system

Abstract

We provide the system. [Solution] Means for monitoring the operating status and progress of the equipment being used, A means of dynamically reallocating tasks, A means for optimizing the performance of a device using a reinforcement learning algorithm, A means of adjusting settings based on user feedback, A means to equalize the load on equipment within a data center, A system that includes this.

Description

Technical Field

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

Background Art

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

Summary of the Invention

Problems to be Solved by the Invention

[0004] In an artificial intelligence system including a plurality of agents, the roles and loads of each agent are not appropriately managed, and as a result, the capabilities of the agents are not fully exerted and the overall performance deteriorates. Furthermore, there is a problem that the accuracy of task execution does not improve because there is no mechanism for efficiently promoting cooperation between agents. There is a need for a management method to solve these problems and improve work efficiency.

Means for Solving the Problems

[0005] To address the above challenges, the present invention provides means for monitoring the operational status and progress of agents, thereby enabling real-time monitoring of each agent's status. It also includes means for dynamically reassigning tasks, allowing for adjustment of workload imbalances. Furthermore, it incorporates means for optimizing agent performance using reinforcement learning algorithms, enabling continuous improvement. It provides means for adjusting settings based on user feedback, allowing for flexible operation in response to user requests. This enables efficient management of cooperation between agents, and by combining this with an alert function from an anomaly detection system, agent performance can be maximized.

[0006] An "agent" is a program or system designed to perform a specific task, and it operates autonomously using artificial intelligence.

[0007] "Operational status" refers to information that shows what tasks the agent is currently performing and how much resources it is using.

[0008] "Progress" is an indicator that represents the degree to which an agent has completed their assigned tasks, or the percentage of the task that has been achieved relative to the target.

[0009] "Monitoring" refers to the act of observing the performance and status of an agent in real time and collecting the necessary data.

[0010] "Task reallocation" refers to the process of transferring tasks that were handled by a particular agent to another agent, and is a technique used to balance the workload.

[0011] Reinforcement learning is a machine learning technique in which an agent interacts with its environment, repeatedly engaging in trial and error to select the optimal course of action.

[0012] "Optimization" is the process of adjusting the settings and resource allocation of each component to maximize the overall efficiency and performance of a system.

[0013] "Feedback" refers to information such as opinions and requests obtained from users, and is used to help operate and improve the system.

[0014] "Configuration adjustment" refers to the operation of changing system and agent parameters according to user requirements and operational conditions to achieve an optimal state.

[0015] "Cooperative system" refers to an organized framework and division of roles among multiple agents to carry out tasks together.

[0016] "Anomaly detection" refers to a series of techniques used to immediately identify and notify of problems when an agent exhibits behavior outside its normal operating range.

[0017] An "alert" is a warning message sent to users or administrators when an anomaly or situation requiring attention occurs in a system or agent. [Brief explanation of the drawing]

[0018] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6]It is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] It is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] It is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] It shows an emotion map to which multiple emotions are mapped. [Figure 10] It shows an emotion map to which multiple emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

Embodiments for Carrying Out the Invention

[0019] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0020] First, the language used in the following description will be explained.

[0021] In the following embodiments, the signed processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Furthermore, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include CPU (Central Processing Unit), GPU (Graphics Processing Unit), GPGPU (General-Purpose computing on Graphics Processing Units), and APU (Accelerated Processing Unit).

[0022] In the following embodiments, signed RAM (Random Access Memory) is a memory that temporarily stores information and is used as work memory by the processor.

[0023] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0024] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0025] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0026] [First Embodiment]

[0027] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0028] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0029] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0030] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0031] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0032] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0033] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0034] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0035] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0036] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0037] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0038] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0039] This invention provides a method for efficiently managing a multi-agent system utilizing artificial intelligence and optimizing task execution. Specifically, it includes a mechanism in which a server monitors the operational status and progress of each agent in real time and dynamically reassigns tasks as needed. It also has a function to continuously improve agent performance through reinforcement learning and adjust system settings based on user feedback.

[0040] The server periodically collects and analyzes information from each agent. This information includes the operational status and task progress of each agent, and immediately generates an alert if an anomaly is detected. The server uses this information to redistribute tasks among agents. For example, if an agent is overloaded, the server will assign some of its tasks to agents with less load.

[0041] The server uses reinforcement learning algorithms to analyze past data and optimize task distribution among agents. This process refines the agents' cooperation over time, improving overall performance.

[0042] Users can provide feedback through their devices. This feedback is analyzed by the server and reflected in the system settings. For example, if a user instructs the server to prioritize tasks in a certain category, the server will immediately implement task allocation that takes this into account. This entire process is automated, minimizing user intervention and enabling efficient operation.

[0043] This invention enables the optimal coordination of multiple AI agents used within an enterprise, allowing various tasks to be performed efficiently and quickly. Immediate response and optimization processes based on anomaly detection reduce wasted resources. This provides a mechanism for improving overall productivity.

[0044] The following describes the processing flow.

[0045] Step 1:

[0046] The server collects operational status and progress data from each agent. This includes which tasks each agent is currently running, the amount of resources being used, and the degree of task completion. The server stores this information in a central database.

[0047] Step 2:

[0048] The server analyzes the collected data to check for any anomalies. If an anomaly is detected, the server generates an alert and notifies the administrator. This allows problems to be detected early and corrective actions to be taken.

[0049] Step 3:

[0050] The server determines whether tasks need to be dynamically reallocated based on the agent load. If the load is uneven, the server shifts some tasks to other agents with less load, thereby maintaining overall balance.

[0051] Step 4:

[0052] The server activates a reinforcement learning algorithm and learns from past performance data. This algorithm generates a new optimization strategy, which will be used for future task assignments.

[0053] Step 5:

[0054] Users send feedback to the server using their devices. This feedback may include changes to the priority of specific tasks or new requests. The server receives this feedback and adjusts the agent settings. This ensures that user requests are efficiently reflected in system operations.

[0055] (Example 1)

[0056] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0057] In recent years, many companies and organizations have implemented systems in which multiple agents simultaneously perform various tasks. However, there is a need for efficient operation of these agents, dynamic task reassignment, early detection of anomalies, and flexible system configuration that takes user feedback into consideration. Conventional systems have relied on manual work and fixed settings, and have suffered from a lack of real-time adaptability.

[0058] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0059] In this invention, the server includes means for periodically collecting the operating status and work progress of agents, means for analyzing the collected data and detecting operational anomalies, and means for dynamically redistributing tasks based on the analysis results. This enables efficient operation of agents and flexible and adaptive system operation.

[0060] An "agent" is a unit of software or hardware designed to perform a specific task.

[0061] "Operational status" refers to information indicating the progress of the tasks being performed by the agent and the resource usage.

[0062] "Work progress" is an indicator that shows the extent to which the tasks assigned to the agent have been completed.

[0063] "Collecting" means periodically acquiring and accumulating specific information as numerical data or logs.

[0064] "Analyzing" means evaluating collected data using computational processing and algorithms to extract useful information and patterns.

[0065] "Detecting anomalies" means discovering conditions or performance problems that deviate from normal operation.

[0066] "Dynamic reallocation" means flexibly rearranging existing tasks and resources to suit new states and conditions.

[0067] "Reinforcement learning technology" is a method for improving performance by enabling a system to learn optimal actions based on its experience.

[0068] "User input information" refers to data related to instructions and feedback that users provide to the system.

[0069] "Optimizing settings" means adjusting parameters and configurations to achieve efficient system operation.

[0070] This invention provides a solution for effectively managing a multi-agent system utilizing artificial intelligence and appropriately allocating tasks. The server periodically collects information on the operational status and work progress from each agent. This allows the server to understand the overall system status in real time. The collected information is acquired as data packets via the network and stored in a database.

[0071] The server applies reinforcement learning algorithms to analyze the collected data and optimize task allocation among agents. During this process, the server evaluates past task performance results and learns practical task allocation strategies from successful and unsuccessful examples. This analysis can be performed using data analysis programming languages ​​such as Python or R.

[0072] Furthermore, the server can immediately generate alerts if it detects any abnormalities in agent operation and reassign tasks as needed. For example, if an agent is overloaded, it can transfer tasks from that agent to other agents to maintain overall performance. This reassignment is automated, improving efficiency across the entire enterprise.

[0073] Users can provide feedback to the system through their terminals. Specifically, users input task priorities and new instructions using a GUI, and this data is sent to the server. The server analyzes this feedback in real time and can adjust system settings as needed. This optimization of settings allows the system to flexibly respond to user needs.

[0074] As a concrete example, consider a large-scale data analysis project where a sudden increase in computational demands causes some agents to reach their limits. The server detects this situation and immediately redistributes tasks to other agents to resolve the problem. Furthermore, if a user provides feedback requesting priority processing of a specific data category, the server adjusts task priorities accordingly. Such applications are conceivable.

[0075] Examples of prompts for a generative AI model include the following:

[0076] "You are the server managing AI agents within the company. Check the operational status of Agent A, and if it is overloaded, redistribute tasks to Agent B. When doing so, use a reinforcement learning algorithm to determine the optimal distribution and adjust the settings based on user feedback."

[0077] The system realized by this invention enables the effective management of multiple agents within a company or organization, allowing them to perform tasks efficiently and flexibly with minimal intervention.

[0078] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0079] Step 1:

[0080] The server periodically collects information on the operational status and work progress from each agent. In this step, it receives data packets from agents via the network. The specific input is log data of the agents' operational status and progress, and by saving this to a database, it obtains output that is ready for the next analysis process.

[0081] Step 2:

[0082] The server analyzes the collected data to detect operational anomalies. The operational data saved in Step 1 is used as input, and a programming language (e.g., Python) or statistical software is used for analysis. The server analyzes the agent's CPU usage and memory usage, and if these exceed the threshold, it determines it to be an anomaly and generates an alert. This alert serves as a basis for deciding on reallocation.

[0083] Step 3:

[0084] The server dynamically redistributes tasks based on the analysis results. The input consists of anomaly information detected in step 2 and the agent load status. The server uses a reinforcement learning algorithm to evaluate the performance of each agent and transfers tasks to agents with less load. This process outputs an efficient task redistribution.

[0085] Step 4:

[0086] Users provide feedback through their terminals, which is then analyzed by the server. The input consists of user instructions regarding configuration changes and task priorities. The server receives and analyzes this feedback information to adjust system settings as needed. This feedback-based optimization is the final output, improving the system's operational efficiency.

[0087] In this way, the system collects, analyzes, assigns tasks to, and incorporates feedback in real time, continuously optimizing the entire system.

[0088] (Application Example 1)

[0089] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0090] Efficiently preventing load imbalances and malfunctions in data centers is challenging. Conventional systems struggle to accurately understand the operating status of equipment, making it difficult to respond quickly to equipment experiencing heavy loads. Furthermore, dynamic adjustments based on individual feedback are difficult, hindering overall system optimization.

[0091] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0092] In this invention, the server includes means for monitoring the operating status and progress of the equipment being used, means for dynamically reassigning tasks, and means for optimizing the performance of the equipment using a reinforcement learning algorithm. This enables efficient equalization of the load within the data center, prevention of equipment malfunctions, and optimal system adjustments incorporating user feedback.

[0093] "Equipment used" refers to the collective term for electronic devices such as servers and computers that are operated within a data center.

[0094] "Operating status" refers to information about the current operating status and resource consumption of the equipment being used.

[0095] "Progress" is an indicator that shows the extent to which a task or work has been completed.

[0096] "Work" refers to the procedures for processing tasks and processes within a data center.

[0097] "Load" refers to the pressure exerted by the amount of data and computational load on the equipment being used.

[0098] "Reassignment" refers to the process of redistributing tasks or resources to different devices.

[0099] A "reinforcement learning algorithm" is a type of machine learning that learns the optimal action through trial and error.

[0100] "Performance" refers to the ability and efficiency with which the equipment used can perform a given task.

[0101] "Optimization" is the process of adjusting a system to maximize its overall efficiency and effectiveness.

[0102] "An anomaly" refers to an unexpected state that deviates from normal operation.

[0103] "Feedback" refers to information provided based on users' experiences and requests.

[0104] "Equalization" refers to a state where resources and workloads are distributed evenly to eliminate imbalances.

[0105] The system that realizes this invention is primarily composed of a server. The server is programmed using Python and TENSORFLOW® and collects and monitors the operating status and progress of equipment used in the data center in real time. The server analyzes this information and evaluates the workload on each piece of equipment. If the workload is concentrated on a particular piece of equipment, it dynamically reassigns the work to equalize the overall load.

[0106] The reinforcement learning algorithm is implemented using TensorFlow, and the server optimizes the performance of the equipment. This enables optimal task allocation based on historical data. Furthermore, the server can dynamically adjust system settings based on user feedback, new requests, and improvements.

[0107] As a concrete example, suppose a data center has a group of servers processing a critical analytical task, and some of the servers become overloaded. In this case, the system on the servers immediately monitors the situation and reassigns tasks from the overloaded servers to the less loaded servers. This ensures overall processing efficiency and prevents problems from occurring.

[0108] An example of a prompt for a generative AI model is as follows: "Please suggest the optimal method and parameter settings for building a reinforcement learning model that monitors server load in a data center and optimizes the task."

[0109] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0110] Step 1:

[0111] The server collects operational status and progress data from the devices being used in real time. Inputs include CPU usage, memory usage, and I / O wait times provided by each device, which are then imported into an internal database. Outputs are the latest operational status data for each device. Specifically, data is periodically retrieved from each device using an internal communication protocol.

[0112] Step 2:

[0113] The server analyzes the collected data and evaluates the load status of each device. The input is operational status data from step 1, and the output is a list showing the load status of each device. This process uses statistical methods to calculate the load and identify devices that may be overloaded. Specifically, it sets thresholds using an anomaly detection algorithm and lists devices whose load has been exceeded.

[0114] Step 3:

[0115] The server reallocates tasks from heavily loaded devices to other less-loaded devices. The input is the load status list obtained in step 2. The output is the task list for each device after reallocation. In this process, dynamic programming techniques are used to equalize the load. Specifically, the server plans and executes the movement of tasks from heavily loaded devices to less loaded ones.

[0116] Step 4:

[0117] The server continuously improves the performance of each device by applying reinforcement learning algorithms. Inputs are progress reports and user feedback, and output is a model of optimized task allocation. The server uses TensorFlow to learn from historical data and generate optimal agent behavior patterns. Specifically, it analyzes performance data and updates the optimal course of action.

[0118] Step 5:

[0119] Terminal users review the results of system configurations optimized by the server and provide feedback as needed. Inputs are optimized task allocations and feedback requests from the server, while output is user feedback information. Specifically, configuration verification and feedback input are performed via a GUI interface.

[0120] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0121] This invention provides a method to improve system efficiency and user satisfaction by combining an emotion engine that recognizes user emotions with a multi-agent system that utilizes artificial intelligence. A server acts as the central point, monitoring the operational status and progress of agents in real time and dynamically reassigning tasks as needed. Agent performance is optimized using reinforcement learning algorithms.

[0122] An emotion engine is used to analyze the user's emotional state on the device. The collected emotional information is sent to the server and reflected in task priorities and agent settings. Based on this emotional information, the server adjusts the cooperation between agents to achieve more adaptive task execution. In addition, system settings can be adjusted in conjunction with the emotion engine based on user feedback.

[0123] For example, if the emotion engine detects that a user is experiencing stress, the server reassigns high-load tasks to another agent, reducing the user's workload. This improves the user experience.

[0124] Furthermore, if a user expresses satisfaction or dissatisfaction, the server will adjust task allocation among agents accordingly. This setting can be flexibly changed according to user requests and the work environment. If an anomaly is detected, the server will immediately issue an alert and pinpoint the source of the problem.

[0125] This system enables AI agents used within a company to work efficiently and harmoniously, adapting to various situations while performing tasks. Through integration with an emotion engine, the system achieves more human-like interactions, improving overall productivity and user satisfaction.

[0126] The following describes the processing flow.

[0127] Step 1:

[0128] The server collects operational status and progress data from each agent. This includes the current task status, resources being used, and the degree of work completion. The server analyzes this data and monitors the agent's performance.

[0129] Step 2:

[0130] The device monitors the user's actions and inputs, and uses an emotion engine to analyze the user's emotional state in real time. The device then sends these analysis results to the server.

[0131] Step 3:

[0132] Based on the user's emotional information, the server adjusts the priority and allocation of agent tasks. For example, if a user is stressed, the server will reallocate tasks to reduce the load on them.

[0133] Step 4:

[0134] The server executes a reinforcement learning algorithm, leveraging historical performance data and user sentiment information to optimize the overall system efficiency. This allows the agent to act more adaptively in subsequent task assignments.

[0135] Step 5:

[0136] Users provide feedback using their devices. This feedback may include task priorities and satisfaction with the process. The server aggregates this feedback and uses it to improve system settings and agent collaboration.

[0137] Step 6:

[0138] The server uses an anomaly detection module to immediately generate alerts and take appropriate action if problems occur in the operation of agents or the system as a whole. This is important for speeding up problem resolution and maintaining system stability.

[0139] (Example 2)

[0140] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0141] In recent years, AI-powered business support systems have become widespread. However, these systems fail to adequately address users' emotional states, limiting their potential for improving work efficiency and user satisfaction. Furthermore, the optimization of cooperation between programs within the system is insufficient, making it difficult to maximize the overall effectiveness of business operations. In addition, they lack the functionality to respond immediately when an anomaly occurs, highlighting the need for rapid problem resolution.

[0142] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0143] In this invention, the server includes means for analyzing the user's emotional state, means for dynamically adjusting task priorities based on the analyzed emotional information, means for monitoring a group of autonomous programs with multiple roles and reassigning roles as needed, means for improving the operational efficiency of the program group using machine learning algorithms, and means for modifying the system configuration based on user input. This enables flexible task management in response to the user's emotions, thereby improving operational efficiency and maximizing user satisfaction.

[0144] "Methods for analyzing a user's emotional state" refer to technologies that detect emotions in real time from the user's facial expressions, tone of voice, etc., and analyze emotional trends based on that data.

[0145] "A means of dynamically adjusting task priorities based on analyzed emotional information" refers to a method that utilizes the results of analyzing user emotional data to constantly optimize the order in which tasks are performed in a way that does not burden the user.

[0146] A "group of autonomous programs with multiple roles" is a collection of software that performs different functions independently, but is designed to work together as a whole to achieve a specific objective.

[0147] "Methods for improving the operational efficiency of a group of programs using machine learning algorithms" refers to techniques that use artificial intelligence algorithms to analyze the operational data within a program and improve its efficiency based on that analysis.

[0148] "Means of modifying system configuration based on user input" refers to the process of adjusting and improving system settings and operation based on user feedback and operation history.

[0149] This invention is a multi-agent system that recognizes the user's emotional state in real time and optimizes the system's operation based on that recognition. The following describes a specific implementation of this system.

[0150] The main components consist of a server, a terminal, and a user. The server functions as the core of the system, and the terminal collects user emotion data via an emotion engine. The terminal is equipped with image recognition software and voice analysis tools to analyze the user's voice and facial expressions, specifically using libraries such as OpenCV. These tools receive data through the user interface and send it to the server.

[0151] The server uses machine learning algorithms to analyze the received sentiment data. At this stage, it monitors the agent's activity and dynamically optimizes task priorities based on the analysis results. By applying reinforcement learning algorithms, the cooperative relationships between agents are automatically adjusted. For example, if the analysis indicates that the user is experiencing stress, the server reduces the user's workload by assigning high-load tasks to other agents.

[0152] Furthermore, the terminal provides the user with feedback on instructions from the server and changes in status, and adjusts the overall system settings as needed. For this purpose, a dashboard application is used to provide visual information. The server collects user feedback and uses it to adjust system settings for further optimization.

[0153] A concrete example of how this system works is when a user is stressed due to project deadlines. In this situation, the emotion engine detects the stress level, and the server quickly reassigns tasks. This allows the project to progress smoothly, and the user can achieve better results.

[0154] An example of a prompt might be, "How should the system reassign tasks when the user is experiencing stress?" Using such prompts, it's possible to leverage generative AI models to further improve and adapt the system.

[0155] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0156] Step 1:

[0157] The device collects the user's emotions using an emotion engine. Here, the user's facial expressions are captured by a camera, and their voice is recorded by a microphone. Image data and audio data are acquired as input, and image recognition software and audio analysis tools are used to analyze them. Specifically, OpenCV is used to analyze facial expressions from the image data, and the voice tone is analyzed by the audio analysis tool to estimate the emotional state. Emotional data is generated as a result of the analysis, and this becomes the output for the next step.

[0158] Step 2:

[0159] The device sends the emotion data obtained in Step 1 to the server. It takes emotion data as input and sends it to the server using a digital communication protocol. Encrypted data transmission is performed to ensure data reliability and security. The transmitted emotion data is received by the server and used in the next step.

[0160] Step 3:

[0161] The server analyzes received sentiment data and dynamically adjusts task priorities. Using sentiment data submitted by users as input, it employs a reinforcement learning algorithm for task management. Specifically, it uses sentiment data to evaluate the user's stress level and satisfaction level, and revises the allocation of work tasks. The output is a list of adjusted task priorities, allowing for task reallocation between agents.

[0162] Step 4:

[0163] The server adjusts the cooperation between agents, reflecting task priorities. The priority list generated in step 3 is used as input. The server monitors the agents' operational status and reassigns tasks as needed. Specifically, it reduces the workload of heavily loaded agents and moves tasks to agents with greater processing power. The output provides the adjusted task assignments for each agent.

[0164] Step 5:

[0165] The terminal provides feedback to the user. As input, it receives adjusted task information sent from the server. Specifically, it implements a function to inform the user of the task's current status and changes through a visual dashboard. As output, information is presented in a user-friendly format, enabling effective interaction between the system and the user.

[0166] (Application Example 2)

[0167] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0168] In modern work environments, the emotions and stress levels of human workers significantly impact productivity and work efficiency. However, traditional agent systems and automated work environments have struggled to manage tasks while considering user emotions and workload. Therefore, there is a need to improve overall work efficiency while preventing worker overload.

[0169] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0170] In this invention, the server includes means for monitoring the operational status and progress of agents, means for dynamically reassigning tasks, means for analyzing the user's emotional state, and means for adjusting the workload and assigning auxiliary tasks based on the emotional state. This enables improved user experience within the work environment and increased overall work efficiency.

[0171] An "agent" is an artificial intelligence or automated software unit designed to perform a specific task.

[0172] "Operating status" refers to the situation or state in which an agent is actually performing a task.

[0173] "Progress" is an indicator that shows how much of the assigned task an agent has completed.

[0174] "Means for dynamically reallocating tasks" refers to a function that flexibly adjusts the distribution of tasks among agents according to the situation.

[0175] A "reinforcement learning algorithm" is a learning method that helps an agent select the best course of action based on its experience.

[0176] "User feedback" refers to users' opinions and evaluations of how the system works.

[0177] "Means for analyzing emotional states" refer to methods and technologies for identifying a user's emotions and evaluating their state.

[0178] "Means of adjusting workload and assigning auxiliary tasks" refers to technologies that optimize task allocation by taking into account the user's emotional state and provide additional or modified tasks to specific agents.

[0179] The system for implementing this invention employs a combination of technologies to monitor the operational status and progress of agents in real time and to analyze the user's emotions. The server manages the activities of multiple agents and dynamically reassigns tasks as needed. In doing so, it analyzes the emotional state from the user's facial expressions and voice using a terminal worn by the user, such as smart glasses or a wearable device. The software used is OpenCV, a computer vision technology, and TensorFlow, which is suitable for emotion recognition based on it.

[0180] The server receives this data and adjusts the workload using a reinforcement learning algorithm based on sentiment information. It also has an anomaly detection function, and if any malfunction or anomaly occurs in the agent's operation, it can immediately notify using the alert function.

[0181] As a concrete example, in a factory production line, workers' stress levels can be determined in real time by analyzing their facial expressions using smart glasses. Based on this, a server automatically redistributes tasks from specific agents to others, maintaining a balance in the workload. As a result, the burden on workers is reduced, and efficiency can be maintained at a higher level.

[0182] An example of a prompt for a generative AI model is, "What is a way to dynamically adjust the workload by performing facial expression analysis in the work environment based on data collected from a smart glasses device?" Using this prompt allows the system to adapt more to human emotions, supporting smooth execution in real-world situations.

[0183] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0184] Step 1:

[0185] The device captures the user's facial expressions and movements using sensors. Input includes image and audio data acquired from the device's camera and microphone. The device preprocesses this data using OpenCV to extract features for emotion recognition. The output is feature data.

[0186] Step 2:

[0187] The device inputs feature data into a TensorFlow model to estimate the user's emotional state. TensorFlow uses the pre-trained model to analyze the input data and outputs emotion labels. The output includes emotion labels such as "joy" and "anger" along with their confidence scores.

[0188] Step 3:

[0189] The server receives sentiment labels sent from the terminal and matches them against the agent's current task load data. The inputs are the user's sentiment information and the agent's task data. Based on this data, the server runs a reinforcement learning algorithm. The output is a task reassignment instruction.

[0190] Step 4:

[0191] The server sends task reassignment instructions to the agents, dynamically changing the agents' task configurations. The server then records the newly configured tasks and coordinates cooperation between agents. The input is the reassignment instructions, and the output is the updated task schedule.

[0192] Step 5:

[0193] Users provide feedback on improvements made by the system. This feedback is collected from the terminal and sent to the server. The server stores this feedback for future reinforcement learning and adjusts the overall system settings. The input is user feedback information, and the output is updated training data information.

[0194] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0195] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0196] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0197] [Second Embodiment]

[0198] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0199] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0200] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0201] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0202] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0203] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0204] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0205] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0206] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0207] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0208] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0209] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0210] This invention provides a method for efficiently managing a multi-agent system utilizing artificial intelligence and optimizing task execution. Specifically, it includes a mechanism in which a server monitors the operational status and progress of each agent in real time and dynamically reassigns tasks as needed. It also has a function to continuously improve agent performance through reinforcement learning and adjust system settings based on user feedback.

[0211] The server periodically collects and analyzes information from each agent. This information includes the operational status and task progress of each agent, and immediately generates an alert if an anomaly is detected. The server uses this information to redistribute tasks among agents. For example, if an agent is overloaded, the server will assign some of its tasks to agents with less load.

[0212] The server uses reinforcement learning algorithms to analyze past data and optimize task distribution among agents. This process refines the agents' cooperation over time, improving overall performance.

[0213] Users can provide feedback through their devices. This feedback is analyzed by the server and reflected in the system settings. For example, if a user instructs the server to prioritize tasks in a certain category, the server will immediately implement task allocation that takes this into account. This entire process is automated, minimizing user intervention and enabling efficient operation.

[0214] This invention enables the optimal coordination of multiple AI agents used within an enterprise, allowing various tasks to be performed efficiently and quickly. Immediate response and optimization processes based on anomaly detection reduce wasted resources. This provides a mechanism for improving overall productivity.

[0215] The following describes the processing flow.

[0216] Step 1:

[0217] The server collects operational status and progress data from each agent. This includes which tasks each agent is currently running, the amount of resources being used, and the degree of task completion. The server stores this information in a central database.

[0218] Step 2:

[0219] The server analyzes the collected data to check for any anomalies. If an anomaly is detected, the server generates an alert and notifies the administrator. This allows problems to be detected early and corrective actions to be taken.

[0220] Step 3:

[0221] The server determines whether tasks need to be dynamically reallocated based on the agent load. If the load is uneven, the server shifts some tasks to other agents with less load, thereby maintaining overall balance.

[0222] Step 4:

[0223] The server activates a reinforcement learning algorithm and learns from past performance data. This algorithm generates a new optimization strategy, which will be used for future task assignments.

[0224] Step 5:

[0225] Users send feedback to the server using their devices. This feedback may include changes to the priority of specific tasks or new requests. The server receives this feedback and adjusts the agent settings. This ensures that user requests are efficiently reflected in system operations.

[0226] (Example 1)

[0227] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0228] In recent years, many companies and organizations have implemented systems in which multiple agents simultaneously perform various tasks. However, there is a need for efficient operation of these agents, dynamic task reassignment, early detection of anomalies, and flexible system configuration that takes user feedback into consideration. Conventional systems have relied on manual work and fixed settings, and have suffered from a lack of real-time adaptability.

[0229] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0230] In this invention, the server includes means for periodically collecting the operating status and work progress of agents, means for analyzing the collected data and detecting operational anomalies, and means for dynamically redistributing tasks based on the analysis results. This enables efficient operation of agents and flexible and adaptive system operation.

[0231] An "agent" is a unit of software or hardware designed to perform a specific task.

[0232] "Operational status" refers to information indicating the progress of the tasks being performed by the agent and the resource usage.

[0233] "Work progress" is an indicator that shows the extent to which the tasks assigned to the agent have been completed.

[0234] "Collecting" means periodically acquiring and accumulating specific information as numerical data or logs.

[0235] "Analyzing" means evaluating collected data using computational processing and algorithms to extract useful information and patterns.

[0236] "Detecting anomalies" means discovering conditions or performance problems that deviate from normal operation.

[0237] "Dynamic reallocation" means flexibly rearranging existing tasks and resources to suit new states and conditions.

[0238] "Reinforcement learning technology" is a method for improving performance by enabling a system to learn optimal actions based on its experience.

[0239] "User input information" refers to data related to instructions and feedback that users provide to the system.

[0240] "Optimizing settings" means adjusting parameters and configurations to achieve efficient system operation.

[0241] This invention provides a solution for effectively managing a multi-agent system utilizing artificial intelligence and appropriately allocating tasks. The server periodically collects information on the operational status and work progress from each agent. This allows the server to understand the overall system status in real time. The collected information is acquired as data packets via the network and stored in a database.

[0242] The server applies reinforcement learning algorithms to analyze the collected data and optimize task allocation among agents. During this process, the server evaluates past task performance results and learns practical task allocation strategies from successful and unsuccessful examples. This analysis can be performed using data analysis programming languages ​​such as Python or R.

[0243] Furthermore, the server can immediately generate alerts if it detects any abnormalities in agent operation and reassign tasks as needed. For example, if an agent is overloaded, it can transfer tasks from that agent to other agents to maintain overall performance. This reassignment is automated, improving efficiency across the entire enterprise.

[0244] Users can provide feedback to the system through their terminals. Specifically, users input task priorities and new instructions using a GUI, and this data is sent to the server. The server analyzes this feedback in real time and can adjust system settings as needed. This optimization of settings allows the system to flexibly respond to user needs.

[0245] As a concrete example, consider a large-scale data analysis project where a sudden increase in computational demands causes some agents to reach their limits. The server detects this situation and immediately redistributes tasks to other agents to resolve the problem. Furthermore, if a user provides feedback requesting priority processing of a specific data category, the server adjusts task priorities accordingly. Such applications are conceivable.

[0246] Examples of prompts for a generative AI model include the following:

[0247] "You are the server managing AI agents within the company. Check the operational status of Agent A, and if it is overloaded, redistribute tasks to Agent B. When doing so, use a reinforcement learning algorithm to determine the optimal distribution and adjust the settings based on user feedback."

[0248] The system realized by this invention enables the effective management of multiple agents within a company or organization, allowing them to perform tasks efficiently and flexibly with minimal intervention.

[0249] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0250] Step 1:

[0251] The server periodically collects information on the operational status and work progress from each agent. In this step, it receives data packets from agents via the network. The specific input is log data of the agents' operational status and progress, and by saving this to a database, it obtains output that is ready for the next analysis process.

[0252] Step 2:

[0253] The server analyzes the collected data to detect operational anomalies. The operational data saved in Step 1 is used as input, and a programming language (e.g., Python) or statistical software is used for analysis. The server analyzes the agent's CPU usage and memory usage, and if these exceed the threshold, it determines it to be an anomaly and generates an alert. This alert serves as a basis for deciding on reallocation.

[0254] Step 3:

[0255] The server dynamically redistributes tasks based on the analysis results. The input consists of anomaly information detected in step 2 and the agent load status. The server uses a reinforcement learning algorithm to evaluate the performance of each agent and transfers tasks to agents with less load. This process outputs an efficient task redistribution.

[0256] Step 4:

[0257] Users provide feedback through their terminals, which is then analyzed by the server. The input consists of user instructions regarding configuration changes and task priorities. The server receives and analyzes this feedback information to adjust system settings as needed. This feedback-based optimization is the final output, improving the system's operational efficiency.

[0258] In this way, the system collects, analyzes, assigns tasks to, and incorporates feedback in real time, continuously optimizing the entire system.

[0259] (Application Example 1)

[0260] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0261] Efficiently preventing load imbalances and malfunctions in data centers is challenging. Conventional systems struggle to accurately understand the operating status of equipment, making it difficult to respond quickly to equipment experiencing heavy loads. Furthermore, dynamic adjustments based on individual feedback are difficult, hindering overall system optimization.

[0262] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0263] In this invention, the server includes means for monitoring the operating status and progress of the equipment being used, means for dynamically reassigning tasks, and means for optimizing the performance of the equipment using a reinforcement learning algorithm. This enables efficient equalization of the load within the data center, prevention of equipment malfunctions, and optimal system adjustments incorporating user feedback.

[0264] "Equipment used" refers to the collective term for electronic devices such as servers and computers that are operated within a data center.

[0265] "Operating status" refers to information about the current operating status and resource consumption of the equipment being used.

[0266] "Progress" is an indicator that shows the extent to which a task or work has been completed.

[0267] "Work" refers to the procedures for processing tasks and processes within a data center.

[0268] "Load" refers to the pressure exerted by the amount of data and computational load on the equipment being used.

[0269] "Reassignment" refers to the process of redistributing tasks or resources to different devices.

[0270] A "reinforcement learning algorithm" is a type of machine learning that learns the optimal action through trial and error.

[0271] "Performance" refers to the ability and efficiency with which the equipment used can perform a given task.

[0272] "Optimization" is the process of adjusting a system to maximize its overall efficiency and effectiveness.

[0273] "An anomaly" refers to an unexpected state that deviates from normal operation.

[0274] "Feedback" refers to information provided based on users' experiences and requests.

[0275] "Equalization" refers to a state where resources and workloads are distributed evenly to eliminate imbalances.

[0276] The system that realizes this invention is primarily server-based. The server is programmed using Python and TensorFlow and collects and monitors the operating status and progress of equipment used within the data center in real time. The server analyzes this information and evaluates the workload on each piece of equipment. If the workload is concentrated on a particular piece of equipment, it dynamically reassigns tasks to equalize the overall load.

[0277] The reinforcement learning algorithm is implemented using TensorFlow, and the server optimizes the performance of the equipment. This enables optimal task allocation based on historical data. Furthermore, the server can dynamically adjust system settings based on user feedback, new requests, and improvements.

[0278] As a specific example, in a data center, assume that a certain server group is in the process of handling an important analysis task and some of the servers have become overloaded. In this case, the system on the server immediately monitors the situation and reassigns tasks from the overloaded servers to the lightly loaded servers. This ensures the overall processing efficiency and prevents abnormalities from occurring.

[0279] An example of a prompt sentence for the generative AI model can be described as follows. "Please propose an optimal method and parameter settings for constructing a reinforcement learning model that monitors the server load in a data center and optimizes tasks."

[0280] The flow of the specific process in Application Example 1 will be described using FIG. 12.

[0281] Step 1:

[0282] The server collects the operating status and progress data from the devices in use in real time. The inputs are the CPU usage rate, memory usage rate, I / O waiting time, etc. provided by each device in use, and these are imported into the internal database. The output is the latest information on the operating status of each device. Specifically, data is periodically obtained from each device using the internal communication protocol.

[0283] Step 2:

[0284] The server analyzes the collected data and evaluates the load status of each device. The input is the data on the operating status from Step 1, and the output is a list indicating the load status of each device. In this process, statistical methods are used to calculate the load and identify devices that may be overloaded. As a specific operation, a threshold is set using an anomaly detection algorithm, and devices with excessive load are listed.

[0285] Step 3:

[0286] The server reallocates tasks from devices with concentrated loads to other devices with available capacity. The input is the load status list obtained in step 2. The output is the task list for each device after reallocation. In this process, a dynamic programming approach is used to equalize the loads. Specifically, tasks are planned and executed to move from devices with high loads to those with low loads.

[0287] Step 4:

[0288] The server applies a reinforcement learning algorithm to continuously improve the performance of each device. The input is the progress report of the tasks and user feedback, and the output is an optimized task allocation model. The server uses TensorFlow to learn from past data and generate the optimal operation pattern of the agent. Specifically, the performance data is analyzed and the optimal action plan is updated.

[0289] Step 5:

[0290] The end-user checks the results of the system settings optimized by the server and provides feedback if necessary. The input is the optimized task allocation and the feedback request from the server, and the output is the feedback information from the user. Specifically, setting confirmation and feedback input are performed via the GUI interface.

[0291] Furthermore, an emotion engine for estimating the user's emotion may be combined. That is, the specific processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform specific processing using the user's emotion.

[0292] This invention provides a method to improve system efficiency and user satisfaction by combining an emotion engine that recognizes user emotions with a multi-agent system that utilizes artificial intelligence. A server acts as the central point, monitoring the operational status and progress of agents in real time and dynamically reassigning tasks as needed. Agent performance is optimized using reinforcement learning algorithms.

[0293] An emotion engine is used to analyze the user's emotional state on the device. The collected emotional information is sent to the server and reflected in task priorities and agent settings. Based on this emotional information, the server adjusts the cooperation between agents to achieve more adaptive task execution. In addition, system settings can be adjusted in conjunction with the emotion engine based on user feedback.

[0294] For example, if the emotion engine detects that a user is experiencing stress, the server reassigns high-load tasks to another agent, reducing the user's workload. This improves the user experience.

[0295] Furthermore, if a user expresses satisfaction or dissatisfaction, the server will adjust task allocation among agents accordingly. This setting can be flexibly changed according to user requests and the work environment. If an anomaly is detected, the server will immediately issue an alert and pinpoint the source of the problem.

[0296] This system enables AI agents used within a company to work efficiently and harmoniously, adapting to various situations while performing tasks. Through integration with an emotion engine, the system achieves more human-like interactions, improving overall productivity and user satisfaction.

[0297] The following describes the processing flow.

[0298] Step 1:

[0299] The server collects the operation status and progress data from each agent. This includes the status of the current task, the resources in use, and the degree of completion of the work. The server analyzes this data and monitors the operation status of the agents.

[0300] Step 2:

[0301] The terminal monitors the user's actions and inputs, and uses an emotion engine to analyze the user's emotional state in real time. The terminal sends this analysis result to the server.

[0302] Step 3:

[0303] Based on the obtained user emotion information, the server adjusts the priority and allocation of the agents' tasks. For example, when the user is feeling stressed, the server performs reallocation to reduce high-load tasks.

[0304] Step 4:

[0305] The server executes a reinforcement learning algorithm, utilizes past performance data and user emotion information to optimize the overall efficiency of the system. As a result, the agents will operate more adaptively in the next task allocation.

[0306] Step 5:

[0307] The user provides feedback using the terminal. The feedback may include the priority of the task and the satisfaction with the processing. The server aggregates this feedback and reflects it in the system settings and the cooperation mechanism of the agents.

[0308] Step 6:

[0309] The server uses an anomaly detection module. When problems occur in the operation of the agents or the system as a whole, it immediately generates an alert and takes appropriate actions. This is important for increasing the speed of problem solving and maintaining the stability of the system.

[0310] (Example 2)

[0311] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0312] In recent years, AI-powered business support systems have become widespread. However, these systems fail to adequately address users' emotional states, limiting their potential for improving work efficiency and user satisfaction. Furthermore, the optimization of cooperation between programs within the system is insufficient, making it difficult to maximize the overall effectiveness of business operations. In addition, they lack the functionality to respond immediately when an anomaly occurs, highlighting the need for rapid problem resolution.

[0313] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0314] In this invention, the server includes means for analyzing the user's emotional state, means for dynamically adjusting task priorities based on the analyzed emotional information, means for monitoring a group of autonomous programs with multiple roles and reassigning roles as needed, means for improving the operational efficiency of the program group using machine learning algorithms, and means for modifying the system configuration based on user input. This enables flexible task management in response to the user's emotions, thereby improving operational efficiency and maximizing user satisfaction.

[0315] "Methods for analyzing a user's emotional state" refer to technologies that detect emotions in real time from the user's facial expressions, tone of voice, etc., and analyze emotional trends based on that data.

[0316] "A means of dynamically adjusting task priorities based on analyzed emotional information" refers to a method that utilizes the results of analyzing user emotional data to constantly optimize the order in which tasks are performed in a way that does not burden the user.

[0317] A "group of autonomous programs with multiple roles" is a collection of software that performs different functions independently, but is designed to work together as a whole to achieve a specific objective.

[0318] "Methods for improving the operational efficiency of a group of programs using machine learning algorithms" refers to techniques that use artificial intelligence algorithms to analyze the operational data within a program and improve its efficiency based on that analysis.

[0319] "Means of modifying system configuration based on user input" refers to the process of adjusting and improving system settings and operation based on user feedback and operation history.

[0320] This invention is a multi-agent system that recognizes the user's emotional state in real time and optimizes the system's operation based on that recognition. The following describes a specific implementation of this system.

[0321] The main components consist of a server, a terminal, and a user. The server functions as the core of the system, and the terminal collects user emotion data via an emotion engine. The terminal is equipped with image recognition software and voice analysis tools to analyze the user's voice and facial expressions, specifically using libraries such as OpenCV. These tools receive data through the user interface and send it to the server.

[0322] The server uses machine learning algorithms to analyze the received sentiment data. At this stage, it monitors the agent's activity and dynamically optimizes task priorities based on the analysis results. By applying reinforcement learning algorithms, the cooperative relationships between agents are automatically adjusted. For example, if the analysis indicates that the user is experiencing stress, the server reduces the user's workload by assigning high-load tasks to other agents.

[0323] Furthermore, the terminal provides the user with feedback on instructions from the server and changes in status, and adjusts the overall system settings as needed. For this purpose, a dashboard application is used to provide visual information. The server collects user feedback and uses it to adjust system settings for further optimization.

[0324] A concrete example of how this system works is when a user is stressed due to project deadlines. In this situation, the emotion engine detects the stress level, and the server quickly reassigns tasks. This allows the project to progress smoothly, and the user can achieve better results.

[0325] An example of a prompt might be, "How should the system reassign tasks when the user is experiencing stress?" Using such prompts, it's possible to leverage generative AI models to further improve and adapt the system.

[0326] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0327] Step 1:

[0328] The device collects the user's emotions using an emotion engine. Here, the user's facial expressions are captured by a camera, and their voice is recorded by a microphone. Image data and audio data are acquired as input, and image recognition software and audio analysis tools are used to analyze them. Specifically, OpenCV is used to analyze facial expressions from the image data, and the voice tone is analyzed by the audio analysis tool to estimate the emotional state. Emotional data is generated as a result of the analysis, and this becomes the output for the next step.

[0329] Step 2:

[0330] The device sends the emotion data obtained in Step 1 to the server. It takes emotion data as input and sends it to the server using a digital communication protocol. Encrypted data transmission is performed to ensure data reliability and security. The transmitted emotion data is received by the server and used in the next step.

[0331] Step 3:

[0332] The server analyzes received sentiment data and dynamically adjusts task priorities. Using sentiment data submitted by users as input, it employs a reinforcement learning algorithm for task management. Specifically, it uses sentiment data to evaluate the user's stress level and satisfaction level, and revises the allocation of work tasks. The output is a list of adjusted task priorities, allowing for task reallocation between agents.

[0333] Step 4:

[0334] The server adjusts the cooperation between agents, reflecting task priorities. The priority list generated in step 3 is used as input. The server monitors the agents' operational status and reassigns tasks as needed. Specifically, it reduces the workload of heavily loaded agents and moves tasks to agents with greater processing power. The output provides the adjusted task assignments for each agent.

[0335] Step 5:

[0336] The terminal provides feedback to the user. As input, it receives adjusted task information sent from the server. Specifically, it implements a function to inform the user of the task's current status and changes through a visual dashboard. As output, information is presented in a user-friendly format, enabling effective interaction between the system and the user.

[0337] (Application Example 2)

[0338] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0339] In modern work environments, the emotions and stress levels of human workers significantly impact productivity and work efficiency. However, traditional agent systems and automated work environments have struggled to manage tasks while considering user emotions and workload. Therefore, there is a need to improve overall work efficiency while preventing worker overload.

[0340] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0341] In this invention, the server includes means for monitoring the operational status and progress of agents, means for dynamically reassigning tasks, means for analyzing the user's emotional state, and means for adjusting the workload and assigning auxiliary tasks based on the emotional state. This enables improved user experience within the work environment and increased overall work efficiency.

[0342] An "agent" is an artificial intelligence or automated software unit designed to perform a specific task.

[0343] "Operating status" refers to the situation or state in which an agent is actually performing a task.

[0344] "Progress" is an indicator that shows how much of the assigned task an agent has completed.

[0345] "Means for dynamically reallocating tasks" refers to a function that flexibly adjusts the distribution of tasks among agents according to the situation.

[0346] A "reinforcement learning algorithm" is a learning method that helps an agent select the best course of action based on its experience.

[0347] "User feedback" refers to users' opinions and evaluations of how the system works.

[0348] "Means for analyzing emotional states" refer to methods and technologies for identifying a user's emotions and evaluating their state.

[0349] "Means of adjusting workload and assigning auxiliary tasks" refers to technologies that optimize task allocation by taking into account the user's emotional state and provide additional or modified tasks to specific agents.

[0350] The system for implementing this invention employs a combination of technologies to monitor the operational status and progress of agents in real time and to analyze the user's emotions. The server manages the activities of multiple agents and dynamically reassigns tasks as needed. In doing so, it analyzes the emotional state from the user's facial expressions and voice using a terminal worn by the user, such as smart glasses or a wearable device. The software used is OpenCV, a computer vision technology, and TensorFlow, which is suitable for emotion recognition based on it.

[0351] The server receives this data and adjusts the workload using a reinforcement learning algorithm based on sentiment information. It also has an anomaly detection function, and if any malfunction or anomaly occurs in the agent's operation, it can immediately notify using the alert function.

[0352] As a concrete example, in a factory production line, workers' stress levels can be determined in real time by analyzing their facial expressions using smart glasses. Based on this, a server automatically redistributes tasks from specific agents to others, maintaining a balance in the workload. As a result, the burden on workers is reduced, and efficiency can be maintained at a higher level.

[0353] An example of a prompt for a generative AI model is, "What is a way to dynamically adjust the workload by performing facial expression analysis in the work environment based on data collected from a smart glasses device?" Using this prompt allows the system to adapt more to human emotions, supporting smooth execution in real-world situations.

[0354] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0355] Step 1:

[0356] The device captures the user's facial expressions and movements using sensors. Input includes image and audio data acquired from the device's camera and microphone. The device preprocesses this data using OpenCV to extract features for emotion recognition. The output is feature data.

[0357] Step 2:

[0358] The device inputs feature data into a TensorFlow model to estimate the user's emotional state. TensorFlow uses the pre-trained model to analyze the input data and outputs emotion labels. The output includes emotion labels such as "joy" and "anger" along with their confidence scores.

[0359] Step 3:

[0360] The server receives sentiment labels sent from the terminal and matches them against the agent's current task load data. The inputs are the user's sentiment information and the agent's task data. Based on this data, the server runs a reinforcement learning algorithm. The output is a task reassignment instruction.

[0361] Step 4:

[0362] The server sends task reassignment instructions to the agents, dynamically changing the agents' task configurations. The server then records the newly configured tasks and coordinates cooperation between agents. The input is the reassignment instructions, and the output is the updated task schedule.

[0363] Step 5:

[0364] Users provide feedback on improvements made by the system. This feedback is collected from the terminal and sent to the server. The server stores this feedback for future reinforcement learning and adjusts the overall system settings. The input is user feedback information, and the output is updated training data information.

[0365] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0366] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0367] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0368] [Third Embodiment]

[0369] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0370] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0371] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0372] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0373] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0374] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0375] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0376] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0377] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0378] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0379] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0380] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0381] This invention provides a method for efficiently managing a multi-agent system utilizing artificial intelligence and optimizing task execution. Specifically, it includes a mechanism in which a server monitors the operational status and progress of each agent in real time and dynamically reassigns tasks as needed. It also has a function to continuously improve agent performance through reinforcement learning and adjust system settings based on user feedback.

[0382] The server periodically collects and analyzes information from each agent. This information includes the operational status and task progress of each agent, and immediately generates an alert if an anomaly is detected. The server uses this information to redistribute tasks among agents. For example, if an agent is overloaded, the server will assign some of its tasks to agents with less load.

[0383] The server uses reinforcement learning algorithms to analyze past data and optimize task distribution among agents. This process refines the agents' cooperation over time, improving overall performance.

[0384] Users can provide feedback through their devices. This feedback is analyzed by the server and reflected in the system settings. For example, if a user instructs the server to prioritize tasks in a certain category, the server will immediately implement task allocation that takes this into account. This entire process is automated, minimizing user intervention and enabling efficient operation.

[0385] This invention enables the optimal coordination of multiple AI agents used within an enterprise, allowing various tasks to be performed efficiently and quickly. Immediate response and optimization processes based on anomaly detection reduce wasted resources. This provides a mechanism for improving overall productivity.

[0386] The following describes the processing flow.

[0387] Step 1:

[0388] The server collects operational status and progress data from each agent. This includes which tasks each agent is currently running, the amount of resources being used, and the degree of task completion. The server stores this information in a central database.

[0389] Step 2:

[0390] The server analyzes the collected data to check for any anomalies. If an anomaly is detected, the server generates an alert and notifies the administrator. This allows problems to be detected early and corrective actions to be taken.

[0391] Step 3:

[0392] The server determines whether tasks need to be dynamically reallocated based on the agent load. If the load is uneven, the server shifts some tasks to other agents with less load, thereby maintaining overall balance.

[0393] Step 4:

[0394] The server activates a reinforcement learning algorithm and learns from past performance data. This algorithm generates a new optimization strategy, which will be used for future task assignments.

[0395] Step 5:

[0396] Users send feedback to the server using their devices. This feedback may include changes to the priority of specific tasks or new requests. The server receives this feedback and adjusts the agent settings. This ensures that user requests are efficiently reflected in system operations.

[0397] (Example 1)

[0398] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0399] In recent years, many companies and organizations have implemented systems in which multiple agents simultaneously perform various tasks. However, there is a need for efficient operation of these agents, dynamic task reassignment, early detection of anomalies, and flexible system configuration that takes user feedback into consideration. Conventional systems have relied on manual work and fixed settings, and have suffered from a lack of real-time adaptability.

[0400] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0401] In this invention, the server includes means for periodically collecting the operating status and work progress of agents, means for analyzing the collected data and detecting operational anomalies, and means for dynamically redistributing tasks based on the analysis results. This enables efficient operation of agents and flexible and adaptive system operation.

[0402] An "agent" is a unit of software or hardware designed to perform a specific task.

[0403] "Operational status" refers to information indicating the progress of the tasks being performed by the agent and the resource usage.

[0404] "Work progress" is an indicator that shows the extent to which the tasks assigned to the agent have been completed.

[0405] "Collecting" means periodically acquiring and accumulating specific information as numerical data or logs.

[0406] "Analyzing" means evaluating collected data using computational processing and algorithms to extract useful information and patterns.

[0407] "Detecting anomalies" means discovering conditions or performance problems that deviate from normal operation.

[0408] "Dynamic reallocation" means flexibly rearranging existing tasks and resources to suit new states and conditions.

[0409] "Reinforcement learning technology" is a method for improving performance by enabling a system to learn optimal actions based on its experience.

[0410] "User input information" refers to data related to instructions and feedback that users provide to the system.

[0411] "Optimizing settings" means adjusting parameters and configurations to achieve efficient system operation.

[0412] This invention provides a solution for effectively managing a multi-agent system utilizing artificial intelligence and appropriately allocating tasks. The server periodically collects information on the operational status and work progress from each agent. This allows the server to understand the overall system status in real time. The collected information is acquired as data packets via the network and stored in a database.

[0413] The server applies reinforcement learning algorithms to analyze the collected data and optimize task allocation among agents. During this process, the server evaluates past task performance results and learns practical task allocation strategies from successful and unsuccessful examples. This analysis can be performed using data analysis programming languages ​​such as Python or R.

[0414] Furthermore, the server can immediately generate alerts if it detects any abnormalities in agent operation and reassign tasks as needed. For example, if an agent is overloaded, it can transfer tasks from that agent to other agents to maintain overall performance. This reassignment is automated, improving efficiency across the entire enterprise.

[0415] Users can provide feedback to the system through their terminals. Specifically, users input task priorities and new instructions using a GUI, and this data is sent to the server. The server analyzes this feedback in real time and can adjust system settings as needed. This optimization of settings allows the system to flexibly respond to user needs.

[0416] As a concrete example, consider a large-scale data analysis project where a sudden increase in computational demands causes some agents to reach their limits. The server detects this situation and immediately redistributes tasks to other agents to resolve the problem. Furthermore, if a user provides feedback requesting priority processing of a specific data category, the server adjusts task priorities accordingly. Such applications are conceivable.

[0417] Examples of prompts for a generative AI model include the following:

[0418] "You are the server managing AI agents within the company. Check the operational status of Agent A, and if it is overloaded, redistribute tasks to Agent B. When doing so, use a reinforcement learning algorithm to determine the optimal distribution and adjust the settings based on user feedback."

[0419] The system realized by this invention enables the effective management of multiple agents within a company or organization, allowing them to perform tasks efficiently and flexibly with minimal intervention.

[0420] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0421] Step 1:

[0422] The server periodically collects information on the operational status and work progress from each agent. In this step, it receives data packets from agents via the network. The specific input is log data of the agents' operational status and progress, and by saving this to a database, it obtains output that is ready for the next analysis process.

[0423] Step 2:

[0424] The server analyzes the collected data to detect operational anomalies. The operational data saved in Step 1 is used as input, and a programming language (e.g., Python) or statistical software is used for analysis. The server analyzes the agent's CPU usage and memory usage, and if these exceed the threshold, it determines it to be an anomaly and generates an alert. This alert serves as a basis for deciding on reallocation.

[0425] Step 3:

[0426] The server dynamically redistributes tasks based on the analysis results. The input consists of anomaly information detected in step 2 and the agent load status. The server uses a reinforcement learning algorithm to evaluate the performance of each agent and transfers tasks to agents with less load. This process outputs an efficient task redistribution.

[0427] Step 4:

[0428] Users provide feedback through their terminals, which is then analyzed by the server. The input consists of user instructions regarding configuration changes and task priorities. The server receives and analyzes this feedback information to adjust system settings as needed. This feedback-based optimization is the final output, improving the system's operational efficiency.

[0429] In this way, the system collects, analyzes, assigns tasks to, and incorporates feedback in real time, continuously optimizing the entire system.

[0430] (Application Example 1)

[0431] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0432] Efficiently preventing load imbalances and malfunctions in data centers is challenging. Conventional systems struggle to accurately understand the operating status of equipment, making it difficult to respond quickly to equipment experiencing heavy loads. Furthermore, dynamic adjustments based on individual feedback are difficult, hindering overall system optimization.

[0433] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0434] In this invention, the server includes means for monitoring the operating status and progress of the equipment being used, means for dynamically reassigning tasks, and means for optimizing the performance of the equipment using a reinforcement learning algorithm. This enables efficient equalization of the load within the data center, prevention of equipment malfunctions, and optimal system adjustments incorporating user feedback.

[0435] "Equipment used" refers to the collective term for electronic devices such as servers and computers that are operated within a data center.

[0436] "Operating status" refers to information about the current operating status and resource consumption of the equipment being used.

[0437] "Progress" is an indicator that shows the extent to which a task or work has been completed.

[0438] "Work" refers to the procedures for processing tasks and processes within a data center.

[0439] "Load" refers to the pressure exerted by the amount of data and computational load on the equipment being used.

[0440] "Reassignment" refers to the process of redistributing tasks or resources to different devices.

[0441] A "reinforcement learning algorithm" is a type of machine learning that learns the optimal action through trial and error.

[0442] "Performance" refers to the ability and efficiency with which the equipment used can perform a given task.

[0443] "Optimization" is the process of adjusting a system to maximize its overall efficiency and effectiveness.

[0444] "An anomaly" refers to an unexpected state that deviates from normal operation.

[0445] "Feedback" refers to information provided based on users' experiences and requests.

[0446] "Equalization" refers to a state where resources and workloads are distributed evenly to eliminate imbalances.

[0447] The system that realizes this invention is primarily server-based. The server is programmed using Python and TensorFlow and collects and monitors the operating status and progress of equipment used within the data center in real time. The server analyzes this information and evaluates the workload on each piece of equipment. If the workload is concentrated on a particular piece of equipment, it dynamically reassigns tasks to equalize the overall load.

[0448] The reinforcement learning algorithm is implemented using TensorFlow, and the server optimizes the performance of the equipment. This enables optimal task allocation based on historical data. Furthermore, the server can dynamically adjust system settings based on user feedback, new requests, and improvements.

[0449] As a concrete example, suppose a data center has a group of servers processing a critical analytical task, and some of the servers become overloaded. In this case, the system on the servers immediately monitors the situation and reassigns tasks from the overloaded servers to the less loaded servers. This ensures overall processing efficiency and prevents problems from occurring.

[0450] An example of a prompt for a generative AI model is as follows: "Please suggest the optimal method and parameter settings for building a reinforcement learning model that monitors server load in a data center and optimizes the task."

[0451] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0452] Step 1:

[0453] The server collects operational status and progress data from the devices being used in real time. Inputs include CPU usage, memory usage, and I / O wait times provided by each device, which are then imported into an internal database. Outputs are the latest operational status data for each device. Specifically, data is periodically retrieved from each device using an internal communication protocol.

[0454] Step 2:

[0455] The server analyzes the collected data and evaluates the load status of each device. The input is operational status data from step 1, and the output is a list showing the load status of each device. This process uses statistical methods to calculate the load and identify devices that may be overloaded. Specifically, it sets thresholds using an anomaly detection algorithm and lists devices whose load has been exceeded.

[0456] Step 3:

[0457] The server reallocates tasks from heavily loaded devices to other less-loaded devices. The input is the load status list obtained in step 2. The output is the task list for each device after reallocation. In this process, dynamic programming techniques are used to equalize the load. Specifically, the server plans and executes the movement of tasks from heavily loaded devices to less loaded ones.

[0458] Step 4:

[0459] The server continuously improves the performance of each device by applying reinforcement learning algorithms. Inputs are progress reports and user feedback, and output is a model of optimized task allocation. The server uses TensorFlow to learn from historical data and generate optimal agent behavior patterns. Specifically, it analyzes performance data and updates the optimal course of action.

[0460] Step 5:

[0461] Terminal users review the results of system configurations optimized by the server and provide feedback as needed. Inputs are optimized task allocations and feedback requests from the server, while output is user feedback information. Specifically, configuration verification and feedback input are performed via a GUI interface.

[0462] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0463] This invention provides a method to improve system efficiency and user satisfaction by combining an emotion engine that recognizes user emotions with a multi-agent system that utilizes artificial intelligence. A server acts as the central point, monitoring the operational status and progress of agents in real time and dynamically reassigning tasks as needed. Agent performance is optimized using reinforcement learning algorithms.

[0464] An emotion engine is used to analyze the user's emotional state on the device. The collected emotional information is sent to the server and reflected in task priorities and agent settings. Based on this emotional information, the server adjusts the cooperation between agents to achieve more adaptive task execution. In addition, system settings can be adjusted in conjunction with the emotion engine based on user feedback.

[0465] For example, if the emotion engine detects that a user is experiencing stress, the server reassigns high-load tasks to another agent, reducing the user's workload. This improves the user experience.

[0466] Furthermore, if a user expresses satisfaction or dissatisfaction, the server will adjust task allocation among agents accordingly. This setting can be flexibly changed according to user requests and the work environment. If an anomaly is detected, the server will immediately issue an alert and pinpoint the source of the problem.

[0467] This system enables AI agents used within a company to work efficiently and harmoniously, adapting to various situations while performing tasks. Through integration with an emotion engine, the system achieves more human-like interactions, improving overall productivity and user satisfaction.

[0468] The following describes the processing flow.

[0469] Step 1:

[0470] The server collects operational status and progress data from each agent. This includes the current task status, resources being used, and the degree of work completion. The server analyzes this data and monitors the agent's performance.

[0471] Step 2:

[0472] The device monitors the user's actions and inputs, and uses an emotion engine to analyze the user's emotional state in real time. The device then sends these analysis results to the server.

[0473] Step 3:

[0474] Based on the user's emotional information, the server adjusts the priority and allocation of agent tasks. For example, if a user is stressed, the server will reallocate tasks to reduce the load on them.

[0475] Step 4:

[0476] The server executes a reinforcement learning algorithm, leveraging historical performance data and user sentiment information to optimize the overall system efficiency. This allows the agent to act more adaptively in subsequent task assignments.

[0477] Step 5:

[0478] Users provide feedback using their devices. This feedback may include task priorities and satisfaction with the process. The server aggregates this feedback and uses it to improve system settings and agent collaboration.

[0479] Step 6:

[0480] The server uses an anomaly detection module to immediately generate alerts and take appropriate action if problems occur in the operation of agents or the system as a whole. This is important for speeding up problem resolution and maintaining system stability.

[0481] (Example 2)

[0482] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0483] In recent years, AI-powered business support systems have become widespread. However, these systems fail to adequately address users' emotional states, limiting their potential for improving work efficiency and user satisfaction. Furthermore, the optimization of cooperation between programs within the system is insufficient, making it difficult to maximize the overall effectiveness of business operations. In addition, they lack the functionality to respond immediately when an anomaly occurs, highlighting the need for rapid problem resolution.

[0484] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0485] In this invention, the server includes means for analyzing the user's emotional state, means for dynamically adjusting task priorities based on the analyzed emotional information, means for monitoring a group of autonomous programs with multiple roles and reassigning roles as needed, means for improving the operational efficiency of the program group using machine learning algorithms, and means for modifying the system configuration based on user input. This enables flexible task management in response to the user's emotions, thereby improving operational efficiency and maximizing user satisfaction.

[0486] "Methods for analyzing a user's emotional state" refer to technologies that detect emotions in real time from the user's facial expressions, tone of voice, etc., and analyze emotional trends based on that data.

[0487] "A means of dynamically adjusting task priorities based on analyzed emotional information" refers to a method that utilizes the results of analyzing user emotional data to constantly optimize the order in which tasks are performed in a way that does not burden the user.

[0488] A "group of autonomous programs with multiple roles" is a collection of software that performs different functions independently, but is designed to work together as a whole to achieve a specific objective.

[0489] "Methods for improving the operational efficiency of a group of programs using machine learning algorithms" refers to techniques that use artificial intelligence algorithms to analyze the operational data within a program and improve its efficiency based on that analysis.

[0490] "Means of modifying system configuration based on user input" refers to the process of adjusting and improving system settings and operation based on user feedback and operation history.

[0491] This invention is a multi-agent system that recognizes the user's emotional state in real time and optimizes the system's operation based on that recognition. The following describes a specific implementation of this system.

[0492] The main components consist of a server, a terminal, and a user. The server functions as the core of the system, and the terminal collects user emotion data via an emotion engine. The terminal is equipped with image recognition software and voice analysis tools to analyze the user's voice and facial expressions, specifically using libraries such as OpenCV. These tools receive data through the user interface and send it to the server.

[0493] The server uses machine learning algorithms to analyze the received sentiment data. At this stage, it monitors the agent's activity and dynamically optimizes task priorities based on the analysis results. By applying reinforcement learning algorithms, the cooperative relationships between agents are automatically adjusted. For example, if the analysis indicates that the user is experiencing stress, the server reduces the user's workload by assigning high-load tasks to other agents.

[0494] Furthermore, the terminal provides the user with feedback on instructions from the server and changes in status, and adjusts the overall system settings as needed. For this purpose, a dashboard application is used to provide visual information. The server collects user feedback and uses it to adjust system settings for further optimization.

[0495] A concrete example of how this system works is when a user is stressed due to project deadlines. In this situation, the emotion engine detects the stress level, and the server quickly reassigns tasks. This allows the project to progress smoothly, and the user can achieve better results.

[0496] An example of a prompt might be, "How should the system reassign tasks when the user is experiencing stress?" Using such prompts, it's possible to leverage generative AI models to further improve and adapt the system.

[0497] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0498] Step 1:

[0499] The device collects the user's emotions using an emotion engine. Here, the user's facial expressions are captured by a camera, and their voice is recorded by a microphone. Image data and audio data are acquired as input, and image recognition software and audio analysis tools are used to analyze them. Specifically, OpenCV is used to analyze facial expressions from the image data, and the voice tone is analyzed by the audio analysis tool to estimate the emotional state. Emotional data is generated as a result of the analysis, and this becomes the output for the next step.

[0500] Step 2:

[0501] The device sends the emotion data obtained in Step 1 to the server. It takes emotion data as input and sends it to the server using a digital communication protocol. Encrypted data transmission is performed to ensure data reliability and security. The transmitted emotion data is received by the server and used in the next step.

[0502] Step 3:

[0503] The server analyzes received sentiment data and dynamically adjusts task priorities. Using sentiment data submitted by users as input, it employs a reinforcement learning algorithm for task management. Specifically, it uses sentiment data to evaluate the user's stress level and satisfaction level, and revises the allocation of work tasks. The output is a list of adjusted task priorities, allowing for task reallocation between agents.

[0504] Step 4:

[0505] The server adjusts the cooperation between agents, reflecting task priorities. The priority list generated in step 3 is used as input. The server monitors the agents' operational status and reassigns tasks as needed. Specifically, it reduces the workload of heavily loaded agents and moves tasks to agents with greater processing power. The output provides the adjusted task assignments for each agent.

[0506] Step 5:

[0507] The terminal provides feedback to the user. As input, it receives adjusted task information sent from the server. Specifically, it implements a function to inform the user of the task's current status and changes through a visual dashboard. As output, information is presented in a user-friendly format, enabling effective interaction between the system and the user.

[0508] (Application Example 2)

[0509] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0510] In modern work environments, the emotions and stress levels of human workers significantly impact productivity and work efficiency. However, traditional agent systems and automated work environments have struggled to manage tasks while considering user emotions and workload. Therefore, there is a need to improve overall work efficiency while preventing worker overload.

[0511] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0512] In this invention, the server includes means for monitoring the operational status and progress of agents, means for dynamically reassigning tasks, means for analyzing the user's emotional state, and means for adjusting the workload and assigning auxiliary tasks based on the emotional state. This enables improved user experience within the work environment and increased overall work efficiency.

[0513] An "agent" is an artificial intelligence or automated software unit designed to perform a specific task.

[0514] "Operating status" refers to the situation or state in which an agent is actually performing a task.

[0515] "Progress" is an indicator that shows how much of the assigned task an agent has completed.

[0516] "Means for dynamically reallocating tasks" refers to a function that flexibly adjusts the distribution of tasks among agents according to the situation.

[0517] A "reinforcement learning algorithm" is a learning method that helps an agent select the best course of action based on its experience.

[0518] "User feedback" refers to users' opinions and evaluations of how the system works.

[0519] "Means for analyzing emotional states" refer to methods and technologies for identifying a user's emotions and evaluating their state.

[0520] "Means of adjusting workload and assigning auxiliary tasks" refers to technologies that optimize task allocation by taking into account the user's emotional state and provide additional or modified tasks to specific agents.

[0521] The system for implementing this invention employs a combination of technologies to monitor the operational status and progress of agents in real time and to analyze the user's emotions. The server manages the activities of multiple agents and dynamically reassigns tasks as needed. In doing so, it analyzes the emotional state from the user's facial expressions and voice using a terminal worn by the user, such as smart glasses or a wearable device. The software used is OpenCV, a computer vision technology, and TensorFlow, which is suitable for emotion recognition based on it.

[0522] The server receives this data and adjusts the workload using a reinforcement learning algorithm based on sentiment information. It also has an anomaly detection function, and if any malfunction or anomaly occurs in the agent's operation, it can immediately notify using the alert function.

[0523] As a concrete example, in a factory production line, workers' stress levels can be determined in real time by analyzing their facial expressions using smart glasses. Based on this, a server automatically redistributes tasks from specific agents to others, maintaining a balance in the workload. As a result, the burden on workers is reduced, and efficiency can be maintained at a higher level.

[0524] An example of a prompt for a generative AI model is, "What is a way to dynamically adjust the workload by performing facial expression analysis in the work environment based on data collected from a smart glasses device?" Using this prompt allows the system to adapt more to human emotions, supporting smooth execution in real-world situations.

[0525] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0526] Step 1:

[0527] The device captures the user's facial expressions and movements using sensors. Input includes image and audio data acquired from the device's camera and microphone. The device preprocesses this data using OpenCV to extract features for emotion recognition. The output is feature data.

[0528] Step 2:

[0529] The device inputs feature data into a TensorFlow model to estimate the user's emotional state. TensorFlow uses the pre-trained model to analyze the input data and outputs emotion labels. The output includes emotion labels such as "joy" and "anger" along with their confidence scores.

[0530] Step 3:

[0531] The server receives sentiment labels sent from the terminal and matches them against the agent's current task load data. The inputs are the user's sentiment information and the agent's task data. Based on this data, the server runs a reinforcement learning algorithm. The output is a task reassignment instruction.

[0532] Step 4:

[0533] The server sends task reassignment instructions to the agents, dynamically changing the agents' task configurations. The server then records the newly configured tasks and coordinates cooperation between agents. The input is the reassignment instructions, and the output is the updated task schedule.

[0534] Step 5:

[0535] Users provide feedback on improvements made by the system. This feedback is collected from the terminal and sent to the server. The server stores this feedback for future reinforcement learning and adjusts the overall system settings. The input is user feedback information, and the output is updated training data information.

[0536] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0537] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0538] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0539] [Fourth Embodiment]

[0540] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0541] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0542] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0543] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0544] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0545] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0546] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0547] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0548] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0549] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0550] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0551] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0552] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0553] This invention provides a method for efficiently managing a multi-agent system utilizing artificial intelligence and optimizing task execution. Specifically, it includes a mechanism in which a server monitors the operational status and progress of each agent in real time and dynamically reassigns tasks as needed. It also has a function to continuously improve agent performance through reinforcement learning and adjust system settings based on user feedback.

[0554] The server periodically collects and analyzes information from each agent. This information includes the operational status and task progress of each agent, and immediately generates an alert if an anomaly is detected. The server uses this information to redistribute tasks among agents. For example, if an agent is overloaded, the server will assign some of its tasks to agents with less load.

[0555] The server uses reinforcement learning algorithms to analyze past data and optimize task distribution among agents. This process refines the agents' cooperation over time, improving overall performance.

[0556] Users can provide feedback through their devices. This feedback is analyzed by the server and reflected in the system settings. For example, if a user instructs the server to prioritize tasks in a certain category, the server will immediately implement task allocation that takes this into account. This entire process is automated, minimizing user intervention and enabling efficient operation.

[0557] This invention enables the optimal coordination of multiple AI agents used within an enterprise, allowing various tasks to be performed efficiently and quickly. Immediate response and optimization processes based on anomaly detection reduce wasted resources. This provides a mechanism for improving overall productivity.

[0558] The following describes the processing flow.

[0559] Step 1:

[0560] The server collects operational status and progress data from each agent. This includes which tasks each agent is currently running, the amount of resources being used, and the degree of task completion. The server stores this information in a central database.

[0561] Step 2:

[0562] The server analyzes the collected data to check for any anomalies. If an anomaly is detected, the server generates an alert and notifies the administrator. This allows problems to be detected early and corrective actions to be taken.

[0563] Step 3:

[0564] The server determines whether tasks need to be dynamically reallocated based on the agent load. If the load is uneven, the server shifts some tasks to other agents with less load, thereby maintaining overall balance.

[0565] Step 4:

[0566] The server activates a reinforcement learning algorithm and learns from past performance data. This algorithm generates a new optimization strategy, which will be used for future task assignments.

[0567] Step 5:

[0568] Users send feedback to the server using their devices. This feedback may include changes to the priority of specific tasks or new requests. The server receives this feedback and adjusts the agent settings. This ensures that user requests are efficiently reflected in system operations.

[0569] (Example 1)

[0570] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0571] In recent years, many companies and organizations have implemented systems in which multiple agents simultaneously perform various tasks. However, there is a need for efficient operation of these agents, dynamic task reassignment, early detection of anomalies, and flexible system configuration that takes user feedback into consideration. Conventional systems have relied on manual work and fixed settings, and have suffered from a lack of real-time adaptability.

[0572] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0573] In this invention, the server includes means for periodically collecting the operating status and work progress of agents, means for analyzing the collected data and detecting operational anomalies, and means for dynamically redistributing tasks based on the analysis results. This enables efficient operation of agents and flexible and adaptive system operation.

[0574] An "agent" is a unit of software or hardware designed to perform a specific task.

[0575] "Operational status" refers to information indicating the progress of the tasks being performed by the agent and the resource usage.

[0576] "Work progress" is an indicator that shows the extent to which the tasks assigned to the agent have been completed.

[0577] "Collecting" means periodically acquiring and accumulating specific information as numerical data or logs.

[0578] "Analyzing" means evaluating collected data using computational processing and algorithms to extract useful information and patterns.

[0579] "Detecting anomalies" means discovering conditions or performance problems that deviate from normal operation.

[0580] "Dynamic reallocation" means flexibly rearranging existing tasks and resources to suit new states and conditions.

[0581] "Reinforcement learning technology" is a method for improving performance by enabling a system to learn optimal actions based on its experience.

[0582] "User input information" refers to data related to instructions and feedback that users provide to the system.

[0583] "Optimizing settings" means adjusting parameters and configurations to achieve efficient system operation.

[0584] This invention provides a solution for effectively managing a multi-agent system utilizing artificial intelligence and appropriately allocating tasks. The server periodically collects information on the operational status and work progress from each agent. This allows the server to understand the overall system status in real time. The collected information is acquired as data packets via the network and stored in a database.

[0585] The server applies reinforcement learning algorithms to analyze the collected data and optimize task allocation among agents. During this process, the server evaluates past task performance results and learns practical task allocation strategies from successful and unsuccessful examples. This analysis can be performed using data analysis programming languages ​​such as Python or R.

[0586] Furthermore, the server can immediately generate alerts if it detects any abnormalities in agent operation and reassign tasks as needed. For example, if an agent is overloaded, it can transfer tasks from that agent to other agents to maintain overall performance. This reassignment is automated, improving efficiency across the entire enterprise.

[0587] Users can provide feedback to the system through their terminals. Specifically, users input task priorities and new instructions using a GUI, and this data is sent to the server. The server analyzes this feedback in real time and can adjust system settings as needed. This optimization of settings allows the system to flexibly respond to user needs.

[0588] As a concrete example, consider a large-scale data analysis project where a sudden increase in computational demands causes some agents to reach their limits. The server detects this situation and immediately redistributes tasks to other agents to resolve the problem. Furthermore, if a user provides feedback requesting priority processing of a specific data category, the server adjusts task priorities accordingly. Such applications are conceivable.

[0589] Examples of prompts for a generative AI model include the following:

[0590] "You are the server managing AI agents within the company. Check the operational status of Agent A, and if it is overloaded, redistribute tasks to Agent B. When doing so, use a reinforcement learning algorithm to determine the optimal distribution and adjust the settings based on user feedback."

[0591] The system realized by this invention enables the effective management of multiple agents within a company or organization, allowing them to perform tasks efficiently and flexibly with minimal intervention.

[0592] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0593] Step 1:

[0594] The server periodically collects information on the operational status and work progress from each agent. In this step, it receives data packets from agents via the network. The specific input is log data of the agents' operational status and progress, and by saving this to a database, it obtains output that is ready for the next analysis process.

[0595] Step 2:

[0596] The server analyzes the collected data to detect operational anomalies. The operational data saved in Step 1 is used as input, and a programming language (e.g., Python) or statistical software is used for analysis. The server analyzes the agent's CPU usage and memory usage, and if these exceed the threshold, it determines it to be an anomaly and generates an alert. This alert serves as a basis for deciding on reallocation.

[0597] Step 3:

[0598] The server dynamically redistributes tasks based on the analysis results. The input consists of anomaly information detected in step 2 and the agent load status. The server uses a reinforcement learning algorithm to evaluate the performance of each agent and transfers tasks to agents with less load. This process outputs an efficient task redistribution.

[0599] Step 4:

[0600] Users provide feedback through their terminals, which is then analyzed by the server. The input consists of user instructions regarding configuration changes and task priorities. The server receives and analyzes this feedback information to adjust system settings as needed. This feedback-based optimization is the final output, improving the system's operational efficiency.

[0601] In this way, the system collects, analyzes, assigns tasks to, and incorporates feedback in real time, continuously optimizing the entire system.

[0602] (Application Example 1)

[0603] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0604] Efficiently preventing load imbalances and malfunctions in data centers is challenging. Conventional systems struggle to accurately understand the operating status of equipment, making it difficult to respond quickly to equipment experiencing heavy loads. Furthermore, dynamic adjustments based on individual feedback are difficult, hindering overall system optimization.

[0605] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0606] In this invention, the server includes means for monitoring the operating status and progress of the equipment being used, means for dynamically reassigning tasks, and means for optimizing the performance of the equipment using a reinforcement learning algorithm. This enables efficient equalization of the load within the data center, prevention of equipment malfunctions, and optimal system adjustments incorporating user feedback.

[0607] "Equipment used" refers to the collective term for electronic devices such as servers and computers that are operated within a data center.

[0608] "Operating status" refers to information about the current operating status and resource consumption of the equipment being used.

[0609] "Progress" is an indicator that shows the extent to which a task or work has been completed.

[0610] "Work" refers to the procedures for processing tasks and processes within a data center.

[0611] "Load" refers to the pressure exerted by the amount of data and computational load on the equipment being used.

[0612] "Reassignment" refers to the process of redistributing tasks or resources to different devices.

[0613] A "reinforcement learning algorithm" is a type of machine learning that learns the optimal action through trial and error.

[0614] "Performance" refers to the ability and efficiency with which the equipment used can perform a given task.

[0615] "Optimization" is the process of adjusting a system to maximize its overall efficiency and effectiveness.

[0616] "An anomaly" refers to an unexpected state that deviates from normal operation.

[0617] "Feedback" refers to information provided based on users' experiences and requests.

[0618] "Equalization" refers to a state where resources and workloads are distributed evenly to eliminate imbalances.

[0619] The system that realizes this invention is primarily server-based. The server is programmed using Python and TensorFlow and collects and monitors the operating status and progress of equipment used within the data center in real time. The server analyzes this information and evaluates the workload on each piece of equipment. If the workload is concentrated on a particular piece of equipment, it dynamically reassigns tasks to equalize the overall load.

[0620] The reinforcement learning algorithm is implemented using TensorFlow, and the server optimizes the performance of the equipment. This enables optimal task allocation based on historical data. Furthermore, the server can dynamically adjust system settings based on user feedback, new requests, and improvements.

[0621] As a concrete example, suppose a data center has a group of servers processing a critical analytical task, and some of the servers become overloaded. In this case, the system on the servers immediately monitors the situation and reassigns tasks from the overloaded servers to the less loaded servers. This ensures overall processing efficiency and prevents problems from occurring.

[0622] An example of a prompt for a generative AI model is as follows: "Please suggest the optimal method and parameter settings for building a reinforcement learning model that monitors server load in a data center and optimizes the task."

[0623] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0624] Step 1:

[0625] The server collects operational status and progress data from the devices being used in real time. Inputs include CPU usage, memory usage, and I / O wait times provided by each device, which are then imported into an internal database. Outputs are the latest operational status data for each device. Specifically, data is periodically retrieved from each device using an internal communication protocol.

[0626] Step 2:

[0627] The server analyzes the collected data and evaluates the load status of each device. The input is operational status data from step 1, and the output is a list showing the load status of each device. This process uses statistical methods to calculate the load and identify devices that may be overloaded. Specifically, it sets thresholds using an anomaly detection algorithm and lists devices whose load has been exceeded.

[0628] Step 3:

[0629] The server reallocates tasks from heavily loaded devices to other less-loaded devices. The input is the load status list obtained in step 2. The output is the task list for each device after reallocation. In this process, dynamic programming techniques are used to equalize the load. Specifically, the server plans and executes the movement of tasks from heavily loaded devices to less loaded ones.

[0630] Step 4:

[0631] The server continuously improves the performance of each device by applying reinforcement learning algorithms. Inputs are progress reports and user feedback, and output is a model of optimized task allocation. The server uses TensorFlow to learn from historical data and generate optimal agent behavior patterns. Specifically, it analyzes performance data and updates the optimal course of action.

[0632] Step 5:

[0633] Terminal users review the results of system configurations optimized by the server and provide feedback as needed. Inputs are optimized task allocations and feedback requests from the server, while output is user feedback information. Specifically, configuration verification and feedback input are performed via a GUI interface.

[0634] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0635] This invention provides a method to improve system efficiency and user satisfaction by combining an emotion engine that recognizes user emotions with a multi-agent system that utilizes artificial intelligence. A server acts as the central point, monitoring the operational status and progress of agents in real time and dynamically reassigning tasks as needed. Agent performance is optimized using reinforcement learning algorithms.

[0636] An emotion engine is used to analyze the user's emotional state on the device. The collected emotional information is sent to the server and reflected in task priorities and agent settings. Based on this emotional information, the server adjusts the cooperation between agents to achieve more adaptive task execution. In addition, system settings can be adjusted in conjunction with the emotion engine based on user feedback.

[0637] For example, if the emotion engine detects that a user is experiencing stress, the server reassigns high-load tasks to another agent, reducing the user's workload. This improves the user experience.

[0638] Furthermore, if a user expresses satisfaction or dissatisfaction, the server will adjust task allocation among agents accordingly. This setting can be flexibly changed according to user requests and the work environment. If an anomaly is detected, the server will immediately issue an alert and pinpoint the source of the problem.

[0639] This system enables AI agents used within a company to work efficiently and harmoniously, adapting to various situations while performing tasks. Through integration with an emotion engine, the system achieves more human-like interactions, improving overall productivity and user satisfaction.

[0640] The following describes the processing flow.

[0641] Step 1:

[0642] The server collects operational status and progress data from each agent. This includes the current task status, resources being used, and the degree of work completion. The server analyzes this data and monitors the agent's performance.

[0643] Step 2:

[0644] The device monitors the user's actions and inputs, and uses an emotion engine to analyze the user's emotional state in real time. The device then sends these analysis results to the server.

[0645] Step 3:

[0646] Based on the user's emotional information, the server adjusts the priority and allocation of agent tasks. For example, if a user is stressed, the server will reallocate tasks to reduce the load on them.

[0647] Step 4:

[0648] The server executes a reinforcement learning algorithm, leveraging historical performance data and user sentiment information to optimize the overall system efficiency. This allows the agent to act more adaptively in subsequent task assignments.

[0649] Step 5:

[0650] Users provide feedback using their devices. This feedback may include task priorities and satisfaction with the process. The server aggregates this feedback and uses it to improve system settings and agent collaboration.

[0651] Step 6:

[0652] The server uses an anomaly detection module to immediately generate alerts and take appropriate action if problems occur in the operation of agents or the system as a whole. This is important for speeding up problem resolution and maintaining system stability.

[0653] (Example 2)

[0654] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0655] In recent years, AI-powered business support systems have become widespread. However, these systems fail to adequately address users' emotional states, limiting their potential for improving work efficiency and user satisfaction. Furthermore, the optimization of cooperation between programs within the system is insufficient, making it difficult to maximize the overall effectiveness of business operations. In addition, they lack the functionality to respond immediately when an anomaly occurs, highlighting the need for rapid problem resolution.

[0656] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0657] In this invention, the server includes means for analyzing the user's emotional state, means for dynamically adjusting task priorities based on the analyzed emotional information, means for monitoring a group of autonomous programs with multiple roles and reassigning roles as needed, means for improving the operational efficiency of the program group using machine learning algorithms, and means for modifying the system configuration based on user input. This enables flexible task management in response to the user's emotions, thereby improving operational efficiency and maximizing user satisfaction.

[0658] "Methods for analyzing a user's emotional state" refer to technologies that detect emotions in real time from the user's facial expressions, tone of voice, etc., and analyze emotional trends based on that data.

[0659] "A means of dynamically adjusting task priorities based on analyzed emotional information" refers to a method that utilizes the results of analyzing user emotional data to constantly optimize the order in which tasks are performed in a way that does not burden the user.

[0660] A "group of autonomous programs with multiple roles" is a collection of software that performs different functions independently, but is designed to work together as a whole to achieve a specific objective.

[0661] "Methods for improving the operational efficiency of a group of programs using machine learning algorithms" refers to techniques that use artificial intelligence algorithms to analyze the operational data within a program and improve its efficiency based on that analysis.

[0662] "Means of modifying system configuration based on user input" refers to the process of adjusting and improving system settings and operation based on user feedback and operation history.

[0663] This invention is a multi-agent system that recognizes the user's emotional state in real time and optimizes the system's operation based on that recognition. The following describes a specific implementation of this system.

[0664] The main components consist of a server, a terminal, and a user. The server functions as the core of the system, and the terminal collects user emotion data via an emotion engine. The terminal is equipped with image recognition software and voice analysis tools to analyze the user's voice and facial expressions, specifically using libraries such as OpenCV. These tools receive data through the user interface and send it to the server.

[0665] The server uses machine learning algorithms to analyze the received sentiment data. At this stage, it monitors the agent's activity and dynamically optimizes task priorities based on the analysis results. By applying reinforcement learning algorithms, the cooperative relationships between agents are automatically adjusted. For example, if the analysis indicates that the user is experiencing stress, the server reduces the user's workload by assigning high-load tasks to other agents.

[0666] Furthermore, the terminal provides the user with feedback on instructions from the server and changes in status, and adjusts the overall system settings as needed. For this purpose, a dashboard application is used to provide visual information. The server collects user feedback and uses it to adjust system settings for further optimization.

[0667] A concrete example of how this system works is when a user is stressed due to project deadlines. In this situation, the emotion engine detects the stress level, and the server quickly reassigns tasks. This allows the project to progress smoothly, and the user can achieve better results.

[0668] An example of a prompt might be, "How should the system reassign tasks when the user is experiencing stress?" Using such prompts, it's possible to leverage generative AI models to further improve and adapt the system.

[0669] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0670] Step 1:

[0671] The device collects the user's emotions using an emotion engine. Here, the user's facial expressions are captured by a camera, and their voice is recorded by a microphone. Image data and audio data are acquired as input, and image recognition software and audio analysis tools are used to analyze them. Specifically, OpenCV is used to analyze facial expressions from the image data, and the voice tone is analyzed by the audio analysis tool to estimate the emotional state. Emotional data is generated as a result of the analysis, and this becomes the output for the next step.

[0672] Step 2:

[0673] The device sends the emotion data obtained in Step 1 to the server. It takes emotion data as input and sends it to the server using a digital communication protocol. Encrypted data transmission is performed to ensure data reliability and security. The transmitted emotion data is received by the server and used in the next step.

[0674] Step 3:

[0675] The server analyzes received sentiment data and dynamically adjusts task priorities. Using sentiment data submitted by users as input, it employs a reinforcement learning algorithm for task management. Specifically, it uses sentiment data to evaluate the user's stress level and satisfaction level, and revises the allocation of work tasks. The output is a list of adjusted task priorities, allowing for task reallocation between agents.

[0676] Step 4:

[0677] The server adjusts the cooperation between agents, reflecting task priorities. The priority list generated in step 3 is used as input. The server monitors the agents' operational status and reassigns tasks as needed. Specifically, it reduces the workload of heavily loaded agents and moves tasks to agents with greater processing power. The output provides the adjusted task assignments for each agent.

[0678] Step 5:

[0679] The terminal provides feedback to the user. As input, it receives adjusted task information sent from the server. Specifically, it implements a function to inform the user of the task's current status and changes through a visual dashboard. As output, information is presented in a user-friendly format, enabling effective interaction between the system and the user.

[0680] (Application Example 2)

[0681] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0682] In modern work environments, the emotions and stress levels of human workers significantly impact productivity and work efficiency. However, traditional agent systems and automated work environments have struggled to manage tasks while considering user emotions and workload. Therefore, there is a need to improve overall work efficiency while preventing worker overload.

[0683] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0684] In this invention, the server includes means for monitoring the operational status and progress of agents, means for dynamically reassigning tasks, means for analyzing the user's emotional state, and means for adjusting the workload and assigning auxiliary tasks based on the emotional state. This enables improved user experience within the work environment and increased overall work efficiency.

[0685] An "agent" is an artificial intelligence or automated software unit designed to perform a specific task.

[0686] "Operating status" refers to the situation or state in which an agent is actually performing a task.

[0687] "Progress" is an indicator that shows how much of the assigned task an agent has completed.

[0688] "Means for dynamically reallocating tasks" refers to a function that flexibly adjusts the distribution of tasks among agents according to the situation.

[0689] A "reinforcement learning algorithm" is a learning method that helps an agent select the best course of action based on its experience.

[0690] "User feedback" refers to users' opinions and evaluations of how the system works.

[0691] "Means for analyzing emotional states" refer to methods and technologies for identifying a user's emotions and evaluating their state.

[0692] "Means of adjusting workload and assigning auxiliary tasks" refers to technologies that optimize task allocation by taking into account the user's emotional state and provide additional or modified tasks to specific agents.

[0693] The system for implementing this invention employs a combination of technologies to monitor the operational status and progress of agents in real time and to analyze the user's emotions. The server manages the activities of multiple agents and dynamically reassigns tasks as needed. In doing so, it analyzes the emotional state from the user's facial expressions and voice using a terminal worn by the user, such as smart glasses or a wearable device. The software used is OpenCV, a computer vision technology, and TensorFlow, which is suitable for emotion recognition based on it.

[0694] The server receives this data and adjusts the workload using a reinforcement learning algorithm based on sentiment information. It also has an anomaly detection function, and if any malfunction or anomaly occurs in the agent's operation, it can immediately notify using the alert function.

[0695] As a concrete example, in a factory production line, workers' stress levels can be determined in real time by analyzing their facial expressions using smart glasses. Based on this, a server automatically redistributes tasks from specific agents to others, maintaining a balance in the workload. As a result, the burden on workers is reduced, and efficiency can be maintained at a higher level.

[0696] An example of a prompt for a generative AI model is, "What is a way to dynamically adjust the workload by performing facial expression analysis in the work environment based on data collected from a smart glasses device?" Using this prompt allows the system to adapt more to human emotions, supporting smooth execution in real-world situations.

[0697] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0698] Step 1:

[0699] The device captures the user's facial expressions and movements using sensors. Input includes image and audio data acquired from the device's camera and microphone. The device preprocesses this data using OpenCV to extract features for emotion recognition. The output is feature data.

[0700] Step 2:

[0701] The device inputs feature data into a TensorFlow model to estimate the user's emotional state. TensorFlow uses the pre-trained model to analyze the input data and outputs emotion labels. The output includes emotion labels such as "joy" and "anger" along with their confidence scores.

[0702] Step 3:

[0703] The server receives sentiment labels sent from the terminal and matches them against the agent's current task load data. The inputs are the user's sentiment information and the agent's task data. Based on this data, the server runs a reinforcement learning algorithm. The output is a task reassignment instruction.

[0704] Step 4:

[0705] The server sends task reassignment instructions to the agents, dynamically changing the agents' task configurations. The server then records the newly configured tasks and coordinates cooperation between agents. The input is the reassignment instructions, and the output is the updated task schedule.

[0706] Step 5:

[0707] Users provide feedback on improvements made by the system. This feedback is collected from the terminal and sent to the server. The server stores this feedback for future reinforcement learning and adjusts the overall system settings. The input is user feedback information, and the output is updated training data information.

[0708] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0709] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0710] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0711] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0712] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0713] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0714] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0715] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0716] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0717] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0718] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0719] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[0720] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0721] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[0722] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0723] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0724] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0725] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0726] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0727] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0728] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[0729] The following is further disclosed regarding the embodiments described above.

[0730] (Claim 1)

[0731] A means of monitoring the operational status and progress of agents,

[0732] A means of dynamically reassigning tasks,

[0733] A means for optimizing the performance of an agent using a reinforcement learning algorithm,

[0734] A means of adjusting settings based on user feedback,

[0735] A system that includes this.

[0736] (Claim 2)

[0737] The system according to claim 1, further comprising means for managing cooperation between agents and changing priorities to promote efficiency.

[0738] (Claim 3)

[0739] The system according to claim 1, further comprising means for alerting an agent to an anomaly by an anomaly detection system.

[0740] "Example 1"

[0741] (Claim 1)

[0742] A means of periodically collecting the agent's operating status and work progress,

[0743] A means for analyzing collected data and detecting operational anomalies,

[0744] A means of dynamically redistributing tasks based on analysis results,

[0745] A means of improving the agent's task assignment using reinforcement learning techniques,

[0746] A means of optimizing settings based on user input information,

[0747] A system that includes this.

[0748] (Claim 2)

[0749] The system according to claim 1, further comprising means for changing the priority of tasks and ensuring effective operation in order to facilitate cooperation between agents.

[0750] (Claim 3)

[0751] The system according to claim 1, further comprising means for early warning of operational abnormalities of the agent by an anomaly detection function.

[0752] "Application Example 1"

[0753] (Claim 1)

[0754] Means for monitoring the operating status and progress of the equipment being used,

[0755] A means of dynamically reallocating tasks,

[0756] A means for optimizing the performance of a device using a reinforcement learning algorithm,

[0757] A means of adjusting settings based on user feedback,

[0758] A means to equalize the load on equipment within a data center,

[0759] A system that includes this.

[0760] (Claim 2)

[0761] The system according to claim 1, further comprising means for managing collaborative systems and changing priorities to promote efficiency.

[0762] (Claim 3)

[0763] The system according to claim 1, further comprising means for alerting equipment abnormalities using an anomaly detection system.

[0764] "Example 2 of combining an emotion engine"

[0765] (Claim 1)

[0766] A means of analyzing the emotional state of users,

[0767] A means of dynamically adjusting task priorities based on analyzed emotional information,

[0768] A means of monitoring a group of autonomous programs with multiple roles and reassigning those roles as needed,

[0769] A means of improving the operational efficiency of a group of programs using machine learning algorithms,

[0770] A means of modifying the system configuration based on user input,

[0771] A system that includes this.

[0772] (Claim 2)

[0773] The system according to claim 1, further comprising means for controlling the cooperative relationships between groups of programs and for changing their order in order to improve the effectiveness of operations.

[0774] (Claim 3)

[0775] The system according to claim 1, further comprising means for notifying of a malfunction in a group of programs by an anomaly detection function.

[0776] "Application example 2 when combining with an emotional engine"

[0777] (Claim 1)

[0778] A means of monitoring the operational status and progress of agents,

[0779] A means of dynamically reassigning tasks,

[0780] A means for optimizing the performance of an agent using a reinforcement learning algorithm,

[0781] A means of adjusting settings based on user feedback,

[0782] A means of analyzing the user's emotional state,

[0783] A means of adjusting workload and assigning auxiliary tasks based on emotional state,

[0784] A system that includes this.

[0785] (Claim 2)

[0786] The system according to claim 1, further comprising means for managing cooperation between agents and changing priorities to promote efficiency.

[0787] (Claim 3)

[0788] The system according to claim 1, further comprising means for alerting an agent to an anomaly by an anomaly detection system. [Explanation of symbols]

[0789] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. Means for monitoring the operating status and progress of the equipment being used, A means of dynamically reallocating tasks, A means for optimizing the performance of a device using a reinforcement learning algorithm, A means of adjusting settings based on user feedback, A means to equalize the load on equipment within a data center, A system that includes this.

2. The system according to claim 1, further comprising means for managing the collaborative system and changing priorities to promote efficiency.

3. The system according to claim 1, further comprising means for alerting equipment abnormalities using an anomaly detection system.

Images