Information processing device and information processing method

Abstract

Provided are an information processing device and an information processing method that can identify required knowledge in accordance with an occurred event, for supporting work that requires combined knowledge. A database search device receives, as an input, an event that has occurred at a work site, and outputs a work instruction to support work. The database search device has a knowledge domain analysis unit that receives information about work, and selects, on the basis of content dynamically determined from the received information, a knowledge domain required for understanding the input event from among a plurality of knowledge domains.

Description

Information processing device and information processing method

[0001] The present invention relates to an information processing device and an information processing method.

[0002] Today, various applications and solutions are being offered to improve productivity at production sites. For example, by connecting equipment at production sites to a network and utilizing sensors, robots, and other devices, it is possible to aggregate on-site work information in real time to a higher-level system. Furthermore, efforts are being made to improve productivity by utilizing generative AI to solve work issues and obtain information for improving work efficiency.

[0003] Retrieval-Augmented Generation (RAG) is attracting attention as a business application of generative AI. It searches external knowledge databases and embeds information related to the input prompt to output an answer that incorporates external knowledge. By utilizing the quoted information, RAG can perform various tasks such as information extraction, translation, similar sentence generation, and summarization.

[0004] On the other hand, since there are multiple external knowledge databases such as manuals, and it is necessary to select an appropriate external knowledge database depending on the events that occur at the production site, applying RAG requires man-hours for determining the selection of the external knowledge database.A conventional publicly known example of a technology for acquiring appropriate knowledge from multiple knowledge groups has been disclosed (Patent Document 1).

[0005] Japanese Patent Application Laid-Open No. 2000-200197

[0006] In the above-mentioned conventional example, knowledge is selected based on relevance, but since events and situations occurring at manufacturing sites and other workplaces change dynamically due to multiple factors such as differences in workers, environments, and locations, it is not always possible to select an appropriate knowledge database based on past relevance. Another issue is that the relevance to creating search queries for utilizing generative AI cannot be taken into consideration.

[0007] The present invention has been made in view of the above background. That is, one of the objects of the present invention is to provide an information processing device and an information processing method that can identify the necessary knowledge in response to an event that has occurred when supporting a task that requires complex knowledge.

[0008] In order to solve the above-mentioned problems and achieve the above-mentioned object, the information processing device of the present invention is an information processing device that outputs work support information for supporting work, and is equipped with: a storage device that stores a knowledge database table that manages the relationship between a knowledge database and a work site evaluation value that changes dynamically according to work progress, and the relationship between the knowledge database and work content that does not change dynamically according to work progress; and a calculation device that, when receiving as input data dynamic attribute information, which is information about work that changes dynamically according to work progress, and static attribute information, which is information about work that does not change dynamically according to work progress, for an event that occurs at the work site, calculates the work site evaluation value based on the dynamic attribute information, calculates static work content based on the static attribute information, and selects the knowledge database required to output the work support information based on the knowledge database table, the work site evaluation value, and the static work content.

[0009] The information processing method of the present invention is an information processing method by an information processing device that includes a storage device storing a knowledge database table that manages the relationship between a knowledge database and a dynamically changing evaluation value of a work site, and the relationship between the knowledge database and static work content, and an arithmetic device, and performs the following steps: accepting, as input data, dynamic attribute information, which is information about work that dynamically changes depending on differences in the worker and / or work environment, and static attribute information, which is information about work that does not change depending on differences in the worker and / or work environment, for an event that occurs at the work site; determining an evaluation value of the work site based on the dynamic attribute information; determining the static work content based on the static attribute information; and selecting the knowledge database required to output work support information based on the knowledge database table, the static work content, and the evaluation value of the work site.

[0010] According to the present invention, for task support that requires complex knowledge, it is possible to identify the necessary knowledge according to the event that has occurred. Note that the effects described herein are not necessarily limited to those described herein, and may be any of the effects described in this disclosure.

[0011] FIG. 1 is a configuration diagram showing a system including a database search device according to an embodiment of the present invention. FIG. 2 is an explanatory diagram showing an example of creating a related knowledge search query according to an embodiment of the present invention. FIG. 3A is an explanatory diagram showing an example of creating an integrated search query according to an embodiment of the present invention. FIG. 3B is an explanatory diagram showing an example of creating a work instruction according to an embodiment of the present invention. FIG. 4 is an explanatory diagram showing an example of a knowledge database table according to an embodiment of the present invention. FIG. 5 is a flowchart explaining an example of an operation according to an embodiment of the present invention. FIG. 6A is an explanatory diagram showing an example of a method for evaluating dynamic attributes according to an embodiment of the present invention. FIG. 6B is an explanatory diagram showing an example of a method for evaluating dynamic attributes according to an embodiment of the present invention. FIG. 6C is an explanatory diagram showing an example of a method for evaluating dynamic attributes according to an embodiment of the present invention. FIG. 6D is an explanatory diagram showing an example of a method for evaluating dynamic attributes according to an embodiment of the present invention. FIG. 6E is an explanatory diagram showing an example of a method for evaluating dynamic attributes according to an embodiment of the present invention. FIG. 7 is an explanatory diagram showing an example of knowledge database table selection according to an embodiment of the present invention. FIG. 8 is an explanatory diagram showing an example of a knowledge database setting interface according to an embodiment of the present invention. FIG. 9 is an explanatory diagram showing an example of a log of integrated search query creation according to an embodiment of the present invention. FIG. 10 is a diagram showing an example of the hardware configuration of a computer applied to the database search device.

[0012] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Note that the following description and drawings are merely examples for explaining the present invention, and some omissions and simplifications have been made as appropriate for clarity of explanation. The present invention can also be implemented in various other forms. Furthermore, unless otherwise specified, each component may be singular or plural.

[0013] In the following description, various types of information may be described using expressions such as "table," but the various types of information may also be expressed using other data structures. Furthermore, when describing identification information, expressions such as "ID," "name," and "number" are used, but these are interchangeable and other expressions may also be used. Furthermore, in the following description, processing may be described using a functional block as the subject, but the subject of the processing may be a CPU or a device instead of a functional block.

[0014] Elements with reference numbers having branch numbers (for example, knowledge database 20-1, knowledge database 20-2, etc.) are represented by adding the higher number of the branch number to the element when there is no need to particularly distinguish between them. For example, knowledge databases 20-1 to 20-3 are referred to as "knowledge database 20" when there is no need to particularly distinguish between them.

[0015] As used herein, a "knowledge domain" refers to a collection of knowledge or information related to a particular field, theme, subject, etc. A knowledge domain may be expressed in the form of a database, instruction manual, web page, manual, video, image, etc.

[0016] A "knowledge database" is a database that collects and organizes knowledge and information about a particular field, theme, or subject, making it easily accessible.

[0017] At a work site, the presence of both skilled and beginner workers can lead to variations in work level. Furthermore, the work environment, such as the location of work and the temperature and humidity of the work area, can change from time to time. A mechanism for absorbing these variations in work level and work environment and providing work support will be described below as one embodiment of the present invention.

[0018] The overall configuration of this embodiment will be described below with reference to Figures 1 to 4. Figure 1 is a configuration diagram showing a database search system including a database search device 10 according to one embodiment of the present invention. Figure 2 is an explanatory diagram showing an example of creating a related knowledge search statement according to one embodiment of the present invention. Figure 3 is an explanatory diagram showing an example of creating an integrated search statement according to one embodiment of the present invention. Figure 4 is an explanatory diagram showing an example of a knowledge database table according to one embodiment of the present invention.

[0019] First, the overall configuration will be described with reference to Fig. 1. A database search system including a database search device 10 of this embodiment is composed of, for example, the database search device 10 that acquires related knowledge from a plurality of knowledge databases 20 based on the contents of input data and outputs work support information (e.g., work instructions 51) for supporting work, and a generation AI engine 30 that outputs answers to questions input to the database search device 10 for creating related knowledge search statements 45 or for acquiring work instructions, as shown in Fig. 1.

[0020] The data to be input to the database search device 10 includes sensor data and work results 41 that change dynamically depending on the worker and work environment, and work attributes 42 that indicate work content and work conditions that are independent of the worker and work environment.

[0021] The database search device 10 outputs, as output data, work instructions 51 that indicate instructions and advice for improving the work efficiency and productivity of the worker based on the work results.

[0022] The database search device 10 includes a knowledge domain analysis unit 101 that selects a knowledge database 20 required for outputting work instructions based on input sensor data / work results 41 and work attributes 42, and outputs a related knowledge search statement 45 for acquiring related knowledge 46 from each knowledge database 20, a related knowledge acquisition unit 102 that acquires the related knowledge 46 from each knowledge database 20, and a knowledge integration unit 103 that integrates multiple related knowledge 46 and creates an integrated search statement 47.

[0023] The knowledge domain analysis unit 101 includes a knowledge database table 1011 that holds information about the knowledge database 20 that can be used to output work instructions, a necessary knowledge domain extraction unit 1012 that selects the knowledge database 20 that is required to output work instructions based on the input sensor data / work results 41, work attributes 42, and knowledge database table 1011, and a related knowledge search statement creation unit 1013 that creates a related knowledge search statement 45 for acquiring related knowledge 46 for the knowledge database 20 selected by the necessary knowledge domain extraction unit 1012.

[0024] The data input and output by the knowledge domain analysis unit 101 will be described using FIG. 2 . The knowledge domain analysis unit 101 receives sensor data / work history 41 and work attributes 42 as input data. In this embodiment, the sensor data / work history 41 includes the work time, the number of errors during the work, whether or not there are abnormalities in the tools used, and a temperature / humidity discomfort index calculated from temperature and humidity values. Furthermore, in this embodiment, the work attributes 42 include the target product, the work process, the tools used, and the work area. The sensor data / work history 41 is information about the work that depends on the worker and / or the work environment and changes dynamically as the work progresses. This information is sometimes referred to as "dynamic attribute information." The work attributes 42 are information indicating work attributes that do not change as the work progresses, such as work content and work conditions that are independent of the worker or the work environment. This information is sometimes referred to as "static attribute information." These are merely examples, and should be determined based on the requirements of the work site.

[0025] The necessary knowledge domain extraction unit 1012 in the knowledge domain analysis unit 101 creates a work report 43 indicating the details of the work based on the sensor data / work results 41 and the work attributes 42. The work report 43 includes the details of the work, the work time, the number of errors, and whether or not there were any tool abnormalities.

[0026] Additionally, the necessary knowledge domain extraction unit 1012 selects a necessary knowledge database 20 (knowledge domain) by referring to the knowledge database table 1011. An example of a method for selecting the knowledge database 20 will be described later.

[0027] Next, the related knowledge search statement creation unit 1013 creates a related knowledge search query 44 for the selected knowledge database 20. The related knowledge search query 44 follows a predetermined format and includes a prerequisite such as "You are an excellent manufacturing site manager," a command statement instructing the creation of a query statement to query the selected knowledge database 20, and a work report 43. Figure 2 shows an example in which the assembly manual for product A, the field work manual for beginners, and the product A configuration manual are selected as knowledge databases 20, and the command statement instructs the rephrasing of the query statement for each knowledge database 20.

[0028] The related knowledge search statement creation unit 1013 sends the related knowledge search question 44 to the generation AI engine 30-2 and receives a related knowledge search statement 45 as an answer from the generation AI engine 30-2. Figure 2 shows an example in which three types of knowledge databases 20 are selected, and three types of related knowledge search statements 45 are received. The related knowledge search statement creation unit 1013 sends the received related knowledge search statement 45 to the related knowledge acquisition unit 102 connected to the corresponding knowledge database 20.

[0029] 3A shows an example of related knowledge 46 acquired by the related knowledge acquisition unit 102 and an integrated search statement 47 created by the knowledge integration unit 103. The related knowledge acquisition unit 102 transmits the related knowledge search statement 45 received from the related knowledge search statement creation unit 1013 to the knowledge database 20, and receives related knowledge 46 from the knowledge database 20 as a response.

[0030] In FIG. 3A , for example, a related knowledge search statement 45, "When assembling product A, can you name three tasks that take a long time and three tasks with complex procedures?" is sent to a knowledge database 20-1 for a product A assembly manual, and a response, "The tasks that take a long time are installing part A, tightening screws, and installing a chassis. The tasks with complex procedures are installing part A, installing part B, and installing part C," is received. The response content is related knowledge 46, and detailed information about the tasks can be obtained. Similarly, in FIG. 3A , points that beginners should pay attention to can be obtained from a knowledge database 20-2 for a field work manual for beginners, and details about the configuration of product A can be obtained from a knowledge database 20-3 for a product A configuration manual, thereby obtaining prerequisite knowledge for understanding events that occur at the work site and work results.

[0031] The related knowledge acquisition unit 102 transmits the acquired related knowledge 46 to the knowledge integration unit 103. The knowledge domain analysis unit 101 also transmits the work report 43 to the knowledge integration unit 103. The knowledge integration unit 103 combines the received multiple pieces of related knowledge 46 and the contents of the work report 43 into a single search statement to create an integrated search statement 47. The integrated search statement 47 includes the premise information "You are an excellent manufacturing site educator" and the instruction statement "Please give me some advice on how to reduce the work time for the following work report," requesting work instructions and work support methods, and also includes the contents of the work report 43 and the related knowledge 46 as related information. By adding the related knowledge 46 to the contents of the work report 43, which can be determined from the input sensor data / work results 41 and work attributes 42, and inputting the information into the generation AI engine 30-1, highly accurate work instructions and work support methods can be obtained.

[0032] The knowledge integration unit 103 inputs the integrated search statement 47 into the generation AI engine 30-1, thereby obtaining work instructions 51 from the integrated search statement 47, as shown in FIG. 3B . This integrated search statement 47 uses the "Product A Assembly Manual," "Beginner's Field Work Manual," and "Product A Configuration Manual" as the knowledge database 20, as described above, and is a search statement intended to obtain an answer that integrates information (related knowledge 46) obtained from these knowledge databases 20. The work instructions 51 are work support information (work instructions, advice) that mainly includes information that integrates the "Product A Assembly Manual," "Beginner's Field Work Manual," and "Product A Configuration Manual." In the example of the work instructions 51 in FIG. 3B , the contents of the major points in the work instructions 51 are created based on the "Beginner's Field Work Manual." Point 3 incorporates information from the "Product A Configuration Manual" as supplementary explanation. Point 4 incorporates information from the "Product A Assembly Manual" as supplementary explanation.

[0033] Next, an example of the knowledge database table 1011 that the necessary knowledge area extraction unit 1012 uses to select the necessary knowledge database 20 will be described with reference to Fig. 4. The knowledge database table 1011 is composed of information indicating the contents of the knowledge database 20 and the conditions for requiring the knowledge database 20, as shown in Fig. 4, for example, and stores the knowledge database name, dynamic attributes, and static attributes linked to the knowledge database ID.

[0034] The knowledge database name indicates documents and data that may be referred to when performing work, such as instructions, work manuals, operation manuals, and layout information of the work place.

[0035] Dynamic attributes indicate conditions that require the relevant knowledge database 20 for events that can be determined from sensor data and work performance 41, which change from time to time. In this embodiment, the following dynamic attributes are stored: work level, work time, error frequency, tool abnormality, and environmental discomfort. Each dynamic attribute is entered as a value evaluated on a three-level scale (1 to 3) or as a binary value (yes or no). If there is no association with a dynamic attribute, no value is entered; in Figure 4, this is indicated by "-" (-). Work level indicates beginner, intermediate, or expert level. Work time and error frequency indicate whether the performance is standard, good, or poor compared to predetermined standard work time and error frequency. Tool abnormality indicates whether the sensor data of the tool being used indicates an abnormal value. Environmental discomfort is evaluated on a three-level scale based on temperature and humidity, indicating whether the work environment based on temperature and humidity is standard, good, or poor. The dynamic attribute items are merely examples, and other items may be stored.

[0036] Static attributes indicate conditions related to work content that are independent of the worker or work environment and can be obtained from work attributes 42. In this embodiment, the work process, tools used, work area, and target product are stored as static attributes. The work process indicates the relevant work step, such as assembly or inspection. The tools used indicate the tools used in the work. The work area indicates the location of the work. The target product indicates the name of the product manufactured at the work site. The items of static attributes are merely examples, and other items may also be stored.

[0037] Dynamic attributes and static attributes are obtained from the input sensor data / work results 41 and work attributes 42, and related knowledge can be obtained by selecting the knowledge database 20 associated with the corresponding dynamic attributes and static attributes from the knowledge database table 1011.

[0038] An example of a work instruction creation flowchart in this embodiment is shown in Fig. 5. First, the knowledge domain analysis unit 101 receives the sensor data / work results 41 and the work attributes 42 (step S001).

[0039] Next, the knowledge domain analysis unit 101 determines the evaluation values ​​of the dynamic attributes {task level, task time, number of errors, and environmental discomfort level} from the sensor data and task results 41 (step S002). The evaluation method will be described later.

[0040] Next, the necessary knowledge domain extraction unit 1012 of the knowledge domain analysis unit 101 refers to the knowledge database table 1011 based on the dynamic attribute (evaluation value of the dynamic attribute) and the task attribute 42, and selects the knowledge database 20 (step S003).

[0041] Next, the knowledge domain analysis unit 101 creates a work report 43 from the sensor data / work results 41 and the work attributes 42, and creates a related knowledge search query 44 (step S004).

[0042] Next, the related knowledge search statement creation unit 1013 of the knowledge domain analysis unit 101 inputs the related knowledge search question 44 to the generation AI engine 30-2 and creates a related knowledge search statement 45 corresponding to the knowledge database 20 selected in step S003 (step S005).

[0043] Next, the related knowledge retrieval statement creation unit 1013 of the knowledge domain analysis unit 101 transmits each related knowledge retrieval statement 45 to the related knowledge acquisition unit 102 connected to the corresponding knowledge database 20, and each related knowledge acquisition unit 102 inputs the related knowledge 45 into the corresponding knowledge database 20 to acquire related knowledge 46 (step S006). The related knowledge acquisition unit 102 transmits the acquired related knowledge 46 to the knowledge integration unit 103.

[0044] Next, the knowledge integration unit 103 combines the related knowledge 46 and the work report 43 to create an integrated search statement 47 (step S007).

[0045] Finally, the knowledge integration unit 103 inputs the integrated search statement 47 to the generation AI engine 30-1 and obtains the work instruction 51 (step S008).

[0046] The evaluation value determination method in step S002 will be described using Figures 6A to 6E. Figure 6A shows an evaluation value determination logic table 601 for work time. Judgment conditions are stored for the relevant product and work process, and a three-level evaluation value is determined based on the judgment conditions. Figure 6B shows an evaluation value determination logic table 602 for error frequency. As with the evaluation value determination logic table 601 for work time, judgment conditions are stored for the relevant product and work process, and a three-level evaluation value is determined based on the judgment conditions. Figure 6C shows an evaluation value determination logic table 603 for tool abnormalities. The evaluation value is determined as a binary value of 1 or 0 based on the presence or absence of a tool abnormality. Figure 6D shows an evaluation value determination logic table 604 for environmental discomfort. Judgment conditions based on the discomfort index of temperature and humidity are stored, and a three-level evaluation value is determined based on the judgment conditions.

[0047] The target product and work process are identified from the work attributes 42, and the values ​​of the sensor data / work performance 41 are compared with the evaluation value determination logic table 601 for work time, the evaluation value determination logic table 602 for error frequency, the evaluation value determination logic table 603 for tool abnormality, and the evaluation value determination logic table 604 for environmental discomfort, thereby making it possible to determine the evaluation values ​​for work time, error frequency, tool abnormality, and environmental discomfort.

[0048] The evaluation value of the work level is determined using the evaluation values ​​of the work time, error frequency, and tool abnormality, which depend on the proficiency of the worker. Fig. 6E shows a logic table 605 for determining the evaluation value of the worker level. The table stores the judgment conditions for the total of the evaluation values ​​of the work time, error frequency, and tool abnormality, and determines the evaluation value in three stages: if the total value is high, the worker is judged as an expert; if the total value is low, the worker is judged as a beginner; and if the total value is intermediate, the worker is judged as an intermediate worker.

[0049] 6A to 6E are merely examples, and other evaluation methods may be used. For example, evaluation may be performed on a five-point scale, or the values ​​of multiple pieces of sensor data and work performance 41 may be weighted and then added together to determine the evaluation value.

[0050] Next, an example of selecting the knowledge database 20 will be described with reference to Fig. 7. In the example of Fig. 7, the sensor data / work performance 41 indicates that the work time is 15 minutes, the number of errors is 2, the tool abnormality is none, and the temperature and humidity discomfort index is 65. The work attributes 42 indicate that the target product is product A, the work process is assembly, the tools used are none, and the work area is assembly booth 1.

[0051] Since it is known from the task attributes 42 that the target product is Product A and the task process is assembly, the task evaluation value is determined for the assembly task of Product A by referring to the task evaluation value determination logic table 601, the task evaluation value determination logic table 602, the tool abnormality evaluation value determination logic table 603, the environmental discomfort evaluation value determination logic table 604, and the worker level evaluation value determination logic table 605 shown in Figures 6A to 6E. In the example of Figure 7, the following evaluation values ​​are obtained: task level: 1 (beginner), task time: 1 (long), error frequency: 3 (standard), tool abnormality: 1 (none), environmental discomfort: 3 (good).

[0052] The obtained evaluation value is set as a dynamic attribute and the task attribute 42 is set as a static attribute, and the knowledge database table 1011 is referenced to select a location where the dynamic attribute and the static attribute can be ANDed. In the example of Figure 7, three types of knowledge databases 20-1 to 20-3 are selected: the product A configuration manual, the product A assembly manual, and the beginner's field work manual. As described above, the evaluation value is determined using the sensor data / task results 41 and the task attribute 42, and by using the evaluation value to reference the knowledge database table 1011, it is possible to determine which areas of knowledge are required as knowledge related to the events indicated by the sensor data / task results 41 and the task attribute 42.

[0053] 8 shows an example of the knowledge database setting interface 1014. The knowledge database setting interface 1014 is an interface (user interface (GUI (Graphical User Interface)) used by the system administrator 81 to register information in the knowledge database table 1011.

[0054] The knowledge database setting interface 1014 has a function for inputting information to be registered in the knowledge database table 1011, and includes a knowledge database name input field 1401, a worker level evaluation value selection field 1411, a work time evaluation value selection field 1412, an error frequency evaluation value selection field 1413, a tool abnormality evaluation value selection field 1414, an environmental discomfort evaluation value selection field 1415, a work process attribute input field 1421, a tool used attribute input field 1422, a work area attribute input field 1423, and a target product attribute input field 1424.

[0055] The name of the knowledge database to be registered is input in the knowledge database name input field 1401. The evaluation values ​​of the dynamic attributes to be associated with the knowledge database 20 to be registered are selected in the worker level evaluation value selection field 1411, work time evaluation value selection field 1412, error frequency evaluation value selection field 1413, tool abnormality evaluation value selection field 1414, and environment discomfort evaluation value selection field 1415. Information on static attributes to be associated with the knowledge database 20 to be registered is input in the work process attribute input field 1421, tool used attribute input field 1422, work area attribute input field 1423, and target product attribute input field 1424. After inputting, by pressing the register button 1402, the input information is registered in the knowledge database table 1011.

[0056] FIG. 9 shows an example of an integrated search statement creation log. The integrated search statement creation log 901 records input logs of the sensor data / work history 41 and work attributes 42 described in FIGS. 2 and 3 . It also records evaluation results based on the sensor data / work history 41 and the selection results of the knowledge database 20 using the evaluation results. Additionally, it records creation logs of the related knowledge search query 44, related knowledge search statement 45, related knowledge 46, and integrated search statement 47. From the integrated search statement creation log 901, it is possible to confirm which knowledge database 20 was selected for each input data, how the related knowledge 46 was acquired, and how it was combined into the integrated search statement 47. In the example of FIG. 9 , the data is recorded in JavaScript Object Notation (JSON) format, but other data formats may also be used.

[0057] As described above, according to this embodiment, the necessary knowledge database 20 is identified from input data indicating the event that occurred and the work results, relevant information is obtained from the relevant knowledge database 20, and a question to be input to the generation AI engine is created, thereby making it possible to obtain appropriate work instructions.

[0058] That is, according to this embodiment, for work support that requires complex knowledge, it is possible to identify the necessary knowledge in accordance with the event that has occurred and obtain appropriate work instructions.

[0059] In addition, it is possible to reduce the amount of work required to check the necessary information to support work, thereby contributing to a reduction in personnel resources and work time.

[0060] Furthermore, according to this embodiment, it is possible to improve the accuracy of work support utilizing the generation AI by integrating and using multiple types of knowledge databases 20.

[0061] Furthermore, the above-described configurations, functional units, processing units, processing means, etc. may be partially or entirely implemented in hardware, for example, by designing them as integrated circuits. The above-described configurations, functions, etc. may also be implemented in software, with a processor interpreting and executing a program that implements each function. Information such as the programs, tables, and files that implement each function can be stored in a memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD.

[0062] 10 is a diagram showing an example of the hardware configuration of a computer applied to the database search device 10. The computer 1000 may be referred to as an "information processing device." The computer 1000 includes a CPU 1001, a ROM 1002, a RAM 1003, a non-volatile storage device 1004 that can read and write data, a network interface 1005, and an input / output interface 1006. These are connected to each other via a bus 1007 so that they can communicate with each other.

[0063] The CPU 1001 is a computing device that loads various programs (not shown) stored in the ROM 1002 and / or storage device 1004 into the RAM 1003 and executes the programs loaded into the RAM 1003, thereby realizing the various functions of the functional blocks described above. The RAM 1003 is loaded with the various programs to be executed by the CPU 1001 as described above, and temporarily stores data used when the CPU 1001 executes the various programs. The ROM 1002 and / or storage device 1004 are non-volatile storage media, and various programs are stored in the ROM 1002 and / or storage device 1004.

[0064] The network interface 1005 is an interface for connecting the computer 1000 to a network. The input / output interface 1006 is an interface for connecting the computer 1000 to an operation device and a display (display device) capable of displaying images.

[0065] Instead of the computer 1000, a hardware device in which part or all of the computer 1000 is configured using an FPGA (Field Programmable Gate Array) or the like may be used.

[0066] Furthermore, the above-described arrangement of the various functional units, various processing units, and various databases of the database search device 10 and the system including the database search device 10 is merely an example. The arrangement of the various functional units, various processing units, and various databases can be changed to an optimal arrangement in terms of the performance, processing efficiency, communication efficiency, etc. of the hardware and software provided in these devices.

[0067] Furthermore, the configuration (schema, etc.) of the database that stores the various types of data described above can be flexibly changed from the perspective of efficient use of resources, improved processing efficiency, improved access efficiency, improved search efficiency, and the like.

[0068] Furthermore, in the above-described embodiment, a manufacturing site was described as an example, but the present invention is not limited to this. For example, advertisements and guidance presented to customers at a commercial facility may be changed according to the customer's movements and characteristics. In this case, the above-described processes are replaced with products and advertisements / guidances. Furthermore, the knowledge database 20 is a system that handles customer service manuals, product descriptions, etc.

[0069] In the above embodiment, the work instructions 51 (work support information) may be accumulated (stored) in a database. In this case, a work support document such as an existing manual may be automatically revised based on the work instructions 51 (work support information) accumulated in the database. For example, information including the work instructions 51 accumulated in the database and the existing manual (work support document) may be input to an AI engine (generation AI engine), which may then output a revised manual (work support document).

[0070] Furthermore, a work support document such as a new manual may be automatically generated based on the work instructions stored in the database. For example, information including the work instructions 51 (work support information) stored in the database may be input to an AI engine (generation AI engine) to output a new manual (work support document).

[0071] The present invention can also have the following configuration.

[0072] [1] A database search device including an information processing device that receives input data including information about an event that has occurred at a work site and outputs work support information to support work, wherein the information processing device is capable of accessing a plurality of knowledge areas, and is configured to select, from the plurality of knowledge areas, the knowledge area that is necessary for outputting the work support information and for understanding the event, based on content dynamically determined from the input data.

[0073] [2] A database search system having: a generating AI engine; an information processing device that receives input data including information about an event that occurred at a work site and outputs work support information using the generating AI engine; and a plurality of knowledge areas that can be accessed by the information processing device, wherein the information processing device is configured to select, from the plurality of knowledge areas, the knowledge area that is necessary for outputting the work support information and for understanding the event, based on content dynamically determined from the input data.

[0074] [3] A work support device including an information processing device that receives input data including information about an event that has occurred at a work site and outputs work support information, wherein the information processing device is configured to be able to access a plurality of knowledge areas, and to select, from the plurality of knowledge areas based on content dynamically determined from the input data, a knowledge area that is necessary for outputting the work support information and for understanding the event, obtain related knowledge from the selected knowledge area, and create the work support information based on the related knowledge.

[0075] [4] The work support device according to [3], wherein the information processing device is configured to store the output work support information in a storage device, and update an existing work support document or create a new work document based on the stored work support information.

[0076] 10...database search system, 20...knowledge database, 30...generation AI engine, 41...sensor data / work results, 42...work attributes, 43...work report, 44...question sentence for related knowledge search, 45...related knowledge search sentence, 46...related knowledge, 47...integrated search sentence, 91...system administrator, 101...knowledge domain analysis unit, 102...related knowledge acquisition unit, 103...knowledge integration unit, 1011...knowledge database table, 1012...necessary knowledge domain extraction unit, 1013...related knowledge search sentence creation unit, 1014...knowledge database setting interface

Claims

1. An information processing device that outputs work support information to support work, comprising: a storage device that stores a knowledge database table that manages the relationship between a knowledge database and an evaluation value of a work site that changes dynamically according to work progress, and the relationship between the knowledge database and work content that does not change dynamically according to work progress; and a calculation device that, when receiving as input data dynamic attribute information, which is information about work that changes dynamically according to work progress, and static attribute information, which is information about work that does not change dynamically according to work progress, for an event that occurred at the work site, calculates an evaluation value of the work site based on the dynamic attribute information, calculates static work content based on the static attribute information, and selects the knowledge database required to output the work support information based on the knowledge database table, the evaluation value of the work site, and the static work content.

2. An information processing device according to claim 1, wherein the dynamic attribute information is sensor data and work results, and the static attribute information is work attributes.

3. An information processing device according to claim 2, wherein the arithmetic unit is configured to create a related knowledge search statement for acquiring related knowledge from the selected knowledge database.

4. An information processing device according to claim 3, wherein the arithmetic unit is configured to: use the related knowledge search statement to acquire the related knowledge from the selected knowledge database; integrate the acquired related knowledge to create an integrated search statement for searching the work support information; and create the work support information using the integrated search statement.

5. An information processing device according to claim 4, wherein the arithmetic unit is configured to have a log function that enables confirmation of the selection results of the knowledge database, the creation record of the related knowledge search statement, and the creation record of the integrated search statement.

6. An information processing device according to claim 3, wherein the arithmetic device is configured to: create a work report based on the dynamic attribute information and the static attribute information; create a related knowledge search query from the work report and the selected knowledge database; create a related knowledge search query from the related knowledge search query to obtain the related knowledge from the knowledge database; and use the related knowledge search query to obtain the related knowledge required to output the work support information from the knowledge database.

7. An information processing device according to claim 6, wherein the calculation device creates an integrated search statement for searching the work support information by integrating the acquired related knowledge and the work report, and creates the work support information using the integrated search statement.

8. An information processing device according to claim 7, wherein the arithmetic device uses a generation AI engine to create the related knowledge search statement for retrieving the related knowledge from the knowledge database from the related knowledge search question statement, and uses the generation AI engine to create the work support information from the integrated search statement.

9. An information processing device according to claim 8, wherein the arithmetic unit is configured to store the output work support information in a storage device, and update an existing work support document or create a new work document based on the stored work support information.

10. An information processing device according to claim 1, wherein the arithmetic unit is configured to be able to register in the knowledge database table the relationship between the knowledge database and the dynamically changing evaluation value of the work site, and the relationship between the knowledge database and the static work content, via a user interface into which information can be input.

11. An information processing method using an information processing device that includes a storage device storing a knowledge database table that manages the relationship between a knowledge database and a dynamically changing evaluation value of a work site, and the relationship between the knowledge database and static work content, and an arithmetic device, the information processing method performing the following steps: accepting, as input data, dynamic attribute information, which is information about work that dynamically changes depending on differences in the worker and / or work environment, and static attribute information, which is information about work that does not change depending on differences in the worker and / or work environment, for events that occur at the work site; determining an evaluation value of the work site based on the dynamic attribute information; determining the static work content based on the static attribute information; and selecting the knowledge database required to output work support information based on the knowledge database table, the static work content, and the evaluation value of the work site.

Images