Database creation system and deterioration diagnosis system

Abstract

We provide a database creation system that can flexibly handle data files created in various formats. [Solution] The present invention is a database creation system (1) characterized by comprising: an acquisition unit (4) that acquires an inspection data file (26) of a building or building equipment; an object identification unit (6) that inputs the information of the inspection data file into a large-scale language processing model (28) and identifies an object to be diagnosed for deterioration; a generation unit (8) that generates a prompt corresponding to the identified object; a data extraction unit (10) that inputs the generated prompt into the large-scale language processing model to extract inspection result information from the inspection data file and store it in a database (30); an inspection item matching unit (12) that matches the inspection items essential for deterioration diagnosis with the extracted inspection result information; and an information deficiency notification unit (14) that notifies the system when essential inspection result information is missing.

Description

Technical Field

[0001] The present invention relates to a database creation system and a deterioration diagnosis system, and particularly to a database creation system for creating a database storing data used for deterioration diagnosis of a building or building equipment, and a deterioration diagnosis system for performing deterioration diagnosis of a building or building equipment.

Background Art

[0002] Japanese Patent Application Laid-Open No. 2017-134502 (Patent Document 1) describes a document management device, a document management method, and a program. This document management device includes a storage unit, a format determination unit, an attribute assignment unit, and a storage location management unit, and the storage unit stores a format correspondence table. Further, the format determination unit compares the description location of information on a document with the description location indicated by the format correspondence table, and stores the correspondence between the attribute associated with the description location indicated by the format correspondence table determined to satisfy the condition and the extracted information from the description location of the document data. The storage location management unit determines a storage location based on the attributed information obtained by adding information representing an attribute corresponding to the extracted information to the information described in the extracted information, and stores the document data.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, the document management device described in Patent Document 1 requires that a format mapping table be stored in the storage unit. Therefore, in order to handle documents in various formats containing the necessary information, it is necessary to prepare a format mapping table for all types of documents to be managed. As a result, the document management device described in Patent Document 1 requires updating the format mapping table each time it tries to handle a new type of document.

[0005] In particular, when applying a document management system to documents containing inspection results for buildings or building equipment, and attempting to create a database storing data used for deterioration diagnosis, there is a problem in that the document formats to be handled are not standardized, making the creation of format-matching tables extremely complicated. That is, although buildings and building equipment are inspected and recorded on a daily basis, the format of the documents describing individual inspection results is not standardized, making it difficult to utilize the results of daily inspections for deterioration diagnosis. In addition, a single document containing inspection results for buildings or building equipment often does not include all the inspection items necessary for deterioration diagnosis, and in some cases, the creation of format-matching tables itself is difficult.

[0006] Therefore, the present invention aims to provide a database creation system and a deterioration diagnosis system that can flexibly handle data files created in various formats and reliably extract inspection result information related to necessary inspection items. [Means for solving the problem]

[0007] To solve the above-mentioned problems, the present invention provides a database creation system for creating a database that stores data used for deterioration diagnosis of buildings or building equipment, comprising: an inspection data file acquisition unit that acquires an inspection data file containing information on the inspection results of buildings or building equipment; an object identification unit that inputs the information contained in the inspection data file acquired by the inspection data file acquisition unit into a large-scale language processing model and identifies an object to be diagnosed for deterioration; a prompt generation unit that generates a prompt corresponding to the object identified by the object identification unit; a data extraction unit that inputs the prompt generated by the prompt generation unit into a large-scale language processing model to extract information on inspection results related to inspection items necessary for deterioration diagnosis of the object from the inspection data file and stores it in the database; an inspection item matching unit that matches the inspection items essential for deterioration diagnosis of the object with the information on inspection results extracted by the data extraction unit; and an information shortage notification unit that notifies that there is insufficient information necessary for deterioration diagnosis when information on inspection results related to inspection items essential for deterioration diagnosis of the object is missing.

[0008] According to the present invention configured in this manner, a large-scale language processing model is used to extract information on inspection results related to inspection items necessary for diagnosing deterioration of an object from an inspection data file and store it in a database. This makes it possible to easily extract necessary data from inspection data files created in various formats and store it in the database. Furthermore, according to the present invention configured as described above, the object identification unit identifies the object to be diagnosed for deterioration, and a prompt corresponding to the identified object is generated and input to the large-scale language processing model. This ensures that the necessary information can be reliably extracted from the inspection data file according to the object. Moreover, according to the present invention configured as described above, the inspection item matching unit compares the inspection items essential for diagnosing deterioration of the object with the inspection result information extracted by the data extraction unit. This ensures that it is possible to reliably identify when inspection result information is missing for inspection items essential for deterioration diagnosis, and to reliably store the necessary information in the database.

[0009] Furthermore, the present invention is a deterioration diagnosis system for performing deterioration diagnosis of buildings or building equipment, characterized by comprising: an inspection data file acquisition unit that acquires an inspection data file containing information on the inspection results of buildings or building equipment; an object identification unit that inputs the information contained in the inspection data file acquired by the inspection data file acquisition unit into a large-scale language processing model and identifies an object to be diagnosed for deterioration; a prompt generation unit that generates a prompt corresponding to the object identified by the object identification unit; a data extraction unit that inputs the prompt generated by the prompt generation unit into a large-scale language processing model to extract information on inspection results related to inspection items necessary for the deterioration diagnosis of the object from the inspection data file and stores it in a database; a judgment value generation unit that generates judgment values ​​to be used for deterioration diagnosis of buildings or building equipment based on the inspection result information extracted by the data extraction unit; a deterioration diagnosis unit that performs deterioration diagnosis of buildings or building equipment based on the judgment values ​​generated by the judgment value generation unit; and a diagnosis result output unit that outputs the results of deterioration diagnosis by the deterioration diagnosis unit.

[0010] According to the present invention configured in this way, a large-scale language processing model is used to extract information on inspection results related to inspection items necessary for diagnosing deterioration of an object from an inspection data file and store it in a database. This makes it possible to easily extract necessary data from inspection data files created in various formats and store it in the database. Furthermore, in the present invention configured as described above, the data extraction unit extracts information on inspection results from the inspection data file, and the judgment value generation unit generates judgment values ​​based on the extracted inspection result information. Therefore, useful information can be generated as judgment values ​​for deterioration diagnosis from an inspection data file 26 containing information on inspection results performed at various times, enabling accurate deterioration diagnosis.

[0011] In the deterioration diagnosis system of the present invention, preferably, the system further includes a diagnosis accuracy calculation unit that calculates the diagnostic accuracy of the deterioration diagnosis of a building or building equipment based on the inspection result information extracted by the data extraction unit.

[0012] With the present invention configured in this way, the diagnostic accuracy calculation unit calculates the diagnostic accuracy of the deterioration diagnosis of a building or building equipment, so that the user can easily understand the accuracy of the deterioration diagnosis and effectively utilize the results of the deterioration diagnosis.

[0013] In the deterioration diagnosis system of the present invention, preferably, the data extraction unit is configured to extract information related to repairs to a building or building equipment, and information on the date and / or time the repairs were carried out, and the judgment value generation unit generates judgment values ​​with respect to diagnostic items related to repairs, based on inspection results from the date or time the repairs were carried out on or after.

[0014] Generally, buildings and building equipment undergo repairs, and the degree of deterioration changes significantly as a result of these repairs. According to the present invention configured as described above, the judgment value generation unit generates judgment values ​​with respect to diagnostic items related to repairs, based on inspection results from the date or time the repairs were performed onward. Therefore, deterioration diagnoses that accurately reflect the status of repairs can be performed.

[0015] In the deterioration diagnosis system of the present invention, preferably, the inspection result information extracted by the data extraction unit includes information on the date and / or time the inspection was performed.

[0016] With the present invention configured in this way, the inspection result information extracted by the data extraction unit includes information on the date and / or time the inspection was performed, so that the status of inspections can be grasped in chronological order.

[0017] The deterioration diagnosis system of the present invention preferably further includes a report generation unit that generates a report of the inspection results of a building or building equipment at predetermined intervals using the inspection result information extracted by the data extraction unit and stored in the database.

[0018] According to the present invention configured as described above, since the report creation unit creates a report on the inspection results of a building or building equipment at regular intervals, it is possible to confirm, based on this report, that the information in the inspection data file obtained through daily inspections is stored in the database without any deficiencies. Further, since the inspection results are created in the form of a report, the created report can be directly utilized as inspection documents.

Effects of the Invention

[0019] According to the database creation system and the deterioration diagnosis system of the present invention, it is possible to flexibly handle data files created in various formats and reliably extract information on inspection results regarding necessary inspection items.

Brief Description of the Drawings

[0020] [Figure 1] It is a block diagram showing the entire database creation system and deterioration diagnosis system according to an embodiment of the present invention. [Figure 2] It is a flowchart for explaining the processing procedure in the database creation system and the deterioration diagnosis system according to an embodiment of the present invention. [Figure 3] In an embodiment of the present invention, it is a diagram showing an example of an inspection data file input to the deterioration diagnosis system and data generated based thereon. [Figure 4] It is a flowchart of a subroutine called from step S6 of the flowchart shown in FIG. 2. [Figure 5] It is a flowchart of a subroutine called from step S8 of the flowchart shown in FIG. 2. [Figure 6] In an embodiment of the present invention, it is a diagram showing an example of a determination value generated from inspection results stored in a database. [Figure 7] In an embodiment of the present invention, it is a diagram showing an example of the display of the result of deterioration diagnosis. [Figure 8]In an embodiment of the present invention, it is a diagram showing an example of the display of the deterioration diagnosis result. [Figure 9] In an embodiment of the present invention, it is a diagram showing a modification example of the data format stored in the database.

Mode for Carrying Out the Invention

[0021] Next, a database creation system and a deterioration diagnosis system according to an embodiment of the present invention will be described with reference to the accompanying drawings. The database creation system of the present embodiment is a system that creates a database storing data used for the deterioration diagnosis of a building or building equipment. Further, the deterioration diagnosis system of the present embodiment is a system that creates a database in the same procedure as the database creation system of the present embodiment and performs the deterioration diagnosis of a building or building equipment using the stored data.

[0022] FIG. 1 is a block diagram showing the entire database creation system and deterioration diagnosis system according to an embodiment of the present invention. FIG. 2 is a flowchart for explaining the processing procedure in the database creation system and deterioration diagnosis system according to an embodiment of the present invention.

[0023] As shown in FIG. 1, a deterioration diagnosis system 1 according to an embodiment of the present invention includes an inspection data file acquisition unit 4, an object identification unit 6, a prompt generation unit 8, a data extraction unit 10, an inspection item collation unit 12, and an information shortage notification unit 14. Further, the deterioration diagnosis system 1 includes a determination value generation unit 16, a deterioration diagnosis unit 18, a diagnosis accuracy calculation unit 20, a diagnosis result output unit 22, and a report creation unit 24. Further, the inspection data file acquisition unit 4, the object identification unit 6, the prompt generation unit 8, the data extraction unit 10, the inspection item collation unit 12, and the information shortage notification unit 14, which form a part of the deterioration diagnosis system 1, constitute a database creation system 2 according to an embodiment of the present invention. <​​The database creation system 2 of this embodiment is configured to acquire an inspection data file 26, input the information contained in the acquired inspection data file into a large-scale language processing model 28, and create a database 30 that stores data used for deterioration diagnosis of buildings or building equipment. Furthermore, the deterioration diagnosis system 1 of this embodiment is configured to generate judgment values ​​for deterioration diagnosis items of buildings or building equipment based on the inspection result information stored in the database 30, perform the deterioration diagnosis, and output the deterioration diagnosis results to a display 32 or a printer 34. Specifically, each functional unit of the deterioration diagnosis system 1 and the database creation system 2 is realized by a microprocessor, memory, interface circuit, and software that operates them (these are not shown).

[0025] Next, with reference to Figure 2, the information processing procedure in the deterioration diagnosis system 1 of this embodiment will be described. First, in step S1 of Figure 2, the inspection data file acquisition unit 4 acquires the inspection data file 26. The inspection data file 26 is an electronic file containing information on the inspection results for buildings and building equipment. For example, the inspection data file 26 can be an electronic file containing information on maintenance materials created during routine inspections of buildings and building equipment that are performed periodically. This inspection data file 26 can be acquired from an external storage device or the cloud (not shown) connected to the deterioration diagnosis system 1.

[0026] Furthermore, if the maintenance documents are in paper form, as a preprocessing step, the documents are scanned (not shown), and the scanned image data is subjected to optical character recognition (OCR) processing to convert it into text data and generate the inspection data file 26. It is desirable that the text data converted from the paper documents be in a data structure and format that can retain information such as the positional relationship of characters on the paper and the format of the list, such as in CSV, ML, or JSON format.

[0027] Next, in step S2, the information contained in the inspection data file 26 acquired by the inspection data file acquisition unit 4 is input into the large-scale language processing model 28 by the object identification unit 6. Then, the large-scale language processing model 28 identifies the objects that should be subjected to deterioration diagnosis from the information contained in the inspection data file 26.

[0028] Figure 3 shows an example of an inspection data file input to the deterioration diagnosis system 1, and the data generated based on it. As shown in column A of Figure 3, when electronic data containing information on the inspection results of building equipment is acquired as an inspection data file 26, this information is input to the large-scale language processing model 28 by the object identification unit 6. The large-scale language processing model 28 analyzes the input inspection data file 26 and organizes and outputs the information contained in the inspection data file 26, as shown in column B of Figure 3. In the example shown in column B of Figure 3, the information on equipment A and equipment B contained in the inspection data file 26 is organized as pump information. Based on the output from the large-scale language processing model 28, the object identification unit 6 identifies the object as a pump.

[0029] Next, in step S3 of Figure 2, the prompt generation unit 8 generates a prompt corresponding to the object identified by the object identification unit 6. That is, the prompt generation unit 8 has a table of inspection items necessary for deterioration diagnosis for each object whose deterioration should be diagnosed, and generates prompts so that these inspection items are extracted by the large-scale language processing model 28. For example, the prompt generation unit 8 stores a template prompt for obtaining the deterioration status of each inspection item necessary for deterioration diagnosis for each type of object to be diagnosed. Then, the prompt template may be selected according to the object identified by the object identification unit 6.

[0030] In the example shown in Figure 3, as indicated in column C, Please categorize the following items into one of the following four categories: 0 for no problem, 1 for problem, unknown for unknown, and repaired if repair work has been carried out. Exterior rust and cracks, Unusual noise, Water leakage.” The prompt 8 generates the following prompt.

[0031] Next, in step S4, the prompt generated in step S3 is input to the large-scale language processing model 28 by the data extraction unit 10, and information on inspection results related to the inspection items necessary for diagnosing the deterioration of the object is extracted from the inspection data file 26. The extracted inspection result information is stored in the database 30. In the example shown in Figure 3, as shown in column D, for "Equipment 1" of the pump, the inspection result information extracted is "0 (none)" for external rust and cracks, "1 (present)" for abnormal noise, and "unknown" for water leakage ("Equipment 2" is not shown). The extracted inspection result information may also include information on the date and / or time the inspection was performed.

[0032] In the example above, the large-scale language processing model 28 assigns the description of "✓" for appearance in the inspection data file 26 shown in column A of Figure 3 to "0 (none)" and the description of "△" for abnormal sound to "1 (present)". Alternatively, if the inspection data file 26 contains the sentence "Rust has occurred", the large-scale language processing model 28 understands the content and assigns the rust / cracks in appearance to "1 (present)". Furthermore, if there is no evaluation or mention of rust / cracks in appearance in the inspection data file 26, it is assigned to "unknown".

[0033] Furthermore, in step S5, the large-scale language processing model 28 extracts information about repairs to the object, as well as the date and / or time the repairs were performed, from the inspection data file 26. The extracted inspection result information is stored in the database 30. For example, if the daily report for April 4, 2023, obtained as the inspection data file 26, contains the entry "Abnormal noise was detected from equipment A. This was addressed by replacing parts," the large-scale language processing model 28 extracts information indicating that "repairs" were performed on inspection item 2 on the same day and stores it in the database 30. In the example shown in Figure 3, as shown in column E, the results of deterioration diagnoses performed on the object to date, the details of repairs, and the inspection result information extracted from the inspection data file 26 are stored in the database 30 along with the date and / or time.

[0034] Next, in step S6, a matching process for inspection items is performed on the inspection results that have been extracted and stored in the database 30. Specifically, in step S6, the inspection item matching unit 12 executes the flowchart shown in Figure 4 as a subroutine. The inspection item matching unit 12 executes the flowchart shown in Figure 4 to match the inspection items essential for diagnosing the deterioration of the object with the information of the inspection results extracted by the data extraction unit 10.

[0035] First, in step S21 of Figure 4, the inspection item number n is set to 1. Next, in step S22, it is determined whether all the inspection results information for the first inspection item, inspection item 1, performed within the target period is "unknown". If there are any inspection results other than "unknown" performed within the target period, the process in the flowchart proceeds to step S25; if all inspection results are "unknown", it proceeds to step S23. Here, "target period" refers to the period from the latest deterioration diagnosis and most recent repair stored in the database 30 (however, if the time is also recorded, the date and time may be used) after the date of the most recent deterioration diagnosis and repair within the predetermined period (2 years in this embodiment). If neither deterioration diagnosis nor repair was performed within the predetermined period, the predetermined period is considered the "target period". This makes it possible to exclude older inspection results and perform highly reliable diagnoses. (Note that the inspection result information illustrated in Figures 3 and 6 is all assumed to be within the predetermined period.)

[0036] Next, in step S23, it is determined whether the inspection items for which all inspection results were "unknown" are essential inspection items for deterioration diagnosis. Specifically, the inspection item matching unit 12 has an essential item table (not shown) in which essential inspection items are set for each type of object, and it is determined whether the inspection items for which all inspection results were "unknown" are set as essential inspection items. If they are not essential inspection items, the process in the flowchart proceeds to step S25; if they are essential inspection items, it proceeds to step S24.

[0037] In the example shown in column E of Figure 3, inspection item 1, "external rust, cracks, etc.," was diagnosed as "1" (deterioration present) during the previous deterioration diagnosis on April 4, 2022. Repairs were then carried out on October 4, 2023, and all subsequent inspection records show it as "unknown." Similarly, inspection item n was diagnosed as "1" (deterioration present) during the deterioration diagnosis on April 4, 2022, and all subsequent inspection records show it as "unknown." Therefore, for example, if inspection item 1 is set as a mandatory inspection item in the mandatory items table (not shown), the flowchart proceeds to step S24 because all inspection results since the repair on October 4, 2023 are "unknown."

[0038] In step S24, the information deficiency notification unit 14 displays on the display 32 that there is insufficient information for the inspection results stored in the database 30. For example, the information deficiency notification unit 14 displays the message "For pump equipment 1, the inspection results for inspection item 1 have not been entered" on the display 32. Notification of insufficient information can also be made by output from the printer 34, generation of a matching result file (not shown), or output of sound. In the example in column 3E of Figure, since all inspection results for inspection item n are "unknown" after the previous deterioration diagnosis, if inspection item n is a required inspection item, a notification will be made that there is insufficient information for the inspection results.

[0039] Next, in step S25, 1 is added to the inspection item number n. For example, after the processing in steps S22 to S24 is performed for inspection item 1 (n=1), the value of n is updated to 2. Furthermore, in step S26, it is determined whether processing has been completed for all inspection items. If processing has been completed for all inspection items, the process in the flowchart shown in Figure 4 is completed, and the process returns to the flowchart shown in Figure 2. On the other hand, if processing has not been completed for all inspection items, the process in the flowchart returns to step S22, and processing is performed for the next inspection item. For example, after processing for inspection item 1 (n=1) is completed, processing for inspection item 2 (n=2) is performed.

[0040] Once the flowchart shown in Figure 4 is completed, step S7 of the flowchart in Figure 2 is executed. In step S7, information on the inspection results is entered for the inspection items that were notified as lacking in step S24 of Figure 4. Specifically, when the user is notified in step S24 of Figure 4 that there is insufficient information on the inspection results, they search for the inspection data file 26 containing the missing inspection result information and have the inspection data file acquisition unit 4 acquire (read) it. The user can also input the missing inspection result information using a keyboard (not shown) (for example, "0 (no deterioration)", "1 (deterioration present)", "unknown", etc.). If there is no information on the missing inspection result, the user inputs that there is no information using a keyboard (not shown).

[0041] Thus, according to the database creation system 2 (deterioration diagnosis system 1) of this embodiment, if information on inspection results for inspection items essential for deterioration diagnosis of the target object is missing, notification is received, allowing the necessary information to be reliably collected and deterioration diagnosis to be performed. Furthermore, by securing information on inspection results for inspection items essential for deterioration diagnosis of the target object, a minimum level of diagnostic accuracy in deterioration diagnosis can be guaranteed.

[0042] Next, in step S8, the judgment value generation unit 16 generates judgment values ​​for deterioration diagnosis items of the building or building equipment based on the inspection result information extracted by the data extraction unit 10 and stored in the database 30. That is, the judgment value generation unit 16 synthesizes the information of numerous inspection results stored in the database 30 to determine the current state of the object for each inspection item and generates a judgment value. Since the inspection results stored in the database 30 may have been improved by subsequent repairs, inspection results prior to the repairs may not be usable. Also, since the latest inspection result information is divided into multiple documents and stored in different columns of the database 30, it is necessary to synthesize these to determine the current deterioration status.

[0043] Figure 5 is a flowchart of the subroutine called from step S8 of the flowchart shown in Figure 2. Figure 6 is a diagram showing an example of a judgment value generated from the inspection results stored in the database 30.

[0044] First, in step S31 of Figure 5, the inspection item number n is set to 1. Next, in step S32, it is determined whether there is a "1" among the inspection results stored in the database 30 for the first inspection item, inspection item 1, performed within the target period. That is, in step S32, if no repairs have been carried out since the last deterioration diagnosis performed within the predetermined period, the period from that deterioration diagnosis onward is considered the target period, and it is determined whether there is a "1" among the inspection results within the target period. Also, if repairs have been carried out since the last deterioration diagnosis performed within the predetermined period, the period from that repair onward is considered the target period, and it is determined whether there is a "1" among the inspection results within the target period. Furthermore, if neither a deterioration diagnosis nor repairs have been carried out within the predetermined period, the predetermined period is considered the target period, and it is determined whether there is a "1" among the inspection results within the target period. If there is a "1" among the inspection results, the process in the flowchart proceeds to step S34; if there is no "1" among the inspection results, it proceeds to step S33.

[0045] In the example shown in column A of Figure 6, for inspection item 1, the result of the three inspections since the previous deterioration diagnosis on April 4, 2022, has been "1," so the process in the flowchart proceeds to step S34. In step S34, "1" is set as the judgment value for inspection item 1 (diagnosis item 1). Next, in step S37, 1 is added to the inspection item number n. Furthermore, in step S38, it is determined whether the processing for all inspection items has been completed. If the processing for all inspection items has been completed, the process in the flowchart shown in Figure 5 is completed, and the process returns to the flowchart shown in Figure 2. On the other hand, if the processing for all inspection items has not been completed, the process in the flowchart returns to step S32, and the processing for the next inspection item is executed.

[0046] In the example shown in column A of Figure 6, after a judgment value is set for inspection item 1, step S32 is executed for inspection item 2. For inspection item 2, "abnormal noise," it is determined that there are no "1"s in the inspection results after the repair was carried out on April 4, 2023. Therefore, the process in the flowchart proceeds to step S33, where it is determined whether or not there are "0"s after the repair. Since there are three "0"s in the inspection results after the repair for inspection item 2, the process in the flowchart proceeds to step S35, and "0" is set as the judgment value for inspection item 2 (diagnostic item 2).

[0047] Similarly, for inspection item 3, "water leakage," since one inspection result has been recorded as "1" since the previous deterioration diagnosis, the judgment value is set to "1." In the example shown in column A of Figure 6, although the most recent inspection result for inspection item 3 is "0," there is data showing an inspection result of "1" since the previous deterioration diagnosis, so the judgment value for inspection item 3 is set to "1." Thus, in this embodiment, the judgment value is set by selecting the worst value from the inspection result information recorded since the previous deterioration diagnosis or the most recent repair as the basis for judgment. However, the judgment value can be set using any method in accordance with the deterioration diagnosis method, such as calculating the average value of the inspection results.

[0048] Furthermore, for inspection item n, since all inspection results are "unknown" after the previous deterioration diagnosis, the process in the flowchart proceeds from step S32 to S33 to S36, and "unknown" is set as the judgment value. As a result, judgment values ​​are generated for each diagnostic item, as shown in column B of Figure 6. Next, once the processing for all inspection items is complete, the process returns to the flowchart shown in Figure 2.

[0049] Once the processing shown in the flowchart in Figure 5 is completed, the process returns to the flowchart in Figure 2, and step S9 is executed. In step S9, the accuracy of the deterioration diagnosis for the target building or building equipment is calculated by the diagnosis accuracy calculation unit 20.

[0050] First, for each building or building equipment undergoing deterioration assessment, multiple diagnostic items are set, and each of these diagnostic items is assigned a weight according to its importance. The accuracy of the deterioration assessment is then calculated by dividing the sum of the weights assigned to the diagnostic items for which the judgment value is "0" or "1" by the sum of the weights assigned to all diagnostic items.

[0051] For example, in the example shown in column B of Figure 6, there are four diagnostic items: 1, 2, 3, and n. If the weights c set for these four diagnostic items are 0.3, 0.1, 0.4, and 0.2 respectively, the accuracy R of the deterioration diagnosis is: The result is calculated as JPEG2026110209000002.jpg9150. In other words, the more diagnostic items there are with an unknown judgment value, and the higher the importance of the diagnostic items with an unknown judgment value (the larger the weight c), the lower the accuracy R of the deterioration diagnosis becomes.

[0052] Next, in step S10, the deterioration diagnosis unit 18 diagnoses the deterioration of the object using the judgment values ​​of each deterioration diagnosis item generated in step S8. Various methods are known for diagnosing the deterioration of buildings or building equipment, and the deterioration of the object can be diagnosed using any deterioration diagnosis method.

[0053] For example, the Building and Equipment Long-life Association (BELCA) has established diagnostic items for each type of building equipment, and sets judgment values ​​of 0 (no problem) or 1 (deterioration, problem present). Deterioration diagnosis can be performed by accumulating these evaluations. Specifically, the deterioration state of a building or building equipment can be diagnosed by multiplying each judgment value by a weight set according to the importance of each diagnostic item and accumulating these values.

[0054] Furthermore, in step S11, the results of the deterioration diagnosis performed by the deterioration diagnosis unit 18 are output by the diagnosis result output unit 22. The diagnosis result output unit 22 also outputs the value of the diagnostic accuracy R calculated in step S9 along with the deterioration diagnosis results. Specifically, in step S11, the diagnosis result output unit 22 displays the results of the deterioration diagnosis of the building or building equipment on the display 32. The deterioration diagnosis results can also be output from the printer 34, generated as a deterioration diagnosis result file (not shown), or output as audio. Next, in step S12, the results of the deterioration diagnosis performed by the deterioration diagnosis unit 18 are stored in the database 30 along with the date the deterioration diagnosis was performed, and the processing of the flowchart shown in Figure 2 is completed.

[0055] Figures 7 and 8 show examples of how deterioration diagnosis results are displayed. In the example shown in Figure 7, the total number of malfunctions occurring in each of the multiple buildings that underwent deterioration assessments is displayed. The number of pieces of equipment that have exceeded their service life is also displayed for each building.

[0056] Furthermore, it is possible to filter the deterioration diagnosis results for multiple buildings or building equipment, narrowing down the information and displaying the deterioration diagnosis results according to the type of deterioration. Figure 8 shows an example of extracting and displaying electrical equipment in Building A that is experiencing abnormal heat generation. This makes it possible to grasp an overview of the deterioration situation and quickly find a list of managed items that are experiencing the same problem. As a result, it becomes easier to formulate a repair plan.

[0057] Furthermore, in this embodiment, the information stored in the database 30 over a certain period can also be displayed as a deterioration report (weekly report, monthly report). That is, the report creation unit 24 uses the inspection result information extracted by the data extraction unit 10 and stored in the database 30 to create a report of the inspection results of the building or building equipment at predetermined intervals. This makes it easy to confirm whether or not the information has been stored in the database 30 without any defects. Also, by displaying it in the form of a report, it can be used as a document to be submitted as is.

[0058] Furthermore, the present invention can be configured to display the source materials of the information stored in the database 30, along with images and links to their storage locations. It can also be configured to immediately display photographs if they are included in the inspection data file 26. By simultaneously displaying the source materials in this way, it is easy to confirm that there are no deficiencies in the storage of the database 30.

[0059] According to the database creation system 2 of the embodiment of the present invention, a large-scale language processing model 28 is used to extract information on inspection results related to inspection items necessary for diagnosing deterioration of an object from the inspection data file 26 and store it in the database 30. Therefore, necessary data can be easily extracted from inspection data files 26 created in various formats and stored in the database 30.

[0060] Furthermore, according to the database creation system 2 of this embodiment, the object identification unit 6 identifies the object to be diagnosed for deterioration, and a prompt corresponding to the identified object is generated and input to the large-scale language processing model 28. This ensures that the necessary information can be reliably extracted from the inspection data file 26 according to the object. Moreover, according to the database creation system 2 of this embodiment, the inspection item matching unit 12 matches the inspection items essential for diagnosing the deterioration of the object with the inspection result information extracted by the data extraction unit 10. This ensures that it is possible to reliably identify when inspection result information is missing for inspection items essential for deterioration diagnosis, and to reliably store the necessary information in the database 30.

[0061] Furthermore, according to the deterioration diagnosis system 1 of the embodiment of the present invention, a large-scale language processing model 28 is used to extract information on inspection results related to inspection items necessary for deterioration diagnosis of the target object from the inspection data file 26 and store it in the database 30. This makes it possible to easily extract necessary data from inspection data files 26 created in various formats and store it in the database 30. In addition, in the deterioration diagnosis system 1 of this embodiment, the data extraction unit 10 extracts information on inspection results from the inspection data file 26, and the judgment value generation unit 16 generates judgment values ​​based on the extracted inspection result information. Therefore, useful information can be generated as judgment values ​​for deterioration diagnosis from the inspection data file 26 which contains information on inspection results performed at various times, enabling accurate deterioration diagnosis.

[0062] Furthermore, according to the deterioration diagnosis system 1 of this embodiment, the diagnostic accuracy calculation unit 20 calculates the diagnostic accuracy of the deterioration diagnosis of the building or building equipment, so that the user can easily understand the accuracy of the deterioration diagnosis and effectively utilize the results of the deterioration diagnosis.

[0063] Furthermore, according to the deterioration diagnosis system 1 of this embodiment, the judgment value generation unit 16 generates judgment values ​​with respect to diagnostic items related to repairs based on inspection results from the date or time the repairs were performed onward, so that deterioration diagnoses that accurately reflect the status of repairs can be performed.

[0064] Furthermore, according to the deterioration diagnosis system 1 of this embodiment, the inspection result information extracted by the data extraction unit 10 includes information on the date and / or time the inspection was performed, so the status of inspections can be grasped in chronological order.

[0065] Furthermore, according to the deterioration diagnosis system 1 of this embodiment, the report creation unit 24 creates a report of the inspection results of the building or building equipment at predetermined intervals. This report allows for confirmation that the information in the inspection data file 26 obtained through daily inspections is stored in the database 30 without any deficiencies. In addition, since the inspection results are created in report format, the created report can be used directly as an inspection report document.

[0066] Although embodiments of the present invention have been described above, various modifications can be made to the embodiments described above. For example, in the embodiments described above, data was stored in the database for each document entered as an inspection data file, but as a modification, as shown in Figure 9, the database can also store data in a format that groups information by device. In addition to the example shown in Figure 9, data can be stored in the database in any format.

[0067] Furthermore, in the above-described embodiment, when extracting the deterioration status from the inspection data file (step S4 in Figure 2), the inspection result information was categorized as "0", "1", or "unknown" and stored in the database. However, the format of the data stored in the database can be appropriately set according to the deterioration diagnosis method to be performed. For example, the inspection result information can also be categorized as "○", "×", "△", or a numerical value of three or more levels and stored in the database.

[0068] Furthermore, if the inspection result information is stored in the database in a format other than "0", "1", or "unknown", the present invention can also be configured so that when calculating the judgment value from that information (step S8 in Figure 2), the judgment value of "0", "1", or "unknown" is generated. For example, if the inspection result information is stored in the database in the format of "○", "×", or "△", "○" can be set to judgment value "0", and "×" and "△" can be set to judgment value "1". Also, for example, if the inspection result information is stored in the database on a scale of 1 to 10, "7" to "10" can be set to judgment value "0", and "1" to "6" can be set to judgment value "1".

[0069] Furthermore, in the embodiments described above, the judgment value used for deterioration diagnosis was set to "0", "1", or "unknown". However, the judgment value generated by the judgment value generation unit can be set appropriately according to the deterioration diagnosis method used. [Explanation of Symbols]

[0070] 1. Deterioration Diagnosis System 2. Database Creation System 4. Inspection data file acquisition unit 6. Object Identification Section 8. Prompt generation unit 10 Data Extraction Unit 12. Inspection Item Verification Unit 14 Insufficient Information Notification Department 16. Judgment Value Generation Unit 18. Deterioration Diagnosis Department 20 Diagnostic accuracy calculation unit 22 Diagnostic Result Output Unit 24 Report Preparation Department 26 Inspection data file 28 Large-scale language processing models 30 databases 32 displays 34 Printers

Claims

1. A database creation system that creates a database containing data used for diagnosing deterioration of buildings or building equipment, An inspection data file acquisition unit acquires an inspection data file containing information on the inspection results of a building or building equipment, The inspection data file acquired by this inspection data file acquisition unit inputs the information contained in the inspection data file into a large-scale language processing model, and the object identification unit identifies the object to be subjected to deterioration diagnosis. A prompt generation unit generates a prompt corresponding to the object identified by this object identification unit, The prompts generated by this prompt generation unit are input to a large-scale language processing model, and the data extraction unit extracts information on inspection results related to inspection items necessary for diagnosing deterioration of the object from the inspection data file and stores it in a database. An inspection item matching unit compares the inspection items essential for diagnosing deterioration of the object with the inspection result information extracted by the data extraction unit mentioned above. When information regarding inspection results for inspection items essential for diagnosing deterioration of an object is missing, the system includes an information deficiency notification unit that notifies the system that there is insufficient information necessary for the deterioration diagnosis. A database creation system characterized by having the following features.

2. A deterioration diagnosis system for diagnosing deterioration of buildings or building equipment, An inspection data file acquisition unit acquires an inspection data file containing information on the inspection results of a building or building equipment, The inspection data file acquired by this inspection data file acquisition unit inputs the information contained in the inspection data file into a large-scale language processing model, and the object identification unit identifies the object to be subjected to deterioration diagnosis. A prompt generation unit generates a prompt corresponding to the object identified by this object identification unit, The prompts generated by this prompt generation unit are input to a large-scale language processing model, and the data extraction unit extracts information on inspection results related to inspection items necessary for diagnosing deterioration of the object from the inspection data file and stores it in a database. Based on the inspection results information extracted by this data extraction unit, a judgment value generation unit generates judgment values ​​used for deterioration diagnosis of buildings or building equipment, Based on the judgment values ​​generated by this judgment value generation unit, a deterioration diagnosis unit performs a deterioration diagnosis of the building or building equipment. This includes a diagnostic result output unit that outputs the results of the deterioration diagnosis performed by the deterioration diagnosis unit, A deterioration diagnosis system characterized by having the following features.

3. Furthermore, the deterioration diagnosis system according to claim 2 further comprises a diagnosis accuracy calculation unit that calculates the diagnostic accuracy of deterioration diagnosis of a building or building equipment based on the inspection result information extracted by the data extraction unit.

4. The deterioration diagnosis system according to claim 2, wherein the data extraction unit is configured to extract information relating to repairs to a building or building equipment, and information relating to the date and / or time the repairs were carried out, and the judgment value generation unit generates a judgment value with respect to the diagnostic items related to the repairs, based on the inspection results information from the date or time the repairs were carried out on or after.

5. The deterioration diagnosis system according to claim 2, wherein the inspection result information extracted by the data extraction unit includes information on the date and / or time the inspection was performed.

6. Furthermore, the deterioration diagnosis system according to claim 2, further comprising a report creation unit that uses the inspection result information extracted by the data extraction unit and stored in the database to create an inspection result report of a building or building equipment at predetermined intervals.

Images