Design support device, design support method, and design support program
The design support device addresses the issue of late-stage corrections by using AI and knowledge graphs to predict and classify design warnings, enhancing the efficiency of building design by alerting designers early.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- OHBAYASHI GUMI LTD
- Filing Date
- 2024-12-09
- Publication Date
- 2026-06-19
AI Technical Summary
Existing building design methods fail to alert designers of potential concerns early in the design phase, leading to time-consuming corrections in later stages.
A design support device that includes a control unit to obtain specifications, predict cautionary information, classify it by design type, and output warnings to user devices, utilizing AI and knowledge graphs to analyze building plans and regional conditions.
Supports efficient building design by alerting designers to potential concerns early, enabling timely corrections and improving the design process efficiency.
Smart Images
Figure 2026100333000001_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to a design support device, a design support method, and a design support program for supporting a design based on a plan for a building.
Background Art
[0002] Techniques for proposing a building plan in accordance with customer requirements have been studied (see Patent Document 1). In the technique described in this document, the control unit displays the site range specified by the user on a map, and within the site range, displays the building range specified by the user on the map. Then, a building shape corresponding to the building range is displayed. In this case, the building shape is displayed based on the length of one or more spans set in advance.
[0003] Also, techniques for obtaining defect information of a target building in consideration of the performance information of buildings constructed in the past have been studied (see Patent Document 2). In the technique described in this document, the acquisition unit of the building information processing device acquires the design information of the target building. Next, the defect information acquisition unit inputs the design information acquired by the acquisition unit into a learned model that has been previously learned based on learning data in which the design information of the learning building is associated with the defect information representing the defects of the learning building, and acquires the defect information of the target building.
Prior Art Documents
Patent Documents
[0004]
Patent Document 1
Patent Document 2
Summary of the Invention
Problems to be Solved by the Invention
[0005] Patent Document 1 describes a method for proposing a building outline to a client before the detailed design of the building is completed. Patent Document 2 describes a method for presenting information about potential defects in a building during the design phase. However, if concerns are included during the design phase, failing to alert the designer at an early stage will result in time-consuming corrections in later stages. [Means for solving the problem]
[0006] A design support device for solving the above problems includes a control unit connected to a user device. The control unit obtains specifications from the user device from information contained in the materials to be considered for the building, obtains related information related to the specifications, predicts cautionary information for the design of the building, classifies the predicted cautionary information according to the design type, and outputs it to the user device responsible for the design type. [Effects of the Invention]
[0007] This disclosure can support efficient design. [Brief explanation of the drawing]
[0008] [Figure 1] This is an explanatory diagram of the system according to the embodiment. [Figure 2] This is an explanatory diagram of the hardware configuration of the embodiment. [Figure 3] This is an explanatory diagram of the processing procedure of the embodiment. [Figure 4] This is an explanatory diagram of the information used in the embodiment. [Figure 5] This is an explanatory diagram of the information used in the embodiment. [Modes for carrying out the invention]
[0009] The following describes one embodiment of the design support device, design support method, and design support program, using Figures 1 to 5. This embodiment describes a case in which concerns that may arise during the detailed review stage (subsequent stage) are output in the plan summary created during the planning stage (preliminary stage) before detailed review. The plan summary includes materials such as proposals and presentation materials used in architectural competitions, etc., to convey the overall concept and design direction of the project to stakeholders such as the client (ordering party) before detailed design begins. As shown in Figure 1, the design support device A1 of this embodiment comprises a user device 10, a support device 20, and an AI system 30 that are interconnected via a network.
[0010] (Description of hardware configuration) Figure 2 illustrates the hardware configuration of the information processing device H10, which constitutes the user device 10, the support device 20, and the AI system 30. The information processing device H10 includes a communication device H11, an input device H12, a display device H13, a storage device H14, and a processor H15. Note that this hardware configuration is just one example, and it can be implemented with other hardware.
[0011] Communication device H11 is an interface that establishes a communication path with other devices and performs data transmission and reception. The input device H12 is a device that accepts input of various types of information. The display device H13 is a display or the like that displays various types of information.
[0012] The storage device H14 is a storage device that stores data and various programs for executing various functions of the user device 10, the support device 20, and the AI system 30. Examples of storage devices H14 include ROM, RAM, and hard disks.
[0013] Processor H15 controls each process in user device 10, support device 20, and AI system 30 by using programs and data stored in storage device H14. Examples of processor H15 include, for example, a CPU or an MPU. This processor H15 expands a program stored in a ROM or the like to a RAM and executes various processes for each process.
[0014] Processor H15 is not limited to performing software processing for all processes it executes. For example, processor H15 may include a dedicated hardware circuit (e.g., an application-specific integrated circuit: ASIC) that performs hardware processing for at least a part of the processes it executes. That is, processor H15 may be configured as follows.
[0015] [1] One or more processors operating according to a computer program (software) [2] One or more dedicated hardware circuits that execute at least a part of various processes [3] A circuit (circuitry) including combinations thereof The processor includes a CPU and memories such as a RAM and a ROM, and the memories store program codes or instructions configured to cause the CPU to execute processes. The memory, that is, the computer-readable medium, includes any available medium that can be accessed by a general-purpose or dedicated computer.
[0016] (Design support device A1) The design support device A1 will be described with reference to FIG. 1. The user device 10 is a computer terminal used by the user.
[0017] The support device 20 is a computer system that supports the confirmation of materials under consideration. The support device 20 includes a control unit 21, a case information storage unit 22 (knowledge database), a regional information storage unit 23, and a consideration target storage unit 24.
[0018] The control unit 21 performs processes (processes including an acquisition stage, a recognition stage, a pointing-out stage, etc.) described later. By executing a processing program for this purpose, the control unit 21 functions as an acquisition unit 211, a recognition unit 212, a pointing-out unit 213, etc.
[0019] The acquisition unit 211 executes a process of acquiring a project outline document (document to be considered). In the present embodiment, a file to be considered provided to a client (orderer) is acquired. The file to be considered can use a digital document created by a document creation application, a presentation creation application, etc., or a scanned document created by scanning a paper medium document.
[0020] The recognition unit 212 executes a recognition process (character recognition process, image recognition process, etc.) for recognizing characters and images included in the file to be considered. By this recognition process, the information (characters and images) included in the file to be considered is texturized, and characteristic words (words) representing the characteristics of the building are extracted. The pointing-out unit 213 executes a process of extracting concerns in subsequent processes such as the design of the building using the characteristic words recognized in the file to be considered.
[0021] Case information storage unit 22 records case management information about attention cases. This case management information is recorded when an attention-worthy case is registered. The case management information includes a knowledge graph generated from characteristic words extracted from the design document, etc. of the attention case. In this knowledge graph, a graph structure linked by edges indicating relationships is used with characteristic words as nodes. In this graph structure, it is composed of an ontology (concept) and an instance (word) that systematize knowledge and know-how. Thereby, a structure indicating the relevance between concepts is formed.
[0022] This knowledge graph collects cases of warnings and criticisms from reports, meeting minutes, etc., classifies these cases into categories such as design type, location, cause, scope of impact, and countermeasures, and defines entities (nodes). Then, these entities are incorporated into the graph as nodes, and their relationships as edges. This knowledge graph is generated using past warning cases and is predictive information for predicting new concerns. In this embodiment, conditions applicable to the building specifications included in the plan summary can be acquired as knowledge.
[0023] The regional information storage unit 23 stores regional management information regarding regional specific conditions in the area where the building is constructed. This regional management information is recorded when regional specific conditions in the area are registered. This regional management information includes information about regional specific conditions for a regional identifier. These region-specific conditions include design requirements that take into account local laws and regulations, seismic resistance, environmental protection and ecological considerations, climate and weather, ground conditions, infrastructure and transportation conditions, and disaster prevention measures.
[0024] Local laws and regulations include design requirements regarding building height, appearance, use, and sunlight, which are restricted depending on the area where the construction is planned. For example, in areas with tourist attractions or historical streetscapes, design requirements regarding building design and height will be in accordance with landscape regulations.
[0025] Seismic resistance requirements include design conditions for foundation work in earthquake-prone areas, taking into account the strength and geological conditions of the local ground. Environmental protection and ecological considerations include design conditions that address the protection of the natural environment and preservation of green spaces in the area where the construction is planned.
[0026] Climate and weather conditions include design requirements (ventilation performance, snow-resistant structure, etc.) that are appropriate to the climate and weather conditions of the planned construction site (e.g., temperature and snowfall). Ground conditions include design conditions for ground improvement work, etc., depending on the stability of the ground at the planned construction site (e.g., soft ground or possibility of liquefaction).
[0027] Infrastructure and traffic conditions include design conditions that take into account the infrastructure status such as water, electricity, gas, and sewage systems, as well as traffic conditions at the planned construction site. Disaster prevention measures include design conditions such as building placement and height, evacuation routes, elevated structures, and drainage plans, depending on the area's risk of landslides and flooding.
[0028] The storage unit 24 for items under consideration records management information for items under consideration, which confirms the presence or absence of concerns. This management information for items under consideration is recorded when the materials for items under consideration are acquired. This management information for items under consideration includes the item identification, the file under consideration, concerns, and information about the relevant parties.
[0029] The subject of consideration identifier is an identifier used to identify the subject of consideration. The files under consideration are those obtained from user device 10. The concerns are those identified based on the files under review. Stakeholder information refers to information that identifies the stakeholders (design personnel) involved in this concern.
[0030] The AI system 30 is a computer system that utilizes generative AI technology based on a large-scale language model that generates responses to prompts. In this embodiment, the AI system 30 predicts the presence or absence of concerns based on prompts in a case under consideration and outputs a response (warning information) regarding whether or not a warning is necessary.
[0031] (Design support processing) Next, we will explain the design support process using Figure 3. First, the control unit 21 of the support device 20 executes the process of acquiring the documents to be reviewed (step S11). Specifically, the acquisition unit 211 of the control unit 21 displays an upload screen on the user device 10. This upload screen includes an input field for the documents to be reviewed. When a file to be reviewed (in this case, a plan summary) is entered into the input field for the documents to be reviewed on the user device 10, the acquisition unit 211 of the control unit 21 acquires the file to be reviewed from the user device 10 and assigns a document to be reviewed identifier to it. Then, the acquisition unit 211 records the document to be reviewed management information, including the file to be reviewed, in the document to be reviewed storage unit 24, associating it with the document to be reviewed identifier.
[0032] Next, the control unit 21 of the support device 20 performs character recognition processing (step S12). Specifically, the recognition unit 212 of the control unit 21 extracts the characters contained in the file under consideration as text. In this case, the recognition unit 212 extracts feature words that will become nodes and edges based on the morphemes contained in the text.
[0033] As shown in Figure 4, in this embodiment, the recognition unit 212 obtains characteristic words 511 extracted from text regarding the specifications of the building from each item such as the project overview and future outlook described in the document under consideration 500.
[0034] Next, the control unit 21 of the support device 20 performs image recognition processing (step S13). Specifically, the recognition unit 212 of the control unit 21 represents images such as diagrams and photographs contained in the file under consideration as text. Then, the recognition unit 212 obtains characteristic words about the building specifications from the morphemes contained in the represented text.
[0035] As shown in Figure 4, in this embodiment, characteristic words 512 are extracted from images regarding the specifications of the building from various drawings and photographs such as maps, exterior drawings, floor plans, and interior designs described in the document under consideration 500.
[0036] Next, the control unit 21 of the support device 20 performs the process of acquiring location information (step S14). Specifically, the recognition unit 212 of the control unit 21 identifies the address from the location information contained in the file under consideration. Then, the recognition unit 212 acquires the regional attributes of the area containing this address as feature words from the regional information storage unit 23. As shown in Figure 4, in this embodiment, location information 520 is obtained from the location (address) described in the document under consideration 500.
[0037] Next, the control unit 21 of the support device 20 executes the process of acquiring the graph under consideration (step S15). Specifically, the identification unit 213 of the control unit 21 generates a knowledge graph (graph under consideration) by linking nodes with edges using the characteristic words acquired from the document under consideration 500. As shown in Figure 4, in this embodiment, the graph under consideration 530 is obtained using an ontology with the feature words 511, 512 and location information 520.
[0038] Next, the control unit 21 of the support device 20 performs knowledge acquisition processing using the knowledge graph (step S16). Specifically, the identification unit 213 of the control unit 21 searches the generated target graph in the knowledge graph recorded in the case information storage unit 22. Then, the identification unit 213 uses the extracted knowledge graph to identify knowledge (design conditions applied to the building) as related information.
[0039] As shown in Figure 4, in this embodiment, knowledge 540 is acquired using the graph under consideration. This knowledge 540 becomes cautionary information regarding design conditions corresponding to the knowledge graph. As shown in Figure 5, a search target graph is generated using characteristic words contained in the document under consideration 500, and the knowledge graph related to this search target graph is obtained as knowledge 540.
[0040] Next, the control unit 21 of the support device 20 performs AI input processing, including knowledge (step S17). Specifically, the identification unit 213 of the control unit 21 inputs the identified knowledge, along with the documents under consideration, into the AI system 30. In this case, it instructs the system to output concerns according to the files under consideration. The AI system 30 may also receive a graph under consideration as input.
[0041] Next, the control unit 21 of the support device 20 performs AI output acquisition processing (step S18). Specifically, the recognition unit 212 of the control unit 21 obtains a response from the AI system 30. This response includes information on concerns and design types (architectural, structural, and equipment) in relation to the building's condition, according to the design conditions. The concerns include items that may not meet the design conditions.
[0042] Next, the control unit 21 of the support device 20 performs user type identification processing (step S19). Specifically, the identification unit 213 of the control unit 21 sets the type of person in charge (design person in charge) according to the design type. In this embodiment, depending on the design type, architectural design, structural design, and equipment design are assigned as the types of person in charge.
[0043] As shown in Figure 4, in this embodiment, caution information 551 (for architectural design), caution information 552 (for structural design), and caution information 553 (for equipment design) are acquired. Note that caution information 551 to 553 may include common concerns. Also, concerns specific to the design type may be included only in the caution information specific to the design type. In this case, caution information including concerns corresponding to the type of person in charge will be generated.
[0044] Next, the control unit 21 of the support device 20 performs weighting processing (step S20). Specifically, the identification unit 213 of the control unit 21 classifies the concerns according to the magnitude of their impact on subsequent processes (e.g., design changes). Here, the identification unit 213 uses the case information storage unit 22 to identify the scope of influence of the design conditions of the concerns. The identification unit 213 then assigns a higher weight to concerns with a wider scope of influence. For example, it identifies words related to the overall concept according to the attributes of the characteristic words. Concerns related to concepts, etc., included in the materials under consideration are assigned a higher weight because they have a greater impact on other designs. In addition, the weighting may be changed according to the type of person in charge (architectural design, structural design, equipment design) depending on the type of design.
[0045] Next, the control unit 21 of the support device 20 performs the output processing of caution information (step S21). Specifically, the identification unit 213 of the control unit 21 outputs caution information, including concerns corresponding to the type of person in charge, to each user device 10 for architectural design, structural design, and equipment design. In this case, the identification unit 213 changes the display format of the concerns according to the weighting level. For example, by outputting the caution information in order of highest weighting, it prompts the user to pay attention.
[0046] (Operation of this embodiment) Since cautionary information is generated using the plan summary document before detailed review, the person in charge will be alerted at an early stage.
[0047] (Effects of this embodiment) (1) In this embodiment, the control unit 21 of the support device 20 performs character recognition processing (step S12). This makes it possible to obtain information about buildings from the text contained in the document under consideration.
[0048] (2) In this embodiment, the control unit 21 of the support device 20 performs image recognition processing (step S13). This makes it possible to obtain information about buildings from images included in the materials under consideration.
[0049] (3) In this embodiment, the control unit 21 of the support device 20 performs location information acquisition processing (step S14). This makes it possible to acquire area-specific cautionary information according to the location where the building is constructed.
[0050] (4) In this embodiment, the control unit 21 of the support device 20 executes the process of acquiring the graph under consideration (step S15). This makes it possible to identify the relationships between the texts acquired from the data under consideration.
[0051] (5) In this embodiment, the control unit 21 of the support device 20 performs knowledge acquisition processing using the knowledge graph (step S16). This makes it possible to grasp the relationships between individual pieces of information and the overall picture from the connections between pieces of information.
[0052] (6) In this embodiment, the control unit 21 of the support device 20 performs AI input processing (step S17) and AI output acquisition processing (step S18), including knowledge. As a result, it is possible to generate warning information that prompts attention based on the knowledge obtained by the knowledge graph.
[0053] (7) In this embodiment, the control unit 21 of the support device 20 performs user type identification processing (step S19) and warning information output processing (step S21). This allows the user to be assigned to a person who can address the concerns based on the warning content. (8) In this embodiment, the control unit 21 of the support device 20 performs weighting processing (step S20). This makes it possible to set the priority of the warning information.
[0054] This embodiment can be implemented with the following modifications. This embodiment and the following modifications can be combined with each other to the extent that they do not contradict each other technically.
[0055] • In the above embodiment, a user device 10 and a support device 20 were used, but the hardware configuration is not limited to these. In the above embodiment, the plan summary is used as the document to be examined. However, the document to be examined is not limited to the plan summary, as long as it is a document that affects subsequent processes.
[0056] In the above embodiment, the case information storage unit 22 is used as the knowledge database. The knowledge required for construction is not limited to past cases, but may also include the knowledge, experience, know-how of each designer, and laws and regulations applicable to the building.
[0057] In the above embodiment, the control unit 21 of the support device 20 performs character recognition processing (step S12). Here, the address of the building is identified in the document under consideration. Identifying the building address is not limited to the recognition process, but may also be obtained from, for example, the user device 10. In this case, the upload screen includes an address input field for entering the address where the building will be constructed.
[0058] In the above embodiment, the control unit 21 of the support device 20 performs the process of acquiring location information (step S14). The acquisition of location information may be performed after the process of acquiring the graph under consideration (step S15). In this case, the AI input process is performed including the knowledge including location information in addition to the knowledge acquired by the knowledge graph (step S17).
[0059] In the above embodiment, the control unit 21 of the support device 20 performs knowledge acquisition processing using a knowledge graph (step S16). The knowledge acquisition processing is not limited to the use of a knowledge graph. For example, vector search, keyword search, hybrid search, etc., can be used. In vector search, documents are converted into vector representations and the similarity with the query is calculated to search for highly relevant documents. In keyword search, a search is performed based on the match between keywords included in the query and words in the documents. In hybrid search, a search is performed by combining vector search and keyword search to take advantage of the benefits of both.
[0060] In the above embodiment, the control unit 21 of the support device 20 performs knowledge acquisition processing using the knowledge graph (step S16). Here, the case information storage unit 22 is searched. Alternatively, or in addition to this, a search may be performed on a network site.
[0061] In the above embodiment, the control unit 21 of the support device 20 performs the output processing of warning information (step S21). Here, the warning information is output in order of highest weighting. Changing the display format of concerns is not limited to changing the display order. For example, it may also be done by changing the display color.
[0062] In the above embodiment, the control unit 21 of the support device 20 performs weighting processing (step S20). Here, the case information storage unit 22 is used to identify the scope of influence of the design conditions of the concern. Alternatively, based on caution information, higher weighting may be applied depending on the likelihood that other specifications included in the document under consideration will be changed. In this case, the control unit 21 of the support device 20 includes a prompt to the AI system 30 to predict the impact on other specifications and perform weighting. [Explanation of symbols]
[0063] A1...Design support device, 10...User device, 20...Support device, 21...Control unit, 211...Acquisition unit, 212...Recognition unit, 213...Identification unit, 22...Case information storage unit, 23...Regional information storage unit, 24...Consideration target storage unit, 30...AI system.
Claims
1. A design support device comprising a control unit connected to a user device, The control unit, From the aforementioned user device, specifications are obtained from information contained in the documents under consideration regarding the building. Obtain relevant information related to the above specifications, Including the aforementioned specifications and related information, the attention information for the design of the building is predicted, A design support device characterized by classifying the predicted cautionary information according to the design type and outputting it to the user device responsible for the design type.
2. The design support device according to claim 1, characterized in that the control unit uses prediction information generated using past cases for the purpose of the prediction.
3. The design support device according to claim 1, characterized in that the control unit inputs the specifications and related information to the generating AI to generate the caution information.
4. The control unit, The aforementioned cautionary information is classified according to its impact on subsequent processes. The design support device according to claim 1, characterized in that the display format of the warning information is changed based on the classification.
5. The control unit, We obtain location information included in the aforementioned documents under consideration, The design support device according to claim 1, characterized in that it outputs the warning information based on regional information that affects the building, in accordance with the location information.
6. A method for assisting design using a design support device equipped with a control unit connected to a user device, The control unit, From the aforementioned user device, specifications are obtained from information contained in the documents under consideration regarding the building. Obtain relevant information related to the above specifications, Including the aforementioned specifications and related information, the attention information for the design of the building is predicted, A design support method characterized by classifying the predicted cautionary information according to the design type and outputting it to the user device responsible for the design type.
7. A design support program that uses a design support device equipped with a control unit connected to a user device, The control unit, From the aforementioned user device, specifications are obtained from information contained in the documents under consideration regarding the building. Obtain relevant information related to the above specifications, Including the aforementioned specifications and related information, the attention information for the design of the building is predicted, A design support program characterized by classifying the predicted cautionary information according to the design type and functioning as a means to output it to the user device responsible for that design type.