data is in which each editing element in the architectural drawings fits the specifications of the finishing schedule data, and serves as training data for the AI training.

Next, the operation of this embodiment will be described.
FIG. 6 is a flowchart showing the trained model generation process.

FIG. 7 is a block diagram showing the process of generating and using a trained model.
The trained model generation process is a process executed to generate a trained model, and when executed by the CPU 30, the process proceeds to step S100 to execute a finishing table data analysis process, as shown in Fig. 6. In the finishing table data analysis process, the finishing table data is read from the storage device 42, and information on the finishing elements is extracted from the read finishing table data.

Then, the process proceeds to step S102. In step S102, a CAD data analysis process is executed. In the CAD data analysis process, CAD data of the plan, detailed floor plan, and other architectural drawings is read from the storage device 42, and information regarding the editing elements is extracted from the read CAD data.

Then, the process proceeds to step S104. In step S104, as shown in FIG. 7, a training dataset is generated based on the information extracted in steps S100 and S102, and the process proceeds to step S106, where the generated training dataset is input to a training program and a trained model is generated by the training program. The training program includes pre-learning parameters and hyperparameters, and performs training based on the input training dataset and hyperparameters to update the pre-learning parameters. The trained model is then output as the training result.

The trained model includes trained parameters in which pre-trained parameters have been updated by training, and an inference program. The inference program inputs finishing table data and edit elements, and based on the trained parameters, estimates the finishing elements of the finishing table data and the matching elements corresponding to the edit elements (edit elements that match the specifications of the finishing elements) from the input finishing table data and edit elements, and outputs the estimated matching elements. The input to the inference program is not limited to the finishing table data itself, and may be made by extracting finishing elements of the finishing table data that correspond to the input edit elements, as long as the correspondence can be specified by ID or the like. Note that the relationship between the input edit elements and the matching elements is determined by AI training, and therefore, although it shows a similar tendency to the contents of past finishing table data and CAD data, there is ambiguity in that they do not necessarily match exactly. However, this ambiguity can be reduced by the amount of training data and the training accuracy.

FIG. 8 is a flowchart showing the matching element estimation process.
The matching element estimation process is executed in response to a request from a user, and when executed by the CPU 30, the process first proceeds to step S150 as shown in Fig. 8. In step S150, the CAD software determines whether or not any of the editable elements in the architectural drawing has been designated by a user operation, and if it is determined that an edit element has been designated (YES), the process proceeds to step S152.

In step S152, the finishing table data corresponding to the CAD data currently being edited is read from the memory device 42, the edit elements specified in step S150 are obtained, and the process proceeds to step S154, where the trained model in the memory device 42 is used to estimate matching elements from the read finishing table data and the obtained edit elements. The estimation is performed by inputting the finishing table data and edit elements into the trained model and obtaining the matching elements output from the trained model.

Then, the process proceeds to step S156, where it is determined whether the edited element specified in step S150 and the matching element estimated in step S154 are inconsistent in terms of identification or content. If it is determined that they are inconsistent (YES), the process proceeds to step S158, where a message is displayed on the display device 44 indicating that there is a defect in the edited element, and the process ends.

On the other hand, if it is determined in step S156 that the edit element and the matching element do not mismatch (NO), or if it is determined in step S150 that an edit element is not specified (NO), the process ends.

[Example of creating a plan drawing]
Next, the operation of changing the editing elements will be described taking the case of creating a plan view as an example. It is assumed that the finishing table data corresponding to the CAD data of the plan view currently being edited is the content shown in FIG.

FIG. 9 is a plan view before the editing elements are changed.
FIG. 10 is a plan view after the editing element has been changed.

In the CAD software, the designer specifies edit element 400 for the floor of the area marked "internal corridor" in the plan of FIG. 9. At this time, edit element 400 is set as "vinyl floor tile (C)," whereas the corresponding specification element (third line) in the finishing schedule data of FIG. 2 is "tile carpet (A), bordered." Therefore, when the finishing schedule data of FIG. 2 and edit element 400 are input into the trained model, the matching element "tile carpet (A), bordered" is estimated, and a message 402 is displayed indicating that there is a defect in edit element 400. Following the message 402, the designer changes edit element 400 to "tile carpet (A), bordered," as shown in FIG. 10.

Next, in the CAD software, the designer specifies edit element 404 for the wall of the area marked "internal corridor" in the plan of FIG. 9. At this time, edit element 404 is set to "vinyl cross (B)," whereas the corresponding specification element (third line) in the finishing schedule data of FIG. 2 is "vinyl cross (A)." Therefore, when the finishing schedule data of FIG. 2 and edit element 404 are input into the trained model, the matching element "vinyl cross (A)" is estimated, and a message 406 is displayed indicating that there is a defect in edit element 404. The designer follows the message 406 and changes edit element 404 to "vinyl cross (A)," as shown in FIG. 10.

Next, in the CAD software, the designer specifies edit element 408 for the ceiling of the area marked "internal corridor" in the plan of FIG. 9. At this time, edit element 408 is set as "vinyl cloth (B)," whereas the corresponding specification element (third line) in the finishing schedule data of FIG. 2 is "vinyl cloth (A)." Therefore, when the finishing schedule data of FIG. 2 and edit element 408 are input into the trained model, the matching element "vinyl cloth (A)" is estimated, and a message 406 is displayed indicating that there is a defect in edit element 408. The designer follows the message 406 and changes edit element 408 to "vinyl cloth (A)," as shown in FIG. 10.

Similarly, in the CAD software, the designer specifies the ceiling height (edit element) 410 of the ceiling in the area marked "Windbreak" in the plan of Figure 9. Since the ceiling height 410 also does not conform to the specifications of the finishing schedule data in Figure 2, a message 412 is displayed indicating that there is a defect in the ceiling height 410. Following the message 412, the designer changes the ceiling height 410 to 2450 [mm], as shown in Figure 10.

In addition, the designer specifies the effective dimension 414 of the area marked "shared corridor" in the plan of FIG. 9 in the CAD software. Since the effective dimension 414 does not comply with building-related laws and regulations, a message 416 is displayed indicating that the effective dimension 414 is incomplete. Whether or not it complies with building-related laws and regulations can be determined, for example, by the technology of Patent Document 1. The effective dimension 414 can be changed by adjusting the position of the right wall or the left wall, but since it is difficult for AI to determine which should be prioritized, such editing elements 414 are editing elements that require human judgment to change, and are not changed automatically by AI, but are left to the designer's judgment and the designer changes the position of the right or left wall.

[Example of detailed floor plan drawing]
Next, the operation of changing the editing elements will be described taking the case of creating a detailed plan view as an example. It is assumed that the finishing table data corresponding to the CAD data of the detailed plan view currently being edited is the content shown in FIG.

FIG. 11 is a detailed plan view before the editing element is changed.
FIG. 12 is a plan view detail after the editing element has been changed.

In the CAD software, the designer specifies the tag (edit element) 420 of the area displayed as "Bathroom/Dressing Room" in the detailed floor plan of FIG. 11. At this time, the tag 420 is set as "L4JB," whereas the corresponding specification element (line 17) in the finishing schedule data of FIG. 2 is "L4," "GBt12.5," and "Vinyl Wall (C)" (= "L4JA" from the wall type diagram of FIG. 5). Therefore, when the finishing schedule data and tag 420 of FIG. 2 are input into the trained model, the matching element "L4JA" is estimated, and a message 422 is displayed indicating that there is a problem with the tag 420. The designer follows the message 422 and changes the tag 420 to "L4JA," as shown in FIG. 12.

Next, in the CAD software, the designer specifies the wall tag 424 of the area displayed as "LD" in the detailed plan view of FIG. 11. At this time, the tag 424 is set as "L4LA", whereas the corresponding specification element (line 13) in the finishing table data of FIG. 2 is "L4", "GBt12.5 * Surface base is waterproof GB", and "Vinyl cloth (C)" (= "L4DA" from the wall type diagram of FIG. 5). Therefore, when the finishing table data and tag 424 of FIG. 2 are input into the trained model, the matching element "L4DA" is estimated, and a message 426 is displayed indicating that there is a defect in the tag 424. The designer follows the message 426 and changes the tag 424 to "L4DA" as shown in FIG. 12.

Next, the effects of this embodiment will be described.
In this embodiment, finishing schedule data and edited elements are acquired, and a trained model is used to estimate suitable elements from the acquired finishing schedule data and edited elements, which has been trained based on information about the finishing elements listed in the finishing schedule and information about suitable elements listed in the architectural drawings that match the specifications of the finishing elements.If there is a mismatch between the acquired edited elements and the estimated suitable elements, a message is displayed indicating that there is a defect in the edited elements.

This is expected to allow editing elements to conform to the specifications of the finishing list.
In this embodiment, step S152 corresponds to the element information acquisition means of the first or second invention, step S154 corresponds to the estimation means of the first or second invention, and step S158 corresponds to the output means of the second invention.

Second Embodiment
Next, a second embodiment of the present invention will be described with reference to Figures 13 and 14.

This embodiment differs from the first embodiment in that it displays other edit elements that should be changed when an edit element is changed. When an edit element is changed, other edit elements may no longer conform to the specifications of the finishing table. For example, when the ceiling height 410 in FIG. 9 is changed, it becomes necessary to change the height of the space above the ceiling, etc. In this embodiment, the other edit elements that should be changed when an edit element is changed are displayed, thereby aiming to support the designer in making changes to the other edit elements that should be changed as well. Below, only the differences from the first embodiment will be explained, and explanations of the overlapping parts will be omitted.

First, the data structure of the storage device 42 will be described.
The memory device 42 stores CAD data for plan views, detailed floor plans, and other architectural drawings, including CAD data before an editing element is changed (hereinafter referred to as "CAD data before change") and CAD data after an editing element is changed (hereinafter referred to as "CAD data after change").

The storage device 42 stores data of the trained model. The trained model is trained based on information about editable elements in architectural drawings and information about other editable elements that have been changed as a result of changing those editable elements. The pre-change CAD data and post-change CAD data are used for training the AI, so the storage device 42 stores a large amount of data that has been created in the past.

Next, the operation of this embodiment will be described.
FIG. 13 is a block diagram showing the process of generating and using a trained model.

The trained model generation process is a process executed to generate a trained model, and when executed by the CPU 30, the pre-change CAD data and the post-change CAD data are read from the storage device 42, and information regarding the edited elements is extracted from the read CAD data.

Then, as shown in FIG. 13, a learning dataset is generated based on the extracted information, the generated learning dataset is input to a learning program, and a trained model is generated by the learning program. The inference program of the trained model inputs an edit element that has been changed among the editable edit elements in the architectural drawing, estimates from the input edit element other edit elements that should be changed in conjunction with the change to that edit element based on the trained parameters, and outputs the estimated edit elements as related elements.

FIG. 14 is a flowchart showing the related element estimation process.
The related element estimation process is executed in response to a request from a user, and when executed by the CPU 30, the process first proceeds to step S200 as shown in Fig. 14. In step S200, it is determined whether or not any of the editable elements in the architectural drawing has been changed in the CAD software, and if it is determined that the edit element has been changed (YES), the process proceeds to step S202.

In step S202, the changed edited element is acquired, and the process proceeds to step S204, where the related element is estimated from the acquired edited element using the trained model in the storage device 42. The estimation is performed by inputting the edited element into the trained model and acquiring the related element output from the trained model.

Then, the process proceeds to step S206, where other editing elements to be changed are displayed in a specific manner (e.g., highlighted) based on the estimated related elements, and the process ends.

On the other hand, if it is determined in step S200 that the edit element has not been changed (NO), the process ends.

Next, the effects of this embodiment will be described.
In this embodiment, the edited element that has been changed is obtained, and related elements are estimated from the obtained edited element using a trained model that has been trained based on information about editable edited elements in an architectural drawing and information about other edited elements that have been changed as a result of the change to that edited element, and the other edited elements that should be changed are displayed in a specific manner based on the estimated related elements.

This allows other editing elements that need to be changed to also be changed.
In this embodiment, step S202 corresponds to the element information acquisition means of invention 4, step S204 corresponds to the estimation means of invention 4 or 5, and step S206 corresponds to the output means of invention 5.

Third embodiment
Next, a third embodiment of the present invention will be described with reference to Figures 15 and 16.

This embodiment differs from the second embodiment in that it estimates editing elements that require human judgment to change. When an editing element is changed, it is desirable to automatically change other editing elements in conjunction with the change, but editing elements that require human judgment to change must be excluded from the targets of automatic change. For example, when changing the effective dimension 414 in FIG. 9, the designer decides whether to change the right or left wall position. In this embodiment, other editing elements that should be changed in conjunction with a change in an editing element are automatically changed, and editing elements that require human judgment to change are excluded from the targets of automatic change, with the aim of supporting the user in understanding whether or not other editing elements and their contents should be changed by a human. Below, only the differences from the first and second embodiments will be described, and overlapping parts will be omitted.

First, the data structure of the storage device 42 will be described.
The storage device 42 stores judgment necessity data in which edit elements are registered in association with information on whether a change requires human judgment. Edit elements that require human judgment for change may be set by human judgment. Also, edit elements that have two or more change patterns before and after the change may be extracted based on the pre-change CAD data and the post-change CAD data, and the extracted edit elements may be set as edit elements that require human judgment for change.

The storage device 42 stores data of the trained model. The trained model is trained based on information about editable editing elements in architectural drawings, information about other editing elements and their contents that have been changed as a result of changing the editing elements, and judgment necessity information about whether or not a human judgment is required to change the other editing elements. The pre-change CAD data, post-change CAD data, and judgment necessity data are used for training the AI, so the storage device 42 stores a large amount of data created in the past.

Next, the operation of this embodiment will be described.
FIG. 15 is a block diagram showing the process of generating and using a trained model.

The trained model generation process is a process executed to generate a trained model, and when executed by the CPU 30, the pre-change CAD data, post-change CAD data, and judgment necessity data are read from the storage device 42, and information regarding the edit elements and judgment necessity information are extracted from the read CAD data and judgment necessity data.

Then, as shown in FIG. 15, a learning dataset is generated based on the extracted information, the generated learning dataset is input to a learning program, and a trained model is generated by the learning program. The inference program of the trained model inputs an edit element that has been changed among the editable edit elements in the architectural drawing, and estimates from the input edit element, based on the trained parameters, other edit elements that should be changed in conjunction with the change to that edit element and information on whether or not a decision is required, and outputs the estimated edit elements as related elements.

FIG. 16 is a flowchart showing the related element estimation process.
The related element estimation process is executed in response to a request from a user, and when executed by the CPU 30, the process first proceeds to step S250 as shown in Fig. 16. In step S250, it is determined whether or not any of the editable elements in the architectural drawing has been changed in the CAD software, and if it is determined that the edit element has been changed (YES), the process proceeds to step S252.

In step S252, the changed edited element is acquired, and the process proceeds to step S254, where the related elements and the decision necessity information are estimated from the acquired edited element using the trained model in the storage device 42. The estimation is performed by inputting the edited element into the trained model, and acquiring the related elements and the decision necessity information output from the trained model.

Then, the process proceeds to step S256, where it is determined based on the estimated judgment necessity information whether or not human judgment is required to change the estimated related elements. If it is determined that human judgment is not required to change the related elements (NO), the process proceeds to step S258, where the other edited elements to be changed are changed based on the estimated related elements so that their contents become the contents of the related elements, and the process ends.

On the other hand, if it is determined in step S256 that changing the related elements requires human judgment (YES), the process proceeds to step S260, where other edited elements to be changed based on the estimated related elements are displayed in a specific manner (e.g., highlighted), and the process ends.

On the other hand, if it is determined in step S250 that the edit element has not been changed (NO), the process ends.

Next, the effects of this embodiment will be described.
In this embodiment, an edit element that has been changed among editable edit elements in an architectural drawing is obtained, and related elements and judgment necessity information are estimated from the obtained edit element using a trained model that has been trained based on information about the editable edit element in the architectural drawing, information about other edit elements and their contents that have been changed as a result of the change to that edit element, and judgment necessity information regarding whether human judgment is required to change the other edit elements.If it is determined based on the estimated judgment necessity information that human judgment is required to change the related element, the other edit elements are displayed in a specific manner based on the estimated related elements, and if it is determined that human judgment is not required to change the related elements, the other edit elements are changed based on the estimated related elements.

This allows the person to understand whether other editing elements and content need to be changed or not.

In this embodiment, step S252 corresponds to the element information acquisition means of invention 7 or 9, step S254 corresponds to the estimation means of inventions 7 to 11, step S258 corresponds to the change means of invention 11, and step S260 corresponds to the output means of invention 8 or 10.

[Modifications]
In the first embodiment described above, defects were displayed for edit elements specified by user operation, but this is not limited to this. It is also possible to adopt a configuration in which all edit elements in the drawing are scanned and defects are displayed all at once for edit elements that have defects.

In addition, in the first embodiment and its modified example, a message is displayed indicating that there is a defect in the edited element, but this is not limiting, and a configuration can also be adopted in which the indicated edited element is changed based on the estimated matching element. In this case, the trained model used is one that has been trained based on information about the finishing element listed in the finishing table, as well as information about matching elements and their contents that are listed in the architectural drawing and that match the specifications of the finishing element.

In addition, in the second embodiment and its modified example, the other edit elements to be changed are displayed in a specific manner based on the estimated related elements, but this is not limiting, and a configuration in which the other edit elements to be changed are changed based on the estimated related elements can also be adopted. In this case, the trained model used is one that has been trained using a trained model that has been trained based on information about editable edit elements in architectural drawings and information about other edit elements and their contents that have been changed as a result of the change in the edit element.

In addition, in the third embodiment and its modified examples, other edit elements that should be changed are changed or displayed in a specific manner based on the estimated related elements, but this is not limited to the above. If the configuration does not involve changes, the trained model can be one that has been trained based on information about editable edit elements in architectural drawings, information about other edit elements that have been changed as a result of the change to that edit element, and judgment necessity information regarding whether or not human judgment is required to change the other edit elements. Also, if the configuration does not involve display in a specific manner (for example, a configuration that only estimates whether or not human judgment is required), the trained model can be one that has been trained based on information about editable edit elements in architectural drawings and judgment necessity information regarding whether or not human judgment is required to change the other edit elements that have been changed as a result of the change to that edit element).

In addition, in the second and third embodiments and their variations, a trained model is used that has been trained based on information about editable edit elements in architectural drawings and information about other edit elements that have been changed as a result of changes to those edit elements. However, this is not limiting, and a trained model that has been trained based on information about editable edit elements in architectural drawings and information about other edit elements that should be changed as a result of changes to those edit elements can also be used. In other words, instead of pre-change CAD data and post-change CAD data, correspondence information, such as changing edit element B when edit element A is changed, can be used to train the AI.

In addition, in the first to third embodiments and their variations, a trained model is generated and the trained model is used, but this is not limiting, and a configuration that utilizes a large language model can also be adopted. Specifically, for example, the following configuration can be adopted.

The drawing creation support device 100 is connected to a large-scale language model server having a large-scale language model via a network. The large-scale language model server is configured with the same hardware configuration as the drawing creation support device 100. The large-scale language model is a deep learning model that pre-learns from a huge amount of data what is called a language model that models human spoken words based on their occurrence probability. When the large-scale language model server receives a request, it uses the large-scale language model to statistically estimate the generation probability of the next word from the sentence included in the received request, and transmits the estimation result to the request source. As the large-scale language model, for example, known technologies described on the Internet sites "https://chatgpt-lab.com/n/n418d3aa56f0b" and "https://agirobots.com/chatgpt-mechanism-and-problem/" can be adopted.

The following inventions A1 to A3 or B1 to B8 can be applied to the drawing creation support device 100.

[Invention A1] A request input means for inputting a request including finishing element information on finishing elements described in a finishing table and editing element information on editable editing elements in an architectural drawing, and including a request for generating matching element information on matching elements described in the architectural drawing that match the specifications of the finishing elements, into a large-scale language model;
The language model further includes an element information acquiring means for acquiring matching element information output from the large-scale language model in response to the request.

[Invention A2] In Invention A1,
The system further comprises an output means for outputting information on an edited element related to the edited element information based on the adapted element information acquired by the element information acquisition means.

The embodiments of inventions A1 and A2 will be described with reference to FIG. 9. This is a modification of the first embodiment, in which the process of step S154 is replaced with a process of acquiring matching element information from a large-scale language model server.

When the designer designates, in the CAD software, an edit element 400 for the floor of the area marked "internal corridor" in the plan of FIG. 9, the drawing creation support device 100 sends a request to the large-scale language model server. The request includes the finishing schedule data and edit element 400 of FIG. 2, and also includes a request to generate a matching element for the edit element 400. In response to this request, the large-scale language model server outputs matching element information. When the drawing creation support device 100 receives the matching element information, a message 402 is displayed indicating that there is a defect in the edit element 400 based on the received matching element information.

[Invention A3] In Invention A1 or A2,
The request includes finish element information on a finish element described in a finish schedule and edit element information on an editable edit element in an architectural drawing, and includes a request to generate conforming element information on a conforming element and its contents that conforms to the specifications of the finish element and is described in the architectural drawing;
The system further includes a change means for changing an edit element related to the edit element information based on the suitable element information acquired by the element information acquisition means.

The following describes the embodiments of inventions A1 and A3 using FIG. 9 as an example. These are modifications of the first embodiment described above.

When the designer designates, in the CAD software, an edit element 400 for the floor of the area marked "internal corridor" in the plan of FIG. 9, the drawing creation support device 100 sends a request to the large-scale language model server. The request includes the finishing schedule data and edit element 400 of FIG. 2, and also includes a request to generate a matching element for the edit element 400. In response to this request, the large-scale language model server outputs matching element information. When the drawing creation support device 100 receives the matching element information, it changes the edit element 400 based on the received matching element information.

[Invention B1] A request input means for inputting a request including changed element information on an edited element among editable edit elements in an architectural drawing that has been changed, and including a request for generating related element information on other edit elements that should be changed in conjunction with the change in the edited element, into a large-scale language model;
The method further comprises: acquiring element information about the related elements output from the large-scale language model in response to the request.

[Invention B2] In Invention B1,
The device further comprises an output means for outputting information on other edit elements related to the related element information based on the related element information acquired by the element information acquisition means.

The embodiments of inventions B1 and B2 will be described with reference to FIG. 9. This is a modification of the second embodiment, in which the process of step S202 is replaced with a process of acquiring related element information from a large-scale language model server.

When the designer changes the edit element 400 of the floor in the area marked "internal corridor" in the plan of Figure 9, the drawing creation support device 100 sends a request to the large-scale language model server. The request includes the edit element 400, and also includes a request to generate related elements of the edit element 400. In response to this request, the large-scale language model server outputs related element information. When the drawing creation support device 100 receives the related element information, other edit elements that should be changed in conjunction with the change to the edit element 400 are displayed in a specific manner based on the received related element information.

[Invention B3] In Invention B1 or B2,
the request includes changed element information on an edit element that has been changed among editable edit elements in the architectural drawing, and includes a request to generate other edit elements that should be changed in conjunction with the change in the edit element and related element information on the contents of the other edit elements;
The editing device further includes a change means for changing, based on the related element information acquired by the element information acquisition means, another edit element related to the related element information.

The following describes the embodiments of inventions B1 and B3 using FIG. 9 as an example. These are modifications of the second embodiment described above.

When the designer changes the edit element 400 of the floor in the area marked "internal corridor" in the plan of Figure 9, the drawing creation support device 100 sends a request to the large-scale language model server. The request includes the edit element 400, and also includes a request to generate related elements of the edit element 400. In response to this request, the large-scale language model server outputs related element information. When the drawing creation support device 100 receives the related element information, it changes other edit elements that should be changed in conjunction with the change to the edit element 400 based on the received related element information.

[Invention B4] A request input means for inputting a request including changed element information on an edited element among editable edit elements in an architectural drawing that has been changed, and a request for generating judgment necessity information on whether a human judgment is required for changing other edit elements that should be changed in conjunction with the change of the edited element, into a large-scale language model;
The system further comprises a judgment necessity information acquisition means for acquiring judgment necessity information output from the large-scale language model in response to the request.

[Invention B5] In Invention B4,
The apparatus further comprises an output means for outputting information on whether or not a human judgment is required for changing another editing element related to the judgment necessity information, based on the judgment necessity information acquired by the judgment necessity information acquisition means.

[Invention B6] A request input means for inputting a request including changed element information on an edited element that has been changed among editable edit elements in an architectural drawing, and also including a request for generating other edit elements that should be changed in conjunction with the change in the edited element and related element information on the contents of the other edit elements, and judgment necessity information on whether or not human judgment is required to change the other edit elements, into the large-scale language model;
The method further comprises an element information acquiring means for acquiring related element information and judgment necessity information outputted from the large-scale language model in response to the request.

[Invention B7] In Invention B6,
If it is determined based on the judgment necessity information acquired by the element information acquisition means that human judgment is required to change other editing elements related to the related element information, the device is provided with an output means for outputting information regarding the other editing elements based on the related element information acquired by the element information acquisition means.

[Invention B8] In Invention B6 or B7,
If it is determined based on the judgment necessity information acquired by the element information acquisition means that no human judgment is required to change other editing elements related to the related element information, a modification means is provided for changing the other editing elements based on the related element information acquired by the element information acquisition means.

The following describes embodiments of inventions B4 to B8 using FIG. 9 as an example. This is a modification of the third embodiment, in which the process of step S254 is replaced with a process of acquiring related element information and judgment necessity information from a large-scale language model server.

When the designer changes the edit element 400 of the floor in the area marked "internal corridor" in the plan of FIG. 9, the drawing creation support device 100 sends a request to the large-scale language model server. The request includes the edit element 400, and also includes a request to generate related elements of the edit element 400 and judgment necessity information regarding whether human judgment is required to change the related elements. In response to this request, the large-scale language model server outputs the related element information and the judgment necessity information. When the drawing creation support device 100 receives the related element information and the judgment necessity information, it determines whether human judgment is required to change the related elements based on the received related element information. As a result, if it is determined that human judgment is required, the related elements are displayed in a specific manner based on the received related element information. On the other hand, if it is determined that human judgment is not required, the related elements are changed based on the received related element information.

In addition, in the first to third embodiments and their variations, the device is realized as a single device, but the present invention is not limited to this and can also be realized as a network system. As an example of a network system, some or all of the functions of the drawing creation support device 100 can be configured as a virtual server on a server that provides cloud computing services.

In addition, in the first to third embodiments and their variations, the drawing creation support device 100 is configured to use the storage device 42, but this is not limiting, and it can also be configured to use an external storage device such as a database server.

In addition, in the above first to third embodiments and their modifications, the processes shown in the flowcharts of Figures 6, 8, 14, and 16 are executed by executing a program that is pre-stored in ROM 32. However, this is not limiting, and the program showing these procedures may be loaded into RAM 34 from a storage medium on which the program is stored and executed.

Here, storage media refers to semiconductor storage media such as RAM and ROM, magnetic storage media such as FD and HD, optically readable storage media such as CD, CDV, LD and DVD, and magnetic storage/optically readable storage media such as MO, and includes any storage media that can be read by a computer, regardless of the reading method (electronic, magnetic, optical, etc.).

Moreover, the first to third embodiments and the modifications thereof can be applied to each other.
Furthermore, the present invention is not limited to the above-described first to third embodiments and their modifications, but can also be applied to other cases without departing from the spirit of the present invention.

100...Drawing creation support device, 30...CPU, 32...ROM, 34...RAM, 38...I/F, 39...bus, 40...input device, 42...storage device, 44...display device, 400, 404, 408, 410...editing elements, 402, 406, 412, 416, 422, 426...defect display, 414...effective dimensions, 420, 424...tags

Claims (11)

An element information acquiring means for acquiring finishing element information relating to finishing elements described in a finishing table and editing element information relating to editable editing elements in an architectural drawing;
An architectural drawing creation support system comprising: an estimation means for estimating the suitable element information from the finishing element information and edited element information acquired by the element information acquisition means using a trained model that has been trained based on information regarding the finishing elements described in the finishing table and suitable element information regarding suitable elements that are compatible with the specifications of the finishing elements and are described in the architectural drawing. In claim 1,
An architectural drawing creation support system comprising: an output means for outputting information regarding edited elements related to edited element information acquired by the element information acquisition means based on the compatible element information estimated by the estimation means. In any one of claims 1 and 2,
The trained model is trained based on information on the finishing elements described in the finishing table and information on the matching elements and their contents described in the architectural drawings that conform to the specifications of the finishing elements,
An architectural drawing creation support system comprising a modification means for modifying an edit element related to the edit element information acquired by the element information acquisition means based on the suitable element information estimated by the estimation means. An element information acquisition means for acquiring changed element information regarding an edit element that has been changed among editable edit elements in an architectural drawing;
An architectural drawing creation support system comprising: an estimation means for estimating the related element information from the changed element information acquired by the element information acquisition means using a trained model that has been trained based on information regarding editable edit elements in an architectural drawing and related element information regarding other edit elements that should be changed or have been changed due to a change to the edit element. In claim 4,
An architectural drawing production support system comprising: an output means for outputting information regarding other edit elements related to the related element information based on the related element information estimated by the estimation means. In any one of claims 4 and 5,
The trained model is trained based on information about editable edit elements in architectural drawings, and related element information about other edit elements that should be changed or have been changed due to a change in the edit element and their contents,
An architectural drawing production support system comprising: a change means for changing other edit elements related to the related element information based on the related element information estimated by the estimation means. An element information acquisition means for acquiring changed element information regarding an edit element that has been changed among editable edit elements in an architectural drawing;
An architectural drawing creation support system comprising: an estimation means for estimating judgment necessity information from changed element information acquired by the element information acquisition means using a trained model that has been trained based on information regarding editable edit elements in an architectural drawing and judgment necessity information regarding whether human judgment is required to change other edit elements that should be changed or have been changed due to a change to the edit element. In claim 7,
An architectural drawing creation support system characterized by comprising an output means for outputting information regarding whether or not human judgment is required to change other editing elements related to the judgment necessity information estimated by the estimation means, based on the judgment necessity information estimated by the estimation means. An element information acquisition means for acquiring changed element information regarding an edit element that has been changed among editable edit elements in an architectural drawing;
An architectural drawing creation support system comprising: an estimation means for estimating the related element information and the judgment necessity information from the changed element information acquired by the element information acquisition means, using a trained model that has been trained based on information regarding editable edit elements in an architectural drawing, related element information regarding other edit elements and their contents that should be changed or have been changed due to a change to the edit element, and judgment necessity information regarding whether human judgment is required to change the other edit elements. In claim 9,
An architectural drawing creation support system characterized by having an output means for outputting information regarding the other edited elements based on the related element information estimated by the estimation means when it is determined that human judgment is required to change other edited elements related to the related element information based on the judgment necessity information estimated by the estimation means. In any one of claims 9 and 10,
An architectural drawing creation support system characterized by having a modification means for modifying other editing elements related to the related element information based on the related element information estimated by the estimation means when it is determined that no human judgment is required to change the other editing elements related to the related element information based on the judgment necessity information estimated by the estimation means.

Images