Information processing device, information processing method, and information processing program

Abstract

To provide an information processing device and the like that can reduce the workload of an inspector.SOLUTION: An information processing device registers inspection result information including an inspection result of an inspection object, in a database. The information processing device includes: a knowledge management unit that generates a query condition corresponding to the inspection result information, and retrieves a knowledge corresponding to the query condition from the database; and a notification unit that outputs for notification the knowledge retrieved by the knowledge management unit to an inspector terminal device.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to an information processing device, an information processing method, and an information processing program. [Background technology]

[0002] For example, an inspection support system is widely known in which an inspector inspects each piece of equipment within a plant that is subject to inspection, and then transfers the inspection results to a server device using a tablet device that the inspector brings with him or her, and the server device then manages the inspection results of each piece of equipment.

[0003] FIG. 10 is an explanatory diagram showing an example of a conventional inspection support system 100. The inspection support system 100 shown in FIG. 10 includes a tablet terminal 110, a PC (Personal Computer) terminal 120, a server device 130, and a communication network 140. The tablet terminal 110 is, for example, a terminal device carried by an inspector who inspects equipment to be inspected in a plant. The PC terminal 120 is, for example, a terminal device located in an office. The server device 130 is an information processing device that manages the entire inspection support system 100. The communication network 140 is a communication network that transmits and receives information between the tablet terminal 110, the PC terminal 120, and the server device 130.

[0004] The tablet terminal 110 has a tablet communication unit 111, a tablet input unit 112, a tablet display unit 113, a tablet memory unit 114, and a tablet control unit 115. The tablet communication unit 111 is a communication interface that communicates with a communication network 140. The tablet input unit 112 is an input interface, such as a touch panel, that inputs various types of information. The tablet input unit 112 inputs the inspection results of the equipment to be inspected. The tablet display unit 113 displays various types of information. The tablet memory unit 114 stores various types of information. The tablet control unit 115 controls the entire tablet terminal 110.

[0005] The PC terminal 120 has a PC-side communication unit 121, a PC-side input unit 122, a PC-side display unit 123, a PC-side memory unit 124, and a PC-side control unit 125. The PC-side communication unit 121 is a communication interface that communicates with the communication network 140. The PC-side input unit 122 is an input interface that inputs various information such as commands and data. The PC-side display unit 123 is an output interface that displays various information. The PC-side memory unit 124 stores various information. The PC-side control unit 125 controls the entire PC terminal 120.

[0006] The server device 130 has a server-side communication unit 131, a database 132, and a server-side control unit 133. The server-side communication unit 131 is a communication interface that communicates with the communication network 140. The database 132 stores various information. The database 132 manages various information such as the inspection results for each device to be inspected, the contents of incidents and countermeasures, manuals, etc. The server-side control unit 133 controls the entire server device 130.

[0007] 11 is a sequence diagram showing an example of the processing operations of the inspection support system 100 (shown in FIG. 10) related to input processing, unresolved incident registration processing, and search processing. The inspection support system 100 has input processing, unresolved incident registration processing, and search processing.

[0008] The input process is a process of registering inspection result information input from the tablet terminal 110 in the database 132. The unresolved incident registration process is a process of inputting the details and countermeasures of an unresolved incident from the PC terminal 120 to the server device 130, and registering the input details and countermeasures of the unresolved incident in the database 132. The search process is a simple search process of performing a simple search using a keyword from the PC terminal 120 in the database 132 and returning the search results to the PC terminal 120.

[0009] An inspector brings tablet terminal 110 to a site within a plant and inspects the equipment to be inspected. First, as an input process, the inspector inputs inspection result information, including inspection results such as the status and temperature of the equipment to be inspected confirmed on-site, from tablet input unit 112 to tablet control unit 115 (step S201). The inspection result information includes the name of the equipment to be inspected, the inspection results, etc. When the tablet control unit 115 inputs the inspection result information, it transfers the inspection result information to server device 130 via communication network 140 (step S202).

[0010] Then, the server-side control unit 133 in the server device 130 processes the received inspection result information, transfers the processed inspection result information to the database 132 (step S203), and registers the inspection result information in the database 132 (step S204). As a result, the inspection result information input from the tablet terminal 110 can be registered in the database 132.

[0011] Furthermore, as an unresolved incident registration process, for example, when the server-side control unit 133 has registered inspection result information in the database 132, it notifies the PC-side control unit 125 in the PC terminal 120 of an unresolved incident registration request (step S211). The PC-side control unit 125 displays the inspection result information of the inspection target on the PC-side display unit 123 as an unresolved incident registration request. As a result, the user of the PC terminal 120 looks at the name of the device to be inspected and the inspection results, and inputs the content of the incident and countermeasures to the PC-side control unit 125 via the PC-side input unit 122 (step S212). Note that the content of the incident may be, for example, the content and cause of the inspection result, and the countermeasures for the incident may be, for example, the content of the countermeasures taken in response to the inspection result. Furthermore, when the PC-side control unit 125 has input the content of the incident and countermeasures, it notifies the server-side control unit 133 of a request to register the content of the incident and countermeasures (step S213).

[0012] When the server-side control unit 133 receives a request to register the incident details and countermeasures, it transfers the incident details and countermeasures to the database 132 in association with the inspection result information (step S214). Then, the server-side control unit 133 registers the incident details and countermeasures in the database 132 (step S215). As a result, the incident details and countermeasures input from the PC terminal 120 can be registered in the database 132.

[0013] Furthermore, as a simple search process, the PC-side control unit 125 inputs a simple search request including a keyword through the PC-side input unit 122 and notifies the server-side control unit 133 of the input simple search request (step S221). When the server-side control unit 133 receives the simple search request from the PC-side control unit 125, it notifies the database 132 of the query conditions for the simple search (step S222).

[0014] The server-side control unit 133 acquires information corresponding to the query conditions of the simple search from the database 132 as search results (step S223). Furthermore, the server-side control unit 133 transfers the search results acquired from the database 132 to the PC-side control unit 125 (step S224). The PC-side control unit 125 displays the search results transferred from the server-side control unit 133 on the PC-side display unit 123. As a result, the user of the PC terminal 120 can check the search results of the simple search by looking at the display screen of the PC-side display unit 123. [Prior art documents] [Non-patent literature]

[0015] [Non-Patent Document 1] Yokogawa Electric Corporation, "Digital Transformation in the Field (Field Assistant)", [online], Yokogawa Electric Corporation, [Retrieved February 3, 2024], Internet <URL:https: / / www.yokogawa.co.jp / solutions / solutions / oprex-transformation / oprex-operational-risk-management / field-assistant-jp / #%E9%96%A2%E9%80%A3%E6%83%85%E5%A0%B1__%E5%B0%8E%E5%85%A5%E4%BA%8B%E4%BE%8B> Summary of the Invention [Problem to be solved by the invention]

[0016] However, the conventional inspection support system 100 is a one-way communication system in which inspection result information is input from the tablet terminal 110 and registered in the database 132. In other words, during an inspection, the tablet terminal 110 cannot search for the details of incidents or countermeasures, such as past cases, that correspond to the inspection result information. As a result, the inspector cannot recognize the details of the incident or countermeasures in the inspection result information, which places a heavy workload on the inspector.

[0017] In one aspect, an object is to provide an information processing device or the like that can reduce the workload of an inspector. [Means for solving the problem]

[0018] In one aspect, the information processing device disclosed in the present application registers inspection result information including inspection results of an inspection target in a database. The information processing device has a knowledge management unit that generates a query condition corresponding to the inspection result information and searches the database for knowledge corresponding to the query condition, and a notification unit that notifies an inspector terminal device of the knowledge searched by the knowledge management unit. [Effects of the Invention]

[0019] According to one aspect, the workload of the inspector can be reduced. [Brief explanation of the drawings]

[0020] [Figure 1] FIG. 1 is an explanatory diagram showing an example of an inspection support system according to this embodiment. [Figure 2] FIG. 2 is a sequence diagram showing an example of the processing operation of the inspection support system related to input processing. [Figure 3] FIG. 3 is a flowchart showing an example of the processing operation of the tablet control unit related to tablet input processing. [Figure 4] FIG. 4 is a flowchart showing an example of the processing operation of the server device related to the inspection result registration processing. [Figure 5] FIG. 5 is a sequence diagram showing an example of the processing operation of the inspection support system related to the unresolved incident registration processing. [Figure 6] FIG. 6 is a flowchart showing an example of the processing operation of the server device related to the unresolved incident registration processing. [Figure 7] FIG. 7 is a sequence diagram showing an example of the processing operation of the inspection support system related to the search processing. [Figure 8] FIG. 8 is a flowchart showing an example of the processing operation of the server device related to the search processing. [Figure 9] FIG. 9 is an explanatory diagram illustrating an example of an information processing device that executes an information processing program. [Figure 10] FIG. 10 is an explanatory diagram showing an example of a conventional inspection support system. [Figure 11] FIG. 11 is a sequence diagram showing an example of the processing operations of the inspection support system relating to input processing, unresolved incident registration processing, and search processing. DETAILED DESCRIPTION OF THE INVENTION

[0021] Hereinafter, embodiments of the information processing device and the like disclosed in the present application will be described in detail with reference to the drawings. Note that the disclosed technology is not limited to these embodiments. Furthermore, the embodiments described below may be modified as appropriate within the scope of not causing any contradiction.

[0022] <Example> FIG. 1 is an explanatory diagram showing an example of an inspection support system 1 according to the present embodiment. The inspection support system 1 shown in FIG. 1 includes a tablet terminal 2, a PC (Personal Computer) terminal 3, a server device 4, and a communication network 5. The tablet terminal 2 is, for example, an inspector terminal device carried by an inspector who inspects equipment to be inspected in a plant. The equipment to be inspected is, for example, field equipment such as a pressure transmitter, a temperature transmitter, a multivariable transmitter, a flow meter, or an induction heating power supply for an epitaxial growth apparatus. The PC terminal 3 is, for example, a terminal device such as a computer located in an office. The server device 4 is an information processing device such as a computer that manages the entire inspection support system 1. The communication network 5 is, for example, a communication network such as the Internet or a LAN (Local Area Network) that transmits and receives information between the tablet terminal 110, the PC terminal 120, and the server device 130.

[0023] <Tablet device configuration> The tablet terminal 2 has a tablet-side communication unit 21, a tablet-side input unit 22, an evaluation input unit 23, a tablet-side output unit 24, a tablet-side storage unit 25, and a tablet-side control unit 26. The tablet-side communication unit 21 is a communication interface that communicates with the communication network 5.

[0024] The tablet-side input unit 22 is an input interface, such as a touch panel, for inputting various types of information. The tablet-side input unit 22 inputs inspection result information for the equipment to be inspected. The inspection result information is information including the name of the equipment to be inspected, the inspection results, etc. The tablet-side input unit 22 is a manual input method for inputting the inspection results, for example, by keyboard input, pull-down menu, dial input, etc. The tablet-side input unit 22 also has a keyboard input method or a fixed phrase input method for inputting a string of characters as the inspection result. The keyboard input method is, for example, a method for inputting characters using a normal keyboard. The fixed phrase input method is, for example, a method for inputting the status of the equipment to be inspected using fixed phrases prepared in advance by the inspection support system 1, using pull-down menus, radio buttons, etc.

[0025] The tablet-side output unit 24 is an output interface, such as a display unit or an audio output unit, that outputs various types of information. The tablet-side output unit 24 notifies and outputs knowledge for the inspection result information, which will be described later. The knowledge is, for example, information on the content of an incident for the inspection result and countermeasures. The content of the incident is, for example, information on the state of the inspection result and its cause. The countermeasures for the incident are, for example, information on countermeasures to improve the state of the inspection result or suppression methods.

[0026] The evaluation input unit 23 is an input interface for evaluating the effectiveness of the knowledge content, for example, when the inspector finds the knowledge content useful, by inputting a high evaluation to the tablet-side control unit 26. The inspector looks at the knowledge content for the inspection result information output from the tablet-side output unit 24 and inputs an evaluation of the knowledge through the evaluation input unit 23.

[0027] The tablet-side storage unit 25 stores various types of information. The tablet-side control unit 26 controls the entire tablet terminal 2. The tablet-side control unit 26 notifies the server device 4 of the knowledge evaluation according to the evaluation of the knowledge by the evaluation input unit 23. The knowledge evaluation is, for example, information evaluated by the inspector on the validity of the content of the notified knowledge, for example, a five-level evaluation, with the highest evaluation being 5, the lowest evaluation being 1, and the evaluation result indicating that the validity was higher in ascending order from 1 to 5. Note that the evaluation is not limited to a five-level evaluation and can be changed as appropriate.

[0028] <Configuration of the PC terminal> The PC terminal 3 includes a PC-side communication unit ३१, a PC-side input unit ३२, a PC-side output unit ३३, a PC-side storage unit ३४, and a PC-side control unit ३५. The PC-side communication unit ३१ is a communication interface for communicating with the communication network 5. The PC-side input unit ३२ is an input interface for inputting various types of information such as commands and data. The PC-side output unit ३३ is an output interface such as a display unit and a voice output unit for outputting various types of information. The PC-side storage unit ३४ stores various types of information. The PC-side control unit ३५ controls the entire PC terminal 3.

[0029] <Configuration of the server device> The server device 4 includes a server-side communication unit 41, a database 42, a knowledge management unit 43, and a server-side control unit 44. The server-side communication unit 41 is a communication interface for communicating with the communication network 5. The server-side control unit 44 controls the entire server device 4. The database 42 stores various types of information. The database 42 registers and manages various types of information such as inspection results for each device to be inspected, the content and countermeasures of incidents, natural language such as manuals, and past data.

[0030] The knowledge management unit 43 includes a search unit 43A, a determination unit 43B, an acquisition unit 43C, and a learning unit 43D. The search unit 43A generates query conditions for semantic search, such as past cases, corresponding to the inspection result information from the tablet terminal 2. Semantic search is a technique used, for example, in natural language processing, to search the database 42 for highly relevant information (knowledge) based on the context and meaning of a query, even if the query is ambiguous. Natural language processing is a search process using, for example, Word2Vec semantic search or BERT similar sentence search. The search unit 43A searches the database 42 for knowledge corresponding to the query conditions, such as the content of the incident and countermeasures, and transfers the search result, knowledge, to the server-side control unit 44.

[0031] The determination unit 43B determines whether the numerical value of the inspection result corresponding to the inspection result information is abnormal, for example, using machine learning processing. An abnormality is a numerical abnormality that occurs when the numerical value of the inspection result deviates from a threshold or range. If the numerical value of the inspection result is abnormal, the search unit 43A generates a semantic search query condition corresponding to the situation in which the numerical value is abnormal, and searches the database 42 for knowledge corresponding to the query condition. For example, if the ammeter to be inspected is abnormal and exceeds 100 A, the search unit 43A generates a semantic search query condition for searching for knowledge for a situation in which the ammeter exceeded 100 A.

[0032] The determination unit 43B determines whether a time-series change in the numerical values ​​of the inspection results corresponding to the inspection result information is abnormal, for example, using machine learning processing. An abnormal time-series change is, for example, an abnormality when the progression of the numerical values ​​changes suddenly. When the time-series change in the numerical values ​​of the inspection results is abnormal, the search unit 43A generates a query condition for semantic search corresponding to the situation where the time-series change in the numerical values ​​is abnormal, and searches the database 42 for knowledge corresponding to the query condition. For example, when the time-series change in the ammeter to be inspected is abnormal, where the current was 50 A the day before but has now exceeded 100 A, the search unit 43A generates a query condition for semantic search for knowledge related to the situation where the time-series change from 50 A to 100 A.

[0033] The server-side control unit 44 has an alarm unit 44A. The alarm unit 44A outputs an alarm of knowledge that is a search result of the knowledge management unit 43 to the tablet terminal 2. The alarm unit 44A determines whether or not the knowledge that is a search result of the knowledge management unit 43 is valid. The validity of knowledge is determined by whether or not the knowledge corresponding to the inspection result information that is a search result of the knowledge management unit 43 is highly relevant to the inspection target. If the knowledge is valid, the alarm unit 44A outputs an alarm of the knowledge to the tablet terminal 2. Furthermore, if the knowledge is not valid, the alarm unit 44A prohibits the output of an alarm of the knowledge to the tablet terminal 2.

[0034] When the server-side control unit 44 receives the evaluation result of the knowledge related to the inspection result from the tablet terminal 2, it transfers the received evaluation result of the knowledge to the knowledge management unit 43. The acquisition unit 43C of the knowledge management unit 43 acquires the evaluation result of the knowledge related to the inspection result.

[0035] The learning unit 43D re-learns the query conditions used to search for the knowledge of the inspection results according to the evaluation result of the knowledge of the inspection results acquired by the acquisition unit 43C. For example, if the evaluation result of the knowledge of the inspection results is the highest rating of "5", the learning unit 43D re-learns the query conditions so that the knowledge of the inspection results will be ranked higher in the search. By re-learning the query conditions, it is possible to improve the accuracy of subsequent knowledge searches.

[0036] <Input processing behavior> 2 is a sequence diagram showing an example of the processing operation of the inspection support system 1 related to input processing. The input processing is a process of registering the inspection result information input from the tablet terminal 2 in the database 42, performing a semantic search for knowledge related to the inspection result information in the database 42, and notifying the tablet terminal 2 of the knowledge result of the search.

[0037] In the inspection support system 1, an inspector brings a tablet terminal 2 to the site and inspects the equipment to be inspected. In FIG. 2, the tablet control unit 26 inputs inspection result information, such as the status and temperature of the equipment to be inspected confirmed on-site, through the tablet input unit 22 (step S11). An example of the inspection result information input in step S11 is "There is an abnormal noise coming from the horizontal epitaxial growth device." For ease of explanation, "horizontal epitaxial growth device" is an abbreviation for "Yokogawa Electric's induction heating power supply for epitaxial growth device," but it is assumed that the inspector inputs the abbreviation through the tablet input unit 22.

[0038] The tablet-side control unit 26 notifies the server-side control unit 44 in the server device 4 of the input inspection result information via the communication network 5 (step S11A). As described above, the inspection result information notified in step S11A is "There is an abnormal noise coming from the horizontal epitaxial film." The server-side control unit 44 then processes the received inspection result information, transfers the processed inspection result information to the database 42 (step S12), and registers the inspection result information in the database 42 (step S12A).

[0039] The server-side control unit 44 generates search data for past cases of the received inspection result information and transfers the search data for past cases to the knowledge management unit 43 (step S13). The search data generated in step S13 is "There is an abnormal noise coming from the horizontal epi."

[0040] When the search unit 43A in the knowledge management unit 43 receives the search data of past cases, it generates a query condition for semantic search from the search data of past cases using natural language processing (step S14). In step S14, the search unit 43A recognizes that "Yokoepi" means "induction heating power supply for epitaxial growth apparatus manufactured by Yokogawa Electric Corporation," and generates a query condition that "There is an abnormal noise coming from the induction heating power supply for epitaxial growth apparatus manufactured by Yokogawa Electric Corporation."

[0041] The search unit 43A requests query conditions for the semantic search from the database 42 (step S15), and searches the database 42 for knowledge of inspection result information corresponding to the query conditions (step S16). In step S16, the search unit 43A obtains search results for knowledge related to the query condition "There is an abnormal noise coming from the induction heating power supply for an epitaxial growth device manufactured by Yokogawa Electric Corporation," such as "1. An abnormal noise was heard from the Yokogawa epitaxial power supply on February 10, 2001. The solution was..." and "2. An abnormal noise was heard from the Yokogawa epitaxial on March 4, 2020. No solution was found and it recovered naturally."

[0042] The search unit 43A transfers the searched knowledge of the inspection result information to the server-side control unit 44 (step S17). In step S17, taking into consideration the display screen of the tablet-side output unit 24, it may be configured to transfer only the knowledge of the search results for which there is a solution, such as "A strange noise was heard from the Yokogawa epitaxial power supply on February 10, 2001. The solution is..."

[0043] The notification unit 44A in the server-side control unit 44 analyzes the validity of the knowledge of the inspection result information (step S17A). Based on the analysis result of the validity of the knowledge, the notification unit 44A outputs the knowledge to the tablet terminal 2 as a notification (step S18). Then, the tablet-side control unit 26 of the tablet terminal 2 outputs the knowledge of the inspection result information from the server-side control unit 44 as a notification from the tablet-side output unit 24 (step S19). In step S19, for example, the notification content is "An abnormal noise was detected from the Yokogawa epitaxial power supply on February 10, 2001. The corrective action is..." The inspector can check the content of the knowledge of the inspection result output from the tablet-side output unit 24, confirm the content of the incident of the inspection result, and take measures to deal with the incident of the inspection result.

[0044] The tablet-side control unit 26 inputs an evaluation according to the content of the knowledge of the inspection result through the evaluation input unit 23 (step S20). As the evaluation in step S20, for example, an evaluation of "4" is input. The tablet-side control unit 26 notifies the server-side control unit 44 of the knowledge evaluation according to the content of the knowledge (step S20A). The server-side control unit 44 transfers the knowledge evaluation to the knowledge management unit 43 (step S21).

[0045] The acquisition unit 43C in the knowledge management unit 43 receives a knowledge evaluation according to the content of the knowledge. The learning unit 43D in the knowledge management unit 43 re-learns the query conditions for semantic search to search for knowledge of the inspection results according to the knowledge evaluation of the inspection results (step S22). In step S22, according to the knowledge evaluation of "4", the query conditions that will bring the knowledge "A strange noise was heard from the Yokogawa epitaxial power supply on February 10, 2001. The solution is..." to the top of the search results are re-learned.

[0046] 3 is a flow diagram showing an example of the processing operation of the tablet control unit 26 related to tablet input processing. In FIG. 3, the tablet control unit 26 determines whether or not the inspection result of the inspection target has been input via the tablet input unit 22 (step S51).

[0047] When the tablet-side control unit 26 receives the inspection result of the inspection target (step S51: Yes), it notifies the server device 4 of the inspection result (step S52). The tablet-side control unit 26 also determines whether or not knowledge of the inspection result has been received from the server device 4 (step S53). The knowledge of the inspection result includes the name of the device to be inspected, the inspection result, the details of the incident, and countermeasures.

[0048] When the tablet-side control unit 26 receives the knowledge of the inspection result (step S53; Yes), it outputs the knowledge of the inspection result to the tablet-side output unit 24 (step S54). As a result, the inspector can view the content of the knowledge of the inspection result output by the tablet-side output unit 24 and perform the corresponding work for the inspection result.

[0049] Furthermore, the tablet-side control unit 26 determines whether or not a knowledge evaluation has been input through the evaluation input unit 23 (step S55). If a knowledge evaluation has been input (step S55: Yes), the tablet-side control unit 26 transfers the knowledge evaluation according to the evaluation input to the server device 4 (step S56), and ends the processing operation shown in FIG.

[0050] If the tablet-side control unit 26 has not input the inspection result for the inspection target (step S51: No), it ends the processing operation shown in Fig. 3. If the tablet-side control unit 26 has not received knowledge of the inspection result (step S53: No), it ends the processing operation shown in Fig. 3. If the tablet-side control unit 26 has not input an evaluation of the knowledge (step S55: No), it ends the processing operation shown in Fig. 3.

[0051] 4 is a flow diagram showing an example of the processing operation of the server device 4 related to the inspection result registration processing. In FIG. 4, the server-side control unit 44 in the server device 4 determines whether or not inspection result information has been received from the tablet terminal 2 (step S61).

[0052] When the server-side control unit 44 receives the inspection result information from the tablet terminal 2 (step S61: Yes), it registers the inspection result information in the database 42 (step S62). After registering the inspection result information in the database 42, the server-side control unit 44 notifies the knowledge management unit 43 of search data of past cases related to the inspection results (step S63).

[0053] The search unit 43A in the knowledge management unit 43 generates query conditions for semantic search from the search data of past cases using natural language processing (step S64). The search unit 43A requests the generated query conditions from the database 42 (step S65). The search unit 43A determines whether knowledge corresponding to the query conditions has been searched for in the database 42 (step S66).

[0054] If the search unit 43A finds knowledge of the inspection results that corresponds to the query conditions (step S66: Yes), the search unit 43A transfers the found knowledge of the inspection results to the server-side control unit 44 (step S67).

[0055] The notification unit 44A in the server-side control unit 44 analyzes the validity of the received knowledge of the inspection results (step S68). The validity is determined based on whether the content of the received knowledge is highly related to the inspection results. The notification unit 44A determines whether the knowledge of the inspection results is valid based on the analysis result (step S69).

[0056] If the knowledge of the inspection result is valid (step S69: Yes), the notification unit 44A notifies the tablet terminal 2 of the knowledge of the inspection result (step S70).

[0057] Furthermore, the server-side control unit 44 determines whether or not a knowledge evaluation has been received from the tablet terminal 2 (step S71). If a knowledge evaluation has been received (step S71: Yes), the server-side control unit 44 transfers the knowledge evaluation to the knowledge management unit 43 (step S72).

[0058] The learning unit 43D in the knowledge management unit 43 re-learns the query conditions for the semantic search based on the knowledge evaluation received from the server-side control unit 44 so that the knowledge of the inspection results will be ranked higher in the search (step S73), and then terminates the processing operation shown in Figure 4.

[0059] If the server-side control unit 44 has not received inspection result information from the tablet terminal 2 (step S61: No), it terminates the processing operation shown in Fig. 4. If the knowledge of the inspection result is invalid (step S69: No), it terminates the processing operation shown in Fig. 4. If the server-side control unit 44 has not received a knowledge evaluation from the tablet terminal 2 (step S71: No), it terminates the processing operation shown in Fig. 4. If the search unit 43A has not searched for knowledge of the inspection result that corresponds to the query condition (step S66: No), it terminates the processing operation shown in Fig. 4.

[0060] The server device 4 generates query conditions for a semantic search using natural language processing that correspond to the inspection result information (natural language), and searches the database 42 for knowledge that corresponds to the query conditions. The server device 4 then determines whether the searched knowledge is valid, and if the knowledge is valid, outputs a notification of the knowledge to the tablet terminal 11. As a result, the inspector can view the knowledge of the inspection results (contents of the incident and countermeasures) on-site while inspecting the equipment to be inspected, and respond to the inspection results.

[0061] The server device 4 acquires the evaluation result of the knowledge corresponding to the inspection result information from the tablet terminal 11, and re-learns the query condition corresponding to the inspection result information based on the acquired evaluation result, thereby improving the accuracy of searching for knowledge for the inspection result.

[0062] In step S64, a case where a query condition for a semantic search is generated using natural language processing when the inspection result is a natural sentence is illustrated. However, the inspection result is not limited to a natural sentence, and may be a numerical value only. In this case, the determination unit 43B of the knowledge management unit 43 determines whether the numerical value of the inspection result is abnormal, for example, using machine learning processing. If the numerical value of the inspection result is abnormal, the search unit 43A generates a query condition for a semantic search corresponding to the situation where the numerical value is abnormal. As a result, knowledge corresponding to the situation where the numerical value of the inspection result is abnormal can be searched.

[0063] Furthermore, the determination unit 43B determines whether the time-series change in the numerical values ​​of the inspection results is abnormal, for example, using machine learning processing. If the time-series change in the numerical values ​​of the inspection results is abnormal, the search unit 43A generates a query condition for semantic search corresponding to the situation where the time-series change in the numerical values ​​is abnormal. As a result, it is possible to search for knowledge corresponding to the situation where the time-series change in the numerical values ​​of the inspection results is abnormal.

[0064] <Unresolved incident registration process> 5 is a sequence diagram showing an example of the processing operation of the inspection support system 1 related to the unresolved incident registration processing. The unresolved incident registration processing is a processing in which the content of the incident and the countermeasures for the inspection result information are input from the PC terminal 3, and the content of the incident and the countermeasures for the inspection result information are registered in the database 42.

[0065] 5, for example, when the server-side control unit 44 registers the inspection result information in the database 42, it notifies the PC terminal 3 of an unresolved incident registration request (step S31). Note that the unresolved incident registration request may be, for example, "An unresolved incident has been registered. The content is 'Liquid leakage has occurred from the XX valve.'"

[0066] The PC-side control unit 35 in the PC terminal 3 outputs the inspection result information as an unresolved incident registration request to the PC-side output unit 33. As a result, the user of the PC terminal 3 checks the inspection result information and considers the details of the incident and countermeasures as a result of the inspection. The PC-side control unit 35 then inputs incident information including the details of the incident and countermeasures as a result of the consideration via the PC-side input unit 32 (step S32), and transfers the input incident information to the server-side control unit 44 (step S33).

[0067] When the server-side control unit 44 receives the incident information, it transfers the incident information to the database 42 in association with the inspection result (step S34), and registers the incident information (contents and countermeasures) of the inspection result in the database 42 (step S35). As a result, the contents of the incident and countermeasures of the inspection result can be registered in the database 42 from the PC terminal 3.

[0068] 6 is a flow diagram showing an example of the processing operation of the server device 4 related to the unresolved incident registration processing. In FIG. 6, after the inspection result information from the tablet terminal 2 is registered in the database 42, the server-side control unit 44 in the server device 4 determines whether or not the inspection result is an unresolved incident (step S81).

[0069] If the inspection result is an unresolved incident (step S81: Yes), the server-side control unit 44 notifies the PC terminal 3 of a request to register the unresolved incident of the inspection result (step S82).

[0070] The server-side control unit 44 determines whether or not the content of the unresolved incident and the countermeasures have been received from the PC terminal 3 (step S83). If the content of the unresolved incident and the countermeasures have been received from the PC terminal 3 (step S83: Yes), the server-side control unit 44 registers the content of the incident and the countermeasures of the received inspection results in the database 42 (step S84), and ends the processing operation shown in FIG.

[0071] If the inspection result is not an unresolved incident (step S81: No), the server-side control unit 44 ends the processing operation shown in Fig. 6. If the server-side control unit 44 has not received the details of the unresolved incident and the countermeasures from the PC terminal 3 (step S83: No), the server-side control unit 44 ends the processing operation shown in Fig. 6.

[0072] <Search process operation> 7 is a sequence diagram showing an example of the processing operation related to search processing of the inspection support system 1. The search processing is a natural language search process in which a semantic search is performed on the database 42 from the PC terminal 3 using keywords of natural language processing, and the search results are returned to the PC terminal 3.

[0073] 7, the PC-side control unit 35 inputs a query request in natural language via the PC-side input unit 32 and notifies the server-side control unit 44 of the input query request in natural language (step S41). An example of the query request in natural language in step S41 is, for example, "There is a strange noise coming from the horizontal epi."

[0074] When the server-side control unit 44 receives a query request in natural language from the PC-side control unit 35, it transfers the query request in natural language to the knowledge management unit 43 (step S42). An example of the query request in natural language in step S42 is "There is a strange noise coming from the horizontal epi."

[0075] When a query request in natural language is detected from the server-side control unit 44, the search unit 43A in the knowledge management unit 43 uses natural language processing to generate query conditions for semantic search from the search data of the query request in natural language (step S43). In step S43, the search unit 43A recognizes that "Yokoepi" means "induction heating power supply for epitaxial growth apparatus manufactured by Yokogawa Electric Corporation," and generates a query condition that "an abnormal noise is coming from the induction heating power supply for epitaxial growth apparatus manufactured by Yokogawa Electric Corporation."

[0076] The search unit 43A requests query conditions for the semantic search from the database 42 (step S44), and searches the database 42 for knowledge corresponding to the query conditions (step S45). In step S45, the search unit 43A obtains search results for knowledge related to the query condition "There is an abnormal noise coming from the induction heating power supply for an epitaxial growth device manufactured by Yokogawa Electric Corporation," such as "1. An abnormal noise was coming from the Yokogawa epitaxial power supply on February 10, 2001. The solution was..." and "2. An abnormal noise was generated from the Yokogawa epitaxial on March 4, 2020. No solution was found and it recovered naturally."

[0077] The search unit 43A transfers knowledge corresponding to the searched query conditions to the server-side control unit 44 (step S46). In step S46, for example, knowledge such as "1. On February 10, 2001, an abnormal noise was heard from the Yokogawa epitaxial power supply. The solution was to..." and "2. On March 4, 2020, an abnormal noise was heard from the Yokogawa epitaxial power supply. No solution was found and it was fixed naturally." is transferred.

[0078] The server-side control unit 44 transfers the knowledge corresponding to the query conditions to the PC-side control unit 35 (step S47). The PC-side control unit 35 outputs the knowledge corresponding to the query conditions transferred from the server-side control unit 44 to the PC-side output unit 33 (step S48). In step S48, for example, knowledge such as "1. An unusual noise was heard from the Yokogawa epitaxial power supply on February 10, 2001. The solution was..." and "2. An unusual noise was heard from the Yokogawa epitaxial power supply on March 4, 2020. No solution was found and it was fixed naturally" are output from the PC-side output unit 33. As a result, the user of the PC terminal 3 can recognize the knowledge that is the query result in response to the query request output from the PC-side output unit 33.

[0079] 8 is a flow diagram showing an example of a processing operation related to search processing of the server device 4. In Fig. 8, the server-side control unit 44 determines whether or not a query request in natural language has been received from the PC terminal 3 (step S91).

[0080] When the server-side control unit 44 receives a query request (step S91: Yes), it notifies the knowledge management unit 43 of the search data of the query request (step S92). The search unit 43A in the knowledge management unit 43 generates query conditions for semantic search from the search data using natural language processing or machine learning (step S93). The search unit 43A requests the generated query conditions from the database 42 (step S94). The search unit 43A determines whether knowledge corresponding to the query conditions has been searched for from the database 42 (step S95).

[0081] If the search unit 43A finds the knowledge of the query request corresponding to the query condition (step S95: Yes), the search unit 43A transfers the acquired knowledge of the query request to the server-side control unit 44 (step S96).

[0082] The notification unit 44A in the server-side control unit 44 analyzes the validity of the knowledge of the received query request (step S97). Based on the analysis result, the notification unit 44A determines whether the knowledge of the query request is valid (step S98).

[0083] If the knowledge of the inquiry request is valid (step S98: Yes), the server-side control unit 44 notifies the PC terminal 3 of the knowledge of the inquiry request (step S99), and ends the processing operation shown in FIG.

[0084] If the server-side control unit 44 has not received a query request in natural language from the PC terminal 3 (step S91: No), the server-side control unit 44 ends the processing operation shown in Fig. 8. If the server-side control unit 44 has not searched for knowledge corresponding to the query conditions (step S95: No), the server-side control unit 44 notifies the PC terminal 3 of knowledge NG (step S100) and ends the processing operation shown in Fig. 8. Note that knowledge NG is, for example, a message indicating that the search was not successful. If the knowledge in the query request is not valid (step S98: No), the server-side control unit 44 proceeds to the processing of step S100 to notify the PC terminal 3 of knowledge NG. Note that knowledge NG is, for example, a message indicating that the knowledge is not valid.

[0085] The server device 4 generates a query condition for a semantic search using natural language processing in response to a query request in natural language, and searches the database 42 for knowledge corresponding to the query condition. The server device 4 then determines whether the searched knowledge is valid, and if the knowledge is valid, outputs the knowledge to the PC terminal 3. As a result, the user of the PC terminal 3 can recognize the knowledge corresponding to the query request.

[0086] In this embodiment, the server device 4 generates a query condition corresponding to the inspection result information input from the tablet terminal 2, searches the database 42 for knowledge corresponding to the query condition, and notifies and outputs the searched knowledge to the tablet terminal 2. As a result, the inspector on the tablet terminal 2 can check the knowledge related to the inspection result information and immediately respond to the inspection result on-site, thereby reducing the workload. Furthermore, even if the inspector does not have much experience or skills, he or she can recognize the content of the incident and countermeasures on-site, and quickly resolve the problem or prevent the problem from occurring.

[0087] The server device 4 generates query conditions for semantic search using natural language processing corresponding to the inspection result information, and searches for natural language knowledge corresponding to the query conditions from the database 42. As a result, the inspector can check the natural language knowledge related to the inspection result information and immediately respond to the inspection results on-site, thereby reducing the workload.

[0088] The server device 4 uses machine learning processing to determine whether the numerical value of the inspection result corresponding to the inspection result information is abnormal. If the numerical value of the inspection result is abnormal, the server device 4 generates a query condition for semantic search corresponding to the situation in which the numerical value is abnormal, and searches the database 42 for knowledge corresponding to the query condition. As a result, even if the inspection result is only a numerical value, if the numerical value is abnormal, the inspector can check the knowledge in natural language related to the situation in which the numerical value is abnormal and immediately respond to the inspection result on site, thereby reducing the workload.

[0089] The server device 4 uses machine learning processing to determine whether or not the time-series change in the numerical values ​​of the inspection results corresponding to the inspection result information is abnormal. If the time-series change in the numerical values ​​of the inspection results is abnormal, the server device 4 generates a query condition for semantic search corresponding to the situation in which the time-series change is abnormal, and searches the database 42 for knowledge corresponding to the query condition. As a result, even if the inspection results are only numerical values, if the time-series change in the numerical values ​​is abnormal, the inspector can check the natural language knowledge related to the situation in which the time-series change is abnormal and immediately respond to the inspection results on-site, thereby reducing the workload.

[0090] The server device 4 acquires the evaluation result of the knowledge corresponding to the inspection result information from the tablet terminal 2, and re-learns the query condition corresponding to the inspection result information based on the acquired evaluation result of the knowledge. As a result, it is possible to improve the accuracy of searching for knowledge for the inspection result.

[0091] The server device 4 determines whether the searched knowledge is valid, and if the knowledge is valid, outputs a notification of the knowledge to the tablet terminal 2. As a result, the inspector can obtain valid knowledge.

[0092] Furthermore, the inspection support system 1 can also predict failures by detecting time-series changes in the numerical values ​​and character strings of the inspection results.The inspection support system 1 is equipped with a knowledge evaluation function, which gradually prevents irrelevant past cases from being displayed.

[0093] In the inspection support system 1, during patrol inspections, in addition to inspections using existing numerical values, character strings and registered standard phrases can also be entered, allowing the system to search for relevant cases from natural language such as registered knowledge and manuals, and by reporting them, it can present prevention and countermeasures for accidents / breakdowns, etc.

[0094] For ease of explanation, a manual input method has been described as an example of a method for inputting inspection result information, including the inspection results of the equipment to be inspected, into the server device 4, in which the inspection result information is input into the server device 4 from the tablet input unit 22 of the tablet terminal 2. However, the input method is not limited to the manual input method, and an automatic input method, semi-automatic input method, or the like may also be used, and can be changed as appropriate. The automatic input method is a method in which data is sent directly from the equipment to be inspected to the server device 4, and the inspection results are automatically input into the server device 4. The semi-automatic input method is a method in which the inspection results are semi-automatically input into the server device 4 using, for example, QR Code (registered trademark), NFC (Near Field Communication), Bluetooth (registered trademark), instrument reading AI, or the like.

[0095] If the inspection result is a change in a character string, or if a change in the text over time is detected, such as a similar phenomenon (text) being recorded every week, the knowledge management unit 43 generates a query condition related to that phenomenon. Then, the knowledge management unit 43 notifies the tablet terminal 2 of knowledge corresponding to the query condition.

[0096] When the inspection result is a hybrid of character strings and numbers, the knowledge management unit 43 generates query conditions for past cases, etc., using a hybrid search method that combines character strings and numbers instead of using only character strings or numbers. Then, the knowledge management unit 43 notifies the tablet terminal 2 of knowledge that corresponds to the query conditions.

[0097] The knowledge management unit 43 can perform searches and notifications with fewer omissions by incorporating not only numerical values ​​but also a hybrid of strings and numerical values, natural language processing technology, and semantic searches in addition to simple string searches. Furthermore, by using machine learning processing, it is possible to predict failures by detecting changes in the time series of numerical values ​​and strings.

[0098] Also, the notification unit 44A has been exemplified as notifying the tablet terminal 2 of the knowledge of the inspection results, but this is not limited to the tablet terminal 2, and the knowledge of the inspection results may be notified to the PC terminal 3, and other methods can be used. Furthermore, other methods can be used, such as calling a telephone or radio and notifying using a voice synthesizer, or notifying by email, chat, and other methods, and other methods can be used.

[0099] In addition, although an example has been given of registering the details of an incident and countermeasures in the database 42 from a PC terminal 3, this is not limited to this, and for example, registration may also be possible from a tablet terminal 2, and changes can be made as appropriate.

[0100] Furthermore, all or any part of the processing functions performed by each device may be realized by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware using wired logic.

[0101] The various processes described in this embodiment can be realized by executing a program prepared in advance on an information processing device. Therefore, an example of an information processing device that executes a program having the same functions as those in the above embodiment will be described below. Fig. 9 is an explanatory diagram showing an example of an information processing device that executes an information processing program.

[0102] 9 includes a communication unit 201, a hard disk drive (HDD) 202, a read only memory (ROM) 203, and a random access memory (RAM) 204. The information processing device 200 further includes a CPU 205 and a bus 206.

[0103] The information processing device 200 is, for example, a computer. The HDD 102 is a database that registers inspection result information including inspection results of inspection targets.

[0104] An information processing program that performs the same functions as those of the above-described embodiment is pre-stored in ROM 203. The information processing program may be stored in a recording medium that can be read by a drive (not shown) instead of ROM 203. The recording medium may be, for example, a portable recording medium such as a CD-ROM, DVD, USB memory, or SD card, or a semiconductor memory such as a flash memory. As shown in FIG. 9, the information processing program is a management program 203A and a notification program 203B. The management program 203A and the notification program 203B may be integrated or separated as appropriate.

[0105] Then, the CPU 205 reads out these programs 203A and 203B from the ROM 203, and deploys and executes each of the read programs on the RAM 204. Then, as shown in Fig. 9, the CPU 205 executes each of the programs 203A and 203B on the RAM 204, thereby functioning as a management process 204A and a notification process 204B.

[0106] The CPU 205 generates a query condition corresponding to the inspection result information and searches for knowledge corresponding to the query condition from the HDD 202. The CPU 205 notifies the inspector terminal device of the searched knowledge. As a result, the inspector at the inspector terminal device can check the knowledge related to the inspection result information and immediately respond to the inspection result on-site, thereby reducing the workload.

[0107] In addition, some examples of combinations in the techniques of the present disclosure are described below.

[0108] (1) An information processing device that registers inspection result information, including inspection results of inspection objects, in a database, and has: a knowledge management unit that generates query conditions corresponding to the inspection result information and searches the database for knowledge corresponding to the query conditions; and a notification unit that notifies an inspector terminal device of the knowledge searched by the knowledge management unit.

[0109] (2) The information processing device described in (1), wherein the knowledge management unit has an acquisition unit that acquires an evaluation result of the knowledge corresponding to the inspection result information from the inspector terminal device, and a learning unit that re-learns the query condition corresponding to the inspection result information based on the evaluation result acquired by the acquisition unit.

[0110] (3) The information processing device according to (1) or (2), wherein the knowledge management unit generates a query condition for a semantic search corresponding to the inspection result information and searches the database for the knowledge corresponding to the query condition.

[0111] (4) An information processing device described in any one of (1) to (3), wherein the knowledge management unit generates a query condition for semantic search using natural language processing corresponding to the inspection result information, and searches the database for the knowledge corresponding to the query condition.

[0112] (5) An information processing device described in any one of (1) to (4), wherein the knowledge management unit determines whether the numerical value of the inspection result corresponding to the inspection result information is abnormal, generates a semantic search query condition corresponding to a situation in which the numerical value of the inspection result is abnormal, and searches the database for the knowledge corresponding to the query condition.

[0113] (6) An information processing device described in any one of (1) to (5), wherein the knowledge management unit determines whether the time series change of the numerical value of the inspection result corresponding to the inspection result information is abnormal, generates a semantic search query condition corresponding to a situation in which the time series change of the numerical value of the inspection result is abnormal, and searches the database for the knowledge corresponding to the query condition.

[0114] (7) An information processing device described in any one of (1) to (6), wherein the notification unit determines whether the knowledge searched by the knowledge management unit is valid, and if the knowledge is valid, outputs the knowledge to the inspector terminal device.

[0115] (8) The information processing device according to any one of (1) to (7), wherein the information processing device inputs the inspection result information from the inspector terminal device.

[0116] (9) An information processing method in which an information processing device registers inspection result information including inspection results of an inspection object in a database, generates query conditions corresponding to the inspection result information, searches the database for knowledge corresponding to the query conditions, and notifies the inspector terminal device of the searched knowledge, thereby executing each of the following processes.

[0117] (10) An information processing program that causes a processor of an information processing device that registers inspection result information, including inspection results of inspection objects, in a database to execute the following processes: generate query conditions corresponding to the inspection result information, search the database for knowledge corresponding to the query conditions, and notify the inspector terminal device of the searched knowledge. [Explanation of symbols]

[0118] 1. Inspection support system 2. Tablet devices 3 PC terminals 4. Server equipment 23 Evaluation input section 42 databases 43 Knowledge Management Department 43A Search Department 43B Judgment section 43C Acquisition Department 43D Learning Department 44 Server-side control unit 44A Notification Department

Claims

1. An information processing device that registers inspection result information including inspection results of inspection objects in a database, a knowledge management unit that generates a query condition corresponding to the inspection result information and searches the database for knowledge corresponding to the query condition; a notification unit that notifies the knowledge searched by the knowledge management unit to an inspector terminal device; An information processing device comprising:

2. The knowledge management unit an acquisition unit that acquires an evaluation result of the knowledge corresponding to the inspection result information from the inspector terminal device; a learning unit that re-learns the query condition corresponding to the inspection result information based on the evaluation result acquired by the acquisition unit; 2. The information processing apparatus according to claim 1, further comprising:

3. The knowledge management unit The information processing apparatus according to claim 1 , further comprising: generating a query condition for a semantic search corresponding to the inspection result information; and searching the database for the knowledge corresponding to the query condition.

4. The knowledge management unit 4. The information processing apparatus according to claim 3, wherein a query condition for the semantic search using natural language processing corresponding to the inspection result information is generated, and the knowledge corresponding to the query condition is searched for from the database.

5. The knowledge management unit 5. The information processing device according to claim 4, wherein the information processing device determines whether or not a numerical value of the inspection result corresponding to the inspection result information is abnormal, generates a query condition for the semantic search corresponding to a situation in which the numerical value of the inspection result is abnormal, and searches the database for the knowledge corresponding to the query condition.

6. The knowledge management unit 5. The information processing device according to claim 4, wherein the information processing device determines whether a time series change in a numerical value of the inspection result corresponding to the inspection result information is abnormal, generates a query condition for the semantic search corresponding to a situation in which the time series change in the numerical value of the inspection result is abnormal, and searches the database for the knowledge corresponding to the query condition.

7. The notification unit 2. The information processing apparatus according to claim 1, wherein the knowledge management unit determines whether the searched knowledge is valid, and if the knowledge is valid, outputs a notification of the knowledge to the inspector terminal device.

8. The information processing device includes:

2. The information processing apparatus according to claim 1, wherein the inspection result information is input from the inspector terminal device.

9. an information processing device that registers inspection result information including inspection results of inspection objects in a database; generating a query condition corresponding to the inspection result information, and searching the database for knowledge corresponding to the query condition; The retrieved knowledge is output to the inspector terminal device. An information processing method characterized by executing each process.

10. A processor of an information processing device that registers inspection result information including inspection results of inspection objects in a database, generating a query condition corresponding to the inspection result information, and searching the database for knowledge corresponding to the query condition; The retrieved knowledge is output to the inspector terminal device. An information processing program that causes each process to be executed.

Images