system

Abstract

We provide the system. [Solution] A safety confirmation system for on-site work, A method for users to upload photos taken on-site to a server via their device, A means for sending images received by a server to a generative artificial intelligence for analysis, A means for generative artificial intelligence to identify potential dangers and points of caution within an image and generate a report, A means for the server to send the generated report to the terminal and for the user to view it, A means for users to incorporate safety check items based on reports into the pre-work toolbox meeting, A system that includes this.

Description

Technical Field

[0001] The technology of the present disclosure relates to a system.

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, and includes steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance that responds to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Ensuring the safety of on-site work is extremely important. However, conventional safety confirmation methods are manual, with many subjective elements and limitations in the accuracy of safety confirmation. Therefore, even in high-risk sites such as base station construction, accidents are likely to occur. Also, in the toolbox meeting (TBM-KY) before on-site work, it is difficult to fully grasp and share potential risks. Therefore, there is a need for a system that can objectively and quickly identify potential risks at the site and improve the accuracy of safety confirmation.

Means for Solving the Problems

[0005] The present invention is a safety confirmation system for on-site work, comprising the following means: means for a user to upload photos taken on-site to a server via a terminal; means for the server to transmit the received images to a generative artificial intelligence for analysis; means for the generative artificial intelligence to identify potential hazards and points of caution in the images and generate a report; means for the server to transmit the generated report to a terminal for the user to view it; and means for the user to reflect safety confirmation items based on the report in a pre-work toolbox meeting. This system enables the objective and rapid identification of potential hazards on-site and facilitates sharing via TBM-KY. Including images that visually show hazardous areas as a result of the analysis in the report generated by the server makes it easier to understand visually. Furthermore, including means for the generative artificial intelligence to detect objects, people, and environmental elements in the images and compare them with safety standards based on these enables more accurate safety confirmation.

[0006] The "On-site Safety Confirmation System" is a system in which a generative artificial intelligence system identifies potential hazards and points of caution based on photos taken by the user at the work site and generates a report.

[0007] A "user" refers to a person performing work on-site or a person supervising that work, and is someone who uses this system to perform safety checks.

[0008] "Terminal" refers to a device (e.g., smartphone, tablet, etc.) that a user uses to take photos on-site and upload them to a server.

[0009] A "server" is a computing system that receives image data sent by users, transmits it to a generative artificial intelligence system for analysis, generates a report, and sends it to the terminal.

[0010] "Generative artificial intelligence" is an AI technology that detects objects, people, and environmental elements within an image, identifies potential hazards and points of caution based on these, and generates a report.

[0011] A "report" is a document that summarizes the risks, points of caution, and countermeasures identified as a result of analysis by a generative artificial intelligence, and is provided to the user.

[0012] A "toolbox meeting" is a short meeting held before work begins where on-site workers gather to share details about the work and safety check items.

[0013] "Risk" refers to factors that could potentially cause accidents or injuries when performing work on-site.

[0014] "Points to note" refers to matters and measures that require particular attention during work, and is information intended to prevent accidents and problems. [Brief explanation of the drawing]

[0015] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] Shows an emotion map to which a plurality of emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

Embodiments for Carrying Out the Invention

[0016] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0017] First, the terms used in the following description will be explained.

[0018] In the following embodiments, the numbered processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of a plurality of arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of a plurality of types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0019] In the following embodiments, the numbered RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0020] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0021] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0022] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0023] [First Embodiment]

[0024] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0025] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0026] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0027] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0028] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0029] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0030] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0031] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0033] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0034] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0035] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0036] This invention relates to a safety verification system for on-site work, and provides a system in which a generative artificial intelligence identifies potential hazards and points of caution based on photographs taken by the user at the work site and generates a report.

[0037] System Configuration

[0038] This system consists of the following elements:

[0039] 1. Terminal

[0040] This is a device for users to take photos on-site.

[0041] It has a function to upload the photos taken to a server.

[0042] 2. Server

[0043] Receive photo data sent from the device.

[0044] Image analysis is performed using generative artificial intelligence.

[0045] A report is generated based on the analysis results.

[0046] Send the report to the terminal.

[0047] 3. Generative Artificial Intelligence

[0048] It is an artificial intelligence embedded in a server that detects objects, people, and environmental elements within images.

[0049] Identify potential hazards and points of caution.

[0050] Automatically generate reports.

[0051] Program processing

[0052] 1. Upload image

[0053] Users take photos of areas requiring safety checks using their smartphones or tablets at the site.

[0054] The device uploads the photos it takes to the server using a dedicated app.

[0055] 2. Image reception

[0056] The server receives and stores image data transmitted from terminals via the internet.

[0057] 3. Image Analysis

[0058] The server sends the stored images to the generative artificial intelligence.

[0059] Generative artificial intelligence detects and analyzes objects, people, and environmental elements within an image.

[0060] Based on the detected information, it is compared with pre-set safety standards to identify potential hazards and points of caution.

[0061] 4. Report generation

[0062] The server generates a report based on the analysis results from the generative artificial intelligence.

[0063] The report includes specific hazardous areas, proposed countermeasures, and images that visually represent these hazardous areas.

[0064] 5. Report distribution

[0065] The server sends the generated report to the terminal.

[0066] The terminal receives the report and displays it for the user to view.

[0067] 6. Reflection in TBM-KY

[0068] Based on the report content, the user adds safety check items to the TBM-KY record.

[0069] The report contents are shared in a meeting before on-site work begins, ensuring everyone is aware of the risks.

[0070] Specific example

[0071] Example 1: Image upload

[0072] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[0073] Specific example 2: Image analysis

[0074] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs to be reinforced."

[0075] Specific example 3: Report generation

[0076] The server generates a report based on the analysis results. It creates a report stating, "Unstable areas of the scaffolding have been detected. Reinforcement is required," and also inserts a photo of the dangerous areas circled in red as part of the analysis results.

[0077] Specific Example 4: Report Distribution and TBM-KY Reflection

[0078] The server generates a report and sends it to the user's terminal. The user views the report, adds "scaffolding reinforcement" as an item requiring it to TBM-KY, and notifies everyone.

[0079] This invention provides an effective system for field workers to quickly and accurately identify potential hazards at a work site, thereby improving the accuracy of safety checks and preventing accidents from occurring.

[0080] The following describes the processing flow.

[0081] Step 1:

[0082] Users take photos on-site. They use smartphones or tablets to take pictures of areas where safety checks are necessary.

[0083] Step 2:

[0084] The device saves the photos, and the app uploads them to the server. When the user selects photos taken using the dedicated app and presses the upload button, the app compresses the image files and sends them to the server via the internet.

[0085] Step 3:

[0086] The server receives the image. The server receives the uploaded image data via the internet and saves it to the specified directory.

[0087] Step 4:

[0088] The server sends the stored image to the generative artificial intelligence. The server then passes the received image data to the analysis module, which begins image analysis.

[0089] Step 5:

[0090] Generative artificial intelligence performs image analysis. The AI ​​detects objects, people, and environmental elements within the image and identifies potential hazards and points of caution based on these. For example, it can determine unstable areas of scaffolding or the proximity of power lines.

[0091] Step 6:

[0092] The server generates a report based on the analysis results. Upon receiving the analysis results from the generative artificial intelligence, the server automatically generates a report that includes the identified hazardous areas and suggested countermeasures. The report includes images showing the hazardous areas with red frames as part of the analysis results.

[0093] Step 7:

[0094] The server sends the generated report to the terminal. The server sends the generated report to the user's terminal and notifies the user of the report's arrival via push notification.

[0095] Step 8:

[0096] The device displays the report. The user opens a dedicated app and views the report sent from the server. The report includes specific hazardous areas, suggested countermeasures, and images visually illustrating the hazardous areas.

[0097] Step 9:

[0098] The user reflects the information in TBM-KY. Based on the report, the user adds safety check items to the pre-work toolbox meeting (TBM-KY) and shares them with all field workers.

[0099] By following the steps outlined above, potential hazards at the site are identified objectively and quickly. Through sharing this information via TBM-KY (Track Byte-Marketing and Hazard Prediction), everyone becomes aware of the hazards and takes appropriate measures, thereby improving the accuracy of safety checks.

[0100] (Example 1)

[0101] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0102] In on-site work, it is essential to quickly and accurately identify potential hazards and efficiently conduct safety checks. However, currently, many sites rely on paper-based checklists and experience, which can lead to overlooking hazards and inadequate countermeasures. Furthermore, identifying individual hazardous areas and responding quickly becomes difficult, making it challenging to ensure overall site safety.

[0103] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0104] In this invention, the server includes means for uploading photos taken by the user on-site to the server via a terminal; means for the server to store the received image data in a database and transmit it to a generative artificial intelligence; means for the generative artificial intelligence to detect objects, people, and environmental elements in the image and identify potential hazards and points of caution based on this information by comparing it with safety standards; means for the server to generate a report based on the analysis results and transmit it to the terminal of a designated user; means for the terminal to receive the generated report and display it so that the user can view it; and means for the user to reflect safety confirmation items in a toolbox meeting based on the contents of the report. This makes it possible to quickly and accurately identify potential hazards in on-site work and to take prompt countermeasures against the identified hazardous areas.

[0105] A "user" is an individual or group engaged in on-site work who uses the system to perform safety checks.

[0106] A "device" refers to a smartphone, tablet, or other electronic device used by a user to upload photos taken on-site to a server.

[0107] A "server" is a computer system that receives photo data sent from a terminal, requests image analysis from a generative artificial intelligence, and generates a report based on that analysis.

[0108] "Generative artificial intelligence" refers to artificial intelligence programs that have algorithms to detect objects, people, and environmental elements within an image and identify potential dangers and points of caution.

[0109] "Image data" refers to photographic data taken by a user on their device and uploaded to the server.

[0110] A "report" is a document that includes specific hazardous areas and recommended countermeasures, based on the analysis results of a generative artificial intelligence.

[0111] A "toolbox meeting" is a meeting held before work begins where workers gather to check safety procedures and share information about the work to be done.

[0112] This invention relates to a safety verification system for on-site work, providing a system in which a generative artificial intelligence identifies potential hazards and points of caution based on photographs taken by the user at the worksite and generates a report. The components of this system are a terminal, a server, and a generative artificial intelligence.

[0113] System components

[0114] 1. Terminal

[0115] A terminal refers to a device such as a smartphone or tablet used by a user to record photos taken on-site.

[0116] The device requires the installation of a dedicated application. This dedicated application has functions for compressing, selecting, and uploading photos.

[0117] 2. Server

[0118] A server is a computer system that receives image data sent by users via the internet and stores it in a database.

[0119] The server has a generative artificial intelligence (AI) built in, and it sends the received image data to the AI.

[0120] The server generates a report based on the analysis results of the generative artificial intelligence and sends it to the user's terminal.

[0121] 3. Generative Artificial Intelligence

[0122] Generative artificial intelligence is an AI algorithm embedded in a server that has the function of detecting objects, people, and environmental elements within an image.

[0123] Generative artificial intelligence identifies potential hazards and points of caution by comparing them with safety standards and automatically generates reports.

[0124] System Operation Instructions

[0125] Users take photos of areas requiring safety checks on-site and upload the photo data to a server using a dedicated app. Upon receiving the image data, the server requests analysis from a generative artificial intelligence (AI). The AI ​​analyzes the image data and identifies potential hazards and points of caution. Based on the analysis results, the server generates a report and sends it back to the user's terminal. Users view the report and incorporate the findings into toolbox meetings.

[0126] Specific example

[0127] Example 1: Image upload

[0128] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[0129] Specific example 2: Image analysis

[0130] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs reinforcement."

[0131] Specific example 3: Report generation

[0132] The server generates a report based on the analysis results. The report includes a message stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," along with a photograph showing the dangerous areas circled in red as a result of the analysis.

[0133] Specific Example 4: Report Distribution and TBM-KY Reflection

[0134] The server generates a report and sends it to the user's terminal. The user views the report, adds "reinforcement of scaffolding" as an item requiring it to TBM-KY (Toolbox Meeting Keep and Yado), and notifies everyone.

[0135] This invention provides an effective system for field workers to quickly and accurately identify potential hazards at a work site, thereby improving the accuracy of safety checks and preventing accidents from occurring.

[0136] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0137] Step 1:

[0138] Users take photos of areas requiring safety checks using their smartphones or tablets at the site. After taking the photos, they select them using a dedicated app on their device and press the upload button. The device compresses the selected photos to prepare them for efficient use of network bandwidth. The input is the captured photo data, and the output is the compressed photo data.

[0139] Step 2:

[0140] The device sends compressed photo data to the server. The server receives the photo data via the internet and temporarily stores it in a database. The input is the compressed photo data, and the output is the image data stored in the database. The server returns a response to the device confirming that the image was saved successfully.

[0141] Step 3:

[0142] The server retrieves image data stored in the database and sends it to the generative artificial intelligence (AI). The input is the image data stored in the database, and the output is the data sent to the generative AI. Next, the generative AI receives the image and executes an object detection algorithm. Specifically, it identifies objects, people, and environmental elements in the image and returns this information to the server in an encoded format.

[0143] Step 4:

[0144] The generative artificial intelligence sends detection results from the image back to the server. This response includes data on identified objects, people, environmental elements, and potential hazardous areas. The input is image data to be analyzed, and the output is detection result data. Based on the received analysis data, the server compares it with established safety standards to identify hazardous areas and points of caution. The output is an interim report summarizing the hazardous areas and points of caution.

[0145] Step 5:

[0146] The server generates a report document using the analysis results of the generative artificial intelligence. The report includes details of detected hazardous areas, recommended countermeasures, and images highlighting the hazardous areas in red. The input is interim report data, and the output is a report document (in PDF or HTML format).

[0147] Step 6:

[0148] The server sends the generated report to the designated user's device. The device receives the report and notifies the user via a dedicated app. The input is the report document, and the output is the report notification to the user's device. The user views the report using the dedicated app to check for specific instructions and countermeasures.

[0149] Step 7:

[0150] Based on the generated report, the user adds the necessary safety check items to the TBM-KY (Toolbox Meeting Keep and Yado) record. The user shares the contents of this report with other workers during the pre-work meeting so that everyone is aware of the hazards. The input is the report content, and the output is the updated TBM-KY record.

[0151] (Application Example 1)

[0152] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0153] Traditional on-site safety check systems primarily rely on manual inspections, which suffers from the time and effort required to identify potential hazards. Furthermore, while many robots operate within factories, and there is a need to improve the efficiency of safety checks, the lack of robots themselves capable of performing safety checks of the work environment hinders progress in this area as well.

[0154] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0155] This invention includes a server comprising means for uploading photos taken by a user on-site to the server via a terminal, means for the server to transmit the received images to a generative artificial intelligence for analysis, means for the generative artificial intelligence to identify potential hazards and points of caution in the images and generate a report, means for the server to transmit the generated report to a terminal for the user to view, means for the user to reflect safety confirmation items based on the report in a pre-work toolbox meeting, means for a robot to take photos of the work area and upload those images to the server, means for the server to analyze images inside the production facility and identify potential hazardous areas, and means for the server to transmit the generated report to a work manager's terminal for viewing. This enables the rapid identification of potential hazardous areas and points of caution in on-site work and work areas within factories, improving the efficiency and accuracy of safety checks.

[0156] A "user" is a worker or manager who uses the safety verification system.

[0157] "The site" refers to factories, construction sites, and other workplaces.

[0158] A "device" is a device used by a user to take photos and upload them to a server, and includes smartphones, tablets, and other similar devices.

[0159] A "server" is a computer system that receives and analyzes image data sent by users and generates reports.

[0160] "Generative artificial intelligence" is an artificial intelligence technology that analyzes objects and environmental elements within an image to identify potential hazards and points of caution.

[0161] A "report" is a document containing information analyzed by a generative artificial intelligence system, including information on hazardous areas and countermeasures.

[0162] A "toolbox meeting" is a pre-work safety check meeting.

[0163] A "robot" is an automated work machine used in a factory, and is a device equipped with the image capture function of the present invention.

[0164] "Work area" refers to the area or section where robots and workers operate.

[0165] "Uploading" refers to the act of sending images taken with a device or robot to a server.

[0166] "Analysis" refers to the process by which generative artificial intelligence processes image data to identify potential hazards and points of caution.

[0167] A "manager" is a person responsible for safety management in a factory or work site.

[0168] A "production facility" refers to a place where products are produced, including factories and manufacturing sites.

[0169] "Potential hazard areas" are locations identified through image analysis where accidents or injuries are likely to occur.

[0170] "Visually representing images" are images that highlight dangerous areas or add supplementary explanations based on the analysis results.

[0171] This invention provides a system that streamlines safety checks in factories and on-sites, analyzing user-submitted photographs to identify potential hazards and generate reports. Specific embodiments of this system are described below.

[0172] First, the user takes photos of the work area and equipment using a device such as a smartphone or tablet. A dedicated application is installed on this device, and the captured images are uploaded to the server via this application. The uploaded image data is received and stored by the server. The stored images are then sent to a generative artificial intelligence.

[0173] The server analyzes the received images using generative artificial intelligence. During this process, the AI ​​detects objects, people, and environmental elements within the image and identifies potential hazards and points of caution based on pre-defined safety criteria. This generative AI utilizes libraries such as TENSORFLOW®.

[0174] Next, the server automatically generates a report based on the analysis results from the generative artificial intelligence. This report contains information on identified hazardous areas and suggested countermeasures. It also includes images that visually represent the hazardous areas based on the analysis results. Libraries such as ReportLab can be used to create the report.

[0175] The generated report is sent from the server to the user's terminal, where the user can view it. Furthermore, based on this report, the user adds safety check items to the pre-work toolbox meeting and informs everyone.

[0176] Furthermore, a dedicated application can be installed on robots used within the factory, allowing them to take photos of the work area using their onboard cameras. The robots upload the captured images to a server, where they are similarly analyzed by generative artificial intelligence. The reports generated based on the analysis results are sent to the work supervisor's terminal, enabling them to review the information immediately.

[0177] Specific example

[0178] 1. Image capture: Users take photos of specific areas within the factory using their smartphones. After taking the photos, they upload the images to the server using a dedicated application.

[0179] 2. Example of a prompt:

[0180] Analyze the images of this factory to identify potential hazards.

[0181] 3. Image Analysis: Generative artificial intelligence analyzes photographs to detect unstable scaffolding, unsecured equipment, and other issues.

[0182] 4. Report generation: The server generates a report stating, "Unstable scaffolding detected. Reinforcement is required," and includes an image with the unstable area highlighted in red.

[0183] 5. Report Distribution: The server generates reports and sends them to the user's smartphone for viewing. This allows for quick implementation of necessary safety measures before actual work begins.

[0184] The above describes the basic configuration for carrying out the present invention. This system makes on-site work and safety checks within factories more efficient and enables accident prevention.

[0185] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0186] Step 1:

[0187] The user takes photos of the work area and equipment using a device (smartphone or tablet). Specifically, this involves launching the application and capturing the desired scene in camera mode. In this case, the input is the captured image, and the output is the image file saved on the device.

[0188] Step 2:

[0189] The captured photos are uploaded to the server via a dedicated app. Specifically, pressing the "upload button" in the application sends the image data to the server. In this process, the input is the image file on the device, and the output is the image data transferred to the server.

[0190] Step 3:

[0191] The server saves the received image data and sends it to the generative artificial intelligence. Specifically, it saves the received image data as a temporary file and passes its path to the generative AI. In this case, the input is the image data sent from the terminal, and the output is the provision of the path to the image data to the generative AI.

[0192] Step 4:

[0193] Generative artificial intelligence analyzes image data to identify potential hazards and points of caution. Specifically, it applies object detection algorithms within images to identify hazardous areas. In this process, the input is image data transmitted from a server, and the output is hazardous area information as a result of the analysis.

[0194] Step 5:

[0195] The server generates a report based on the analysis results from the generative artificial intelligence. Specifically, it uses a report generation library to create a page that visually displays an overview of the analysis results and areas of concern. In this process, the input is the analysis result data from the generative artificial intelligence, and the output is the completed report file.

[0196] Step 6:

[0197] The server sends the generated report to the user's device, and the user views it. Specifically, the server sends the report file to the user's device via email or in-app notification, and the user receives the notification and opens the report. In this process, the input is the generated report file, and the output is the report displayed on the user's device.

[0198] Step 7:

[0199] Based on the report, the user incorporates safety check items into the pre-work toolbox meeting. Specifically, this involves reviewing the report, sharing the information in the meeting, and taking necessary safety measures. In this process, the input is the submitted report, and the output is the safety check information shared within the meeting.

[0200] Step 8:

[0201] The robot takes pictures of the work area and uploads the images to a server. Specifically, the robot's camera function is automatically activated, images are captured, and uploaded to the server via the network. In this process, the input is the image data captured by the robot, and the output is the image data transferred to the server.

[0202] Step 9:

[0203] The server analyzes images from inside the production facility to identify potential hazards. Specifically, it uses generative artificial intelligence to analyze objects and environmental elements within the work area and highlight hazards. In this process, the input is image data transmitted from the robot, and the output is information about the identified hazards.

[0204] Step 10:

[0205] The server sends the generated report to the work manager's terminal, ensuring that the manager can review it immediately. Specifically, the server delivers the report, including the analysis results, to the work manager's terminal via email or a dedicated application, allowing the manager to view the report content immediately. In this case, the input is the generated report, and the output is the report displayed on the work manager's terminal.

[0206] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0207] This invention combines a safety confirmation system for on-site work with an emotion engine that recognizes user emotions. In addition to a system that uses generative artificial intelligence to identify potential hazards and points of caution based on photos taken by the user at the worksite and generates a report, it also provides a function that performs safety confirmation while taking the user's emotional state into consideration.

[0208] System Configuration

[0209] This system consists of the following elements:

[0210] 1. Terminal

[0211] This is a device for users to take photos on-site.

[0212] It has a function to upload the photos taken to a server.

[0213] It implements an emotion engine and has the ability to analyze the user's voice and facial expressions.

[0214] 2. Server

[0215] Receive photo data sent from the device.

[0216] Image analysis is performed using generative artificial intelligence.

[0217] A report is generated based on the analysis results.

[0218] Send the report to the terminal.

[0219] 3. Generative Artificial Intelligence

[0220] It is an artificial intelligence embedded in a server that detects objects, people, and environmental elements within images.

[0221] Identify potential hazards and points of caution.

[0222] Automatically generate reports.

[0223] 4. Emotional Engine

[0224] The system analyzes the user's voice and facial expressions to identify their emotional state.

[0225] Depending on the emotional state, the generated reports will include additional warnings and alerts.

[0226] Program processing

[0227] 1. Upload image

[0228] Users take photos of areas requiring safety checks using their smartphones or tablets at the site.

[0229] The device uploads the photos it takes to the server using a dedicated app.

[0230] 2. Image reception

[0231] The server receives image data sent from the terminal via the internet and saves it to the specified directory.

[0232] 3. Image Analysis

[0233] The server sends the stored images to the generative artificial intelligence.

[0234] Generative artificial intelligence detects and analyzes objects, people, and environmental elements within an image.

[0235] Based on the detected information, it is compared with pre-set safety standards to identify potential hazards and points of caution.

[0236] 4. Emotion analysis

[0237] The device analyzes the user's voice and facial expressions during shooting using an emotion engine to identify the user's emotional state.

[0238] For example, if signs of anxiety or tension are detected, that information is sent to the server.

[0239] 5. Report generation

[0240] The server generates a report based on the analysis results from the generative artificial intelligence.

[0241] The report includes specific hazardous areas, proposed countermeasures, and images that visually represent these hazardous areas.

[0242] Based on the user's emotional state identified by the emotion engine, the report will include additional warnings and alerts as needed.

[0243] 6. Report distribution

[0244] The server sends the generated report to the terminal.

[0245] The terminal receives the report and displays it for the user to view.

[0246] 7. Reflection in TBM-KY

[0247] Based on the report content, the user adds safety check items to the TBM-KY record.

[0248] The report contents are shared in a meeting before on-site work begins, ensuring everyone is aware of the risks.

[0249] Specific example

[0250] Example 1: Image upload

[0251] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[0252] Specific example 2: Image analysis

[0253] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs to be reinforced."

[0254] Example 3: Sentiment Analysis

[0255] The device uses an emotion engine to analyze the user's voice and facial expressions while they are taking photos, and detects if the user is feeling anxious. This information is then sent to the server.

[0256] Specific Example 4: Report Generation

[0257] The server generates a report based on the analysis results. In addition to a report stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," it also includes a warning that takes into account the user's level of anxiety.

[0258] Specific Example 5: Report Distribution and TBM-KY Reflection

[0259] The server generates a report and sends it to the user's terminal. The user views the report, adds "scaffolding reinforcement" as an item requiring it to TBM-KY, and notifies everyone.

[0260] This invention aims to further improve safety by combining an emotion engine to enable safety checks that also take into account the emotional state of on-site workers.

[0261] The following describes the processing flow.

[0262] Step 1:

[0263] Users take photos on-site. They use smartphones or tablets to take pictures of areas where safety checks are necessary.

[0264] Step 2:

[0265] The device saves the photos, and the app uploads them to the server. When the user selects photos taken using the dedicated app and presses the upload button, the app compresses the image files and sends them to the server via the internet.

[0266] Step 3:

[0267] The device collects user voice and facial expression data. While taking or uploading photos, the device uses its built-in camera and microphone to collect user voice and facial expression data.

[0268] Step 4:

[0269] The device passes the collected voice and facial expression data to the emotion engine. The emotion engine analyzes the data and identifies the user's emotional state. For example, it can determine from the voice whether the user is nervous or not.

[0270] Step 5:

[0271] The server receives the image. The server receives the uploaded image data via the internet and saves it to the specified directory.

[0272] Step 6:

[0273] The server sends the stored image to the generative artificial intelligence. The server then passes the received image data to the analysis module, which begins image analysis.

[0274] Step 7:

[0275] Generative artificial intelligence performs image analysis. The AI ​​detects objects, people, and environmental elements within the image and identifies potential hazards and points of caution based on these. For example, it can determine unstable areas of scaffolding or the proximity of power lines.

[0276] Step 8:

[0277] The server generates a report based on the analysis results. Upon receiving the analysis results from the generative artificial intelligence, the server automatically generates a report that includes the identified hazardous areas and suggested countermeasures. The report includes images showing the hazardous areas with red frames as part of the analysis results.

[0278] Step 9:

[0279] The server receives emotion data from the emotion engine. If the user is in a stressed state, the server uses this information to include additional warnings and alerts in the report.

[0280] Step 10:

[0281] The server sends the generated report to the terminal. The server sends the generated report to the user's terminal and notifies the user of the report's arrival via push notification.

[0282] Step 11:

[0283] The device displays the report. The user opens a dedicated app and views the report sent from the server. The report includes specific hazardous areas, suggested countermeasures, and images visually representing the hazardous areas. It also includes additional warnings based on the user's emotional state.

[0284] Step 12:

[0285] The user reflects it in the TBM-KY. Based on the content of the report, the user adds the safety confirmation items to the toolbox meeting (TBM-KY) before on-site work and shares them with all on-site workers.

[0286] In the above steps, by objectively and quickly identifying potential risks at the site and taking appropriate measures considering the emotional state of the user, the accuracy of safety confirmation is improved.

[0287] (Example 2)

[0288] Next, Example 2 will be described. In the following description, the data processing device 12 is referred to as a "server", and the smart device 14 is referred to as a "terminal".

[0289] In the conventional safety confirmation system for on-site work, since the emotional state of the user is not considered, the stress and anxiety felt by the worker cannot be appropriately reflected, and there is a possibility of overlooking potential risks or taking inappropriate measures. Furthermore, since these systems do not have sufficient means to visually show the analysis results, there is a problem that it is difficult for the user to intuitively understand specific dangerous locations. Therefore, the object of this invention is to solve these problems and realize more accurate safety confirmation and countermeasures while considering the emotional state of the worker.

[0290] The specific processing by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0291] In this invention, the server includes means for uploading images taken by the user on-site to the server via a terminal, means for the server to save the received image data to a specified directory, means for the server to transmit the saved image data to a generative artificial intelligence for analysis, means for the generative artificial intelligence to detect objects, people, and environmental elements in the images and generate a report based on these to identify potential hazards and points of caution, means for the terminal to transmit the user's voice and facial expression data to an emotion analysis engine to identify the user's emotional state, means for the server to include additional warnings and alerts in the report based on the user's emotional state, means for the server to transmit the generated report to the terminal for the user to view it, and means for the user to reflect safety confirmation items based on the generated report in a pre-work meeting. This makes it possible to perform safety checks that take into account the user's emotional state, and by visually indicating specific hazardous areas, more intuitive and effective safety measures can be realized.

[0292] "Users" refer to workers and managers who take images on-site and use the system.

[0293] A "terminal" is a device used by a user that has the function of taking images on-site and sending the data to a server. Specific examples include smartphones and tablets.

[0294] A "server" refers to a computer system that processes data received from a terminal, works with generative artificial intelligence and emotion analysis engines to generate analysis results, and sends reports to the terminal.

[0295] "Generative artificial intelligence" refers to machine learning models or algorithms that analyze image data to detect objects, people, and environmental elements within the image, and identify potential dangers and points of caution.

[0296] An "emotion analysis engine" refers to an algorithm or technology that analyzes a user's voice and facial expression data to identify the user's emotional state.

[0297] "Image data" refers to photographs and videos taken by users on-site, which are uploaded to the server and used for analysis.

[0298] A "directory" refers to a specific folder or part of storage used to store data received within a server.

[0299] A "report" refers to a document that integrates the results of image analysis by generative artificial intelligence and sentiment analysis by a sentiment analysis engine, including potential risks, points of caution, and countermeasures.

[0300] A "meeting" refers to a planning meeting held by users before starting work, where safety checks are conducted based on the contents of reports, such as TBM-KY (Tool Box Meeting - Kiken Yochi, Hazard Prediction Activity).

[0301] "Warning" refers to additional warnings or alerts included in reports based on the user's emotional state, providing information that encourages workers to pay close attention to specific hazards.

[0302] "Analysis results" refer to data provided by generative artificial intelligence and emotion analysis engines, including detected risk areas, countermeasures against them, and information on the user's emotional state.

[0303] This invention relates to a safety verification system for on-site work, and by combining it with a function to recognize the user's emotions, it achieves more effective safety measures. This system consists of the following hardware and software.

[0304] System Configuration

[0305] This system includes the following main components:

[0306] 1. Terminal

[0307] It is a device for users to take photos on-site. Specific examples include smartphones and tablets.

[0308] Install a dedicated application for uploading the captured image data to the server.

[0309] It has a function of sending the user's voice and facial expression data to the emotion analysis engine to identify the user's emotion.

[0310] 2. Server

[0311] Save the image data received from the terminal in the specified directory.

[0312] Install and analyze software that operates as a generative artificial intelligence (e.g., the API of OpenAI (registered trademark)).

[0313] Integrate the analysis results using software that operates as an emotion analysis engine (e.g., emotion recognition API).

[0314] Generate a report and send it to the terminal.

[0315] 3. Generative Artificial Intelligence

[0316] It is a machine learning model incorporated in the server that detects objects, people, and environmental elements in the image.

[0317] Identify potential risks and precautions and automatically generate a report.

[0318] The software to be used is the API of OpenAI or similar analysis tools.

[0319] 4. Emotion Analysis Engine

[0320] Analyze audio and video data to identify the user's emotional state.

[0321] Based on emotional states, the generated reports will include additional warnings and alerts.

[0322] Program processing

[0323] The program for this system processes the information in the following order:

[0324] 1. Upload image

[0325] Users take pictures of areas requiring safety checks using their smartphones or tablets at the site.

[0326] The device uploads captured images to a server using a dedicated app. This app has the functionality to transmit image data over the internet.

[0327] 2. Image reception

[0328] The server receives image data sent from the terminal via the internet and saves it to a specified directory. Specifically, it saves files to cloud storage or local storage.

[0329] 3. Image Analysis

[0330] The server sends the stored image data to a generative artificial intelligence system. For example, a Python script on the server can be used to send the image data to an API.

[0331] Generative artificial intelligence analyzes received image data to detect objects, people, and environmental elements within the image. Specifically, it uses computer vision technology to identify unstable areas of scaffolding or lack of safety equipment.

[0332] 4. Emotion analysis

[0333] The device transmits the user's voice and facial expressions to an emotion analysis engine to identify the user's emotional state. This collects data using the smartphone's camera and microphone.

[0334] The terminal sends the analysis results to the server, where they are used as additional security management information.

[0335] 5. Report generation

[0336] The server generates a report based on analysis results from a generative artificial intelligence and the sentiment analysis engine. This report includes a detailed explanation of the hazardous areas, specific countermeasures, and warnings that reflect the user's emotional state.

[0337] The report will be generated in PDF format or another suitable format and saved to the specified directory.

[0338] 6. Report distribution

[0339] The server sends the generated report to the device. For example, it notifies the user of the report using email or push notifications via a dedicated app.

[0340] The terminal displays the received report, allowing the user to review it.

[0341] 7. Reflection in TBM-KY

[0342] Users conduct pre-work meetings based on reports during safety meetings (e.g., TBM-KY) to confirm hazardous areas and countermeasures.

[0343] For example, report contents can be shared on a smartphone screen, and necessary safety measures can be discussed.

[0344] Specific example

[0345] Example 1: Image upload

[0346] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[0347] Specific example 2: Image analysis

[0348] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs to be reinforced."

[0349] Example 3: Sentiment Analysis

[0350] The device uses an emotion analysis engine to analyze the user's voice and facial expressions while they are taking photos, and detects if the user is feeling anxious. This information is then sent to the server.

[0351] Specific Example 4: Report Generation

[0352] The server generates a report based on the analysis results. In addition to a report stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," it also includes a warning that takes into account the user's level of anxiety.

[0353] Specific Example 5: Report Distribution and TBM-KY Reflection

[0354] The server generates a report and sends it to the user's terminal. The user views the report, adds "scaffolding reinforcement" as an item requiring it to TBM-KY, and notifies everyone.

[0355] Example of a prompt:

[0356] "Identify potential hazards in this image and propose safety measures."

[0357] This enables safety checks for on-site work that take into account the user's emotional state, and by visually indicating specific hazardous areas, it allows for more intuitive and effective safety measures.

[0358] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0359] Step 1: Take and upload images

[0360] Users take pictures of areas requiring safety checks using their smartphones or tablets at the site.

[0361] Input: Image taken by the user.

[0362] Specific operation: The user takes pictures of the site using their smartphone's camera app and selects those images within the dedicated app.

[0363] The device uploads images to the server using a dedicated app. This app transmits image data via HTTPS over the internet.

[0364] Output: Image data uploaded to the server.

[0365] Step 2: Receiving and saving image data

[0366] The server receives image data sent from the terminal via the internet.

[0367] Input: Uploaded image data.

[0368] Specific operation: The server receives image data via a dedicated API and saves it to the directory " / images / uploads". The save location is recorded in the database.

[0369] Output: Image files saved in the specified directory.

[0370] Step 3: Perform image analysis

[0371] The server transmits the stored image data to the generative artificial intelligence system.

[0372] Input: The path to the image file saved in the directory.

[0373] Specific operation: A Python script on the server is used to send the image data path to the API.

[0374] Generative artificial intelligence analyzes image data to detect objects, people, and environmental elements.

[0375] Output: Analysis results (list of detected objects, people, and environmental elements).

[0376] Step 4: Identifying potential hazards

[0377] Generative artificial intelligence identifies potential risks and points of caution based on the analysis results.

[0378] Input: A list of detected objects, people, and environmental elements.

[0379] Specific operation: The generative artificial intelligence uses an analysis algorithm to compare the detected elements with safety standards and identify hazardous areas and points of caution.

[0380] Output: List of hazardous areas and points to note.

[0381] Step 5: Perform sentiment analysis

[0382] The device transmits the user's voice and facial expressions during shooting to an emotion analysis engine.

[0383] Input: User's voice data and facial expression data.

[0384] Specific operation: The device's microphone and camera are used to collect voice and facial expression data, which is then sent to an emotion analysis engine via a dedicated app.

[0385] The emotion analysis engine analyzes this data to identify the user's emotional state.

[0386] Output: User's emotional state (anxiety, tension, relief, etc.).

[0387] Step 6: Generate the report

[0388] The server generates a report based on the results of generative artificial intelligence and emotion analysis engines.

[0389] Input: List of hazards and points of caution, user's emotional state.

[0390] Specific operation: The server uses a web framework such as Django to integrate the analysis results and create a report. The report includes risk areas, countermeasures, and additional warnings based on the user's emotional state. The report is generated in PDF format, etc.

[0391] Output: The generated report file.

[0392] Step 7: Report Distribution

[0393] The server sends the generated report to the terminal.

[0394] Input: The generated report file.

[0395] Specific operation: The server uses a dedicated API to send report files to the terminal. For example, it might use email or push notifications.

[0396] The device displays received reports within a dedicated app, allowing users to review them.

[0397] Output: The report file delivered to the terminal.

[0398] Step 8: Reflection in TBM-KY

[0399] The user updates the TBM-KY record based on the generated report.

[0400] Input: Report file delivered to the terminal.

[0401] Specific actions: Users view reports in safety meetings and add necessary safety check items to the TBM-KY sheet. This is then shared with everyone, and specific safety measures are discussed.

[0402] Output: Updated TBM-KY record.

[0403] (Application Example 2)

[0404] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0405] Conventional safety confirmation systems for on-site work often failed to consider the user's psychological state, resulting in situations where appropriate warnings and responses could not be provided. Furthermore, because potential hazards and points of caution were identified solely through image analysis, safety measures based on the user's emotional changes were insufficient. This led to problems where optimal safety checks could not be performed, increasing the risk of workplace accidents.

[0406] In Application Example 2, the identification processing by the identification processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for uploading photos taken by the user at the site to the server via a terminal, means for the server to transmit the received images to a generative artificial intelligence for analysis, means for the generative artificial intelligence to identify potential dangers and points of caution in the images and generate a report, means for the server to transmit the generated report to the terminal for the user to view, means for the terminal to identify the user's emotional state using an emotion engine that analyzes the user's voice and facial expressions, and means for the emotion engine to transmit the analysis results to the server and include additional warnings and alerts in the report. This makes it possible to perform a comprehensive safety check that takes the user's emotions into consideration and reduces the risk of occupational accidents.

[0407] A "site work safety confirmation system" is a system for ensuring safety during on-site work by checking the conditions of the work site, identifying potential hazards and points of caution, and implementing safety measures.

[0408] A "terminal" is a device that allows users to take photos on-site and upload them to a server.

[0409] A "server" is a device that receives image data transmitted from a terminal, analyzes it using generative artificial intelligence, generates a report based on the analysis results, and then transmits that report back to the terminal.

[0410] "Generative artificial intelligence" is an artificial intelligence that analyzes transmitted image data, detects objects, people, and environmental elements within the image, and has the function of identifying potential dangers and points of caution.

[0411] A "report" is a document created based on the results of analysis by a generative artificial intelligence system, which describes potential hazards, points of caution, and countermeasures to address them.

[0412] An "emotion engine" is software or hardware that analyzes a user's voice and facial expressions to identify the user's emotional state.

[0413] A "toolbox meeting" is a pre-work meeting where workers gather to share information about the day's tasks, precautions, and safety checks.

[0414] "Safety confirmation items" are matters and procedures that must be checked to ensure safety during on-site work, and are shared at toolbox meetings.

[0415] The following describes an embodiment for carrying out this invention. This system allows users to upload photos taken with a smartphone or other device to a server for safety checks during on-site work, and a generative artificial intelligence analyzes the photo data. Furthermore, safety is further enhanced by analyzing the user's emotional state using an emotion engine and reflecting it in the report.

[0416] Hardware and software to use

[0417] Device: Use a device such as a smartphone or tablet that has the ability to take photos on-site and upload them to a server.

[0418] Server: This device receives image data, performs analysis using generative artificial intelligence, generates a report, and sends it to the terminal.

[0419] Generative artificial intelligence: This uses artificial intelligence implemented on a server to analyze image data and identify potential dangers and points of caution.

[0420] Emotion engine: This is software or hardware implemented in the device that analyzes the user's voice and facial expressions to identify their emotional state.

[0421] Data processing and data calculation

[0422] 1. Upload image:

[0423] The user takes a photo of the site with their device. The device then uploads this photo to the server.

[0424] 2. Image analysis:

[0425] The server receives image data and sends it to a generative artificial intelligence system to detect objects, people, and environmental elements within the image. This allows for the identification of potential hazards and points of caution.

[0426] 3. Emotion analysis:

[0427] The device's emotion engine analyzes the user's voice and facial expressions to identify their emotional state. When emotions such as anxiety or tension are detected, that information is sent to the server.

[0428] 4. Report generation:

[0429] The server generates a report based on the analysis results of the generative artificial intelligence and the emotion engine. The report includes specific areas of risk and countermeasures, as well as additional warnings and alerts tailored to the user's emotional state.

[0430] 5. Report distribution:

[0431] The server generates a report, which is then sent to the user's terminal for the user to view.

[0432] Specific example

[0433] Example 1: Image upload

[0434] Users take photos of equipment inside the factory and upload them to the server using a dedicated app.

[0435] Specific example 2: Image analysis

[0436] The server uses a generative artificial intelligence system to analyze the equipment photos it receives, identifying unstable parts and missing safety devices.

[0437] Example 3: Sentiment Analysis

[0438] The device uses an emotion engine to analyze the user's voice and facial expressions to detect their level of tension. For example, audio data of a user sighing while taking a photo can be analyzed.

[0439] Specific Example 4: Report Generation

[0440] The server generates a report stating, "An unstable area has been detected. Reinforcement is needed," and includes alerts tailored to the user's level of anxiety.

[0441] Specific example 5: Report distribution

[0442] The user receives the report and incorporates safety check items into the pre-work meeting based on it.

[0443] Examples of prompts to input into a generative AI model:

[0444] input_image: "factory_image.jpg"

[0445] input_emotion_data: "operator_audio.wav"

[0446] This system enables advanced safety checks that take into account the user's emotional state, thereby improving safety in on-site work.

[0447] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0448] Step 1:

[0449] Image capture and upload:

[0450] The user takes photos of the problem area at the site using a smartphone or tablet. The captured images are uploaded to the server via a dedicated app. Specifically, the user launches the camera app and captures a still image. After taking the photo, pressing the upload button transfers the photo data from a specified directory on the device to the server. At this time, an image file, for example "factory_image.jpg", is generated and uploaded.

[0451] Input: Captured photo data, user operation

[0452] Output: Image file sent to the server

[0453] Step 2:

[0454] Image reception and saving:

[0455] The server receives image data transmitted from the terminal via the internet. The received data is stored in a designated directory on the server. This ensures that the data necessary for the next analysis process is available on the server.

[0456] Input: Image file (factory_image.jpg)

[0457] Output: Image files stored on the server

[0458] Step 3:

[0459] Image analysis:

[0460] The server sends the stored image to a generative artificial intelligence (AI). The AI ​​detects and analyzes objects, people, and environmental elements within the image. Here, the machine learning model recognizes and identifies hazardous elements in the image. For example, it might point out unstable parts of scaffolding or missing safety devices.

[0461] Input: Saved image file

[0462] Output: Analysis results (data on hazardous areas and points to note)

[0463] Step 4:

[0464] Emotion analysis:

[0465] The device analyzes the user's voice and facial expressions during filming using an emotion engine to identify the user's emotional state. Based on information obtained from audio data and camera footage, it determines anxiety, tension, and other emotional states. For example, it analyzes the user's voice during filming, such as sighs or trembling voices, to assess the degree of anxiety.

[0466] Input: Voice data, facial expression data

[0467] Output: Emotional state data (anxiety, tension, etc.)

[0468] Step 5:

[0469] Report generation:

[0470] The server generates a report based on image analysis results from a generative artificial intelligence and emotional state data from an emotion engine. The report includes specific hazards detected, countermeasures, and additional warnings and alerts based on the user's emotional state. For example, it may include statements such as, "An unstable section of the scaffolding has been detected. Reinforcement is required," as well as warnings such as, "The operator is feeling anxious. Please re-check the situation."

[0471] Input: Image analysis results, emotional state data

[0472] Output: Generated report

[0473] Step 6:

[0474] Report distribution:

[0475] The server sends the generated report to the user's device. The device receives the report and displays it so the user can view it. For example, a dedicated app might receive a notification, and an interface might be presented within the app that allows the user to view the report.

[0476] Input: Generated report

[0477] Output: Report displayed on the user terminal

[0478] Step 7:

[0479] Reflection in TBM-KY:

[0480] Based on the report received on their device, users add safety confirmation items to the TBM-KY (Toolbox Meeting Hazard Prediction) record. The hazards are then shared with everyone at the actual meeting, and safety measures are implemented. For example, an item such as "scaffolding reinforcement is needed" might be discussed at the meeting, and specific countermeasures are communicated to everyone.

[0481] Input: Report content

[0482] Output: Updated TBM-KY records, implementation of safety measures.

[0483] This enables detailed safety checks that take into account the user's emotional state and provides rapid feedback, significantly improving safety in on-site work.

[0484] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0485] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0486] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0487] [Second Embodiment]

[0488] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0489] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0490] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0491] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0492] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0493] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0494] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0495] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0496] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0497] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0498] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0499] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0500] This invention relates to a safety verification system for on-site work, and provides a system in which a generative artificial intelligence identifies potential hazards and points of caution based on photographs taken by the user at the work site and generates a report.

[0501] System Configuration

[0502] This system consists of the following elements:

[0503] 1. Terminal

[0504] This is a device for users to take photos on-site.

[0505] It has a function to upload the photos taken to a server.

[0506] 2. Server

[0507] Receive photo data sent from the device.

[0508] Image analysis is performed using generative artificial intelligence.

[0509] A report is generated based on the analysis results.

[0510] Send the report to the terminal.

[0511] 3. Generative Artificial Intelligence

[0512] It is an artificial intelligence embedded in a server that detects objects, people, and environmental elements within images.

[0513] Identify potential hazards and points of caution.

[0514] Automatically generate reports.

[0515] Program processing

[0516] 1. Upload image

[0517] Users take photos of areas requiring safety checks using their smartphones or tablets at the site.

[0518] The device uploads the photos it takes to the server using a dedicated app.

[0519] 2. Image reception

[0520] The server receives and stores image data transmitted from terminals via the internet.

[0521] 3. Image Analysis

[0522] The server sends the stored images to the generative artificial intelligence.

[0523] Generative artificial intelligence detects and analyzes objects, people, and environmental elements within an image.

[0524] Based on the detected information, it is compared with pre-set safety standards to identify potential hazards and points of caution.

[0525] 4. Report generation

[0526] The server generates a report based on the analysis results from the generative artificial intelligence.

[0527] The report includes specific hazardous areas, proposed countermeasures, and images that visually represent these hazardous areas.

[0528] 5. Report distribution

[0529] The server sends the generated report to the terminal.

[0530] The terminal receives the report and displays it for the user to view.

[0531] 6. Reflection in TBM-KY

[0532] Based on the report content, the user adds safety check items to the TBM-KY record.

[0533] The report contents are shared in a meeting before on-site work begins, ensuring everyone is aware of the risks.

[0534] Specific example

[0535] Example 1: Image upload

[0536] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[0537] Specific example 2: Image analysis

[0538] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs to be reinforced."

[0539] Specific example 3: Report generation

[0540] The server generates a report based on the analysis results. It creates a report stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," and also inserts a photo of the dangerous areas circled in red as part of the analysis results.

[0541] Specific Example 4: Report Distribution and TBM-KY Reflection

[0542] The server generates a report and sends it to the user's terminal. The user views the report, adds "scaffolding reinforcement" as an item requiring it to TBM-KY, and notifies everyone.

[0543] This invention provides an effective system for field workers to quickly and accurately identify potential hazards at a work site, thereby improving the accuracy of safety checks and preventing accidents from occurring.

[0544] The following describes the processing flow.

[0545] Step 1:

[0546] Users take photos on-site. They use smartphones or tablets to take pictures of areas where safety checks are necessary.

[0547] Step 2:

[0548] The device saves the photos, and the app uploads them to the server. When the user selects photos taken using the dedicated app and presses the upload button, the app compresses the image files and sends them to the server via the internet.

[0549] Step 3:

[0550] The server receives the image. The server receives the uploaded image data via the internet and saves it to the specified directory.

[0551] Step 4:

[0552] The server sends the stored image to the generative artificial intelligence. The server then passes the received image data to the analysis module, which begins image analysis.

[0553] Step 5:

[0554] Generative artificial intelligence performs image analysis. The AI ​​detects objects, people, and environmental elements within the image and identifies potential hazards and points of caution based on these. For example, it can determine unstable areas of scaffolding or the proximity of power lines.

[0555] Step 6:

[0556] The server generates a report based on the analysis results. Upon receiving the analysis results from the generative artificial intelligence, the server automatically generates a report that includes the identified hazardous areas and suggested countermeasures. The report includes images showing the hazardous areas with red frames as part of the analysis results.

[0557] Step 7:

[0558] The server sends the generated report to the terminal. The server sends the generated report to the user's terminal and notifies the user of the report's arrival via push notification.

[0559] Step 8:

[0560] The device displays the report. The user opens a dedicated app and views the report sent from the server. The report includes specific hazardous areas, suggested countermeasures, and images visually illustrating the hazardous areas.

[0561] Step 9:

[0562] The user reflects the information in TBM-KY. Based on the report, the user adds safety check items to the toolbox meeting (TBM-KY) before on-site work and shares them with all on-site workers.

[0563] By following the steps outlined above, potential hazards at the site are identified objectively and quickly. Through sharing this information via TBM-KY (Track Byte-Marketing - Hazard Prediction), everyone becomes aware of the hazards and takes appropriate measures, thereby improving the accuracy of safety checks.

[0564] (Example 1)

[0565] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0566] In on-site work, it is essential to quickly and accurately identify potential hazards and efficiently conduct safety checks. However, currently, many sites rely on paper-based checklists and experience, which can lead to overlooking hazards and inadequate countermeasures. Furthermore, identifying individual hazardous areas and responding quickly becomes difficult, making it challenging to ensure overall site safety.

[0567] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0568] In this invention, the server includes means for uploading photos taken by the user on-site to the server via a terminal; means for the server to store the received image data in a database and transmit it to a generative artificial intelligence; means for the generative artificial intelligence to detect objects, people, and environmental elements in the image and identify potential hazards and points of caution based on this information by comparing it with safety standards; means for the server to generate a report based on the analysis results and transmit it to the terminal of a designated user; means for the terminal to receive the generated report and display it so that the user can view it; and means for the user to reflect safety confirmation items in a toolbox meeting based on the contents of the report. This makes it possible to quickly and accurately identify potential hazards in on-site work and to take prompt countermeasures against the identified hazardous areas.

[0569] A "user" is an individual or group engaged in on-site work who uses the system to perform safety checks.

[0570] A "terminal" refers to a smartphone, tablet, or other electronic device used by a user to upload photos taken on-site to a server.

[0571] A "server" is a computer system that receives photo data sent from a terminal, requests image analysis from a generative artificial intelligence, and generates a report based on that analysis.

[0572] "Generative artificial intelligence" refers to an artificial intelligence program that has algorithms to detect objects, people, and environmental elements within an image and identify potential dangers and points of caution.

[0573] "Image data" refers to photographic data taken by a user on their device and uploaded to the server.

[0574] A "report" is a document that includes specific hazardous areas and recommended countermeasures, based on the analysis results of a generative artificial intelligence.

[0575] A "toolbox meeting" is a meeting held before work begins where workers gather to check safety procedures and share information about the work to be done.

[0576] This invention relates to a safety verification system for on-site work, providing a system in which a generative artificial intelligence identifies potential hazards and points of caution based on photographs taken by the user at the worksite and generates a report. The components of this system are a terminal, a server, and a generative artificial intelligence.

[0577] System components

[0578] 1. Terminal

[0579] A terminal refers to a device such as a smartphone or tablet used by a user to record photos taken on-site.

[0580] The device requires the installation of a dedicated application. This dedicated application has functions for compressing, selecting, and uploading photos.

[0581] 2. Server

[0582] A server is a computer system that receives image data sent by users via the internet and stores it in a database.

[0583] The server has a generative artificial intelligence (AI) built in, and it sends the received image data to the AI.

[0584] The server generates a report based on the analysis results of the generative artificial intelligence and sends it to the user's terminal.

[0585] 3. Generative Artificial Intelligence

[0586] Generative artificial intelligence is an AI algorithm embedded in a server that has the function of detecting objects, people, and environmental elements within an image.

[0587] Generative artificial intelligence identifies potential hazards and points of caution by comparing them with safety standards and automatically generates reports.

[0588] System Operation Instructions

[0589] Users take photos of areas requiring safety checks on-site and upload the photo data to a server using a dedicated app. Upon receiving the image data, the server requests analysis from a generative artificial intelligence (AI). The AI ​​analyzes the image data and identifies potential hazards and points of caution. Based on the analysis results, the server generates a report and sends it back to the user's terminal. Users view the report and incorporate the findings into toolbox meetings.

[0590] Specific example

[0591] Example 1: Image upload

[0592] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[0593] Specific example 2: Image analysis

[0594] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs reinforcement."

[0595] Specific example 3: Report generation

[0596] The server generates a report based on the analysis results. The report includes a message stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," along with a photograph showing the dangerous areas circled in red as a result of the analysis.

[0597] Specific Example 4: Report Distribution and TBM-KY Reflection

[0598] The server generates a report and sends it to the user's terminal. The user views the report, adds "reinforcement of scaffolding" as an item requiring it to TBM-KY (Toolbox Meeting Keep and Yado), and notifies everyone.

[0599] This invention provides an effective system for field workers to quickly and accurately identify potential hazards at a work site, thereby improving the accuracy of safety checks and preventing accidents from occurring.

[0600] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0601] Step 1:

[0602] Users take photos of areas requiring safety checks using their smartphones or tablets at the site. After taking the photos, they select them using a dedicated app on their device and press the upload button. The device compresses the selected photos to prepare them for efficient use of network bandwidth. The input is the captured photo data, and the output is the compressed photo data.

[0603] Step 2:

[0604] The device sends compressed photo data to the server. The server receives the photo data via the internet and temporarily stores it in a database. The input is the compressed photo data, and the output is the image data stored in the database. The server returns a response to the device confirming that the image was saved successfully.

[0605] Step 3:

[0606] The server retrieves image data stored in the database and sends it to the generative artificial intelligence (AI). The input is the image data stored in the database, and the output is the data sent to the generative AI. Next, the generative AI receives the image and executes an object detection algorithm. Specifically, it identifies objects, people, and environmental elements in the image and returns this information to the server in an encoded format.

[0607] Step 4:

[0608] The generative artificial intelligence sends detection results from the image back to the server. This response includes data on identified objects, people, environmental elements, and potential hazardous areas. The input is image data to be analyzed, and the output is detection result data. Based on the received analysis data, the server compares it with established safety standards to identify hazardous areas and points of caution. The output is an interim report summarizing the hazardous areas and points of caution.

[0609] Step 5:

[0610] The server generates a report document using the analysis results of the generative artificial intelligence. The report includes details of detected hazardous areas, recommended countermeasures, and images highlighting the hazardous areas in red. The input is interim report data, and the output is a report document (in PDF or HTML format).

[0611] Step 6:

[0612] The server sends the generated report to the designated user's device. The device receives the report and notifies the user via a dedicated app. The input is the report document, and the output is the report notification to the user's device. The user views the report using the dedicated app to check for specific instructions and countermeasures.

[0613] Step 7:

[0614] Based on the generated report, the user adds the necessary safety check items to the TBM-KY (Toolbox Meeting Keep and Yado) record. The user shares the contents of this report with other workers during the pre-work meeting so that everyone is aware of the hazards. The input is the report content, and the output is the updated TBM-KY record.

[0615] (Application Example 1)

[0616] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0617] Traditional on-site safety check systems primarily rely on manual inspections, which suffers from the time and effort required to identify potential hazards. Furthermore, while many robots operate within factories, and there is a need to improve the efficiency of safety checks, the lack of robots themselves capable of performing safety checks of the work environment hinders progress in this area as well.

[0618] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0619] This invention includes a server comprising means for uploading photos taken by a user on-site to the server via a terminal, means for the server to transmit the received images to a generative artificial intelligence for analysis, means for the generative artificial intelligence to identify potential hazards and points of caution in the images and generate a report, means for the server to transmit the generated report to a terminal for the user to view, means for the user to reflect safety confirmation items based on the report in a pre-work toolbox meeting, means for a robot to take photos of the work area and upload those images to the server, means for the server to analyze images inside the production facility and identify potential hazardous areas, and means for the server to transmit the generated report to a work manager's terminal for viewing. This enables the rapid identification of potential hazardous areas and points of caution in on-site work and work areas within factories, improving the efficiency and accuracy of safety checks.

[0620] A "user" is a worker or manager who uses the safety verification system.

[0621] "The site" refers to factories, construction sites, and other workplaces.

[0622] A "device" is a device used by a user to take photos and upload them to a server, and includes smartphones, tablets, and other similar devices.

[0623] A "server" is a computer system that receives and analyzes image data sent by users and generates reports.

[0624] "Generative artificial intelligence" is an artificial intelligence technology that analyzes objects and environmental elements within an image to identify potential hazards and points of caution.

[0625] A "report" is a document containing information analyzed by a generative artificial intelligence system, including information on hazardous areas and countermeasures.

[0626] A "toolbox meeting" is a pre-work safety check meeting.

[0627] A "robot" is an automated work machine used in a factory, and is a device equipped with the image capture function of the present invention.

[0628] "Work area" refers to the area or section where robots and workers operate.

[0629] "Uploading" refers to the act of sending images taken with a device or robot to a server.

[0630] "Analysis" refers to the process by which generative artificial intelligence processes image data to identify potential hazards and points of caution.

[0631] A "manager" is a person responsible for safety management in a factory or work site.

[0632] A "production facility" refers to a place where products are produced, including factories and manufacturing sites.

[0633] "Potential hazard areas" are locations identified through image analysis where accidents or injuries are likely to occur.

[0634] "Visually representing images" are images that highlight dangerous areas or add supplementary explanations based on the analysis results.

[0635] This invention provides a system that streamlines safety checks in factories and on-sites, analyzing user-submitted photographs to identify potential hazards and generate reports. Specific embodiments of this system are described below.

[0636] First, the user takes photos of the work area and equipment using a device such as a smartphone or tablet. A dedicated application is installed on this device, and the captured images are uploaded to the server via this application. The uploaded image data is received and stored by the server. The stored images are then sent to a generative artificial intelligence.

[0637] The server analyzes the received images using generative artificial intelligence (AI). During this process, the AI ​​detects objects, people, and environmental elements within the image, and identifies potential hazards and points of caution based on pre-defined safety criteria. This generative AI utilizes libraries such as TensorFlow.

[0638] Next, the server automatically generates a report based on the analysis results from the generative artificial intelligence. This report contains information on identified hazardous areas and suggested countermeasures. It also includes images that visually represent the hazardous areas based on the analysis results. Libraries such as ReportLab can be used to create the report.

[0639] The generated report is sent from the server to the user's terminal, where the user can view it. Furthermore, based on this report, the user adds safety check items to the pre-work toolbox meeting and informs everyone.

[0640] Furthermore, a dedicated application can be installed on robots used within the factory, allowing them to take photos of the work area using their onboard cameras. The robots upload the captured images to a server, where they are similarly analyzed by generative artificial intelligence. The reports generated based on the analysis results are sent to the work supervisor's terminal, enabling them to review the information immediately.

[0641] Specific example

[0642] 1. Image Capture: Users take photos of specific areas within the factory using their smartphones. After taking the photos, they upload the images to the server using a dedicated application.

[0643] 2. Example of a prompt:

[0644] Analyze the images of this factory to identify potential hazards.

[0645] 3. Image Analysis: Generative artificial intelligence analyzes photographs to detect unstable scaffolding, unsecured equipment, and other issues.

[0646] 4. Report generation: The server generates a report stating, "Unstable scaffolding detected. Reinforcement is required," and includes an image with the unstable area highlighted in red.

[0647] 5. Report Distribution: The server generates reports and sends them to the user's smartphone for viewing. This allows for quick implementation of necessary safety measures before actual work begins.

[0648] The above describes the basic configuration for carrying out the present invention. This system makes on-site work and safety checks within factories more efficient and enables accident prevention.

[0649] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0650] Step 1:

[0651] The user takes photos of the work area and equipment using a device (smartphone or tablet). Specifically, this involves launching the application and capturing the desired scene in camera mode. In this case, the input is the captured image, and the output is the image file saved on the device.

[0652] Step 2:

[0653] The captured photos are uploaded to the server via a dedicated app. Specifically, pressing the "upload button" in the application sends the image data to the server. In this process, the input is the image file on the device, and the output is the image data transferred to the server.

[0654] Step 3:

[0655] The server saves the received image data and sends it to the generative artificial intelligence. Specifically, it saves the received image data as a temporary file and passes its path to the generative AI. In this case, the input is the image data sent from the terminal, and the output is the provision of the path to the image data to the generative AI.

[0656] Step 4:

[0657] Generative artificial intelligence analyzes image data to identify potential hazards and points of caution. Specifically, it applies object detection algorithms within images to identify hazardous areas. In this process, the input is image data transmitted from a server, and the output is hazardous area information as a result of the analysis.

[0658] Step 5:

[0659] The server generates a report based on the analysis results from the generative artificial intelligence. Specifically, it uses a report generation library to create a page that visually displays an overview of the analysis results and areas of concern. In this process, the input is the analysis result data from the generative artificial intelligence, and the output is the completed report file.

[0660] Step 6:

[0661] The server sends the generated report to the user's device, and the user views it. Specifically, the server sends the report file to the user's device via email or in-app notification, and the user receives the notification and opens the report. In this process, the input is the generated report file, and the output is the report displayed on the user's device.

[0662] Step 7:

[0663] Based on the report, the user incorporates safety check items into the pre-work toolbox meeting. Specifically, this involves reviewing the report, sharing the information in the meeting, and taking necessary safety measures. In this process, the input is the submitted report, and the output is the safety check information shared within the meeting.

[0664] Step 8:

[0665] The robot takes pictures of the work area and uploads the images to a server. Specifically, the robot's camera function is automatically activated, images are captured, and uploaded to the server via the network. In this process, the input is the image data captured by the robot, and the output is the image data transferred to the server.

[0666] Step 9:

[0667] The server analyzes images from inside the production facility to identify potential hazards. Specifically, it uses generative artificial intelligence to analyze objects and environmental elements within the work area and highlight hazards. In this process, the input is image data transmitted from the robot, and the output is information about the identified hazards.

[0668] Step 10:

[0669] The server sends the generated report to the work manager's terminal, ensuring that the manager can review it immediately. Specifically, the server delivers the report, including the analysis results, to the work manager's terminal via email or a dedicated application, allowing the manager to view the report content immediately. In this case, the input is the generated report, and the output is the report displayed on the work manager's terminal.

[0670] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0671] This invention combines a safety confirmation system for on-site work with an emotion engine that recognizes user emotions. In addition to a system that uses generative artificial intelligence to identify potential hazards and points of caution based on photos taken by the user at the worksite and generates a report, it also provides a function that performs safety confirmation while taking the user's emotional state into consideration.

[0672] System Configuration

[0673] This system consists of the following elements:

[0674] 1. Terminal

[0675] This is a device for users to take photos on-site.

[0676] It has a function to upload the photos taken to a server.

[0677] It implements an emotion engine and has the ability to analyze the user's voice and facial expressions.

[0678] 2. Server

[0679] Receive photo data sent from the device.

[0680] Image analysis is performed using generative artificial intelligence.

[0681] A report is generated based on the analysis results.

[0682] Send the report to the terminal.

[0683] 3. Generative Artificial Intelligence

[0684] It is an artificial intelligence embedded in a server that detects objects, people, and environmental elements within images.

[0685] Identify potential hazards and points of caution.

[0686] Automatically generate reports.

[0687] 4. Emotional Engine

[0688] The system analyzes the user's voice and facial expressions to identify their emotional state.

[0689] Depending on the emotional state, the generated reports will include additional warnings and alerts.

[0690] Program processing

[0691] 1. Upload image

[0692] Users take photos of areas requiring safety checks using their smartphones or tablets at the site.

[0693] The device uploads the photos it takes to the server using a dedicated app.

[0694] 2. Image reception

[0695] The server receives image data sent from the terminal via the internet and saves it to the specified directory.

[0696] 3. Image Analysis

[0697] The server sends the stored images to the generative artificial intelligence.

[0698] Generative artificial intelligence detects and analyzes objects, people, and environmental elements within an image.

[0699] Based on the detected information, it is compared with pre-set safety standards to identify potential hazards and points of caution.

[0700] 4. Emotion analysis

[0701] The device analyzes the user's voice and facial expressions during shooting using an emotion engine to identify the user's emotional state.

[0702] For example, if signs of anxiety or tension are detected, that information is sent to the server.

[0703] 5. Report generation

[0704] The server generates a report based on the analysis results from the generative artificial intelligence.

[0705] The report includes specific hazardous areas, proposed countermeasures, and images that visually represent these hazardous areas.

[0706] Based on the user's emotional state identified by the emotion engine, the report will include additional warnings and alerts as needed.

[0707] 6. Report distribution

[0708] The server sends the generated report to the terminal.

[0709] The terminal receives the report and displays it for the user to view.

[0710] 7. Reflection in TBM-KY

[0711] Based on the report content, the user adds safety check items to the TBM-KY record.

[0712] The report contents are shared in a meeting before on-site work begins, ensuring everyone is aware of the risks.

[0713] Specific example

[0714] Example 1: Image upload

[0715] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[0716] Specific example 2: Image analysis

[0717] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs to be reinforced."

[0718] Example 3: Sentiment Analysis

[0719] The device uses an emotion engine to analyze the user's voice and facial expressions while they are taking photos, and detects if the user is feeling anxious. This information is then sent to the server.

[0720] Specific Example 4: Report Generation

[0721] The server generates a report based on the analysis results. In addition to a report stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," it also includes a warning that takes into account the user's level of anxiety.

[0722] Specific Example 5: Report Distribution and TBM-KY Reflection

[0723] The server generates a report and sends it to the user's terminal. The user views the report, adds "scaffolding reinforcement" as an item requiring it to TBM-KY, and notifies everyone.

[0724] This invention aims to further improve safety by combining an emotion engine to enable safety checks that also take into account the emotional state of on-site workers.

[0725] The following describes the processing flow.

[0726] Step 1:

[0727] Users take photos on-site. They use smartphones or tablets to take pictures of areas where safety checks are necessary.

[0728] Step 2:

[0729] The device saves the photos, and the app uploads them to the server. When the user selects photos taken using the dedicated app and presses the upload button, the app compresses the image files and sends them to the server via the internet.

[0730] Step 3:

[0731] The device collects user voice and facial expression data. While taking or uploading photos, the device uses its built-in camera and microphone to collect user voice and facial expression data.

[0732] Step 4:

[0733] The device passes the collected voice and facial expression data to the emotion engine. The emotion engine analyzes the data and identifies the user's emotional state. For example, it can determine from the voice whether the user is nervous or not.

[0734] Step 5:

[0735] The server receives the image. The server receives the uploaded image data via the internet and saves it to the specified directory.

[0736] Step 6:

[0737] The server sends the stored image to the generative artificial intelligence. The server then passes the received image data to the analysis module, which begins image analysis.

[0738] Step 7:

[0739] Generative artificial intelligence performs image analysis. The AI ​​detects objects, people, and environmental elements within the image and identifies potential hazards and points of caution based on these. For example, it can determine unstable areas of scaffolding or the proximity of power lines.

[0740] Step 8:

[0741] The server generates a report based on the analysis results. Upon receiving the analysis results from the generative artificial intelligence, the server automatically generates a report that includes the identified hazardous areas and suggested countermeasures. The report includes images showing the hazardous areas with red frames as part of the analysis results.

[0742] Step 9:

[0743] The server receives emotion data from the emotion engine. If the user is in a stressed state, the server uses this information to include additional warnings and alerts in the report.

[0744] Step 10:

[0745] The server sends the generated report to the terminal. The server sends the generated report to the user's terminal and notifies the user of the report's arrival via push notification.

[0746] Step 11:

[0747] The device displays the report. The user opens a dedicated app and views the report sent from the server. The report includes specific hazardous areas, suggested countermeasures, and images visually representing the hazardous areas. It also includes additional warnings based on the user's emotional state.

[0748] Step 12:

[0749] The user reflects the information in TBM-KY. Based on the report, the user adds safety check items to the pre-work toolbox meeting (TBM-KY) and shares them with all field workers.

[0750] By following the steps outlined above, we can objectively and quickly identify potential hazards on-site and take appropriate measures that take into account the emotional state of users, thereby improving the accuracy of safety checks.

[0751] (Example 2)

[0752] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0753] Conventional on-site safety confirmation systems do not take into account the emotional state of the user, and therefore cannot adequately reflect the stress and anxiety felt by workers, potentially leading to overlooking potential hazards or implementing inappropriate countermeasures. Furthermore, these systems lack sufficient means of visually displaying analysis results, making it difficult for users to intuitively understand specific hazardous areas. Therefore, this invention aims to solve these problems and achieve more accurate safety confirmation and countermeasures while taking into account the emotional state of workers.

[0754] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0755] In this invention, the server includes means for uploading images taken by the user on-site to the server via a terminal, means for the server to save the received image data to a specified directory, means for the server to transmit the saved image data to a generative artificial intelligence for analysis, means for the generative artificial intelligence to detect objects, people, and environmental elements in the images and generate a report based on these to identify potential hazards and points of caution, means for the terminal to transmit the user's voice and facial expression data to an emotion analysis engine to identify the user's emotional state, means for the server to include additional warnings and alerts in the report based on the user's emotional state, means for the server to transmit the generated report to the terminal for the user to view it, and means for the user to reflect safety confirmation items based on the generated report in a pre-work meeting. This makes it possible to perform safety checks that take into account the user's emotional state, and by visually indicating specific hazardous areas, more intuitive and effective safety measures can be realized.

[0756] "Users" refer to workers and managers who take images on-site and use the system.

[0757] A "terminal" is a device used by a user that has the function of taking images on-site and sending the data to a server. Specific examples include smartphones and tablets.

[0758] A "server" refers to a computer system that processes data received from a terminal, works with generative artificial intelligence and emotion analysis engines to generate analysis results, and sends reports to the terminal.

[0759] "Generative artificial intelligence" refers to machine learning models or algorithms that analyze image data to detect objects, people, and environmental elements within the image, and identify potential dangers and points of caution.

[0760] An "emotion analysis engine" refers to an algorithm or technology that analyzes a user's voice and facial expression data to identify the user's emotional state.

[0761] "Image data" refers to photographs and videos taken by users on-site, which are uploaded to the server and used for analysis.

[0762] A "directory" refers to a specific folder or part of storage used to store data received within a server.

[0763] A "report" refers to a document that integrates the results of image analysis by generative artificial intelligence and sentiment analysis by a sentiment analysis engine, including potential risks, points of caution, and countermeasures.

[0764] A "meeting" refers to a planning meeting held by users before starting work, and is a forum for safety checks based on the contents of reports, such as TBM-KY (Tool Box Meeting - Kiken Yochi, Hazard Prediction Activity).

[0765] "Warning" refers to additional warnings or alerts included in reports based on the user's emotional state, providing information that encourages workers to pay close attention to specific hazards.

[0766] "Analysis results" refer to data provided by generative artificial intelligence and emotion analysis engines, including detected risk areas, countermeasures against them, and information on the user's emotional state.

[0767] This invention relates to a safety verification system for on-site work, and by combining it with a function to recognize the user's emotions, it achieves more effective safety measures. This system consists of the following hardware and software.

[0768] System Configuration

[0769] This system includes the following main components:

[0770] 1. Terminal

[0771] These are devices that users use to take photos and videos on-site. Specific examples include smartphones and tablets.

[0772] Install a dedicated application for uploading captured image data to the server.

[0773] It has the function of sending the user's voice and facial expression data to an emotion analysis engine to identify the user's emotions.

[0774] 2. Server

[0775] Saves image data received from the terminal to the specified directory.

[0776] Install and analyze software that operates as a generative artificial intelligence (for example, the OpenAI API).

[0777] The analysis results are integrated using software that acts as an emotion analysis engine (e.g., an emotion recognition API).

[0778] Generate a report and send it to the terminal.

[0779] 3. Generative Artificial Intelligence

[0780] This is a machine learning model embedded in the server that detects objects, people, and environmental elements within images.

[0781] Identify potential risks and points of caution, and automatically generate reports.

[0782] The software used will be the OpenAI API and similar analysis tools.

[0783] 4. Emotion Analysis Engine

[0784] Analyze audio and video data to identify the user's emotional state.

[0785] Based on emotional states, the generated reports will include additional warnings and alerts.

[0786] Program processing

[0787] The program for this system processes the information in the following order:

[0788] 1. Upload image

[0789] Users take pictures of areas requiring safety checks using their smartphones or tablets at the site.

[0790] The device uploads captured images to a server using a dedicated app. This app has the functionality to transmit image data over the internet.

[0791] 2. Image reception

[0792] The server receives image data sent from the terminal via the internet and saves it to a specified directory. Specifically, it saves files to cloud storage or local storage.

[0793] 3. Image Analysis

[0794] The server sends the stored image data to the generative artificial intelligence system. For example, a Python script on the server can be used to send the image data to the API.

[0795] Generative artificial intelligence analyzes received image data to detect objects, people, and environmental elements within the image. Specifically, it uses computer vision technology to identify unstable areas of scaffolding or lack of safety equipment.

[0796] 4. Emotion analysis

[0797] The device transmits the user's voice and facial expressions to an emotion analysis engine to identify the user's emotional state. This collects data using the smartphone's camera and microphone.

[0798] The terminal sends the analysis results to the server, where they are used as additional security management information.

[0799] 5. Report generation

[0800] The server generates a report based on analysis results from a generative artificial intelligence and the sentiment analysis engine. This report includes a detailed explanation of the hazardous areas, specific countermeasures, and warnings that reflect the user's emotional state.

[0801] The report will be generated in PDF format or another suitable format and saved to the specified directory.

[0802] 6. Report distribution

[0803] The server sends the generated report to the device. For example, it notifies the user of the report using email or push notifications via a dedicated app.

[0804] The terminal displays the received report, allowing the user to review it.

[0805] 7. Reflection in TBM-KY

[0806] Users conduct pre-work meetings based on reports during safety meetings (e.g., TBM-KY) to confirm hazardous areas and countermeasures.

[0807] For example, report contents can be shared on a smartphone screen, and necessary safety measures can be discussed.

[0808] Specific example

[0809] Example 1: Image upload

[0810] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[0811] Specific example 2: Image analysis

[0812] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs to be reinforced."

[0813] Example 3: Sentiment Analysis

[0814] The device uses an emotion analysis engine to analyze the user's voice and facial expressions while they are taking photos, and detects if the user is feeling anxious. This information is then sent to the server.

[0815] Specific Example 4: Report Generation

[0816] The server generates a report based on the analysis results. In addition to a report stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," it also includes a warning that takes into account the user's level of anxiety.

[0817] Specific Example 5: Report Distribution and TBM-KY Reflection

[0818] The server generates a report and sends it to the user's terminal. The user views the report, adds "scaffolding reinforcement" as an item requiring it to TBM-KY, and notifies everyone.

[0819] Example of a prompt:

[0820] "Identify potential hazards in this image and propose safety measures."

[0821] This enables safety checks for on-site work that take into account the user's emotional state, and by visually indicating specific hazardous areas, it allows for more intuitive and effective safety measures.

[0822] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0823] Step 1: Take and upload images

[0824] Users take pictures of areas requiring safety checks using their smartphones or tablets at the site.

[0825] Input: Image taken by the user.

[0826] Specific operation: The user takes pictures of the site using their smartphone's camera app and selects those images within the dedicated app.

[0827] The device uploads images to the server using a dedicated app. This app transmits image data via HTTPS over the internet.

[0828] Output: Image data uploaded to the server.

[0829] Step 2: Receiving and saving image data

[0830] The server receives image data sent from the terminal via the internet.

[0831] Input: Uploaded image data.

[0832] Specific operation: The server receives image data via a dedicated API and saves it to the directory " / images / uploads". The save location is recorded in the database.

[0833] Output: Image files saved in the specified directory.

[0834] Step 3: Perform image analysis

[0835] The server transmits the stored image data to the generative artificial intelligence system.

[0836] Input: The path to the image file saved in the directory.

[0837] Specific operation: A Python script on the server is used to send the image data path to the API.

[0838] Generative artificial intelligence analyzes image data to detect objects, people, and environmental elements.

[0839] Output: Analysis results (list of detected objects, people, and environmental elements).

[0840] Step 4: Identifying potential hazards

[0841] Generative artificial intelligence identifies potential risks and points of caution based on the analysis results.

[0842] Input: A list of detected objects, people, and environmental elements.

[0843] Specific operation: The generative artificial intelligence uses an analysis algorithm to compare the detected elements with safety standards and identify hazardous areas and points of caution.

[0844] Output: List of hazardous areas and points to note.

[0845] Step 5: Perform sentiment analysis

[0846] The device transmits the user's voice and facial expressions during shooting to an emotion analysis engine.

[0847] Input: User's voice data and facial expression data.

[0848] Specific operation: The device's microphone and camera are used to collect voice and facial expression data, which is then sent to an emotion analysis engine via a dedicated app.

[0849] The emotion analysis engine analyzes this data to identify the user's emotional state.

[0850] Output: User's emotional state (anxiety, tension, relief, etc.).

[0851] Step 6: Generate the report

[0852] The server generates a report based on the results of generative artificial intelligence and emotion analysis engines.

[0853] Input: List of hazards and points of caution, user's emotional state.

[0854] Specific operation: The server uses a web framework such as Django to integrate the analysis results and create a report. The report includes risk areas, countermeasures, and additional warnings based on the user's emotional state. The report is generated in PDF format, etc.

[0855] Output: The generated report file.

[0856] Step 7: Report Distribution

[0857] The server sends the generated report to the terminal.

[0858] Input: The generated report file.

[0859] Specific operation: The server uses a dedicated API to send report files to the terminal. For example, it might use email or push notifications.

[0860] The device displays received reports within a dedicated app, allowing users to review them.

[0861] Output: The report file delivered to the terminal.

[0862] Step 8: Reflection in TBM-KY

[0863] The user updates the TBM-KY record based on the generated report.

[0864] Input: Report file delivered to the terminal.

[0865] Specific actions: Users view reports in safety meetings and add necessary safety check items to the TBM-KY sheet. This is then shared with everyone, and specific safety measures are discussed.

[0866] Output: Updated TBM-KY record.

[0867] (Application Example 2)

[0868] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0869] Conventional safety confirmation systems for on-site work often failed to consider the user's psychological state, resulting in situations where appropriate warnings and responses could not be provided. Furthermore, because potential hazards and points of caution were identified solely through image analysis, safety measures based on the user's emotional changes were insufficient. This led to problems where optimal safety checks could not be performed, increasing the risk of workplace accidents.

[0870] In Application Example 2, the identification processing by the identification processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for uploading photos taken by the user at the site to the server via a terminal, means for the server to transmit the received images to a generative artificial intelligence for analysis, means for the generative artificial intelligence to identify potential dangers and points of caution in the images and generate a report, means for the server to transmit the generated report to the terminal for the user to view, means for the terminal to identify the user's emotional state using an emotion engine that analyzes the user's voice and facial expressions, and means for the emotion engine to transmit the analysis results to the server and include additional warnings and alerts in the report. This makes it possible to perform a comprehensive safety check that takes the user's emotions into consideration and reduces the risk of occupational accidents.

[0871] A "site work safety confirmation system" is a system for ensuring safety during on-site work by checking the conditions of the work site, identifying potential hazards and points of caution, and implementing safety measures.

[0872] A "terminal" is a device that allows users to take photos on-site and upload them to a server.

[0873] A "server" is a device that receives image data transmitted from a terminal, analyzes it using generative artificial intelligence, generates a report based on the analysis results, and then transmits that report back to the terminal.

[0874] "Generative artificial intelligence" is an artificial intelligence that analyzes transmitted image data, detects objects, people, and environmental elements within the image, and has the function of identifying potential dangers and points of caution.

[0875] A "report" is a document created based on the results of analysis by a generative artificial intelligence system, which describes potential hazards, points of caution, and countermeasures to address them.

[0876] An "emotion engine" is software or hardware that analyzes a user's voice and facial expressions to identify the user's emotional state.

[0877] A "toolbox meeting" is a pre-work meeting where workers gather to share information about the day's tasks, precautions, and safety checks.

[0878] "Safety confirmation items" are matters and procedures that must be checked to ensure safety during on-site work, and are shared at toolbox meetings.

[0879] The following describes an embodiment for carrying out this invention. This system allows users to upload photos taken with a smartphone or other device to a server for safety checks during on-site work, and a generative artificial intelligence analyzes the photo data. Furthermore, safety is further enhanced by analyzing the user's emotional state using an emotion engine and reflecting it in the report.

[0880] Hardware and software to use

[0881] Device: Use a device such as a smartphone or tablet that has the ability to take photos on-site and upload them to a server.

[0882] Server: This device receives image data, performs analysis using generative artificial intelligence, generates a report, and sends it to the terminal.

[0883] Generative artificial intelligence: This uses artificial intelligence implemented on a server to analyze image data and identify potential dangers and points of caution.

[0884] Emotion engine: This is software or hardware implemented in the device that analyzes the user's voice and facial expressions to identify their emotional state.

[0885] Data processing and data calculation

[0886] 1. Upload image:

[0887] The user takes a photo of the site with their device. The device then uploads this photo to the server.

[0888] 2. Image analysis:

[0889] The server receives image data and sends it to a generative artificial intelligence system to detect objects, people, and environmental elements within the image. This allows for the identification of potential hazards and points of caution.

[0890] 3. Emotion analysis:

[0891] The device's emotion engine analyzes the user's voice and facial expressions to identify their emotional state. When emotions such as anxiety or tension are detected, that information is sent to the server.

[0892] 4. Report generation:

[0893] The server generates a report based on the analysis results of the generative artificial intelligence and the emotion engine. The report includes specific areas of risk and countermeasures, as well as additional warnings and alerts tailored to the user's emotional state.

[0894] 5. Report distribution:

[0895] The server generates a report, which is then sent to the user's terminal for the user to view.

[0896] Specific example

[0897] Example 1: Image upload

[0898] Users take photos of equipment inside the factory and upload them to the server using a dedicated app.

[0899] Specific example 2: Image analysis

[0900] The server uses a generative artificial intelligence to analyze the photos of the equipment it receives, identifying unstable parts and missing safety devices.

[0901] Example 3: Sentiment Analysis

[0902] The device uses an emotion engine to analyze the user's voice and facial expressions to detect their level of tension. For example, audio data of a user sighing while taking a photo can be analyzed.

[0903] Specific Example 4: Report Generation

[0904] The server generates a report stating, "An unstable area has been detected. Reinforcement is needed," and includes alerts tailored to the user's level of anxiety.

[0905] Specific example 5: Report distribution

[0906] The user receives the report and incorporates safety check items into the pre-work meeting based on it.

[0907] Examples of prompts to input into a generative AI model:

[0908] input_image: "factory_image.jpg"

[0909] input_emotion_data: "operator_audio.wav"

[0910] This system enables advanced safety checks that take into account the user's emotional state, thereby improving safety in on-site work.

[0911] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0912] Step 1:

[0913] Image capture and upload:

[0914] The user takes photos of the problem area at the site using a smartphone or tablet. The captured images are uploaded to the server via a dedicated app. Specifically, the user launches the camera app and captures a still image. After taking the photo, pressing the upload button transfers the photo data from a specified directory on the device to the server. At this time, an image file, for example "factory_image.jpg", is generated and uploaded.

[0915] Input: Captured photo data, user operation

[0916] Output: Image file sent to the server

[0917] Step 2:

[0918] Image reception and saving:

[0919] The server receives image data transmitted from the terminal via the internet. The received data is stored in a designated directory on the server. This ensures that the data necessary for the next analysis process is available on the server.

[0920] Input: Image file (factory_image.jpg)

[0921] Output: Image files stored on the server

[0922] Step 3:

[0923] Image analysis:

[0924] The server sends the stored image to a generative artificial intelligence (AI). The AI ​​detects and analyzes objects, people, and environmental elements within the image. Here, the machine learning model recognizes and identifies hazardous elements in the image. For example, it might point out unstable parts of scaffolding or missing safety devices.

[0925] Input: Saved image file

[0926] Output: Analysis results (data on hazardous areas and points to note)

[0927] Step 4:

[0928] Emotion analysis:

[0929] The device analyzes the user's voice and facial expressions during filming using an emotion engine to identify the user's emotional state. Based on information obtained from audio data and camera footage, it determines anxiety, tension, and other emotional states. For example, it analyzes the user's voice during filming, such as sighs or trembling voices, to assess the degree of anxiety.

[0930] Input: Voice data, facial expression data

[0931] Output: Emotional state data (anxiety, tension, etc.)

[0932] Step 5:

[0933] Report generation:

[0934] The server generates a report based on image analysis results from a generative artificial intelligence and emotional state data from an emotion engine. The report includes specific hazards detected, countermeasures, and additional warnings and alerts based on the user's emotional state. For example, it may include statements such as, "An unstable section of the scaffolding has been detected. Reinforcement is required," as well as warnings such as, "The operator is feeling anxious. Please re-check the situation."

[0935] Input: Image analysis results, emotional state data

[0936] Output: Generated report

[0937] Step 6:

[0938] Report distribution:

[0939] The server sends the generated report to the user's device. The device receives the report and displays it so the user can view it. For example, a dedicated app might receive a notification, and an interface might be presented within the app that allows the user to view the report.

[0940] Input: Generated report

[0941] Output: Report displayed on the user terminal

[0942] Step 7:

[0943] Reflection in TBM-KY:

[0944] Based on the report received on their device, users add safety confirmation items to the TBM-KY (Toolbox Meeting Hazard Prediction) record. The hazards are then shared with everyone at the actual meeting, and safety measures are implemented. For example, an item such as "scaffolding reinforcement is needed" might be discussed at the meeting, and specific countermeasures are communicated to everyone.

[0945] Input: Report content

[0946] Output: Updated TBM-KY records, implementation of safety measures.

[0947] This enables detailed safety checks that take into account the user's emotional state and provides rapid feedback, significantly improving safety in on-site work.

[0948] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0949] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0950] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0951] [Third Embodiment]

[0952] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0953] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0954] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0955] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0956] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0957] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0958] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0959] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0960] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0961] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0962] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0963] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0964] This invention relates to a safety verification system for on-site work, and provides a system in which a generative artificial intelligence identifies potential hazards and points of caution based on photographs taken by the user at the work site and generates a report.

[0965] System Configuration

[0966] This system consists of the following elements:

[0967] 1. Terminal

[0968] This is a device for users to take photos on-site.

[0969] It has a function to upload the photos taken to a server.

[0970] 2. Server

[0971] Receive photo data sent from the device.

[0972] Image analysis is performed using generative artificial intelligence.

[0973] A report is generated based on the analysis results.

[0974] Send the report to the terminal.

[0975] 3. Generative Artificial Intelligence

[0976] It is an artificial intelligence embedded in a server that detects objects, people, and environmental elements within images.

[0977] Identify potential hazards and points of caution.

[0978] Automatically generate reports.

[0979] Program processing

[0980] 1. Upload image

[0981] Users take photos of areas requiring safety checks using their smartphones or tablets at the site.

[0982] The device uploads the photos it takes to the server using a dedicated app.

[0983] 2. Image reception

[0984] The server receives and stores image data transmitted from terminals via the internet.

[0985] 3. Image Analysis

[0986] The server sends the stored images to the generative artificial intelligence.

[0987] Generative artificial intelligence detects and analyzes objects, people, and environmental elements within an image.

[0988] Based on the detected information, it is compared with pre-set safety standards to identify potential hazards and points of caution.

[0989] 4. Report generation

[0990] The server generates a report based on the analysis results from the generative artificial intelligence.

[0991] The report includes specific hazardous areas, proposed countermeasures, and images that visually represent these hazardous areas.

[0992] 5. Report distribution

[0993] The server sends the generated report to the terminal.

[0994] The terminal receives the report and displays it for the user to view.

[0995] 6. Reflection in TBM-KY

[0996] Based on the report content, the user adds safety check items to the TBM-KY record.

[0997] The report contents are shared in a meeting before on-site work begins, ensuring everyone is aware of the risks.

[0998] Specific example

[0999] Example 1: Image upload

[1000] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[1001] Specific example 2: Image analysis

[1002] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs to be reinforced."

[1003] Specific example 3: Report generation

[1004] The server generates a report based on the analysis results. It creates a report stating, "Unstable areas of the scaffolding have been detected. Reinforcement is required," and also inserts a photo of the dangerous areas circled in red as part of the analysis results.

[1005] Specific Example 4: Report Distribution and TBM-KY Reflection

[1006] The server generates a report and sends it to the user's terminal. The user views the report, adds "scaffolding reinforcement" as an item requiring it to TBM-KY, and notifies everyone.

[1007] This invention provides an effective system for field workers to quickly and accurately identify potential hazards at a work site, thereby improving the accuracy of safety checks and preventing accidents from occurring.

[1008] The following describes the processing flow.

[1009] Step 1:

[1010] Users take photos on-site. They use smartphones or tablets to take pictures of areas where safety checks are necessary.

[1011] Step 2:

[1012] The device saves the photos, and the app uploads them to the server. When the user selects photos taken using the dedicated app and presses the upload button, the app compresses the image files and sends them to the server via the internet.

[1013] Step 3:

[1014] The server receives the image. The server receives the uploaded image data via the internet and saves it to the specified directory.

[1015] Step 4:

[1016] The server sends the stored image to the generative artificial intelligence. The server then passes the received image data to the analysis module, which begins image analysis.

[1017] Step 5:

[1018] Generative artificial intelligence performs image analysis. The AI ​​detects objects, people, and environmental elements within the image and identifies potential hazards and points of caution based on these. For example, it can determine unstable areas of scaffolding or the proximity of power lines.

[1019] Step 6:

[1020] The server generates a report based on the analysis results. Upon receiving the analysis results from the generative artificial intelligence, the server automatically generates a report that includes the identified hazardous areas and suggested countermeasures. The report includes images showing the hazardous areas with red frames as part of the analysis results.

[1021] Step 7:

[1022] The server sends the generated report to the terminal. The server sends the generated report to the user's terminal and notifies the user of the report's arrival via push notification.

[1023] Step 8:

[1024] The device displays the report. The user opens a dedicated app and views the report sent from the server. The report includes specific hazardous areas, suggested countermeasures, and images visually illustrating the hazardous areas.

[1025] Step 9:

[1026] The user reflects the information in TBM-KY. Based on the report, the user adds safety check items to the toolbox meeting (TBM-KY) before on-site work and shares them with all on-site workers.

[1027] By following the steps outlined above, potential hazards at the site are identified objectively and quickly. Through sharing this information via TBM-KY (Track Byte-Marketing - Hazard Prediction), everyone becomes aware of the hazards and takes appropriate measures, thereby improving the accuracy of safety checks.

[1028] (Example 1)

[1029] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1030] In on-site work, it is essential to quickly and accurately identify potential hazards and efficiently conduct safety checks. However, currently, many sites rely on paper-based checklists and experience, which can lead to overlooking hazards and inadequate countermeasures. Furthermore, identifying individual hazardous areas and responding quickly becomes difficult, making it challenging to ensure overall site safety.

[1031] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1032] In this invention, the server includes means for uploading photos taken by the user on-site to the server via a terminal; means for the server to store the received image data in a database and transmit it to a generative artificial intelligence; means for the generative artificial intelligence to detect objects, people, and environmental elements in the image and identify potential hazards and points of caution based on this information by comparing it with safety standards; means for the server to generate a report based on the analysis results and transmit it to the terminal of a designated user; means for the terminal to receive the generated report and display it so that the user can view it; and means for the user to reflect safety confirmation items in a toolbox meeting based on the contents of the report. This makes it possible to quickly and accurately identify potential hazards in on-site work and to take prompt countermeasures against the identified hazardous areas.

[1033] A "user" is an individual or group engaged in on-site work who uses the system to perform safety checks.

[1034] A "terminal" refers to a smartphone, tablet, or other electronic device used by a user to upload photos taken on-site to a server.

[1035] A "server" is a computer system that receives photo data sent from a terminal, requests image analysis from a generative artificial intelligence, and generates a report based on that analysis.

[1036] "Generative artificial intelligence" refers to an artificial intelligence program that has algorithms to detect objects, people, and environmental elements within an image and identify potential dangers and points of caution.

[1037] "Image data" refers to photographic data taken by a user on their device and uploaded to the server.

[1038] A "report" is a document that includes specific hazardous areas and recommended countermeasures, based on the analysis results of a generative artificial intelligence.

[1039] A "toolbox meeting" is a meeting held before work begins where workers gather to check safety procedures and share information about the work to be done.

[1040] This invention relates to a safety verification system for on-site work, providing a system in which a generative artificial intelligence identifies potential hazards and points of caution based on photographs taken by the user at the worksite and generates a report. The components of this system are a terminal, a server, and a generative artificial intelligence.

[1041] System components

[1042] 1. Terminal

[1043] A terminal refers to a device such as a smartphone or tablet used by a user to record photos taken on-site.

[1044] The device requires the installation of a dedicated application. This dedicated application has functions for compressing, selecting, and uploading photos.

[1045] 2. Server

[1046] A server is a computer system that receives image data sent by users via the internet and stores it in a database.

[1047] The server has a generative artificial intelligence (AI) built in, and it sends the received image data to the AI.

[1048] The server generates a report based on the analysis results of the generative artificial intelligence and sends it to the user's terminal.

[1049] 3. Generative Artificial Intelligence

[1050] Generative artificial intelligence is an AI algorithm embedded in a server that has the function of detecting objects, people, and environmental elements within an image.

[1051] Generative artificial intelligence identifies potential hazards and points of caution by comparing them with safety standards and automatically generates reports.

[1052] System Operation Instructions

[1053] Users take photos of areas requiring safety checks on-site and upload the photo data to a server using a dedicated app. Upon receiving the image data, the server requests analysis from a generative artificial intelligence (AI). The AI ​​analyzes the image data and identifies potential hazards and points of caution. Based on the analysis results, the server generates a report and sends it back to the user's terminal. Users view the report and incorporate the findings into toolbox meetings.

[1054] Specific example

[1055] Example 1: Image upload

[1056] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[1057] Specific example 2: Image analysis

[1058] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs reinforcement."

[1059] Specific example 3: Report generation

[1060] The server generates a report based on the analysis results. The report includes a message stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," along with a photograph showing the dangerous areas circled in red as a result of the analysis.

[1061] Specific Example 4: Report Distribution and TBM-KY Reflection

[1062] The server generates a report and sends it to the user's terminal. The user views the report, adds "reinforcement of scaffolding" as an item requiring it to TBM-KY (Toolbox Meeting Keep and Yado), and notifies everyone.

[1063] This invention provides an effective system for field workers to quickly and accurately identify potential hazards at a work site, thereby improving the accuracy of safety checks and preventing accidents from occurring.

[1064] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1065] Step 1:

[1066] Users take photos of areas requiring safety checks using their smartphones or tablets at the site. After taking the photos, they select them using a dedicated app on their device and press the upload button. The device compresses the selected photos to prepare them for efficient use of network bandwidth. The input is the captured photo data, and the output is the compressed photo data.

[1067] Step 2:

[1068] The device sends compressed photo data to the server. The server receives the photo data via the internet and temporarily stores it in a database. The input is the compressed photo data, and the output is the image data stored in the database. The server returns a response to the device confirming that the image was saved successfully.

[1069] Step 3:

[1070] The server retrieves image data stored in the database and sends it to the generative artificial intelligence (AI). The input is the image data stored in the database, and the output is the data sent to the generative AI. Next, the generative AI receives the image and executes an object detection algorithm. Specifically, it identifies objects, people, and environmental elements in the image and returns this information to the server in an encoded format.

[1071] Step 4:

[1072] The generative artificial intelligence sends detection results from the image back to the server. This response includes data on identified objects, people, environmental elements, and potential hazardous areas. The input is image data to be analyzed, and the output is detection result data. Based on the received analysis data, the server compares it with established safety standards to identify hazardous areas and points of caution. The output is an interim report summarizing the hazardous areas and points of caution.

[1073] Step 5:

[1074] The server generates a report document using the analysis results of the generative artificial intelligence. The report includes details of detected hazardous areas, recommended countermeasures, and images highlighting the hazardous areas in red. The input is interim report data, and the output is a report document (in PDF or HTML format).

[1075] Step 6:

[1076] The server sends the generated report to the designated user's device. The device receives the report and notifies the user via a dedicated app. The input is the report document, and the output is the report notification to the user's device. The user views the report using the dedicated app to check for specific instructions and countermeasures.

[1077] Step 7:

[1078] Based on the generated report, the user adds the necessary safety check items to the TBM-KY (Toolbox Meeting Keep and Yado) record. The user shares the contents of this report with other workers during the pre-work meeting so that everyone is aware of the hazards. The input is the report content, and the output is the updated TBM-KY record.

[1079] (Application Example 1)

[1080] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1081] Traditional on-site safety check systems primarily rely on manual inspections, which suffers from the time and effort required to identify potential hazards. Furthermore, while many robots operate within factories, and there is a need to improve the efficiency of safety checks, the lack of robots themselves capable of performing safety checks of the work environment hinders progress in this area as well.

[1082] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1083] This invention includes a server comprising means for uploading photos taken by a user on-site to the server via a terminal, means for the server to transmit the received images to a generative artificial intelligence for analysis, means for the generative artificial intelligence to identify potential hazards and points of caution in the images and generate a report, means for the server to transmit the generated report to a terminal for the user to view, means for the user to reflect safety confirmation items based on the report in a pre-work toolbox meeting, means for a robot to take photos of the work area and upload those images to the server, means for the server to analyze images inside the production facility and identify potential hazardous areas, and means for the server to transmit the generated report to a work manager's terminal for viewing. This enables the rapid identification of potential hazardous areas and points of caution in on-site work and work areas within factories, improving the efficiency and accuracy of safety checks.

[1084] A "user" refers to a worker or manager who uses the safety verification system.

[1085] "The site" refers to factories, construction sites, and other workplaces.

[1086] A "device" is a device used by a user to take photos and upload them to a server, and includes smartphones, tablets, and other similar devices.

[1087] A "server" is a computer system that receives and analyzes image data sent by users and generates reports.

[1088] "Generative artificial intelligence" is an artificial intelligence technology that analyzes objects and environmental elements within an image to identify potential hazards and points of caution.

[1089] A "report" is a document containing information analyzed by a generative artificial intelligence system, including information on hazardous areas and countermeasures.

[1090] A "toolbox meeting" is a pre-work safety check meeting.

[1091] A "robot" is an automated work machine used in a factory, and is a device equipped with the image capture function of the present invention.

[1092] "Work area" refers to the area or section where robots and workers operate.

[1093] "Uploading" refers to the act of sending images taken with a device or robot to a server.

[1094] "Analysis" refers to the process by which generative artificial intelligence processes image data to identify potential hazards and points of caution.

[1095] A "manager" is a person responsible for safety management in a factory or work site.

[1096] A "production facility" refers to a place where products are produced, including factories and manufacturing sites.

[1097] "Potential hazard areas" are locations identified through image analysis where accidents or injuries are likely to occur.

[1098] "Visually representing images" are images that highlight dangerous areas or add supplementary explanations based on the analysis results.

[1099] This invention provides a system that streamlines safety checks in factories and on-sites, analyzing user-submitted photographs to identify potential hazards and generate reports. Specific embodiments of this system are described below.

[1100] First, the user takes photos of the work area and equipment using a device such as a smartphone or tablet. A dedicated application is installed on this device, and the captured images are uploaded to the server via this application. The uploaded image data is received and stored by the server. The stored images are then sent to a generative artificial intelligence system.

[1101] The server analyzes the received images using generative artificial intelligence (AI). During this process, the AI ​​detects objects, people, and environmental elements within the image, and identifies potential hazards and points of caution based on pre-defined safety criteria. This generative AI utilizes libraries such as TensorFlow.

[1102] Next, the server automatically generates a report based on the analysis results from the generative artificial intelligence. This report contains information on identified hazardous areas and suggested countermeasures. It also includes images that visually represent the hazardous areas based on the analysis results. Libraries such as ReportLab can be used to create the report.

[1103] The generated report is sent from the server to the user's terminal, where the user can view it. Furthermore, based on this report, the user adds safety check items to the pre-work toolbox meeting and informs everyone.

[1104] Furthermore, a dedicated application can be installed on robots used within the factory, allowing them to take photos of the work area using their onboard cameras. The robots upload the captured images to a server, where they are similarly analyzed by generative artificial intelligence. The reports generated based on the analysis results are sent to the work supervisor's terminal, enabling them to review the information immediately.

[1105] Specific example

[1106] 1. Image capture: Users take photos of specific areas within the factory using their smartphones. After taking the photos, they upload the images to the server using a dedicated application.

[1107] 2. Example of a prompt:

[1108] Analyze the images of this factory to identify potential hazards.

[1109] 3. Image Analysis: Generative artificial intelligence analyzes photographs to detect unstable scaffolding, unsecured equipment, and other issues.

[1110] 4. Report generation: The server generates a report stating, "Unstable scaffolding detected. Reinforcement is required," and includes an image with the unstable area highlighted in red.

[1111] 5. Report Distribution: The server generates reports and sends them to the user's smartphone for viewing. This allows for quick implementation of necessary safety measures before actual work begins.

[1112] The above describes the basic configuration for carrying out the present invention. This system makes on-site work and safety checks within factories more efficient and enables accident prevention.

[1113] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1114] Step 1:

[1115] The user takes photos of the work area and equipment using a device (smartphone or tablet). Specifically, the user launches the application and captures the desired scene in camera mode. In this case, the input is the captured image, and the output is the image file saved on the device.

[1116] Step 2:

[1117] The captured photos are uploaded to the server via a dedicated app. Specifically, pressing the "upload button" in the application sends the image data to the server. In this process, the input is the image file on the device, and the output is the image data transferred to the server.

[1118] Step 3:

[1119] The server saves the received image data and sends it to the generative artificial intelligence. Specifically, it saves the received image data as a temporary file and passes its path to the generative AI. In this case, the input is the image data sent from the terminal, and the output is the provision of the path to the image data to the generative AI.

[1120] Step 4:

[1121] Generative artificial intelligence analyzes image data to identify potential hazards and points of caution. Specifically, it applies object detection algorithms within images to identify hazardous areas. In this process, the input is image data transmitted from a server, and the output is information about hazardous areas as a result of the analysis.

[1122] Step 5:

[1123] The server generates a report based on the analysis results from the generative artificial intelligence. Specifically, it uses a report generation library to create a page that visually displays an overview of the analysis results and areas of concern. In this process, the input is the analysis result data from the generative artificial intelligence, and the output is the completed report file.

[1124] Step 6:

[1125] The server sends the generated report to the user's device, and the user views it. Specifically, the server sends the report file to the user's device via email or in-app notification, and the user receives the notification and opens the report. In this process, the input is the generated report file, and the output is the report displayed on the user's device.

[1126] Step 7:

[1127] Based on the report, the user incorporates safety check items into the pre-work toolbox meeting. Specifically, this involves reviewing the report, sharing the information in the meeting, and taking necessary safety measures. In this process, the input is the submitted report, and the output is the safety check information shared within the meeting.

[1128] Step 8:

[1129] The robot takes pictures of the work area and uploads the images to a server. Specifically, the robot's camera function is automatically activated, images are captured, and uploaded to the server via the network. In this process, the input is the image data captured by the robot, and the output is the image data transferred to the server.

[1130] Step 9:

[1131] The server analyzes images from inside the production facility to identify potential hazards. Specifically, it uses generative artificial intelligence to analyze objects and environmental elements within the work area and highlight hazards. In this process, the input is image data transmitted from the robot, and the output is information about the identified hazards.

[1132] Step 10:

[1133] The server sends the generated report to the work manager's terminal, ensuring that the manager can review it immediately. Specifically, the server delivers the report, including the analysis results, to the work manager's terminal via email or a dedicated application, allowing the manager to view the report content immediately. In this case, the input is the generated report, and the output is the report displayed on the work manager's terminal.

[1134] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1135] This invention combines a safety confirmation system for on-site work with an emotion engine that recognizes user emotions. In addition to a system that uses generative artificial intelligence to identify potential hazards and points of caution based on photos taken by the user at the worksite and generates a report, it also provides a function that performs safety confirmation while taking the user's emotional state into consideration.

[1136] System Configuration

[1137] This system consists of the following elements:

[1138] 1. Terminal

[1139] This is a device for users to take photos on-site.

[1140] It has a function to upload the photos taken to a server.

[1141] It implements an emotion engine and has the ability to analyze the user's voice and facial expressions.

[1142] 2. Server

[1143] Receive photo data sent from the device.

[1144] Image analysis is performed using generative artificial intelligence.

[1145] A report is generated based on the analysis results.

[1146] Send the report to the terminal.

[1147] 3. Generative Artificial Intelligence

[1148] It is an artificial intelligence embedded in a server that detects objects, people, and environmental elements within images.

[1149] Identify potential hazards and points of caution.

[1150] Automatically generate reports.

[1151] 4. Emotional Engine

[1152] The system analyzes the user's voice and facial expressions to identify their emotional state.

[1153] Depending on the emotional state, the generated reports will include additional warnings and alerts.

[1154] Program processing

[1155] 1. Upload image

[1156] Users take photos of areas requiring safety checks using their smartphones or tablets at the site.

[1157] The device uploads the photos it takes to the server using a dedicated app.

[1158] 2. Image reception

[1159] The server receives image data sent from the terminal via the internet and saves it to the specified directory.

[1160] 3. Image Analysis

[1161] The server sends the stored images to the generative artificial intelligence.

[1162] Generative artificial intelligence detects and analyzes objects, people, and environmental elements within an image.

[1163] Based on the detected information, it is compared with pre-set safety standards to identify potential hazards and points of caution.

[1164] 4. Emotion analysis

[1165] The device analyzes the user's voice and facial expressions during shooting using an emotion engine to identify the user's emotional state.

[1166] For example, if signs of anxiety or tension are detected, that information is sent to the server.

[1167] 5. Report generation

[1168] The server generates a report based on the analysis results from the generative artificial intelligence.

[1169] The report includes specific hazardous areas, proposed countermeasures, and images that visually represent these hazardous areas.

[1170] Based on the user's emotional state identified by the emotion engine, the report will include additional warnings and alerts as needed.

[1171] 6. Report distribution

[1172] The server sends the generated report to the terminal.

[1173] The terminal receives the report and displays it for the user to view.

[1174] 7. Reflection in TBM-KY

[1175] Based on the report content, the user adds safety check items to the TBM-KY record.

[1176] The report contents are shared in a meeting before on-site work begins, ensuring everyone is aware of the risks.

[1177] Specific example

[1178] Example 1: Image upload

[1179] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[1180] Specific example 2: Image analysis

[1181] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs to be reinforced."

[1182] Example 3: Sentiment Analysis

[1183] The device uses an emotion engine to analyze the user's voice and facial expressions while they are taking photos, and detects if the user is feeling anxious. This information is then sent to the server.

[1184] Specific Example 4: Report Generation

[1185] The server generates a report based on the analysis results. In addition to a report stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," it also includes a warning that takes into account the user's level of anxiety.

[1186] Specific Example 5: Report Distribution and TBM-KY Reflection

[1187] The server generates a report and sends it to the user's terminal. The user views the report, adds "scaffolding reinforcement" as an item requiring it to TBM-KY, and notifies everyone.

[1188] This invention aims to further improve safety by combining an emotion engine to enable safety checks that also take into account the emotional state of on-site workers.

[1189] The following describes the processing flow.

[1190] Step 1:

[1191] Users take photos on-site. They use smartphones or tablets to take pictures of areas where safety checks are necessary.

[1192] Step 2:

[1193] The device saves the photos, and the app uploads them to the server. When the user selects photos taken using the dedicated app and presses the upload button, the app compresses the image files and sends them to the server via the internet.

[1194] Step 3:

[1195] The device collects user voice and facial expression data. While taking or uploading photos, the device uses its built-in camera and microphone to collect user voice and facial expression data.

[1196] Step 4:

[1197] The device passes the collected voice and facial expression data to the emotion engine. The emotion engine analyzes the data and identifies the user's emotional state. For example, it can determine from the voice whether the user is nervous or not.

[1198] Step 5:

[1199] The server receives the image. The server receives the uploaded image data via the internet and saves it to the specified directory.

[1200] Step 6:

[1201] The server sends the stored image to the generative artificial intelligence. The server then passes the received image data to the analysis module, which begins image analysis.

[1202] Step 7:

[1203] Generative artificial intelligence performs image analysis. The AI ​​detects objects, people, and environmental elements within the image and identifies potential hazards and points of caution based on these. For example, it can determine unstable areas of scaffolding or the proximity of power lines.

[1204] Step 8:

[1205] The server generates a report based on the analysis results. Upon receiving the analysis results from the generative artificial intelligence, the server automatically generates a report that includes the identified hazardous areas and suggested countermeasures. The report includes images showing the hazardous areas with red frames as part of the analysis results.

[1206] Step 9:

[1207] The server receives emotion data from the emotion engine. If the user is in a stressed state, the server uses this information to include additional warnings and alerts in the report.

[1208] Step 10:

[1209] The server sends the generated report to the terminal. The server sends the generated report to the user's terminal and notifies the user of the report's arrival via push notification.

[1210] Step 11:

[1211] The device displays the report. The user opens a dedicated app and views the report sent from the server. The report includes specific hazardous areas, suggested countermeasures, and images visually representing the hazardous areas. It also includes additional warnings based on the user's emotional state.

[1212] Step 12:

[1213] The user reflects the information in TBM-KY. Based on the report, the user adds safety check items to the toolbox meeting (TBM-KY) before on-site work and shares them with all on-site workers.

[1214] By following the steps outlined above, we can objectively and quickly identify potential hazards on-site and take appropriate measures that take into account the emotional state of users, thereby improving the accuracy of safety checks.

[1215] (Example 2)

[1216] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1217] Conventional on-site safety confirmation systems do not take into account the emotional state of the user, and therefore cannot adequately reflect the stress and anxiety felt by workers, potentially leading to overlooking potential hazards or implementing inappropriate countermeasures. Furthermore, these systems lack sufficient means of visually displaying analysis results, making it difficult for users to intuitively understand specific hazardous areas. Therefore, this invention aims to solve these problems and achieve more accurate safety confirmation and countermeasures while taking into account the emotional state of workers.

[1218] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1219] In this invention, the server includes means for uploading images taken by the user on-site to the server via a terminal, means for the server to save the received image data to a specified directory, means for the server to transmit the saved image data to a generative artificial intelligence for analysis, means for the generative artificial intelligence to detect objects, people, and environmental elements in the images and generate a report based on these to identify potential hazards and points of caution, means for the terminal to transmit the user's voice and facial expression data to an emotion analysis engine to identify the user's emotional state, means for the server to include additional warnings and alerts in the report based on the user's emotional state, means for the server to transmit the generated report to the terminal for the user to view it, and means for the user to reflect safety confirmation items based on the generated report in a pre-work meeting. This makes it possible to perform safety checks that take into account the user's emotional state, and by visually indicating specific hazardous areas, more intuitive and effective safety measures can be realized.

[1220] "Users" refer to workers and managers who take images on-site and use the system.

[1221] A "terminal" is a device used by a user that has the function of taking images on-site and sending the data to a server. Specific examples include smartphones and tablets.

[1222] A "server" refers to a computer system that processes data received from a terminal, works with generative artificial intelligence and emotion analysis engines to generate analysis results, and sends reports to the terminal.

[1223] "Generative artificial intelligence" refers to machine learning models or algorithms that analyze image data to detect objects, people, and environmental elements within the image, and to identify potential dangers and points of caution.

[1224] An "emotion analysis engine" refers to an algorithm or technology that analyzes a user's voice and facial expression data to identify the user's emotional state.

[1225] "Image data" refers to photographs and videos taken by users on-site, which are uploaded to the server and used for analysis.

[1226] A "directory" refers to a specific folder or part of storage used to store data received within a server.

[1227] A "report" refers to a document that integrates the results of image analysis by generative artificial intelligence and sentiment analysis by a sentiment analysis engine, including potential risks, points of caution, and countermeasures.

[1228] A "meeting" refers to a planning meeting held by users before starting work, where safety checks are conducted based on the contents of reports, such as TBM-KY (Tool Box Meeting - Kiken Yochi, Hazard Prediction Activity).

[1229] "Warning" refers to additional warnings or alerts included in reports based on the user's emotional state, providing information that encourages workers to pay close attention to specific hazards.

[1230] "Analysis results" refer to data provided by generative artificial intelligence and emotion analysis engines, including detected risk areas, countermeasures against them, and information on the user's emotional state.

[1231] This invention relates to a safety verification system for on-site work, and by combining it with a function to recognize the user's emotions, it achieves more effective safety measures. This system consists of the following hardware and software.

[1232] System Configuration

[1233] This system includes the following main components:

[1234] 1. Terminal

[1235] These are devices that users use to take photos and videos on-site. Specific examples include smartphones and tablets.

[1236] Install a dedicated application for uploading captured image data to the server.

[1237] It has the function of sending the user's voice and facial expression data to an emotion analysis engine to identify the user's emotions.

[1238] 2. Server

[1239] Saves image data received from the terminal to the specified directory.

[1240] Install and analyze software that operates as a generative artificial intelligence (for example, the OpenAI API).

[1241] The analysis results are integrated using software that acts as an emotion analysis engine (e.g., an emotion recognition API).

[1242] Generate a report and send it to the terminal.

[1243] 3. Generative Artificial Intelligence

[1244] This is a machine learning model embedded in the server that detects objects, people, and environmental elements within images.

[1245] Identify potential risks and points of caution, and automatically generate reports.

[1246] The software used will be the OpenAI API and similar analysis tools.

[1247] 4. Emotion Analysis Engine

[1248] Analyze audio and video data to identify the user's emotional state.

[1249] Based on emotional states, the generated reports will include additional warnings and alerts.

[1250] Program processing

[1251] The program for this system processes the information in the following order:

[1252] 1. Upload image

[1253] Users take pictures of areas requiring safety checks using their smartphones or tablets at the site.

[1254] The device uploads captured images to a server using a dedicated app. This app has the functionality to transmit image data over the internet.

[1255] 2. Image reception

[1256] The server receives image data sent from the terminal via the internet and saves it to a specified directory. Specifically, it saves files to cloud storage or local storage.

[1257] 3. Image Analysis

[1258] The server sends the stored image data to a generative artificial intelligence system. For example, a Python script on the server can be used to send the image data to an API.

[1259] Generative artificial intelligence analyzes received image data to detect objects, people, and environmental elements within the image. Specifically, it uses computer vision technology to identify unstable areas of scaffolding or lack of safety equipment.

[1260] 4. Emotion analysis

[1261] The device transmits the user's voice and facial expressions to an emotion analysis engine to identify the user's emotional state. This collects data using the smartphone's camera and microphone.

[1262] The terminal sends the analysis results to the server, which is then used as additional security management information.

[1263] 5. Report generation

[1264] The server generates a report based on analysis results from a generative artificial intelligence and the sentiment analysis engine. This report includes detailed descriptions of hazardous areas, specific countermeasures, and warnings that reflect the user's emotional state.

[1265] The report will be generated in PDF format or another suitable format and saved to the specified directory.

[1266] 6. Report distribution

[1267] The server sends the generated report to the device. For example, it notifies the user of the report using email or push notifications via a dedicated app.

[1268] The terminal displays the received report, allowing the user to review it.

[1269] 7. Reflection in TBM-KY

[1270] Users conduct pre-work meetings based on reports during safety meetings (e.g., TBM-KY) to confirm hazardous areas and countermeasures.

[1271] For example, report contents can be shared on a smartphone screen, and necessary safety measures can be discussed.

[1272] Specific example

[1273] Example 1: Image upload

[1274] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[1275] Specific example 2: Image analysis

[1276] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs to be reinforced."

[1277] Example 3: Sentiment Analysis

[1278] The device uses an emotion analysis engine to analyze the user's voice and facial expressions while they are taking photos, and detects if the user is feeling anxious. This information is then sent to the server.

[1279] Specific Example 4: Report Generation

[1280] The server generates a report based on the analysis results. In addition to a report stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," it also includes a warning that takes into account the user's level of anxiety.

[1281] Specific Example 5: Report Distribution and TBM-KY Reflection

[1282] The server generates a report and sends it to the user's terminal. The user views the report, adds "scaffolding reinforcement" as an item requiring it to TBM-KY, and notifies everyone.

[1283] Example of a prompt:

[1284] "Identify potential hazards in this image and propose safety measures."

[1285] This enables safety checks for on-site work that take into account the user's emotional state, and by visually indicating specific hazardous areas, it allows for more intuitive and effective safety measures.

[1286] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1287] Step 1: Take and upload images

[1288] Users take pictures of areas requiring safety checks using their smartphones or tablets at the site.

[1289] Input: Image taken by the user.

[1290] Specific operation: The user takes pictures of the site using their smartphone's camera app and selects those images within the dedicated app.

[1291] The device uploads images to the server using a dedicated app. This app transmits image data via HTTPS over the internet.

[1292] Output: Image data uploaded to the server.

[1293] Step 2: Receiving and saving image data

[1294] The server receives image data sent from the terminal via the internet.

[1295] Input: Uploaded image data.

[1296] Specific operation: The server receives image data via a dedicated API and saves it to the directory " / images / uploads". The save location is recorded in the database.

[1297] Output: Image files saved in the specified directory.

[1298] Step 3: Perform image analysis

[1299] The server transmits the stored image data to the generative artificial intelligence system.

[1300] Input: The path to the image file saved in the directory.

[1301] Specific operation: A Python script on the server is used to send the image data path to the API.

[1302] Generative artificial intelligence analyzes image data to detect objects, people, and environmental elements.

[1303] Output: Analysis results (list of detected objects, people, and environmental elements).

[1304] Step 4: Identifying potential hazards

[1305] Generative artificial intelligence identifies potential risks and points of caution based on the analysis results.

[1306] Input: A list of detected objects, people, and environmental elements.

[1307] Specific operation: The generative artificial intelligence uses an analysis algorithm to compare the detected elements with safety standards and identify hazardous areas and points of caution.

[1308] Output: List of hazardous areas and points to note.

[1309] Step 5: Perform sentiment analysis

[1310] The device transmits the user's voice and facial expressions during shooting to an emotion analysis engine.

[1311] Input: User's voice data and facial expression data.

[1312] Specific operation: The device's microphone and camera are used to collect voice and facial expression data, which is then sent to an emotion analysis engine via a dedicated app.

[1313] The emotion analysis engine analyzes this data to identify the user's emotional state.

[1314] Output: User's emotional state (anxiety, tension, relief, etc.).

[1315] Step 6: Generate the report

[1316] The server generates a report based on the results of generative artificial intelligence and emotion analysis engines.

[1317] Input: List of hazards and points of caution, user's emotional state.

[1318] Specific operation: The server uses a web framework such as Django to integrate the analysis results and create a report. The report includes risk areas, countermeasures, and additional warnings based on the user's emotional state. The report is generated in PDF format, etc.

[1319] Output: The generated report file.

[1320] Step 7: Report Distribution

[1321] The server sends the generated report to the terminal.

[1322] Input: The generated report file.

[1323] Specific operation: The server uses a dedicated API to send report files to the terminal. For example, it might use email or push notifications.

[1324] The device displays received reports within a dedicated app, allowing users to review them.

[1325] Output: The report file delivered to the terminal.

[1326] Step 8: Reflection in TBM-KY

[1327] The user updates the TBM-KY record based on the generated report.

[1328] Input: Report file delivered to the terminal.

[1329] Specific actions: Users view reports in safety meetings and add necessary safety check items to the TBM-KY sheet. This is then shared with everyone, and specific safety measures are discussed.

[1330] Output: Updated TBM-KY record.

[1331] (Application Example 2)

[1332] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1333] Conventional safety confirmation systems for on-site work often failed to consider the user's psychological state, resulting in situations where appropriate warnings and responses could not be provided. Furthermore, because potential hazards and points of caution were identified solely through image analysis, safety measures based on the user's emotional changes were insufficient. This led to problems where optimal safety checks could not be performed, increasing the risk of workplace accidents.

[1334] In Application Example 2, the identification processing by the identification processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for uploading photos taken by the user at the site to the server via a terminal, means for the server to transmit the received images to a generative artificial intelligence for analysis, means for the generative artificial intelligence to identify potential dangers and points of caution in the images and generate a report, means for the server to transmit the generated report to the terminal for the user to view, means for the terminal to identify the user's emotional state using an emotion engine that analyzes the user's voice and facial expressions, and means for the emotion engine to transmit the analysis results to the server and include additional warnings and alerts in the report. This makes it possible to perform a comprehensive safety check that takes the user's emotions into consideration and reduces the risk of occupational accidents.

[1335] A "site work safety confirmation system" is a system for ensuring safety during on-site work by checking the conditions of the work site, identifying potential hazards and points of caution, and implementing safety measures.

[1336] A "terminal" is a device that allows users to take photos on-site and upload them to a server.

[1337] A "server" is a device that receives image data transmitted from a terminal, analyzes it using generative artificial intelligence, generates a report based on the analysis results, and then transmits that report back to the terminal.

[1338] "Generative artificial intelligence" is an artificial intelligence that analyzes transmitted image data, detects objects, people, and environmental elements within the image, and has the function of identifying potential dangers and points of caution.

[1339] A "report" is a document created based on the results of analysis by a generative artificial intelligence system, which describes potential hazards, points of caution, and countermeasures to address them.

[1340] An "emotion engine" is software or hardware that analyzes a user's voice and facial expressions to identify the user's emotional state.

[1341] A "toolbox meeting" is a pre-work meeting where workers gather to share information about the day's tasks, precautions, and safety checks.

[1342] "Safety confirmation items" are matters and procedures that must be checked to ensure safety during on-site work, and are shared at toolbox meetings.

[1343] The following describes an embodiment for carrying out this invention. This system allows users to upload photos taken with a smartphone or other device to a server for safety checks during on-site work, and a generative artificial intelligence analyzes the photo data. Furthermore, safety is further enhanced by analyzing the user's emotional state using an emotion engine and reflecting it in the report.

[1344] Hardware and software to use

[1345] Device: Use a device such as a smartphone or tablet that has the ability to take photos on-site and upload them to a server.

[1346] Server: This device receives image data, performs analysis using generative artificial intelligence, generates a report, and sends it to the terminal.

[1347] Generative artificial intelligence: This uses artificial intelligence implemented on a server to analyze image data and identify potential dangers and points of caution.

[1348] Emotion engine: This is software or hardware implemented in the device that analyzes the user's voice and facial expressions to identify their emotional state.

[1349] Data processing and data calculation

[1350] 1. Upload image:

[1351] The user takes a photo of the site with their device. The device then uploads this photo to the server.

[1352] 2. Image analysis:

[1353] The server receives image data and sends it to a generative artificial intelligence system to detect objects, people, and environmental elements within the image. This allows for the identification of potential hazards and points of caution.

[1354] 3. Emotion analysis:

[1355] The device's emotion engine analyzes the user's voice and facial expressions to identify their emotional state. When emotions such as anxiety or tension are detected, that information is sent to the server.

[1356] 4. Report generation:

[1357] The server generates a report based on the analysis results of the generative artificial intelligence and the emotion engine. The report includes specific areas of risk and countermeasures, as well as additional warnings and alerts tailored to the user's emotional state.

[1358] 5. Report distribution:

[1359] The server generates a report, which is then sent to the user's terminal for the user to view.

[1360] Specific example

[1361] Example 1: Image upload

[1362] Users take photos of equipment inside the factory and upload them to the server using a dedicated app.

[1363] Specific example 2: Image analysis

[1364] The server uses a generative artificial intelligence to analyze the photos of the equipment it receives, identifying unstable parts and missing safety devices.

[1365] Example 3: Sentiment Analysis

[1366] The device uses an emotion engine to analyze the user's voice and facial expressions to detect their level of tension. For example, audio data of a user sighing while taking a photo can be analyzed.

[1367] Specific Example 4: Report Generation

[1368] The server generates a report stating, "An unstable area has been detected. Reinforcement is needed," and includes alerts tailored to the user's level of anxiety.

[1369] Specific example 5: Report distribution

[1370] The user receives the report and incorporates safety check items into the pre-work meeting based on it.

[1371] Examples of prompts to input into a generative AI model:

[1372] input_image: "factory_image.jpg"

[1373] input_emotion_data: "operator_audio.wav"

[1374] This system enables advanced safety checks that take into account the user's emotional state, thereby improving safety in on-site work.

[1375] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1376] Step 1:

[1377] Image capture and upload:

[1378] The user takes photos of the problem area at the site using a smartphone or tablet. The captured images are uploaded to the server via a dedicated app. Specifically, the user launches the camera app and captures a still image. After taking the photo, pressing the upload button transfers the photo data from a specified directory on the device to the server. At this time, an image file, for example "factory_image.jpg", is generated and uploaded.

[1379] Input: Captured photo data, user operation

[1380] Output: Image file sent to the server

[1381] Step 2:

[1382] Image reception and saving:

[1383] The server receives image data transmitted from the terminal via the internet. The received data is stored in a designated directory on the server. This ensures that the data necessary for the next analysis process is available on the server.

[1384] Input: Image file (factory_image.jpg)

[1385] Output: Image files stored on the server

[1386] Step 3:

[1387] Image analysis:

[1388] The server sends the stored image to a generative artificial intelligence (AI). The AI ​​detects and analyzes objects, people, and environmental elements within the image. Here, the machine learning model recognizes and identifies hazardous elements in the image. For example, it might point out unstable parts of scaffolding or missing safety devices.

[1389] Input: Saved image file

[1390] Output: Analysis results (data on hazardous areas and points to note)

[1391] Step 4:

[1392] Emotion analysis:

[1393] The device analyzes the user's voice and facial expressions during filming using an emotion engine to identify the user's emotional state. Based on information obtained from audio data and camera footage, it determines anxiety, tension, and other emotional states. For example, it analyzes the user's voice during filming, such as sighs or trembling voices, to assess the degree of anxiety.

[1394] Input: Voice data, facial expression data

[1395] Output: Emotional state data (anxiety, tension, etc.)

[1396] Step 5:

[1397] Report generation:

[1398] The server generates a report based on image analysis results from a generative artificial intelligence and emotional state data from an emotion engine. The report includes specific hazards detected, countermeasures, and additional warnings and alerts based on the user's emotional state. For example, it may include statements such as, "An unstable section of the scaffolding has been detected. Reinforcement is required," as well as warnings such as, "The operator is feeling anxious. Please re-check the situation."

[1399] Input: Image analysis results, emotional state data

[1400] Output: Generated report

[1401] Step 6:

[1402] Report distribution:

[1403] The server sends the generated report to the user's device. The device receives the report and displays it so the user can view it. For example, a dedicated app might receive a notification, and an interface might be presented within the app that allows the user to view the report.

[1404] Input: Generated report

[1405] Output: Report displayed on the user terminal

[1406] Step 7:

[1407] Reflection in TBM-KY:

[1408] Based on the report received on their device, users add safety confirmation items to the TBM-KY (Toolbox Meeting Hazard Prediction) record. The hazards are then shared with everyone at the actual meeting, and safety measures are implemented. For example, an item such as "scaffolding reinforcement is needed" might be discussed at the meeting, and specific countermeasures are communicated to everyone.

[1409] Input: Report content

[1410] Output: Updated TBM-KY records, implementation of safety measures.

[1411] This enables detailed safety checks that take into account the user's emotional state and provides rapid feedback, significantly improving safety in on-site work.

[1412] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1413] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1414] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[1415] [Fourth Embodiment]

[1416] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1417] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1418] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1419] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[1420] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1421] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1422] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1423] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[1424] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1425] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1426] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1427] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1428] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1429] This invention relates to a safety verification system for on-site work, and provides a system in which a generative artificial intelligence identifies potential hazards and points of caution based on photographs taken by the user at the work site and generates a report.

[1430] System Configuration

[1431] This system consists of the following elements:

[1432] 1. Terminal

[1433] This is a device for users to take photos on-site.

[1434] It has a function to upload the photos taken to a server.

[1435] 2. Server

[1436] Receive photo data sent from the device.

[1437] Image analysis is performed using generative artificial intelligence.

[1438] A report is generated based on the analysis results.

[1439] Send the report to the terminal.

[1440] 3. Generative Artificial Intelligence

[1441] It is an artificial intelligence embedded in a server that detects objects, people, and environmental elements within images.

[1442] Identify potential hazards and points of caution.

[1443] Automatically generate reports.

[1444] Program processing

[1445] 1. Upload image

[1446] Users take photos of areas requiring safety checks using their smartphones or tablets at the site.

[1447] The device uploads the photos it takes to the server using a dedicated app.

[1448] 2. Image reception

[1449] The server receives and stores image data transmitted from terminals via the internet.

[1450] 3. Image Analysis

[1451] The server sends the stored images to the generative artificial intelligence.

[1452] Generative artificial intelligence detects and analyzes objects, people, and environmental elements within an image.

[1453] Based on the detected information, it is compared with pre-set safety standards to identify potential hazards and points of caution.

[1454] 4. Report generation

[1455] The server generates a report based on the analysis results from the generative artificial intelligence.

[1456] The report includes specific hazardous areas, proposed countermeasures, and images that visually represent these hazardous areas.

[1457] 5. Report distribution

[1458] The server sends the generated report to the terminal.

[1459] The terminal receives the report and displays it for the user to view.

[1460] 6. Reflection in TBM-KY

[1461] Based on the report content, the user adds safety check items to the TBM-KY record.

[1462] The report contents are shared in a meeting before on-site work begins, ensuring everyone is aware of the risks.

[1463] Specific example

[1464] Example 1: Image upload

[1465] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[1466] Specific example 2: Image analysis

[1467] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs to be reinforced."

[1468] Specific example 3: Report generation

[1469] The server generates a report based on the analysis results. It creates a report stating, "Unstable areas of the scaffolding have been detected. Reinforcement is required," and also inserts a photo of the dangerous areas circled in red as part of the analysis results.

[1470] Specific Example 4: Report Distribution and TBM-KY Reflection

[1471] The server generates a report and sends it to the user's terminal. The user views the report, adds "scaffolding reinforcement" as an item requiring it to TBM-KY, and notifies everyone.

[1472] This invention provides an effective system for field workers to quickly and accurately identify potential hazards at a work site, thereby improving the accuracy of safety checks and preventing accidents from occurring.

[1473] The following describes the processing flow.

[1474] Step 1:

[1475] Users take photos on-site. They use smartphones or tablets to take pictures of areas where safety checks are necessary.

[1476] Step 2:

[1477] The device saves the photos, and the app uploads them to the server. When the user selects photos taken using the dedicated app and presses the upload button, the app compresses the image files and sends them to the server via the internet.

[1478] Step 3:

[1479] The server receives the image. The server receives the uploaded image data via the internet and saves it to the specified directory.

[1480] Step 4:

[1481] The server sends the stored image to the generative artificial intelligence. The server then passes the received image data to the analysis module, which begins image analysis.

[1482] Step 5:

[1483] Generative artificial intelligence performs image analysis. The AI ​​detects objects, people, and environmental elements within the image and identifies potential hazards and points of caution based on these. For example, it can determine unstable areas of scaffolding or the proximity of power lines.

[1484] Step 6:

[1485] The server generates a report based on the analysis results. Upon receiving the analysis results from the generative artificial intelligence, the server automatically generates a report that includes the identified hazardous areas and suggested countermeasures. The report includes images showing the hazardous areas with red frames as part of the analysis results.

[1486] Step 7:

[1487] The server sends the generated report to the terminal. The server sends the generated report to the user's terminal and notifies the user of the report's arrival via push notification.

[1488] Step 8:

[1489] The device displays the report. The user opens a dedicated app and views the report sent from the server. The report includes specific hazardous areas, suggested countermeasures, and images visually illustrating the hazardous areas.

[1490] Step 9:

[1491] The user reflects the information in TBM-KY. Based on the report, the user adds safety check items to the toolbox meeting (TBM-KY) before on-site work and shares them with all on-site workers.

[1492] By following the steps outlined above, potential hazards at the site are identified objectively and quickly. Through sharing this information via TBM-KY (Track Byte-Marketing - Hazard Prediction), everyone becomes aware of the hazards and takes appropriate measures, thereby improving the accuracy of safety checks.

[1493] (Example 1)

[1494] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1495] In on-site work, it is essential to quickly and accurately identify potential hazards and efficiently conduct safety checks. However, currently, many sites rely on paper-based checklists and experience, which can lead to overlooking hazards and inadequate countermeasures. Furthermore, identifying individual hazardous areas and responding quickly becomes difficult, making it challenging to ensure overall site safety.

[1496] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1497] In this invention, the server includes means for uploading photos taken by the user on-site to the server via a terminal; means for the server to store the received image data in a database and transmit it to a generative artificial intelligence; means for the generative artificial intelligence to detect objects, people, and environmental elements in the image and identify potential hazards and points of caution based on this information by comparing it with safety standards; means for the server to generate a report based on the analysis results and transmit it to the terminal of a designated user; means for the terminal to receive the generated report and display it so that the user can view it; and means for the user to reflect safety confirmation items in a toolbox meeting based on the contents of the report. This makes it possible to quickly and accurately identify potential hazards in on-site work and to take prompt countermeasures against the identified hazardous areas.

[1498] A "user" is an individual or group engaged in on-site work who uses the system to perform safety checks.

[1499] A "terminal" refers to a smartphone, tablet, or other electronic device used by a user to upload photos taken on-site to a server.

[1500] A "server" is a computer system that receives photo data sent from a terminal, requests image analysis from a generative artificial intelligence, and generates a report based on that analysis.

[1501] "Generative artificial intelligence" refers to an artificial intelligence program that has algorithms to detect objects, people, and environmental elements within an image and identify potential dangers and points of caution.

[1502] "Image data" refers to photographic data taken by a user on their device and uploaded to the server.

[1503] A "report" is a document that includes specific hazardous areas and recommended countermeasures, based on the analysis results of a generative artificial intelligence.

[1504] A "toolbox meeting" is a meeting held before work begins where workers gather to check safety procedures and share information about the work to be done.

[1505] This invention relates to a safety verification system for on-site work, providing a system in which a generative artificial intelligence identifies potential hazards and points of caution based on photographs taken by the user at the worksite and generates a report. The components of this system are a terminal, a server, and a generative artificial intelligence.

[1506] System components

[1507] 1. Terminal

[1508] A terminal refers to a device such as a smartphone or tablet used by a user to record photos taken on-site.

[1509] The device requires the installation of a dedicated application. This dedicated application has functions for compressing, selecting, and uploading photos.

[1510] 2. Server

[1511] A server is a computer system that receives image data sent by users via the internet and stores it in a database.

[1512] The server has a generative artificial intelligence (AI) built in, and it sends the received image data to the AI.

[1513] The server generates a report based on the analysis results of the generative artificial intelligence and sends it to the user's terminal.

[1514] 3. Generative Artificial Intelligence

[1515] Generative artificial intelligence is an AI algorithm embedded in a server that has the function of detecting objects, people, and environmental elements within an image.

[1516] Generative artificial intelligence identifies potential hazards and points of caution by comparing them with safety standards and automatically generates reports.

[1517] System Operation Instructions

[1518] Users take photos of areas requiring safety checks on-site and upload the photo data to a server using a dedicated app. Upon receiving the image data, the server requests analysis from a generative artificial intelligence (AI). The AI ​​analyzes the image data and identifies potential hazards and points of caution. Based on the analysis results, the server generates a report and sends it back to the user's terminal. Users view the report and incorporate the findings into toolbox meetings.

[1519] Specific example

[1520] Example 1: Image upload

[1521] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[1522] Specific example 2: Image analysis

[1523] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs reinforcement."

[1524] Specific example 3: Report generation

[1525] The server generates a report based on the analysis results. The report includes a message stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," along with a photograph showing the dangerous areas circled in red as a result of the analysis.

[1526] Specific Example 4: Report Distribution and TBM-KY Reflection

[1527] The server generates a report and sends it to the user's terminal. The user views the report, adds "reinforcement of scaffolding" as an item requiring it to TBM-KY (Toolbox Meeting Keep and Yado), and notifies everyone.

[1528] This invention provides an effective system for field workers to quickly and accurately identify potential hazards at a work site, thereby improving the accuracy of safety checks and preventing accidents from occurring.

[1529] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1530] Step 1:

[1531] Users take photos of areas requiring safety checks using their smartphones or tablets at the site. After taking the photos, they select them using a dedicated app on their device and press the upload button. The device compresses the selected photos to prepare them for efficient use of network bandwidth. The input is the captured photo data, and the output is the compressed photo data.

[1532] Step 2:

[1533] The device sends compressed photo data to the server. The server receives the photo data via the internet and temporarily stores it in a database. The input is the compressed photo data, and the output is the image data stored in the database. The server returns a response to the device confirming that the image was saved successfully.

[1534] Step 3:

[1535] The server retrieves image data stored in the database and sends it to the generative artificial intelligence (AI). The input is the image data stored in the database, and the output is the data sent to the generative AI. Next, the generative AI receives the image and executes an object detection algorithm. Specifically, it identifies objects, people, and environmental elements in the image and returns this information to the server in an encoded format.

[1536] Step 4:

[1537] The generative artificial intelligence sends detection results from the image back to the server. This response includes data on identified objects, people, environmental elements, and potential hazardous areas. The input is image data to be analyzed, and the output is detection result data. Based on the received analysis data, the server compares it with established safety standards to identify hazardous areas and points of caution. The output is an interim report summarizing the hazardous areas and points of caution.

[1538] Step 5:

[1539] The server generates a report document using the analysis results of the generative artificial intelligence. The report includes details of detected hazardous areas, recommended countermeasures, and images highlighting the hazardous areas in red. The input is interim report data, and the output is a report document (in PDF or HTML format).

[1540] Step 6:

[1541] The server sends the generated report to the designated user's device. The device receives the report and notifies the user via a dedicated app. The input is the report document, and the output is the report notification to the user's device. The user views the report using the dedicated app to check for specific instructions and countermeasures.

[1542] Step 7:

[1543] Based on the generated report, the user adds the necessary safety check items to the TBM-KY (Toolbox Meeting Keep and Yado) record. The user shares the contents of this report with other workers during the pre-work meeting so that everyone is aware of the hazards. The input is the report content, and the output is the updated TBM-KY record.

[1544] (Application Example 1)

[1545] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1546] Traditional on-site safety check systems primarily rely on manual inspections, which suffers from the time and effort required to identify potential hazards. Furthermore, while many robots operate within factories, and there is a need to improve the efficiency of safety checks, the lack of robots themselves capable of performing safety checks of the work environment hinders progress in this area as well.

[1547] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1548] This invention includes a server comprising means for uploading photos taken by a user on-site to the server via a terminal, means for the server to transmit the received images to a generative artificial intelligence for analysis, means for the generative artificial intelligence to identify potential hazards and points of caution in the images and generate a report, means for the server to transmit the generated report to a terminal for the user to view, means for the user to reflect safety confirmation items based on the report in a pre-work toolbox meeting, means for a robot to take photos of the work area and upload those images to the server, means for the server to analyze images inside the production facility and identify potential hazardous areas, and means for the server to transmit the generated report to a work manager's terminal for viewing. This enables the rapid identification of potential hazardous areas and points of caution in on-site work and work areas within factories, improving the efficiency and accuracy of safety checks.

[1549] A "user" refers to a worker or manager who uses the safety verification system.

[1550] "The site" refers to factories, construction sites, and other workplaces.

[1551] A "device" is a device used by a user to take photos and upload them to a server, and includes smartphones, tablets, and other similar devices.

[1552] A "server" is a computer system that receives and analyzes image data sent by users and generates reports.

[1553] "Generative artificial intelligence" is an artificial intelligence technology that analyzes objects and environmental elements within an image to identify potential hazards and points of caution.

[1554] A "report" is a document containing information analyzed by a generative artificial intelligence system, including information on hazardous areas and countermeasures.

[1555] A "toolbox meeting" is a pre-work safety check meeting.

[1556] A "robot" is an automated work machine used in a factory, and is a device equipped with the image capture function of the present invention.

[1557] "Work area" refers to the area or section where robots and workers operate.

[1558] "Uploading" refers to the act of sending images taken with a device or robot to a server.

[1559] "Analysis" refers to the process by which generative artificial intelligence processes image data to identify potential hazards and points of caution.

[1560] A "manager" is a person responsible for safety management in a factory or work site.

[1561] A "production facility" refers to a place where products are produced, including factories and manufacturing sites.

[1562] "Potential hazard areas" are locations identified through image analysis where accidents or injuries are likely to occur.

[1563] "Visually representing images" are images that highlight dangerous areas or add supplementary explanations based on the analysis results.

[1564] This invention provides a system that streamlines safety checks in factories and on-sites, analyzing user-submitted photographs to identify potential hazards and generate reports. Specific embodiments of this system are described below.

[1565] First, the user takes photos of the work area and equipment using a device such as a smartphone or tablet. A dedicated application is installed on this device, and the captured images are uploaded to the server via this application. The uploaded image data is received and stored by the server. The stored images are then sent to a generative artificial intelligence system.

[1566] The server analyzes the received images using generative artificial intelligence (AI). During this process, the AI ​​detects objects, people, and environmental elements within the image, and identifies potential hazards and points of caution based on pre-defined safety criteria. This generative AI utilizes libraries such as TensorFlow.

[1567] Next, the server automatically generates a report based on the analysis results from the generative artificial intelligence. This report contains information on identified hazardous areas and suggested countermeasures. It also includes images that visually represent the hazardous areas based on the analysis results. Libraries such as ReportLab can be used to create the report.

[1568] The generated report is sent from the server to the user's terminal, where the user can view it. Furthermore, based on this report, the user adds safety check items to the pre-work toolbox meeting and informs everyone.

[1569] Furthermore, a dedicated application can be installed on robots used within the factory, allowing them to take photos of the work area using their onboard cameras. The robots upload the captured images to a server, where they are similarly analyzed by generative artificial intelligence. The reports generated based on the analysis results are sent to the work supervisor's terminal, enabling them to review the information immediately.

[1570] Specific example

[1571] 1. Image capture: Users take photos of specific areas within the factory using their smartphones. After taking the photos, they upload the images to the server using a dedicated application.

[1572] 2. Example of a prompt:

[1573] Analyze the images of this factory to identify potential hazards.

[1574] 3. Image Analysis: Generative artificial intelligence analyzes photographs to detect unstable scaffolding, unsecured equipment, and other issues.

[1575] 4. Report generation: The server generates a report stating, "Unstable scaffolding detected. Reinforcement is required," and includes an image with the unstable area highlighted in red.

[1576] 5. Report Distribution: The server generates reports and sends them to the user's smartphone for viewing. This allows for quick implementation of necessary safety measures before actual work begins.

[1577] The above describes the basic configuration for carrying out the present invention. This system makes on-site work and safety checks within factories more efficient and enables accident prevention.

[1578] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1579] Step 1:

[1580] The user takes photos of the work area and equipment using a device (smartphone or tablet). Specifically, the user launches the application and captures the desired scene in camera mode. In this case, the input is the captured image, and the output is the image file saved on the device.

[1581] Step 2:

[1582] The captured photos are uploaded to the server via a dedicated app. Specifically, pressing the "upload button" in the application sends the image data to the server. In this process, the input is the image file on the device, and the output is the image data transferred to the server.

[1583] Step 3:

[1584] The server saves the received image data and sends it to the generative artificial intelligence. Specifically, it saves the received image data as a temporary file and passes its path to the generative AI. In this case, the input is the image data sent from the terminal, and the output is the provision of the path to the image data to the generative AI.

[1585] Step 4:

[1586] Generative artificial intelligence analyzes image data to identify potential hazards and points of caution. Specifically, it applies object detection algorithms within images to identify hazardous areas. In this process, the input is image data transmitted from a server, and the output is information about hazardous areas as a result of the analysis.

[1587] Step 5:

[1588] The server generates a report based on the analysis results from the generative artificial intelligence. Specifically, it uses a report generation library to create a page that visually displays an overview of the analysis results and areas of concern. In this process, the input is the analysis result data from the generative artificial intelligence, and the output is the completed report file.

[1589] Step 6:

[1590] The server sends the generated report to the user's device, and the user views it. Specifically, the server sends the report file to the user's device via email or in-app notification, and the user receives the notification and opens the report. In this process, the input is the generated report file, and the output is the report displayed on the user's device.

[1591] Step 7:

[1592] Based on the report, the user incorporates safety check items into the pre-work toolbox meeting. Specifically, this involves reviewing the report, sharing the information in the meeting, and taking necessary safety measures. In this process, the input is the submitted report, and the output is the safety check information shared within the meeting.

[1593] Step 8:

[1594] The robot takes pictures of the work area and uploads the images to a server. Specifically, the robot's camera function is automatically activated, images are captured, and uploaded to the server via the network. In this process, the input is the image data captured by the robot, and the output is the image data transferred to the server.

[1595] Step 9:

[1596] The server analyzes images from inside the production facility to identify potential hazards. Specifically, it uses generative artificial intelligence to analyze objects and environmental elements within the work area and highlight hazards. In this process, the input is image data transmitted from the robot, and the output is information about the identified hazards.

[1597] Step 10:

[1598] The server sends the generated report to the work manager's terminal, ensuring that the manager can review it immediately. Specifically, the server delivers the report, including the analysis results, to the work manager's terminal via email or a dedicated application, allowing the manager to view the report content immediately. In this case, the input is the generated report, and the output is the report displayed on the work manager's terminal.

[1599] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1600] This invention combines a safety confirmation system for on-site work with an emotion engine that recognizes user emotions. In addition to a system that uses generative artificial intelligence to identify potential hazards and points of caution based on photos taken by the user at the worksite and generates a report, it also provides a function that performs safety confirmation while taking the user's emotional state into consideration.

[1601] System Configuration

[1602] This system consists of the following elements:

[1603] 1. Terminal

[1604] This is a device for users to take photos on-site.

[1605] It has a function to upload the photos taken to a server.

[1606] It implements an emotion engine and has the ability to analyze the user's voice and facial expressions.

[1607] 2. Server

[1608] Receive photo data sent from the device.

[1609] Image analysis is performed using generative artificial intelligence.

[1610] A report is generated based on the analysis results.

[1611] Send the report to the terminal.

[1612] 3. Generative Artificial Intelligence

[1613] It is an artificial intelligence embedded in a server that detects objects, people, and environmental elements within images.

[1614] Identify potential hazards and points of caution.

[1615] Automatically generate reports.

[1616] 4. Emotional Engine

[1617] The system analyzes the user's voice and facial expressions to identify their emotional state.

[1618] Depending on the emotional state, the generated reports will include additional warnings and alerts.

[1619] Program processing

[1620] 1. Upload image

[1621] Users take photos of areas requiring safety checks using their smartphones or tablets at the site.

[1622] The device uploads the photos it takes to the server using a dedicated app.

[1623] 2. Image reception

[1624] The server receives image data sent from the terminal via the internet and saves it to the specified directory.

[1625] 3. Image Analysis

[1626] The server sends the stored images to the generative artificial intelligence.

[1627] Generative artificial intelligence detects and analyzes objects, people, and environmental elements within an image.

[1628] Based on the detected information, it is compared with pre-set safety standards to identify potential hazards and points of caution.

[1629] 4. Emotion analysis

[1630] The device analyzes the user's voice and facial expressions during shooting using an emotion engine to identify the user's emotional state.

[1631] For example, if signs of anxiety or tension are detected, that information is sent to the server.

[1632] 5. Report generation

[1633] The server generates a report based on the analysis results from the generative artificial intelligence.

[1634] The report includes specific hazardous areas, proposed countermeasures, and images that visually represent these hazardous areas.

[1635] Based on the user's emotional state identified by the emotion engine, the report will include additional warnings and alerts as needed.

[1636] 6. Report distribution

[1637] The server sends the generated report to the terminal.

[1638] The terminal receives the report and displays it for the user to view.

[1639] 7. Reflection in TBM-KY

[1640] Based on the report content, the user adds safety check items to the TBM-KY record.

[1641] The report contents are shared in a meeting before on-site work begins, ensuring everyone is aware of the risks.

[1642] Specific example

[1643] Example 1: Image upload

[1644] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[1645] Specific example 2: Image analysis

[1646] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs to be reinforced."

[1647] Example 3: Sentiment Analysis

[1648] The device uses an emotion engine to analyze the user's voice and facial expressions while they are taking photos, and detects if the user is feeling anxious. This information is then sent to the server.

[1649] Specific Example 4: Report Generation

[1650] The server generates a report based on the analysis results. In addition to a report stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," it also includes a warning that takes into account the user's level of anxiety.

[1651] Specific Example 5: Report Distribution and TBM-KY Reflection

[1652] The server generates a report and sends it to the user's terminal. The user views the report, adds "scaffolding reinforcement" as an item requiring it to TBM-KY, and notifies everyone.

[1653] This invention aims to further improve safety by combining an emotion engine to enable safety checks that also take into account the emotional state of on-site workers.

[1654] The following describes the processing flow.

[1655] Step 1:

[1656] Users take photos on-site. They use smartphones or tablets to take pictures of areas where safety checks are necessary.

[1657] Step 2:

[1658] The device saves the photos, and the app uploads them to the server. When the user selects photos taken using the dedicated app and presses the upload button, the app compresses the image files and sends them to the server via the internet.

[1659] Step 3:

[1660] The device collects user voice and facial expression data. While taking or uploading photos, the device uses its built-in camera and microphone to collect user voice and facial expression data.

[1661] Step 4:

[1662] The device passes the collected voice and facial expression data to the emotion engine. The emotion engine analyzes the data and identifies the user's emotional state. For example, it can determine from the voice whether the user is nervous or not.

[1663] Step 5:

[1664] The server receives the image. The server receives the uploaded image data via the internet and saves it to the specified directory.

[1665] Step 6:

[1666] The server sends the stored image to the generative artificial intelligence. The server then passes the received image data to the analysis module, which begins image analysis.

[1667] Step 7:

[1668] Generative artificial intelligence performs image analysis. The AI ​​detects objects, people, and environmental elements within the image and identifies potential hazards and points of caution based on these. For example, it can determine unstable areas of scaffolding or the proximity of power lines.

[1669] Step 8:

[1670] The server generates a report based on the analysis results. Upon receiving the analysis results from the generative artificial intelligence, the server automatically generates a report that includes the identified hazardous areas and suggested countermeasures. The report includes images showing the hazardous areas with red frames as part of the analysis results.

[1671] Step 9:

[1672] The server receives emotion data from the emotion engine. If the user is in a stressed state, the server uses this information to include additional warnings and alerts in the report.

[1673] Step 10:

[1674] The server sends the generated report to the terminal. The server sends the generated report to the user's terminal and notifies the user of the report's arrival via push notification.

[1675] Step 11:

[1676] The device displays the report. The user opens a dedicated app and views the report sent from the server. The report includes specific hazardous areas, suggested countermeasures, and images visually representing the hazardous areas. It also includes additional warnings based on the user's emotional state.

[1677] Step 12:

[1678] The user reflects the information in TBM-KY. Based on the report, the user adds safety check items to the toolbox meeting (TBM-KY) before on-site work and shares them with all on-site workers.

[1679] By following the steps outlined above, we can objectively and quickly identify potential hazards on-site and take appropriate measures that take into account the emotional state of users, thereby improving the accuracy of safety checks.

[1680] (Example 2)

[1681] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1682] Conventional on-site safety confirmation systems do not take into account the emotional state of the user, and therefore cannot adequately reflect the stress and anxiety felt by workers, potentially leading to overlooking potential hazards or implementing inappropriate countermeasures. Furthermore, these systems lack sufficient means of visually displaying analysis results, making it difficult for users to intuitively understand specific hazardous areas. Therefore, this invention aims to solve these problems and achieve more accurate safety confirmation and countermeasures while taking into account the emotional state of workers.

[1683] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1684] In this invention, the server includes means for uploading images taken by the user on-site to the server via a terminal, means for the server to save the received image data to a specified directory, means for the server to transmit the saved image data to a generative artificial intelligence for analysis, means for the generative artificial intelligence to detect objects, people, and environmental elements in the images and generate a report based on these to identify potential hazards and points of caution, means for the terminal to transmit the user's voice and facial expression data to an emotion analysis engine to identify the user's emotional state, means for the server to include additional warnings and alerts in the report based on the user's emotional state, means for the server to transmit the generated report to the terminal for the user to view it, and means for the user to reflect safety confirmation items based on the generated report in a pre-work meeting. This makes it possible to perform safety checks that take into account the user's emotional state, and by visually indicating specific hazardous areas, more intuitive and effective safety measures can be realized.

[1685] "Users" refer to workers and managers who take images on-site and use the system.

[1686] A "terminal" is a device used by a user that has the function of taking images on-site and sending the data to a server. Specific examples include smartphones and tablets.

[1687] A "server" refers to a computer system that processes data received from a terminal, works with generative artificial intelligence and emotion analysis engines to generate analysis results, and sends reports to the terminal.

[1688] "Generative artificial intelligence" refers to machine learning models or algorithms that analyze image data to detect objects, people, and environmental elements within the image, and to identify potential dangers and points of caution.

[1689] An "emotion analysis engine" refers to an algorithm or technology that analyzes a user's voice and facial expression data to identify the user's emotional state.

[1690] "Image data" refers to photographs and videos taken by users on-site, which are uploaded to the server and used for analysis.

[1691] A "directory" refers to a specific folder or part of storage used to store data received within a server.

[1692] A "report" refers to a document that integrates the results of image analysis by generative artificial intelligence and sentiment analysis by a sentiment analysis engine, including potential risks, points of caution, and countermeasures.

[1693] A "meeting" refers to a planning meeting held by users before starting work, where safety checks are conducted based on the contents of reports, such as TBM-KY (Tool Box Meeting - Kiken Yochi, Hazard Prediction Activity).

[1694] "Warning" refers to additional warnings or alerts included in reports based on the user's emotional state, providing information that encourages workers to pay close attention to specific hazards.

[1695] "Analysis results" refer to data provided by generative artificial intelligence and emotion analysis engines, including detected risk areas, countermeasures against them, and information on the user's emotional state.

[1696] This invention relates to a safety verification system for on-site work, and by combining it with a function to recognize the user's emotions, it achieves more effective safety measures. This system consists of the following hardware and software.

[1697] System Configuration

[1698] This system includes the following main components:

[1699] 1. Terminal

[1700] These are devices that users use to take photos and videos on-site. Specific examples include smartphones and tablets.

[1701] Install a dedicated application for uploading captured image data to the server.

[1702] It has the function of sending the user's voice and facial expression data to an emotion analysis engine to identify the user's emotions.

[1703] 2. Server

[1704] Saves image data received from the terminal to the specified directory.

[1705] Install and analyze software that operates as a generative artificial intelligence (for example, the OpenAI API).

[1706] The analysis results are integrated using software that acts as an emotion analysis engine (e.g., an emotion recognition API).

[1707] Generate a report and send it to the terminal.

[1708] 3. Generative Artificial Intelligence

[1709] This is a machine learning model embedded in the server that detects objects, people, and environmental elements within images.

[1710] Identify potential risks and points of caution, and automatically generate reports.

[1711] The software used will be the OpenAI API and similar analysis tools.

[1712] 4. Emotion Analysis Engine

[1713] Analyze audio and video data to identify the user's emotional state.

[1714] Based on emotional states, the generated reports will include additional warnings and alerts.

[1715] Program processing

[1716] The program for this system processes the information in the following order:

[1717] 1. Upload image

[1718] Users take pictures of areas requiring safety checks using their smartphones or tablets at the site.

[1719] The device uploads captured images to a server using a dedicated app. This app has the functionality to transmit image data over the internet.

[1720] 2. Image reception

[1721] The server receives image data sent from the terminal via the internet and saves it to a specified directory. Specifically, it saves files to cloud storage or local storage.

[1722] 3. Image Analysis

[1723] The server sends the stored image data to a generative artificial intelligence system. For example, a Python script on the server can be used to send the image data to an API.

[1724] Generative artificial intelligence analyzes received image data to detect objects, people, and environmental elements within the image. Specifically, it uses computer vision technology to identify unstable areas of scaffolding or lack of safety equipment.

[1725] 4. Emotion analysis

[1726] The device transmits the user's voice and facial expressions to an emotion analysis engine to identify the user's emotional state. This collects data using the smartphone's camera and microphone.

[1727] The terminal sends the analysis results to the server, which is then used as additional security management information.

[1728] 5. Report generation

[1729] The server generates a report based on analysis results from a generative artificial intelligence and the sentiment analysis engine. This report includes detailed descriptions of hazardous areas, specific countermeasures, and warnings that reflect the user's emotional state.

[1730] The report will be generated in PDF format or another suitable format and saved to the specified directory.

[1731] 6. Report distribution

[1732] The server sends the generated report to the device. For example, it notifies the user of the report using email or push notifications via a dedicated app.

[1733] The terminal displays the received report, allowing the user to review it.

[1734] 7. Reflection in TBM-KY

[1735] Users conduct pre-work meetings based on reports during safety meetings (e.g., TBM-KY) to confirm hazardous areas and countermeasures.

[1736] For example, report contents can be shared on a smartphone screen, and necessary safety measures can be discussed.

[1737] Specific example

[1738] Example 1: Image upload

[1739] The user takes photos of the scaffolding at the site. They select the photos using a dedicated app on their device and press the upload button to send them to the server.

[1740] Specific example 2: Image analysis

[1741] The server sends the received photos of the scaffolding to a generative artificial intelligence. The generative AI identifies unstable parts of the scaffolding and points out that "the scaffolding needs to be reinforced."

[1742] Example 3: Sentiment Analysis

[1743] The device uses an emotion analysis engine to analyze the user's voice and facial expressions while they are taking photos, and detects if the user is feeling anxious. This information is then sent to the server.

[1744] Specific Example 4: Report Generation

[1745] The server generates a report based on the analysis results. In addition to a report stating, "Unstable sections of the scaffolding have been detected. Reinforcement is required," it also includes a warning that takes into account the user's level of anxiety.

[1746] Specific Example 5: Report Distribution and TBM-KY Reflection

[1747] The server generates a report and sends it to the user's terminal. The user views the report, adds "scaffolding reinforcement" as an item requiring it to TBM-KY, and notifies everyone.

[1748] Example of a prompt:

[1749] "Identify potential hazards in this image and propose safety measures."

[1750] This enables safety checks for on-site work that take into account the user's emotional state, and by visually indicating specific hazardous areas, it allows for more intuitive and effective safety measures.

[1751] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1752] Step 1: Take and upload images

[1753] Users take pictures of areas requiring safety checks using their smartphones or tablets at the site.

[1754] Input: Image taken by the user.

[1755] Specific operation: The user takes pictures of the site using their smartphone's camera app and selects those images within the dedicated app.

[1756] The device uploads images to the server using a dedicated app. This app transmits image data via HTTPS over the internet.

[1757] Output: Image data uploaded to the server.

[1758] Step 2: Receiving and saving image data

[1759] The server receives image data sent from the terminal via the internet.

[1760] Input: Uploaded image data.

[1761] Specific operation: The server receives image data via a dedicated API and saves it to the directory " / images / uploads". The save location is recorded in the database.

[1762] Output: Image files saved in the specified directory.

[1763] Step 3: Perform image analysis

[1764] The server transmits the stored image data to the generative artificial intelligence system.

[1765] Input: The path to the image file saved in the directory.

[1766] Specific operation: A Python script on the server is used to send the image data path to the API.

[1767] Generative artificial intelligence analyzes image data to detect objects, people, and environmental elements.

[1768] Output: Analysis results (list of detected objects, people, and environmental elements).

[1769] Step 4: Identifying potential hazards

[1770] Generative artificial intelligence identifies potential risks and points of caution based on the analysis results.

[1771] Input: A list of detected objects, people, and environmental elements.

[1772] Specific operation: The generative artificial intelligence uses an analysis algorithm to compare the detected elements with safety standards and identify hazardous areas and points of caution.

[1773] Output: List of hazardous areas and points to note.

[1774] Step 5: Perform sentiment analysis

[1775] The device transmits the user's voice and facial expressions during shooting to an emotion analysis engine.

[1776] Input: User's voice data and facial expression data.

[1777] Specific operation: The device's microphone and camera are used to collect voice and facial expression data, which is then sent to an emotion analysis engine via a dedicated app.

[1778] The emotion analysis engine analyzes this data to identify the user's emotional state.

[1779] Output: User's emotional state (anxiety, tension, relief, etc.).

[1780] Step 6: Generate the report

[1781] The server generates a report based on the results of generative artificial intelligence and emotion analysis engines.

[1782] Input: List of hazards and points of caution, user's emotional state.

[1783] Specific operation: The server uses a web framework such as Django to integrate the analysis results and create a report. The report includes risk areas, countermeasures, and additional warnings based on the user's emotional state. The report is generated in PDF format, etc.

[1784] Output: The generated report file.

[1785] Step 7: Report Distribution

[1786] The server sends the generated report to the terminal.

[1787] Input: The generated report file.

[1788] Specific operation: The server uses a dedicated API to send report files to the terminal. For example, it might use email or push notifications.

[1789] The device displays received reports within a dedicated app, allowing users to review them.

[1790] Output: The report file delivered to the terminal.

[1791] Step 8: Reflection in TBM-KY

[1792] The user updates the TBM-KY record based on the generated report.

[1793] Input: Report file delivered to the terminal.

[1794] Specific actions: Users view reports in safety meetings and add necessary safety check items to the TBM-KY sheet. This is then shared with everyone, and specific safety measures are discussed.

[1795] Output: Updated TBM-KY record.

[1796] (Application Example 2)

[1797] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1798] Conventional safety confirmation systems for on-site work often failed to consider the user's psychological state, resulting in situations where appropriate warnings and responses could not be provided. Furthermore, because potential hazards and points of caution were identified solely through image analysis, safety measures based on the user's emotional changes were insufficient. This led to problems where optimal safety checks could not be performed, increasing the risk of workplace accidents.

[1799] In Application Example 2, the identification processing by the identification processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for uploading photos taken by the user at the site to the server via a terminal, means for the server to transmit the received images to a generative artificial intelligence for analysis, means for the generative artificial intelligence to identify potential dangers and points of caution in the images and generate a report, means for the server to transmit the generated report to the terminal for the user to view, means for the terminal to identify the user's emotional state using an emotion engine that analyzes the user's voice and facial expressions, and means for the emotion engine to transmit the analysis results to the server and include additional warnings and alerts in the report. This makes it possible to perform a comprehensive safety check that takes the user's emotions into consideration and reduces the risk of occupational accidents.

[1800] A "site work safety confirmation system" is a system for ensuring safety during on-site work by checking the conditions of the work site, identifying potential hazards and points of caution, and implementing safety measures.

[1801] A "terminal" is a device that allows users to take photos on-site and upload them to a server.

[1802] A "server" is a device that receives image data transmitted from a terminal, analyzes it using generative artificial intelligence, generates a report based on the analysis results, and then transmits that report back to the terminal.

[1803] "Generative artificial intelligence" is an artificial intelligence that analyzes transmitted image data, detects objects, people, and environmental elements within the image, and has the function of identifying potential dangers and points of caution.

[1804] A "report" is a document created based on the results of analysis by a generative artificial intelligence system, which describes potential hazards, points of caution, and countermeasures to address them.

[1805] An "emotion engine" is software or hardware that analyzes a user's voice and facial expressions to identify the user's emotional state.

[1806] A "toolbox meeting" is a pre-work meeting where workers gather to share information about the day's tasks, precautions, and safety checks.

[1807] "Safety confirmation items" are matters and procedures that must be checked to ensure safety during on-site work, and are shared at toolbox meetings.

[1808] The following describes an embodiment for carrying out this invention. This system allows users to upload photos taken with a smartphone or other device to a server for safety checks during on-site work, and a generative artificial intelligence analyzes the photo data. Furthermore, safety is further enhanced by analyzing the user's emotional state using an emotion engine and reflecting it in the report.

[1809] Hardware and software to use

[1810] Device: Use a device such as a smartphone or tablet that has the ability to take photos on-site and upload them to a server.

[1811] Server: This device receives image data, performs analysis using generative artificial intelligence, generates a report, and sends it to the terminal.

[1812] Generative artificial intelligence: This uses artificial intelligence implemented on a server to analyze image data and identify potential dangers and points of caution.

[1813] Emotion engine: This is software or hardware implemented in the device that analyzes the user's voice and facial expressions to identify their emotional state.

[1814] Data processing and data calculation

[1815] 1. Upload image:

[1816] The user takes a photo of the site with their device. The device then uploads this photo to the server.

[1817] 2. Image analysis:

[1818] The server receives image data and sends it to a generative artificial intelligence system to detect objects, people, and environmental elements within the image. This allows for the identification of potential hazards and points of caution.

[1819] 3. Emotion analysis:

[1820] The device's emotion engine analyzes the user's voice and facial expressions to identify their emotional state. When emotions such as anxiety or tension are detected, that information is sent to the server.

[1821] 4. Report generation:

[1822] The server generates a report based on the analysis results of the generative artificial intelligence and the emotion engine. The report includes specific areas of risk and countermeasures, as well as additional warnings and alerts tailored to the user's emotional state.

[1823] 5. Report distribution:

[1824] The server generates a report, which is then sent to the user's terminal for the user to view.

[1825] Specific example

[1826] Example 1: Image upload

[1827] Users take photos of equipment inside the factory and upload them to the server using a dedicated app.

[1828] Specific example 2: Image analysis

[1829] The server uses a generative artificial intelligence system to analyze the equipment photos it receives, identifying unstable parts and missing safety devices.

[1830] Example 3: Sentiment Analysis

[1831] The device uses an emotion engine to analyze the user's voice and facial expressions to detect their level of tension. For example, audio data of a user sighing while taking a photo can be analyzed.

[1832] Specific Example 4: Report Generation

[1833] The server generates a report stating, "An unstable area has been detected. Reinforcement is needed," and includes alerts tailored to the user's level of anxiety.

[1834] Specific example 5: Report distribution

[1835] The user receives the report and incorporates safety check items into the pre-work meeting based on it.

[1836] Examples of prompts to input into a generative AI model:

[1837] input_image: "factory_image.jpg"

[1838] input_emotion_data: "operator_audio.wav"

[1839] This system enables advanced safety checks that take into account the user's emotional state, thereby improving safety in on-site work.

[1840] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1841] Step 1:

[1842] Image capture and upload:

[1843] The user takes photos of the problem area at the site using a smartphone or tablet. The captured images are uploaded to the server via a dedicated app. Specifically, the user launches the camera app and captures a still image. After taking the photo, pressing the upload button transfers the photo data from a specified directory on the device to the server. At this time, an image file, for example "factory_image.jpg", is generated and uploaded.

[1844] Input: Captured photo data, user operation

[1845] Output: Image file sent to the server

[1846] Step 2:

[1847] Image reception and saving:

[1848] The server receives image data transmitted from the terminal via the internet. The received data is stored in a designated directory on the server. This ensures that the data necessary for the next analysis process is available on the server.

[1849] Input: Image file (factory_image.jpg)

[1850] Output: Image files stored on the server

[1851] Step 3:

[1852] Image analysis:

[1853] The server sends the stored image to a generative artificial intelligence (AI). The AI ​​detects and analyzes objects, people, and environmental elements within the image. Here, the machine learning model recognizes and identifies hazardous elements in the image. For example, it might point out unstable parts of scaffolding or missing safety devices.

[1854] Input: Saved image file

[1855] Output: Analysis results (data on hazardous areas and points to note)

[1856] Step 4:

[1857] Emotion analysis:

[1858] The device analyzes the user's voice and facial expressions during filming using an emotion engine to identify the user's emotional state. Based on information obtained from audio data and camera footage, it determines anxiety, tension, and other emotional states. For example, it analyzes the user's voice during filming, such as sighs or trembling voices, to assess the degree of anxiety.

[1859] Input: Voice data, facial expression data

[1860] Output: Emotional state data (anxiety, tension, etc.)

[1861] Step 5:

[1862] Report generation:

[1863] The server generates a report based on image analysis results from a generative artificial intelligence and emotional state data from an emotion engine. The report includes specific hazards detected, countermeasures, and additional warnings and alerts based on the user's emotional state. For example, it may include statements such as, "An unstable section of the scaffolding has been detected. Reinforcement is required," as well as warnings such as, "The operator is feeling anxious. Please re-check the situation."

[1864] Input: Image analysis results, emotional state data

[1865] Output: Generated report

[1866] Step 6:

[1867] Report distribution:

[1868] The server sends the generated report to the user's device. The device receives the report and displays it so the user can view it. For example, a dedicated app might receive a notification, and an interface might be presented within the app that allows the user to view the report.

[1869] Input: Generated report

[1870] Output: Report displayed on the user terminal

[1871] Step 7:

[1872] Reflection in TBM-KY:

[1873] Based on the report received on their device, users add safety confirmation items to the TBM-KY (Toolbox Meeting Hazard Prediction) record. The hazards are then shared with everyone at the actual meeting, and safety measures are implemented. For example, an item such as "scaffolding reinforcement is needed" might be discussed at the meeting, and specific countermeasures are communicated to everyone.

[1874] Input: Report content

[1875] Output: Updated TBM-KY records, implementation of safety measures.

[1876] This enables detailed safety checks that take into account the user's emotional state and provides rapid feedback, significantly improving safety in on-site work.

[1877] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1878] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1879] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1880] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1881] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. In the upper and lower directions of the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. Also, the upper side of the concentric circles is where "pleasant" emotions are located, and the lower side is where "unpleasant" emotions are located. In this way, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1882] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1883] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1884] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1885] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1886] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[1887] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[1888] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[1889] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[1890] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1891] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[1892] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[1893] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[1894] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[1895] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[1896] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[1897] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[1898] The following is further disclosed regarding the embodiments described above.

[1899] (Claim 1)

[1900] It is a safety confirmation system for on-site work.

[1901] A method for users to upload photos taken on-site to a server via their device,

[1902] A means for sending images received by a server to a generative artificial intelligence for analysis,

[1903] A means for generative artificial intelligence to identify potential dangers and points of caution within an image and generate a report,

[1904] A means for the server to send the generated report to the terminal and for the user to view it,

[1905] A means for users to incorporate safety check items based on reports into the pre-work toolbox meeting,

[1906] A system that includes this.

[1907] (Claim 2)

[1908] The system according to claim 1, wherein the server-generated report includes images that visually indicate dangerous areas as a result of the analysis.

[1909] (Claim 3)

[1910] The system according to claim 1, comprising means for a generative artificial intelligence to detect objects, people, and environmental elements in an image and compare them with safety standards based on these.

[1911] "Example 1"

[1912] (Claim 1)

[1913] A method for users to upload photos taken on-site to a server via their device,

[1914] A means for the server to store the received image data in a database and transmit it to a generative artificial intelligence,

[1915] A generative artificial intelligence system detects objects, people, and environmental elements within an image, and based on this information, identifies potential hazards and points of caution by comparing them with safety standards.

[1916] A means by which the server generates a report based on the analysis results and sends it to the terminal of a specified user,

[1917] A means by which the terminal receives the generated report and displays it so that the user can view it,

[1918] A means for users to incorporate safety check items into toolbox meetings based on the report content,

[1919] A system that includes this.

[1920] (Claim 2)

[1921] The system according to claim 1, comprising means of including in the server-generated report an image visually indicating the hazardous areas as an analysis result.

[1922] (Claim 3)

[1923] The system according to claim 1, comprising means for a generative artificial intelligence to detect objects, people, and environmental elements in an image and compare them with safety standards based on these.

[1924] "Application Example 1"

[1925] (Claim 1)

[1926] A method for users to upload photos taken on-site to a server via their device,

[1927] A means for sending images received by a server to a generative artificial intelligence for analysis,

[1928] A means for generative artificial intelligence to identify potential dangers and points of caution within an image and generate a report,

[1929] A means for the server to send the generated report to the terminal and for the user to view it,

[1930] A means for users to incorporate safety check items based on reports into the pre-work toolbox meeting,

[1931] A method for a robot to take pictures of the work area and upload those images to a server,

[1932] A server analyzes images of the inside of the production facility to identify potential hazardous areas,

[1933] A means of sending the reports generated by the server to the work administrator's terminal so that they can be viewed,

[1934] A system that includes this.

[1935] (Claim 2)

[1936] The system according to claim 1, wherein the server-generated report includes images that visually indicate dangerous areas as a result of the analysis.

[1937] (Claim 3)

[1938] The system according to claim 1, comprising means for a generative artificial intelligence to detect objects, people, and environmental elements in an image and compare them with safety standards based on these.

[1939] "Example 2 of combining an emotion engine"

[1940] (Claim 1)

[1941] A method for users to upload images they have taken on-site to a server via their device,

[1942] A means for the server to save received image data to a specified directory,

[1943] A means for the server to transmit stored image data to a generative artificial intelligence for analysis,

[1944] A means for generating a report by using generative artificial intelligence to detect objects, people, and environmental elements in an image, and to identify potential hazards and points of caution based on these.

[1945] A means by which the terminal transmits the user's voice and facial expression data to an emotion analysis engine to identify the user's emotional state,

[1946] The server has a means to include additional warnings and alerts in the report based on the user's emotional state.

[1947] A means for the server to send the generated report to the terminal and for the user to view it,

[1948] A means of incorporating safety check items into pre-work meetings based on user-generated reports,

[1949] A system that includes this.

[1950] (Claim 2)

[1951] The system according to claim 1, wherein the server-generated report includes images that visually indicate dangerous areas a...

Claims

1. It is a safety confirmation system for on-site work. A method for users to upload photos taken on-site to a server via their device, A means for sending images received by a server to a generative artificial intelligence for analysis, A means for generative artificial intelligence to identify potential dangers and points of caution within an image and generate a report, A means for the server to send a report it generates to a terminal and for the user to view it, A means for users to incorporate safety check items based on reports into the pre-work toolbox meeting, A system that includes this.

2. The system according to claim 1, wherein the server-generated report includes images that visually indicate dangerous areas as a result of the analysis.

3. The system according to claim 1, comprising means for a generative artificial intelligence to detect objects, people, and environmental elements in an image and compare them with safety standards based on these.

Images