Recommendation method, recommendation program, and recommendation device

Abstract

To improve work efficiency of users.SOLUTION: A recommendation device disclosed herein comprises an acquisition unit, identification unit, determination unit, and presentation unit. The acquisition unit acquires trouble information on troubles occurring in a plant. The identification unit identifies maintenance plans associated to troubles similar to troubles indicated by the trouble information from among pieces of maintenance plan information accumulated in advance in association with troubles on the basis of a trend of maintenance plans selected by a user in the past. The presentation unit presents the maintenance plans identified through the identification process.SELECTED DRAWING: Figure 4

Description

Technical Field

[0001] The present invention relates to a recommendation method, a recommendation program, and a recommendation device.

Background Art

[0002] Techniques for assisting a user's decision-making based on history are known. For example, in the field of content distribution via the Internet, a technique for recommending content that is expected to be liked by the user based on the user's viewing history is used. Also, in the field of retail, a technique for recommending products that are likely to be purchased by the user based on the user's purchase history is used.

[0003] Techniques for assisting a user's decision-making based on such history are also useful in the industrial field. For example, a technique for automatically generating a maintenance plan for industrial machines in a plant based on the results of inspections performed in the past is known. The plant is, for example, an oil plant, a petrochemical plant, a chemical plant, or a gas plant. In a plant, facilities including industrial machines operate, and chemical products and the like are produced.

[0004] Furthermore, a technique for updating failure diagnosis rules for maintenance parts for restoring equipment failures from maintenance case information accumulated in the past based on newly added maintenance cases is known.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0006] However, conventional technologies have a problem in that they sometimes fail to improve the efficiency of users' work.

[0007] For example, by prioritizing and recommending maintenance plans that are realistically feasible for a given defect from among several effective maintenance plans, users can quickly make decisions about which maintenance plan to adopt.

[0008] Conversely, if a maintenance plan is recommended that is effective in addressing a problem but is practically difficult to implement, the user will have to perform extra work such as narrowing down the recommended maintenance plan.

[0009] On the other hand, conventional technology cannot update the rules regarding which conservation plans should be prioritized and recommended.

[0010] One aspect of this project is to provide recommendation methods, recommendation programs, and recommendation devices that can improve the efficiency of users' work. [Means for solving the problem]

[0011] One aspect of the recommendation method is characterized in that a computer acquires malfunction information, which is information about malfunctions that have occurred in the plant, identifies maintenance plans that are similar to the malfunctions indicated by the malfunction information from among the maintenance plan information that has been stored in advance and associated with the malfunctions, based on the trends of maintenance plans that have been previously selected by the user, and performs a process to present the maintenance plan identified by the identification process.

[0012] The recommendation program relating to one aspect is characterized by having a computer acquire malfunction information, which is information about malfunctions that have occurred in the plant, identify maintenance plans that are similar to the malfunctions indicated by the malfunction information from among the maintenance plan information that has been stored in advance and associated with the malfunctions, based on the trends of maintenance plans that the user has selected in the past, and then execute a process to present the maintenance plan identified by the identification process.

[0013] The recommendation device for one side has an acquisition unit that acquires defect information which is the information of defects generated in the plant, a specification unit that specifies, based on the tendency of the maintenance plan selected by the user in the past, a maintenance plan associated with a defect similar to the defect indicated by the defect information from the information of the maintenance plan stored in advance in association with the defect, and a presentation unit that presents the maintenance plan specified by the specification unit. It is characterized by having these components.

Effect of the Invention

[0014] According to one embodiment, the efficiency of the user's work can be improved.

Brief Description of the Drawings

[0015] [Figure 1] It is a diagram showing a configuration example of a recommendation system according to the first embodiment. [Figure 2] It is a diagram showing the flow of the recommendation process of the recommendation system. [Figure 3] It is a diagram showing the flow of the update process of the recommendation system. [Figure 4] It is a diagram showing a configuration example of the recommendation device. [Figure 5] It is a diagram showing an example of a defect DB. [Figure 6] It is a diagram showing an example of a maintenance plan DB. [Figure 7] It is a diagram showing an example of a danger information DB. [Figure 8] It is a diagram showing an example of a selection history DB. [Figure 9] It is a diagram showing an example of a defect report screen. [Figure 10] It is a diagram showing an example of a maintenance plan search screen. [Figure 11] It is a diagram showing an example of a rating table. [Figure 12] It is a diagram showing an example of a rating table. [Figure 13] It is a diagram showing an example of a maintenance plan search screen. [Figure 14]This figure shows an example of a maintenance plan search screen. [Figure 15] This figure shows an example of a graph display screen. [Figure 16] This flowchart shows the flow of the recommendation process for the recommendation device. [Figure 17] This is a flowchart showing the process for updating recommended devices. [Figure 18] This is a diagram illustrating an example hardware configuration. [Modes for carrying out the invention]

[0016] The embodiments of the recommendation method, recommendation program, and recommendation apparatus disclosed herein will be described in detail below with reference to the drawings. However, the invention of this application is not limited by the embodiments described herein. Furthermore, the same elements are denoted by the same reference numerals, and redundant descriptions are omitted as appropriate. Also, each embodiment can be combined as appropriate within the bounds of consistency.

[0017] The configuration of the recommendation system according to the first embodiment will be explained using Figure 1. Figure 1 is a diagram showing an example of the configuration of the recommendation system according to the first embodiment.

[0018] As shown in Figure 1, the recommendation system 1 includes a recommendation device 10, an administrator terminal device 20, an engineer terminal device 30, and an information provision server 40. The recommendation device 10 and the information provision server 40 are, for example, server devices. The administrator terminal device 20 and the engineer terminal device 30 are, for example, stationary or portable terminal devices such as personal computers and smartphones.

[0019] The recommendation device 10, the administrator terminal device 20, the engineer terminal device 30, and the information provision server 40 are connected to each other via network N. For example, network N is the internet. Alternatively, network N may be a LAN (Local Area Network) built within a facility such as a plant.

[0020] Here, Recommendation System 1 is a system for recommending maintenance plans for a plant. For example, Recommendation System 1 presents multiple maintenance plans for a plant malfunction. The plant manager can then actually adopt one of the presented maintenance plans. In this way, Recommendation System 1 can support the plant manager's decision-making and improve the efficiency of the manager's work.

[0021] The plant in this embodiment is, for example, an oil plant, a petrochemical plant, a chemical plant, or a gas plant. When the plant is in operation, products such as LNG (liquefied natural gas), resins (plastics, nylon, etc.) and chemical products are obtained.

[0022] Furthermore, the plant includes factory facilities, machinery facilities, production facilities, power generation facilities, storage facilities, and facilities at the wellhead for extracting oil, natural gas, etc. In addition to equipment for generating products, the plant is also equipped with field instruments for acquiring information about the plant's condition.

[0023] Field devices include sensor devices, operating devices, and alarm devices. For example, sensor devices include pressure sensors, temperature sensors, pH sensors, speed sensors, and acceleration sensors. For example, operating devices include valves, pumps, and fans, which are driven by motors and actuators. For example, alarm devices include lamps and speakers.

[0024] Various malfunctions can occur in a plant. These malfunctions are abnormalities in facilities, equipment, field instruments, etc., and are events that have an adverse effect on the plant. These adverse effects include the cessation and delay of product production, and increased costs (e.g., monetary costs, human costs, time costs).

[0025] When a malfunction occurs, the plant manager develops a maintenance plan to resolve the problem. This maintenance plan may include, for example, repairing, cleaning, or replacing parts of the malfunctioning equipment. Note that maintenance plans may also be implemented to prevent malfunctions, regardless of whether a malfunction has occurred.

[0026] Figure 2 shows the flow of the recommendation system's recommendation process. Administrator U2 and Engineer U3 in Figure 2 are users. Administrator U2 develops maintenance plans for plant malfunctions. Engineer U3 operates the plant equipment. Engineer U3 also reports any plant malfunctions they discover.

[0027] Recommendation System 1 receives bug reports from Engineer U3. Based on the nature of the bug, Recommendation System 1 also presents potential maintenance plans to Administrator U2.

[0028] As shown in Figure 2, engineer U3 reports the plant malfunction to the recommendation device 10 via the engineer's terminal device 30 (step S1). For example, the engineer's terminal device 30 transmits the malfunction information entered by engineer U3 to the recommendation device 10.

[0029] Furthermore, malfunction reports are not limited to engineer U3; they may also be made by other users, including administrator U2. For example, administrator U2 can report plant malfunctions to the recommendation device 10 via the administrator terminal device 20.

[0030] When a malfunction is reported, administrator U2 inputs additional information via the administrator terminal device 20 (step S2). This additional information is used to determine the priority of the maintenance plan candidates presented by the recommendation device 10.

[0031] Furthermore, if a malfunction is reported, the information provision server 40 provides additional information to the recommendation device 10 (step S3). For example, the information provision server 40 provides additional information in response to a request from the recommendation device 10.

[0032] For example, additional information provided by Administrator U2 might indicate what information regarding the maintenance plan is considered important. For instance, additional information provided by Administrator U2 might indicate whether or not the costs associated with the maintenance plan are considered important, and to what extent.

[0033] On the other hand, the additional information provided by the information server 40 includes external information that is not directly related to the maintenance plan. For example, the additional information provided by the information server 40 includes weather information, information on the plant's operating plan, etc.

[0034] Next, the recommendation device 10 adjusts the method for calculating similarity based on the selection history (step S4). The selection history is information that stores which maintenance plans were selected from among the maintenance plans recommended by the recommendation system 1. For example, the recommendation device 10 updates the weights for calculating similarity based on the selection history. Details of the method for calculating and adjusting similarity will be described later.

[0035] Subsequently, the recommendation device 10 identifies maintenance plans similar to the reported defect using the adjusted similarity calculation method (step S5). First, the recommendation device 10 extracts defects similar to the reported defect from the information that associates defect information with maintenance plans, which has been stored in advance. Furthermore, the recommendation device 10 identifies maintenance plans associated with the extracted defects.

[0036] The information linking the malfunction information to the maintenance plan may be stored in the recommendation device 10 in advance, or it may be input to the recommendation device 10 when step S4 is executed.

[0037] Here, the recommendation device 10 determines the priority of the identified maintenance plan based on the additional information obtained from the administrator U2 and the information provision server 40 (step S6). Then, the recommendation device 10 presents the maintenance plan according to the priority (step S7).

[0038] The presented maintenance plan is selected by the user (e.g., administrator U2). The recommendation device 10 updates the selection history according to the selected maintenance plan.

[0039] Figure 3 shows the flow of the update process for the recommendation system. As shown in Figure 3, administrator U2 selects one of the presented maintenance plans (step S11). For example, administrator U2 performs an operation to select one of the maintenance plans displayed on the administrator terminal device 20 (e.g., by clicking on an option).

[0040] The recommendation device 10 reflects the selection result made by administrator U2 in the selection history (step S12). For example, the recommendation device 10 adds the current selection result to a table or the like that stores the selection results.

[0041] [Functional Configuration] The configuration of the recommendation device 10 will be explained using Figure 4. Figure 4 is a diagram showing an example of the configuration of the recommendation device. As shown in Figure 4, the recommendation device 10 has a communication unit 11, a storage unit 12, and a control unit 13. Note that the functional units of the recommendation device 10 are not limited to those shown in the figure, and may also have functional units such as an interface for sending and receiving data with output devices such as a display and a speaker.

[0042] The communication unit 11 is a processing unit that controls communication with other devices, and is implemented, for example, by a communication interface. For example, the communication unit 11 controls communication with the administrator terminal device 20, the engineer terminal device 30, and the information provision server 40.

[0043] The storage unit 12 is a processing unit that stores various data and various programs executed by the control unit 13. The storage unit 12 is implemented by, for example, memory and a hard disk. This storage unit 12 stores various data generated in the processing executed by the recommendation device 10, such as data obtained in the process of the control unit 13 executing various processes and processing results obtained as a result of executing various processes.

[0044] The memory unit 12 stores the malfunction DB 121, maintenance plan DB 122, hazard information DB 123, and selection history DB 124.

[0045] The defect database 121 stores information about defects that have occurred. Figure 5 shows an example of the defect database. As shown in Figure 5, the defect database 121 includes the following items: "Defect ID", "Date and Time", "Location", "Team", "Severity", "Problem", and "Detailed Information".

[0046] The "Defect ID" field is information used to identify the defect. The "Date and Time" field is the date and time the defect occurred, or the date and time the defect information was registered in Defect DB121. The "Location" field indicates the location within the plant where the defect occurred. The "Team" field indicates the team associated with the facility, equipment, or device affected by the defect.

[0047] The "Importance" field indicates the severity of the issue. For example, "Importance" can be one of the following: "Very High," "High," "Medium," or "Low."

[0048] The "Problem" field provides an overview of the bug and is described by the user who registers (or reports) the bug. The "Details" field provides detailed information about the bug and is described by the user who registers (or reports) the bug.

[0049] For example, Figure 5 shows that the malfunction identified as "DF0151" occurred at the date and time "2022 / 9 / 1 13:01" in location "Area_1", the team responsible for the malfunction is "Team_1", ​​and the severity is "low". Figure 5 also shows that the summary of the malfunction identified as "DF0151" is "gas leak", and the detailed information is "gas leak in Area_1".

[0050] The maintenance plan DB122 stores information on completed maintenance plans. Figure 6 shows an example of the maintenance plan DB. As shown in Figure 6, the maintenance plan DB122 includes the following items: "Maintenance Plan ID", "Date and Time", "Type of Failure", "Cause of Failure", "Solution", "Component", "Required Preparations", "Safety Requirements", "Number of Personnel", "Time Required", "Cost", and "Failure ID".

[0051] The "Maintenance Plan ID" field is information used to identify the maintenance plan. The "Date and Time" field is the date and time the maintenance plan was implemented (the start or completion date and time), or the date and time the maintenance plan information was registered in the Maintenance Plan DB122.

[0052] The item "Type of Defect" indicates the type of defect corresponding to the maintenance plan. For example, a defect corresponding to the maintenance plan is a defect that the maintenance plan aims to resolve, or a defect that is expected to be prevented from occurring by the maintenance plan. The item "Cause" indicates the cause of the defect corresponding to the maintenance plan. The item "Solution" indicates the specific solution for the defect corresponding to the maintenance plan. The item "Component" is a component that has a close relationship with the maintenance plan or the defect corresponding to the maintenance plan.

[0053] The item "Required Preparations" contains information on the items, skills, and procedures necessary for the maintenance plan. The item "Safety Requirements" contains the safety requirements that must be met in the maintenance plan. The item "Number of Personnel" is the number of personnel required to implement the maintenance plan (personnel cost). The item "Time Required" is the time required to implement the maintenance plan (time cost). The item "Cost" is the cost incurred to implement the maintenance plan (monetary cost). The item "Defect ID" is information used to associate the maintenance plan with the defect.

[0054] For example, Figure 6 shows that the maintenance plan identified as "MT0211" was implemented on the date and time "2022 / 9 / 1 14:21". Figure 6 also shows that the maintenance plan identified as "MT0211" corresponds to the malfunction identified as "DF0151".

[0055] Furthermore, Figure 6 shows that the type of malfunction corresponding to the maintenance plan identified as "MT0211" is "leakage," the cause is "aging deterioration," the solution is "repair and improvement," and the component is "valve."

[0056] Furthermore, Figure 6 shows that the preparations required for the maintenance plan identified as "MT0211" are "scaffolding installation, containers," and the safety requirements are "gas detection, harness wearing."

[0057] Furthermore, Figure 6 shows that the maintenance plan identified as "MT0211" required "5 personnel," took "6 hours," and cost "¥1,500,000."

[0058] Hazard Information DB123 stores information on hazards related to the subject of the maintenance plan. Figure 7 shows an example of the Hazard Information DB. As shown in Figure 7, Hazard Information DB123 includes the items "Subject," "Level of Hazard," "Type of Hazard," and "Safety Requirements."

[0059] The "Target" item is information used to identify the object that indicates a hazard. For example, the "Target" item is information that identifies the location, facility, equipment, machinery, or component within a plant.

[0060] The "Risk Level" item indicates the magnitude of the risk involved. For example, the "Risk Level" can be one of the following: "High," "Medium," or "Low."

[0061] The item "Type of Hazard" contains information about the types of hazards that may occur in the subject. The item "Safety Requirements" contains the safety requirements that are required depending on the type of hazard.

[0062] For example, Figure 7 shows that the risk level of "Area_1" is "medium," the type of risk that may occur is related to "height," and the safety requirement is "wearing a harness."

[0063] The selection history DB124 stores the selection results of maintenance plans. Figure 8 shows an example of the selection history DB. As shown in Figure 8, the selection history DB124 includes the following items: "Date and Time", "Maintenance Plan ID", "User Skill Level", "Overall Ranking", "Number of Personnel Ranking", "Required Time Ranking", and "Cost Ranking".

[0064] The "Date and Time" field is the date and time the maintenance plan was selected. The "Maintenance Plan ID" field is information used to identify the maintenance plan.

[0065] "User proficiency" refers to the proficiency level of the user who selected the maintenance plan. Here, proficiency refers to the proficiency in developing the maintenance plan, and is assumed to be stored separately in a master table that manages user information, for example.

[0066] Furthermore, the "overall ranking," "number of personnel ranking," "required time ranking," and "cost ranking" are the rankings at the time the conservation plan was presented.

[0067] As described below, the recommendation device 10 determines the priority of each identified maintenance plan. The recommendation device 10 then presents maintenance plans according to the determined priority. Maintenance plans with higher priority are more likely to be selected by the user. For example, when the recommendation device 10 presents a list of maintenance plans, it places maintenance plans with higher priority higher up in the list. As a result, maintenance plans with higher priority are more likely to be selected by the user.

[0068] The "overall ranking" is a ranking based on priority. The higher the priority, the higher the "overall ranking."

[0069] As explained in Figure 6, the maintenance plan includes items such as "number of personnel," "time required," and "cost." The "ranking by number of personnel," "ranking by time required," and "ranking by cost" indicate the ranking of the "number of personnel," "time required," and "cost" of the maintenance plan within the presented maintenance plans. Note that the "ranking by number of personnel," "ranking by time required," and "ranking by cost" are higher (higher) as the "number of personnel," "time required," and "cost" of the maintenance plan decrease.

[0070] For example, Figure 8 shows that the maintenance plan identified as "MT0211" was selected on "2022 / 8 / 1 10:11" by a user with a skill level of "medium". Figure 6 also shows that the maintenance plan "MT0211" had an overall ranking of "1st", a personnel ranking of "1st", a required time ranking of "2nd", and a cost ranking of "1st".

[0071] For example, Figure 8 shows that the maintenance plan identified as "MT0215" was selected on "2022 / 8 / 10 15:56" by a user with a "high" skill level. Figure 6 also shows that the maintenance plan "MT0215" had an overall ranking of "3rd", a personnel ranking of "3rd", a required time ranking of "5th", and a cost ranking of "1st".

[0072] Here, maintenance plan "MT0211" was likely selected by the user because it was ranked "1st" overall. On the other hand, maintenance plan "MT0215" was selected despite being ranked "5th" overall. Furthermore, maintenance plan "MT0215" was selected despite being ranked "1st" in cost, even though it was ranked "5th" overall. From this, it can be inferred that the user prioritized cost over overall priority when selecting maintenance plan "MT0215".

[0073] In the example in Figure 8, one selection result is shown for each maintenance plan "MT0211," "MT0213," and "MT0215," but multiple selection results may be registered for each maintenance plan in the selection history DB124.

[0074] Returning to Figure 4, the control unit 13 is the processing unit that oversees the entire recommendation device 10. The control unit 13 is implemented by, for example, a processor. The control unit 13 includes a display control unit 131, an acquisition unit 132, a specification unit 133, a determination unit 134, and an update unit 135.

[0075] The display control unit 131 controls the display of the screen. The display control unit 131 can display the screen on the administrator terminal device 20 and the engineer terminal device 30.

[0076] For example, the display control unit 131 functions as a web server. That is, the display control unit 131 generates data (for example, an HTML file) for displaying the screen. The display control unit 131 then transmits the generated data to the administrator terminal device 20 and the engineer terminal device 30.

[0077] The administrator terminal device 20 and the engineer terminal device 30 display a screen based on the data received from the display control unit 131 in a web browser or the like.

[0078] Furthermore, the administrator terminal 20 and the engineer terminal 30 may display their screens using a dedicated application for the recommendation system 1 instead of a web browser. In this case, the display control unit 131 transmits data in a format compatible with the application.

[0079] The acquisition unit 132 acquires defect information, which is information about defects that have occurred in the plant, and additional information that is different from the defect information.

[0080] The defect information is the defect information reported to the recommendation system 1. For example, the defect information is input via a defect report screen that the display control unit 131 displays on the engineer's terminal device 30.

[0081] The bug reporting screen will be explained using Figure 9. Figure 9 is a diagram showing an example of the bug reporting screen.

[0082] As shown in Figure 9, the bug report screen 61 displays a pull-down list 61a, a pull-down list 61b, a pull-down list 61c, a text box 61d, a text box 61e, and a button 61f.

[0083] A dropdown list displays one or more options. The user then selects one of the options displayed by the dropdown list.

[0084] The pull-down list 61a displays "Area_1", "Area_2", "Area_3", and "Area_4" as predetermined candidate locations within the plant.

[0085] The pull-down list 61b displays "Team_1", ​​"Team_2", "Team_3", and "Team_4" as predefined team candidates.

[0086] The pull-down list 61c displays "Very High," "High," "Medium," and "Low" as predetermined importance level options.

[0087] Text boxes 61d and 61e accept text input from the user. However, the number of characters that can be entered into text box 61d is set to be less than the number of characters that can be entered into text box 61e.

[0088] When button 61f is pressed, the contents entered in each object (pull-down list and text box) on the malfunction report screen 61 are sent to the recommendation device 10. This allows the acquisition unit 132 to acquire the malfunction information.

[0089] For example, if button 61f is pressed in the state shown in Figure 9, the acquisition unit 132 acquires malfunction information where the location is "Area_2", the team is "Team_3", the severity is "Medium", the problem is "Gas leak", and the detailed information is "Carbon dioxide leaking from the tank".

[0090] When a malfunction is reported, the acquisition unit 132 acquires additional information. This additional information is used to determine the priority of the maintenance plan candidates presented by the recommendation device 10. For example, the additional information is input via a maintenance plan search screen displayed on the administrator terminal device 20 by the display control unit 131.

[0091] The maintenance plan search screen will be explained using Figure 10. Figure 10 shows an example of the maintenance plan search screen.

[0092] As shown in Figure 10, the maintenance plan search screen 62 displays a display area 62a, a text box 621b, a seek bar 622b, a text box 621c, a seek bar 622c, a text box 621d, a seek bar 622d, a checkbox 62e, a checkbox 62f, a checkbox 62g, and a button 62h.

[0093] The seek bar specifies parameters by the position of the box on the bar. The user can adjust the position of the box on the bar.

[0094] The text box 621b is where the specified value for the additional information "Number of personnel" is entered. The seek bar 622b specifies the degree of importance given to the additional information "Number of personnel".

[0095] The text box 621c is where the specified value for the additional information "allowable time" is entered. The seek bar 622c specifies the degree of importance given to the additional information "allowable time".

[0096] The text box 621d is where the specified value for the additional information "Cost" is entered. The seek bar 622d specifies the degree of importance given to the additional information "Cost".

[0097] Checkbox 62e specifies whether to consider the additional information "risk information". Checkbox 62f specifies whether to consider the additional information "external information". Checkbox 62e specifies whether to consider the additional information "past selection history".

[0098] When button 61f is pressed, the process of identifying and presenting a maintenance plan is initiated based on the malfunction information and additional information entered in the maintenance plan search screen 62.

[0099] For example, if button 62h is pressed in the state shown in Figure 10, the acquisition unit 132 acquires information that the specified values ​​for the additional information "number of personnel," "allowable time," and "cost" are not specified, the degree of importance is set to the default value (100%), and the additional information "risk information," "external information," and "past selection history" are not considered.

[0100] The specific unit 133 identifies a maintenance plan that is similar to the defect indicated by the defect information, from among the maintenance plan information that has been stored in advance and associated with the defect.

[0101] Here, the specific unit 133 refers to the selection history DB 124 and calculates the weights (hereinafter referred to as item scores) for each of the "number of personnel," "required time," and "cost" of the maintenance plan. The item scores affect the calculation of similarity. The specific unit 133 adjusts the similarity calculation method based on the calculation of the item scores.

[0102] First, the identification unit 133 narrows down the selection results used to calculate the item score using the "Date and Time" field in the selection history DB 124. For example, the identification unit 133 narrows down the selection results to those where the "Date and Time" field falls within the past three months from the current date and time. Alternatively, the identification unit 133 may use the selection results for the entire period to calculate the item score.

[0103] For example, if the selection history DB 124 shows a tendency for cost to be given importance, the specific unit 133 will increase the cost item score. Also, the specific unit 133 will increase the influence of the selection result on the item score the higher the skill level of the user who selected the maintenance plan.

[0104] For example, the identification unit 133 refers to the selection results (records in the selection history DB 124) and calculates the sum of each item by subtracting the rank of each item from the overall rank. Then, the identification unit 133 multiplies the sum of the items by a coefficient according to the user's proficiency level to obtain the item score. For example, if the user's proficiency level is "high", the identification unit 133 multiplies the sum of the items by "2", if the user's proficiency level is "medium", it multiplies the sum of the items by "1", and if the user's proficiency level is "low", it multiplies the sum of the items by "1".

[0105] For example, based on the first row of Figure 8 (maintenance plan "MT0211"), the specific unit 133 calculates the item score for "number of personnel" as (1-1) × 1.5 = 0, the item score for "time required" as (1-2) × 1.5 = -1.5, the item score for "time required" as (1-2) × 1.5 = -1.5, and the cost score as (1-1) × 1.5 = 0.

[0106] For example, based on the second row of Figure 8 (maintenance plan "MT0213"), the specific unit 133 calculates the item score for "number of personnel" as (1-2)×1=-1, the item score for "time required" as (1-2)×1.5=-1.5, the item score for "time required" as (1-2)×1.5=-1.5, and the cost score as (1-1)×1.5=0.

[0107] Thus, the identification unit 133 can identify a maintenance plan based on the trends in the number of personnel, allowable time, and cost of maintenance plans previously selected by the user. Furthermore, the identification unit 133 can identify a maintenance plan based on the skill level of the user who previously selected the maintenance plan.

[0108] Next, the specific unit 133 converts the defect information and the defect DB 121 into a rating table. The rating table will be explained using Figures 11 and 12. Figures 11 and 12 are diagrams showing an example of a rating table.

[0109] Here, we assume that the location of the malfunction information is "Area_2", the team is "Team_3", the severity is "Medium", the problem is "Gas leak", and the details are "Carbon dioxide leaking from the tank".

[0110] The specific unit 133 obtains the rating table shown in Figure 11 by converting the defect information. As shown in Figure 11, the rating table reflects the "location," "team," and "severity" included in the defect information.

[0111] In the rating table, items included in the defect information have a value of "5," while items not included in the defect information have a value of "1."

[0112] Since the location of the defect information is "Area_2", the values ​​for "Area_1", "Area_3", and "Area_4" in the rating table are "1", and the value for "Area_2" is "5".

[0113] Rating tables are represented as vectors. For example, the rating table shown in Figure 11 corresponds to the vector (1,5,1,1,1,1,5,1,1,1,5,1).

[0114] Furthermore, the specific unit 133 obtains the rating table shown in Figure 12 by converting the defect DB 121. Figure 12 shows the rating tables for defects "DF0151" and "DF0152".

[0115] The rating table for defect "DF0151" corresponds to the vector (5,1,1,1,5,1,1,1,1,1,1,5). The rating table for defect "DF0152" corresponds to the vector (5,1,1,1,1,1,5,1,1,1,1,5).

[0116] The identification unit 133 calculates the similarity between the rating table obtained by converting the defect information and the rating table obtained by converting the defect DB 121. At this time, the identification unit 133 treats the rating tables as vectors and calculates the cosine similarity between the vectors using equation (1).

[0117]

number

[0118] However, A and B are vectors for which similarity is calculated. Also, for example, A i is the i-th element of vector A. n is the number of elements in the vector. In the examples in Figures 11 and 12, n = 12.

[0119] Furthermore, the specific unit 133 calculates the similarity between defects using TF-IDF as shown in equation (2), based on the "problem" and "detailed information" sections of the defect information and the defect DB 121, which are text information that has not been converted into a rating table. Here, the "problem" and "detailed information" for each defect are considered to be documents.

[0120]

number

[0121] However, t is the number of occurrences of the target word in the "Problem" and "Detailed Information" sections of the defect information. n is the number of documents stored in the defect DB121. df is the number of documents stored in the defect DB121 that contain the target word.

[0122] Let the similarity calculated by equation (1) be Sim1, and the similarity calculated by equation (2) be Sim2. The identification unit 133 then calculates the total similarity S D Calculate it as shown in equation (3).

[0123]

number

[0124] However, w is a pre-set weight. w is a value between 0 and 1, and may be, for example, 0.5.

[0125] Note that the method for calculating the similarity between defects is not limited to that described here. The specific unit 133 can calculate the similarity between defects using any known method for calculating the similarity between data.

[0126] Furthermore, the specific part 133 is S D The final similarity is calculated based on the pre-calculated item scores. For example, the specific unit 133 ensures that items with larger total item scores have a greater impact on the similarity.

[0127] For example, the identification unit 133 sums the item scores for each item and identifies the item with the highest total item score. Then, the smaller the value of the identified item, the higher the similarity score.

[0128] For example, if the item with the highest total item score is "cost", the specific unit 133 will determine the S of each defect. D The final similarity is calculated by multiplying this by a coefficient that increases as the cost of the corresponding conservation plan decreases (for example, the reciprocal of the cost).

[0129] Note that the similarity calculated by the specific unit 133 here is not the pure similarity between the data, but for convenience, we will refer to it as similarity here.

[0130] The identification unit 133 extracts a certain number of defects from the defect DB 121 in order of decreasing similarity. For example, the identification unit 133 extracts approximately 3 to 20 defects.

[0131] Furthermore, the identification unit 133 identifies the maintenance plan corresponding to the extracted defect from among the maintenance plans stored in the maintenance plan DB 122. For example, the identification unit 133 identifies the maintenance plan corresponding to the defect based on the "Defect ID" item in the maintenance plan DB 122.

[0132] The determination unit 134 determines the priority of each maintenance plan identified by the identification unit 133 based on the additional information. The display control unit 131 then presents the maintenance plans according to their priority.

[0133] For example, if the item with the highest total item score is "Expenses", S D Even if the defect was the largest, if the "cost" of the corresponding maintenance plan is very large, it may not be extracted by the specific unit 133. Conversely, S D Even if the value becomes very large and indicates a malfunction, if the "cost" of the corresponding maintenance plan is very small, it may be extracted by the specific unit 133.

[0134] Furthermore, the factors that should be considered in relation to user selection tendencies are not limited to "number of personnel," "time required," and "cost."

[0135] For example, the identification unit 133 identifies a maintenance plan based on the risk trend associated with the maintenance plan. The identification unit 133 can obtain the risk level of the maintenance plan corresponding to the malfunction by referring to the risk information DB 123.

[0136] Furthermore, the identification unit 133 can calculate an item score for "risk level" by referring to the maintenance plan DB 122 and the selection history DB and comparing the overall ranking with the risk level ranking. If the user shows a tendency to place importance on risk level, the identification unit 133 can reflect the risk level in the similarity calculation.

[0137] Since accidents during the implementation of a maintenance plan pose a particularly high risk, the specific unit 133 may always reflect the degree of risk in the similarity calculation, making it less likely for users to be presented with high-risk maintenance plans. In this case, the specific unit 133 may drastically reduce the similarity in accordance with the degree of risk, ignoring the selection trends of other items.

[0138] Once the priority is determined by the decision unit 134, the display control unit 131 displays the button 62i and the search result 62j on the maintenance plan search screen 62, as shown in Figure 13. Figure 13 is a diagram showing an example of the maintenance plan search screen.

[0139] The search result 62j displays the maintenance plans identified by the identification unit 133 in descending order of priority determined by the decision unit 134. In the example in Figure 13, the identification unit 133 identified the maintenance plans "MT0213," "MT0211," and "MT0215." The decision unit 134 then determined the highest priority for "MT0213," followed by "MT0211" and "MT0215," in descending order of priority.

[0140] The acquisition unit 132 acquires requests entered by the user, or constraints generated based on those requests, as additional information. Users can enter requests via the maintenance plan search screen 62. The acquisition unit 132 generates constraints based on the input content of the maintenance plan search screen 62.

[0141] For example, the acquisition unit 132 acquires, as additional information, the degree to which each of the personnel number, allowable time, and cost associated with the maintenance plan is given importance. It is desirable that the personnel number, allowable time, and cost all be small.

[0142] For example, if the values ​​of the maintenance plan (number of personnel, allowable time, or cost) exceed the specified values ​​entered on the maintenance plan search screen 62, the decision unit 134 will reduce the priority of the maintenance plan. In this case, the decision unit 134 will reduce the priority further the larger the value indicated by the seek bar.

[0143] Here, let a1 be the number of personnel required for a certain maintenance plan, a2 be the required time, and a3 be the cost. The decision unit 134 calculates the priority of this maintenance plan as 1 / ((w1×a1)×(w2×a2)×(w3×a3)).

[0144] The seek bar displays values ​​from 0% to 200% (0 to 2). The determination unit 134 sets weights w1, w3, and w3 to be equal to the values ​​indicated by the seek bar. However, if a1, a2, and a3 do not exceed the specified value, the determination unit 134 limits weights w1, w3, and w3 to 1 or less, and if a1, a2, and a3 exceed the specified value, it may further increase weights w1, w3, and w3. In other words, the determination unit 134 considers the specified value as the upper limit, and if the value of an item exceeds the upper limit, it penalizes the weight (for example, by increasing the degree of increase).

[0145] In other words, when a specified value is given, the decision unit 134 suppresses the reduction in priority for items in the maintenance plan, such as the number of personnel, allowable time, or cost, that do not exceed the specified value, while significantly reducing the priority for items that do exceed the specified value.

[0146] In the example in Figure 13, the specified cost value "1,000,000" is entered. The seek bar values ​​for the number of personnel and allowable time are "100%". The seek bar value for cost is "200%".

[0147] For example, the values ​​for each item in the maintenance plan "MT0123" are (a1, a2, a3) = (4, 8, 800000). Therefore, in the example in Figure 13, the decision unit 134 calculates the priority by setting (w1, w2, w3) = (1, 1, 1). In this case, since the cost does not exceed the specified value, the decision unit 134 sets the cost weight w3 to the upper limit value of "1".

[0148] For example, the values ​​for each item in the maintenance plan "MT0121" are (a1, a2, a3) = (5, 6, 1500000). Therefore, in the example in Figure 13, the decision unit 134 calculates the priority by setting (w1, w2, w3) = (1, 1, 4). In this case, since the cost exceeds the specified value, the decision unit 134 sets the cost weight w3 to "4", which is twice "2".

[0149] Furthermore, if a specified value is entered, the determination unit 134 may change the upper limit depending on the user's degree of importance (the value of the seek bar). For example, if the value of the seek bar is "100%", the determination unit 134 will use the specified value as the upper limit, as explained above.

[0150] On the other hand, the determination unit 134 may change the upper limit value according to the value of the seek bar. For example, the determination unit 134 may decrease the upper limit value (increase the penalty) as the value of the seek bar increases. For example, if the value of the seek bar is r (0 ≤ r²), the determination unit 134 calculates the upper limit value as "specified value × (1 + 0.2 × (1 - r))". In this case, the upper limit value will vary between 0.8 times and 1.2 times the specified value.

[0151] For example, if the specified time is "10 hours" and the seek bar value is "0%", then r=0, and the determination unit 134 calculates the upper limit as "10 hours × (1 + (0.2 × (1 - 0%))) = 12 hours".

[0152] For example, if the specified cost is "1,000,000" and the seek bar value is "200%", then r=2, and the determination unit 134 calculates the upper limit as "1,000,000 × (1 + (0.2 × (1-2))) = 800,000".

[0153] The decision unit 134 can determine priority by considering information about hazards. The acquisition unit 132 acquires information about hazards associated with the maintenance plan as additional information.

[0154] If checkbox 62e is checked, the acquisition unit 132 refers to the hazard information DB 123 and obtains the hazard level corresponding to the maintenance plan. For example, the location of defect "DF0151" corresponding to maintenance plan "MT0123" is "Area_1". Therefore, the decision unit 134 obtains "Medium" as the hazard level for maintenance plan "MT0123" from the hazard information DB 123. The decision unit 134 reduces the priority of the maintenance plan the higher the obtained hazard level.

[0155] The decision unit 134 can determine priority based on the number of times each maintenance plan has been used in the past and the skill level of the user who used it. In this case, the acquisition unit 132 acquires information based on the number of times each maintenance plan has been used in the past and the skill level of the user who used it as additional information.

[0156] If checkbox 62f is checked, the acquisition unit 132 acquires the selection history for each maintenance plan from the selection history DB 124.

[0157] The decision unit 134 prioritizes maintenance plans that have been selected more frequently or more often by users with a high level of skill.

[0158] Furthermore, the decision unit 134 can acquire additional information not only from the user but also from external sources. For example, the acquisition unit 132 can acquire weather information or plant operation plans as additional information from the information provision server 40.

[0159] For example, if weather information indicates rain or a typhoon, the decision unit 134 may lower the priority of maintenance plans that involve outdoor work. In that case, the decision unit 134 may also increase the priority of maintenance plans that can be completed before the rain or typhoon begins.

[0160] Furthermore, for example, if the plant's operation plan indicates that the maintenance plan for the entire plant will be carried out within a certain number of days, the decision unit 134 will reduce the priority of maintenance plans that are similar to those included in the overall maintenance plan.

[0161] Furthermore, for example, the decision unit 134 determines whether or not degraded operation of the plant is permitted for a certain period, based on the plant's operational plan and the inventory status of replacement parts. The decision unit 134 may also obtain from the information provision server 40 the periods during which degraded operation of the plant is permitted and periods during which it is not permitted. The decision unit 134 then reduces the priority of maintenance plans that involve degraded operation during periods when degraded operation is not permitted.

[0162] Furthermore, for example, if the plant's operational plan indicates that product production cannot be stopped, the decision unit 134 will prioritize maintenance plans that do not require the plant to be shut down. The decision unit 134 determines whether or not it is necessary to shut down the plant based on items such as "necessary preparations" and "safety requirements" in the maintenance plan.

[0163] Furthermore, for example, the decision unit 134 prioritizes emergency maintenance plans during periods when plant shutdown and reduced-scale operation are not permitted. For example, regardless of user input, the decision unit 134 prioritizes emergency maintenance plans by emphasizing the required time.

[0164] In this way, the decision unit 134 obtains the user's requests via the maintenance plan search screen 62. The seek bar value and specified value are obtained as the user's requests. The decision unit 134 also generates constraint conditions based on the obtained user requests. The constraint conditions are conditions for determining the priority according to the user's requests.

[0165] Furthermore, the determination unit 134 may change the priority according to the item score calculated by the identification unit 133. For example, the determination unit 134 increases the priority of items with a larger total item score by decreasing the weights w1, w3, and w3 type. This allows items that users tend to value to have a greater influence on the priority.

[0166] Figure 14 shows an example of a maintenance plan search screen. As shown in Figure 14, if additional information different from that in Figure 13 is entered, the priority calculated by the determination unit 134 changes, and the display order of maintenance plans in the search results 62j changes.

[0167] In this way, the display control unit 131 presents maintenance plans according to priority. The display control unit 131 displays maintenance plans with higher priority higher up in the list. The display control unit 131 may also hide maintenance plans with a priority below a certain threshold. Furthermore, the display control unit 131 may highlight a certain number of maintenance plans with higher priority by differentiating the font and style.

[0168] When the button 62i shown in Figures 13 and 14 is pressed, the display control unit 131 causes the administrator terminal device 20 to display a graph display screen. Figure 15 shows an example of the graph display screen.

[0169] As shown in Figure 15, in the graph display screen 63, shapes representing maintenance plans such as "MT0211," "MT0213," and "MT0215" are placed on a coordinate plane represented by two axes.

[0170] The horizontal axis in the graph display screen 63 represents the cost of the maintenance plan. The vertical axis in the graph display screen 63 represents the man-hours obtained by multiplying the number of personnel and the required time for the maintenance plan. The tolerance value is a value calculated based on additional information. The display control unit 131 may use the specified value entered in the maintenance plan search screen 62 as the tolerance value, or it may use a value calculated from the specified value as the tolerance value.

[0171] Furthermore, the density of the patterns in each figure representing the maintenance plan indicates the degree of risk. The display control unit 131 displays a higher density of patterns for maintenance plans with a greater degree of risk, based on the risk information DB 123.

[0172] According to the graph display screen 63, the user can identify the maintenance plan with the smallest values ​​for items such as the number of personnel, required time, and cost, regardless of priority. In addition, the display control unit 131 may display the maintenance plan with the smallest values ​​for items such as the number of personnel, required time, and cost on the maintenance plan search screen 62, separately from the search results 62j.

[0173] The update unit 135 records information about the maintenance plan selected by the user from among the maintenance plans presented by the processing. Specifically, the update unit 135 updates the selection history DB 124 according to the selection result.

[0174] For example, the priority of search result 62j in Figure 13 corresponds to the overall ranking. Also, let's assume that the three items "MT0211", "MT0213", and "MT0215" were identified by the identification unit 133.

[0175] For example, if "MT0211" is selected in Figure 13, the update unit 135 adds a record to the selection history DB 124 with the following characteristics: "Maintenance Plan ID" is "MT0211", "Overall Ranking" is "2nd", "Number of Personnel Ranking" is "2nd", "Required Time Ranking" is "1st", and "Cost Ranking" is "2nd".

[0176] [Process Flow] Figure 16 is a flowchart showing the flow of the recommendation process of the recommendation device. As shown in Figure 16, first, the recommendation device 10 acquires defect information (step S101). The defect information is input, for example, via the engineer's terminal device 30.

[0177] Next, the recommendation device 10 acquires additional information (step S102). For example, the processing from step S102 onward is initiated in response to a request from the administrator terminal device 20.

[0178] The recommendation device 10 converts the acquired fault history and accumulated maintenance plan history into a rating table (step S103).

[0179] Here, the recommendation device 10 calculates the weight of each element to be used for similarity calculation (for example, the item score for each item) based on the selection history DB 124 (step S104). Then, the recommendation device 10 extracts defects similar to the defects indicated by the defect information based on the rating table (step S105). Furthermore, the recommendation device 10 identifies the maintenance plan associated with the extracted defects (step S106).

[0180] Here, the recommendation device 10 determines the priority of each identified maintenance plan based on the additional information (step S107). For example, the recommendation device 10 determines the priority based on the number of personnel, time required, and cost of the maintenance plan.

[0181] The recommendation device 10 then presents the identified maintenance plans according to their priority (step S108). For example, the recommendation device 10 places maintenance plans with higher priority higher up in the list displayed on the screen.

[0182] Figure 17 is a flowchart showing the flow of the recommendation process of the recommendation device. As shown in Figure 17, first, the recommendation device 10 receives the selection result of the maintenance plan (step S201). For example, the recommendation device 10 receives the selection result via the maintenance plan search screen 62.

[0183] Next, the recommendation device 10 adds the selection result to the selection history DB (step S202). This ensures that the selected maintenance plan will influence the maintenance plan recommended in the future.

[0184] [effect] As described above, the recommendation device 10 of the embodiment has an acquisition unit 132, a specification unit 133, and a determination unit 134. The acquisition unit 132 acquires defect information, which is information about defects that have occurred in the plant, and additional information that is different from the defect information. The specification unit 133 identifies maintenance plans that are similar to the defects indicated by the defect information from among the maintenance plan information that has been stored in advance and associated with defects. The determination unit 134 determines the priority of each identified maintenance plan based on the additional information.

[0185] In this way, the recommendation device 10 can prioritize recommending maintenance plans that are realistically feasible for the user to adopt from among multiple maintenance plans. As a result, according to this embodiment, the efficiency of the user's work is improved.

[0186] Furthermore, the acquisition unit 132 acquires additional information, which consists of requests entered by the user or constraints generated based on those requests. This allows the recommendation device 10 to reflect the user's intentions in the maintenance plan it presents.

[0187] The acquisition unit 132 acquires, as additional information, the degree to which each of the personnel number, allowable time, and cost associated with the maintenance plan is given importance. This allows the recommendation device 10 to present a maintenance plan that is advantageous to the user from the standpoint of human cost, time cost, and monetary cost.

[0188] The acquisition unit 132 acquires information on hazards associated with the maintenance plan as additional information. This allows the recommendation device 10 to prioritize and present maintenance plans that can avoid hazards.

[0189] The acquisition unit 132 acquires additional information based on the number of times each maintenance plan has been used in the past and the skill level of the user who used it. This allows the recommendation device 10 to suggest maintenance plans that have a proven track record of being selected by highly skilled users in the past.

[0190] The acquisition unit 132 acquires weather-related information as supplementary information. This allows the recommendation device 10 to prevent the maintenance plan from being unable to be implemented due to weather conditions.

[0191] The acquisition unit 132 acquires the plant's operation plan as additional information. This allows the recommendation device 10 to assist the user in selecting a maintenance plan that is suitable not only for addressing malfunctions but also for the overall operation of the plant.

[0192] As described above, the recommendation device 10 of the embodiment includes a display control unit 131, an acquisition unit 132, a specification unit 133, a determination unit 134, and an update unit 135. The acquisition unit 132 acquires defect information, which is information about defects that have occurred in the plant. The specification unit 133 identifies maintenance plans that are similar to the defects indicated by the defect information from among the maintenance plan information that has been stored in advance and associated with defects, based on the trends of maintenance plans that the user has selected in the past. The display control unit 131 presents the maintenance plan identified by the identification process.

[0193] In this way, the recommendation device 10 can present a maintenance plan by taking into account the trends of maintenance plans previously selected by the user. As a result, according to this embodiment, a maintenance plan that better matches the user's intentions can be presented, thereby improving the efficiency of the user's work.

[0194] The update unit 135 records information on the maintenance plan selected by the user from among the maintenance plans presented by the presentation process. The identification unit 133 identifies the maintenance plan based on the trends obtained from the information recorded by the recording process. In this way, the recommendation device 10 can improve the accuracy of recommending maintenance plans that match the user's intentions by acquiring the user's selection results as feedback and accumulating that feedback.

[0195] The identification unit 133 identifies a maintenance plan based on the trends in the number of personnel, allowable time, and cost of maintenance plans previously selected by the user. This allows the recommendation device 10 to present a maintenance plan that takes into account how important the user considers human costs, time costs, and monetary costs.

[0196] The identification unit 133 identifies a maintenance plan based on the risk trend associated with the maintenance plan. This allows the recommendation device 10 to present a maintenance plan that matches the user's preference for a safer maintenance plan.

[0197] The identification unit 133 identifies a maintenance plan based on the skill level of users who have previously selected a maintenance plan. This allows the recommendation device 10 to present maintenance plans that have a proven track record of being selected by highly skilled users in the past.

[0198] [system] Unless otherwise specified, the processing procedures, control procedures, specific names, and various data and parameters shown in the above documents and drawings may be changed at will.

[0199] Furthermore, the components of each illustrated device are functionally conceptual and do not necessarily need to be physically configured as shown. In other words, the specific forms of distribution and integration of each device are not limited to those shown. That is, all or part of them can be functionally or physically distributed and integrated in any unit according to various loads, usage conditions, etc.

[0200] Furthermore, each processing function performed by each device may be implemented, in whole or in part, by a CPU and a program executed for analysis by that CPU, or by wired logic hardware.

[0201] [Hardware] Next, an example of the hardware configuration of the recommendation device 10 will be described. Figure 18 is a diagram illustrating an example of the hardware configuration. As shown in Figure 18, the recommendation device 10 includes a communication device 10a, an HDD (Hard Disk Drive) 10b, memory 10c, and a processor 10d. Furthermore, each of the parts shown in Figure 18 is interconnected by a bus or the like.

[0202] The communication device 10a is a network interface card or the like, and communicates with other servers. The HDD 10b stores programs and databases that operate the functions shown in Figure 4.

[0203] The processor 10d operates a process that performs the functions described in Figure 2, etc., by reading a program that performs the same processing as each processing unit shown in Figure 4 from the HDD 10b or the like and loading it into memory 10c. For example, this process performs the same functions as each processing unit of the recommendation device 10. Specifically, the processor 10d reads a program that has the same functions as the display control unit 131, acquisition unit 132, identification unit 133, decision unit 134, and update unit 135, etc., from the HDD 10b or the like. Then, the processor 10d executes a process that performs the same processing as the display control unit 131, acquisition unit 132, identification unit 133, decision unit 134, and update unit 135, etc.

[0204] Thus, the recommendation device 10 operates as a recommendation device that executes a recommendation method by reading and executing a program. Furthermore, the recommendation device 10 can also achieve the same functionality as the embodiment described above by reading the program from the recording medium using a media reader and executing the read program. Note that the program referred to in this other embodiment is not limited to being executed by the recommendation device 10. For example, the present invention can be similarly applied when another computer or server executes the program, or when they collaborate to execute the program.

[0205] This program can be distributed via networks such as the Internet. Furthermore, this program can be recorded on computer-readable storage media such as hard disks, flexible disks (FDs), CD-ROMs, MOs (Magneto-Optical disks), and DVDs (Digital Versatile Discs), and executed by reading the program from these media using a computer.

[0206] Some examples of the combinations of technical features that will be disclosed are listed below.

[0207] (1) Computers We obtain malfunction information, which is information about malfunctions that have occurred in the plant. From the maintenance plan information that has been stored in advance and associated with defects, maintenance plans associated with defects similar to the defects indicated by the defect information are identified based on the trends of maintenance plans previously selected by the user. The maintenance plan identified by the aforementioned process is presented. A recommendation method characterized by performing a process. (2) Further processing is performed to record the information of the maintenance plan selected by the user from among the maintenance plans presented by the aforementioned processing. The recommendation method according to (1), characterized in that the process for identifying the plan identifies the maintenance plan based on the trends obtained from the information recorded by the process for recording the plan. (3) The recommendation method according to (1) or (2), characterized in that the process of identifying a maintenance plan identifies a maintenance plan based on the trends in the number of personnel, allowable time, and cost of maintenance plans previously selected by the user. (4) The recommendation method according to any one of (1) to (3), characterized in that the process to be identified identifies a maintenance plan based on the trend of risk associated with the maintenance plan. (5) The recommendation method according to any one of (1) to (4), characterized in that the process of identifying the maintenance plan identifies the maintenance plan based on the skill level of users who have previously selected a maintenance plan. (6) On the computer, We obtain malfunction information, which is information about malfunctions that have occurred in the plant. From the maintenance plan information that has been stored in advance and associated with defects, maintenance plans associated with defects similar to the defects indicated by the defect information are identified based on the trends of maintenance plans previously selected by the user. The maintenance plan identified by the aforementioned process is presented. A recommendation program characterized by executing a process. (7) An acquisition unit that acquires malfunction information, which is information about malfunctions that have occurred in the plant, A selection unit identifies maintenance plans that are similar to the malfunctions indicated by the malfunction information, based on the trends of maintenance plans previously selected by the user, from among the maintenance plan information that has been stored in advance and associated with malfunctions. A presentation unit that presents the maintenance plan identified by the aforementioned specific unit, A recommendation device characterized by having the following features. [Explanation of symbols]

[0208] 10 Recommendation device 11 Communications Department 12 Storage section 13 Control Unit 61 Bug Report Screen 62. Maintenance Plan Search Screen 63 Graph display screen 61a, 61b, 61c dropdown list 61d, 61e, 621b, 621c, 621d Text boxes 61f, 62h, 62i buttons 62e, 62f, 62g checkboxes 622b, 622c, 622d seek bar 62j search results 121 Defective Database 122 Conservation Plan Database 123 Hazard Information Database 124 Selection History DB 131 Display Control Unit 132 Acquisition Department 133 Specific part 134 Decision Section 135 Update Department

Claims

1. Computers We obtain malfunction information, which is information about the first malfunction that occurred in the plant. Based on the ranking of personnel, required time, and cost of maintenance plans that have been pre-stored and associated with each of the multiple defects, a score for personnel, required time, and cost is calculated for each of them, the similarity between the first defect and each of the multiple defects is calculated, and based on the similarity, a maintenance plan corresponding to the defect extracted from the multiple defects is identified. The maintenance plan identified by the aforementioned process is presented. A recommendation method characterized by performing a process.

2. The recommendation method according to claim 1, wherein the process for identifying the problem involves calculating a score for the degree to which each of the number of personnel, required time, and cost is given importance, based on the number of personnel, required time, and cost of multiple maintenance plans previously selected by the user, and identifying a maintenance plan associated with a problem similar to the problem indicated by the problem information, based on the score.

3. Further processing is performed to record the information of the maintenance plan selected by the user from among the maintenance plans presented by the aforementioned processing. The recommendation method according to claim 1, characterized in that the identifying process identifies a maintenance plan based on trends obtained from the information recorded by the recording process.

4. The recommendation method according to claim 1, characterized in that the process to be identified identifies the maintenance plan based on the trend of risk associated with the maintenance plan.

5. On the computer, We obtain malfunction information, which is information about the first malfunction that occurred in the plant. Based on the ranking of personnel, required time, and cost of maintenance plans that have been pre-stored and associated with each of the multiple defects, a score for personnel, required time, and cost is calculated for each of them, the similarity between the first defect and each of the multiple defects is calculated, and based on the similarity, a maintenance plan corresponding to the defect extracted from the multiple defects is identified. The maintenance plan identified by the aforementioned process is presented. A recommendation program characterized by executing a process.

6. An acquisition unit that acquires information about the first malfunction that occurred in the plant, A special unit calculates scores for the number of personnel, required time, and cost of maintenance plans that have been pre-stored and associated with each of the multiple defects, calculates the similarity between the first defect and each of the multiple defects, and identifies the maintenance plan corresponding to the defect extracted from the multiple defects based on the similarity. A presentation unit that presents the maintenance plan identified by the aforementioned specific unit, A recommendation device characterized by having the following features.

Images