system

Abstract

Provide a system. A method for inputting risk-related data; means for transmitting said data to a server; means for storing said data in a database; means for preprocessing the stored data; means for extracting features from the preprocessed data; A means for analyzing risk based on the extracted feature quantity; A means of applying generative AI to generate risk countermeasures based on the analysis results; means for transmitting the generated countermeasures to a user terminal; A means for the user to review and select countermeasures; a means for feeding back the execution result of the selected countermeasure to the server; and means for updating the model based on said feedback.

Description

[Technical Field]

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

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] In corporate risk management, traditional methods often involve manual data collection, analysis, and countermeasure proposals, resulting in reduced speed and accuracy of decision-making and delays in appropriate countermeasures against risks. This problem is particularly pronounced in large companies that face a wide variety of risks and scales, and is likely to result in potential losses. Another issue is that the process for evaluating and improving risk countermeasures is unclear, making it difficult to form an appropriate feedback loop. [Means for solving the problem]

[0005] In order to solve the above problems, the present invention provides the following means. The system includes a means for a company to input risk-related data and transmit the data to a server, and a means for the server to store the received data in a database. The system then preprocesses the stored data and uses a means for extracting features from the preprocessed data. The system further provides a means for analyzing risk based on the extracted features.

[0006] Based on the analysis results, a generation AI is applied to generate risk countermeasures, and the generated countermeasures are sent to the user's device. The user has a means to check and select the countermeasures using the device. The system further includes a means to feed back the results of the implemented countermeasures to the server, and is equipped with a means to update the model based on this feedback. This enables companies to quickly and accurately assess risks and implement countermeasures, improving the efficiency and effectiveness of risk management.

[0007] "Risk-related data" refers to information about potential risks to a company's operations and activities, including data on supply chain interruptions, legal violations, and employee health status.

[0008] A "server" is a computer system that receives, stores, analyzes, and transmits data to user terminals over a network.

[0009] A "database" is a system in which data managed on a server is stored in an organized manner and can be searched, extracted, and updated.

[0010] "Preprocessing" refers to a series of operations performed on risk-related data to prepare them for analysis, including imputing missing values, removing unnecessary items, and normalizing the data.

[0011] "Features" refer to attributes or elements extracted from preprocessed data that are important for risk assessment.

[0012] "Risk analysis" refers to a series of calculations and analytical methods for evaluating the probability of occurrence and impact of risks based on extracted features.

[0013] "Generative AI" is a system that uses artificial intelligence technology to generate optimal risk countermeasures based on analysis results.

[0014] "User terminal" is a device used by a user to connect to a server, input data, and review and select countermeasures.

[0015] "Feedback" refers to sending information about the results and effectiveness of the measures taken back to the server, thereby improving the model. [Brief explanation of the drawings]

[0016] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10]1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0017] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0018] First, the terms used in the following description will be explained.

[0019] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0020] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0021] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0022] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0023] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0024] [First embodiment]

[0025] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0026] 1, a 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.

[0027] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0028] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0029] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the 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.

[0030] 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 of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0031] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0032] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] 2, in the data processing device 12, a specific process 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" according to the technology of the present 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 process 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.

[0034] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0035] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0036] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0037] This invention relates to a system for efficiently managing a company's operational risks. The system aims to automate the risk management process by integrating data collection, data analysis, risk assessment, countermeasure generation, countermeasure implementation, and feedback.

[0038] System configuration

[0039] Data collection

[0040] The user inputs risk-related data via a terminal.

[0041] The user uses the device to input various risk-related data, such as supply chain interruptions, legal violation risks, and employee health status. The device then verifies this information and sends it to the server.

[0042] Data storage and preprocessing

[0043] The server receives the data and stores it in the database

[0044] The server receives the data sent from the device and stores it in a database, along with a temporary label.

[0045] The server preprocesses the data

[0046] The server performs preprocessing on the stored data, which includes processes such as imputing missing values, removing unnecessary items, and normalizing the data.

[0047] Feature extraction and risk analysis

[0048] The server extracts features

[0049] After preprocessing, the features necessary for risk assessment are extracted from the data. For example, features such as "frequency of past delivery delays," "conditions for legal violations," and "results of employee health assessments" are extracted.

[0050] The server performs risk analysis

[0051] Based on the extracted features, the server performs risk analysis, using machine learning algorithms and statistical models to assess the probability of risk occurrence and its impact.

[0052] Risk countermeasure generation

[0053] The server generates countermeasures using generation AI

[0054] Based on the analysis results, the server applies generative AI to generate optimal risk countermeasures, such as suggesting alternative suppliers, increasing inventory, or implementing employee wellness programs.

[0055] The server sends the countermeasures to the device.

[0056] The generated risk countermeasure plan is sent from the server to the user's terminal, and the user can check the countermeasure plan.

[0057] Implementation of countermeasures and feedback

[0058] The user checks and selects countermeasures via the device.

[0059] The user uses the terminal to check the countermeasures received from the server, select the countermeasure that they think is most appropriate, and the selected countermeasure is then put into action.

[0060] The user enters the execution results into the terminal and sends them to the server.

[0061] After the countermeasures are implemented, the user sends feedback on the results and effects to the server from the terminal, and the effectiveness of the countermeasures is evaluated based on this feedback.

[0062] The server improves the model based on the feedback

[0063] The server stores the collected feedback data in a database and uses it to update the risk assessment model and parameters of the generative AI, making the next risk analysis and countermeasure generation more accurate and effective.

[0064] Specific examples

[0065] For example, when a manufacturing company manages supply chain risks, the user inputs data on delivery delays from key suppliers, supply chain interruption risks, and information on alternative suppliers into their terminals. The server then preprocesses this data, performs risk analysis, and uses generative AI to propose countermeasures. The user can then confirm and implement the proposed alternative supplier selection plan to minimize risk. This process is returned to the server through feedback, improving the accuracy and effectiveness of the entire system.

[0066] In this way, the present invention improves the efficiency of operational risk management for companies, enabling rapid and accurate risk assessment and countermeasure proposals, thereby enabling companies to minimize potential losses from risks.

[0067] The processing flow will be explained below.

[0068] Step 1:

[0069] Users use the terminal to input risk-related data, such as supplier delivery delays, legal violation risk, and employee health status.

[0070] Step 2:

[0071] The terminal validates the entered data, checking for format and content errors as necessary, and sends the validated data to the server.

[0072] Step 3:

[0073] The server stores the data received from the device in a database, where it is appropriately labeled when stored.

[0074] Step 4:

[0075] The server preprocesses the stored data, specifically by imputing missing values, removing unnecessary items, and normalizing the data.

[0076] Step 5:

[0077] The server extracts features necessary for risk assessment from the preprocessed data, such as the frequency of past delivery delays, conditions for legal violations, and employee health assessment results.

[0078] Step 6:

[0079] The server then runs algorithms for risk analysis based on the extracted features. These algorithms include machine learning models and statistical models, and use them to evaluate the probability of risk occurrence and its impact.

[0080] Step 7:

[0081] Based on the results of the risk analysis, the server uses generative AI to generate optimal risk countermeasures, such as proposing alternative suppliers, increasing inventory, or introducing employee health programs.

[0082] Step 8:

[0083] The server transmits the generated countermeasures to the user terminal, where the user can check the countermeasures.

[0084] Step 9:

[0085] The user uses the terminal to check the proposed countermeasures and select the countermeasure to be implemented. The terminal then feeds this selection information back to the server.

[0086] Step 10:

[0087] The user implements the measures selected by the user, such as finding an alternative supplier or increasing inventory within the company.

[0088] Step 11:

[0089] The user uses the terminal to input the results of the measures taken and send them to the server, which then provides feedback on the effectiveness of the measures.

[0090] Step 12:

[0091] The server stores the received feedback in a database, which can then be used for the next risk assessment and countermeasure generation.

[0092] Step 13:

[0093] The server updates the risk assessment model and the parameters of the generation AI based on the feedback data, thereby improving the accuracy and effectiveness of the next analysis and generated countermeasures.

[0094] Example 1

[0095] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0096] In today's corporate environment, efficient management of operational risk is crucial, but traditional manual risk management methods are insufficient. Rapid and accurate risk assessment and countermeasure proposals are therefore required. However, there is no integrated system for data collection, preprocessing, analysis, countermeasure generation, and feedback, making the risk management process fragmented and inefficient. As a result, companies often are unable to respond quickly to potential loss risks. Furthermore, an effective means of automating complex tasks such as preprocessing, feature extraction, risk analysis, and countermeasure generation is needed.

[0097] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0098] In this invention, the server includes a means for inputting risk-related information, a means for transmitting the information to the server, a means for storing the information in a database, a means for preprocessing the stored information, a means for extracting features from the preprocessed information, a means for assessing risk based on the extracted features, a means for applying a generation AI to generate risk countermeasures based on the assessment results, a means for transmitting the generated countermeasures to a user terminal, a means for the user to review and select the countermeasures, a means for feeding back the execution results of the selected countermeasures to the server, and a means for updating the model based on the feedback. This unifies the risk management process, enabling efficient and rapid automation of risk assessment and countermeasure proposals. Furthermore, by continuously improving the model based on feedback, the accuracy of risk assessment and countermeasure proposals can be improved.

[0099] "Risk-related information" refers to data collected to identify and manage potential risks in a company's operations, such as delivery delays, violations of laws and regulations, and employee health conditions.

[0100] A "server" is a computer system that receives, processes, and transmits information over a network.

[0101] A "database" is a system for storing and managing data in an orderly manner and retrieving it as needed.

[0102] "Preprocessing" refers to the process of processing data to prepare risk-related information in an analyzable format, and includes imputing missing values, removing unnecessary items, and normalizing data.

[0103] "Features" are characteristics or attributes of data selected and calculated for risk assessment and are used directly in analysis.

[0104] "Means for assessing risk" refers to the process and technology for calculating and assessing the probability of occurrence and impact of risk based on the extracted features.

[0105] "Generative AI" refers to technologies and algorithms that use artificial intelligence to generate appropriate responses and countermeasures from specific input data.

[0106] "Feedback" refers to information about the results and effectiveness of a proposed measure that is returned to the system and used for the next process improvement.

[0107] "Means for updating the model" refers to technologies and methods that adjust the parameters of the system's risk assessment model and generation AI based on collected feedback data, thereby improving the accuracy of the next risk assessment and countermeasure generation.

[0108] "User terminal" means a device used by a user to enter data, check results, and select countermeasures, and examples include PCs, tablets, and smartphones.

[0109] This invention relates to a system for efficiently managing a company's operational risks. The system aims to integrate and automate a series of processes including data collection, data analysis, risk assessment, countermeasure proposal generation, and implementation and feedback of the countermeasure proposals.

[0110] System configuration

[0111] Data collection

[0112] The user inputs risk-related data via a terminal.

[0113] Users input risk-related information via terminal devices such as PCs, tablets, and smartphones. For example, they enter data such as "supply chain interruptions," "risk of legal violations," and "employee health status" into a dedicated input form. The terminal then verifies this data and sends it to the server.

[0114] Data storage and preprocessing

[0115] The server receives the data and stores it in the database

[0116] The server receives risk-related data sent from the device, performs error checks and format validation, and if there are no problems, stores the data in a database. At this time, the data is temporarily labeled.

[0117] The server preprocesses the data

[0118] The server performs preprocessing on the stored data, including missing value imputation, removal of unnecessary items, data normalization, and outlier detection and correction. The preprocessed data is then used for feature extraction and risk analysis.

[0119] Feature extraction and risk assessment

[0120] The server extracts features

[0121] From the preprocessed data, the server extracts features necessary for risk assessment, such as "frequency of delivery delays in the past year," "conditions for legal violations," and "fluctuations in employee health status." These features are generated as a new dataset.

[0122] The server performs the risk assessment

[0123] The server uses the feature dataset to apply machine learning algorithms (e.g., random forests and support vector machines) and statistical models to analyze the probability of risk occurrence and its impact. The analysis results are stored in a database as risk assessment results.

[0124] Risk countermeasure generation

[0125] The server generates countermeasures using generation AI

[0126] Based on the risk assessment results, the server applies a generative AI model (e.g., GPT-3 (registered trademark), BERT) to generate optimal risk countermeasures. Specific countermeasures are generated, such as "suggesting alternative suppliers," "increasing inventory," and "introducing employee health programs."

[0127] The server sends the countermeasures to the device.

[0128] The generated risk countermeasures are sent from the server to the user's device, where the user can check the countermeasures.

[0129] Implementation of countermeasures and feedback

[0130] The user checks and selects countermeasures via the device.

[0131] The user checks the countermeasures sent via the terminal and selects the countermeasure that they think is most effective. The selected countermeasure is then implemented.

[0132] The user enters the execution results into the terminal and sends them to the server.

[0133] After implementing the measures, the user inputs the results and effects from the device and provides feedback to the server. The feedback is provided as quantitative and qualitative information.

[0134] The server updates the model based on the feedback

[0135] The server stores the feedback data in a database and updates the parameters of the risk assessment model and the generation AI based on this data, making the next risk analysis and countermeasure generation more accurate and effective. The feedback content is also used as training data for the model.

[0136] Specific examples

[0137] For example, when a manufacturing company manages supply chain risks, the user inputs data on delivery delays from key suppliers, supply chain interruption risks, and information on alternative suppliers from their device. The server then preprocesses this data, performs risk analysis, and uses generative AI to propose countermeasures. The user then confirms and implements the proposed alternative supplier selection plan, thereby minimizing risk. This process is returned to the server through feedback, improving the accuracy and effectiveness of the entire system.

[0138] Examples of prompt statements

[0139] Here are some examples of prompts to input to a generative AI model:

[0140] "We are facing an increased risk of supply chain interruption and would like to find alternative suppliers. Our current supplier has repeatedly delayed deliveries and is running low on inventory. Please generate the optimal solution for this."

[0141] By using this prompt, the generative AI can generate specific risk countermeasures and provide them to the user.

[0142] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0143] Step 1:

[0144] The user inputs risk-related data via a terminal.

[0145] Input: Users use a device (e.g., PC, tablet, smartphone) to enter risk-related information (e.g., supply chain interruptions, risk of legal violations, employee health status) into a dedicated input form.

[0146] Data processing: The terminal verifies the format of the input data and reformats it if necessary. It also performs validation to improve data reliability (e.g., checking required fields, verifying data types).

[0147] Output: Sends the formatted data to the server.

[0148] Step 2:

[0149] The server receives the data and stores it in the database

[0150] Input: Risk-related data sent from the device.

[0151] Data processing: The server performs error checks and format validation on the received data. If there are no problems, the data is stored in a temporary buffer.

[0152] Output: Temporarily label the passed data when saving it to the database.

[0153] Step 3:

[0154] The server preprocesses the data

[0155] Input: Risk-related data stored in the database.

[0156] Data processing: The server performs preprocessing on the stored data, including missing value imputation (e.g., mean imputation or mode imputation), removal of unnecessary items, data normalization (e.g., scaling numerical data to the range 0 to 1), and outlier detection and correction.

[0157] Output: Generate a preprocessed dataset in a format that can be used in the next step.

[0158] Step 4:

[0159] The server extracts features

[0160] Input: The preprocessed dataset.

[0161] Data processing: The server selects and extracts features necessary for risk assessment. For example, it actively extracts features such as "frequency of delivery delays in the past year," "conditions for legal violations," and "fluctuations in employee health."

[0162] Output: Generate a feature dataset and prepare it for input into a risk assessment model.

[0163] Step 5:

[0164] The server performs the risk assessment

[0165] Input: Feature dataset.

[0166] Data calculation: The server applies machine learning algorithms (e.g., random forests, support vector machines) and statistical models to calculate the probability and impact of risks, generating an assessment score for each risk.

[0167] Output: The risk assessment results are stored in a database and formatted for use in the next step.

[0168] Step 6:

[0169] The server generates countermeasures using generation AI

[0170] Input: Risk assessment results.

[0171] Data processing: Based on the analysis results, the server applies a generative AI model (e.g., GPT-3, BERT) to generate optimal risk countermeasures. Specifically, a prompt is input into the generative AI, and a countermeasure is obtained as a response.

[0172] Output: Generate a list of the generated risk treatments and format it for delivery to the user.

[0173] Step 7:

[0174] The server sends the countermeasures to the device.

[0175] Input: Generated risk treatment proposals.

[0176] Data transmission: The server sends the countermeasures plan to the user's device. Data is also encoded and encrypted during transmission to ensure secure communication.

[0177] Output: Provide countermeasures in a format that can be displayed on the user's device.

[0178] Step 8:

[0179] The user checks and selects countermeasures via the device.

[0180] Input: Risk treatment proposal sent from the server.

[0181] How it works: The user reviews the proposed countermeasures via their device and selects the countermeasure they think is best. The selected countermeasure is then sent as feedback from the device to the server.

[0182] Output: Information about the countermeasures selected by the user.

[0183] Step 9:

[0184] The user enters the execution results into the terminal and sends them to the server.

[0185] Input: The results of the implemented countermeasures and their effectiveness.

[0186] Data transmission: The user inputs data (quantitative and qualitative information) on the execution results and effects using the terminal and transmits it to the server. The input data is also validated.

[0187] Output: Sending feedback information.

[0188] Step 10:

[0189] The server updates the model based on the feedback

[0190] Input: Feedback data.

[0191] Data calculation: The server uses the feedback data to update the parameters of the risk assessment model and generative AI, specifically adding the feedback data as training data for the model and retraining it.

[0192] Output: Updated risk assessment model and generative AI model.

[0193] The above is a specific processing flow of the risk management system based on the present invention, and provides a detailed explanation of the specific operations and data flow performed at each step.

[0194] (Application example 1)

[0195] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0196] Operational risk management in factories requires rapid and accurate response to a wide range of risk factors. However, in current systems, risk data collection and analysis, and the generation and implementation of countermeasures are carried out separately, and there is no integrated automated system. This makes efficient risk management difficult and sometimes prevents timely responses. Furthermore, there is a lack of systems that can check risks and implement countermeasures in real time, and delays in the communication of information between on-site workers and managers are also an issue.

[0197] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0198] In this invention, the server includes: a means for inputting risk-related data; a means for transmitting the data to the server; a means for storing the data in a database; a means for preprocessing the stored data; a means for extracting features from the preprocessed data; a means for analyzing risks based on the extracted features; a means for applying a generation AI to generate risk countermeasures based on the analysis results; a means for transmitting the generated countermeasures to a user terminal; a means for the user to review and select the countermeasures; a robot control means for automatically executing the selected countermeasure; a means for feeding back the execution results of the user-selected countermeasure to the server; a means for updating the model based on the feedback; and a means for displaying risks in real time via smart glasses or a head-mounted display. This enables integrated and automated operational risk management in factories, enabling rapid and accurate responses to risks. Furthermore, field workers can check risks in real time through the smart glasses or a head-mounted display, enabling appropriate information sharing between them and managers.

[0199] "Risk-related data" refers to information on all risk factors that may occur in the operation of a factory, such as malfunctions of machinery, changes in the work environment, and safety-related events.

[0200] A "server" is a high-performance computing device for managing and analyzing data in a factory.

[0201] A "database" is a system for systematically storing and managing collected data.

[0202] "Preprocessing" refers to processes such as filling in missing values, removing unnecessary items, and normalizing data in order to convert collected data into a format suitable for analysis.

[0203] "Features" are important attributes or indicators extracted from data that are necessary for analysis.

[0204] "Risk analysis" is a process of evaluating the probability of occurrence and the impact of risks based on extracted features.

[0205] "Generative AI" is an artificial intelligence model that automatically generates optimal risk countermeasures based on the results of risk analysis.

[0206] A "user terminal" is an electronic device that allows factory floor workers and managers to access the system and check and input information.

[0207] "Robot control means" is a system that operates and manages the robot to automatically execute the selected countermeasure.

[0208] "Feedback" is the process of returning the results and effects of the actions taken to the server.

[0209] "Smart glasses" are wearable devices that factory floor workers use to check risks in real time.

[0210] A "head-mounted display" is a display device used by factory managers to remotely grasp the situation on-site and issue instructions.

[0211] This invention relates to a system for managing operational risks in factories. Specifically, it provides a method for efficiently managing operational risks in factories by automating a series of processes, from data collection, data preprocessing, risk analysis, countermeasure proposal generation, and countermeasure implementation and feedback.

[0212] Hardware and software used

[0213] Smart glasses: Wearable devices (e.g., Google® Glass® or Vuzix Blade) that allow field workers to see risks in real time.

[0214] Head-mounted display: A display device that allows managers to remotely grasp the situation on-site (e.g., Microsoft (registered trademark) HoloLens (registered trademark) or Oculus Quest).

[0215] Robot: A device for automatically carrying out selected countermeasures (e.g., ABB's Yumi).

[0216] Server: A high-performance computing device that uses databases, machine learning libraries (TENSORFLOW (registered trademark), PyTorch), and data processing libraries (Pandas, Numpy).

[0217] Database: A data management system such as MySQL (registered trademark).

[0218] System configuration and operation details

[0219] 1. Data Collection:

[0220] Through the terminal (smart glasses), field workers input risk-related data, such as abnormal machine operation, temperature changes, abnormal vibrations, and the presence or absence of obstacles. This data is also acquired from the smart glasses' sensors. The collected data is then transmitted to a server via wireless communication.

[0221] 2. Data preprocessing:

[0222] The server stores the received data in a database, and then uses Pandas and Numpy to perform preprocessing such as filling in missing values, removing unnecessary items, and normalizing the data.

[0223] 3. Feature extraction:

[0224] After preprocessing, features necessary for risk assessment are extracted using Scikit-learn and TensorFlow. For example, important features include machine operating hours, ambient temperature and humidity, and changes in the work environment.

[0225] 4. Risk Analysis:

[0226] Based on the extracted features, a deep learning model is used to evaluate the probability of risk occurrence and its impact. This analysis uses machine learning libraries such as TensorFlow and PyTorch.

[0227] 5. Countermeasure generation:

[0228] Generative AI (such as the GPT series) is used to generate optimal risk countermeasures based on the results of risk analysis, such as proposing alternative work procedures or machine repair schedules.

[0229] 6. Implementation and feedback of countermeasures:

[0230] The generated countermeasure proposals are sent to the terminal, where the user (worker or manager) reviews them and selects the most appropriate one. The selected countermeasure is then automatically executed by the robot control system. The results and effects of the implemented countermeasures are fed back to the server and used to improve the next risk assessment model.

[0231] Examples of concrete examples and prompts

[0232] Specific examples

[0233] 1. Example of data entered into smart glasses:

[0234] temperature 35 degrees

[0235] Abnormal vibration

[0236] Check for obstacles in the work area

[0237] 2. Proposed measures:

[0238] Cooling system inspection

[0239] Quickly remove obstacles

[0240] Maintenance for abnormal vibration

[0241] Prompt Sentence Examples

[0242] "The temperature on the factory floor has reached 35 degrees, and abnormal vibrations are occurring. Also, an obstacle has been found in the work area. Please generate optimal countermeasures to address these risks."

[0243] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0244] Step 1:

[0245] Data collection

[0246] Through the terminal (smart glasses), the user (field worker) inputs risk-related data. The input data includes temperature, vibration abnormalities, and the presence or absence of obstacles in the work area. Data is also collected from the smart glasses' sensors. This data is transmitted to the server via wireless communication.

[0247] Input: Risk-related data entered by workers through terminals and sensor data from smart glasses.

[0248] Output: Collected risk-related data is sent to a server.

[0249] Step 2:

[0250] Data Preprocessing

[0251] The server stores the data received from the terminal in a database, and then uses Pandas and Numpy to perform preprocessing such as filling in missing values, removing unnecessary items, and normalizing the data.

[0252] Input: Collected risk-related data.

[0253] Output: Clean, pre-processed data.

[0254] Step 3:

[0255] Feature extraction

[0256] The server uses Scikit-learn and TensorFlow to extract the features necessary for risk assessment from the preprocessed data, including machine operating hours, ambient temperature, humidity, and changes in the work environment.

[0257] Input: Preprocessed data.

[0258] Output: Extracted features.

[0259] Step 4:

[0260] Risk Analysis

[0261] The server uses a deep learning model based on the extracted features to assess the probability and impact of a risk, using machine learning libraries such as TensorFlow and PyTorch.

[0262] Input: Extracted features.

[0263] Output: Probability and impact of risk.

[0264] Step 5:

[0265] Countermeasures generation

[0266] The server uses a generative AI (such as the GPT series) to generate optimal risk countermeasures based on the risk analysis results, including alternative work procedures and machine repair schedules.

[0267] Input: Risk analysis results.

[0268] Output: Generated countermeasures.

[0269] Step 6:

[0270] Implementation of countermeasures and feedback

[0271] The generated countermeasure proposals are sent to the device (smart glasses or head-mounted display), where the user can review them and select the most appropriate one. The selected countermeasure is then automatically implemented by the robot control system. The results and effects of the implemented countermeasures are fed back to the server and used to improve the risk assessment model.

[0272] Input: Generated countermeasures.

[0273] Output: Results of implemented measures and feedback data.

[0274] Specific prompt examples

[0275] "The temperature on the factory floor has reached 35 degrees, and abnormal vibrations are occurring. Also, an obstacle has been found in the work area. Please generate optimal countermeasures to address these risks."

[0276] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0277] This invention relates to a system for improving the accuracy and effectiveness of risk management in a company's operational risk management by combining risk information collection and analysis, the generation and implementation of countermeasures, and an emotion engine that recognizes user emotions.

[0278] System configuration

[0279] The system consists of the following components:

[0280] A means of inputting risk-related data

[0281] A means of sending data to the server

[0282] A means of storing data in a database

[0283] A means of preprocessing the stored data

[0284] A means of extracting features from preprocessed data

[0285] A method for analyzing risk based on extracted features

[0286] A means of applying generative AI to generate risk countermeasures based on analysis results

[0287] A means for transmitting the generated countermeasures to the user terminal

[0288] A means for users to check and select countermeasures

[0289] A means of feeding back the results of the selected measures to the server

[0290] A means of updating the model based on feedback

[0291] Emotion engine that recognizes user emotions

[0292] Program Processing Details

[0293] The user inputs risk-related data via a terminal.

[0294] Users use a terminal to input risk-related data, such as supplier delivery delays, legal violation risks, and employee health status. An emotion engine that recognizes the user's emotions also collects the user's emotional data during this process.

[0295] The device sends input data and emotion data to the server.

[0296] The device sends the input data and the emotion data collected by the emotion engine to the server, where the user's emotion data is stored along with the risk-related data.

[0297] The server stores the data in a database and performs preprocessing.

[0298] The server stores the received data and emotion data in a database, where preprocessing is performed, such as filling in missing values, removing unnecessary items, and normalizing the data.

[0299] The server extracts features and analyzes risks

[0300] From the preprocessed data, the server extracts the features necessary for risk assessment. Based on these features, it runs an algorithm to assess the probability of risk occurrence and its impact.

[0301] The server generates risk countermeasures using generation AI

[0302] Based on the analysis results, the server applies a generative AI to generate risk countermeasures. The generative AI also takes the user's emotional data into account when generating countermeasures, so if the user is feeling excessive stress, for example, it will propose countermeasures to alleviate that situation.

[0303] The server sends the generated countermeasures and emotion data to the user terminal.

[0304] The generated countermeasures are sent from the server to the user's device, where the user can check them. The user's emotional data is also displayed, which can be used as a reference when selecting a countermeasure.

[0305] The user checks the proposed measures and selects one

[0306] The user can then use their device to check the generated countermeasures and select which ones to implement. This process allows the selection of countermeasures that reflect the user's feelings.

[0307] The user executes the selected measure and enters the results into the device.

[0308] After the selected countermeasure is implemented, the user provides feedback on the results and effects of the countermeasure to the server from the terminal. This feedback includes not only the results of the countermeasure but also the user's emotional data.

[0309] The server updates the model based on the feedback

[0310] The server stores the feedback data and emotion data in a database and uses them to update the parameters of the risk assessment model and the generation AI, thereby improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[0311] Specific examples

[0312] For example, when a manufacturing company performs risk management for its supply chain, the user inputs data such as "data regarding delivery delays from major suppliers" and "risk of supply chain interruptions" into a terminal. In addition, an emotion engine simultaneously collects user emotional data such as "stress" and "anxiety." The server preprocesses this data, performs risk analysis, and then uses generative AI to generate countermeasures that take the emotional data into account. The user reviews these countermeasures, selects the most appropriate one, and implements it. After implementation, the results and emotional data are sent back to the server and used to update the model.

[0313] In this way, the present invention can realize risk management that takes into account users' emotions, and effectively minimize potential losses from enterprise risks.

[0314] The processing flow will be explained below.

[0315] Step 1:

[0316] Users use a terminal to input risk-related data, such as supplier delivery delays, legal violation risks, and employee health status. At the same time, an emotion engine analyzes the user's facial expressions and tone of voice to collect emotional data.

[0317] Step 2:

[0318] The device validates the risk-related and emotion data entered and sends it to the server, where validation includes checking the data format and filtering out invalid data.

[0319] Step 3:

[0320] The server stores the risk-related data and emotion data sent from the device in a database, where the data is appropriately labeled.

[0321] Step 4:

[0322] The server performs preprocessing on the stored risk-related data, including filling in missing values, removing unnecessary items, and normalizing the data. The same preprocessing is also performed on the emotion data, preparing it in a format suitable for analysis.

[0323] Step 5:

[0324] The server extracts features necessary for risk assessment from the preprocessed data. Features extracted include, for example, "frequency of past delivery delays," "conditions for legal violations," and "results of employee health assessments." Emotional features such as "stress level," "happiness level," and "anxiety level" are extracted from the emotion data.

[0325] Step 6:

[0326] The server runs a risk analysis algorithm based on the extracted features and emotion features. This algorithm includes machine learning models and statistical models to evaluate the probability of risk occurrence and its impact.

[0327] Step 7:

[0328] Based on the results of the risk analysis, the server uses a generative AI to generate risk countermeasures. The generative AI also takes emotional features into account to generate countermeasures that take into consideration the user's mental health. For example, it may suggest reallocating work to reduce stress or a specific action plan to mitigate risks.

[0329] Step 8:

[0330] The server transmits the generated risk countermeasures and related emotion data to the user terminal, and the user uses the terminal to check the countermeasures.

[0331] Step 9:

[0332] The user checks the proposed countermeasures through the device and selects the most appropriate one. The selected countermeasure is then sent back to the server from the device.

[0333] Step 10:

[0334] Implement the user-selected measures, such as finding alternative suppliers, increasing inventory, or implementing employee wellness programs within the company.

[0335] Step 11:

[0336] The user uses the device to input the results of the measures taken, and the emotion engine also collects the user's emotional data at that time and sends it to the server to provide feedback.

[0337] Step 12:

[0338] The server stores the received feedback and emotion data in a database, which is then used for the next risk assessment and countermeasure generation.

[0339] Step 13:

[0340] The server updates the risk assessment model and the parameters of the generative AI based on the feedback and emotion data, thereby improving the accuracy and effectiveness of the next analysis and generated countermeasures.

[0341] Example 2

[0342] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0343] In operational risk management in companies, there is a need for the collection and analysis of risk information, the generation and implementation of countermeasures, and flexible countermeasure proposals that take user emotions into account. However, conventional systems have difficulty effectively incorporating user emotions, resulting in insufficient accuracy and effectiveness in risk management. Therefore, there is a need to develop a system that takes user emotional data into account, flexibly generates risk countermeasures, and utilizes the results of their implementation as feedback.

[0344] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0345] In this invention, the server includes a means for a user to input risk-related data, a means for transmitting the data and the user's emotion data to the server, and a means for storing the data and the emotion data in a database. This allows the user's emotions to be taken into account in risk analysis and countermeasure generation, enabling more appropriate risk countermeasures. Furthermore, by feeding back the execution results of the selected countermeasures to the server along with the emotion data, the model is updated, improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[0346] "User" refers to a company employee or system user who inputs risk-related data and checks and selects countermeasures.

[0347] A "terminal" is an electronic device through which a user inputs risk-related data and emotion data and transmits the data to a server.

[0348] The "server" is a central management system that receives, stores, analyzes risk-related data and emotion data, and generates countermeasures.

[0349] A "database" is a storage device for storing risk-related data and emotion data received by the server.

[0350] An "emotion engine" is a software or hardware component that recognizes a user's emotions in real time and collects them as data.

[0351] "Preprocessing" refers to the process of imputing missing values, removing unnecessary items, and normalizing the data received.

[0352] "Feature extraction" is the process of extracting elements from data that are important for risk assessment and analysis.

[0353] "Risk analysis" is a process of evaluating the probability of occurrence and the impact of a risk based on the extracted feature quantities.

[0354] "Generative AI" is an artificial intelligence algorithm that automatically generates countermeasures based on the results of risk analysis, taking into account the user's emotions.

[0355] "Feedback" is a process in which the execution results of the measures selected by the user and emotion data are sent to the server.

[0356] "Model updating" is a process in which the parameters of the risk assessment model and generation AI are modified based on feedback data to improve the accuracy and effectiveness of the next analysis and countermeasure generation.

[0357] This invention relates to a system for improving the accuracy and effectiveness of risk management in a company's operational risk management by combining risk information collection and analysis, the generation and implementation of countermeasures, and an emotion engine that recognizes user emotions. Specific embodiments for implementing this invention will be described below.

[0358] Hardware and Software Configuration

[0359] The system consists of the following components:

[0360] Device: An electronic device through which a user enters risk-related data. Examples include a personal computer, tablet, or smartphone.

[0361] Server: A central management system that receives, stores, analyzes, and generates countermeasures. A server with a high-performance processor and large storage capacity is recommended.

[0362] Database: A storage device for saving data received by the server. A relational database management system (RDBMS) is used.

[0363] Emotion engine: A software or hardware component that recognizes a user's emotions in real time. For example, there are emotion recognition systems that use facial recognition technology or voice analysis technology.

[0364] System operation details

[0365] The user inputs risk-related data via a terminal.

[0366] Users input risk-related data from their own devices, such as supplier delivery delays, legal violation risks, and employee health status. In addition, an emotion engine recognizes and collects the user's emotions (e.g., stress, anxiety, etc.) in real time as they input their data.

[0367] The device sends input data and emotion data to the server.

[0368] The terminal transmits the input risk-related data and emotion data to the server, whereby the risk-related data and the user's emotion data are stored.

[0369] The server stores the data in a database and performs preprocessing.

[0370] The server stores the received risk data and emotion data in a database, then performs preprocessing on the stored data, such as filling in missing values, removing unnecessary items, and normalizing the data.

[0371] The server extracts features and analyzes risks

[0372] From the preprocessed data, the server extracts the features necessary for risk assessment, then runs an algorithm to evaluate the probability of risk occurrence and its impact.

[0373] The server generates risk countermeasures using generation AI

[0374] Based on the analysis results, the server applies generative AI to generate risk countermeasures. This also takes into account the user's emotional data, so if the user is feeling excessive stress, for example, countermeasures to alleviate that situation will be proposed.

[0375] The server sends the generated countermeasures and emotion data to the user terminal.

[0376] The generated countermeasures are sent from the server to the user's device, where the user can check them. Emotion data is also displayed, which can be used as a reference when selecting a countermeasure.

[0377] The user checks the proposed measures and selects one

[0378] The user can use the device to check the generated countermeasures and select the countermeasures to be implemented. This process allows the selection of countermeasures that reflect the user's feelings.

[0379] The user executes the selected measure and enters the results into the device.

[0380] After the selected countermeasure is executed, the user sends feedback of the result and effect from the terminal to the server. This feedback includes the result of the execution of the countermeasure and the user's emotional data.

[0381] The server updates the model based on the feedback

[0382] The server stores the feedback data and emotion data in a database and uses them to update the risk assessment model and the parameters of the generation AI, thereby improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[0383] Examples of concrete examples and prompts

[0384] For example, when a manufacturing company performs risk management for its supply chain, the user inputs data such as "data regarding delivery delays from major suppliers" and "risk of supply chain interruptions" into a terminal. In addition, an emotion engine simultaneously collects user emotional data such as "stress" and "anxiety." The server preprocesses this data, performs risk analysis, and then uses generative AI to generate countermeasures that take the emotional data into account. The user reviews these countermeasures, selects the most appropriate one, and implements it. After implementation, the results and emotional data are sent back to the server and used to update the model.

[0385] Prompt Sentence Examples

[0386] "Please provide data on delivery delays from your key suppliers. Also, please rate your current stress level on a scale of 1 to 10."

[0387] "Please provide detailed information about the risk of supply chain interruption and indicate your level of concern about that risk on a scale of 0 to 100."

[0388] In this way, the present invention realizes risk management that takes into account users' emotions, and can effectively minimize potential losses from enterprise risks.

[0389] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0390] Step 1:

[0391] The user inputs risk-related data via a terminal.

[0392] Users use a terminal to input risk-related data. This data includes information such as "supplier delivery delays," "risk of legal violations," and "employee health status." The emotion engine recognizes and collects the user's emotions (such as "stress" or "anxiety") in real time as they input data. Specifically, users enter data using a keyboard or touchscreen, and the camera and microphone analyze facial expressions and tone of voice to collect emotional data.

[0393] Input: Risk-related data, user sentiment data

[0394] Output: None (passed to next step)

[0395] Step 2:

[0396] The device sends input data and emotion data to the server.

[0397] The device sends the risk data entered by the user and the emotion data collected by the emotion engine in a specific format (e.g., JSON format) to the server, where the data is checked for integrity and any necessary encryption is applied.

[0398] Input: Risk-related data, user sentiment data

[0399] Output: Data sent to the server

[0400] Step 3:

[0401] The server stores the data in a database and performs preprocessing.

[0402] The server stores the received risk and emotion data in a database. It then performs preprocessing on the data, such as filling in missing values, removing unnecessary items, and normalizing the data. For example, if there are missing values, they are filled in with the average value, unnecessary items are removed, and numerical data is standardized.

[0403] Input: Data sent to the server

[0404] Output: Preprocessed data

[0405] Step 4:

[0406] The server extracts features and analyzes risks

[0407] From the preprocessed data, the server extracts the features necessary for risk assessment. At this time, it applies data analysis algorithms to evaluate the probability of risk occurrence and its impact. For example, it uses a machine learning model to calculate a risk score and extract important variables necessary for risk assessment.

[0408] Input: Preprocessed data

[0409] Output: Risk analysis results (feature amount, risk occurrence probability, impact)

[0410] Step 5:

[0411] The server generates risk countermeasures using generation AI

[0412] The server applies a generative AI based on the risk analysis results to generate risk countermeasures. This generative AI also takes into account the user's emotional data to create appropriate risk countermeasures. For example, the generative AI can propose multiple countermeasures using the risk score and the user's emotional data as input.

[0413] Input: Risk analysis results, user emotion data

[0414] Output: Risk countermeasure proposal

[0415] Step 6:

[0416] The server sends the generated countermeasures and emotion data to the user terminal.

[0417] The generated countermeasures are sent from the server to the user's device, where the user can check the countermeasures and view the emotional data.

[0418] Input: Risk countermeasures, emotional data

[0419] Output: Countermeasures and emotion data sent to the device

[0420] Step 7:

[0421] The user checks the proposed measures and selects one

[0422] The user can then use their device to review the proposed countermeasures and select the appropriate one. This process also takes into account emotional data, allowing the user to select the countermeasure that best suits their situation. For example, the user can select the countermeasure that seems easiest to implement from the provided countermeasures.

[0423] Input: Countermeasures and emotion data sent to the device

[0424] Output: Selected countermeasures

[0425] Step 8:

[0426] The user executes the selected measure and enters the results into the device.

[0427] The user implements the selected countermeasure and sends feedback on the results and effects from the terminal to the server. This feedback includes the results of the countermeasure and emotional data. For example, the success rate of the countermeasure and the emotional state after implementation can be input.

[0428] Input: Execution results of selected countermeasures and emotion data

[0429] Output: Feedback data sent to the server

[0430] Step 9:

[0431] The server updates the model based on the feedback

[0432] The server stores the received feedback data in a database and uses it to update the parameters of the risk assessment model and the generation AI, thereby improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[0433] Input: Feedback data

[0434] Output: Updated risk assessment model and generative AI

[0435] (Application example 2)

[0436] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0437] Conventional operational risk management systems in companies have the problem that collecting and analyzing risk data takes time, making it difficult to respond in real time. In addition, because they do not take into account the emotions of workers, measures to reduce stress and anxiety are insufficient, making it difficult to improve production efficiency and reduce workers' psychological stress. This has resulted in the accuracy and effectiveness of risk management being insufficient, placing a strain on corporate operations.

[0438] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for inputting risk-related data, means for transmitting the data to the server, means for saving the data in a database, means for preprocessing the saved data, means for extracting features from the preprocessed data, means for analyzing risks based on the extracted features, means for applying a generation AI that generates risk countermeasures based on the analysis results, means for transmitting the generated countermeasures to a user terminal, means for the user to confirm and select the countermeasures, means for feeding back the execution results of the selected countermeasures to the server, means for updating the model based on the feedback, means for collecting risk data and emotion data from industrial robots and worker terminals in real time, means for generating and presenting countermeasures taking the data and emotion data into consideration, and means for displaying the countermeasures on the industrial robot's display device or the worker's terminal. This makes it possible to collect and analyze risk data and worker emotion data in real time and quickly implement highly accurate risk countermeasures.

[0439] "Risk-related data" refers to data necessary to assess a company's operational risks, including the condition of equipment, the progress of processes, and the health status of personnel.

[0440] A "server" is a computer system that receives, stores, and analyzes risk data via a network.

[0441] A "database" is a data storage system for systematically storing risk-related data and emotion data.

[0442] "Preprocessing" is the process of imputing missing values, removing unnecessary items, and normalizing data to prepare it in an analyzable format.

[0443] "Features" are extracted important data points and indicators necessary for risk analysis.

[0444] "Analyzing risk" means evaluating the probability and impact of a risk based on collected data and predicting potential dangers.

[0445] "Generative AI" is an artificial intelligence system that uses machine learning and AI technology to generate risk countermeasures.

[0446] A "user terminal" is a device that a user uses to input risk data and check countermeasure proposals, such as a smartphone or tablet.

[0447] "Feedback" refers to returning data to the server about the results and effects of the measures taken by the user.

[0448] "Updating the model" means improving and adjusting the parameters of the analytical model and generation AI based on feedback data, thereby improving the accuracy of the next analysis and countermeasure generation.

[0449] "Emotion data" is data that quantifies the user's psychological state, such as stress level and anxiety level.

[0450] An "industrial robot" is a robot used to assist or automate work within a factory, and is a mechanical device equipped with various sensors that has the ability to collect risk data in real time.

[0451] System Overview

[0452] The operational risk management system of this invention collects risk factors and worker emotion data in a factory in real time, and analyzes and generates countermeasures based on this data. The system is composed of various components, including industrial robots, worker terminals, servers, databases, and generation AI.

[0453] Data collection

[0454] Users and industrial robots collect risk-related data in real time within the factory, including equipment failure status, the progress of production processes, and the health status of workers. Workers also use smart glasses and terminals to input emotional data such as their stress and anxiety levels.

[0455] Data transmission and storage

[0456] The terminals and industrial robots send the collected risk and emotion data to a server, where it is stored in a database. The server then performs preprocessing on the stored data, such as filling in missing values, removing unnecessary items, and normalizing the data.

[0457] Data analysis

[0458] The server extracts features from the preprocessed data and performs analysis to evaluate the probability of risk occurrence and its impact, using machine learning algorithms and data analysis software (e.g., pandas, numpy, scikit-learn).

[0459] Generate countermeasures

[0460] Based on the analysis results, the server uses a generation AI to generate risk countermeasures. The generation AI also takes into account the user's emotional data when generating countermeasures, so if the user is feeling excessive stress, for example, it will propose countermeasures to alleviate that situation.

[0461] Proposing and implementing countermeasures

[0462] The generated countermeasures are sent from the server to the user's terminal or the display device of the industrial robot, where the user can check the countermeasures and select the most appropriate one. The selected countermeasure is then implemented.

[0463] Feedback and Model Updates

[0464] The user feeds back the results of the countermeasures they have implemented to the server via their device. This feedback includes not only the results of the countermeasures but also the user's emotional data. The server updates the model based on the feedback data and emotional data, improving the accuracy and effectiveness of the next analysis and generation of countermeasures.

[0465] Specific examples

[0466] For example, when a major machine breaks down in a factory, the industrial robot sends real-time data on the situation to a server. At the same time, the worker sends their stress level to the server through smart glasses. The server analyzes this data and uses generative AI to generate appropriate repair plans and measures to reduce worker stress, such as break suggestions. These measures are then displayed on the industrial robot's display and the worker's device.

[0467] Prompt Sentence Examples

[0468] "When equipment failure occurs in a factory, please collect information on the situation in real time and generate optimal repair measures. Please also suggest measures to reduce worker stress."

[0469] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0470] Step 1:

[0471] Users and industrial robots collect risk-related data within factories. Specifically, industrial robots use sensors to obtain information about the equipment's operating status and manufacturing processes, while workers use smart glasses to input their own health and psychological state (stress level, anxiety level). The data input here covers a wide range of topics, including equipment status, production line delays, and worker fatigue levels.

[0472] Step 2:

[0473] The terminal sends the collected risk data and emotion data to a server. The terminal then compiles and packages this data and sends it to the server via the network. The transmitted data is input to the server in the form of information such as the operating status of equipment, production delays, and worker stress levels.

[0474] Step 3:

[0475] The server stores the received data in a database. The server stores the transmitted risk and emotion data in the database and tags them appropriately. The stored data also includes time information and the data source (equipment, worker, etc.).

[0476] Step 4:

[0477] The server preprocesses the stored data by removing unnecessary items, filling in missing values, and normalizing the data. The server first removes meaningless and outliers from the data, then infers and fills in missing data points. Finally, the data is converted into a unified format and prepared for feature extraction.

[0478] Step 5:

[0479] The server extracts features from the preprocessed data. After data normalization is complete, the server uses a machine learning algorithm to extract important features necessary for risk assessment. The output of this process is input data for assessing the probability of risk occurrence and its impact.

[0480] Step 6:

[0481] The server analyzes risks based on the extracted features. Using a risk assessment model based on the features, the server calculates and assesses the probability of risk occurrence and the impact. This analysis clarifies which risks are likely to occur and the extent of their impact.

[0482] Step 7:

[0483] The server uses a generative AI based on the analysis results to generate risk countermeasures. Based on the analysis results, the server uses generative AI (e.g., GPT-3 model) to automatically generate optimal risk countermeasures. Countermeasures that take emotional data into consideration are generated here. The generated countermeasures are specific action plans for the countermeasures.

[0484] Step 8:

[0485] The server sends the generated countermeasure plan to the user's terminal or the display device of the industrial robot. The generated countermeasure plan is sent from the server to the user's terminal or the display device of the industrial robot, where it can be checked by the user or worker. The presented countermeasure plan includes the implementation procedure and expected effects.

[0486] Step 9:

[0487] The user or worker reviews the proposed countermeasures and selects the most appropriate one. After the proposed countermeasures are displayed on a terminal or display device, the user or worker reviews the contents and selects the countermeasure to be implemented. The selected countermeasure is sent back to the server as confirmation of the selection.

[0488] Step 10:

[0489] The user implements the selected measure, and the results are fed back from the terminal to the server. After the user or worker implements the measure, the results, effectiveness, and emotional state at the time are entered into the terminal as feedback and sent to the server. This data is used to evaluate the effectiveness of the measure.

[0490] Step 11:

[0491] The server updates the model based on the feedback data. After receiving the feedback data, the server saves it in the database and updates the parameters of the risk assessment model and the generation AI. This update improves the accuracy of the next analysis and countermeasure generation.

[0492] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the 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.

[0493] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0494] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0495] [Second embodiment]

[0496] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0497] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0498] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0499] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0500] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0501] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0502] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0503] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0504] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[0505] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0506] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0507] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. 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."

[0508] This invention relates to a system for efficiently managing a company's operational risks. The system aims to automate the risk management process by integrating data collection, data analysis, risk assessment, countermeasure generation, countermeasure implementation, and feedback.

[0509] System configuration

[0510] Data collection

[0511] The user inputs risk-related data via a terminal.

[0512] The user uses the device to input various risk-related data, such as supply chain interruptions, legal violation risks, and employee health status. The device then verifies this information and sends it to the server.

[0513] Data storage and preprocessing

[0514] The server receives the data and stores it in the database

[0515] The server receives the data sent from the device and stores it in a database, along with a temporary label.

[0516] The server preprocesses the data

[0517] The server performs preprocessing on the stored data, which includes processes such as imputing missing values, removing unnecessary items, and normalizing the data.

[0518] Feature extraction and risk analysis

[0519] The server extracts features

[0520] After preprocessing, the features necessary for risk assessment are extracted from the data. For example, features such as "frequency of past delivery delays," "conditions for legal violations," and "results of employee health assessments" are extracted.

[0521] The server performs risk analysis

[0522] Based on the extracted features, the server performs risk analysis, using machine learning algorithms and statistical models to assess the probability of risk occurrence and its impact.

[0523] Risk countermeasure generation

[0524] The server generates countermeasures using generation AI

[0525] Based on the analysis results, the server applies generative AI to generate optimal risk countermeasures, such as suggesting alternative suppliers, increasing inventory, or implementing employee wellness programs.

[0526] The server sends the countermeasures to the device.

[0527] The generated risk countermeasure plan is sent from the server to the user's terminal, and the user can check the countermeasure plan.

[0528] Implementation of countermeasures and feedback

[0529] The user checks and selects countermeasures via the device.

[0530] The user uses the terminal to check the countermeasures received from the server, select the countermeasure that they think is most appropriate, and the selected countermeasure is then put into action.

[0531] The user enters the execution results into the terminal and sends them to the server.

[0532] After the countermeasures are implemented, the user sends feedback on the results and effects to the server from the terminal, and the effectiveness of the countermeasures is evaluated based on this feedback.

[0533] The server improves the model based on the feedback

[0534] The server stores the collected feedback data in a database and uses it to update the risk assessment model and parameters of the generative AI, making the next risk analysis and countermeasure generation more accurate and effective.

[0535] Specific examples

[0536] For example, when a manufacturing company manages supply chain risks, the user inputs data on delivery delays from key suppliers, supply chain interruption risks, and information on alternative suppliers into their terminals. The server then preprocesses this data, performs risk analysis, and uses generative AI to propose countermeasures. The user can then confirm and implement the proposed alternative supplier selection plan to minimize risk. This process is returned to the server through feedback, improving the accuracy and effectiveness of the entire system.

[0537] In this way, the present invention improves the efficiency of operational risk management for companies, enabling rapid and accurate risk assessment and countermeasure proposals, thereby enabling companies to minimize potential losses from risks.

[0538] The processing flow will be explained below.

[0539] Step 1:

[0540] Users use the terminal to input risk-related data, such as supplier delivery delays, legal violation risk, and employee health status.

[0541] Step 2:

[0542] The terminal validates the entered data, checking for format and content errors as necessary, and sends the validated data to the server.

[0543] Step 3:

[0544] The server stores the data received from the device in a database, where it is appropriately labeled when stored.

[0545] Step 4:

[0546] The server preprocesses the stored data, specifically by imputing missing values, removing unnecessary items, and normalizing the data.

[0547] Step 5:

[0548] The server extracts features necessary for risk assessment from the preprocessed data, such as the frequency of past delivery delays, conditions for legal violations, and employee health assessment results.

[0549] Step 6:

[0550] The server then runs algorithms for risk analysis based on the extracted features. These algorithms include machine learning models and statistical models, and use them to evaluate the probability of risk occurrence and its impact.

[0551] Step 7:

[0552] Based on the results of the risk analysis, the server uses generative AI to generate optimal risk countermeasures, such as proposing alternative suppliers, increasing inventory, or introducing employee health programs.

[0553] Step 8:

[0554] The server transmits the generated countermeasures to the user terminal, where the user can check the countermeasures.

[0555] Step 9:

[0556] The user uses the terminal to check the proposed countermeasures and select the countermeasure to be implemented. The terminal then feeds this selection information back to the server.

[0557] Step 10:

[0558] The user implements the measures selected by the user, such as finding an alternative supplier or increasing inventory within the company.

[0559] Step 11:

[0560] The user uses the terminal to input the results of the measures taken and send them to the server, which then provides feedback on the effectiveness of the measures.

[0561] Step 12:

[0562] The server stores the received feedback in a database, which can then be used for the next risk assessment and countermeasure generation.

[0563] Step 13:

[0564] The server updates the risk assessment model and the parameters of the generation AI based on the feedback data, thereby improving the accuracy and effectiveness of the next analysis and generated countermeasures.

[0565] Example 1

[0566] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0567] In today's corporate environment, efficient management of operational risk is crucial, but traditional manual risk management methods are insufficient. Rapid and accurate risk assessment and countermeasure proposals are therefore required. However, there is no integrated system for data collection, preprocessing, analysis, countermeasure generation, and feedback, making the risk management process fragmented and inefficient. As a result, companies often are unable to respond quickly to potential loss risks. Furthermore, an effective means of automating complex tasks such as preprocessing, feature extraction, risk analysis, and countermeasure generation is needed.

[0568] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0569] In this invention, the server includes a means for inputting risk-related information, a means for transmitting the information to the server, a means for storing the information in a database, a means for preprocessing the stored information, a means for extracting features from the preprocessed information, a means for assessing risk based on the extracted features, a means for applying a generation AI to generate risk countermeasures based on the assessment results, a means for transmitting the generated countermeasures to a user terminal, a means for the user to review and select the countermeasures, a means for feeding back the execution results of the selected countermeasures to the server, and a means for updating the model based on the feedback. This unifies the risk management process, enabling efficient and rapid automation of risk assessment and countermeasure proposals. Furthermore, by continuously improving the model based on feedback, the accuracy of risk assessment and countermeasure proposals can be improved.

[0570] "Risk-related information" refers to data collected to identify and manage potential risks in a company's operations, such as delivery delays, violations of laws and regulations, and employee health conditions.

[0571] A "server" is a computer system that receives, processes, and transmits information over a network.

[0572] A "database" is a system for storing and managing data in an orderly manner and retrieving it as needed.

[0573] "Preprocessing" refers to the process of processing data to prepare risk-related information in an analyzable format, and includes imputing missing values, removing unnecessary items, and normalizing data.

[0574] "Features" are characteristics or attributes of data selected and calculated for risk assessment and are used directly in analysis.

[0575] "Means for assessing risk" refers to the process and technology for calculating and assessing the probability of occurrence and impact of risk based on the extracted features.

[0576] "Generative AI" refers to technologies and algorithms that use artificial intelligence to generate appropriate responses and countermeasures from specific input data.

[0577] "Feedback" refers to information about the results and effectiveness of a proposed measure that is returned to the system and used for the next process improvement.

[0578] "Means for updating the model" refers to technologies and methods that adjust the parameters of the system's risk assessment model and generation AI based on collected feedback data, thereby improving the accuracy of the next risk assessment and countermeasure generation.

[0579] "User terminal" means a device used by a user to enter data, check results, and select countermeasures, and examples include PCs, tablets, and smartphones.

[0580] This invention relates to a system for efficiently managing a company's operational risks. The system aims to integrate and automate a series of processes including data collection, data analysis, risk assessment, countermeasure proposal generation, and implementation and feedback of the countermeasure proposals.

[0581] System configuration

[0582] Data collection

[0583] The user inputs risk-related data via a terminal.

[0584] Users input risk-related information via terminal devices such as PCs, tablets, and smartphones. For example, they enter data such as "supply chain interruptions," "risk of legal violations," and "employee health status" into a dedicated input form. The terminal then verifies this data and sends it to the server.

[0585] Data storage and preprocessing

[0586] The server receives the data and stores it in the database

[0587] The server receives risk-related data sent from the device, performs error checks and format validation, and if there are no problems, stores the data in a database. At this time, the data is temporarily labeled.

[0588] The server preprocesses the data

[0589] The server performs preprocessing on the stored data, including missing value imputation, removal of unnecessary items, data normalization, and outlier detection and correction. The preprocessed data is then used for feature extraction and risk analysis.

[0590] Feature extraction and risk assessment

[0591] The server extracts features

[0592] From the preprocessed data, the server extracts features necessary for risk assessment, such as "frequency of delivery delays in the past year," "conditions for legal violations," and "fluctuations in employee health status." These features are generated as a new dataset.

[0593] The server performs the risk assessment

[0594] The server uses the feature dataset to apply machine learning algorithms (e.g., random forests and support vector machines) and statistical models to analyze the probability of risk occurrence and its impact. The analysis results are stored in a database as risk assessment results.

[0595] Risk countermeasure generation

[0596] The server generates countermeasures using generation AI

[0597] Based on the risk assessment results, the server applies a generative AI model (e.g., GPT-3, BERT) to generate optimal risk countermeasures, such as suggesting alternative suppliers, increasing inventory, or implementing an employee health program.

[0598] The server sends the countermeasures to the device.

[0599] The generated risk countermeasures are sent from the server to the user's device, where the user can check the countermeasures.

[0600] Implementation of countermeasures and feedback

[0601] The user checks and selects countermeasures via the device.

[0602] The user checks the countermeasures sent via the terminal and selects the countermeasure that they think is most effective. The selected countermeasure is then implemented.

[0603] The user enters the execution results into the terminal and sends them to the server.

[0604] After implementing the measures, the user inputs the results and effects from the device and provides feedback to the server. The feedback is provided as quantitative and qualitative information.

[0605] The server updates the model based on the feedback

[0606] The server stores the feedback data in a database and updates the parameters of the risk assessment model and the generation AI based on this data, making the next risk analysis and countermeasure generation more accurate and effective. The feedback content is also used as training data for the model.

[0607] Specific examples

[0608] For example, when a manufacturing company manages supply chain risks, the user inputs data on delivery delays from key suppliers, supply chain interruption risks, and information on alternative suppliers from their device. The server then preprocesses this data, performs risk analysis, and uses generative AI to propose countermeasures. The user then confirms and implements the proposed alternative supplier selection plan, thereby minimizing risk. This process is returned to the server through feedback, improving the accuracy and effectiveness of the entire system.

[0609] Examples of prompt statements

[0610] Here are some examples of prompts to input to a generative AI model:

[0611] "We are facing an increased risk of supply chain interruption and would like to find alternative suppliers. Our current supplier has repeatedly delayed deliveries and is running low on inventory. Please generate the optimal solution for this."

[0612] By using this prompt, the generative AI can generate specific risk countermeasures and provide them to the user.

[0613] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0614] Step 1:

[0615] The user inputs risk-related data via a terminal.

[0616] Input: Users use a device (e.g., PC, tablet, smartphone) to enter risk-related information (e.g., supply chain interruptions, risk of legal violations, employee health status) into a dedicated input form.

[0617] Data processing: The terminal verifies the format of the input data and reformats it if necessary. It also performs validation to improve data reliability (e.g., checking required fields, verifying data types).

[0618] Output: Sends the formatted data to the server.

[0619] Step 2:

[0620] The server receives the data and stores it in the database

[0621] Input: Risk-related data sent from the device.

[0622] Data processing: The server performs error checks and format validation on the received data. If there are no problems, the data is stored in a temporary buffer.

[0623] Output: Temporarily label the passed data when saving it to the database.

[0624] Step 3:

[0625] The server preprocesses the data

[0626] Input: Risk-related data stored in the database.

[0627] Data processing: The server performs preprocessing on the stored data, including missing value imputation (e.g., mean imputation or mode imputation), removal of unnecessary items, data normalization (e.g., scaling numerical data to the range 0 to 1), and outlier detection and correction.

[0628] Output: Generate a preprocessed dataset in a format that can be used in the next step.

[0629] Step 4:

[0630] The server extracts features

[0631] Input: The preprocessed dataset.

[0632] Data processing: The server selects and extracts features necessary for risk assessment. For example, it actively extracts features such as "frequency of delivery delays in the past year," "conditions for legal violations," and "fluctuations in employee health."

[0633] Output: Generate a feature dataset and prepare it for input into a risk assessment model.

[0634] Step 5:

[0635] The server performs the risk assessment

[0636] Input: Feature dataset.

[0637] Data calculation: The server applies machine learning algorithms (e.g., random forests, support vector machines) and statistical models to calculate the probability and impact of risks, generating an assessment score for each risk.

[0638] Output: The risk assessment results are stored in a database and formatted for use in the next step.

[0639] Step 6:

[0640] The server generates countermeasures using generation AI

[0641] Input: Risk assessment results.

[0642] Data processing: Based on the analysis results, the server applies a generative AI model (e.g., GPT-3, BERT) to generate optimal risk countermeasures. Specifically, a prompt is input into the generative AI, and a countermeasure is obtained as a response.

[0643] Output: Generate a list of the generated risk treatments and format it for delivery to the user.

[0644] Step 7:

[0645] The server sends the countermeasures to the device.

[0646] Input: Generated risk treatment proposals.

[0647] Data transmission: The server sends the countermeasures plan to the user's device. Data is also encoded and encrypted during transmission to ensure secure communication.

[0648] Output: Provide countermeasures in a format that can be displayed on the user's device.

[0649] Step 8:

[0650] The user checks and selects countermeasures via the device.

[0651] Input: Risk treatment proposal sent from the server.

[0652] How it works: The user reviews the proposed countermeasures via their device and selects the countermeasure they think is best. The selected countermeasure is then sent as feedback from the device to the server.

[0653] Output: Information about the countermeasures selected by the user.

[0654] Step 9:

[0655] The user enters the execution results into the terminal and sends them to the server.

[0656] Input: The results of the implemented countermeasures and their effectiveness.

[0657] Data transmission: The user inputs data (quantitative and qualitative information) on the execution results and effects using the terminal and transmits it to the server. The input data is also validated.

[0658] Output: Sending feedback information.

[0659] Step 10:

[0660] The server updates the model based on the feedback

[0661] Input: Feedback data.

[0662] Data calculation: The server uses the feedback data to update the parameters of the risk assessment model and generative AI, specifically adding the feedback data as training data for the model and retraining it.

[0663] Output: Updated risk assessment model and generative AI model.

[0664] The above is a specific processing flow of the risk management system based on the present invention, and provides a detailed explanation of the specific operations and data flow performed at each step.

[0665] (Application example 1)

[0666] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0667] Operational risk management in factories requires rapid and accurate response to a wide range of risk factors. However, in current systems, risk data collection and analysis, and the generation and implementation of countermeasures are carried out separately, and there is no integrated automated system. This makes efficient risk management difficult and sometimes prevents timely responses. Furthermore, there is a lack of systems that can check risks and implement countermeasures in real time, and delays in the communication of information between on-site workers and managers are also an issue.

[0668] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0669] In this invention, the server includes: a means for inputting risk-related data; a means for transmitting the data to the server; a means for storing the data in a database; a means for preprocessing the stored data; a means for extracting features from the preprocessed data; a means for analyzing risks based on the extracted features; a means for applying a generation AI to generate risk countermeasures based on the analysis results; a means for transmitting the generated countermeasures to a user terminal; a means for the user to review and select the countermeasures; a robot control means for automatically executing the selected countermeasure; a means for feeding back the execution results of the user-selected countermeasure to the server; a means for updating the model based on the feedback; and a means for displaying risks in real time via smart glasses or a head-mounted display. This enables integrated and automated operational risk management in factories, enabling rapid and accurate responses to risks. Furthermore, field workers can check risks in real time through the smart glasses or a head-mounted display, enabling appropriate information sharing between them and managers.

[0670] "Risk-related data" refers to information on all risk factors that may occur in the operation of a factory, such as malfunctions of machinery, changes in the work environment, and safety-related events.

[0671] A "server" is a high-performance computing device for managing and analyzing data in a factory.

[0672] A "database" is a system for systematically storing and managing collected data.

[0673] "Preprocessing" refers to processes such as filling in missing values, removing unnecessary items, and normalizing data in order to convert collected data into a format suitable for analysis.

[0674] "Features" are important attributes or indicators extracted from data that are necessary for analysis.

[0675] "Risk analysis" is a process of evaluating the probability of occurrence and the impact of risks based on extracted features.

[0676] "Generative AI" is an artificial intelligence model that automatically generates optimal risk countermeasures based on the results of risk analysis.

[0677] A "user terminal" is an electronic device that allows factory floor workers and managers to access the system and check and input information.

[0678] "Robot control means" is a system that operates and manages the robot to automatically execute the selected countermeasure.

[0679] "Feedback" is the process of returning the results and effects of the actions taken to the server.

[0680] "Smart glasses" are wearable devices that factory floor workers use to check risks in real time.

[0681] A "head-mounted display" is a display device used by factory managers to remotely grasp the situation on-site and issue instructions.

[0682] This invention relates to a system for managing operational risks in factories. Specifically, it provides a method for efficiently managing operational risks in factories by automating a series of processes, from data collection, data preprocessing, risk analysis, countermeasure proposal generation, and countermeasure implementation and feedback.

[0683] Hardware and software used

[0684] Smart glasses: Wearable devices (e.g., Google Glass or Vuzix Blade) that allow field workers to see risks in real time.

[0685] Head-mounted display: A display device (such as Microsoft HoloLens or Oculus Quest) that allows managers to remotely grasp the situation on-site.

[0686] Robot: A device for automatically carrying out selected countermeasures (e.g., ABB's Yumi).

[0687] Server: A high-performance computing device that uses databases, machine learning libraries (TensorFlow, PyTorch), and data processing libraries (Pandas, Numpy).

[0688] Database: A data management system such as MySQL.

[0689] System configuration and operation details

[0690] 1. Data Collection:

[0691] Through the terminal (smart glasses), field workers input risk-related data, such as abnormal machine operation, temperature changes, abnormal vibrations, and the presence or absence of obstacles. This data is also acquired from the smart glasses' sensors. The collected data is then transmitted to a server via wireless communication.

[0692] 2. Data preprocessing:

[0693] The server stores the received data in a database, and then uses Pandas and Numpy to perform preprocessing such as filling in missing values, removing unnecessary items, and normalizing the data.

[0694] 3. Feature extraction:

[0695] After preprocessing, features necessary for risk assessment are extracted using Scikit-learn and TensorFlow. For example, important features include machine operating hours, ambient temperature and humidity, and changes in the work environment.

[0696] 4. Risk Analysis:

[0697] Based on the extracted features, a deep learning model is used to evaluate the probability of risk occurrence and its impact. This analysis uses machine learning libraries such as TensorFlow and PyTorch.

[0698] 5. Countermeasure generation:

[0699] Generative AI (such as the GPT series) is used to generate optimal risk countermeasures based on the results of risk analysis, such as proposing alternative work procedures or machine repair schedules.

[0700] 6. Implementation and feedback of countermeasures:

[0701] The generated countermeasure proposals are sent to the terminal, where the user (worker or manager) reviews them and selects the most appropriate one. The selected countermeasure is then automatically executed by the robot control system. The results and effects of the implemented countermeasures are fed back to the server and used to improve the next risk assessment model.

[0702] Examples of concrete examples and prompts

[0703] Specific examples

[0704] 1. Example of data entered into smart glasses:

[0705] temperature 35 degrees

[0706] Abnormal vibration

[0707] Check for obstacles in the work area

[0708] 2. Proposed measures:

[0709] Cooling system inspection

[0710] Quickly remove obstacles

[0711] Maintenance for abnormal vibration

[0712] Prompt Sentence Examples

[0713] "The temperature on the factory floor has reached 35 degrees, and abnormal vibrations are occurring. Also, an obstacle has been found in the work area. Please generate optimal countermeasures to address these risks."

[0714] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0715] Step 1:

[0716] Data collection

[0717] Through the terminal (smart glasses), the user (field worker) inputs risk-related data. The input data includes temperature, vibration abnormalities, and the presence or absence of obstacles in the work area. Data is also collected from the smart glasses' sensors. This data is transmitted to the server via wireless communication.

[0718] Input: Risk-related data entered by workers through terminals and sensor data from smart glasses.

[0719] Output: Collected risk-related data is sent to a server.

[0720] Step 2:

[0721] Data Preprocessing

[0722] The server stores the data received from the terminal in a database, and then uses Pandas and Numpy to perform preprocessing such as filling in missing values, removing unnecessary items, and normalizing the data.

[0723] Input: Collected risk-related data.

[0724] Output: Clean, pre-processed data.

[0725] Step 3:

[0726] Feature extraction

[0727] The server uses Scikit-learn and TensorFlow to extract the features necessary for risk assessment from the preprocessed data, including machine operating hours, ambient temperature, humidity, and changes in the work environment.

[0728] Input: Preprocessed data.

[0729] Output: Extracted features.

[0730] Step 4:

[0731] Risk Analysis

[0732] The server uses a deep learning model based on the extracted features to assess the probability and impact of a risk, using machine learning libraries such as TensorFlow and PyTorch.

[0733] Input: Extracted features.

[0734] Output: Probability and impact of risk.

[0735] Step 5:

[0736] Countermeasures generation

[0737] The server uses a generative AI (such as the GPT series) to generate optimal risk countermeasures based on the risk analysis results, including alternative work procedures and machine repair schedules.

[0738] Input: Risk analysis results.

[0739] Output: Generated countermeasures.

[0740] Step 6:

[0741] Implementation of countermeasures and feedback

[0742] The generated countermeasure proposals are sent to the device (smart glasses or head-mounted display), where the user can review them and select the most appropriate one. The selected countermeasure is then automatically implemented by the robot control system. The results and effects of the implemented countermeasures are fed back to the server and used to improve the risk assessment model.

[0743] Input: Generated countermeasures.

[0744] Output: Results of implemented measures and feedback data.

[0745] Specific prompt examples

[0746] "The temperature on the factory floor has reached 35 degrees, and abnormal vibrations are occurring. Also, an obstacle has been found in the work area. Please generate optimal countermeasures to address these risks."

[0747] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0748] This invention relates to a system for improving the accuracy and effectiveness of risk management in a company's operational risk management by combining risk information collection and analysis, the generation and implementation of countermeasures, and an emotion engine that recognizes user emotions.

[0749] System configuration

[0750] The system consists of the following components:

[0751] A means of inputting risk-related data

[0752] A means of sending data to the server

[0753] A means of storing data in a database

[0754] A means of preprocessing the stored data

[0755] A means of extracting features from preprocessed data

[0756] A method for analyzing risk based on extracted features

[0757] A means of applying generative AI to generate risk countermeasures based on analysis results

[0758] A means for transmitting the generated countermeasures to the user terminal

[0759] A means for users to check and select countermeasures

[0760] A means of feeding back the results of the selected measures to the server

[0761] A means of updating the model based on feedback

[0762] Emotion engine that recognizes user emotions

[0763] Program Processing Details

[0764] The user inputs risk-related data via a terminal.

[0765] Users use a terminal to input risk-related data, such as supplier delivery delays, legal violation risks, and employee health status. An emotion engine that recognizes the user's emotions also collects the user's emotional data during this process.

[0766] The device sends input data and emotion data to the server.

[0767] The device sends the input data and the emotion data collected by the emotion engine to the server, where the user's emotion data is stored along with the risk-related data.

[0768] The server stores the data in a database and performs preprocessing.

[0769] The server stores the received data and emotion data in a database, where preprocessing is performed, such as filling in missing values, removing unnecessary items, and normalizing the data.

[0770] The server extracts features and analyzes risks

[0771] From the preprocessed data, the server extracts the features necessary for risk assessment. Based on these features, it runs an algorithm to assess the probability of risk occurrence and its impact.

[0772] The server generates risk countermeasures using generation AI

[0773] Based on the analysis results, the server applies a generative AI to generate risk countermeasures. The generative AI also takes the user's emotional data into account when generating countermeasures, so if the user is feeling excessive stress, for example, it will propose countermeasures to alleviate that situation.

[0774] The server sends the generated countermeasures and emotion data to the user terminal.

[0775] The generated countermeasures are sent from the server to the user's device, where the user can check them. The user's emotional data is also displayed, which can be used as a reference when selecting a countermeasure.

[0776] The user checks the proposed measures and selects one

[0777] The user can then use their device to check the generated countermeasures and select which ones to implement. This process allows the selection of countermeasures that reflect the user's feelings.

[0778] The user executes the selected measure and enters the results into the device.

[0779] After the selected countermeasure is implemented, the user provides feedback on the results and effects of the countermeasure to the server from the terminal. This feedback includes not only the results of the countermeasure but also the user's emotional data.

[0780] The server updates the model based on the feedback

[0781] The server stores the feedback data and emotion data in a database and uses them to update the parameters of the risk assessment model and the generation AI, thereby improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[0782] Specific examples

[0783] For example, when a manufacturing company performs risk management for its supply chain, the user inputs data such as "data regarding delivery delays from major suppliers" and "risk of supply chain interruptions" into a terminal. In addition, an emotion engine simultaneously collects user emotional data such as "stress" and "anxiety." The server preprocesses this data, performs risk analysis, and then uses generative AI to generate countermeasures that take the emotional data into account. The user reviews these countermeasures, selects the most appropriate one, and implements it. After implementation, the results and emotional data are sent back to the server and used to update the model.

[0784] In this way, the present invention can realize risk management that takes into account users' emotions, and effectively minimize potential losses from enterprise risks.

[0785] The processing flow will be explained below.

[0786] Step 1:

[0787] Users use a terminal to input risk-related data, such as supplier delivery delays, legal violation risks, and employee health status. At the same time, an emotion engine analyzes the user's facial expressions and tone of voice to collect emotional data.

[0788] Step 2:

[0789] The device validates the risk-related and emotion data entered and sends it to the server, where validation includes checking the data format and filtering out invalid data.

[0790] Step 3:

[0791] The server stores the risk-related data and emotion data sent from the device in a database, where the data is appropriately labeled.

[0792] Step 4:

[0793] The server performs preprocessing on the stored risk-related data, including filling in missing values, removing unnecessary items, and normalizing the data. The same preprocessing is also performed on the emotion data, preparing it in a format suitable for analysis.

[0794] Step 5:

[0795] The server extracts features necessary for risk assessment from the preprocessed data. Features extracted include, for example, "frequency of past delivery delays," "conditions for legal violations," and "results of employee health assessments." Emotional features such as "stress level," "happiness level," and "anxiety level" are extracted from the emotion data.

[0796] Step 6:

[0797] The server runs a risk analysis algorithm based on the extracted features and emotion features. This algorithm includes machine learning models and statistical models to evaluate the probability of risk occurrence and its impact.

[0798] Step 7:

[0799] Based on the results of the risk analysis, the server uses a generative AI to generate risk countermeasures. The generative AI also takes emotional features into account to generate countermeasures that take into consideration the user's mental health. For example, it may suggest reallocating work to reduce stress or a specific action plan to mitigate risks.

[0800] Step 8:

[0801] The server transmits the generated risk countermeasures and related emotion data to the user terminal, and the user uses the terminal to check the countermeasures.

[0802] Step 9:

[0803] The user checks the proposed countermeasures through the device and selects the most appropriate one. The selected countermeasure is then sent back to the server from the device.

[0804] Step 10:

[0805] Implement the user-selected measures, such as finding alternative suppliers, increasing inventory, or implementing employee wellness programs within the company.

[0806] Step 11:

[0807] The user uses the device to input the results of the measures taken, and the emotion engine also collects the user's emotional data at that time and sends it to the server to provide feedback.

[0808] Step 12:

[0809] The server stores the received feedback and emotion data in a database, which is then used for the next risk assessment and countermeasure generation.

[0810] Step 13:

[0811] The server updates the risk assessment model and the parameters of the generative AI based on the feedback and emotion data, thereby improving the accuracy and effectiveness of the next analysis and generated countermeasures.

[0812] Example 2

[0813] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0814] In operational risk management in companies, there is a need for the collection and analysis of risk information, the generation and implementation of countermeasures, and flexible countermeasure proposals that take user emotions into account. However, conventional systems have difficulty effectively incorporating user emotions, resulting in insufficient accuracy and effectiveness in risk management. Therefore, there is a need to develop a system that takes user emotional data into account, flexibly generates risk countermeasures, and utilizes the results of their implementation as feedback.

[0815] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0816] In this invention, the server includes a means for a user to input risk-related data, a means for transmitting the data and the user's emotion data to the server, and a means for storing the data and the emotion data in a database. This allows the user's emotions to be taken into account in risk analysis and countermeasure generation, enabling more appropriate risk countermeasures. Furthermore, by feeding back the execution results of the selected countermeasures to the server along with the emotion data, the model is updated, improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[0817] "User" refers to a company employee or system user who inputs risk-related data and checks and selects countermeasures.

[0818] A "terminal" is an electronic device through which a user inputs risk-related data and emotion data and transmits the data to a server.

[0819] The "server" is a central management system that receives, stores, analyzes risk-related data and emotion data, and generates countermeasures.

[0820] A "database" is a storage device for storing risk-related data and emotion data received by the server.

[0821] An "emotion engine" is a software or hardware component that recognizes a user's emotions in real time and collects them as data.

[0822] "Preprocessing" refers to the process of imputing missing values, removing unnecessary items, and normalizing the data received.

[0823] "Feature extraction" is the process of extracting elements from data that are important for risk assessment and analysis.

[0824] "Risk analysis" is a process of evaluating the probability of occurrence and the impact of a risk based on the extracted feature quantities.

[0825] "Generative AI" is an artificial intelligence algorithm that automatically generates countermeasures based on the results of risk analysis, taking into account the user's emotions.

[0826] "Feedback" is a process in which the execution results of the measures selected by the user and emotion data are sent to the server.

[0827] "Model updating" is a process in which the parameters of the risk assessment model and generation AI are modified based on feedback data to improve the accuracy and effectiveness of the next analysis and countermeasure generation.

[0828] This invention relates to a system for improving the accuracy and effectiveness of risk management in a company's operational risk management by combining risk information collection and analysis, the generation and implementation of countermeasures, and an emotion engine that recognizes user emotions. Specific embodiments for implementing this invention will be described below.

[0829] Hardware and Software Configuration

[0830] The system consists of the following components:

[0831] Device: An electronic device through which a user enters risk-related data. Examples include a personal computer, tablet, or smartphone.

[0832] Server: A central management system that receives, stores, analyzes, and generates countermeasures. A server with a high-performance processor and large storage capacity is recommended.

[0833] Database: A storage device for saving data received by the server. A relational database management system (RDBMS) is used.

[0834] Emotion engine: A software or hardware component that recognizes a user's emotions in real time. For example, there are emotion recognition systems that use facial recognition technology or voice analysis technology.

[0835] System operation details

[0836] The user inputs risk-related data via a terminal.

[0837] Users input risk-related data from their own devices, such as supplier delivery delays, legal violation risks, and employee health status. In addition, an emotion engine recognizes and collects the user's emotions (e.g., stress, anxiety, etc.) in real time as they input their data.

[0838] The device sends input data and emotion data to the server.

[0839] The terminal transmits the input risk-related data and emotion data to the server, whereby the risk-related data and the user's emotion data are stored.

[0840] The server stores the data in a database and performs preprocessing.

[0841] The server stores the received risk data and emotion data in a database, then performs preprocessing on the stored data, such as filling in missing values, removing unnecessary items, and normalizing the data.

[0842] The server extracts features and analyzes risks

[0843] From the preprocessed data, the server extracts the features necessary for risk assessment, then runs an algorithm to evaluate the probability of risk occurrence and its impact.

[0844] The server generates risk countermeasures using generation AI

[0845] Based on the analysis results, the server applies generative AI to generate risk countermeasures. This also takes into account the user's emotional data, so if the user is feeling excessive stress, for example, countermeasures to alleviate that situation will be proposed.

[0846] The server sends the generated countermeasures and emotion data to the user terminal.

[0847] The generated countermeasures are sent from the server to the user's device, where the user can check them. Emotion data is also displayed, which can be used as a reference when selecting a countermeasure.

[0848] The user checks the proposed measures and selects one

[0849] The user can use the device to check the generated countermeasures and select the countermeasures to be implemented. This process allows the selection of countermeasures that reflect the user's feelings.

[0850] The user executes the selected measure and enters the results into the device.

[0851] After the selected countermeasure is executed, the user sends feedback of the result and effect from the terminal to the server. This feedback includes the result of the execution of the countermeasure and the user's emotional data.

[0852] The server updates the model based on the feedback

[0853] The server stores the feedback data and emotion data in a database and uses them to update the risk assessment model and the parameters of the generation AI, thereby improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[0854] Examples of concrete examples and prompts

[0855] For example, when a manufacturing company performs risk management for its supply chain, the user inputs data such as "data regarding delivery delays from major suppliers" and "risk of supply chain interruptions" into a terminal. In addition, an emotion engine simultaneously collects user emotional data such as "stress" and "anxiety." The server preprocesses this data, performs risk analysis, and then uses generative AI to generate countermeasures that take the emotional data into account. The user reviews these countermeasures, selects the most appropriate one, and implements it. After implementation, the results and emotional data are sent back to the server and used to update the model.

[0856] Prompt Sentence Examples

[0857] "Please provide data on delivery delays from your key suppliers. Also, please rate your current stress level on a scale of 1 to 10."

[0858] "Please provide detailed information about the risk of supply chain interruption and indicate your level of concern about that risk on a scale of 0 to 100."

[0859] In this way, the present invention realizes risk management that takes into account users' emotions, and can effectively minimize potential losses from enterprise risks.

[0860] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0861] Step 1:

[0862] The user inputs risk-related data via a terminal.

[0863] Users use a terminal to input risk-related data. This data includes information such as "supplier delivery delays," "risk of legal violations," and "employee health status." The emotion engine recognizes and collects the user's emotions (such as "stress" or "anxiety") in real time as they input data. Specifically, users enter data using a keyboard or touchscreen, and the camera and microphone analyze facial expressions and tone of voice to collect emotional data.

[0864] Input: Risk-related data, user sentiment data

[0865] Output: None (passed to next step)

[0866] Step 2:

[0867] The device sends input data and emotion data to the server.

[0868] The device sends the risk data entered by the user and the emotion data collected by the emotion engine in a specific format (e.g., JSON format) to the server, where the data is checked for integrity and any necessary encryption is applied.

[0869] Input: Risk-related data, user sentiment data

[0870] Output: Data sent to the server

[0871] Step 3:

[0872] The server stores the data in a database and performs preprocessing.

[0873] The server stores the received risk and emotion data in a database. It then performs preprocessing on the data, such as filling in missing values, removing unnecessary items, and normalizing the data. For example, if there are missing values, they are filled in with the average value, unnecessary items are removed, and numerical data is standardized.

[0874] Input: Data sent to the server

[0875] Output: Preprocessed data

[0876] Step 4:

[0877] The server extracts features and analyzes risks

[0878] From the preprocessed data, the server extracts the features necessary for risk assessment. At this time, it applies data analysis algorithms to evaluate the probability of risk occurrence and its impact. For example, it uses a machine learning model to calculate a risk score and extract important variables necessary for risk assessment.

[0879] Input: Preprocessed data

[0880] Output: Risk analysis results (feature amount, risk occurrence probability, impact)

[0881] Step 5:

[0882] The server generates risk countermeasures using generation AI

[0883] The server applies a generative AI based on the risk analysis results to generate risk countermeasures. This generative AI also takes into account the user's emotional data to create appropriate risk countermeasures. For example, the generative AI can propose multiple countermeasures using the risk score and the user's emotional data as input.

[0884] Input: Risk analysis results, user emotion data

[0885] Output: Risk countermeasure proposal

[0886] Step 6:

[0887] The server sends the generated countermeasures and emotion data to the user terminal.

[0888] The generated countermeasures are sent from the server to the user's device, where the user can check the countermeasures and view the emotional data.

[0889] Input: Risk countermeasures, emotional data

[0890] Output: Countermeasures and emotion data sent to the device

[0891] Step 7:

[0892] The user checks the proposed measures and selects one

[0893] The user can then use their device to review the proposed countermeasures and select the appropriate one. This process also takes into account emotional data, allowing the user to select the countermeasure that best suits their situation. For example, the user can select the countermeasure that seems easiest to implement from the provided countermeasures.

[0894] Input: Countermeasures and emotion data sent to the device

[0895] Output: Selected countermeasures

[0896] Step 8:

[0897] The user executes the selected measure and enters the results into the device.

[0898] The user implements the selected countermeasure and sends feedback on the results and effects from the terminal to the server. This feedback includes the results of the countermeasure and emotional data. For example, the success rate of the countermeasure and the emotional state after implementation can be input.

[0899] Input: Execution results of selected countermeasures and emotion data

[0900] Output: Feedback data sent to the server

[0901] Step 9:

[0902] The server updates the model based on the feedback

[0903] The server stores the received feedback data in a database and uses it to update the parameters of the risk assessment model and the generation AI, thereby improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[0904] Input: Feedback data

[0905] Output: Updated risk assessment model and generative AI

[0906] (Application example 2)

[0907] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0908] Conventional operational risk management systems in companies have the problem that collecting and analyzing risk data takes time, making it difficult to respond in real time. In addition, because they do not take into account the emotions of workers, measures to reduce stress and anxiety are insufficient, making it difficult to improve production efficiency and reduce workers' psychological stress. This has resulted in the accuracy and effectiveness of risk management being insufficient, placing a strain on corporate operations.

[0909] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for inputting risk-related data, means for transmitting the data to the server, means for saving the data in a database, means for preprocessing the saved data, means for extracting features from the preprocessed data, means for analyzing risks based on the extracted features, means for applying a generation AI that generates risk countermeasures based on the analysis results, means for transmitting the generated countermeasures to a user terminal, means for the user to confirm and select the countermeasures, means for feeding back the execution results of the selected countermeasures to the server, means for updating the model based on the feedback, means for collecting risk data and emotion data from industrial robots and worker terminals in real time, means for generating and presenting countermeasures taking the data and emotion data into consideration, and means for displaying the countermeasures on the industrial robot's display device or the worker's terminal. This makes it possible to collect and analyze risk data and worker emotion data in real time and quickly implement highly accurate risk countermeasures.

[0910] "Risk-related data" refers to data necessary to assess a company's operational risks, including the condition of equipment, the progress of processes, and the health status of personnel.

[0911] A "server" is a computer system that receives, stores, and analyzes risk data via a network.

[0912] A "database" is a data storage system for systematically storing risk-related data and emotion data.

[0913] "Preprocessing" is the process of imputing missing values, removing unnecessary items, and normalizing data to prepare it in an analyzable format.

[0914] "Features" are extracted important data points and indicators necessary for risk analysis.

[0915] "Analyzing risk" means evaluating the probability and impact of a risk based on collected data and predicting potential dangers.

[0916] "Generative AI" is an artificial intelligence system that uses machine learning and AI technology to generate risk countermeasures.

[0917] A "user terminal" is a device that a user uses to input risk data and check countermeasure proposals, such as a smartphone or tablet.

[0918] "Feedback" refers to returning data to the server about the results and effects of the measures taken by the user.

[0919] "Updating the model" means improving and adjusting the parameters of the analytical model and generation AI based on feedback data, thereby improving the accuracy of the next analysis and countermeasure generation.

[0920] "Emotion data" is data that quantifies the user's psychological state, such as stress level and anxiety level.

[0921] An "industrial robot" is a robot used to assist or automate work within a factory, and is a mechanical device equipped with various sensors that has the ability to collect risk data in real time.

[0922] System Overview

[0923] The operational risk management system of this invention collects risk factors and worker emotion data in a factory in real time, and analyzes and generates countermeasures based on this data. The system is composed of various components, including industrial robots, worker terminals, servers, databases, and generation AI.

[0924] Data collection

[0925] Users and industrial robots collect risk-related data in real time within the factory, including equipment failure status, the progress of production processes, and the health status of workers. Workers also use smart glasses and terminals to input emotional data such as their stress and anxiety levels.

[0926] Data transmission and storage

[0927] The terminals and industrial robots send the collected risk and emotion data to a server, where it is stored in a database. The server then performs preprocessing on the stored data, such as filling in missing values, removing unnecessary items, and normalizing the data.

[0928] Data analysis

[0929] The server extracts features from the preprocessed data and performs analysis to evaluate the probability of risk occurrence and its impact, using machine learning algorithms and data analysis software (e.g., pandas, numpy, scikit-learn).

[0930] Generate countermeasures

[0931] Based on the analysis results, the server uses a generation AI to generate risk countermeasures. The generation AI also takes into account the user's emotional data when generating countermeasures, so if the user is feeling excessive stress, for example, it will propose countermeasures to alleviate that situation.

[0932] Proposing and implementing countermeasures

[0933] The generated countermeasures are sent from the server to the user's terminal or the display device of the industrial robot, where the user can check the countermeasures and select the most appropriate one. The selected countermeasure is then implemented.

[0934] Feedback and Model Updates

[0935] The user feeds back the results of the countermeasures they have implemented to the server via their device. This feedback includes not only the results of the countermeasures but also the user's emotional data. The server updates the model based on the feedback data and emotional data, improving the accuracy and effectiveness of the next analysis and generation of countermeasures.

[0936] Specific examples

[0937] For example, when a major machine breaks down in a factory, the industrial robot sends real-time data on the situation to a server. At the same time, the worker sends their stress level to the server through smart glasses. The server analyzes this data and uses generative AI to generate appropriate repair plans and measures to reduce worker stress, such as break suggestions. These measures are then displayed on the industrial robot's display and the worker's device.

[0938] Prompt Sentence Examples

[0939] "When equipment failure occurs in a factory, please collect information on the situation in real time and generate optimal repair measures. Please also suggest measures to reduce worker stress."

[0940] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0941] Step 1:

[0942] Users and industrial robots collect risk-related data within factories. Specifically, industrial robots use sensors to obtain information about the equipment's operating status and manufacturing processes, while workers use smart glasses to input their own health and psychological state (stress level, anxiety level). The data input here covers a wide range of topics, including equipment status, production line delays, and worker fatigue levels.

[0943] Step 2:

[0944] The terminal sends the collected risk data and emotion data to a server. The terminal then compiles and packages this data and sends it to the server via the network. The transmitted data is input to the server in the form of information such as the operating status of equipment, production delays, and worker stress levels.

[0945] Step 3:

[0946] The server stores the received data in a database. The server stores the transmitted risk and emotion data in the database and tags them appropriately. The stored data also includes time information and the data source (equipment, worker, etc.).

[0947] Step 4:

[0948] The server preprocesses the stored data by removing unnecessary items, filling in missing values, and normalizing the data. The server first removes meaningless and outliers from the data, then infers and fills in missing data points. Finally, the data is converted into a unified format and prepared for feature extraction.

[0949] Step 5:

[0950] The server extracts features from the preprocessed data. After data normalization is complete, the server uses a machine learning algorithm to extract important features necessary for risk assessment. The output of this process is input data for assessing the probability of risk occurrence and its impact.

[0951] Step 6:

[0952] The server analyzes risks based on the extracted features. Using a risk assessment model based on the features, the server calculates and assesses the probability of risk occurrence and the impact. This analysis clarifies which risks are likely to occur and the extent of their impact.

[0953] Step 7:

[0954] The server uses a generative AI based on the analysis results to generate risk countermeasures. Based on the analysis results, the server uses generative AI (e.g., GPT-3 model) to automatically generate optimal risk countermeasures. Countermeasures that take emotional data into consideration are generated here. The generated countermeasures are specific action plans for the countermeasures.

[0955] Step 8:

[0956] The server sends the generated countermeasure plan to the user's terminal or the display device of the industrial robot. The generated countermeasure plan is sent from the server to the user's terminal or the display device of the industrial robot, where it can be checked by the user or worker. The presented countermeasure plan includes the implementation procedure and expected effects.

[0957] Step 9:

[0958] The user or worker reviews the proposed countermeasures and selects the most appropriate one. After the proposed countermeasures are displayed on a terminal or display device, the user or worker reviews the contents and selects the countermeasure to be implemented. The selected countermeasure is sent back to the server as confirmation of the selection.

[0959] Step 10:

[0960] The user implements the selected measure, and the results are fed back from the terminal to the server. After the user or worker implements the measure, the results, effectiveness, and emotional state at the time are entered into the terminal as feedback and sent to the server. This data is used to evaluate the effectiveness of the measure.

[0961] Step 11:

[0962] The server updates the model based on the feedback data. After receiving the feedback data, the server saves it in the database and updates the parameters of the risk assessment model and the generation AI. This update improves the accuracy of the next analysis and countermeasure generation.

[0963] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0964] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0965] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0966] [Third embodiment]

[0967] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0968] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0969] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0970] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0971] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0972] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0973] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0974] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0975] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[0976] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0977] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0978] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. 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."

[0979] This invention relates to a system for efficiently managing a company's operational risks. The system aims to automate the risk management process by integrating data collection, data analysis, risk assessment, countermeasure generation, countermeasure implementation, and feedback.

[0980] System configuration

[0981] Data collection

[0982] The user inputs risk-related data via a terminal.

[0983] The user uses the device to input various risk-related data, such as supply chain interruptions, legal violation risks, and employee health status. The device then verifies this information and sends it to the server.

[0984] Data storage and preprocessing

[0985] The server receives the data and stores it in the database

[0986] The server receives the data sent from the device and stores it in a database, along with a temporary label.

[0987] The server preprocesses the data

[0988] The server performs preprocessing on the stored data, which includes processes such as imputing missing values, removing unnecessary items, and normalizing the data.

[0989] Feature extraction and risk analysis

[0990] The server extracts features

[0991] After preprocessing, the features necessary for risk assessment are extracted from the data. For example, features such as "frequency of past delivery delays," "conditions for legal violations," and "results of employee health assessments" are extracted.

[0992] The server performs risk analysis

[0993] Based on the extracted features, the server performs risk analysis, using machine learning algorithms and statistical models to assess the probability of risk occurrence and its impact.

[0994] Risk countermeasure generation

[0995] The server generates countermeasures using generation AI

[0996] Based on the analysis results, the server applies generative AI to generate optimal risk countermeasures, such as suggesting alternative suppliers, increasing inventory, or implementing employee wellness programs.

[0997] The server sends the countermeasures to the device.

[0998] The generated risk countermeasure plan is sent from the server to the user's terminal, and the user can check the countermeasure plan.

[0999] Implementation of countermeasures and feedback

[1000] The user checks and selects countermeasures via the device.

[1001] The user uses the terminal to check the countermeasures received from the server, select the countermeasure that they think is most appropriate, and the selected countermeasure is then put into action.

[1002] The user enters the execution results into the terminal and sends them to the server.

[1003] After the countermeasures are implemented, the user sends feedback on the results and effects to the server from the terminal, and the effectiveness of the countermeasures is evaluated based on this feedback.

[1004] The server improves the model based on the feedback

[1005] The server stores the collected feedback data in a database and uses it to update the risk assessment model and parameters of the generative AI, making the next risk analysis and countermeasure generation more accurate and effective.

[1006] Specific examples

[1007] For example, when a manufacturing company manages supply chain risks, the user inputs data on delivery delays from key suppliers, supply chain interruption risks, and information on alternative suppliers into their terminals. The server then preprocesses this data, performs risk analysis, and uses generative AI to propose countermeasures. The user can then confirm and implement the proposed alternative supplier selection plan to minimize risk. This process is returned to the server through feedback, improving the accuracy and effectiveness of the entire system.

[1008] In this way, the present invention improves the efficiency of operational risk management for companies, enabling rapid and accurate risk assessment and countermeasure proposals, thereby enabling companies to minimize potential losses from risks.

[1009] The processing flow will be explained below.

[1010] Step 1:

[1011] Users use the terminal to input risk-related data, such as supplier delivery delays, legal violation risk, and employee health status.

[1012] Step 2:

[1013] The terminal validates the entered data, checking for format and content errors as necessary, and sends the validated data to the server.

[1014] Step 3:

[1015] The server stores the data received from the device in a database, where it is appropriately labeled when stored.

[1016] Step 4:

[1017] The server preprocesses the stored data, specifically by imputing missing values, removing unnecessary items, and normalizing the data.

[1018] Step 5:

[1019] The server extracts features necessary for risk assessment from the preprocessed data, such as the frequency of past delivery delays, conditions for legal violations, and employee health assessment results.

[1020] Step 6:

[1021] The server then runs algorithms for risk analysis based on the extracted features. These algorithms include machine learning models and statistical models, and use them to evaluate the probability of risk occurrence and its impact.

[1022] Step 7:

[1023] Based on the results of the risk analysis, the server uses generative AI to generate optimal risk countermeasures, such as proposing alternative suppliers, increasing inventory, or introducing employee health programs.

[1024] Step 8:

[1025] The server transmits the generated countermeasures to the user terminal, where the user can check the countermeasures.

[1026] Step 9:

[1027] The user uses the terminal to check the proposed countermeasures and select the countermeasure to be implemented. The terminal then feeds this selection information back to the server.

[1028] Step 10:

[1029] The user implements the measures selected by the user, such as finding an alternative supplier or increasing inventory within the company.

[1030] Step 11:

[1031] The user uses the terminal to input the results of the measures taken and send them to the server, which then provides feedback on the effectiveness of the measures.

[1032] Step 12:

[1033] The server stores the received feedback in a database, which can then be used for the next risk assessment and countermeasure generation.

[1034] Step 13:

[1035] The server updates the risk assessment model and the parameters of the generation AI based on the feedback data, thereby improving the accuracy and effectiveness of the next analysis and generated countermeasures.

[1036] Example 1

[1037] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1038] In today's corporate environment, efficient management of operational risk is crucial, but traditional manual risk management methods are insufficient. Rapid and accurate risk assessment and countermeasure proposals are therefore required. However, there is no integrated system for data collection, preprocessing, analysis, countermeasure generation, and feedback, making the risk management process fragmented and inefficient. As a result, companies often are unable to respond quickly to potential loss risks. Furthermore, an effective means of automating complex tasks such as preprocessing, feature extraction, risk analysis, and countermeasure generation is needed.

[1039] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1040] In this invention, the server includes a means for inputting risk-related information, a means for transmitting the information to the server, a means for storing the information in a database, a means for preprocessing the stored information, a means for extracting features from the preprocessed information, a means for assessing risk based on the extracted features, a means for applying a generation AI to generate risk countermeasures based on the assessment results, a means for transmitting the generated countermeasures to a user terminal, a means for the user to review and select the countermeasures, a means for feeding back the execution results of the selected countermeasures to the server, and a means for updating the model based on the feedback. This unifies the risk management process, enabling efficient and rapid automation of risk assessment and countermeasure proposals. Furthermore, by continuously improving the model based on feedback, the accuracy of risk assessment and countermeasure proposals can be improved.

[1041] "Risk-related information" refers to data collected to identify and manage potential risks in a company's operations, such as delivery delays, violations of laws and regulations, and employee health conditions.

[1042] A "server" is a computer system that receives, processes, and transmits information over a network.

[1043] A "database" is a system for storing and managing data in an orderly manner and retrieving it as needed.

[1044] "Preprocessing" refers to the process of processing data to prepare risk-related information in an analyzable format, and includes imputing missing values, removing unnecessary items, and normalizing data.

[1045] "Features" are characteristics or attributes of data selected and calculated for risk assessment and are used directly in analysis.

[1046] "Means for assessing risk" refers to the process and technology for calculating and assessing the probability of occurrence and impact of risk based on the extracted features.

[1047] "Generative AI" refers to technologies and algorithms that use artificial intelligence to generate appropriate responses and countermeasures from specific input data.

[1048] "Feedback" refers to information about the results and effectiveness of a proposed measure that is returned to the system and used for the next process improvement.

[1049] "Means for updating the model" refers to technologies and methods that adjust the parameters of the system's risk assessment model and generation AI based on collected feedback data, thereby improving the accuracy of the next risk assessment and countermeasure generation.

[1050] "User terminal" means a device used by a user to enter data, check results, and select countermeasures, and examples include PCs, tablets, and smartphones.

[1051] This invention relates to a system for efficiently managing a company's operational risks. The system aims to integrate and automate a series of processes including data collection, data analysis, risk assessment, countermeasure proposal generation, and implementation and feedback of the countermeasure proposals.

[1052] System configuration

[1053] Data collection

[1054] The user inputs risk-related data via a terminal.

[1055] Users input risk-related information via terminal devices such as PCs, tablets, and smartphones. For example, they enter data such as "supply chain interruptions," "risk of legal violations," and "employee health status" into a dedicated input form. The terminal then verifies this data and sends it to the server.

[1056] Data storage and preprocessing

[1057] The server receives the data and stores it in the database

[1058] The server receives risk-related data sent from the device, performs error checks and format validation, and if there are no problems, stores the data in a database. At this time, the data is temporarily labeled.

[1059] The server preprocesses the data

[1060] The server performs preprocessing on the stored data, including missing value imputation, removal of unnecessary items, data normalization, and outlier detection and correction. The preprocessed data is then used for feature extraction and risk analysis.

[1061] Feature extraction and risk assessment

[1062] The server extracts features

[1063] From the preprocessed data, the server extracts features necessary for risk assessment, such as "frequency of delivery delays in the past year," "conditions for legal violations," and "fluctuations in employee health status." These features are generated as a new dataset.

[1064] The server performs the risk assessment

[1065] The server uses the feature dataset to apply machine learning algorithms (e.g., random forests and support vector machines) and statistical models to analyze the probability of risk occurrence and its impact. The analysis results are stored in a database as risk assessment results.

[1066] Risk countermeasure generation

[1067] The server generates countermeasures using generation AI

[1068] Based on the risk assessment results, the server applies a generative AI model (e.g., GPT-3, BERT) to generate optimal risk countermeasures, such as suggesting alternative suppliers, increasing inventory, or implementing an employee health program.

[1069] The server sends the countermeasures to the device.

[1070] The generated risk countermeasures are sent from the server to the user's device, where the user can check the countermeasures.

[1071] Implementation of countermeasures and feedback

[1072] The user checks and selects countermeasures via the device.

[1073] The user checks the countermeasures sent via the terminal and selects the countermeasure that they think is most effective. The selected countermeasure is then implemented.

[1074] The user enters the execution results into the terminal and sends them to the server.

[1075] After implementing the measures, the user inputs the results and effects from the device and provides feedback to the server. The feedback is provided as quantitative and qualitative information.

[1076] The server updates the model based on the feedback

[1077] The server stores the feedback data in a database and updates the parameters of the risk assessment model and the generation AI based on this data, making the next risk analysis and countermeasure generation more accurate and effective. The feedback content is also used as training data for the model.

[1078] Specific examples

[1079] For example, when a manufacturing company manages supply chain risks, the user inputs data on delivery delays from key suppliers, supply chain interruption risks, and information on alternative suppliers from their device. The server then preprocesses this data, performs risk analysis, and uses generative AI to propose countermeasures. The user then confirms and implements the proposed alternative supplier selection plan, thereby minimizing risk. This process is returned to the server through feedback, improving the accuracy and effectiveness of the entire system.

[1080] Examples of prompt statements

[1081] Here are some examples of prompts to input to a generative AI model:

[1082] "We are facing an increased risk of supply chain interruption and would like to find alternative suppliers. Our current supplier has repeatedly delayed deliveries and is running low on inventory. Please generate the optimal solution for this."

[1083] By using this prompt, the generative AI can generate specific risk countermeasures and provide them to the user.

[1084] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1085] Step 1:

[1086] The user inputs risk-related data via a terminal.

[1087] Input: Users use a device (e.g., PC, tablet, smartphone) to enter risk-related information (e.g., supply chain interruptions, risk of legal violations, employee health status) into a dedicated input form.

[1088] Data processing: The terminal verifies the format of the input data and reformats it if necessary. It also performs validation to improve data reliability (e.g., checking required fields, verifying data types).

[1089] Output: Sends the formatted data to the server.

[1090] Step 2:

[1091] The server receives the data and stores it in the database

[1092] Input: Risk-related data sent from the device.

[1093] Data processing: The server performs error checks and format validation on the received data. If there are no problems, the data is stored in a temporary buffer.

[1094] Output: Temporarily label the passed data when saving it to the database.

[1095] Step 3:

[1096] The server preprocesses the data

[1097] Input: Risk-related data stored in the database.

[1098] Data processing: The server performs preprocessing on the stored data, including missing value imputation (e.g., mean imputation or mode imputation), removal of unnecessary items, data normalization (e.g., scaling numerical data to the range 0 to 1), and outlier detection and correction.

[1099] Output: Generate a preprocessed dataset in a format that can be used in the next step.

[1100] Step 4:

[1101] The server extracts features

[1102] Input: The preprocessed dataset.

[1103] Data processing: The server selects and extracts features necessary for risk assessment. For example, it actively extracts features such as "frequency of delivery delays in the past year," "conditions for legal violations," and "fluctuations in employee health."

[1104] Output: Generate a feature dataset and prepare it for input into a risk assessment model.

[1105] Step 5:

[1106] The server performs the risk assessment

[1107] Input: Feature dataset.

[1108] Data calculation: The server applies machine learning algorithms (e.g., random forests, support vector machines) and statistical models to calculate the probability and impact of risks, generating an assessment score for each risk.

[1109] Output: The risk assessment results are stored in a database and formatted for use in the next step.

[1110] Step 6:

[1111] The server generates countermeasures using generation AI

[1112] Input: Risk assessment results.

[1113] Data processing: Based on the analysis results, the server applies a generative AI model (e.g., GPT-3, BERT) to generate optimal risk countermeasures. Specifically, a prompt is input into the generative AI, and a countermeasure is obtained as a response.

[1114] Output: Generate a list of the generated risk treatments and format it for delivery to the user.

[1115] Step 7:

[1116] The server sends the countermeasures to the device.

[1117] Input: Generated risk treatment proposals.

[1118] Data transmission: The server sends the countermeasures plan to the user's device. Data is also encoded and encrypted during transmission to ensure secure communication.

[1119] Output: Provide countermeasures in a format that can be displayed on the user's device.

[1120] Step 8:

[1121] The user checks and selects countermeasures via the device.

[1122] Input: Risk treatment proposal sent from the server.

[1123] How it works: The user reviews the proposed countermeasures via their device and selects the countermeasure they think is best. The selected countermeasure is then sent as feedback from the device to the server.

[1124] Output: Information about the countermeasures selected by the user.

[1125] Step 9:

[1126] The user enters the execution results into the terminal and sends them to the server.

[1127] Input: The results of the implemented countermeasures and their effectiveness.

[1128] Data transmission: The user inputs data (quantitative and qualitative information) on the execution results and effects using the terminal and transmits it to the server. The input data is also validated.

[1129] Output: Sending feedback information.

[1130] Step 10:

[1131] The server updates the model based on the feedback

[1132] Input: Feedback data.

[1133] Data calculation: The server uses the feedback data to update the parameters of the risk assessment model and generative AI, specifically adding the feedback data as training data for the model and retraining it.

[1134] Output: Updated risk assessment model and generative AI model.

[1135] The above is a specific processing flow of the risk management system based on the present invention, and provides a detailed explanation of the specific operations and data flow performed at each step.

[1136] (Application example 1)

[1137] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1138] Operational risk management in factories requires rapid and accurate response to a wide range of risk factors. However, in current systems, risk data collection and analysis, and the generation and implementation of countermeasures are carried out separately, and there is no integrated automated system. This makes efficient risk management difficult and sometimes prevents timely responses. Furthermore, there is a lack of systems that can check risks and implement countermeasures in real time, and delays in the communication of information between on-site workers and managers are also an issue.

[1139] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1140] In this invention, the server includes: a means for inputting risk-related data; a means for transmitting the data to the server; a means for storing the data in a database; a means for preprocessing the stored data; a means for extracting features from the preprocessed data; a means for analyzing risks based on the extracted features; a means for applying a generation AI to generate risk countermeasures based on the analysis results; a means for transmitting the generated countermeasures to a user terminal; a means for the user to review and select the countermeasures; a robot control means for automatically executing the selected countermeasure; a means for feeding back the execution results of the user-selected countermeasure to the server; a means for updating the model based on the feedback; and a means for displaying risks in real time via smart glasses or a head-mounted display. This enables integrated and automated operational risk management in factories, enabling rapid and accurate responses to risks. Furthermore, field workers can check risks in real time through the smart glasses or a head-mounted display, enabling appropriate information sharing between them and managers.

[1141] "Risk-related data" refers to information on all risk factors that may occur in the operation of a factory, such as malfunctions of machinery, changes in the work environment, and safety-related events.

[1142] A "server" is a high-performance computing device for managing and analyzing data in a factory.

[1143] A "database" is a system for systematically storing and managing collected data.

[1144] "Preprocessing" refers to processes such as filling in missing values, removing unnecessary items, and normalizing data in order to convert collected data into a format suitable for analysis.

[1145] "Features" are important attributes or indicators extracted from data that are necessary for analysis.

[1146] "Risk analysis" is a process of evaluating the probability of occurrence and the impact of risks based on extracted features.

[1147] "Generative AI" is an artificial intelligence model that automatically generates optimal risk countermeasures based on the results of risk analysis.

[1148] A "user terminal" is an electronic device that allows factory floor workers and managers to access the system and check and input information.

[1149] "Robot control means" is a system that operates and manages the robot to automatically execute the selected countermeasure.

[1150] "Feedback" is the process of returning the results and effects of the actions taken to the server.

[1151] "Smart glasses" are wearable devices that factory floor workers use to check risks in real time.

[1152] A "head-mounted display" is a display device used by factory managers to remotely grasp the situation on-site and issue instructions.

[1153] This invention relates to a system for managing operational risks in factories. Specifically, it provides a method for efficiently managing operational risks in factories by automating a series of processes, from data collection, data preprocessing, risk analysis, countermeasure proposal generation, and countermeasure implementation and feedback.

[1154] Hardware and software used

[1155] Smart glasses: Wearable devices (e.g., Google Glass or Vuzix Blade) that allow field workers to see risks in real time.

[1156] Head-mounted display: A display device (such as Microsoft HoloLens or Oculus Quest) that allows managers to remotely grasp the situation on-site.

[1157] Robot: A device for automatically carrying out selected countermeasures (e.g., ABB's Yumi).

[1158] Server: A high-performance computing device that uses databases, machine learning libraries (TensorFlow, PyTorch), and data processing libraries (Pandas, Numpy).

[1159] Database: A data management system such as MySQL.

[1160] System configuration and operation details

[1161] 1. Data Collection:

[1162] Through the terminal (smart glasses), field workers input risk-related data, such as abnormal machine operation, temperature changes, abnormal vibrations, and the presence or absence of obstacles. This data is also acquired from the smart glasses' sensors. The collected data is then transmitted to a server via wireless communication.

[1163] 2. Data preprocessing:

[1164] The server stores the received data in a database, and then uses Pandas and Numpy to perform preprocessing such as filling in missing values, removing unnecessary items, and normalizing the data.

[1165] 3. Feature extraction:

[1166] After preprocessing, features necessary for risk assessment are extracted using Scikit-learn and TensorFlow. For example, important features include machine operating hours, ambient temperature and humidity, and changes in the work environment.

[1167] 4. Risk Analysis:

[1168] Based on the extracted features, a deep learning model is used to evaluate the probability of risk occurrence and its impact. This analysis uses machine learning libraries such as TensorFlow and PyTorch.

[1169] 5. Countermeasure generation:

[1170] Generative AI (such as the GPT series) is used to generate optimal risk countermeasures based on the results of risk analysis, such as proposing alternative work procedures or machine repair schedules.

[1171] 6. Implementation and feedback of countermeasures:

[1172] The generated countermeasure proposals are sent to the terminal, where the user (worker or manager) reviews them and selects the most appropriate one. The selected countermeasure is then automatically executed by the robot control system. The results and effects of the implemented countermeasures are fed back to the server and used to improve the next risk assessment model.

[1173] Examples of concrete examples and prompts

[1174] Specific examples

[1175] 1. Example of data entered into smart glasses:

[1176] temperature 35 degrees

[1177] Abnormal vibration

[1178] Check for obstacles in the work area

[1179] 2. Proposed measures:

[1180] Cooling system inspection

[1181] Quickly remove obstacles

[1182] Maintenance for abnormal vibration

[1183] Prompt Sentence Examples

[1184] "The temperature on the factory floor has reached 35 degrees, and abnormal vibrations are occurring. Also, an obstacle has been found in the work area. Please generate optimal countermeasures to address these risks."

[1185] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1186] Step 1:

[1187] Data collection

[1188] Through the terminal (smart glasses), the user (field worker) inputs risk-related data. The input data includes temperature, vibration abnormalities, and the presence or absence of obstacles in the work area. Data is also collected from the smart glasses' sensors. This data is transmitted to the server via wireless communication.

[1189] Input: Risk-related data entered by workers through terminals and sensor data from smart glasses.

[1190] Output: Collected risk-related data is sent to a server.

[1191] Step 2:

[1192] Data Preprocessing

[1193] The server stores the data received from the terminal in a database, and then uses Pandas and Numpy to perform preprocessing such as filling in missing values, removing unnecessary items, and normalizing the data.

[1194] Input: Collected risk-related data.

[1195] Output: Clean, pre-processed data.

[1196] Step 3:

[1197] Feature extraction

[1198] The server uses Scikit-learn and TensorFlow to extract the features necessary for risk assessment from the preprocessed data, including machine operating hours, ambient temperature, humidity, and changes in the work environment.

[1199] Input: Preprocessed data.

[1200] Output: Extracted features.

[1201] Step 4:

[1202] Risk Analysis

[1203] The server uses a deep learning model based on the extracted features to assess the probability and impact of a risk, using machine learning libraries such as TensorFlow and PyTorch.

[1204] Input: Extracted features.

[1205] Output: Probability and impact of risk.

[1206] Step 5:

[1207] Countermeasures generation

[1208] The server uses a generative AI (such as the GPT series) to generate optimal risk countermeasures based on the risk analysis results, including alternative work procedures and machine repair schedules.

[1209] Input: Risk analysis results.

[1210] Output: Generated countermeasures.

[1211] Step 6:

[1212] Implementation of countermeasures and feedback

[1213] The generated countermeasure proposals are sent to the device (smart glasses or head-mounted display), where the user can review them and select the most appropriate one. The selected countermeasure is then automatically implemented by the robot control system. The results and effects of the implemented countermeasures are fed back to the server and used to improve the risk assessment model.

[1214] Input: Generated countermeasures.

[1215] Output: Results of implemented measures and feedback data.

[1216] Specific prompt examples

[1217] "The temperature on the factory floor has reached 35 degrees, and abnormal vibrations are occurring. Also, an obstacle has been found in the work area. Please generate optimal countermeasures to address these risks."

[1218] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1219] This invention relates to a system for improving the accuracy and effectiveness of risk management in a company's operational risk management by combining risk information collection and analysis, the generation and implementation of countermeasures, and an emotion engine that recognizes user emotions.

[1220] System configuration

[1221] The system consists of the following components:

[1222] A means of inputting risk-related data

[1223] A means of sending data to the server

[1224] A means of storing data in a database

[1225] A means of preprocessing the stored data

[1226] A means of extracting features from preprocessed data

[1227] A method for analyzing risk based on extracted features

[1228] A means of applying generative AI to generate risk countermeasures based on analysis results

[1229] A means for transmitting the generated countermeasures to the user terminal

[1230] A means for users to check and select countermeasures

[1231] A means of feeding back the results of the selected measures to the server

[1232] A means of updating the model based on feedback

[1233] Emotion engine that recognizes user emotions

[1234] Program Processing Details

[1235] The user inputs risk-related data via a terminal.

[1236] Users use a terminal to input risk-related data, such as supplier delivery delays, legal violation risks, and employee health status. An emotion engine that recognizes the user's emotions also collects the user's emotional data during this process.

[1237] The device sends input data and emotion data to the server.

[1238] The device sends the input data and the emotion data collected by the emotion engine to the server, where the user's emotion data is stored along with the risk-related data.

[1239] The server stores the data in a database and performs preprocessing.

[1240] The server stores the received data and emotion data in a database, where preprocessing is performed, such as filling in missing values, removing unnecessary items, and normalizing the data.

[1241] The server extracts features and analyzes risks

[1242] From the preprocessed data, the server extracts the features necessary for risk assessment. Based on these features, it runs an algorithm to assess the probability of risk occurrence and its impact.

[1243] The server generates risk countermeasures using generation AI

[1244] Based on the analysis results, the server applies a generative AI to generate risk countermeasures. The generative AI also takes the user's emotional data into account when generating countermeasures, so if the user is feeling excessive stress, for example, it will propose countermeasures to alleviate that situation.

[1245] The server sends the generated countermeasures and emotion data to the user terminal.

[1246] The generated countermeasures are sent from the server to the user's device, where the user can check them. The user's emotional data is also displayed, which can be used as a reference when selecting a countermeasure.

[1247] The user checks the proposed measures and selects one

[1248] The user can then use their device to check the generated countermeasures and select which ones to implement. This process allows the selection of countermeasures that reflect the user's feelings.

[1249] The user executes the selected measure and enters the results into the device.

[1250] After the selected countermeasure is implemented, the user provides feedback on the results and effects of the countermeasure to the server from the terminal. This feedback includes not only the results of the countermeasure but also the user's emotional data.

[1251] The server updates the model based on the feedback

[1252] The server stores the feedback data and emotion data in a database and uses them to update the parameters of the risk assessment model and the generation AI, thereby improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[1253] Specific examples

[1254] For example, when a manufacturing company performs risk management for its supply chain, the user inputs data such as "data regarding delivery delays from major suppliers" and "risk of supply chain interruptions" into a terminal. In addition, an emotion engine simultaneously collects user emotional data such as "stress" and "anxiety." The server preprocesses this data, performs risk analysis, and then uses generative AI to generate countermeasures that take the emotional data into account. The user reviews these countermeasures, selects the most appropriate one, and implements it. After implementation, the results and emotional data are sent back to the server and used to update the model.

[1255] In this way, the present invention can realize risk management that takes into account users' emotions, and effectively minimize potential losses from enterprise risks.

[1256] The processing flow will be explained below.

[1257] Step 1:

[1258] Users use a terminal to input risk-related data, such as supplier delivery delays, legal violation risks, and employee health status. At the same time, an emotion engine analyzes the user's facial expressions and tone of voice to collect emotional data.

[1259] Step 2:

[1260] The device validates the risk-related and emotion data entered and sends it to the server, where validation includes checking the data format and filtering out invalid data.

[1261] Step 3:

[1262] The server stores the risk-related data and emotion data sent from the device in a database, where the data is appropriately labeled.

[1263] Step 4:

[1264] The server performs preprocessing on the stored risk-related data, including filling in missing values, removing unnecessary items, and normalizing the data. The same preprocessing is also performed on the emotion data, preparing it in a format suitable for analysis.

[1265] Step 5:

[1266] The server extracts features necessary for risk assessment from the preprocessed data. Features extracted include, for example, "frequency of past delivery delays," "conditions for legal violations," and "results of employee health assessments." Emotional features such as "stress level," "happiness level," and "anxiety level" are extracted from the emotion data.

[1267] Step 6:

[1268] The server runs a risk analysis algorithm based on the extracted features and emotion features. This algorithm includes machine learning models and statistical models to evaluate the probability of risk occurrence and its impact.

[1269] Step 7:

[1270] Based on the results of the risk analysis, the server uses a generative AI to generate risk countermeasures. The generative AI also takes emotional features into account to generate countermeasures that take into consideration the user's mental health. For example, it may suggest reallocating work to reduce stress or a specific action plan to mitigate risks.

[1271] Step 8:

[1272] The server transmits the generated risk countermeasures and related emotion data to the user terminal, and the user uses the terminal to check the countermeasures.

[1273] Step 9:

[1274] The user checks the proposed countermeasures through the device and selects the most appropriate one. The selected countermeasure is then sent back to the server from the device.

[1275] Step 10:

[1276] Implement the user-selected measures, such as finding alternative suppliers, increasing inventory, or implementing employee wellness programs within the company.

[1277] Step 11:

[1278] The user uses the device to input the results of the measures taken, and the emotion engine also collects the user's emotional data at that time and sends it to the server to provide feedback.

[1279] Step 12:

[1280] The server stores the received feedback and emotion data in a database, which is then used for the next risk assessment and countermeasure generation.

[1281] Step 13:

[1282] The server updates the risk assessment model and the parameters of the generative AI based on the feedback and emotion data, thereby improving the accuracy and effectiveness of the next analysis and generated countermeasures.

[1283] Example 2

[1284] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1285] In operational risk management in companies, there is a need for the collection and analysis of risk information, the generation and implementation of countermeasures, and flexible countermeasure proposals that take user emotions into account. However, conventional systems have difficulty effectively incorporating user emotions, resulting in insufficient accuracy and effectiveness in risk management. Therefore, there is a need to develop a system that takes user emotional data into account, flexibly generates risk countermeasures, and utilizes the results of their implementation as feedback.

[1286] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1287] In this invention, the server includes a means for a user to input risk-related data, a means for transmitting the data and the user's emotion data to the server, and a means for storing the data and the emotion data in a database. This allows the user's emotions to be taken into account in risk analysis and countermeasure generation, enabling more appropriate risk countermeasures. Furthermore, by feeding back the execution results of the selected countermeasures to the server along with the emotion data, the model is updated, improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[1288] "User" refers to a company employee or system user who inputs risk-related data and checks and selects countermeasures.

[1289] A "terminal" is an electronic device through which a user inputs risk-related data and emotion data and transmits the data to a server.

[1290] The "server" is a central management system that receives, stores, analyzes risk-related data and emotion data, and generates countermeasures.

[1291] A "database" is a storage device for storing risk-related data and emotion data received by the server.

[1292] An "emotion engine" is a software or hardware component that recognizes a user's emotions in real time and collects them as data.

[1293] "Preprocessing" refers to the process of imputing missing values, removing unnecessary items, and normalizing the data received.

[1294] "Feature extraction" is the process of extracting elements from data that are important for risk assessment and analysis.

[1295] "Risk analysis" is a process of evaluating the probability of occurrence and the impact of a risk based on the extracted feature quantities.

[1296] "Generative AI" is an artificial intelligence algorithm that automatically generates countermeasures based on the results of risk analysis, taking into account the user's emotions.

[1297] "Feedback" is a process in which the execution results of the measures selected by the user and emotion data are sent to the server.

[1298] "Model updating" is a process in which the parameters of the risk assessment model and generation AI are modified based on feedback data to improve the accuracy and effectiveness of the next analysis and countermeasure generation.

[1299] This invention relates to a system for improving the accuracy and effectiveness of risk management in a company's operational risk management by combining risk information collection and analysis, the generation and implementation of countermeasures, and an emotion engine that recognizes user emotions. Specific embodiments for implementing this invention will be described below.

[1300] Hardware and Software Configuration

[1301] The system consists of the following components:

[1302] Device: An electronic device through which a user enters risk-related data. Examples include a personal computer, tablet, or smartphone.

[1303] Server: A central management system that receives, stores, analyzes, and generates countermeasures. A server with a high-performance processor and large storage capacity is recommended.

[1304] Database: A storage device for saving data received by the server. A relational database management system (RDBMS) is used.

[1305] Emotion engine: A software or hardware component that recognizes a user's emotions in real time. For example, there are emotion recognition systems that use facial recognition technology or voice analysis technology.

[1306] System operation details

[1307] The user inputs risk-related data via a terminal.

[1308] Users input risk-related data from their own devices, such as supplier delivery delays, legal violation risks, and employee health status. In addition, an emotion engine recognizes and collects the user's emotions (e.g., stress, anxiety, etc.) in real time as they input their data.

[1309] The device sends input data and emotion data to the server.

[1310] The terminal transmits the input risk-related data and emotion data to the server, whereby the risk-related data and the user's emotion data are stored.

[1311] The server stores the data in a database and performs preprocessing.

[1312] The server stores the received risk data and emotion data in a database, then performs preprocessing on the stored data, such as filling in missing values, removing unnecessary items, and normalizing the data.

[1313] The server extracts features and analyzes risks

[1314] From the preprocessed data, the server extracts the features necessary for risk assessment, then runs an algorithm to evaluate the probability of risk occurrence and its impact.

[1315] The server generates risk countermeasures using generation AI

[1316] Based on the analysis results, the server applies generative AI to generate risk countermeasures. This also takes into account the user's emotional data, so if the user is feeling excessive stress, for example, countermeasures to alleviate that situation will be proposed.

[1317] The server sends the generated countermeasures and emotion data to the user terminal.

[1318] The generated countermeasures are sent from the server to the user's device, where the user can check them. Emotion data is also displayed, which can be used as a reference when selecting a countermeasure.

[1319] The user checks the proposed measures and selects one

[1320] The user can use the device to check the generated countermeasures and select the countermeasures to be implemented. This process allows the selection of countermeasures that reflect the user's feelings.

[1321] The user executes the selected measure and enters the results into the device.

[1322] After the selected countermeasure is executed, the user sends feedback of the result and effect from the terminal to the server. This feedback includes the result of the execution of the countermeasure and the user's emotional data.

[1323] The server updates the model based on the feedback

[1324] The server stores the feedback data and emotion data in a database and uses them to update the risk assessment model and the parameters of the generation AI, thereby improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[1325] Examples of concrete examples and prompts

[1326] For example, when a manufacturing company performs risk management for its supply chain, the user inputs data such as "data regarding delivery delays from major suppliers" and "risk of supply chain interruptions" into a terminal. In addition, an emotion engine simultaneously collects user emotional data such as "stress" and "anxiety." The server preprocesses this data, performs risk analysis, and then uses generative AI to generate countermeasures that take the emotional data into account. The user reviews these countermeasures, selects the most appropriate one, and implements it. After implementation, the results and emotional data are sent back to the server and used to update the model.

[1327] Prompt Sentence Examples

[1328] "Please provide data on delivery delays from your key suppliers. Also, please rate your current stress level on a scale of 1 to 10."

[1329] "Please provide detailed information about the risk of supply chain interruption and indicate your level of concern about that risk on a scale of 0 to 100."

[1330] In this way, the present invention realizes risk management that takes into account users' emotions, and can effectively minimize potential losses from enterprise risks.

[1331] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1332] Step 1:

[1333] The user inputs risk-related data via a terminal.

[1334] Users use a terminal to input risk-related data. This data includes information such as "supplier delivery delays," "risk of legal violations," and "employee health status." The emotion engine recognizes and collects the user's emotions (such as "stress" or "anxiety") in real time as they input data. Specifically, users enter data using a keyboard or touchscreen, and the camera and microphone analyze facial expressions and tone of voice to collect emotional data.

[1335] Input: Risk-related data, user sentiment data

[1336] Output: None (passed to next step)

[1337] Step 2:

[1338] The device sends input data and emotion data to the server.

[1339] The device sends the risk data entered by the user and the emotion data collected by the emotion engine in a specific format (e.g., JSON format) to the server, where the data is checked for integrity and any necessary encryption is applied.

[1340] Input: Risk-related data, user sentiment data

[1341] Output: Data sent to the server

[1342] Step 3:

[1343] The server stores the data in a database and performs preprocessing.

[1344] The server stores the received risk and emotion data in a database. It then performs preprocessing on the data, such as filling in missing values, removing unnecessary items, and normalizing the data. For example, if there are missing values, they are filled in with the average value, unnecessary items are removed, and numerical data is standardized.

[1345] Input: Data sent to the server

[1346] Output: Preprocessed data

[1347] Step 4:

[1348] The server extracts features and analyzes risks

[1349] From the preprocessed data, the server extracts the features necessary for risk assessment. At this time, it applies data analysis algorithms to evaluate the probability of risk occurrence and its impact. For example, it uses a machine learning model to calculate a risk score and extract important variables necessary for risk assessment.

[1350] Input: Preprocessed data

[1351] Output: Risk analysis results (feature amount, risk occurrence probability, impact)

[1352] Step 5:

[1353] The server generates risk countermeasures using generation AI

[1354] The server applies a generative AI based on the risk analysis results to generate risk countermeasures. This generative AI also takes into account the user's emotional data to create appropriate risk countermeasures. For example, the generative AI can propose multiple countermeasures using the risk score and the user's emotional data as input.

[1355] Input: Risk analysis results, user emotion data

[1356] Output: Risk countermeasure proposal

[1357] Step 6:

[1358] The server sends the generated countermeasures and emotion data to the user terminal.

[1359] The generated countermeasures are sent from the server to the user's device, where the user can check the countermeasures and view the emotional data.

[1360] Input: Risk countermeasures, emotional data

[1361] Output: Countermeasures and emotion data sent to the device

[1362] Step 7:

[1363] The user checks the proposed measures and selects one

[1364] The user can then use their device to review the proposed countermeasures and select the appropriate one. This process also takes into account emotional data, allowing the user to select the countermeasure that best suits their situation. For example, the user can select the countermeasure that seems easiest to implement from the provided countermeasures.

[1365] Input: Countermeasures and emotion data sent to the device

[1366] Output: Selected countermeasures

[1367] Step 8:

[1368] The user executes the selected measure and enters the results into the device.

[1369] The user implements the selected countermeasure and sends feedback on the results and effects from the terminal to the server. This feedback includes the results of the countermeasure and emotional data. For example, the success rate of the countermeasure and the emotional state after implementation can be input.

[1370] Input: Execution results of selected countermeasures and emotion data

[1371] Output: Feedback data sent to the server

[1372] Step 9:

[1373] The server updates the model based on the feedback

[1374] The server stores the received feedback data in a database and uses it to update the parameters of the risk assessment model and the generation AI, thereby improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[1375] Input: Feedback data

[1376] Output: Updated risk assessment model and generative AI

[1377] (Application example 2)

[1378] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1379] Conventional operational risk management systems in companies have the problem that collecting and analyzing risk data takes time, making it difficult to respond in real time. In addition, because they do not take into account the emotions of workers, measures to reduce stress and anxiety are insufficient, making it difficult to improve production efficiency and reduce workers' psychological stress. This has resulted in the accuracy and effectiveness of risk management being insufficient, placing a strain on corporate operations.

[1380] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for inputting risk-related data, means for transmitting the data to the server, means for saving the data in a database, means for preprocessing the saved data, means for extracting features from the preprocessed data, means for analyzing risks based on the extracted features, means for applying a generation AI that generates risk countermeasures based on the analysis results, means for transmitting the generated countermeasures to a user terminal, means for the user to confirm and select the countermeasures, means for feeding back the execution results of the selected countermeasures to the server, means for updating the model based on the feedback, means for collecting risk data and emotion data from industrial robots and worker terminals in real time, means for generating and presenting countermeasures taking the data and emotion data into consideration, and means for displaying the countermeasures on the industrial robot's display device or the worker's terminal. This makes it possible to collect and analyze risk data and worker emotion data in real time and quickly implement highly accurate risk countermeasures.

[1381] "Risk-related data" refers to data necessary to assess a company's operational risks, including the condition of equipment, the progress of processes, and the health status of personnel.

[1382] A "server" is a computer system that receives, stores, and analyzes risk data via a network.

[1383] A "database" is a data storage system for systematically storing risk-related data and emotion data.

[1384] "Preprocessing" is the process of imputing missing values, removing unnecessary items, and normalizing data to prepare it in an analyzable format.

[1385] "Features" are extracted important data points and indicators necessary for risk analysis.

[1386] "Analyzing risk" means evaluating the probability and impact of a risk based on collected data and predicting potential dangers.

[1387] "Generative AI" is an artificial intelligence system that uses machine learning and AI technology to generate risk countermeasures.

[1388] A "user terminal" is a device that a user uses to input risk data and check countermeasure proposals, such as a smartphone or tablet.

[1389] "Feedback" refers to returning data to the server about the results and effects of the measures taken by the user.

[1390] "Updating the model" means improving and adjusting the parameters of the analytical model and generation AI based on feedback data, thereby improving the accuracy of the next analysis and countermeasure generation.

[1391] "Emotion data" is data that quantifies the user's psychological state, such as stress level and anxiety level.

[1392] An "industrial robot" is a robot used to assist or automate work within a factory, and is a mechanical device equipped with various sensors that has the ability to collect risk data in real time.

[1393] System Overview

[1394] The operational risk management system of this invention collects risk factors and worker emotion data in a factory in real time, and analyzes and generates countermeasures based on this data. The system is composed of various components, including industrial robots, worker terminals, servers, databases, and generation AI.

[1395] Data collection

[1396] Users and industrial robots collect risk-related data in real time within the factory, including equipment failure status, the progress of production processes, and the health status of workers. Workers also use smart glasses and terminals to input emotional data such as their stress and anxiety levels.

[1397] Data transmission and storage

[1398] The terminals and industrial robots send the collected risk and emotion data to a server, where it is stored in a database. The server then performs preprocessing on the stored data, such as filling in missing values, removing unnecessary items, and normalizing the data.

[1399] Data analysis

[1400] The server extracts features from the preprocessed data and performs analysis to evaluate the probability of risk occurrence and its impact, using machine learning algorithms and data analysis software (e.g., pandas, numpy, scikit-learn).

[1401] Generate countermeasures

[1402] Based on the analysis results, the server uses a generation AI to generate risk countermeasures. The generation AI also takes into account the user's emotional data when generating countermeasures, so if the user is feeling excessive stress, for example, it will propose countermeasures to alleviate that situation.

[1403] Proposing and implementing countermeasures

[1404] The generated countermeasures are sent from the server to the user's terminal or the display device of the industrial robot, where the user can check the countermeasures and select the most appropriate one. The selected countermeasure is then implemented.

[1405] Feedback and Model Updates

[1406] The user feeds back the results of the countermeasures they have implemented to the server via their device. This feedback includes not only the results of the countermeasures but also the user's emotional data. The server updates the model based on the feedback data and emotional data, improving the accuracy and effectiveness of the next analysis and generation of countermeasures.

[1407] Specific examples

[1408] For example, when a major machine breaks down in a factory, the industrial robot sends real-time data on the situation to a server. At the same time, the worker sends their stress level to the server through smart glasses. The server analyzes this data and uses generative AI to generate appropriate repair plans and measures to reduce worker stress, such as break suggestions. These measures are then displayed on the industrial robot's display and the worker's device.

[1409] Prompt Sentence Examples

[1410] "When equipment failure occurs in a factory, please collect information on the situation in real time and generate optimal repair measures. Please also suggest measures to reduce worker stress."

[1411] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1412] Step 1:

[1413] Users and industrial robots collect risk-related data within factories. Specifically, industrial robots use sensors to obtain information about the equipment's operating status and manufacturing processes, while workers use smart glasses to input their own health and psychological state (stress level, anxiety level). The data input here covers a wide range of topics, including equipment status, production line delays, and worker fatigue levels.

[1414] Step 2:

[1415] The terminal sends the collected risk data and emotion data to a server. The terminal then compiles and packages this data and sends it to the server via the network. The transmitted data is input to the server in the form of information such as the operating status of equipment, production delays, and worker stress levels.

[1416] Step 3:

[1417] The server stores the received data in a database. The server stores the transmitted risk and emotion data in the database and tags them appropriately. The stored data also includes time information and the data source (equipment, worker, etc.).

[1418] Step 4:

[1419] The server preprocesses the stored data by removing unnecessary items, filling in missing values, and normalizing the data. The server first removes meaningless and outliers from the data, then infers and fills in missing data points. Finally, the data is converted into a unified format and prepared for feature extraction.

[1420] Step 5:

[1421] The server extracts features from the preprocessed data. After data normalization is complete, the server uses a machine learning algorithm to extract important features necessary for risk assessment. The output of this process is input data for assessing the probability of risk occurrence and its impact.

[1422] Step 6:

[1423] The server analyzes risks based on the extracted features. Using a risk assessment model based on the features, the server calculates and assesses the probability of risk occurrence and the impact. This analysis clarifies which risks are likely to occur and the extent of their impact.

[1424] Step 7:

[1425] The server uses a generative AI based on the analysis results to generate risk countermeasures. Based on the analysis results, the server uses generative AI (e.g., GPT-3 model) to automatically generate optimal risk countermeasures. Countermeasures that take emotional data into consideration are generated here. The generated countermeasures are specific action plans for the countermeasures.

[1426] Step 8:

[1427] The server sends the generated countermeasure plan to the user's terminal or the display device of the industrial robot. The generated countermeasure plan is sent from the server to the user's terminal or the display device of the industrial robot, where it can be checked by the user or worker. The presented countermeasure plan includes the implementation procedure and expected effects.

[1428] Step 9:

[1429] The user or worker reviews the proposed countermeasures and selects the most appropriate one. After the proposed countermeasures are displayed on a terminal or display device, the user or worker reviews the contents and selects the countermeasure to be implemented. The selected countermeasure is sent back to the server as confirmation of the selection.

[1430] Step 10:

[1431] The user implements the selected measure, and the results are fed back from the terminal to the server. After the user or worker implements the measure, the results, effectiveness, and emotional state at the time are entered into the terminal as feedback and sent to the server. This data is used to evaluate the effectiveness of the measure.

[1432] Step 11:

[1433] The server updates the model based on the feedback data. After receiving the feedback data, the server saves it in the database and updates the parameters of the risk assessment model and the generation AI. This update improves the accuracy of the next analysis and countermeasure generation.

[1434] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1435] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1436] 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 the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1437] [Fourth embodiment]

[1438] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1439] 7, a 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.

[1440] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1441] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1442] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1443] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1444] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1445] The control object 443 includes a display device, LEDs in the eyes, and motors for driving 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 emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1446] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1447] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[1448] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1449] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1450] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1451] This invention relates to a system for efficiently managing a company's operational risks. The system aims to automate the risk management process by integrating data collection, data analysis, risk assessment, countermeasure generation, countermeasure implementation, and feedback.

[1452] System configuration

[1453] Data collection

[1454] The user inputs risk-related data via a terminal.

[1455] The user uses the device to input various risk-related data, such as supply chain interruptions, legal violation risks, and employee health status. The device then verifies this information and sends it to the server.

[1456] Data storage and preprocessing

[1457] The server receives the data and stores it in the database

[1458] The server receives the data sent from the device and stores it in a database, along with a temporary label.

[1459] The server preprocesses the data

[1460] The server performs preprocessing on the stored data, which includes processes such as imputing missing values, removing unnecessary items, and normalizing the data.

[1461] Feature extraction and risk analysis

[1462] The server extracts features

[1463] After preprocessing, the features necessary for risk assessment are extracted from the data. For example, features such as "frequency of past delivery delays," "conditions for legal violations," and "results of employee health assessments" are extracted.

[1464] The server performs risk analysis

[1465] Based on the extracted features, the server performs risk analysis, using machine learning algorithms and statistical models to assess the probability of risk occurrence and its impact.

[1466] Risk countermeasure generation

[1467] The server generates countermeasures using generation AI

[1468] Based on the analysis results, the server applies generative AI to generate optimal risk countermeasures, such as suggesting alternative suppliers, increasing inventory, or implementing employee wellness programs.

[1469] The server sends the countermeasures to the device.

[1470] The generated risk countermeasure plan is sent from the server to the user's terminal, and the user can check the countermeasure plan.

[1471] Implementation of countermeasures and feedback

[1472] The user checks and selects countermeasures via the device.

[1473] The user uses the terminal to check the countermeasures received from the server, select the countermeasure that they think is most appropriate, and the selected countermeasure is then put into action.

[1474] The user enters the execution results into the terminal and sends them to the server.

[1475] After the countermeasures are implemented, the user sends feedback on the results and effects to the server from the terminal, and the effectiveness of the countermeasures is evaluated based on this feedback.

[1476] The server improves the model based on the feedback

[1477] The server stores the collected feedback data in a database and uses it to update the risk assessment model and parameters of the generative AI, making the next risk analysis and countermeasure generation more accurate and effective.

[1478] Specific examples

[1479] For example, when a manufacturing company manages supply chain risks, the user inputs data on delivery delays from key suppliers, supply chain interruption risks, and information on alternative suppliers into their terminals. The server then preprocesses this data, performs risk analysis, and uses generative AI to propose countermeasures. The user can then confirm and implement the proposed alternative supplier selection plan to minimize risk. This process is returned to the server through feedback, improving the accuracy and effectiveness of the entire system.

[1480] In this way, the present invention improves the efficiency of operational risk management for companies, enabling rapid and accurate risk assessment and countermeasure proposals, thereby enabling companies to minimize potential losses from risks.

[1481] The processing flow will be explained below.

[1482] Step 1:

[1483] Users use the terminal to input risk-related data, such as supplier delivery delays, legal violation risk, and employee health status.

[1484] Step 2:

[1485] The terminal validates the entered data, checking for format and content errors as necessary, and sends the validated data to the server.

[1486] Step 3:

[1487] The server stores the data received from the device in a database, where it is appropriately labeled when stored.

[1488] Step 4:

[1489] The server preprocesses the stored data, specifically by imputing missing values, removing unnecessary items, and normalizing the data.

[1490] Step 5:

[1491] The server extracts features necessary for risk assessment from the preprocessed data, such as the frequency of past delivery delays, conditions for legal violations, and employee health assessment results.

[1492] Step 6:

[1493] The server then runs algorithms for risk analysis based on the extracted features. These algorithms include machine learning models and statistical models, and use them to evaluate the probability of risk occurrence and its impact.

[1494] Step 7:

[1495] Based on the results of the risk analysis, the server uses generative AI to generate optimal risk countermeasures, such as proposing alternative suppliers, increasing inventory, or introducing employee health programs.

[1496] Step 8:

[1497] The server transmits the generated countermeasures to the user terminal, where the user can check the countermeasures.

[1498] Step 9:

[1499] The user uses the terminal to check the proposed countermeasures and select the countermeasure to be implemented. The terminal then feeds this selection information back to the server.

[1500] Step 10:

[1501] The user implements the measures selected by the user, such as finding an alternative supplier or increasing inventory within the company.

[1502] Step 11:

[1503] The user uses the terminal to input the results of the measures taken and send them to the server, which then provides feedback on the effectiveness of the measures.

[1504] Step 12:

[1505] The server stores the received feedback in a database, which can then be used for the next risk assessment and countermeasure generation.

[1506] Step 13:

[1507] The server updates the risk assessment model and the parameters of the generation AI based on the feedback data, thereby improving the accuracy and effectiveness of the next analysis and generated countermeasures.

[1508] Example 1

[1509] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1510] In today's corporate environment, efficient management of operational risk is crucial, but traditional manual risk management methods are insufficient. Rapid and accurate risk assessment and countermeasure proposals are therefore required. However, there is no integrated system for data collection, preprocessing, analysis, countermeasure generation, and feedback, making the risk management process fragmented and inefficient. As a result, companies often are unable to respond quickly to potential loss risks. Furthermore, an effective means of automating complex tasks such as preprocessing, feature extraction, risk analysis, and countermeasure generation is needed.

[1511] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1512] In this invention, the server includes a means for inputting risk-related information, a means for transmitting the information to the server, a means for storing the information in a database, a means for preprocessing the stored information, a means for extracting features from the preprocessed information, a means for assessing risk based on the extracted features, a means for applying a generation AI to generate risk countermeasures based on the assessment results, a means for transmitting the generated countermeasures to a user terminal, a means for the user to review and select the countermeasures, a means for feeding back the execution results of the selected countermeasures to the server, and a means for updating the model based on the feedback. This unifies the risk management process, enabling efficient and rapid automation of risk assessment and countermeasure proposals. Furthermore, by continuously improving the model based on feedback, the accuracy of risk assessment and countermeasure proposals can be improved.

[1513] "Risk-related information" refers to data collected to identify and manage potential risks in a company's operations, such as delivery delays, violations of laws and regulations, and employee health conditions.

[1514] A "server" is a computer system that receives, processes, and transmits information over a network.

[1515] A "database" is a system for storing and managing data in an orderly manner and retrieving it as needed.

[1516] "Preprocessing" refers to the process of processing data to prepare risk-related information in an analyzable format, and includes imputing missing values, removing unnecessary items, and normalizing data.

[1517] "Features" are characteristics or attributes of data selected and calculated for risk assessment and are used directly in analysis.

[1518] "Means for assessing risk" refers to the process and technology for calculating and assessing the probability of occurrence and impact of risk based on the extracted features.

[1519] "Generative AI" refers to technologies and algorithms that use artificial intelligence to generate appropriate responses and countermeasures from specific input data.

[1520] "Feedback" refers to information about the results and effectiveness of a proposed measure that is returned to the system and used for the next process improvement.

[1521] "Means for updating the model" refers to technologies and methods that adjust the parameters of the system's risk assessment model and generation AI based on collected feedback data, thereby improving the accuracy of the next risk assessment and countermeasure generation.

[1522] "User terminal" means a device used by a user to enter data, check results, and select countermeasures, and examples include PCs, tablets, and smartphones.

[1523] This invention relates to a system for efficiently managing a company's operational risks. The system aims to integrate and automate a series of processes including data collection, data analysis, risk assessment, countermeasure proposal generation, and implementation and feedback of the countermeasure proposals.

[1524] System configuration

[1525] Data collection

[1526] The user inputs risk-related data via a terminal.

[1527] Users input risk-related information via terminal devices such as PCs, tablets, and smartphones. For example, they enter data such as "supply chain interruptions," "risk of legal violations," and "employee health status" into a dedicated input form. The terminal then verifies this data and sends it to the server.

[1528] Data storage and preprocessing

[1529] The server receives the data and stores it in the database

[1530] The server receives risk-related data sent from the device, performs error checks and format validation, and if there are no problems, stores the data in a database. At this time, the data is temporarily labeled.

[1531] The server preprocesses the data

[1532] The server performs preprocessing on the stored data, including missing value imputation, removal of unnecessary items, data normalization, and outlier detection and correction. The preprocessed data is then used for feature extraction and risk analysis.

[1533] Feature extraction and risk assessment

[1534] The server extracts features

[1535] From the preprocessed data, the server extracts features necessary for risk assessment, such as "frequency of delivery delays in the past year," "conditions for legal violations," and "fluctuations in employee health status." These features are generated as a new dataset.

[1536] The server performs the risk assessment

[1537] The server uses the feature dataset to apply machine learning algorithms (e.g., random forests and support vector machines) and statistical models to analyze the probability of risk occurrence and its impact. The analysis results are stored in a database as risk assessment results.

[1538] Risk countermeasure generation

[1539] The server generates countermeasures using generation AI

[1540] Based on the risk assessment results, the server applies a generative AI model (e.g., GPT-3, BERT) to generate optimal risk countermeasures, such as suggesting alternative suppliers, increasing inventory, or implementing an employee health program.

[1541] The server sends the countermeasures to the device.

[1542] The generated risk countermeasures are sent from the server to the user's device, where the user can check the countermeasures.

[1543] Implementation of countermeasures and feedback

[1544] The user checks and selects countermeasures via the device.

[1545] The user checks the countermeasures sent via the terminal and selects the countermeasure that they think is most effective. The selected countermeasure is then implemented.

[1546] The user enters the execution results into the terminal and sends them to the server.

[1547] After implementing the measures, the user inputs the results and effects from the device and provides feedback to the server. The feedback is provided as quantitative and qualitative information.

[1548] The server updates the model based on the feedback

[1549] The server stores the feedback data in a database and updates the parameters of the risk assessment model and the generation AI based on this data, making the next risk analysis and countermeasure generation more accurate and effective. The feedback content is also used as training data for the model.

[1550] Specific examples

[1551] For example, when a manufacturing company manages supply chain risks, the user inputs data on delivery delays from key suppliers, supply chain interruption risks, and information on alternative suppliers from their device. The server then preprocesses this data, performs risk analysis, and uses generative AI to propose countermeasures. The user then confirms and implements the proposed alternative supplier selection plan, thereby minimizing risk. This process is returned to the server through feedback, improving the accuracy and effectiveness of the entire system.

[1552] Examples of prompt statements

[1553] Here are some examples of prompts to input to a generative AI model:

[1554] "We are facing an increased risk of supply chain interruption and would like to find alternative suppliers. Our current supplier has repeatedly delayed deliveries and is running low on inventory. Please generate the optimal solution for this."

[1555] By using this prompt, the generative AI can generate specific risk countermeasures and provide them to the user.

[1556] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1557] Step 1:

[1558] The user inputs risk-related data via a terminal.

[1559] Input: Users use a device (e.g., PC, tablet, smartphone) to enter risk-related information (e.g., supply chain interruptions, risk of legal violations, employee health status) into a dedicated input form.

[1560] Data processing: The terminal verifies the format of the input data and reformats it if necessary. It also performs validation to improve data reliability (e.g., checking required fields, verifying data types).

[1561] Output: Sends the formatted data to the server.

[1562] Step 2:

[1563] The server receives the data and stores it in the database

[1564] Input: Risk-related data sent from the device.

[1565] Data processing: The server performs error checks and format validation on the received data. If there are no problems, the data is stored in a temporary buffer.

[1566] Output: Temporarily label the passed data when saving it to the database.

[1567] Step 3:

[1568] The server preprocesses the data

[1569] Input: Risk-related data stored in the database.

[1570] Data processing: The server performs preprocessing on the stored data, including missing value imputation (e.g., mean imputation or mode imputation), removal of unnecessary items, data normalization (e.g., scaling numerical data to the range 0 to 1), and outlier detection and correction.

[1571] Output: Generate a preprocessed dataset in a format that can be used in the next step.

[1572] Step 4:

[1573] The server extracts features

[1574] Input: The preprocessed dataset.

[1575] Data processing: The server selects and extracts features necessary for risk assessment. For example, it actively extracts features such as "frequency of delivery delays in the past year," "conditions for legal violations," and "fluctuations in employee health."

[1576] Output: Generate a feature dataset and prepare it for input into a risk assessment model.

[1577] Step 5:

[1578] The server performs the risk assessment

[1579] Input: Feature dataset.

[1580] Data calculation: The server applies machine learning algorithms (e.g., random forests, support vector machines) and statistical models to calculate the probability and impact of risks, generating an assessment score for each risk.

[1581] Output: The risk assessment results are stored in a database and formatted for use in the next step.

[1582] Step 6:

[1583] The server generates countermeasures using generation AI

[1584] Input: Risk assessment results.

[1585] Data processing: Based on the analysis results, the server applies a generative AI model (e.g., GPT-3, BERT) to generate optimal risk countermeasures. Specifically, a prompt is input into the generative AI, and a countermeasure is obtained as a response.

[1586] Output: Generate a list of the generated risk treatments and format it for delivery to the user.

[1587] Step 7:

[1588] The server sends the countermeasures to the device.

[1589] Input: Generated risk treatment proposals.

[1590] Data transmission: The server sends the countermeasures plan to the user's device. Data is also encoded and encrypted during transmission to ensure secure communication.

[1591] Output: Provide countermeasures in a format that can be displayed on the user's device.

[1592] Step 8:

[1593] The user checks and selects countermeasures via the device.

[1594] Input: Risk treatment proposal sent from the server.

[1595] How it works: The user reviews the proposed countermeasures via their device and selects the countermeasure they think is best. The selected countermeasure is then sent as feedback from the device to the server.

[1596] Output: Information about the countermeasures selected by the user.

[1597] Step 9:

[1598] The user enters the execution results into the terminal and sends them to the server.

[1599] Input: The results of the implemented countermeasures and their effectiveness.

[1600] Data transmission: The user inputs data (quantitative and qualitative information) on the execution results and effects using the terminal and transmits it to the server. The input data is also validated.

[1601] Output: Sending feedback information.

[1602] Step 10:

[1603] The server updates the model based on the feedback

[1604] Input: Feedback data.

[1605] Data calculation: The server uses the feedback data to update the parameters of the risk assessment model and generative AI, specifically adding the feedback data as training data for the model and retraining it.

[1606] Output: Updated risk assessment model and generative AI model.

[1607] The above is a specific processing flow of the risk management system based on the present invention, and provides a detailed explanation of the specific operations and data flow performed at each step.

[1608] (Application example 1)

[1609] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1610] Operational risk management in factories requires rapid and accurate response to a wide range of risk factors. However, in current systems, risk data collection and analysis, and the generation and implementation of countermeasures are carried out separately, and there is no integrated automated system. This makes efficient risk management difficult and sometimes prevents timely responses. Furthermore, there is a lack of systems that can check risks and implement countermeasures in real time, and delays in the communication of information between on-site workers and managers are also an issue.

[1611] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1612] In this invention, the server includes: a means for inputting risk-related data; a means for transmitting the data to the server; a means for storing the data in a database; a means for preprocessing the stored data; a means for extracting features from the preprocessed data; a means for analyzing risks based on the extracted features; a means for applying a generation AI to generate risk countermeasures based on the analysis results; a means for transmitting the generated countermeasures to a user terminal; a means for the user to review and select the countermeasures; a robot control means for automatically executing the selected countermeasure; a means for feeding back the execution results of the user-selected countermeasure to the server; a means for updating the model based on the feedback; and a means for displaying risks in real time via smart glasses or a head-mounted display. This enables integrated and automated operational risk management in factories, enabling rapid and accurate responses to risks. Furthermore, field workers can check risks in real time through the smart glasses or a head-mounted display, enabling appropriate information sharing between them and managers.

[1613] "Risk-related data" refers to information on all risk factors that may occur in the operation of a factory, such as malfunctions of machinery, changes in the work environment, and safety-related events.

[1614] A "server" is a high-performance computing device for managing and analyzing data in a factory.

[1615] A "database" is a system for systematically storing and managing collected data.

[1616] "Preprocessing" refers to processes such as filling in missing values, removing unnecessary items, and normalizing data in order to convert collected data into a format suitable for analysis.

[1617] "Features" are important attributes or indicators extracted from data that are necessary for analysis.

[1618] "Risk analysis" is a process of evaluating the probability of occurrence and the impact of risks based on extracted features.

[1619] "Generative AI" is an artificial intelligence model that automatically generates optimal risk countermeasures based on the results of risk analysis.

[1620] A "user terminal" is an electronic device that allows factory floor workers and managers to access the system and check and input information.

[1621] "Robot control means" is a system that operates and manages the robot to automatically execute the selected countermeasure.

[1622] "Feedback" is the process of returning the results and effects of the actions taken to the server.

[1623] "Smart glasses" are wearable devices that factory floor workers use to check risks in real time.

[1624] A "head-mounted display" is a display device used by factory managers to remotely grasp the situation on-site and issue instructions.

[1625] This invention relates to a system for managing operational risks in factories. Specifically, it provides a method for efficiently managing operational risks in factories by automating a series of processes, from data collection, data preprocessing, risk analysis, countermeasure proposal generation, and countermeasure implementation and feedback.

[1626] Hardware and software used

[1627] Smart glasses: Wearable devices (e.g., Google Glass or Vuzix Blade) that allow field workers to see risks in real time.

[1628] Head-mounted display: A display device (such as Microsoft HoloLens or Oculus Quest) that allows managers to remotely grasp the situation on-site.

[1629] Robot: A device for automatically carrying out selected countermeasures (e.g., ABB's Yumi).

[1630] Server: A high-performance computing device that uses databases, machine learning libraries (TensorFlow, PyTorch), and data processing libraries (Pandas, Numpy).

[1631] Database: A data management system such as MySQL.

[1632] System configuration and operation details

[1633] 1. Data Collection:

[1634] Through the terminal (smart glasses), field workers input risk-related data, such as abnormal machine operation, temperature changes, abnormal vibrations, and the presence or absence of obstacles. This data is also acquired from the smart glasses' sensors. The collected data is then transmitted to a server via wireless communication.

[1635] 2. Data preprocessing:

[1636] The server stores the received data in a database, and then uses Pandas and Numpy to perform preprocessing such as filling in missing values, removing unnecessary items, and normalizing the data.

[1637] 3. Feature extraction:

[1638] After preprocessing, features necessary for risk assessment are extracted using Scikit-learn and TensorFlow. For example, important features include machine operating hours, ambient temperature and humidity, and changes in the work environment.

[1639] 4. Risk Analysis:

[1640] Based on the extracted features, a deep learning model is used to evaluate the probability of risk occurrence and its impact. This analysis uses machine learning libraries such as TensorFlow and PyTorch.

[1641] 5. Countermeasure generation:

[1642] Generative AI (such as the GPT series) is used to generate optimal risk countermeasures based on the results of risk analysis, such as proposing alternative work procedures or machine repair schedules.

[1643] 6. Implementation and feedback of countermeasures:

[1644] The generated countermeasure proposals are sent to the terminal, where the user (worker or manager) reviews them and selects the most appropriate one. The selected countermeasure is then automatically executed by the robot control system. The results and effects of the implemented countermeasures are fed back to the server and used to improve the next risk assessment model.

[1645] Examples of concrete examples and prompts

[1646] Specific examples

[1647] 1. Example of data entered into smart glasses:

[1648] temperature 35 degrees

[1649] Abnormal vibration

[1650] Check for obstacles in the work area

[1651] 2. Proposed measures:

[1652] Cooling system inspection

[1653] Quickly remove obstacles

[1654] Maintenance for abnormal vibration

[1655] Prompt Sentence Examples

[1656] "The temperature on the factory floor has reached 35 degrees, and abnormal vibrations are occurring. Also, an obstacle has been found in the work area. Please generate optimal countermeasures to address these risks."

[1657] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1658] Step 1:

[1659] Data collection

[1660] Through the terminal (smart glasses), the user (field worker) inputs risk-related data. The input data includes temperature, vibration abnormalities, and the presence or absence of obstacles in the work area. Data is also collected from the smart glasses' sensors. This data is transmitted to the server via wireless communication.

[1661] Input: Risk-related data entered by workers through terminals and sensor data from smart glasses.

[1662] Output: Collected risk-related data is sent to a server.

[1663] Step 2:

[1664] Data Preprocessing

[1665] The server stores the data received from the terminal in a database, and then uses Pandas and Numpy to perform preprocessing such as filling in missing values, removing unnecessary items, and normalizing the data.

[1666] Input: Collected risk-related data.

[1667] Output: Clean, pre-processed data.

[1668] Step 3:

[1669] Feature extraction

[1670] The server uses Scikit-learn and TensorFlow to extract the features necessary for risk assessment from the preprocessed data, including machine operating hours, ambient temperature, humidity, and changes in the work environment.

[1671] Input: Preprocessed data.

[1672] Output: Extracted features.

[1673] Step 4:

[1674] Risk Analysis

[1675] The server uses a deep learning model based on the extracted features to assess the probability and impact of a risk, using machine learning libraries such as TensorFlow and PyTorch.

[1676] Input: Extracted features.

[1677] Output: Probability and impact of risk.

[1678] Step 5:

[1679] Countermeasures generation

[1680] The server uses a generative AI (such as the GPT series) to generate optimal risk countermeasures based on the risk analysis results, including alternative work procedures and machine repair schedules.

[1681] Input: Risk analysis results.

[1682] Output: Generated countermeasures.

[1683] Step 6:

[1684] Implementation of countermeasures and feedback

[1685] The generated countermeasure proposals are sent to the device (smart glasses or head-mounted display), where the user can review them and select the most appropriate one. The selected countermeasure is then automatically implemented by the robot control system. The results and effects of the implemented countermeasures are fed back to the server and used to improve the risk assessment model.

[1686] Input: Generated countermeasures.

[1687] Output: Results of implemented measures and feedback data.

[1688] Specific prompt examples

[1689] "The temperature on the factory floor has reached 35 degrees, and abnormal vibrations are occurring. Also, an obstacle has been found in the work area. Please generate optimal countermeasures to address these risks."

[1690] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1691] This invention relates to a system for improving the accuracy and effectiveness of risk management in a company's operational risk management by combining risk information collection and analysis, the generation and implementation of countermeasures, and an emotion engine that recognizes user emotions.

[1692] System configuration

[1693] The system consists of the following components:

[1694] A means of inputting risk-related data

[1695] A means of sending data to the server

[1696] A means of storing data in a database

[1697] A means of preprocessing the stored data

[1698] A means of extracting features from preprocessed data

[1699] A method for analyzing risk based on extracted features

[1700] A means of applying generative AI to generate risk countermeasures based on analysis results

[1701] A means for transmitting the generated countermeasures to the user terminal

[1702] A means for users to check and select countermeasures

[1703] A means of feeding back the results of the selected measures to the server

[1704] A means of updating the model based on feedback

[1705] Emotion engine that recognizes user emotions

[1706] Program Processing Details

[1707] The user inputs risk-related data via a terminal.

[1708] Users use a terminal to input risk-related data, such as supplier delivery delays, legal violation risks, and employee health status. An emotion engine that recognizes the user's emotions also collects the user's emotional data during this process.

[1709] The device sends input data and emotion data to the server.

[1710] The device sends the input data and the emotion data collected by the emotion engine to the server, where the user's emotion data is stored along with the risk-related data.

[1711] The server stores the data in a database and performs preprocessing.

[1712] The server stores the received data and emotion data in a database, where preprocessing is performed, such as filling in missing values, removing unnecessary items, and normalizing the data.

[1713] The server extracts features and analyzes risks

[1714] From the preprocessed data, the server extracts the features necessary for risk assessment. Based on these features, it runs an algorithm to assess the probability of risk occurrence and its impact.

[1715] The server generates risk countermeasures using generation AI

[1716] Based on the analysis results, the server applies a generative AI to generate risk countermeasures. The generative AI also takes the user's emotional data into account when generating countermeasures, so if the user is feeling excessive stress, for example, it will propose countermeasures to alleviate that situation.

[1717] The server sends the generated countermeasures and emotion data to the user terminal.

[1718] The generated countermeasures are sent from the server to the user's device, where the user can check them. The user's emotional data is also displayed, which can be used as a reference when selecting a countermeasure.

[1719] The user checks the proposed measures and selects one

[1720] The user can then use their device to check the generated countermeasures and select which ones to implement. This process allows the selection of countermeasures that reflect the user's feelings.

[1721] The user executes the selected measure and enters the results into the device.

[1722] After the selected countermeasure is implemented, the user provides feedback on the results and effects of the countermeasure to the server from the terminal. This feedback includes not only the results of the countermeasure but also the user's emotional data.

[1723] The server updates the model based on the feedback

[1724] The server stores the feedback data and emotion data in a database and uses them to update the parameters of the risk assessment model and the generation AI, thereby improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[1725] Specific examples

[1726] For example, when a manufacturing company performs risk management for its supply chain, the user inputs data such as "data regarding delivery delays from major suppliers" and "risk of supply chain interruptions" into a terminal. In addition, an emotion engine simultaneously collects user emotional data such as "stress" and "anxiety." The server preprocesses this data, performs risk analysis, and then uses generative AI to generate countermeasures that take the emotional data into account. The user reviews these countermeasures, selects the most appropriate one, and implements it. After implementation, the results and emotional data are sent back to the server and used to update the model.

[1727] In this way, the present invention can realize risk management that takes into account users' emotions, and effectively minimize potential losses from enterprise risks.

[1728] The processing flow will be explained below.

[1729] Step 1:

[1730] Users use a terminal to input risk-related data, such as supplier delivery delays, legal violation risks, and employee health status. At the same time, an emotion engine analyzes the user's facial expressions and tone of voice to collect emotional data.

[1731] Step 2:

[1732] The device validates the risk-related and emotion data entered and sends it to the server, where validation includes checking the data format and filtering out invalid data.

[1733] Step 3:

[1734] The server stores the risk-related data and emotion data sent from the device in a database, where the data is appropriately labeled.

[1735] Step 4:

[1736] The server performs preprocessing on the stored risk-related data, including filling in missing values, removing unnecessary items, and normalizing the data. The same preprocessing is also performed on the emotion data, preparing it in a format suitable for analysis.

[1737] Step 5:

[1738] The server extracts features necessary for risk assessment from the preprocessed data. Features extracted include, for example, "frequency of past delivery delays," "conditions for legal violations," and "results of employee health assessments." Emotional features such as "stress level," "happiness level," and "anxiety level" are extracted from the emotion data.

[1739] Step 6:

[1740] The server runs a risk analysis algorithm based on the extracted features and emotion features. This algorithm includes machine learning models and statistical models to evaluate the probability of risk occurrence and its impact.

[1741] Step 7:

[1742] Based on the results of the risk analysis, the server uses a generative AI to generate risk countermeasures. The generative AI also takes emotional features into account to generate countermeasures that take into consideration the user's mental health. For example, it may suggest reallocating work to reduce stress or a specific action plan to mitigate risks.

[1743] Step 8:

[1744] The server transmits the generated risk countermeasures and related emotion data to the user terminal, and the user uses the terminal to check the countermeasures.

[1745] Step 9:

[1746] The user checks the proposed countermeasures through the device and selects the most appropriate one. The selected countermeasure is then sent back to the server from the device.

[1747] Step 10:

[1748] Implement the user-selected measures, such as finding alternative suppliers, increasing inventory, or implementing employee wellness programs within the company.

[1749] Step 11:

[1750] The user uses the device to input the results of the measures taken, and the emotion engine also collects the user's emotional data at that time and sends it to the server to provide feedback.

[1751] Step 12:

[1752] The server stores the received feedback and emotion data in a database, which is then used for the next risk assessment and countermeasure generation.

[1753] Step 13:

[1754] The server updates the risk assessment model and the parameters of the generative AI based on the feedback and emotion data, thereby improving the accuracy and effectiveness of the next analysis and generated countermeasures.

[1755] Example 2

[1756] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1757] In operational risk management in companies, there is a need for the collection and analysis of risk information, the generation and implementation of countermeasures, and flexible countermeasure proposals that take user emotions into account. However, conventional systems have difficulty effectively incorporating user emotions, resulting in insufficient accuracy and effectiveness in risk management. Therefore, there is a need to develop a system that takes user emotional data into account, flexibly generates risk countermeasures, and utilizes the results of their implementation as feedback.

[1758] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1759] In this invention, the server includes a means for a user to input risk-related data, a means for transmitting the data and the user's emotion data to the server, and a means for storing the data and the emotion data in a database. This allows the user's emotions to be taken into account in risk analysis and countermeasure generation, enabling more appropriate risk countermeasures. Furthermore, by feeding back the execution results of the selected countermeasures to the server along with the emotion data, the model is updated, improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[1760] "User" refers to a company employee or system user who inputs risk-related data and checks and selects countermeasures.

[1761] A "terminal" is an electronic device through which a user inputs risk-related data and emotion data and transmits the data to a server.

[1762] The "server" is a central management system that receives, stores, analyzes risk-related data and emotion data, and generates countermeasures.

[1763] A "database" is a storage device for storing risk-related data and emotion data received by the server.

[1764] An "emotion engine" is a software or hardware component that recognizes a user's emotions in real time and collects them as data.

[1765] "Preprocessing" refers to the process of imputing missing values, removing unnecessary items, and normalizing the data received.

[1766] "Feature extraction" is the process of extracting elements from data that are important for risk assessment and analysis.

[1767] "Risk analysis" is a process of evaluating the probability of occurrence and the impact of a risk based on the extracted feature quantities.

[1768] "Generative AI" is an artificial intelligence algorithm that automatically generates countermeasures based on the results of risk analysis, taking into account the user's emotions.

[1769] "Feedback" is a process in which the execution results of the measures selected by the user and emotion data are sent to the server.

[1770] "Model updating" is a process in which the parameters of the risk assessment model and generation AI are modified based on feedback data to improve the accuracy and effectiveness of the next analysis and countermeasure generation.

[1771] This invention relates to a system for improving the accuracy and effectiveness of risk management in a company's operational risk management by combining risk information collection and analysis, the generation and implementation of countermeasures, and an emotion engine that recognizes user emotions. Specific embodiments for implementing this invention will be described below.

[1772] Hardware and Software Configuration

[1773] The system consists of the following components:

[1774] Device: An electronic device through which a user enters risk-related data. Examples include a personal computer, tablet, or smartphone.

[1775] Server: A central management system that receives, stores, analyzes, and generates countermeasures. A server with a high-performance processor and large storage capacity is recommended.

[1776] Database: A storage device for saving data received by the server. A relational database management system (RDBMS) is used.

[1777] Emotion engine: A software or hardware component that recognizes a user's emotions in real time. For example, there are emotion recognition systems that use facial recognition technology or voice analysis technology.

[1778] System operation details

[1779] The user inputs risk-related data via a terminal.

[1780] Users input risk-related data from their own devices, such as supplier delivery delays, legal violation risks, and employee health status. In addition, an emotion engine recognizes and collects the user's emotions (e.g., stress, anxiety, etc.) in real time as they input their data.

[1781] The device sends input data and emotion data to the server.

[1782] The terminal transmits the input risk-related data and emotion data to the server, whereby the risk-related data and the user's emotion data are stored.

[1783] The server stores the data in a database and performs preprocessing.

[1784] The server stores the received risk data and emotion data in a database, then performs preprocessing on the stored data, such as filling in missing values, removing unnecessary items, and normalizing the data.

[1785] The server extracts features and analyzes risks

[1786] From the preprocessed data, the server extracts the features necessary for risk assessment, then runs an algorithm to evaluate the probability of risk occurrence and its impact.

[1787] The server generates risk countermeasures using generation AI

[1788] Based on the analysis results, the server applies generative AI to generate risk countermeasures. This also takes into account the user's emotional data, so if the user is feeling excessive stress, for example, countermeasures to alleviate that situation will be proposed.

[1789] The server sends the generated countermeasures and emotion data to the user terminal.

[1790] The generated countermeasures are sent from the server to the user's device, where the user can check them. Emotion data is also displayed, which can be used as a reference when selecting a countermeasure.

[1791] The user checks the proposed measures and selects one

[1792] The user can use the device to check the generated countermeasures and select the countermeasures to be implemented. This process allows the selection of countermeasures that reflect the user's feelings.

[1793] The user executes the selected measure and enters the results into the device.

[1794] After the selected countermeasure is executed, the user sends feedback of the result and effect from the terminal to the server. This feedback includes the result of the execution of the countermeasure and the user's emotional data.

[1795] The server updates the model based on the feedback

[1796] The server stores the feedback data and emotion data in a database and uses them to update the risk assessment model and the parameters of the generation AI, thereby improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[1797] Examples of concrete examples and prompts

[1798] For example, when a manufacturing company performs risk management for its supply chain, the user inputs data such as "data regarding delivery delays from major suppliers" and "risk of supply chain interruptions" into a terminal. In addition, an emotion engine simultaneously collects user emotional data such as "stress" and "anxiety." The server preprocesses this data, performs risk analysis, and then uses generative AI to generate countermeasures that take the emotional data into account. The user reviews these countermeasures, selects the most appropriate one, and implements it. After implementation, the results and emotional data are sent back to the server and used to update the model.

[1799] Prompt Sentence Examples

[1800] "Please provide data on delivery delays from your key suppliers. Also, please rate your current stress level on a scale of 1 to 10."

[1801] "Please provide detailed information about the risk of supply chain interruption and indicate your level of concern about that risk on a scale of 0 to 100."

[1802] In this way, the present invention realizes risk management that takes into account users' emotions, and can effectively minimize potential losses from enterprise risks.

[1803] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1804] Step 1:

[1805] The user inputs risk-related data via a terminal.

[1806] Users use a terminal to input risk-related data. This data includes information such as "supplier delivery delays," "risk of legal violations," and "employee health status." The emotion engine recognizes and collects the user's emotions (such as "stress" or "anxiety") in real time as they input data. Specifically, users enter data using a keyboard or touchscreen, and the camera and microphone analyze facial expressions and tone of voice to collect emotional data.

[1807] Input: Risk-related data, user sentiment data

[1808] Output: None (passed to next step)

[1809] Step 2:

[1810] The device sends input data and emotion data to the server.

[1811] The device sends the risk data entered by the user and the emotion data collected by the emotion engine in a specific format (e.g., JSON format) to the server, where the data is checked for integrity and any necessary encryption is applied.

[1812] Input: Risk-related data, user sentiment data

[1813] Output: Data sent to the server

[1814] Step 3:

[1815] The server stores the data in a database and performs preprocessing.

[1816] The server stores the received risk and emotion data in a database. It then performs preprocessing on the data, such as filling in missing values, removing unnecessary items, and normalizing the data. For example, if there are missing values, they are filled in with the average value, unnecessary items are removed, and numerical data is standardized.

[1817] Input: Data sent to the server

[1818] Output: Preprocessed data

[1819] Step 4:

[1820] The server extracts features and analyzes risks

[1821] From the preprocessed data, the server extracts the features necessary for risk assessment. At this time, it applies data analysis algorithms to evaluate the probability of risk occurrence and its impact. For example, it uses a machine learning model to calculate a risk score and extract important variables necessary for risk assessment.

[1822] Input: Preprocessed data

[1823] Output: Risk analysis results (feature amount, risk occurrence probability, impact)

[1824] Step 5:

[1825] The server generates risk countermeasures using generation AI

[1826] The server applies a generative AI based on the risk analysis results to generate risk countermeasures. This generative AI also takes into account the user's emotional data to create appropriate risk countermeasures. For example, the generative AI can propose multiple countermeasures using the risk score and the user's emotional data as input.

[1827] Input: Risk analysis results, user emotion data

[1828] Output: Risk countermeasure proposal

[1829] Step 6:

[1830] The server sends the generated countermeasures and emotion data to the user terminal.

[1831] The generated countermeasures are sent from the server to the user's device, where the user can check the countermeasures and view the emotional data.

[1832] Input: Risk countermeasures, emotional data

[1833] Output: Countermeasures and emotion data sent to the device

[1834] Step 7:

[1835] The user checks the proposed measures and selects one

[1836] The user can then use their device to review the proposed countermeasures and select the appropriate one. This process also takes into account emotional data, allowing the user to select the countermeasure that best suits their situation. For example, the user can select the countermeasure that seems easiest to implement from the provided countermeasures.

[1837] Input: Countermeasures and emotion data sent to the device

[1838] Output: Selected countermeasures

[1839] Step 8:

[1840] The user executes the selected measure and enters the results into the device.

[1841] The user implements the selected countermeasure and sends feedback on the results and effects from the terminal to the server. This feedback includes the results of the countermeasure and emotional data. For example, the success rate of the countermeasure and the emotional state after implementation can be input.

[1842] Input: Execution results of selected countermeasures and emotion data

[1843] Output: Feedback data sent to the server

[1844] Step 9:

[1845] The server updates the model based on the feedback

[1846] The server stores the received feedback data in a database and uses it to update the parameters of the risk assessment model and the generation AI, thereby improving the accuracy and effectiveness of the next analysis and countermeasure generation.

[1847] Input: Feedback data

[1848] Output: Updated risk assessment model and generative AI

[1849] (Application example 2)

[1850] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1851] Conventional operational risk management systems in companies have the problem that collecting and analyzing risk data takes time, making it difficult to respond in real time. In addition, because they do not take into account the emotions of workers, measures to reduce stress and anxiety are insufficient, making it difficult to improve production efficiency and reduce workers' psychological stress. This has resulted in the accuracy and effectiveness of risk management being insufficient, placing a strain on corporate operations.

[1852] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for inputting risk-related data, means for transmitting the data to the server, means for saving the data in a database, means for preprocessing the saved data, means for extracting features from the preprocessed data, means for analyzing risks based on the extracted features, means for applying a generation AI that generates risk countermeasures based on the analysis results, means for transmitting the generated countermeasures to a user terminal, means for the user to confirm and select the countermeasures, means for feeding back the execution results of the selected countermeasures to the server, means for updating the model based on the feedback, means for collecting risk data and emotion data from industrial robots and worker terminals in real time, means for generating and presenting countermeasures taking the data and emotion data into consideration, and means for displaying the countermeasures on the industrial robot's display device or the worker's terminal. This makes it possible to collect and analyze risk data and worker emotion data in real time and quickly implement highly accurate risk countermeasures.

[1853] "Risk-related data" refers to data necessary to assess a company's operational risks, including the condition of equipment, the progress of processes, and the health status of personnel.

[1854] A "server" is a computer system that receives, stores, and analyzes risk data via a network.

[1855] A "database" is a data storage system for systematically storing risk-related data and emotion data.

[1856] "Preprocessing" is the process of imputing missing values, removing unnecessary items, and normalizing data to prepare it in an analyzable format.

[1857] "Features" are extracted important data points and indicators necessary for risk analysis.

[1858] "Analyzing risk" means evaluating the probability and impact of a risk based on collected data and predicting potential dangers.

[1859] "Generative AI" is an artificial intelligence system that uses machine learning and AI technology to generate risk countermeasures.

[1860] A "user terminal" is a device that a user uses to input risk data and check countermeasure proposals, such as a smartphone or tablet.

[1861] "Feedback" refers to returning data to the server about the results and effects of the measures taken by the user.

[1862] "Updating the model" means improving and adjusting the parameters of the analytical model and generation AI based on feedback data, thereby improving the accuracy of the next analysis and countermeasure generation.

[1863] "Emotion data" is data that quantifies the user's psychological state, such as stress level and anxiety level.

[1864] An "industrial robot" is a robot used to assist or automate work within a factory, and is a mechanical device equipped with various sensors that has the ability to collect risk data in real time.

[1865] System Overview

[1866] The operational risk management system of this invention collects risk factors and worker emotion data in a factory in real time, and analyzes and generates countermeasures based on this data. The system is composed of various components, including industrial robots, worker terminals, servers, databases, and generation AI.

[1867] Data collection

[1868] Users and industrial robots collect risk-related data in real time within the factory, including equipment failure status, the progress of production processes, and the health status of workers. Workers also use smart glasses and terminals to input emotional data such as their stress and anxiety levels.

[1869] Data transmission and storage

[1870] The terminals and industrial robots send the collected risk and emotion data to a server, where it is stored in a database. The server then performs preprocessing on the stored data, such as filling in missing values, removing unnecessary items, and normalizing the data.

[1871] Data analysis

[1872] The server extracts features from the preprocessed data and performs analysis to evaluate the probability of risk occurrence and its impact, using machine learning algorithms and data analysis software (e.g., pandas, numpy, scikit-learn).

[1873] Generate countermeasures

[1874] Based on the analysis results, the server uses a generation AI to generate risk countermeasures. The generation AI also takes into account the user's emotional data when generating countermeasures, so if the user is feeling excessive stress, for example, it will propose countermeasures to alleviate that situation.

[1875] Proposing and implementing countermeasures

[1876] The generated countermeasures are sent from the server to the user's terminal or the display device of the industrial robot, where the user can check the countermeasures and select the most appropriate one. The selected countermeasure is then implemented.

[1877] Feedback and Model Updates

[1878] The user feeds back the results of the countermeasures they have implemented to the server via their device. This feedback includes not only the results of the countermeasures but also the user's emotional data. The server updates the model based on the feedback data and emotional data, improving the accuracy and effectiveness of the next analysis and generation of countermeasures.

[1879] Specific examples

[1880] For example, when a major machine breaks down in a factory, the industrial robot sends real-time data on the situation to a server. At the same time, the worker sends their stress level to the server through smart glasses. The server analyzes this data and uses generative AI to generate appropriate repair plans and measures to reduce worker stress, such as break suggestions. These measures are then displayed on the industrial robot's display and the worker's device.

[1881] Prompt Sentence Examples

[1882] "When equipment failure occurs in a factory, please collect information on the situation in real time and generate optimal repair measures. Please also suggest measures to reduce worker stress."

[1883] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1884] Step 1:

[1885] Users and industrial robots collect risk-related data within factories. Specifically, industrial robots use sensors to obtain information about the equipment's operating status and manufacturing processes, while workers use smart glasses to input their own health and psychological state (stress level, anxiety level). The data input here covers a wide range of topics, including equipment status, production line delays, and worker fatigue levels.

[1886] Step 2:

[1887] The terminal sends the collected risk data and emotion data to a server. The terminal then compiles and packages this data and sends it to the server via the network. The transmitted data is input to the server in the form of information such as the operating status of equipment, production delays, and worker stress levels.

[1888] Step 3:

[1889] The server stores the received data in a database. The server stores the transmitted risk and emotion data in the database and tags them appropriately. The stored data also includes time information and the data source (equipment, worker, etc.).

[1890] Step 4:

[1891] The server preprocesses the stored data by removing unnecessary items, filling in missing values, and normalizing the data. The server first removes meaningless and outliers from the data, then infers and fills in missing data points. Finally, the data is converted into a unified format and prepared for feature extraction.

[1892] Step 5:

[1893] The server extracts features from the preprocessed data. After data normalization is complete, the server uses a machine learning algorithm to extract important features necessary for risk assessment. The output of this process is input data for assessing the probability of risk occurrence and its impact.

[1894] Step 6:

[1895] The server analyzes risks based on the extracted features. Using a risk assessment model based on the features, the server calculates and assesses the probability of risk occurrence and the impact. This analysis clarifies which risks are likely to occur and the extent of their impact.

[1896] Step 7:

[1897] The server uses a generative AI based on the analysis results to generate risk countermeasures. Based on the analysis results, the server uses generative AI (e.g., GPT-3 model) to automatically generate optimal risk countermeasures. Countermeasures that take emotional data into consideration are generated here. The generated countermeasures are specific action plans for the countermeasures.

[1898] Step 8:

[1899] The server sends the generated countermeasure plan to the user's terminal or the display device of the industrial robot. The generated countermeasure plan is sent from the server to the user's terminal or the display device of the industrial robot, where it can be checked by the user or worker. The presented countermeasure plan includes the implementation procedure and expected effects.

[1900] Step 9:

[1901] The user or worker reviews the proposed countermeasures and selects the most appropriate one. After the proposed countermeasures are displayed on a terminal or display device, the user or worker reviews the contents and selects the countermeasure to be implemented. The selected countermeasure is sent back to the server as confirmation of the selection.

[1902] Step 10:

[1903] The user implements the selected measure, and the results are fed back from the terminal to the server. After the user or worker implements the measure, the results, effectiveness, and emotional state at the time are entered into the terminal as feedback and sent to the server. This data is used to evaluate the effectiveness of the measure.

[1904] Step 11:

[1905] The server updates the model based on the feedback data. After receiving the feedback data, the server saves it in the database and updates the parameters of the risk assessment model and the generation AI. This update improves the accuracy of the next analysis and countermeasure generation.

[1906] 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 control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1907] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1908] 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 the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1909] The emotion identification model 59 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 an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1910] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1911] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1912] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1913] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1914] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs 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 a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I ...

Claims

1. a means for inputting risk-related data; means for transmitting said data to a server; means for storing said data in a database; means for preprocessing the stored data; means for extracting features from the preprocessed data; A means for analyzing risk based on the extracted feature quantity; A means of applying generative AI to generate risk countermeasures based on the analysis results; means for transmitting the generated countermeasures to a user terminal; A means for the user to review and select countermeasures; a means for feeding back the execution result of the selected countermeasure to the server; and means for updating the model based on said feedback.

2. The system of claim 1 further comprising preprocessing means for imputing missing values, removing unnecessary items, and normalizing the data.

3. The system according to claim 1 , further comprising means for evaluating the probability of occurrence and the impact of a risk based on the analysis result.

Images