Fire alarm equipment counting device

Abstract

To accurately and quickly count the number of fire alarm systems shown on a blueprint of a building and to utilize the count for subsequent processing. [Solution] A number counting device 10 according to the present invention comprises a blueprint data storage unit 11 that stores drawing data for architectural blueprints, a data reading unit 12 that reads this drawing data in a state that allows for subsequent analysis, a symbol extraction unit 21 that extracts symbols of fire alarm systems displayed in the drawing data read by the data reading unit, and a number counting unit 24 that counts the number of symbols. The symbol extraction unit 21 and the number counting unit 24 are configured to have AI functions (claim 1). Based on the architectural blueprints, symbols of fire alarm systems are extracted via the symbol extraction unit 21 (using AI), and the number of symbols of fire alarm systems is counted via the number counting unit 24 (using AI), enabling highly accurate and rapid counting of the number of fire alarm systems.

Description

[Technical Field]

[0001] This invention relates to a technology for counting the number of fire alarm devices shown on a building's blueprints, and in particular to a high-precision recognition technology using AI. [Background technology]

[0002] Buildings shall be equipped with the required number of fire alarm systems in accordance with the Fire Service Act.

[0003] Previously, the locations of fire alarm systems were indicated on building blueprints, but the number of systems was counted manually.

[0004] Specifically, as shown in Figure 4, a handwritten design drawing 1 was read using a scanner device 2, and the symbol marks of the fire alarm systems were extracted using a customized dedicated application 3. The number of fire alarm systems was then counted manually 4, and the count results were stored in a recording and input unit 6.

[0005] The design drawing data read via the scanner device 2 may be converted into appropriate data such as PDF5 and saved, and then sent to the dedicated application 3, where the symbol mark of the fire alarm equipment may be extracted using the dedicated application 3.

[0006] Conventionally, the technology for counting the number of articles is known from the following Patent Document 1. Ta.

[0007] This is a technology that captures images of a collection of many items from three directions and calculates the number of items from the images taken at different angles. [Prior art documents] [Patent documents]

[0008] [Patent Document 1] Patent Publication No. 2018-167974 DISCLOSURE OF THE INVENTION [Problem to be solved by the invention]

[0009] The problems with the prior art in counting the number of fire alarm devices are the time loss caused by manual work and the inaccuracy of the counting results.

[0010] Even though the counting was done manually, the number of fire alarm systems shown on the building blueprints was enormous, and counting the symbol marks of the fire alarm systems extracted using a dedicated app took a long time if accuracy was to be ensured.

[0011] Even if you take the time to count, there may be cases where the total result is inaccurate due to counting errors, etc.

[0012] Furthermore, there are also cases where the blueprints are misread when scanned using a scanner, such as when the blueprints are unclear. In such cases, even with analysis software, it is not possible to accurately extract the symbol mark of the fire alarm system.

[0013] Furthermore, since blueprints are hand-drawn drawings, it is not possible to count the number of pieces based on images taken from three directions, as in Patent Document 1.

[0014] Therefore, the purpose of this invention is to count the number of fire alarm systems installed as shown on the design drawings more accurately and in a shorter time, so that the information can be used for subsequent processing. [Means for solving the problem]

[0015] In order to achieve the above-mentioned object, the fire alarm system number counting device of the present invention comprises a blueprint data storage unit that stores drawing data of architectural blueprints, a data reading unit that reads the drawing data stored in the blueprint data storage unit in a state that can be subsequently analyzed, a symbol extraction unit that extracts symbol signs of fire alarm systems displayed in the drawing data read via the data reading unit, and a number counting unit that counts the number of symbol signs (particularly detectors) of fire alarm systems extracted by the symbol extraction unit, and the symbol extraction unit and the number counting unit are configured to have AI functions (Claim 1).

[0016] Based on architectural blueprints, the symbol extraction unit (using AI) extracts the symbol signs of the fire alarm equipment, and the number of symbol signs of the fire alarm equipment (e.g., the number of detectors) is counted via the number counting unit (using AI), enabling highly accurate and rapid counting of the number of fire alarm equipment.

[0017] The system may be provided with an estimate calculation unit that calculates an estimate for the fire alarm equipment shown on the architectural design drawing based on the counting results of the quantity counting unit via a command instruction unit that accepts input of the unit price of the parts (Claim 2).

[0018] The estimated cost varies depending on the unit cost of the fire alarm system components, but the estimate calculation unit executes several types of estimates (including the total cost) via the command instruction unit that accepts the input of the unit cost of the components. The fire alarm system includes a detector, a receiver, an auxiliary receiver, an acoustic device (for example, an emergency bell), etc. The estimate calculation unit generates estimates for any combination, such as an estimate for the entire detector, an estimate for the receiver and auxiliary receiver, an estimate for the acoustic device, or an estimate for the total cost.

[0019] The system may be provided with a legality determination unit that determines whether the location or number of fire alarm equipment displayed on the drawing data of the architectural design drawing read by the data reading unit complies with the Fire Service Act, and the legality determination unit may be configured to have AI functions (Claim 3).

[0020] The legality determination unit, which is configured to have AI functions, uses deep learning and other methods to learn the relationship between the Fire Service Act and fire alarm equipment in a large number of architectural blueprints, and quickly and accurately determines whether the fire alarm equipment in the architectural blueprints read by the data reading unit complies with the Fire Service Act. [Effects of the Invention]

[0021] According to the fire alarm system counting device of the present invention, the number of fire alarm systems installed shown on architectural blueprints can be counted more accurately and in a shorter time. [Brief explanation of the drawings]

[0022] [Figure 1] 1 is a block diagram showing a number counting device for a fire alarm system according to a first embodiment; [Figure 2] FIG. 10 is a block diagram showing a number counting device for a fire alarm system according to a second embodiment. [Figure 3] FIG. 10 is a block diagram showing a number counting device for fire alarm systems according to a third embodiment. [Figure 4] FIG. 1 is a block diagram illustrating a conventionally known device for counting the number of articles. DETAILED DESCRIPTION OF THE INVENTION

[0023] FIG. 1 shows an example of a number counting device 10 for a fire alarm system according to a first embodiment.

[0024] This fire alarm system number counting device 10 can be configured by analyzing and interpreting the blueprints of a building via a PC terminal or the like.

[0025] First, the overall configuration will be described.

[0026] Reference numeral 11 denotes a design drawing data storage unit that stores drawing data of architectural design drawings, and 12 denotes a data reading unit that reads the drawing data stored in the design drawing data storage unit 11 in a state that allows subsequent analysis.

[0027] Reference numeral 20 denotes a component (processing unit) equipped with AI (Artificial Intelligence) functions, 21 denotes a symbol extraction unit that extracts symbol signs of fire alarm equipment (sensors, receivers, sound devices, etc.) displayed in the drawing data read via the data reading unit 12, and 24 denotes a number counting unit that counts the number of symbol signs of fire alarm equipment extracted by the symbol extraction unit 21.

[0028] Reference numeral 14 denotes a quantity recording unit that stores the calculation results of the quantity counting unit 24, and reference numeral 15 denotes a result display unit that sends the number of fire alarm devices stored in the quantity recording unit 14 or the number of fire alarm devices counted by the quantity counting unit 24 to a subsequent stage.

[0029] Reference numeral 16 denotes a peripheral device section such as a PC terminal device, 17 denotes a command instruction section (keyboard device / mouse device / touch panel etc.), and 18 denotes a monitor device.

[0030] The configuration of each part will be explained below.

[0031] The design drawing data storage unit 11 that stores drawing data of architectural design drawings saves image data in an appropriate file format.

[0032] The design drawing data read via a scanner device (not shown) may be saved in an appropriate file format (e.g., JPG, GIF, etc.), or design drawing data created in an appropriate file format (e.g., PDF, etc.) may be imported and saved.

[0033] When importing blueprint data, for example, you can read it via the Internet or via external storage media such as a USB memory stick or SD card. If the blueprints were created in-house, you can also import the blueprint data using the in-house LAN (whether wired or wireless).

[0034] The design data is preferably a two-dimensional figure (for example, a floor plan) showing the location of the fire alarm system. In the case of a multi-story building, it is desirable to use a two-dimensional figure (for example, a floor plan) for each floor.

[0035] Image data may be stored temporarily if it is only used for post-processing. However, since blueprint data is often used repeatedly, it is desirable for the blueprint data storage unit 11 to use a recording medium (such as an HDD or SSD) that can store a large amount of blueprint data for a long period of time.

[0036] The data reading unit 12 reads the drawing data stored in the design drawing data storage unit 11 in a state that allows subsequent analysis. For example, when the drawing data stored in the design drawing data storage unit 11 is captured via a scanner device (not shown), the data reading unit 12 converts the file format into a storage state that allows analysis, etc. in the subsequent processing unit 20 that has an AI function.

[0037] If the processing unit 20 requires a dedicated / proprietary file format, the file format is converted to that format. Also, if subsequent processing is possible in a widely used file format such as PDF, BMP, JPEG, PNG, or TIFF, there is no need to convert the file format. In that case, for example, the file format of the drawing data stored in the design drawing data storage unit 11 may be used as is.

[0038] The data reading unit 12 does not necessarily need to be configured to have an AI function, as long as the processing unit 20 at the subsequent stage is capable of performing analysis using the AI ​​function.

[0039] The symbol extraction unit 21, which extracts the symbol sign of the fire alarm system displayed in the drawing data read by the data reading unit 12, is part of the processing unit 20 equipped with AI functions. There are two main types of marks on fire alarm system detectors: the "NS mark" and the "certification mark." The NS mark is a mark issued by the Japan Fire Equipment Inspection Association to guarantee quality, and currently, the "certification mark," which meets stricter standards, is tending to be used.

[0040] When extracting detectors (heat detectors, smoke detectors, flame detectors, etc.) of fire alarm systems, the symbol extraction unit 21 only needs to extract symbols. Since the number of detectors is not counted manually, there is no need for processing such as highlighting the extracted symbol of the detector. Since there are only a few receivers, auxiliary receivers, and sound equipment (emergency alarm devices), there are cases where the number of receivers, auxiliary receivers, sound equipment (emergency alarm devices), etc. is counted manually. In this case, processing such as highlighting the extracted symbol (receiver, auxiliary receiver, sound equipment) may be performed.

[0041] The symbol extraction unit 21 may have a function of using an AI function (particularly a recognition AI function) to send the extracted symbol sign of the fire alarm system to a subsequent stage as data.

[0042] The AI ​​function of the processing unit 20 may be customized or developed, but paid or free AI tools (AI products) can also be used. Examples include LINE WORKS OCR (LINE WORKS Corporation), MeerGuard Enterprise (OkojoAI Inc.), Asteria AIoT Suite (Asteria Corporation), and StellaController (Phoxter Inc.). Any AI tool that can be applied to image recognition / image processing can be used.

[0043] The function to extract symbols from fire alarm systems is a relatively easy process compared to generation processes, etc. Currently, many well-known / publicly available free AI tools can handle this process with sufficient accuracy.

[0044] The number counting unit 24, which counts the number of symbols of the fire alarm systems extracted by the symbol extraction unit 21, is a part of the processing unit 20 that has an AI function.

[0045] When counting, for example, the number of detectors among the fire alarm systems extracted by the symbol extraction unit 21, the number counting unit 24 can perform counting by adding information based on the display position of the detector symbol. The same applies to receivers, sub-receivers, and audio systems. The number counting unit 24 may count only the number of detectors, or may count all the fire alarm systems, such as detectors, receivers, sub-receivers, audio systems, and fire doors.

[0046] This number counting unit 24 may have a function of using an AI function (particularly a recognition AI function) to send the count result data to a subsequent stage.

[0047] Counting items is a relatively easy process when using AI tools. Currently, there are many well-known / publicly available free AI tools that can perform item counting with a sufficiently high degree of accuracy.

[0048] The number recording unit 14 is a memory that stores the processing results of the number counting unit 24. For example, in the case of a building, the number of fire alarm systems calculated for each floor of the building's blueprint, such as the first, second, third, and nth floors, is stored. Preferably, the building name and floor number can be displayed to facilitate subsequent searches. The building name, floor number, and number of fire alarm system symbols may also be stored in a table format.

[0049] The result display unit 15 outputs the number of fire alarm systems stored in the number recording unit 14 in an appropriate format to the downstream monitor device 18. The display format of the monitor device 18 can be controlled, for example, via the command instructing unit 17 (keyboard device, mouse device, touch panel, etc.). Arrows indicating command instructions from the command instructing unit 17 to the number recording unit 14 or the result display unit 15 are omitted.

[0050] With this configuration, the number of fire alarm systems shown on the design drawings, for example, the number of detectors, can be counted accurately and quickly based on the symbol extraction unit 21 and the number counting unit 24. The number of symbols of fire alarm systems other than detectors may also be counted together with the number of detectors.

[0051] There are various types of fire alarm systems, including smoke detection, heat detection, flame detection, etc. Even if different symbols are used to indicate the type of detector on the blueprint, the symbol extraction unit 21 equipped with AI functions can extract the symbol of the fire alarm system with high accuracy.

[0052] The number counting unit 24 may count the total number of symbols of the extracted fire alarm equipment, or if the symbol extraction unit 21 performs extraction processing by detector type, it may count each symbol by detector type individually.

[0053] A fire alarm system is a complex piece of firefighting equipment consisting of detectors, receivers, sound devices, etc. However, since the receivers and sound devices can be controlled in an integrated manner, the number of receivers and sound devices may be less than the number of detectors.

[0054] FIG. 2 shows a second embodiment using the fire alarm system number counting device 10 according to FIG.

[0055] This second embodiment includes an estimate calculation unit 30 that calculates an estimate of the total price of the fire alarm equipment displayed on the architectural design drawing via a command instruction unit 17 (keyboard device / mouse device, etc.) that accepts input of the unit price of parts based on the counting results of the quantity counting unit 24. In addition to the estimate for the detector, estimates for other fire alarm equipment (receivers, sound devices, etc.) can also be calculated individually and / or in total.

[0056] The estimate calculation unit 30 can calculate an estimate for parts of the fire alarm system. The cost of installation of the fire alarm system can also be calculated by instructing the unit price of installation via the command instruction unit 17.

[0057] FIG. 3 shows a third embodiment using the fire alarm system number counting device 10 according to FIG.

[0058] This third embodiment includes a legality determination unit 42 that determines whether or not the fire alarm systems shown on the architectural design drawings stored in the design drawing data storage unit 11 comply with the Fire Service Act by analyzing the counting results of the number counting unit 24 and information collected via the Internet 40. The legality determination unit 42 has an AI function.

[0059] The Fire Service Act contains provisions regarding fire alarm systems, but some of the specific installation provisions cannot be read from architectural blueprints. This mainly concerns the installation location, which must be determined by the contractor. For example, when installing a detector on the ceiling, it must be spaced an appropriate distance from lighting fixtures so that the lighting fixtures do not interfere with the detection of heat or smoke, or when installing a detector on an interior wall, it must not be too far from the ceiling. These installation location provisions are subject to on-site judgment, making it difficult to determine compliance based on architectural blueprints.

[0060] The legality determination unit 42 in Figure 3 can make determinations based on architectural blueprints, particularly the number of fire alarm systems and their mandatory locations (whether or not there are detectors in rooms, toilets, corridors, kitchens, etc.).

[0061] The Fire Service Act stipulates the number of detectors that must be installed depending on the area of ​​the relevant location, and also specifies the locations where detectors must be installed. These are shown on architectural blueprints, so compliance can be determined. The low-voltage lines that supply power to the detectors must be wired in series, but if they are mistakenly wired in parallel, the detectors will not be able to detect fires, which is obviously a violation of the Fire Service Act.

[0062] The legality determination unit 42, which determines the suitability of the fire alarm equipment shown on the architectural blueprints, is equipped with a deep learning unit 42-1, which accumulates and learns information about the fire alarm equipment shown on the architectural blueprints via the Internet 40, enabling highly accurate determination.

[0063] The deep learning unit 42-1 preferably includes, for example, a law learning unit 42-3 that learns the Fire Service Act through text, and a design drawing data learning unit 42-4 that learns architectural design drawings that are deemed compliant and / or non-compliant.

[0064] Some of the architectural plans that have been determined to be compliant and / or non-compliant include textual descriptions of precedents, such as legal precedents. The textual portion can also be studied via, for example, the legal study unit 42-3.

[0065] The deep learning unit 42-1 deepens learning based on data collected, for example, via the law learning unit 42-3 and the blueprint data learning unit 42-4, and determines whether the fire alarm systems shown on the architectural blueprints stored in the blueprint data storage unit 11 comply with the Fire Service Act. At this time, the unit 42-1 also determines compliance per floor area based on the number of fire alarm systems counted by the number counting unit 24. The determination result is displayed, for example, on the monitor device 18.

[0066] The devices shown in Figures 1 to 3 can use Chat GPT (conversational AI) as the AI ​​function part. Because Chat GPT is also capable of deep learning, it has potential for use as a means of quickly and easily obtaining answers.

[0067] However, Chat GPT tends to use incorrect information and there is a risk of information leakage, so it is not recommended for use in business. [Explanation of symbols]

[0068] 10. (Fire alarm equipment) Counting device 11 Blueprint data storage section 12 Data reading unit 14 Quantity Recording Section 15 Result display area 16 Peripheral Equipment Section 17 Command instruction unit (keyboard device / mouse device / touch panel, etc.) 18 Monitor Device 20 Components with AI functions (processing section) 21 Symbol Extraction Unit 24 Counting section 30 Estimate Calculation Department 40 Internet 42 Legality Determination Department 42-1 Deep Learning Department 42-3 Legal Studies Department 42-4 Blueprint Data Learning Section

Claims

1. a design drawing data storage unit for storing drawing data of architectural design drawings; a data reading unit that reads the drawing data stored in the design drawing data storage unit in a state that allows subsequent analysis; a symbol extraction unit that extracts a symbol sign of the fire alarm system displayed in the drawing data read by the data reading unit; a number counting unit that counts the number of symbols of the fire alarm systems extracted by the symbol extraction unit, The symbol extraction unit and the number counting unit A fire alarm equipment counting device characterized by being configured with AI functionality.

2. Based on the counting result of the counting unit, Through the command instruction section that accepts input of the part unit price, 2. The fire alarm system number counting device according to claim 1, further comprising an estimate calculation unit for calculating an estimate for the fire alarm systems displayed on the architectural design drawings.

3. Based on the drawing data of the architectural design drawing read by the data reading unit, a legality determination unit that determines whether the positions or numbers of fire alarm systems displayed in the drawing data comply with the Fire Service Act; The legality determination department shall:

3. The fire alarm system number counting device according to claim 1, further comprising an AI function.

Images