Support device, program, and method

Abstract

To improve a technology for preventing omission in extraction of hazards or risks in risk assessment.SOLUTION: A support device 100 of the present disclosure is a support device 100 for supporting risk assessment of a production system, comprising: a knowledge graph generation unit 103 which generates a knowledge graph including equipment information, hazard information, and work information; a knowledge graph search unit 110 which generates search queries based on the work information during operation, performs searches on the knowledge graph based on the search queries, and obtains hazards or risks related to the search queries as search results; and a risk assessment list generation unit 111 which generates a risk assessment list in which the work information during operation and the search results are associated each other.SELECTED DRAWING: Figure 8

Description

[Technical Field]

[0001] The present disclosure relates to an assistance device, a program, and a method. [Background technology]

[0002] As a conventional technique aimed at preventing omission of hazards when extracting hazards in risk assessment of production facilities, for example, the technique disclosed in Patent Document 1 has been proposed. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2021-002143 Summary of the Invention [Problem to be solved by the invention]

[0004] In the prior art, risks may not be extracted due to synonyms, spelling variations, abbreviations, language, etc. For example, if "robot" is registered as a hazard (hereinafter also referred to as "hazard"), risks may not be extracted due to some expressions such as android, robot, robot man, bot, automatic welding machine R5, AC-100 (robot equipment number or model number), or painting arm.

[0005] Furthermore, in the prior art, if a hazard exists in addition to the hazards registered for a component or related part in a hazard list, risks due to the other hazards may not be identified. For example, for a robot, if only mechanical hazards are registered in the hazard list, it is possible to identify the risk of being crushed. However, if an electrical hazard exists in the robot's power supply, the risk of electric shock during inspection may not be identified. Furthermore, if a chemical hazard exists in the robot's painting arm, the risk of poisoning from paint solvents during inspection may not be identified.

[0006] Furthermore, in the prior art, when the status of components in the hazard list changes, risks associated with the latest status may not be identified. For example, changes to the layout of a facility such as a plant, equipment upgrades, changes in status or use, changes in raw materials, or part replacements may affect the attributes of hazards (type, effect, cause, location, etc.), which may result in risks not being identified. For example, for raw material pipe X, flammable gas passes through it during the production of product A, so the hazard type is "flammable," the effect is "fire or explosion," and the cause is "spark." Furthermore, toxic gas passes through it during the production of product B, so the hazard type is "chemical," the effect is "poisoning," and the cause is "inhalation." Furthermore, high-temperature steam passes through it during maintenance inspections, so the hazard type is "high temperature," the effect is "burns," and the cause is "contact with the human body." However, in the prior art, even if there is some change in the status of a hazard after the hazards are input at the time of initial setup, the hazard list is not updated unless a new hazard is input again. Therefore, if the type, cause, or effect of a hazard changes, as in the case of raw material pipe X, risks may not be extracted.

[0007] Furthermore, in conventional technology, even when an operation involves the same hazard, the presence or absence of risk varies depending on the worker's operational experience or operating qualifications. However, this difference in the presence or absence of risk is not taken into consideration, which can lead to missed risks. For example, for a "robot" in the hazard list, the risk level for a worker with operating qualifications is "low," while the risk level for a worker without operating qualifications is "high."

[0008] Therefore, an object of the present disclosure is to improve the technology for preventing hazards or risks from being overlooked in risk assessment. [Means for solving the problem]

[0009] In some embodiments, the assistance device comprises: A support device for supporting risk assessment of a production system, comprising: a knowledge graph generation unit that generates a knowledge graph including facility information, hazard information, and work information; a knowledge graph search unit that generates a search query based on work information during operation, searches the knowledge graph based on the search query, and acquires hazards or risks related to the search query as search results; a risk assessment list generation unit that generates a risk assessment list in which the work information during the operation is associated with the search results; Equipped with.

[0010] In this way, hazards or risks are extracted by using a knowledge graph including facility information, hazard information, and work information, thereby improving the technology for preventing omission of hazards or risks in risk assessment.

[0011] In one embodiment, the knowledge graph generator further generates a knowledge graph including human information.

[0012] In this way, hazards or risks can be extracted by further using a knowledge graph including human information, thereby improving the technique for preventing omission of hazards or risks in risk assessment.

[0013] In one embodiment, the assistance device further includes a knowledge graph update unit that, when detecting a change in the facility information, the hazard information, the personnel information, or the work information, updates the knowledge graph based on the change.

[0014] In this way, even if facility information, hazard information, personnel information, or work information changes, the knowledge graph is updated, thereby preventing hazards or risks from being missed from being extracted.

[0015] In one embodiment, the support device further includes an output unit that outputs the risk assessment list in a table format or a graph format.

[0016] In this way, the risk assessment list is output in a table format or a graph format, so that the operator can easily grasp the hazards or risks.

[0017] In some embodiments, the program Computer, It functions as the above support device.

[0018] In this way, hazards or risks are extracted by using a knowledge graph including facility information, hazard information, and work information, thereby improving the technology for preventing omission of hazards or risks in risk assessment.

[0019] In some embodiments, the method comprises: A method for supporting risk assessment of a production system, the method comprising: Generating a knowledge graph including facility information, hazard information, and work information; generating a search query based on work information during operation, searching the knowledge graph based on the search query, and obtaining hazards or risks related to the search query as search results; generating a risk assessment list in which the work information during the operation is associated with the search results; This includes performing the following.

[0020] In this way, hazards or risks are extracted by using a knowledge graph including facility information, hazard information, and work information, thereby improving the technology for preventing omission of hazards or risks in risk assessment. [Effects of the Invention]

[0021] According to the present disclosure, it is possible to improve the technology for preventing hazards or risks from being overlooked in risk assessment. [Brief explanation of the drawings]

[0022] [Figure 1] FIG. 1 is a diagram illustrating a knowledge graph. [Figure 2] FIG. 1 is a diagram illustrating a connected knowledge graph. [Figure 3] FIG. 1 is a diagram illustrating a facility knowledge graph according to an embodiment of the present disclosure. [Figure 4] FIG. 1 is a diagram illustrating a hazard knowledge graph according to an embodiment of the present disclosure. [Figure 5] FIG. 1 illustrates a human resources knowledge graph according to an embodiment of the present disclosure. [Figure 6] FIG. 1 is a diagram illustrating an operation knowledge graph according to an embodiment of the present disclosure. [Figure 7] FIG. 1 is a diagram illustrating a connected knowledge graph according to an embodiment of the present disclosure. [Figure 8] 1 is a diagram illustrating a support device according to an embodiment of the present disclosure. [Figure 9]9 is a diagram illustrating an example of the operation of a knowledge graph search unit included in the support device shown in FIG. 8. FIG. [Figure 10] FIG. 10 is a diagram illustrating a risk assessment list output in a table format according to an embodiment of the present disclosure. [Figure 11] FIG. 10 is a diagram illustrating a risk assessment list output in a graphical format according to an embodiment of the present disclosure. [Figure 12] FIG. 10 is a diagram illustrating a support device according to a modified example of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0023] Hereinafter, an embodiment of the present disclosure will be described with reference to the drawings. In each drawing, the same reference numerals indicate the same or equivalent components.

[0024] The present embodiment aims to improve technology for preventing omission of hazards or risks in risk assessment. Details will be described later, but the present embodiment aims to prevent omission of hazards or risks that are present, either explicitly or implicitly, in work specified by an operator by using a knowledge graph that includes equipment information, hazard information, personnel information, and work information in a production system. Furthermore, the present embodiment aims to promote risk countermeasures by preventing omission of hazards or risks, and to support risk assessment, thereby contributing to the operational safety of a plant.

[0025] (Knowledge Graph) The knowledge graph according to this embodiment will be described with reference to FIGS.

[0026] The basic elements and connections of a knowledge graph will be described with reference to Figures 1 and 2. First, referring to Figure 1, a knowledge graph includes basic elements such as a "subject," a "relation," and an "object." For example, "ABC Company owns the Saitama site" can be expressed in a knowledge graph as "subject: ABC Company," "relation: owns," and "object: Saitama site." Furthermore, for example, "The Saitama site includes an intermediate product production area" can be expressed in a knowledge graph as "subject: Saitama site," "relation: includes," and "object: intermediate product production area." Next, referring to Figure 2, separately defined knowledge graphs A to C are connected into a single knowledge graph by defining their respective "relationships (links)" (shown by dashed dotted lines). This connected knowledge graph (hereinafter also referred to as a "connected knowledge graph") is stored in a graph database, and searches based on various conditions are possible by following the "(relationship) links."

[0027] The facility knowledge graph will be described with reference to FIG. 3. First, the ontology in FIG. 3 is a hierarchical definition of generalized or abstract concepts (classes) of facilities or equipment related to a company's plant. The hierarchical concepts conform to existing industry standards or corporate practices. Each concept has "attributes" as a common property or characteristic. Next, the instances in FIG. 3 are defined to match the same concept. The configuration or attributes of the instances may be predefined or dynamically acquired from the facility information management system 201 (described later). Finally, the knowledge graph in FIG. 3 defines the relationship between the ontologies and instances, associates them on a graph, and is constructed by stacking and linking basic elements. The knowledge graph is saved in a format compatible with a graph database. The format may be, for example, a standard RDF format or may conform to the specified format of the graph database.

[0028] The hazard knowledge graph will be described with reference to Figure 4. First, the ontology in Figure 4 is a hierarchical definition of generalized or abstract concepts (classes) of hazards (i.e., causes of danger) related to production systems in the process industry, etc. Here, the hierarchical concepts follow existing industry safety and risk assessment standards or corporate practices. Each concept also has "attributes" as common properties or characteristics. Next, the instances in Figure 4 are defined to match the same concept. Here, the configuration or attributes of the instances may be predefined or dynamically acquired from the hazard definition information storage system 202, which will be described later. Finally, the knowledge graph in Figure 5 defines the relationship between the ontologies and instances, associates them on a graph, and is constructed by stacking and linking basic elements. Note that the knowledge graph is stored in a format compatible with a graph database. The format may be, for example, a standard RDF format or may follow the format specified for the graph database.

[0029] The human resource knowledge graph will be described with reference to FIG. 5. First, the ontology in FIG. 5 is a hierarchical definition of generalized or abstract concepts (classes) of employees related to plants in a production system. The hierarchical concepts follow existing industry standards or corporate practices. Each concept has "attributes" as a common property or characteristic. Next, the instances (employees) in FIG. 5 are defined to match the same concept. The configuration or attributes of the instances may be predefined or dynamically acquired from the employee information management system 203 (described later). Finally, the knowledge graph in FIG. 5 defines the relationship between the ontologies and instances, associates them on a graph, and is constructed by stacking and linking basic elements. The knowledge graph is stored in a format compatible with a graph database. The format may be, for example, a standard RDF format or may conform to the specified format of the graph database.

[0030] The work knowledge graph will be described with reference to FIG. 6. First, the ontology in FIG. 6 is a hierarchical definition of generalized or abstract concepts (classes) of work related to production systems in the process industry, etc. Here, the hierarchical concepts follow existing industry work standards or corporate work rules, etc. Furthermore, each concept has "attributes" as common properties or characteristics. Next, the instances in FIG. 6 are defined to match the same concept. Here, the configuration or attributes of the instance may be defined in advance or may be dynamically acquired from the work management system 205 described below. Finally, the knowledge graph in FIG. 6 is constructed by defining the relationship between the ontology and the instance, associating them on a graph, and stacking and connecting basic elements. Note that the knowledge graph is saved in a format compatible with a graph database. The format may be, for example, a standard RDF format or may follow the specified format of the graph database.

[0031] By linking the knowledge graphs shown in FIGS. 3 to 6 based on predetermined relationships, a connected knowledge graph, an example of which is shown in FIG. 7, is created.

[0032] (Support device) The configuration of a support device 100 for supporting risk assessment of a production system according to this embodiment will be described with reference to FIG.

[0033] The assistance device 100 includes an external data acquisition unit 101, a data conversion unit 102, a knowledge graph generation unit 103, a knowledge graph update unit 104, a knowledge graph storage unit 105, an input unit 106, a task data acquisition unit 107, a target task selection unit 108, a task data extraction unit 109, a knowledge graph search unit 110, a risk assessment list generation unit 111, an output control unit 112, and an output unit 113.

[0034] The external data acquisition unit 101 receives a request from the knowledge graph generation unit 103 and the knowledge graph update unit 104 to acquire required external data, and acquires the external data from an external information management system 200 .

[0035] The data conversion unit 102 appropriately converts the external data acquired from the external data acquisition unit 101 into a format suitable for the knowledge graph, and passes it to the knowledge graph generation unit 103 .

[0036] The knowledge graph generation unit 103 receives an input from the operator 122 or an input from the data conversion unit 102 and generates a knowledge graph.

[0037] The knowledge graph update unit 104 receives input data from the task data extraction unit 109 (i.e., data extracted from the task data) and searches for data on the knowledge graph related to the input data. Furthermore, if the knowledge graph update unit 104 determines that the knowledge graph needs to be updated based on the search, it requests the external data acquisition unit 101 to acquire data related to the update. Furthermore, the data related to the update acquired from the external data acquisition unit 101 in response to a request from the knowledge graph update unit 104 is overwritten and saved in the knowledge graph storage unit 105 via the data conversion unit 102 and the knowledge graph generation unit 103. When the knowledge graph update unit 104 detects a change in the facility information, hazard information, personnel information, or task information included in the knowledge graph, it updates the knowledge graph based on the change using the method described above. As a result, the knowledge graph is updated based on information dynamically acquired from the external information management system 200.

[0038] The knowledge graph storage unit 105 stores the knowledge graph in a graph database. The knowledge graph storage unit 105 also executes update, deletion, or search requests. For searches, an endpoint is provided that allows a search to be performed on the knowledge graph using a search query. Note that the search query may be a SPARQL query in the case of a knowledge graph of RDF data based on the W3C standard, or a knowledge graph query language specified by the graph database may be used.

[0039] The input unit 106 accepts "knowledge graph generation data" input by the operator 122 and passes it to the knowledge graph generation unit 103. The input unit 106 also accepts "selection of work content" input by the operator 122 and passes it to the target work selection unit 108.

[0040] The task data acquisition unit 107 acquires task data from the external task management system 205 in accordance with the "task content selection" input by the operator 122 and passed to the target task selection unit 108 .

[0041] The target task selection unit 108 receives the "task content selection" input by the operator 122 via the input unit 106, and requests the task data acquisition unit 107 to acquire task data. The target task selection unit 108 also passes the task data related to the request to the task data extraction unit 109.

[0042] The task data extraction unit 109 extracts nouns and verbs as words from task data (structured data format or unstructured text format) using any method, and creates "task content extraction data" shown as an example in column 1051 of Figure 10.

[0043] The knowledge graph search unit 110 generates a search query for data related to the "work content extracted data" based on the "work content extracted data" created by the work data extraction unit 109. The knowledge graph search unit 110 also searches the knowledge graph based on the generated search query. The knowledge graph search unit 110 also obtains hazards or risks related to the search query as search results. The knowledge graph search unit 110 also generates a set of the "work content extracted data" and the search results, temporarily stores it, and passes it to the risk assessment list generation unit 111.

[0044] The risk assessment list generation unit 111 generates a risk assessment list that associates each element of the "work content extracted data" from the knowledge graph search unit 110 with the search results by the knowledge graph search unit 110 related to that element.

[0045] The output control unit 112 executes any process for outputting the risk assessment list created by the risk assessment list creation unit 111 in a table format or a graph format.

[0046] The output unit 113 outputs the results of processing by the output control unit 112. That is, the output unit 113 outputs the risk assessment list created by the risk assessment list generation unit 111 in table format or graph format. FIG. 10 shows an example of output when a risk assessment is output in table format. Column 1051 in FIG. 10 shows a list of activity data extracted by the activity data extraction unit 109. Column 1052 shows related instances having the same concept (previously defined in the knowledge graph) discovered by a search on the knowledge graph. Columns 1053 to 1058 show attributes related to the instances in column 1052. However, these attributes are merely examples, and the present disclosure is not limited to these. For example, attributes include foreign language notation, abbreviations, or variations in terminology spelling. FIG. 11 also shows an example of output when a risk assessment list is output in graph format. Here, a summary ranking of the number of hazards by activity is shown. The graph format may include a summary chart by hazard, facility, or location, a distribution chart, or a summary chart in order of the degree of impact.

[0047] Each component of the assistance device 100 may include a processor such as a central processing unit (CPU) or a graphics processing unit (GPU), a programmable circuit such as a field-programmable gate array (FPGA), or a dedicated circuit such as an application-specific integrated circuit (ASIC). Each component of the assistance device 100 may include a communication interface compatible with any communication standard, such as a mobile communication standard, a wired local area network (LAN) standard, or a wireless LAN standard. Each component of the assistance device 100 may include a semiconductor memory such as a random access memory (RAM) or a read-only memory (ROM), a magnetic memory, or an optical memory. The input unit 106 may include an input interface such as a physical key, a capacitive key, a pointing device, a touch screen, or a microphone. The output unit 113 may include an output interface such as a liquid crystal display (LCD) or an organic electroluminescence (EL) display.

[0048] (External Information Management System) The configuration of an external information management system 200 according to this embodiment will be described with reference to FIG.

[0049] The external information management system 200 includes a facility information management system 201, a hazard definition information storage system 202, an employee information management system 203, other systems 204, and a work management system 205. The external information management system 200 exists on a corporate network other than the support device 100, is connected to the support device 100, and provides necessary data to the support device 100 in response to a request from the support device 100.

[0050] The equipment information management system 201 and the hazard definition information storage system 202 are systems that manage and store information corresponding to concepts on a knowledge graph, such as facility information or equipment information in a company or a plant owned by a company.

[0051] The employee information management system 203 is a system that manages and holds information corresponding to concepts on the knowledge graph, such as employee information.

[0052] The work management system 205 is a system that manages and stores work information that is the subject of risk assessment.

[0053] (Operation of the support device) The operation of the assistance device 100 according to this embodiment will be described with reference to FIGS.

[0054] [Knowledge graph definition before use] Step S100: At the time of initial implementation, the operator 122 defines concepts such as the configuration of equipment (or facilities) owned by the company, the composition of employees, and hazard information (definition information) used on the knowledge graph in a hierarchical format categorized according to industry standards (e.g., international standards) or company practices (e.g., internal regulations) in the knowledge graph generation unit 103 via the input unit 106. This corresponds to the ontology definition shown in Figure 1.

[0055] Step S110: The operator 122 defines the relationship between the company ontology and the instance on the knowledge graph in accordance with the ontology of the knowledge graph in the knowledge graph generation unit 103 via the input unit 106. This corresponds to the definition of the instance shown in FIG. 1. The instance, including its attributes, is acquired from the external information management system 200 via the external data acquisition unit 101.

[0056] Step S120: The operator 122 stores the knowledge graph in the knowledge graph storage unit 105 via the input unit 106.

[0057] [During actual operation (operation)] Step S200: The operator 122 selects an operation to be the target of the risk assessment of the production system in the target operation selection unit 108 via the input unit 106.

[0058] Step S210: The target task selection unit 108 acquires task data related to the task selected by the operator 122 from the task management system 205 via the task data acquisition unit 107, and passes the acquired data to the task data extraction unit 109.

[0059] Step S220: The activity data extraction unit 109 extracts nouns and verbs as words from the activity data, generates “activity content extracted data” an example of which is shown in column 1051 of FIG.

[0060] Step S230: The knowledge graph search unit 110 generates a search query for data related to the "work content extracted data" based on the "work content extracted data." Next, the knowledge graph search unit 110 searches the knowledge graph based on the generated search query. Next, the knowledge graph search unit 110 obtains hazards or risks related to the search query as search results. Next, the knowledge graph search unit 110 generates a set of the "work content extracted data" and the search results, temporarily stores it, and passes it to the risk assessment list generation unit 111.

[0061] Here, the operation of the knowledge graph search unit 110 will be described in detail with reference to FIG.

[0062] Step S231: The knowledge graph search unit 110 receives a list of “task content extracted data” from the task data extraction unit 109.

[0063] Step S232: The knowledge graph search unit 110 extracts one piece of data from the list. For example, the knowledge graph search unit 110 extracts the data indicating "power supply unit" in FIG.

[0064] Step S233: The knowledge graph search unit 110 generates a search query based on the extracted data and sends a search request to the knowledge graph holding unit 105. For example, the knowledge graph search unit 110 generates a query word corresponding to "power supply unit."

[0065] Step S234: The knowledge graph search unit 110 receives data related to the search query from the knowledge graph storage unit 105 as a search result. Here, the search result may be data related to the search query and list data of specific instances and attributes on the knowledge graph related to the search query. For example, the search result may be list data of instances shown as an example in column 1052 of FIG. 10 and attributes shown as an example in columns 1053 to 1058 of FIG. 10.

[0066] Step S235: The knowledge graph search unit 110 generates a set of the data extracted in step S232 and the search results received in step S234, and temporarily stores it.

[0067] Step S236: The knowledge graph search unit 110 determines whether the next data exists in the list received in step S231. If the next data exists, the process returns to step S232. If the next data does not exist, the process proceeds to step S237. For example, in FIG. 10, data indicating "pump" exists as the data following the data indicating "power supply unit," so the process returns to step S232.

[0068] Step S237: The knowledge graph search unit 110 passes the set temporarily saved in step S235 to the risk assessment list generation unit 111.

[0069] In this manner, the knowledge graph search unit 110 generates a search query based on the work information during operation. Subsequently, the knowledge graph search unit 110 searches the knowledge graph based on the search query. Subsequently, the knowledge graph search unit 110 acquires hazards or risks related to the search query as search results. Here, in step S233, the knowledge graph search unit 110 may search from the knowledge graph node first identified by the search query to nodes at a predetermined depth. This may be performed by describing the hierarchical depth range of the search condition property in the search query (such as the W3C standard query language SPARQL) via the input unit 106. For example, if a search query corresponding to "power supply unit" is generated and the predetermined depth is "2," in FIG. 7, the search targets are "electricity," "work hazard," "production safety damage," "source 1," "Jack," "employee 2525," "N123," "Company X," "parts," "pump P-1," "composite work unit," "required equipment," and "equipment." If a node that perfectly matches the expression included in the search query is not found in the knowledge graph, the node with the highest text match may be identified as the search target. In this case, the search target may be narrowed down to the range of other conditions specified by the search query (such as the facility or location).

[0070] Step S240: The risk assessment list generation unit 111 generates a risk assessment list based on the set received from the knowledge graph search unit 110.

[0071] Step S250: The output control unit 112 executes a process for outputting the risk assessment list generated by the risk assessment list generating unit 111 in a table format or a graph format.

[0072] Step S260: The output unit 113 outputs the results of the processing by the output control unit 112. As a result, the risk assessment list is output in the table format shown in FIG. 10 or the graph format shown in FIG.

[0073] According to this embodiment, it is possible to improve the technology for preventing omission of hazards or risks in risk assessment.

[0074] In particular, according to this embodiment, in a pre-work risk assessment, potential hazards involved in the work can be broadly identified, and related information such as causes, countermeasures, examples of similar accidents that can serve as reference, and the degree of impact can be presented, thereby encouraging the elimination or reduction of potential dangers or harms during the work. This makes it possible to support risk assessment and reduce risks.

[0075] Furthermore, according to this embodiment, the corresponding information in the knowledge graph can be updated based on information dynamically acquired from the external information management system 200. Therefore, the latest status of related equipment, the operating environment, human resources, etc. can be reflected in the risk assessment. For example, a pipe corresponding to equipment may previously have carried room-temperature gas (no risk of burns), but after a change in the production process, high-temperature gas (risk of burns) may now be passing through it. Even in this case, the attributes of the pipe corresponding to the equipment are dynamically acquired, making it possible to discover the latest hazards.

[0076] Furthermore, according to this embodiment, the operator 122 can freely create a knowledge graph according to the circumstances and experience of the company, plant, or work site. This allows the operator 122 to reflect his or her own knowledge about risks in the risk assessment, which contributes to the prevention and mitigation of accidents. For example, in the section of the knowledge graph corresponding to "equipment," the operator can add links to reports of related accidents that may occur in the equipment, or add links to the operating know-how of experts, and present these in the risk assessment, thereby preventing or mitigating accidents in the equipment.

[0077] Furthermore, according to this embodiment, even if the target of risk assessment or the type of hazard changes, the support device 100 can be configured to perform flexibly changeable and scalable risk assessment by adding and defining a knowledge graph and changing the connection to the external information management system 200. For example, if a production factory is expanded, the scope of the risk assessment can be easily expanded by adding the facilities or equipment of the production factory to the section corresponding to "equipment" in the knowledge graph. Furthermore, for example, if a new regulated chemical substance is added to the risk assessment due to a company request or a revision of law, etc., the new hazard can be identified in the risk assessment by adding the hazards related to the chemical substance to the knowledge graph and adding information about the chemical substance to the attributes of the equipment that handles the chemical substance.

[0078] Although the present disclosure has been described above based on the drawings and examples, it should be noted that those skilled in the art may make various modifications and alterations based on the present disclosure. Therefore, it should be noted that these modifications and alterations are included within the scope of the present disclosure. For example, the functions included in each component or step can be rearranged so as not to cause a logical inconsistency, and multiple components or steps can be combined or divided into one.

[0079] As a modified example, an embodiment is also possible in which the configuration and operation of the assistance device 100 are distributed among multiple computers that can communicate with each other. For example, an embodiment is also possible in which the knowledge graph generation unit 103 and the knowledge graph storage unit 105 included in the assistance device 100 in FIG. 8 are provided in an external knowledge graph management device 300 different from the assistance device 100-1, as shown in FIG. 12. In this case, the knowledge graph management device 300 may be provided with its own input unit 301. According to this modified example, the knowledge graph management device 300 generates and manages a knowledge graph common to an industry or company, and can also accept requests to use the knowledge graph from multiple assistance devices, such as the assistance device 100-1 for production base A, the assistance device 100-2 for production base B, and the assistance device 100-3 for production base C. Furthermore, according to this modification, a single knowledge graph management device 300 can be used by support devices 100-1, 100-2, and 100-3 for multiple production bases A, B, and C, leading to more efficient investment and use of information systems. Furthermore, according to this modification, a company's knowledge graph can be centrally managed, allowing various pieces of knowledge to be shared within the company. Furthermore, according to this modification, a company with multiple production bases can have a common knowledge graph concept (ontology) within the company, and associate unique information (instances) such as facilities, equipment, employees, or work at each production base with the concept, allowing different production bases to share knowledge.

[0080] The differences between the above-described embodiment and this modification are as follows: In the above-described embodiment, an instance of the knowledge graph is created based on information specific to a single production base (facilities and equipment, hazard definitions, employees, work information, etc.) and linked to an ontology. In contrast, in this modification, an instance is created for each production base based on information specific to each production base and linked to an ontology. Furthermore, while the external information management systems 200-1, 200-2, and 200-3 in the above-described embodiment are systems for managing information within the company, in this modification, when the information is not centrally managed by the company, they are systems for managing information containing information specific to each production base, and when the information is centrally managed by the company, they are the management destination for that information.

[0081] As a modified example, the support device 100 according to the above-described embodiment can be used as a device that supports analyzing the cause of an operational accident from records of past incidents by appropriately changing the content of the knowledge graph. Also, the support device 100 according to the above-described embodiment can be used as a device that supports analyzing the cause of equipment failure diagnosis.

[0082] Also, as a variant, the knowledge graph may not include human information.

[0083] As a modified example, for example, a general-purpose computer may be made to function as the assistance device 100 according to the above-described embodiment. Specifically, a program describing the processing content for realizing each function of the assistance device 100 according to the above-described embodiment is stored in the memory of the general-purpose computer, and the program is read and executed by a processor. Therefore, the present disclosure may also be realized as a program executable by a processor, or a non-transitory computer-readable medium storing the program. [Industrial Applicability]

[0084] The present disclosure can improve techniques for preventing hazards or risks from being overlooked in risk assessments. [Explanation of symbols]

[0085] 100, 100-1, 100-2, 100-3 Support device 101 External data acquisition unit 102 Data conversion unit 103 Knowledge Graph Generation Unit 104 Knowledge Graph Update Department 105 Knowledge graph storage unit 106 Input section 106 107 Work data acquisition unit 108 Target work selection section 109 Work Data Extraction Unit 110 Knowledge Graph Search Unit 111 Risk Assessment List Generation Unit 112 Output control unit 112 113 Output unit 113 122 Operator 200, 200-1, 200-2, 200-3 External information management system 201 Facility Information Management System 202 Hazard definition information storage system 203 Employee Information Management System 204 Other Systems 205 Work Management System 300 Knowledge graph management device 301 Input section

Claims

1. A support device for supporting risk assessment of a production system, comprising: a knowledge graph generation unit that generates a knowledge graph including facility information, hazard information, and work information; a knowledge graph search unit that generates a search query based on work information during operation, searches the knowledge graph based on the search query, and acquires hazards or risks related to the search query as search results; a risk assessment list generation unit that generates a risk assessment list in which the work information during the operation is associated with the search results; an output unit that outputs the risk assessment list in a graph format showing a total ranking of the number of hazards by task; Equipped with The knowledge graph search unit searches the knowledge graph from a node initially identified based on the search query to a node at a predetermined depth.

2. The support device according to claim 1, The knowledge graph generation unit further generates a knowledge graph including human information.

3. 3. The support device according to claim 2, The assistance device further comprises a knowledge graph update unit that, when a change in the facility information, the hazard information, the personnel information, or the work information is detected, updates the knowledge graph based on the change.

4. Computer, A program for causing the assisting device according to any one of claims 1 to 3 to function.

5. A method for supporting risk assessment of a production system, the method comprising: Generating a knowledge graph including facility information, hazard information, and work information; generating a search query based on work information during operation, searching the knowledge graph based on the search query, and obtaining hazards or risks related to the search query as search results; generating a risk assessment list in which the work information during the operation is associated with the search results; outputting the risk assessment list in a graph format showing a total ranking of the number of hazards by task; performing The method, wherein the obtaining includes searching the knowledge graph from a node initially identified based on the search query to nodes at a predetermined depth.

Images