JP7313887B2 - Maintenance improvement support system - Google Patents

Description

The present invention relates to a maintenance improvement support system that operates according to equipment information and user requirements.

In many fields such as infrastructure, railways, industrial equipment, and medical equipment, it is necessary to maintain the specified performance by continuously implementing maintenance after the asset is installed. In maintenance, we collect the status of target assets, apply diagnosis to analyze the presence or absence of abnormalities and problems, and apply appropriate maintenance work.

With the recent development of information technology, it is possible to reduce maintenance costs and improve reliability by implementing maintenance with appropriate content and timing by using IT technology such as diagnostic technology that collects the status of assets with sensors and grasps the current status of assets or predicts future status. There is also an example of achieving significant maintenance efficiency by combining maintenance optimization techniques such as reliability-centered maintenance with IT technology.

However, in order to do so, it is necessary to introduce a maintenance process that is different from conventional maintenance processes linked to diagnostics, and this will lead to major changes in the number of personnel, necessary skills, and maintenance resources such as equipment. In addition, the introduction of new IT technology requires updating the system, and depending on the sensor and analysis method, IT resources such as large-capacity storage and large computing power are required. In addition, it is necessary to accumulate data and know-how to make full use of new technology.

In addition, for example, in order to introduce predictive diagnostic technology that predicts future equipment failures from data, it is necessary to introduce a data collection and visualization system as a preliminary step to accumulate data collection and analysis know-how.

For this reason, in conservation improvement activities, it is necessary to introduce possible combinations of conservation improvement measures in an appropriate order, rather than introducing measures one after another.

Patent Literature 1 discloses a technique for examining maintenance improvement measures. Patent Literature 1 is a system that supports formulation of improvement measures for individual failure modes based on failure mode analysis of target assets.

Patent No. 4237610

There are practical difficulties in determining the content of implementation of conservation improvement measures and the order in which they should be introduced. First, it is necessary that the contents of the improvement measures cover as many important items as possible from among the various maintenance items required for the assets that are the object of maintenance. In addition, only specialists such as maintenance consultants usually have knowledge of what kind of measures are possible for maintenance improvement measures.

In determining the order of introduction of maintenance improvement measures or the IT technology to realize them, it is necessary to introduce the technical prerequisites first. For example, unless a condition monitoring system is introduced and data is accumulated, predictive diagnosis by data analysis is impossible. Furthermore, it is desirable to introduce improvement measures that include core technologies that lead to more improvement measures.

Conservation practitioners' policies need to be reflected in what conservation issues are considered important. Whether it is maintenance cost, responsiveness, or safety depends on the wishes of the maintenance practitioner. In addition, the scale of investment that can be made at one time and the number of years required for improvement also depend on the operator.

Based on the above, it is an object of the present invention to provide a maintenance improvement support system that can present an appropriate maintenance improvement introduction plan based on the user's improvement policy from the maintenance improvement measures prepared in advance based on the knowledge of the target asset.

Based on the above, in the present invention, there are provided an asset knowledge management unit that manages asset knowledge, which is knowledge of asset failures and maintenance, a maintenance improvement measure database that records predefined maintenance improvement measures, a maintenance improvement policy selection unit that receives maintenance improvement policy settings by users, a maintenance improvement measure list generation unit that generates a list of maintenance improvement measures to be implemented from the asset knowledge and maintenance improvement measures, and the user's maintenance improvement policy, and a maintenance improvement plan that has a high implementation efficiency and an easy order from the list of maintenance improvement measures to be implemented. A maintenance improvement support system characterized by having an optimum execution plan generation unit to generate, extracting the maintenance improvement measures that match the characteristics of the asset from the maintenance improvement measures in line with the user's maintenance improvement policy, and determining the order of implementation from the dependency of technology and data requirements between maintenance improvement measures.

By using the present invention, it becomes possible to promote the introduction and promotion of maintenance improvement measures according to the situation of assets and users, and to improve the efficiency of maintenance work.

The figure which shows the structural example of the maintenance improvement support system which concerns on the Example of this invention. FIG. 4 is a diagram showing an example of asset knowledge D1 in the case of wind turbine maintenance. FIG. 4 is a diagram showing an example of maintenance improvement measures D2 accumulated in a maintenance improvement measure database 60 in the case of wind turbine maintenance as an example. FIG. 4 is a diagram showing an example of a KPI list; The figure which shows the example of a component technology list. FIG. 10 is a view showing an example of a maintenance improvement policy input screen 90 displayed on the display section of the input/output section HMI; 4 is a diagram showing a processing flow in a maintenance improvement measure list generation unit 40; FIG. The figure which shows the example of a score calculation result. The figure which shows the example of the measure selection display screen 91 displayed on the display part in input-output part HMI. FIG. 4 is a diagram showing a processing flow in an optimal implementation plan generation unit 50; A diagram showing the relationship between technology and data dependencies and the implementation of measures. FIG. 10 is a diagram showing an example of a maintenance improvement implementation and investment plan creation screen 92 displayed on the display unit of the input/output unit HMI;

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing a configuration example of a maintenance improvement support system according to an embodiment of the present invention. The maintenance improvement support system 100 shown in FIG. 1 is realized by a computer device, and is generally configured by connecting an input/output unit, a database, a calculation unit, and the like to a transmission bus 20 .

FIG. 1 is a diagram schematically showing the functions of a maintenance improvement support system 100 configured by these components: an asset knowledge management unit 10 that manages knowledge of a target asset; a maintenance improvement measure database 60 that records a collection of predefined maintenance improvement measures; a maintenance improvement policy selection unit 30 that accepts the setting of a maintenance improvement policy by a user; From the above list, it can be said that the optimum implementation plan generation unit 50 that generates a plan with a sequence that has a high implementation efficiency of maintenance improvement and is easy to introduce, and the input/output unit HMI that inputs and outputs information with the user M.

The user M in the present invention is a manager or a maintenance implementer who plans and plans maintenance, or an operator who manages assets, and appropriately inputs instructions and data to the maintenance improvement support system 100 via the input/output unit HMI, and is provided by displaying appropriate information from the maintenance improvement support system 100 via the input/output unit HMI.

It should be noted that the present invention is not limited to any particular asset, sensor technology, or analysis technology, but the following description will be made using wind turbine maintenance as an example. Wind turbines need to reduce maintenance downtime as much as possible, as availability affects the amount of electricity generated. In addition, since it is necessary to move to a remote installation point and climb a tower to actually perform inspection and maintenance, it is possible to apply remote monitoring to various parts such as the main shaft bearing, generator, brake, and pitch control. In addition, the application of predictive diagnosis that predicts failures in advance is also conceivable, considering the accident risk at the time of failure. In addition, when monitoring wind farms and remote wind turbines, it is conceivable that a large number of wind turbines must be inspected and maintained at the same time.

The configuration and function of each part in the maintenance improvement support system 100 will be described in detail below according to the flow of processing. In order for the maintenance improvement support system 100 to function, necessary information must be held in the maintenance improvement support system 100 in advance. Therefore, asset knowledge D1 and maintenance improvement measures D2 are prepared in advance.

Of these, the asset knowledge D1 is prepared in advance by the asset knowledge management unit 10 . For this reason, first, the user M inputs knowledge about the structure and failure mode of the target asset or knowledge from the installed condition monitoring system to the asset knowledge management unit 10 through the input/output unit HMI, and stores it as asset knowledge D1 in an appropriate storage area in the maintenance improvement support system 100.

FIG. 2 shows an example of the asset knowledge D1 in the case of wind turbine maintenance. For example, in the example of FIG. 2, asset knowledge is formed for each item of a component D11 that constitutes a wind turbine, a failure mode D13 with an ID (D12), an effect when occurring D14, an inspection measure D15, and a maintenance measure D16.

Specifically, for example, the asset knowledge D1 includes blades, shafts, bearings, gearbox gears, etc., as examples of components D11 that make up a wind turbine, blade cracks (ID=1), blade breakage (ID=2), shaft misalignment (ID=3), bearing scratches/wear (ID=4), grease deterioration (ID=5), and bearing sticking (ID=6) as failure modes D13. When 1) occurs, the downtime is 8 hours and the damage is 1 million yen, and when the blade breakage (ID = 2) occurs, the downtime is 72 hours and the damage is 15 million yen. In addition, methods such as regular checkups, remote diagnosis, and disassembly inspection are individually described as inspection means D15 for each failure mode, and component replacement, oil or grease injection, etc. are individually described as maintenance measures D16 for correcting this failure.

As shown in this example, the asset knowledge is based on the knowledge of failure mode analysis, which includes information such as the component D11 of the target asset, the failure mode D13, and the effect of failure D14. It is preferable that the currently available inspection means D15, the current maintenance action D16, and the time and cost required for each failure mode are also recorded.

Regarding the maintenance improvement measures D2, which are other preparatory information essential for the maintenance improvement support system 100 to function, the maintenance improvement measure database 60 stores a plurality of typical maintenance improvement measures D2 created in advance by maintenance and maintenance IT experts.

FIG. 3 shows an example of maintenance improvement measures D2 accumulated in the maintenance improvement measure database 60 in the case of wind turbine maintenance. Here, maintenance improvement measures D2 are a collection of measures D2a, which are business improvement contents such as operating rate improvement and work load leveling, improvement KPIs (D2b) such as issues and KPIs to be improved by implementation, and elemental technologies D2c such as IT technology and costs required for improvement.

In FIG. 3, for example, in the case where failure prevention is performed as a measure D2a with an ID of 2, operation rate and parts cost should be considered as improvement KPIs (D2b), and condition monitoring and predictive diagnosis should be performed as element technology D2c. Further, in FIG. 3, for example, in the case where work volume leveling is performed as a measure D2a with an ID of 4, labor costs should be considered as an improvement KPI (D2b), and schedule optimization and predictive diagnosis should be performed as elemental technologies D2c are stored in association with each other.

In the maintenance improvement measure database 60, the KPI list Lb of FIG. 4 and the component technology list Lc of FIG. 5 are recorded in relation to the maintenance improvement measure D2. Here, the KPI list Lb is an exhaustive list of the items described in the improvement KPI (D2b) in FIG. 3, and is a list in which individual items such as maintenance response, operation rate, parts cost, labor cost, recording quality, and reliability are described in a form in which IDs are assigned. The component technology list Lc is a comprehensive listing of the items described in the component technology D2c in FIG. 3, and is a list in which individual items such as status monitoring, predictive diagnosis, fault location identification, and schedule optimization are given IDs.

However, especially with respect to the elemental technology list Lc, elemental technologies (condition monitoring, predictive diagnosis, fault location identification, schedule optimization, etc.) necessary for implementing maintenance improvement measures should be saved together with introduction costs, prerequisite technologies, prerequisite data, and generated data. Here, the prerequisite technology and prerequisite data are technology and data that must already be available to the maintenance organization when they introduce the elemental technology. The premise data also records how long or in what amount it must have been accumulated. The example records the required number of years.

The data (prerequisite data, generated data) in the component technology list Lc includes fixed data given from settings or past driving experience, and time-series variable sensor data sequentially given from various sensors in a time-series manner.

The maintenance improvement support system 100 functions as follows on the premise that the various data described above are prepared in advance.

First, the user M inputs the points to be emphasized in implementing the improvement to the maintenance improvement policy selection unit 40 through the input/output unit HMI.

FIG. 6 shows an example of a maintenance improvement policy input screen 90 displayed on the display section of the input/output section HMI. As an example of the maintenance improvement policy input screen 90, a KPI improvement target input section 90a, an investment plan amount input section 90b, and a confirmation button 90c are displayed here. In the illustrated example, the KPI improvement target input section 90a displays the contents of the KPI list Lb of FIG. 4 as KPIs together with IDs, and indicates selection by placing a check mark in the desired improvement column. Here, improvement hopes for utilization rate and part cost are selected. The planned investment amount for each fiscal year is entered in the investment planned amount input section 90b, and after inputting each data, the confirmation button 90c instructs to take in the maintenance improvement policy D3.

As shown in the example of FIG. 6, user M first enters a performance index (KPI) for maintenance to be improved. As more specific examples of KPI, various things such as asset operation rate, reliability, amount of penalty payment, failure investigation success rate, repeat failure rate, etc. are conceivable. In FIG. 6, KPIs are input by selecting them from the KPI list in FIG. 4, but instead of KPIs, desired items for improvement may be input in text, and the KPIs corresponding to the list in FIG. 4 may be automatically selected based on similarity with the KPIs.

User M also enters an annual investment amount that can be spent on implementing maintenance improvements. After inputting each of the above information, the user M presses the confirmation button 90c of the input result. Thereby, the maintenance improvement policy D3 is generated, and the processing of the maintenance improvement measure list generation unit 40 is started.

The maintenance improvement measure list generation unit 40 extracts maintenance improvement measures to be implemented based on the maintenance improvement policy D3 input by the user M, the asset knowledge D1 prepared in advance, and the contents of the maintenance improvement measure D2.

FIG. 7 shows a processing flow in the maintenance improvement measure list generation unit 40. As shown in FIG. According to the processing flow of FIG. 7, after the start condition is established in processing step S100, asset knowledge D1 is first acquired from the asset knowledge management unit 10 in processing step S110. Here, the asset knowledge D1 includes the contents shown in FIG.

Next, in processing step S120, maintenance improvement policy D3 input by user M is obtained from maintenance improvement policy selection unit 30, and maintenance improvement policy D2 is obtained from maintenance improvement policy database 60 in processing step S130. Here, the maintenance improvement policy D3 includes the contents shown in FIG. 6, and the maintenance improvement measure D2 contains the contents shown in FIGS.

Next, in processing step S140, the KPI that user M wishes to improve in maintenance improvement policy D3 (in the case of FIG. 6, the operation rate with ID 2 and the parts cost with ID 3) is matched with maintenance improvement measure D2.

According to this matching, failure prevention (ID=2) and maintenance plan optimization (ID=5) are hits in the maintenance improvement measure D2 of FIG. 3 referred to based on the operation rate selected in FIG. 6, and failure prevention (ID=2) and diagnosis support (ID=3) are hits in the maintenance improvement measure D2 of FIG.

In processing step S140, it is checked whether the KPIs of each maintenance improvement measure D2 and the maintenance improvement policy D3 selected by the user M match each other, and the number of matching KPIs is used as the score of the maintenance improvement measure D2. If the maintenance improvement measure D2 has a weight value indicating the relationship to each KPI instead of the number, the total weight of the matching KPIs may be used as the score. In this case, the relationship between the KPI in the maintenance improvement policy D3 selected by the user M and the maintenance improvement measure D2 can be evaluated in more detail. The score thus obtained in processing step S140 is the KPI score KS.

Next, in processing step S150, it is verified which failures (failure modes D14) included in asset knowledge D1 (FIG. 2) can be improved by individual maintenance improvement measures D2 (FIG. 3). In the technical score setting of the processing step S150, a technical score TS is assigned to the improvement measure D2 according to which part of the asset knowledge D1 can be improved by the elemental technology D2c of each improvement measure D2 in FIG.

As a result, for example, which failure mode in FIG. 2 can be most effectively rectified when condition monitoring, which is the element technology of FIG. 5, is applied, and which failure mode in FIG. 2 can be most effectively rectified when predictive diagnosis is applied.

In the case of the previous example, the matching improvement measures D2 are maintenance plan optimization (ID=5), failure prevention (ID=2), and diagnosis support (ID=3) in FIG. 3. Further, according to the maintenance improvement measure D2 in FIG. Technology D2c is fault localization.

Here, for elemental technology D2c of each improvement measure D2, a technical score TS is set by comparing the technical field (elemental technology list Lc) shown in FIG. 5 with the inspection method and the impact at the time of failure occurrence included in the data of asset knowledge D1 shown in FIG.

In the case of the previous example, in the elemental technology list Lc shown in FIG. 5, reliability-centered storage (ID=5), maintenance plan evaluation (ID=6), condition monitoring (ID=1), predictive diagnosis (ID=2), and failure location identification (ID=3) are compared with the inspection methods and the effects at the time of failure that are included in the data of the asset knowledge D1 shown in FIG.

The technical score TS used in processing step S150 is, for example, the formula (1). Formula (1) is obtained by multiplying weighting coefficients α, β, γ, and δ with S, T, I, and M as the viewpoints of evaluation, and summing them up.
[Number 1]
TS(i,j)=Σ(α×Sj+β/Tj+γ×Ij+δ×Mj) (1)
In equation (1), i is the improvement measure ID, j is the asset knowledge data ID, and Σ is the sum of all the technical elements of each improvement measure. α, β, γ, and δ are constants. For example, α=1, β=6 [months], and γ=δ=5 [hours].

Thus, the formula (1) defines the technical score TS from the four evaluation viewpoints S, T, I, and M, but in carrying out the present invention, some of them may be used or other factors may be added.

Of these, the evaluation viewpoint S focuses on the presence or absence of a sensor. When comparing the data in FIG. 5 and FIG. 2, first, for the sensor in the technical field (component technology list Lc), if the sensor can be used for the inspection method, it is considered that there is commonality, and a point S is added to the score TS. For the point S, the weight α is considered, which is 1 when the sensor is available and 0 otherwise. According to the description of the prerequisite data and generated data items in the elemental technology list Lc shown in FIG. 5, the condition monitoring (ID=1) and predictive diagnosis (ID=2) are recorded using sensors, but sensors are not used for other items.

Furthermore, in the equation (1), the evaluation viewpoints T, I, and M are the periodic inspection period T, the investigation time I, and the treatment time M, respectively. For this reason, it is preferable that each period of periodical inspection cycle T, investigation time I, and treatment time M be included in the time such as the stop time of the occurrence impact D14 in the asset knowledge data D1 of FIG.

Thus, when the technical field plan is executed as the operation of the maintenance improvement system, of the asset knowledge data D1, the time related to maintenance is taken into consideration. For example, if the periodical inspection cycle T is short, or if the investigation time I or the treatment time M is long, it is considered difficult to formulate a maintenance implementation plan, so the technical score TS is given a high score.

On the other hand, when the technical field investigation is executed as the operation of the maintenance improvement system, a long investigation time is given a high score, and as for the technical field treatment, a long treatment time is given a high technical score TS.

As a result, a technical score TS for the selected elemental technology D2c (in the previous example, reliability centered storage (ID=5), maintenance plan evaluation (ID=6), condition monitoring (ID=1), predictive diagnosis (ID=2), and failure location identification (ID=3)) is assigned to each item of the failure mode illustrated in the failure mode D13 of FIG.

Also, at this time, if the technical score TS(i, j) is greater than the threshold ε, C(i)=Σ(TS(i, j)>ε) is set as the coverage of the asset knowledge of the policy i. This indicates how much asset knowledge the policy contributes to improving. Also, the sum of the scores of the individual knowledge data items is taken to obtain the improvement measure score TS(i)=ΣTS(i, j). where Σ is the sum over j.

FIG. 8 shows an example of scores calculated so far. Scores are exemplified as scores TS(i) and coverage C(i) regarding improvement measures for each item of measures D2a in maintenance improvement measures D2 in FIG. Note that this score is for blade cracks with i=1, for example.

In the processing flow of FIG. 7, in a processing step S160 for generating a measure list, a candidate list of measures to be implemented is generated using the calculated score (KPI score KS or technical score TS). Here, the list sorted in the order of the technical score TS(i) is used as the measure candidate list. This is presented to the user M through the input/output unit HMI in the desired measure selection process 170 to allow the user M to select the measure that the user M actually desires to introduce.

FIG. 9 shows an example of the policy selection display screen. FIG. 9 shows an example of a policy selection display screen 91 displayed on the display section of the input/output section HMI. As an example of the policy selection display screen 91, a policy selection display portion 91a, a return portion 91b (return key to the maintenance improvement policy input screen 90), and a confirmation button 91c are displayed here. In the illustrated example, in the measure selection display section 91a, the improvement measures of FIG. 8 are displayed in descending order of the technical score TS(i), along with the improvement KPIs and the elemental technologies along with the IDs. This pattern can be said to be obtained by rearranging the maintenance improvement measure data D2 of FIG. 3 in descending order of the technical score TS(i).

Here, after selecting a measure in the measure selection display section 91a by checking, if the user M confirms it by pressing the confirmation button 91c, the selection result is output as a measure candidate list, and the processing of the maintenance improvement measure list generating section 40 is completed. If the user selects to return to the maintenance improvement policy input screen 90 using the return section 91b, the maintenance improvement policy selection section 30 is restarted. Here, it is assumed that user M has selected and confirmed only failure prevention.

In the processing flow of FIG. 7, processing step S170 is a measure selection process (applying a check mark) using the measure selection display section 91a, and processing step S180 represents a determination process as to whether the return section 91b or the confirmation button 91c is selected.

Returning to FIG. 1, the optimal execution plan generation unit 50 receives the prototype candidate list output from the maintenance improvement measure list generation unit 40 and generates a plan for implementing it in an appropriate order. Here, by using the planned investment amount included in the maintenance improvement policy (FIG. 6) input by the user M and the dependence relationship between the elemental technologies included in the list of the elemental technologies D2a in the maintenance improvement measure database D2 (FIG. 3), the user M draws up an optimal measure introduction plan for realizing maintenance improvement.

The above planning in the maintenance improvement measure list generation unit 40 is based on the following ideas. For example, in order to prevent failures in maintenance improvement measures by using sensor records to predict failures in advance and issue warnings, or to implement predictive diagnosis of elemental technologies, as shown in Fig. 5, it is necessary to introduce condition monitoring technology in advance and collect sensor records related to target assets and failure modes. Also, even if condition monitoring is introduced, predictive diagnosis cannot actually be executed until the data to be analyzed is accumulated, so there is no point in introducing it at the same time. Failure prevention using predictive diagnosis cannot be introduced unless condition monitoring is first introduced and predictive diagnosis becomes possible after a certain period of time. The same can be said for predictive diagnosis, maintenance record improvement of measures, and recording terminals of elemental technologies. Moreover, even if user M considers only failure prevention to be necessary, it is essential to include condition monitoring and analysis in advance, and it is necessary to include condition monitoring and analysis in the implementation plan.

FIG. 10 shows the processing flow of the optimum execution plan generation unit 50. As shown in FIG. According to the processing flow of FIG. 10, after the start condition is established in processing step S200, in processing step S210, first, based on the scores evaluated by the system, a list 41 of maintenance improvement measures desired by user M to be implemented is obtained from the maintenance improvement measure list generation unit 40. This list has the contents shown in FIG. 9, for example.

On the other hand, in the subsequent processing from processing step S220 to processing step S260, the data D2 (maintenance improvement measure) recorded in the maintenance improvement measure database 60 in FIGS.

First, in processing step S220, it is checked whether or not there is a lack of the prerequisite technology described in the elemental technology list Lc of FIG.

Here, it is confirmed whether or not there is a shortage of the underlying technologies in the elemental technology list Lc corresponding to the elemental technology D2c of each measure D2a recorded in the maintenance improvement measure D2. In this example, condition monitoring is required as a prerequisite technology for the predictive diagnosis, which is the elemental technology D2c used for failure prevention in D2a requested by the user M, so it is recorded as an additional required technology.

In processing step S230, it is checked whether all the data in the component technology list Lc corresponding to the component technology D2c of each measure D2a recorded in the maintenance improvement measure database 60 for the measures desired by the user M are sufficient. In the component technology list Lc, sensor records and failure records are required as prerequisite data for predictive diagnosis, so these are recorded.

In processing step S250, in order to secure the essential element technology and data confirmed by the technology dependency check and data dependency check, the maintenance improvement measure database 60 is searched for a maintenance improvement measure D2 containing them. In this example, first, the state monitoring/analysis of the measure D2a in FIG. 3 having the state monitoring (ID=1) in FIG. 5 as the element technology is extracted. Condition monitoring also eliminates the need for this data as it produces a sensor record of the essential data. As for the failure record of the essential data, the recording terminal, which is the elemental technology for generating it, is first extracted, and the maintenance record improvement, which is the improvement measure including this elemental technology, is extracted.

From the processing up to this point, it can be seen that in order to implement the failure prevention measures desired by the user M, it is necessary to implement condition monitoring and failure record improvement as a prerequisite. Although it is possible to satisfy User M's wishes with this, if condition monitoring, failure record improvement, and failure prevention are introduced, there is a possibility that further maintenance improvement measures can be implemented by combining these elemental technologies and generated data, or by adding a small number of technologies and data, and the effects of maintenance improvement can be expanded.

In processing step S260, a search is made to see if there are any other measures that can be implemented by using the elemental technologies and generated data of the measures extracted so far, or by adding a small number of elemental technologies. In this process, measures that use the elemental technology and generated data of the already selected measures are extracted, and processing steps S220 and S230 are repeated for them.

Here, it can be seen that diagnostic support and work load leveling can be realized by introducing fault location support and schedule optimization, respectively, which are elemental technologies. For additional measures, the more elemental technologies and data that are additionally required, the more difficult it is to implement.

In addition, it also recursively searches for policies that are realized on the premise of additional policies. However, elemental technologies and generated data added by additional measures are counted as missing technologies/data R by distinguishing them from existing ones. This process extracts policies whose R is smaller than a certain value, or continues extraction until no more feasible policies are found. Here, the threshold is set to R<=2.

FIG. 11 shows the technology/data dependency relationship and the state of implementation of measures extracted by the processing up to this point. To briefly explain a part of the correlation in FIG. 11, the state monitoring analysis, which is a measure, performs state monitoring as element technology and sensor recording as generated data. In addition, as a measure to improve maintenance records, recording terminals are used as elemental technology, and failure records and maintenance records are recorded as generated data. In addition, the subsequent failure prevention, diagnostic support, and workload leveling also have similar relationships between measures, elemental technologies, and generated data.

In addition, condition monitoring, which is an elemental technology, can be developed, used, and applied to predictive diagnosis in failure prevention, failure location identification in diagnostic support, and schedule optimization in workload leveling. In addition, sensor records as generated data can be applied to predictive diagnosis, which is an elemental technology for failure prevention, and can also be applied to schedule optimization, which is an elemental technology for work load leveling. Other generated data can also be used as data-based judgment materials for other elemental technologies after similar long-term experience and data accumulation. Note that in the drawing, the short data accumulation period is positioned at the top of FIG. 11 and described.

Next, in processing step S270, the implementation schedule is adjusted according to the planned investment amount set by the user M in FIG. Here, from the technology/data dependency relationship already extracted as shown in FIG. 11, the measure (upper part of FIG. 11) that includes the elemental technology/generated data that is the basis of the dependency relationship is first implemented in order, but according to the number of years required for accumulation of the prerequisite data, the time from the introduction of the elemental technology associated with the implementation of the measure to the realization of the elemental technology of the next stage that depends on the generated data and the implementation of the new measure is set. Figure 11 also shows this number of years.

As a result, the order and the minimum number of years required to technically implement each measure have been determined, but in reality, the amount that User M can invest is limited for each year, so it is necessary to adjust the implementation timing of the measures accordingly.

At this time, the implementation timing of each measure is adjusted according to the investable amount of the user M for each year. The actual cost of a measure is the sum of the costs of the elemental technologies constituting the measure that have not been introduced at the implementation stage of the measure. At this time, if the two measures cannot be implemented in the same fiscal year, priority is given to the measure desired by the user M and the measure to provide the premise technology and data for the measure. Otherwise, priority is given to measures with high scores. However, if there are measures that can be implemented within the surplus investable amount, they may be prioritized.

FIG. 12 shows an example of a maintenance improvement implementation and investment plan creation screen 92 displayed on the display section of the input/output section HMI. As an example of the maintenance improvement implementation and investment plan creation screen 92, here, a time-series investment amount display portion 92a, a legend display portion 92c, a confirmation button 90d, and a part of the time-series investment amount display portion 92a are formed. In the illustrated example, the horizontal axis of the time-series investment amount display section 92a shows that the budget is prepared in the order of maintenance record improvement, condition monitoring analysis, and failure prevention diagnostic support as equipment investment items for each year, and the vertical axis displays the investment amount and the achievement status etc. so that they can be compared. In the legend display section 92c, conventions for display in the time-series investment amount display section 92a are written.

In addition, according to the comparison display of the investment amount and the achievement status on the vertical axis of the illustration, it has been approved that additional investment is necessary for reasons such as performance improvement or expansion of the scope of application for the existing condition monitoring analysis, and the plan to accelerate the installation plan for diagnostic support on the horizontal axis is presented. These parts are displayed as the judgment required display portion 92b, and the final decision including the judgment result here is made by the confirm button 90d.

A specific example of the judgment display portion 92b will be described below. When planning the equipment, it is predicted that there may be cases in which it is impossible to implement measures with the amount of money set by user M that can be invested. This may occur when the prerequisite technology for the elemental technology is not developed and essential measures must be implemented in advance. In this case, if it is an essential measure, an instruction is given to make an additional investment. In FIG. 12, since introduction of state monitoring is not possible with the investment amount set by user M, additional investment is urged. In addition, additional investment may make it possible to implement measures at an early stage. In this case, for example, the user M is presented with the necessary investment amount in the case of the shortest feasible time , so that the user M can judge whether or not additional investment is necessary. Also, at this time, by displaying the score of each policy as reference information, the user M can easily judge whether or not the policy is necessary.

While confirming FIG. 12 on the screen through the input/output unit HMI, the user M finally determines the optimum implementation plan for the maintenance improvement measures by adjusting the implementation necessity of the measures, adjusting the start time, and adjusting the investment amount. This determination result is presented to the user M through the input/output unit HMI.

By the above processing, the maintenance improvement support system in the present invention can be realized.

100: Maintenance improvement support system M: User 10: Asset knowledge management unit 60: Maintenance improvement measure database 30: Maintenance improvement policy selection unit 40: Maintenance improvement measure list creation unit 50: Optimal implementation plan creation unit HMI: Input/output unit

Claims (9)

a maintenance improvement policy database that records predefined maintenance improvement measures; a maintenance improvement policy selection unit that accepts setting of a maintenance improvement policy by a user; a maintenance improvement measure list generation unit that generates a list of maintenance improvement measures to be implemented from the asset knowledge and the maintenance improvement measures of the asset and the maintenance improvement policy of the user; A maintenance improvement support system comprising a generator, extracting maintenance improvement measures that match the characteristics of an asset from maintenance improvement measures in line with a user's maintenance improvement policy, and determining the order of implementation from the dependency of technology and data requirements between maintenance improvement measures. The maintenance improvement support system according to claim 1,
The maintenance improvement measure list generation unit selects the maintenance improvement measure as a candidate for implementation from a match between the maintenance KPI that the user desires to improve and the maintenance KPI associated with the maintenance improvement measure stored in advance in the maintenance improvement measure database. The maintenance improvement support system according to claim 1,
The maintenance improvement measure list generation unit selects an improvement measure with high compatibility by measuring the suitability of the maintenance improvement measure that is a candidate for implementation to the target asset characteristics from the asset knowledge input to the asset knowledge management unit, from the fact that the improvement effect on failure inspection means, maintenance measures, and effects is large, or from the fact that the number of failures that are improved by the measure is large. The maintenance improvement support system according to claim 1,
The optimum implementation plan generation unit formulates the implementation order of the maintenance improvement measures in such an order that the elemental technologies and premise data necessary for realizing the maintenance improvement measures that are candidates for implementation are satisfied as much as possible by the elemental technologies and the generated data related to the measures that have already been introduced. The maintenance improvement support system according to claim 4,
In addition to the maintenance improvement measures that the user wants to introduce, the optimum implementation plan generation unit is characterized by realizing a wider range of maintenance improvements by presenting to the user as additional maintenance improvement measures, if there are measures that can be implemented based on the elemental technologies and generated data to be introduced by the maintenance improvement measures already included in the implementation plan. The maintenance improvement support system according to claim 5,
The optimum implementation plan generation unit presents to the user, among the elemental technologies and data that are the premises for the implementation of additional maintenance improvement measures, those that are less satisfied by the existing maintenance improvement measures, by giving priority to presenting them to the user. A maintenance improvement support system characterized by efficiently expanding the scope of maintenance improvement. The maintenance improvement support system according to claim 1,
A maintenance improvement support system, wherein the optimum implementation plan generation unit determines whether or not an improvement measure, which is a candidate for implementation, can be implemented with the user's investment amount by setting an investment amount for the maintenance improvement by the user. The maintenance improvement support system according to claim 7,
A maintenance improvement support system, wherein the optimum implementation plan generation unit prompts the user to change the investment plan when the investment amount planned by the user cannot implement the measures for realizing the maintenance improvement desired by the user. The maintenance improvement support system according to claim 4 or claim 7,
A maintenance improvement support system characterized in that the optimum implementation plan generation unit calculates the time when maintenance improvement measures can be implemented and the costs necessary for realizing them, and by correcting the user's investment plan, it is possible to implement a wider range of maintenance improvements, or if the maintenance improvement can be realized earlier, the user is prompted to change the investment plan.

Images