Information processing device and information processing method

Abstract

An information processing device (100) comprises: an examination information input unit (103) that accepts input of examination information of a product; a target item extraction unit (115) that extracts, as target items, on the basis of the examination information, one or more items from among components constituting the product, events occurring in the components, causes of the events, and effects resulting from the occurrence of the events, and extracts requirements that must be satisfied by the target items; an association information retention unit (111) that retains association information associating two or more items including the target items, from among the components, events, causes, and effects; an association information editing unit (116) that generates first related information, which is a description of the target items, on the basis of the target items and the requirements, identifies items other than the target items on the basis of the first related information, generates second related information by associating the identified other items with the first related information, and generates combined information by combining the second related information with the association information; and an input / output unit (101) that outputs the combined information generated by the association information editing unit (116).

Description

Information processing device and information processing method

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

[0002] In order to maintain the reliability, safety, and maintainability of equipment, it is necessary to write down and analyze the reliability knowledge of the target equipment at the preliminary design stage and the stage of formulating an operation plan.Failure Mode and Effect Analysis (FMEA) and Fault Tree Analysis (FTA) are often used as methods for writing down equipment reliability knowledge.

[0003] Patent Document 1 discloses a technology for creating an FMEA sheet, which describes the method as including the steps of: importing a plurality of documents; morphologically analyzing words in each of the imported documents using a morphological analysis dictionary to divide them into a plurality of morphological words; calculating the co-occurrence frequency of the morphological words in the documents; generating a co-occurrence frequency network using morphological words that have a co-occurrence frequency relationship of at least a certain level among the plurality of morphological words; grouping the plurality of documents using the co-occurrence frequency network; extracting, from each document belonging to the same group, words identical to FMEA words registered in an FMEA word concept dictionary as FMEA words to be used in creating an FMEA sheet; and substituting the extracted FMEA words into the FMEA sheet.

[0004] Japanese Patent Application Laid-Open No. 2008-102758

[0005] It is difficult to comprehensively write out information using Patent Document 1. In reality, it is necessary to combine the creation of failure mode and effect analysis sheets by these automatic processes with manual review and knowledge writing.

[0006] Furthermore, it is difficult to write out reliability information that comprehensively takes into account various operating conditions and external environments. In practice, the conditions for completing a failure mode and effects analysis are often determined by determining in advance the items to be considered in the analysis and then reaching a consensus among various stakeholders involved in the analysis, such as design, operation, and maintenance personnel. However, if too much focus is placed on the items to be considered, other important events may be overlooked. This could result in a loss of reliability or safety of the equipment, which could lead to a serious accident.

[0007] An object of the present invention is to provide an information processing device and an information processing method with improved comprehensiveness of reliability information.

[0008] In order to solve the above-mentioned problems, an information processing device of the present invention comprises: a study information input unit that inputs study information about a product; a target item extraction unit that extracts, based on the study information, one or more of a part that constitutes the product, an event that occurs in the part, a cause of the event, and an impact of the occurrence of the event as target items, and extracts requirements that the target items must satisfy; a linking information holding unit that holds linking information that links two or more items, including the target item, from the part, the event, the cause, and the impact; a linking information editing unit that generates first related information that is a description of the target item based on the target item and the requirements, identifies items other than the target item based on the first related information, generates second related information that associates the identified items with the first related information, and generates combination information that combines the second related information with the linking information; and an output unit that outputs the combination information generated by the linking information editing unit.

[0009] an information processing method according to the present invention, comprising the steps of: an investigation information input unit receiving input of investigation information for a product; a target item extraction unit extracting, based on the investigation information, one or more of a part constituting the product, an event occurring in the part, a cause of the event, and an impact of the occurrence of the event as target items, and extracting requirements to be satisfied by the target items; a linking information holding unit holding linking information linking two or more of the part, the event, the cause, and the impact, including the target item; an association information editing unit generating, based on the target item and the requirements, first association information that describes the target item; an association information editing unit identifying, based on the first association information, items other than the target item, and generating second association information that associates the identified items with the first association information; an association information editing unit generating combination information by combining the second association information with the association information; and an output unit outputting the combination information generated by the association information editing unit. Other means will be described in the detailed description of the invention.

[0010] According to the present invention, it is possible to improve the comprehensiveness of reliability information.

[0011] 1 is a configuration diagram of a support system according to an embodiment of the present invention. FIG. 1 is an example of a screen of an input / output unit. FIG. 2 is a diagram explaining basic information of a linking information storage unit. FIG. 3 is a diagram explaining an example of a failure mode and effect analysis ontology in a linking information storage unit. FIG. 4 is a diagram explaining example linking information in a linking information storage unit. FIG. 5 is a diagram explaining property information of a battery cell. FIG. 6 is a diagram explaining property information of an overcurrent. FIG. 7 is a diagram explaining property information of a "fire due to thermal runaway" node. FIG. 8 is a diagram explaining a prompt database for linking information generation. FIG. 9 is a diagram explaining generation data in JSON format output by a linking information generation unit using a question and answer model. FIG. 10 is a diagram explaining linking information obtained by converting generation data into a graph structure. FIG. 11 is a diagram explaining generation data in JSON format output by a linking information generation unit using a question and answer model. FIG. 12 is a diagram explaining linking information obtained by converting generation data into a graph structure. FIG. 13 is a diagram explaining a failure mode and effect analysis ontology in a linking information storage unit. FIG. 14 is a diagram explaining a prompt statement. 1 is a diagram showing target items and requirements of JSON format data output by a linking information generation unit using a question and answer model. FIG. 2 is a diagram showing target items and requirements of JSON format data output by a first related information generation unit using a question and answer model. FIG. 3 is a diagram showing an example of a prompt sentence. FIG. 4 is a diagram showing first related information. FIG. 5 is a diagram showing an editing procedure database. FIG. 6 is a diagram explaining linking information related to procedure #1. FIG. 7 is a diagram explaining a prompt sentence related to procedure #1. FIG. 8 is a diagram explaining linking information related to procedure #1. FIG. 9 is a diagram showing a prompt sentence related to procedure #2. FIG. 10 is a diagram showing second related information generated in procedure #2. FIG. 11 is a diagram showing a prompt sentence related to procedure #3. FIG. 12 is a diagram showing linking information generated in procedure #3. A flowchart of processing in this embodiment. A flowchart of processing in this embodiment. A block diagram of an information processing device.

[0012] Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings. The information processing device of this embodiment supports the creation of a failure mode and effect analysis sheet as linked information related to the reliability of specific equipment. Here, the failure mode and effect analysis refers to information that links together information on a target component, a failure mode that may occur in this component, the impact of the occurrence of the failure mode, and the cause of the occurrence of the failure mode. The target equipment in this embodiment is, for example, a "battery system" that includes multiple storage batteries. Furthermore, the item to be considered in the failure mode and effect analysis is "design considerations related to fire prevention."

[0013] 1 is a configuration diagram of an information processing device 100 according to this embodiment. The information processing device 100 according to this embodiment is configured to include an input / output unit 101 and a failure mode and effect analysis creation support unit 102. The input / output unit 101 is a user interface that is operated by a user when creating a failure mode and effect analysis. The input / output unit 101 displays information on a display unit, accepts mouse operations and character string input via a keyboard, and is also capable of uploading input data and downloading output data.

[0014] The input / output section 101 includes a study information input section 103, a product name input section 104, an automatic creation instruction section 105, a failure mode and effect analysis display and editing section 106, a download button 107, an upload button 123, a study information statistics output section 108, and a save button 109.

[0015] 2 shows an example of a screen of the input / output unit 101. This screen of the input / output unit 101 is a screen displayed on the display unit of the information processing device 100. The information processing device accepts input of information via the screen of the input / output unit 101 and outputs information.

[0016] The product name input section 104 is a text box for inputting the name of a facility or product to be targeted when writing out the reliability-related linking information. In this embodiment, the product name input section 104 has "battery system" input as the target product name.

[0017] The consideration information input unit 103 is a text box into which the user inputs the desired product consideration information when writing out the reliability-related linking information. The consideration information input unit 103 in this embodiment accepts input in natural language. However, the present invention is not limited to this. The consideration information input unit 103 may be configured to accept input of structured information in the form of specific items and points to be focused on in the items of failure mode and effect analysis.

[0018] The user inputs, for example, "design study regarding fire prevention" in natural language to the study information input unit 103 of this embodiment. In this way, the user enters information regarding the impact that the user wants to prevent as study information into the study information input unit 103. Other information entered into the study information input unit 103 may be information regarding the current study phase of the target facility, such as "design study regarding the cooling system," information regarding maintenance measures to be implemented in the future, such as "introduction of a predictive maintenance system," or information such as "use as educational material."

[0019] The automatic creation instruction unit 105 is an instruction unit that accepts user instructions for the automatic creation of reliability-related linking information to be performed by the failure mode and effect analysis creation support unit 102. The automatic creation instruction unit 105 is configured to include a "Generate component tree" button 201, a "Generate failure modes" button 202, a "Generate effects" button 203, and a "Generate causes" button 204.

[0020] For example, when a user clicks the "Generate Failure Mode" button 202, the failure mode and effect analysis creation support unit 102 (described later) operates to automatically create a failure mode for each component of the battery system and output it to the failure mode and effect analysis display and editing unit 106. The automatic creation instruction unit 105 further includes a target component ID text box 205 and a detailed instruction text box 206.

[0021] The target part ID text box 205 is a text box that accepts input of the target part ID. The user can input the part ID entered in the failure mode effect analysis display editing unit 106 into the target part ID text box 205 and click the "Generate failure mode" button 202 to instruct the generation of a failure mode for the target part.

[0022] The detailed instruction text box 206 is a text box in which the user can input detailed instructions in natural language regarding what information they want to output. The user can specify "output a failure mode when the external temperature is lower than a specified temperature" in the detailed instruction text box 206 and click the "Generate failure mode" button 202 to instruct the generation of a failure mode when the external temperature is lower than a specified temperature.

[0023] The failure mode and effect analysis display editing unit 106 displays the currently created linking information related to reliability in an editable format that is easy for the user to view. In this embodiment, the failure mode and effect analysis display editing unit 106 outputs, in a table format, component names, failure mode names, effect names, cause names, IDs, and combination information indicating the relationships between them. However, the failure mode and effect analysis display editing unit 106 is not limited to this, and the table items may include the contents of other linking information. Furthermore, the failure mode and effect analysis display editing unit 106 may display, instead of a table structure, a graph structure consisting of nodes and edges, or a natural language description. The failure mode and effect analysis display editing unit 106 functions as an output unit that outputs the combination information generated by the linking information editing unit 116.

[0024] The information displayed in the failure mode effect analysis display / edit unit 106 can be edited by the user through on-screen operations. For example, if "over-discharge", which is not output as a failure mode of "battery cell", is added to the table, the data of the linking information, which will be described later, can also be edited and updated.

[0025] The study information statistics output unit 108 can calculate and output related statistics for the study information set in the study information input unit 103 from the linking information on reliability currently being created. Detailed processing will be described later, but in the case of "design study on fire prevention," the target item extraction unit defines the statistics on how many pieces of fire-related impact information have been output and what percentage of the total impact it accounts for, and then calculates and displays the actual statistics.

[0026] The URL text box 124 is a text box for specifying the download destination and upload destination. When the URL (Uniform Resource Locator) of a file is specified in the URL text box 124 and the download button 107 is clicked, the save processing unit 121 downloads the linking information related to reliability stored in the URL to this information processing device. When the URL of a file is specified in the URL text box 124 and the upload button 123 is clicked, the save processing unit 121 uploads the linking information related to reliability to the location specified by the URL. Uploading and downloading are performed using files in a predetermined format designed in advance, such as a CSV (Comma Separated Values) file, a JSON file, a Markdown Mermaid file, or a binary file.

[0027] When the save button 109 is clicked, the save processing unit 121 saves the linking information. The save processing unit 121 of this embodiment stores and saves the linking information in the failure mode and effect analysis database 110.

[0028] 1 , the failure mode and effect analysis creation support unit 102 includes a linking information holding unit 111, a display data conversion unit 112, a save processing unit 121, a generation data conversion unit 114, a linking information generation unit 113, a linking information generation prompt database 122, a target item extraction unit 115, and a linking information editing unit 116.

[0029] The linking information holding unit 111 holds linking information related to reliability. The linking information related to reliability is displayed in the failure mode and effect analysis display and editing unit 106 of the input / output unit 101, used by the study information statistics output unit 108, and also saved by the saving processing unit 121. The linking information related to reliability stored in the linking information holding unit 111 is shown in Figures 3A to 3C and 4A to 4D. The linking information holding unit 111 holds various basic information 301, a failure mode and effect analysis ontology 302 of the target reliability information, and linking information 303.

[0030] 3A is a diagram illustrating basic information 301 in the linking information storage unit 111. The basic information 301 includes information from the product name input unit 104, information from the consideration information input unit 103, and basic information related to writing out reliability information such as a detailed description of the content to be implemented, the name of the person who performed the editing, and the date and time of editing.

[0031] FIG. 3B is a diagram showing an example of the failure mode effect analysis ontology 302 stored in the linking information storage unit 111. FIG. 3C is a diagram showing an example of linking information 303 stored in the linking information storage unit 111. The failure mode effect analysis ontology 302 describes the information items stored in the linking information 303 and the possible relationships between the items. In this embodiment, four types of items, namely, a part 3021, a failure mode 3022, an effect 3023, and a cause 3024, are stored by the failure mode effect analysis. In this embodiment, a parent-child relationship between multiple parts 3021 can be described. In this embodiment, a failure mode 3022 can be described in association with the part 3021, and an effect 3023 and a cause 3024 can be described in association with the failure mode 3022. A causal relationship can also be set between the effect 3023 and the cause 3024.

[0032] In this embodiment, the above is the failure mode and effect analysis ontology 302, but other than this, functions may be linked to components, or functional failure information may be linked to functions. Furthermore, in the case of other analysis methods such as FTA (Fault Tree Analysis), other forms of ontology may be used, such as component names and failure events, or causal chains of failure events and logic circuit information. Furthermore, this ontology does not have to be described in the linking information holding unit 111; it may be defined in the format or query language in the saving processing unit 121, or in the processing executed by the generation data conversion unit 114 or the display data conversion unit 112.

[0033] The linking information 303 is information related to reliability. In this embodiment, data is held in a graph structure described by nodes and edges, but other formats such as JSON or a table structure may also be used. In the linking information 303 of FIG. 3C , among the nodes, component information is represented by a solid line, failure modes by fine dashed lines, effects by coarse dashed lines, and causes by coarse chain lines and hatching.

[0034] In the failure mode and effect analysis ontology 302, a battery module 3032 and a battery management system 3033 are described as child nodes related to the components of a battery system 3031. The child nodes related to the components of this battery module 3032 are a battery cell 3034 and a cooling system 3041. The child nodes related to the failure mode of the battery cell 3034 are a short circuit 3035, an overcurrent 3036, and high temperature operation 3037. The child nodes related to the cause of high temperature operation 3037 are a "fire due to thermal runaway" node 3039 and a "failure in the cooling system" node 3040.

[0035] Furthermore, the child node relating to the parts of the cooling system 3041 is the refrigerant pipe 3042. The child node relating to the failure mode of the refrigerant pipe 3042 is the crack 3043. The child node relating to the cause of the crack 3043 is the aging deterioration 3045. The child node relating to the effect of the crack 3043 is the reduced flow rate 3044. The child node relating to the cause of the reduced flow rate 3044 is the "fault in the cooling system" node 3040.

[0036] For example, a battery module, a battery management system, etc. are described as subcomponents of a battery system according to the failure mode and effect analysis ontology 302. Furthermore, short circuit, overcurrent, and high temperature operation are described as failure modes of a battery cell. Regarding the high temperature operation failure mode, a fire caused by thermal runaway is described as an effect, and a malfunction of the cooling system is described as a cause.

[0037] 4A is a diagram showing property information 4034 of the battery cell 3034. The property information 4034 of the battery cell 3034 is stored in the association information storage unit 111. The property information 4034 of the battery cell 3034 includes a component name row, a component ID row that uniquely identifies the node, and a detailed description row that describes the node.

[0038] 4B is a diagram showing property information 4036 of the overcurrent 3036. The property information 4036 of the overcurrent 3036 is stored in the association information storage unit 111. The property information 4036 of the overcurrent 3036 includes a failure mode name row, a failure mode ID row that uniquely identifies the node, a detailed description row that explains the node, and an occurrence frequency row. The occurrence frequency row stores a multi-level evaluation of the occurrence frequency.

[0039] 4C is a diagram showing property information 4039 of the "fire caused by thermal runaway" node 3039. The property information 4039 of the "fire caused by thermal runaway" node 3039 is stored in the association information storage unit 111. The property information 4039 of the "fire caused by thermal runaway" node 3039 includes an impact name row, an impact ID row that uniquely identifies the node, a detailed description row that explains the node, an impact row, and an impact type row. The impact row stores an evaluation of the magnitude of the impact. The impact type row stores an impact type such as safety, economics, or environmental impact.

[0040] FIG. 4D is a diagram showing property information 4040 of the "cooling system failure" node 3040. The property information 4040 of the "cooling system failure" node 3040 is stored in the association information storage unit 111. The property information 4040 of the "cooling system failure" node 3040 includes a cause name row, a cause ID row that uniquely identifies the node, a detailed description row that explains the node, a detectability row, and a cause type row. The detectability row stores the detectability and detection method. The cause type row stores information indicating whether the cause is internal or external. In addition, a consideration information flag, which will be output later, may be stored as a property.

[0041] 1 , the explanation will be continued. The display data converter 112 converts data between the format displayed in the failure mode and effect analysis display and editing unit 106 and the format of the linking information 303 stored in the linking information holding unit 111. For example, the display data converter 112 of this embodiment converts the linking information 303 held in a graph structure into a table format and displays it in the failure mode and effect analysis display and editing unit 106. The display data converter 112 then converts the data into a predetermined file format and uploads it to the server 130 indicated by an external URL. The display data converter 112 further calculates and outputs statistics to be output to the study information statistics output unit 108 based on the study information flags described below.

[0042] The display data converter 112 further updates the linking information 303 stored in the linking information storage unit 111 based on the content edited by the failure mode and effect analysis display editor 106 and on a file in a predefined format downloaded from the server 130 indicated by an external URL. Note that the display data converter 112 here is not limited to conversion between a graph structure and a table format, and may convert data between any formats as long as it can convert information between the failure mode and effect analysis display editor 106, the external server 130, and the linking information 303.

[0043] When the save button 109 in the input / output unit 101 is clicked, the save processing unit 121 saves the linking information 303 in the linking information holding unit 111 in the failure mode and effect analysis database 110. The failure mode and effect analysis database 110 of this embodiment is assumed to be a graph database. The save processing unit 121 stores the linking information 303 in the failure mode and effect analysis database 110, which is a graph database, using a query language that is set in advance. Note that the failure mode and effect analysis database 110 may also be a relational database. In the same way as when the user clicks the upload button 123, the save processing unit 121 may upload a CSV file or the like to the external server 130.

[0044] The generation data conversion unit 114 performs mutual conversion between the linking information 303 and the output information of the linking information generation unit 113, and mutual conversion between the linking information 303 and prompt data used by the linking information editing unit 116. The generation data conversion unit 114 converts the graph-structured linking information 303 into generation data 601 in JSON format, for example, as shown in FIG. 6A .

[0045] The generation data conversion unit 114 further converts the generation data in JSON format into the format of graph-structured linking information 303. Here, the format that the generation data conversion unit 114 can convert into the format of graph-structured linking information 303 is not limited to the JSON format, but may be any format that allows linking information to be written in a natural language format, such as Markdown Mermaid format. The operation of converting the generation data in JSON format into graph-structured linking information 303 will be described with reference to Figures 6A, 6B, 7A, 7B, 8A, and 8B.

[0046] 5 is a diagram showing details of the linkage information generating prompt database 122. The linkage information generating prompt database 122 stores input instructions 501 that can be associated with the instructions of the automatic creation instruction unit 105, as well as corresponding prompt statements 502 and input information lists 503.

[0047] The prompt sentence 502 corresponding to the "prerequisite prompt" is shown below: "You are an expert in equipment reliability analysis. To conduct a Failure Mode and Effect Analysis (FMEA) for the target {top_component}, please follow the instructions below and submit the output. FMEA is a systematic method for identifying potential defects and failures in the target equipment and evaluating their impact. The purpose of FMEA is to identify risks and consider measures to reduce or eliminate them."

[0048] The prompt statement 502 corresponding to "Generate a parts tree" is shown below: "Please write the parts explosion table with top_component at the top in the following format. Please list the parts taking into consideration the hierarchical relationships between the parts. Parts refer to the individual elements and components that make up a product. Please write in the following format: {format_instructions}"

[0049] When the user clicks the "Generate parts tree" button 201, the linking information generating unit 13 extracts the prompt statement 502 on the first line, which is the premise, and the corresponding prompt statement 502 on the second line.

[0050] The prompt sentence 502 corresponding to "Failure mode generation" is shown below: "The parts explosion table with {top_component} at the top level is shown below. Please list 5 to 10 failure modes related to {component} in this parts explosion in the following format: {format_instructions} A failure mode refers to a specific method or state in which equipment or parts no longer perform their expected functions. A failure mode describes how equipment will stop functioning, in other words, what types of failures may occur. The parts configuration is shown below: {components}"

[0051] When the user clicks the "Generate failure mode" button 202, the linking information generating unit 13 extracts the prompt statement 502 on the first line, which is the premise, and the corresponding prompt statement 502 on the third line.

[0052] The prompt sentence 502 corresponding to "Generate Impact" is shown below: "Below is the parts explosion table with {top_component} at the top level and the failure modes related to {components} in this parts explosion. Please list 5 to 10 possible impacts of each failure mode related to {component} in the following format: {format_instructions} Impact refers to the specific results or effects on the equipment or customers when a failure mode occurs. For example, consider reduced equipment functionality, impact on safety, reduced customer satisfaction, risk of violating laws and regulations, impact on the environment, economic loss, etc. Below is the parts configuration: {components} Below is a list of failure modes: {failure_modes}"

[0053] When the user clicks the "Generate Influence" button 203, the link information generating unit 13 extracts the prompt statement 502 on the first line, which is the premise, and the corresponding prompt statement 502 on the fourth line.

[0054] The prompt sentence 502 corresponding to "Generate Causes" is shown below: "Below is the parts explosion table with {top_component} at the top level and the failure modes related to {components} in this parts explosion. Please list 5 to 10 possible effects of each failure mode related to {component} in the following format. {format_instructions} A cause refers to the fundamental reason or factor that causes a specific failure mode to occur. For example, consider design flaws, manufacturing problems, usage conditions and methods, human factors such as maintenance and operating errors, and external environmental factors. Below is the parts configuration. {components} Below is a list of failure modes. {failure_modes}"

[0055] When the user clicks the "Generate Cause" button 204, the linking information generating unit 13 extracts the prompt statement 502 on the first line, which is the premise, and the corresponding prompt statement 502 on the fifth line.

[0056] The linking information generation unit 113 uses a prompt statement 502 extracted from the linking information generation prompt database 122. In this embodiment, the linking information generation prompt database 122 stores prompt statements corresponding to the "Generate component tree" button 201, the "Generate failure mode" button 202, the "Generate effect" button 203, and the "Generate cause" button 204, as well as prerequisite prompts that are prerequisites for all instructions.

[0057] Without being limited to this, the linking information generation prompt database 122 may prepare prompts for input instructions that are differentiated by case, such as when the target parts are limited. The prompt statement 502 includes variables that are replaced with specific information when used by the linking information generation unit 113, as defined in the input information list 503. Furthermore, the prompt statement 502 is not limited to the example described here, and may be any statement that can obtain the desired output.

[0058] Continuing the explanation, returning to Fig. 1, the linking information generation unit 113 creates generation data in accordance with a user instruction, and stores the data in the linking information 303 in the linking information storage unit 111 via the generation data conversion unit 114.

[0059] The linking information generation unit 113 of this embodiment creates generation data using a question-and-answer model 140 that uses a large-scale language model. Here, the linking information generation unit 113 creates a prompt sentence by retrieving a prompt sentence 502 from the linking information generation prompt database 122 based on a user instruction and substituting an input information list 503. The linking information generation unit 113 inputs the created prompt sentence into the question-and-answer model 140, thereby obtaining generation data in a specified format.

[0060] For example, the operation when the "Generate Part Tree" button 201 is clicked will be described with reference to FIGS. 5, 6A and 6B.

[0061] The linking information generation unit 113 creates a prompt sentence by substituting the character string "battery system", which is the component information input to the product name input unit 104, into the {top_component} portion of the prompt sentence 502 related to "generating a component tree" in FIG. 5. The linking information generation unit 113 inputs the created prompt sentence to the question and answer model 140. As a result, the question and answer model 140 outputs generation data 601 in a JSON (JavaScript Object Notation) format, for example, as shown in FIG. 6A. The linking information generation unit 113 outputs this generation data 601 in the JSON format to the generation data conversion unit 114.

[0062] The generation data conversion unit 114 converts the generation data 601 in JSON format and updates the linking information 303 .

[0063] 6A is a diagram showing generation data 601 in JSON format output by the question and answer model 140. The generation data 601 is in JSON format and indicates that the battery system has a battery module as a subcomponent, the battery module has a battery cell and a cooling system as subcomponents, and the cooling system has a refrigerant pipe as a subcomponent.

[0064] The root node of the generation data 601 has a "battery system" value set for the "component_name" key, and a "system for storing power using a battery" value set for the "description" key. This node also has multiple child nodes set as an array for the "subcomponents" key.

[0065] The child node of the battery system has a "battery module" value set for the "component_name" key, and a "module consisting of multiple battery cells connected in series" value set for the "description" key. This node also has child nodes set as an array for the "subcomponents" key.

[0066] The first child node of the battery module has a "component_name" key with a value of "battery cell" and a "description" key with a value of "minimum replaceable unit of lithium-ion battery." This node also has a "subcomponents" key, which is set to a null array.

[0067] The second child node of the battery module has a "cooling system" value set for the "component_name" key and a "cooling mechanism for managing the temperature of the battery cells" value set for the "description" key. This node also has child nodes set as an array for the "subcomponents" key.

[0068] The child node of the cooling system has a "refrigerant piping" value set for the "component_name" key, and a "piping for circulating refrigerant" value set for the "description" key. This node also has child nodes set as an array for the "subcomponents" key.

[0069] The generation data conversion unit 114 converts this generation data 601 to graph structured linking information 303 and stores it in the linking information holding unit 111 .

[0070] 6B is a diagram showing linking information 303 obtained by converting generation data 601 into a graph structure. In the graph structure of linking information 303, a battery module 3032 is described as a child node related to a component of a battery system 3031. The child nodes related to the components of this battery module 3032 are a battery cell 3034 and a cooling system 3041. The child node related to the component of the cooling system 3041 is a refrigerant pipe 3042.

[0071] The processing performed when the "Generate Failure Mode" button 202 is clicked will be described with reference to FIGS. 5, 7A, and 7B. The linking information generation unit 113 creates a prompt statement by substituting "battery system" for {top_component} in the prompt statement 502 for "generate failure mode" and substituting the names of each component for {component}. For {components}, the data-for-generation conversion unit 114 converts the already created linking information 303 related to the component structure into JSON format. Of the data in this JSON format, the string "battery cell" indicating a child component is substituted. The linking information generation unit 113 inputs the created prompt statement into the question-and-answer model 140. As a result, the question-and-answer model outputs generation data 701 shown in FIG. 7A.

[0072] 7A is a diagram showing generation data 701 in JSON format output by the question-and-answer model 140. The generation data 701 is in JSON format, and three child nodes, namely, overcurrent, short circuit, and high temperature operation, are listed as failure modes of the battery cell. In the root node of the generation data 701, the value "battery cell" is set for the key "component_name". This node further has multiple child nodes set as an array for the key "failure_nodes".

[0073] In the first child node of the battery cell, the value "overcurrent" is set for the key "failure_mode_name", and the value "state in which a current greater than the rated current value flows through the battery cell" is set for the key "description".

[0074] In the second child node of the battery cell, the value "short circuit" is set for the key "failure_mode_name", and the value "the battery cell connection is short circuited, causing current to flow through an unintended path" is set for the key "description".

[0075] The third child node of the battery cell has a value of "high temperature operation" set for the key "failure_mode_name" and a value of "operated at temperatures exceeding the range specified in the specifications" set for the key "description".

[0076] This generation data 701 is output to the generation data conversion unit 114. The generation data conversion unit 114 converts this generation data 701 into the graph structure shown in FIG.

[0077] 7B is a diagram showing an example of linking information 303 obtained by converting generation data 701 into a graph structure. The linking information 303 in Fig. 7B is obtained by converting generation data 701 into a graph structure and updating the linking information 303 in Fig. 6B. In the example of Fig. 7B, information about short circuit 3035, overcurrent 3036, and high temperature operation 3037 has been added and updated as child nodes that are failure modes of battery cell 3034.

[0078] 5, 8A, and 8B, the processing performed when the "Create Impact" button 203 is clicked. When the user clicks the "Create Impact" button 203, the linked information generator 113 creates a prompt statement by substituting the input information list for the prompt statement 502 related to "Create Impact." Specifically, the linked information generator 113 substitutes "battery system" for {top_component}, a character string specifying the data format for {format_instructions}, the name of each component for {component}, the component configuration for {components}, and the failure mode name for {failure_modes}.

[0079] The linking information generation unit 113 inputs the prompt sentence thus created to the question and answer model 140. The question and answer model 140 outputs generation data 801 based on the prompt sentence. The generation data conversion unit 114 converts this generation data 801 into linking information 303.

[0080] 8A is a diagram showing generation data 801 in JSON format output by the question-and-answer model 140. In the root node of the generation data 801, the value "high temperature operation" is set for the key "failure_mode_name", and the value "operated at a temperature exceeding the range defined in the specifications" is set for the key "description". This root node also has the value "fire due to thermal runaway" set for the key "effects" as a single array.

[0081] Fig. 8B is a diagram showing an example of linking information 303 added by converting generation data 801 into a graph structure. Linking information 303 in Fig. 8B is obtained by converting generation data 801 into a graph structure and updating linking information 303 in Fig. 7B. Linking information 303 in Fig. 8B has been updated by adding a "fire due to thermal runaway" node 3039 as a cause of high temperature operation 3037.

[0082] The linking information generating unit 113 may automatically generate the linking information 303 without using the consideration information described in the consideration information input unit 103 .

[0083] Returning to Fig. 1, the explanation will be continued. The target item extraction unit 115 extracts one or more target items that are of particular interest based on the character string "design study related to fire prevention" written in the study information input unit 103, and extracts the requirements that these target items must satisfy. In this embodiment, the question and answer model 140 using a large-scale language model is used to extract the target items on the failure mode and effect analysis ontology 302 and the requirements that these target items must satisfy. Specific examples are shown in Figs. 9A to 9C.

[0084] 9A is a diagram showing the failure mode effect analysis ontology 302 in the linking information storage unit 111. The failure mode effect analysis ontology 302 describes the information items stored in the linking information 303 and the possible relationships between the items. In this embodiment, four types of items are stored by the failure mode effect analysis: part 3021, failure mode 3022, effect 3023, and cause 3024. A parent-child relationship between multiple parts 3021 can be described. A failure mode 3022 can be described linked to the part 3021. An effect 3023 and a cause 3024 can be described linked to the failure mode 3022. A causal relationship can also be set between the effect 3023 and the cause 3024.

[0085] As described above, in this embodiment, the failure mode and effect analysis ontology 302 describes components, failure modes, effects, causes, and the possible relationships between them. However, the failure mode and effect analysis ontology 302 is not limited to this, and it is sufficient if it is information that is handled regarding reliability. From the perspective of the consideration information in the consideration information input unit 103, the target item extraction unit 115 extracts the most noteworthy items from the failure mode and effect analysis ontology 302 among components, failure modes, effects, and causes, and sets them as target items 903.

[0086] FIG. 9B is a diagram showing a prompt sentence 902. Prompt sentence 902 reads as follows: "You are an expert in equipment reliability analysis. To conduct a Failure Mode and Effect Analysis (FMEA) for the target {top_component}, please submit the output according to the instructions below. FMEA is a systematic method for identifying potential defects and failures in the target equipment and evaluating their impact. The purpose of FMEA is to identify risks and consider measures to reduce or eliminate them. When the following consideration information is the purpose of conducting an FMEA, which FMEA item should we focus on? Also, what requirements should the target item satisfy? Consideration item: {purpose_of_consideration} FMEA item candidate: {item_candidate} Output format: {format_instructions}"

[0087] The prompt statement 902 can be substituted by input information, similar to the prompt statement 502 stored in the prompt database 122 for generating linking information. The target item extraction unit 115 substitutes "battery system" for {top_component} of the prompt statement 902, "fire prevention design consideration" for {purpose_of_consideration}, and "component, failure mode, impact, cause" for {item_candidate}, and inputs the result to the question-and-answer model 140 using a large-scale language model. Based on this prompt statement, the question-and-answer model 140 outputs "impact" as the target item 903 and "N or more fire-related impacts are described for each failure mode" as the requirement 904, as shown in FIG. 9C . This requirement 904 is information related to the consideration item obtained from the "impact" information.

[0088] 9C is a diagram showing a target item 903 and a requirement 904 of the JSON format data output by the target item extraction unit 115 using the question and answer model 140. The value of "impact" is set in the target item 903 shown in FIG. 9C, and the value of "N or more fire-related impacts are described for each failure mode" is set in the requirement 904.

[0089] The question and answer model 140 outputs a target item 903 and a requirement 904 from the study information of "design study related to fire prevention" in the study information input unit 103, the prompt sentence 902, and the failure mode and effect analysis ontology 302. The requirement 904 is a character string describing what requirement the target item 903 should satisfy in terms of the study information in the study information input unit 103. The study information statistics output unit 108 may use the target item 903 and the requirement 904 as criteria for calculating statistics.

[0090] 1 , the linking information editing unit 116 edits the linking information 303 based on the consideration information from the consideration information input unit 103 so as to conform to the target items 903 and requirements 904 output by the target item extraction unit 115. The linking information editing unit 116 includes a first related information generation unit 117, an editing procedure database 118, a second related information generation unit 119, and a combination unit 120.

[0091] The first related information generating unit 117 generates specific content of the target item 903 related to the requirement 904. Specific examples are shown in Figures 10A to 10C.

[0092] 10A is a diagram showing a target item 903 and a requirement 904 of JSON format data output by the target item extraction unit 115 using the question and answer model 140. The value of "impact" is set in the target item 903 shown in Fig. 10A, and the value of "N or more fire-related impacts are described for each failure mode" is set in the requirement 904.

[0093] 10B is a diagram showing an example of a prompt sentence 1001 used by the first related information generation unit 117. The prompt sentence 1001 is written as follows: "You are an expert in equipment reliability analysis. To conduct a Failure Mode and Effect Analysis (FMEA) of the target {top_component}, please submit the output according to the instructions shown below. FMEA is a systematic method for identifying potential defects and failures in the target equipment and evaluating their impact. The purpose of FMEA is to identify risks and consider measures to reduce or eliminate them. When the following study information is the purpose of conducting an FMEA, please list and output multiple items for {item} that are related to the following requirements. Requirement: {requirement} Output format: {format_instructions}"

[0094] The first related information generation unit 117 substitutes the requirement 904, "For each failure mode, N or more fire-related impacts are described," for {requirement} of the prompt sentence 1001, and substitutes "impact" of the target item 903 for {item}, and inputs the result to the question-and-answer model 140 using a large-scale language model. The question-and-answer model 140 outputs the first related information 1002 shown in FIG. 10C .

[0095] 10C is a diagram showing first related information 1002 obtained by the first related information generation unit 117. The first related information 1002 is configured to include "fire, explosion, and arc occurrence due to thermal runaway." The first related information 1002 is the result of a study on "design study on fire prevention" entered in the study information input unit 103 regarding the "impact" of the target item.

[0096] 11 is a diagram showing the editing procedure database 118. The editing procedure database 118 stores target items 1101, editing procedures 1102 to be performed by the second related information generation unit 119 or the combination unit 120 for the target items 1101, and prompt statements 1103 for each procedure. The editing procedure database 118 describes the procedure for editing the linking information 303 for each target item 903. The linking information editing unit 116 sequentially executes the editing procedures 1102 for the target items 1101 that match the target items 903. The editing procedure database 118 may also include information on functions to be called and arguments to be input.

[0097] The editing procedure database 118 stores the operation procedures of the second related information generation unit 119 and the operation procedures of the combination unit 120 after the first related information generation unit 117 acquires the first related information 1002. The second related information generation unit 119 generates other items in the related failure mode and effect analysis ontology 302 based on the first related information 1002 generated by the first related information generation unit 117. The combination unit 120 integrates the first related information 1002, the second related information 1302, and the linking information 303 to create new linking information 303.

[0098] An example of an editing procedure when the target item 903 is "impact" is as follows: The combination unit 120 identifies, according to step #1 of "impact" of the target item 1101 in the editing procedure database 118, a failure mode stored in the linking information storage unit 111 that causes the impact of the first associated information 1002. The second associated information generation unit 119 outputs, according to step #2 of "impact" of the target item 1101 to the editing procedure database 118, a failure mode that causes the impact of the first associated information 1002 on the target component, to generate second associated information 1302. The combination unit 120 generates combination information by combining the second associated information 1302 with the linked information 303 edited in step #1, according to step #3 of "impact" of the target item 1101 in the editing procedure database 118.

[0099] These procedures may be manually designed in advance, or may be automatically created from a conventional reliability information creation procedure or information in the failure mode and effect analysis ontology 302 .

[0100] 12A to 12C, the operation of procedure #1 when the target item 903 is "impact" will be described. Fig. 12A is a diagram illustrating the linking information 303 related to procedure #1. The linking information 303 in Fig. 12A is written in a table format and is configured to include a part column, a failure mode column, and an impact column.

[0101] 12B is a diagram illustrating a prompt sentence 1201 for procedure #1. This prompt sentence 1201 is the prompt sentence 1103 for procedure #1 of the “impact” of the target item 1101 in the editing procedure database 118.

[0102] Here, editing is performed using the combination unit 120. The combination unit 120 creates a prompt sentence by substituting information obtained by converting the information of the first related information 1002 and the linking information 303 by the generation data conversion unit 114 into a prompt sentence 1201.

[0103] Note that although the linking information 303 is described in table form in Fig. 12C, it is equivalent to the graph structure described in Fig. 3. The combination unit 120 generates a prompt sentence by substituting the first related information 1002 for {effects_from_requirement} of the prompt sentence 1201, and substituting the linking information 303 converted by the generation data conversion unit 114 for {FMEA_data}. The combination unit 120 then inputs the generated prompt sentence into the question-and-answer model 140 to obtain updated linking information 1202.

[0104] The data in the JSON format output by the question and answer model 140 is converted by the generation data conversion unit 114 into linking information 1202, which has the same data format as the linking information 303. Here, as shown in a consideration information flag column 1204, a consideration information flag is linked to an output related to the consideration information of the consideration information input unit 103, and the flag is held as property information.

[0105] 12C is a diagram illustrating the linking information 1202 related to procedure #1. The linking information 1202 in FIG. 12C is written in a table format, and is configured to include a component column, a failure mode column, an effect column, and a consideration information flag column 1204.

[0106] Here, by inputting the first related information 1002 as an effect related to the fire, the effect of "arc occurrence" is newly output for the failure mode "short circuit." In addition, effects originally included in the linked information 303 that are related to the consideration information in the consideration information input unit 103 are also assigned a consideration information flag.

[0107] <<Procedure #2>> Next, the operation of the second related information generation unit 119 in procedure #2 will be described with reference to FIGS. 13A and 13B.

[0108] 13A is a diagram showing a prompt sentence 1301 for step #2. The prompt sentence 1301 is written as follows: "When the following are considered as effects of the FMEA of {component} in {top_component}, please output the failure modes that cause them. {effects_from_requirement} {format_instructions}"

[0109] The second related information generating unit 119 obtains second related information 1302 by inputting a prompt sentence 1301 into the question and answer model 140 based on the first related information 1002 and the target part name.

[0110] 13B is a diagram showing second related information 1302 related to procedure #2. The second related information 1302 is written in a table format and is configured to include an impact column, a failure mode column, and a component column. The impact column lists "fire due to thermal runaway," "explosion," and "arc generation." The impact corresponds to "fire, explosion, or arc generation due to thermal runaway" listed in the first related information 1002.

[0111] The failure mode column lists "overcurrent," "high temperature operation," and "short circuit," which correspond to the effects of "fire due to thermal runaway," "flammable gas release," which corresponds to the effects of "explosion," and "short circuit" and "poor terminal connection," which correspond to the effects of "arc generation." "Battery cell" is listed as the component that corresponds to these failure modes: "overcurrent," "high temperature operation," "short circuit," "flammable gas release," "short circuit," and "poor terminal connection."

[0112] Specifically, when "fire, explosion, or arc generation due to thermal runaway" is input as first related information 1002, the second related information generation unit 119 outputs possible failure modes of the target component, the battery cell, as second related information 1302.

[0113] <<Procedure #3>> The operation of the combination unit 120 in procedure #3 will be described with reference to FIGS. 14A and 14B.

[0114] 14A is a diagram showing a prompt 1401 for step #3. The prompt 1401 is written as follows: "For the FMEA of {component} in {top_component}, please output the information from the following two FMEAs in a format that makes it clear which of the effects relate to {effect_from_requirement}. {FMEA_data_#1} {FMEA_data_#2}"

[0115] The combination unit 120 receives as input the first association information 1002 and the second association information 1302, and replaces the prompt sentence 1401 with the linking information 1202 updated in step #1 converted by the data for generation conversion unit 114. The combination unit 120 then inputs the generated prompt sentence to the question-and-answer model 140 to obtain data in JSON format. The combination unit 120 outputs the JSON format data generated by the question-and-answer model 140 to the data for generation conversion unit 114. The data for generation conversion unit 114 combines the linking information with the JSON format data output by the combination unit 120 to generate combination information 1402 and stores it in the linking information storage unit 111.

[0116] 14B is a diagram showing combination information 1402 generated in step #3. The combination information 1402 includes a component column, a failure mode column, an effect column, and a review information flag column. The component column lists a battery cell. The failure modes associated with this battery cell include overcurrent, overcharge, overdischarge, high temperature, low temperature, short circuit, flammable gas release, and terminal connection failure.

[0117] The effect of overcurrent is "fire due to thermal runaway" and its statistical information flag is 1. The effect of overcharge is "deterioration due to electrolyte electrolysis" and its statistical information flag is not set. The effect of overdischarge is "early deterioration due to lithium deposition" and its statistical information flag is not set. The effect of high temperature is "fire due to thermal runaway" and its statistical information flag is 1.

[0118] The low temperature effect includes "early deterioration due to lithium deposition" and its statistical information flag is not set. The short circuit effect includes "fire due to thermal runaway" and "arc generation" and its statistical information flag is 1. The flammable gas release effect includes "explosion" and its statistical information flag is 1. The terminal connection failure effect includes "arc generation" and its statistical information flag is 1.

[0119] 15 is a flowchart of the process of this embodiment. First, the user determines whether or not the external server 130 has reliability information data (step S10). If the external server 130 does not have reliability information data (No), the process proceeds to step S12. If the external server 130 has reliability information data (Yes), the user clicks the download button 107 to download the data (step S11), and the process proceeds to step S12.

[0120] Thereafter, the user inputs information into the product name input section 104 and the consideration information input section 103 in the input / output section 101 (step S12). Then, the target item extraction section 115 extracts target items 903 of interest and their requirements 904 in accordance with the input into the consideration information input section 103 (step S13).

[0121] The user instructs the required automatic processing by clicking one of the buttons in the automatic creation instruction unit 105 (step S14). As a result, the linking information generation unit 113 creates linking information 303 and stores it in the linking information storage unit 111 (step S15). It is determined whether the target item of the automatic processing specified by the user matches the target item 903 extracted by the target item extraction unit 115 (step S16).

[0122] If the target item for automatic processing differs from the target item 903 extracted by the target item extraction unit 115 (No), the process proceeds to step S18. If the target item for automatic processing matches the target item 903 extracted by the target item extraction unit 115 (Yes), the linking information editing unit 116 updates the linking information taking into account the consideration information in the consideration information input unit 103 (step S17). For example, if the target item 903 is "impact" and the "generate impact" button 203 is clicked, the linking information editing unit 116 performs processing.

[0123] The first related information generation unit 117 generates first related information 1002, which is a description of a target item related to the consideration information. The second related information generation unit 119 identifies other items other than the target item based on the first related information 1002, and generates second related information 1302 that associates the identified other items with the first related information 1002. The combination unit 120 combines the second related information 1302 with the linking information 303 in the linking information holding unit 111. The combination unit 120 combines the second related information 1302 with the linking information 303 in the linking information holding unit 111, and the linking information editing unit 116 generates combined information.

[0124] The linking information 303 edited by the linking information generator 113 and the linking information editor 116 is displayed in the failure mode and effect analysis display editor 106 (step S18). That is, the combination information obtained by combining the second related information 1302 with the linking information 303 is displayed in the failure mode and effect analysis display editor 106. Furthermore, if the display data converter 112 outputs the study information statistics, the study information statistics output unit 108 outputs them. The user manually edits the data, and the linking information 303 in the linking information storage unit 111 is updated accordingly (step S19). The failure mode and effect analysis creation support unit 102 then determines whether or not any implementation details from the automatic creation instruction unit 105 remain (step S20). If any implementation details from the automatic creation instruction unit 105 remain, the process returns to step S14, and automatic creation is performed again. When all of the implementation contents in the automatic generation instruction unit 105 are completed, the process proceeds to step S20, and the saving processing unit 121 saves the linking information 303 in the failure mode and effect analysis database 110.

[0125] 16 is a block diagram of the information processing device 100. The information processing device 100 includes a CPU (Central Processing Unit) 91, a ROM (Read Only Memory) 92, and a RAM (Random Access Memory) 93. The CPU 91 executes a program (not shown) to realize the functions of the information processing device 100. The ROM 92 is a non-volatile memory and stores, for example, a BIOS (Basic I / O System). The RAM 93 is a volatile memory and stores data temporarily held by the program executed by the CPU 91.

[0126] The information processing device 100 further includes a storage unit 94, a display unit 95, an input unit 96, and a communication interface 97, which are interconnected by a bus 98. The storage unit 94 is a large-capacity storage unit such as a hard disk or SSD (Solid State Drive), and stores programs, data, and the like.

[0127] The display unit 95 is, for example, a liquid crystal display, and displays characters, figures, images, etc. The display unit 95 may be detachable. The input unit 96 is, for example, a transparent touch panel superimposed on the liquid crystal display, and accepts user input operations. The communication interface 97 communicates with, for example, a cloud server. The communication interface 97 may support any communication format, such as a wired NIC (Network Interface Card), a wireless NIC, or a fifth-generation communication module. This completes the description of the embodiment.

[0128] [1] A study information input unit (103) for inputting study information about a product; a target item extraction unit (115) for extracting, based on the study information, one or more of a part constituting the product, an event occurring in the part, a cause of the event, and an impact of the event occurrence as a target item, and extracting requirements that the target item should satisfy; a linking information storage unit (111) for storing linking information linking any two or more items including the target item among the part, the event, the cause, and the impact; and a linking information editing unit (116) for generating first linking information that is a description of the target item based on the target item and the requirements, specifying other items other than the target item based on the first linking information (1002), generating second linking information (1302) that associates the specified other items with the first linking information (1002), and generating combination information that combines the second linking information (1302) with the linking information. an output unit (failure mode and effect analysis display and editing unit 106) that outputs the combination information generated by the linkage information editing unit (116).

[0129] This makes it possible to improve the comprehensiveness of reliability information.

[0130] [2] The information processing device (100) according to claim 1, characterized in that the linking information editing unit (116) is provided with a first related information generation unit (117) that generates first related information (1002), which is a description of the target item, based on the target item and the requirements.

[0131] This makes it possible to generate first related information that describes the target item and includes synonyms of the considered information.

[0132] [3] The information processing device (100) according to claim 2, characterized in that the linked information editing unit (116) is provided with a second related information generation unit (119) that identifies other items other than the target item based on the first related information (1002) and generates second related information (1302) that associates the identified other items with the first related information (1002).

[0133] This makes it possible to generate second related information that associates the identified other items with first related information that is related to the consideration information.

[0134] [4] The information processing device (100) according to claim 3, characterized in that the linked information editing unit (116) includes a combination unit (120) that generates combined information by combining the second related information (1302) with the linked information.

[0135] This allows the second related information to be linked to the linking information, making it possible to comprehensively grasp the relationship between the review information and each item related to it.

[0136] [5] The information processing device (100) according to claim 1, further comprising a linking information generation unit (113) that generates linking information without using the review information, and the linking information storage unit (111) stores the linking information generated by the linking information generation unit (113).

[0137] This allows comprehensive generation of linking information, and therefore comprehensive generation of combination information based on this linking information.

[0138] [6] The information processing device (100) according to claim 1, further comprising an editing procedure database (118) that records the operation procedures of the linking information editing unit (116).

[0139] This makes it possible to appropriately edit the linking information and generate combination information.

[0140] [7] The information processing device (100) according to claim 6, wherein the operation procedure of the linked information editing unit (116) also records the relationship with the target item.

[0141] This makes it possible to easily develop a method for combining the linking information, the first related information, and the second related information to generate combined information.

[0142] [8] The information processing device (100) according to claim 1, wherein the linked information editing unit (116) assigns a consideration information flag to information related to the consideration information among the combination information.

[0143] This makes it possible to easily extract each piece of information related to the review information.

[0144] [9] The information processing device (100) according to claim 8, further comprising: a consideration information statistics output unit (108) that outputs statistics using the consideration information flag.

[0145] This allows the user to easily visually recognize the statistics of information related to the considered information.

[0146]

[10] The information processing device (100) according to claim 1, characterized in that the target item extraction unit (115) uses the review information to extract, as target items, impacts of events occurring in components that constitute the product, and extracts requirements that the target items must satisfy.

[0147] This allows the user to easily understand the requirements that will be affected by the study items.

[0148]

[11] A step in which a study information input unit (103) accepts input of study information for a product; a step in which a target item extraction unit (115) extracts, based on the study information, one or more of a part constituting the product, an event occurring in the part, a cause of the event, and an impact of the occurrence of the event as target items, and extracts requirements to be satisfied by the target items; a step in which a linking information holding unit (111) holds linking information linking any two or more of the part, the event, the cause, and the impact, including the target item; a step in which a linking information editing unit (116) generates first related information that is a description of the target item, based on the target item and the requirements; a step in which the linking information editing unit (116) identifies other items other than the target item based on the first related information (1002), and generates second related information (1302) that associates the identified other items with the first related information (1002); an output unit (failure mode and effect analysis display and editing unit 106) outputting the combination information generated by the linking information editing unit (116).

[0149] This makes it possible to improve the comprehensiveness of reliability information.

[0150] (Modifications) The present invention is not limited to the above-described embodiments and includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those including all of the described configurations. It is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is also possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0151] The above-described configurations, functions, processing units, processing means, etc. may be implemented in part or in whole by hardware such as an integrated circuit. The above-described configurations, functions, etc. may also be implemented by software, with a processor interpreting and executing a program that implements each function. Information such as the program, table, and file that implements each function can be stored in a recording device such as a memory, a hard disk, or an SSD (Solid State Drive), or on a recording medium such as a flash memory card or a DVD (Digital Versatile Disk).

[0152] In each embodiment, the control lines and information lines shown are those considered necessary for explanation, and not all control lines and information lines in the product are necessarily shown. In reality, it can be considered that almost all components are interconnected. Examples of modified examples of the present invention include the following (a) to (c).

[0153] (a) The generation data may not be output using a question-and-answer model that uses a large-scale language model, but may be output using, for example, rule-based natural language processing. (b) The target item extraction unit may select, as the target item, one or more of a part, an event that occurs in the part, a cause of the event, and an impact of the event, but is not limited thereto. (c) The linking information storage unit may store linking information that links two or more items, including the target item extracted by the target item extraction unit, among a part, an event, a cause, and an impact.

[0154] 100 Information processing device 101 Input / output unit 102 Failure mode and effect analysis creation support unit 103 Study information input unit 104 Product name input unit 105 Automatic creation instruction unit 106 Failure mode and effect analysis display and editing unit 107 Download button 108 Study information statistics output unit 109 Save button 110 Failure mode and effect analysis database 111 Linking information storage unit 112 Display data conversion unit 113 Linking information generation unit 114 Generation data conversion unit 115 Target item extraction unit 116 Linking information editing unit 117 First related information generation unit 118 Editing procedure database 119 Second related information generation unit 120 Combination unit 140 Question and answer model

Claims

1. An information processing device comprising: a review information input unit that inputs review information about a product; a target item extraction unit that extracts, based on the review information, one or more of the parts that make up the product, events that occur in the parts, causes of the events, and effects of the events as target items, and extracts requirements that the target items must satisfy; a linking information storage unit that stores linking information that links two or more items, including the target items, from the parts, the events, the causes, and the effects; a linking information editing unit that generates first related information that is a description of the target items based on the target items and the requirements, identifies items other than the target items based on the first related information, generates second related information that links the identified items with the first related information, and generates combination information that combines the second related information with the linking information; and an output unit that outputs the combination information generated by the linking information editing unit.

2. The information processing device according to claim 1, characterized in that the linking information editing unit is provided with a first related information generation unit that generates first related information, which is a description of the target item, based on the target item and the requirements.

3. The information processing device described in claim 2, characterized in that the linked information editing unit is provided with a second related information generation unit that identifies other items other than the target item based on the first related information and generates second related information that associates the identified other items with the first related information.

4. The information processing device according to claim 3, characterized in that the linking information editing unit includes a combination unit that generates combination information by combining the second related information with the linking information.

5. The information processing device according to claim 1, further comprising a linking information generation unit that generates linking information without using the review information, and wherein the linking information storage unit stores the linking information generated by the linking information generation unit.

6. The information processing device according to claim 1, further comprising an editing procedure database that records the operation procedures of the linked information editing unit.

7. The information processing device according to claim 6, wherein the operating procedure of the linked information editing unit also records the relationship between the target items.

8. The information processing device according to claim 1, wherein the linked information editing unit assigns a review information flag to information related to the review information among the combination information.

9. The information processing device according to claim 8, further comprising: a review information statistics output unit that outputs statistics using the review information flag.

10. The information processing device according to claim 1, characterized in that the target item extraction unit uses the review information to extract the impact of events occurring in the components that make up the product as target items, and extracts the requirements that the target items must satisfy.

11. A step in which an examination information input unit accepts input of examination information for a product; a step in which a target item extraction unit extracts, based on the examination information, one or more of the parts constituting the product, the events occurring in the parts, the causes of the events, and the effects of the occurrence of the events as target items, and extracts requirements that the target items must satisfy; a step in which a linking information holding unit holds linking information linking any two or more of the parts, the events, the causes, and the effects, including the target items; a step in which a linking information editing unit generates first related information that is a description of the target items based on the target items and the requirements; a step in which the linking information editing unit identifies other items other than the target items based on the first related information, and generates second related information that associates the identified other items with the first related information; a step in which the linking information editing unit generates combination information by combining the second related information with the linking information; and a step in which an output unit outputs the combination information generated by the linking information editing unit. An information processing method comprising:

Images