JP2024061314A - Business policy evaluation device and business policy evaluation method - Google Patents

Abstract

[Problem] To provide a business policy evaluation device that can rationally and quickly find suitable business policies that satisfy a variety of values from among the countless business policies that exist. [Solution] The business policy evaluation device includes an index designation unit that designates a target index from a plurality of indexes prepared in advance, a business policy extraction unit that extracts business policies related to the target index, an evaluation unit that evaluates the influence of the extracted business policies on the target index, and an output unit that outputs the evaluation result of the evaluation unit. [Selected Figure] Figure 1

Description

The present invention relates to a business policy evaluation device and a business policy evaluation method that support the design and improvement of various business operations, such as maintenance operations.

In recent years, the business environment has changed rapidly, and various industries are being called upon to review and improve their work design in response to the business environment. For example, in maintenance work for infrastructure, railways, industrial equipment, medical equipment, etc., continuous maintenance such as inspection and repair must be carried out to ensure the safe and stable operation of assets (facilities). In this type of maintenance, organizational structures and work standards are designed taking into account the operating status and operational characteristics of assets, as well as the status of maintenance resources such as workers and tools. However, when the business environment changes, it is necessary to take some kind of business measures and review the maintenance work design to suit the new environment.

A business policy construction support system described in Patent Document 1 is known as a conventional technology that supports the construction of such business policies. For example, the abstract of Patent Document 1 discloses a business policy construction support system that "has a business simulator unit that generates a virtual business history by executing a simulation of a service business performed by a service provider under a predetermined business policy, an acquisition unit that acquires virtual business history corresponding to each business policy by having the business simulator unit execute simulations sequentially under different business policies, and an output unit that, when an evaluation value calculated based on the virtual business history acquired by the acquisition unit satisfies a predetermined condition, outputs the business policy corresponding to the virtual business history in association with the evaluation value."

Furthermore, paragraph 0091 of the same document states, "When the business measures input to the business simulator unit 621 have evaluation values that do not satisfy a predetermined condition, the contents of the items identified by the change item information are changed sequentially. The example in FIG. 8(b) shows that the contents of the item identified by the change item information (item number = 1) are changed from P1 to P1' to P1'', and paragraph 0094 states, "In this way, according to the business measure construction support system 131, the business measures are changed sequentially until the calculated evaluation value satisfies the predetermined condition."

JP 2017-208035 A

In recent years, in various business fields, in addition to traditional values such as sales and profits, a variety of values such as the environment, resilience, and human rights have begun to be emphasized. Furthermore, the values that are emphasized are also changing rapidly. For this reason, business operators need to flexibly and continuously transform their business and operational designs in response to the changing values required by the times.

As mentioned above, Patent Document 1 describes a method for sequentially calculating the evaluation value of each business measure through simulation and presenting business measures whose evaluation value satisfies a predetermined condition. With this method, it is necessary to sequentially set possible business measures in advance and simulate each business measure, but there are a wide variety of options for business measures. For example, performance improvement through asset updates, changes to maintenance policies, increasing the number of workers, adding bases, introducing IoT solutions, etc. Furthermore, business measures that combine these are also possible, so there are an infinite number of options for business measures.

Therefore, using the technology of Patent Document 1 to sequentially simulate and evaluate all of the countless options for business measures would take an enormous amount of time. In particular, a highly accurate simulation would require a long evaluation time per case, making it extremely difficult to evaluate business measures for many cases through simulation.

The present invention aims to provide a business policy evaluation device and a business policy evaluation method that can rationally and quickly find suitable business policies that satisfy diverse values from among the countless business policies that exist.

To solve the above problem, a representative embodiment of the present invention is a business policy evaluation device that includes an index designation unit that designates a target index from a plurality of previously prepared indexes, a business policy extraction unit that extracts business policies related to the target index, an evaluation unit that evaluates the influence of the extracted business policies on the target index, and an output unit that outputs the evaluation result of the evaluation unit.

The business policy evaluation device or business policy evaluation method of the present invention can find an appropriate solution rationally and quickly by narrowing down and evaluating suitable policies that improve target indicators from among countless business policy options. Note that issues, configurations, and effects other than those described above will become clear from the explanation of the embodiments below.

FIG. 2 is a functional block diagram of the business policy evaluation device according to the first embodiment. 4 is a process flowchart of the business policy evaluation device according to the first embodiment. 13 is an example of an index setting screen according to the first embodiment. 13 is an example of a business policy list according to the first embodiment. 4 is an example of causal relationship structure information according to the first embodiment; 4 illustrates an example of a business policy extraction process according to the first embodiment. 4 illustrates an example of a business policy extraction process according to the first embodiment. 4 illustrates an example of a business policy extraction process according to the first embodiment. FIG. 11 is a functional block diagram of a business policy evaluation device according to a second embodiment. 13 is a process flowchart of a business policy evaluation device according to a second embodiment. 13 is an example of an output screen of an evaluation result according to Example 2. FIG. 11 is a functional block diagram of a business policy evaluation device according to a third embodiment. 13 is a process flowchart of a business policy evaluation device according to a third embodiment.

Below, an embodiment of the business policy evaluation device of the present invention will be described with reference to the drawings. In each embodiment, maintenance work is exemplified as the business to which the present invention is applied, but the present invention can also be applied to service industries such as finance and transportation, manufacturing industries such as chemicals and automobiles, and agriculture. Maintenance work includes various tasks such as inspection, repair, and maintenance, and assets that are the subject of maintenance work include various items such as facilities, equipment, machinery, and products.

Figure 1 is a functional block diagram of a business policy evaluation device 1 according to a first embodiment of the present invention. As shown here, the business policy evaluation device 1 of this embodiment has an index designation unit 11, a business policy extraction unit 12, an evaluation unit 13, an output unit 14, and an information storage unit 15. The information storage unit 15 stores a business policy list 15a, causal relationship structure data 15b, and evaluation results 15c. Specifically, the business policy evaluation device 1 is a computer equipped with hardware such as a calculation unit such as a CPU, a storage device such as a semiconductor memory, and a communication device. The calculation unit executes a desired program to realize each function of the index designation unit 11, etc., described above, but the following description will omit such well-known technologies as appropriate.

In addition, the business policy evaluation device 1 is connected to an input device 2 such as a keyboard, mouse, or touch panel, and an output device 3 such as an LCD display, allowing the user to input commands to the business policy evaluation device 1 and for the business policy evaluation device 1 to present information to the user.

Next, the process executed by the business policy evaluation device 1 will be explained in sequence according to the flowchart in FIG. 2.

<Step S1>
First, in step S1, the index designation unit 11 designates indexes that the user considers important in conducting business, according to the user's selection described below. An index is also called a KPI (Key Performance Indicator), and is a value that measures the progress of business achievements and the performance of activities related to business. This index can be classified into social issue indexes related to society as a whole, management indexes related to the business status of a company, and activity indexes that represent the status of assets, organizations, resources, etc. in corporate activities. From the viewpoint of improving the user's understanding, it is even better if the activity index is expressed as a quantitative numerical value.

Social issue indicators include, for example, stable energy supply, resilience, green, recycling, human rights, and QoL (Quality of Life). The indicators that are emphasized among these social issue indicators change rapidly depending on the social situation. Management indicators include, for example, profits, sales, free cash flow, stock prices, labor costs, operating costs, parts costs, travel costs, penalties, and employee satisfaction. Activity indicators are diverse and include, for example, operating rate, downtime, number of bases, number of workers, CO2 emissions, number of spares, number of warehouses, and number of maintenance tasks.

Figure 3 shows an example of a target KPI setting screen displayed on the output device 3 (liquid crystal display) when the user selects a target indicator (target KPI) that the user places emphasis on from a number of indicators (KPIs) prepared in advance prior to step S1. This example shows a state in which the user operates the input device 2 (e.g., a mouse) to select the utilization rate, which is one of the activity indicators, as the target indicator. Note that only one target indicator is selected here, but if there are multiple indicators that the user wishes to place emphasis on, the user can select any multiple target indicators. Note that a target indicator is an indicator (KPI) that the user has selected as the one that the user places emphasis on from a number of indicators (KPIs) prepared in advance.

<Step S2>
In step S2, the business policy extraction unit 12 selects a business policy pattern to be evaluated from the business policy list 15a stored in the information storage unit 15. Here, the business policy pattern refers to a change in some design parameter that can be intentionally changed from the current business design. The targets of business design are diverse, including assets, organizations, resources, and technologies (IoT solutions).

Figure 4 is an example of the business measure list 15a. As shown in this example, the business measure list 15a in this embodiment is defined by a combination of business measure patterns and measure parameters. For example, as the measure parameters for business measure pattern #1 "update (replacement) of aging equipment", multiple parameters that affect the evaluation result of "update (replacement) of aging equipment" are registered, such as the number of pieces of equipment to be updated (scope of update), the extent of performance improvement by equipment update, and replacement cost.

As such, since there are many possible settings for policy parameters for even one business policy pattern, the number of business policies that need to be considered for business improvement is enormous. In addition, a composite policy that combines multiple business policy patterns can also be considered as a single policy, so there are countless options for business policies. Furthermore, policies that do not change the business design can also be considered as a business policy, as they are policies that maintain the status quo.

<Step S3>
In step S3, the business measure extraction unit 12 selects business measure patterns that are related to the target indicators specified in step S1, using the causal relationship structure data 15b that defines the causal relationships between business measures and indicators and the causal relationships between indicators. In the following, the structure of the causal relationship structure data 15b will be explained first, and then the process of extracting business measures will be explained.

Figure 5 is an example of the structure of causal relationship structure data 15b. On the left side of the figure, various indicators (KPIs) related to the business are written in boxes, and are arranged hierarchically from the left in the order of social issue indicators, management indicators, and activity indicators. The arrows in the figure indicate causal relationships between indicators, with solid arrows indicating causal relationships of positive effects and dashed arrows indicating causal relationships of negative effects. For example, a solid arrow connects the indicator "maintenance work time" to the indicator "downtime," indicating a causal relationship in which an increase in maintenance work time also increases downtime. Furthermore, a dashed arrow connects the indicator "downtime" to the indicator "uptime rate," indicating a causal relationship in which an increase in downtime decreases uptime.

Each indicator in Figure 5 is connected by causal arrows to all of the indicators one level downstream that would be affected if that indicator changes. As the number of indicators increases, the causal structure becomes more complex, making it difficult for people to grasp all of the causal relationships. In general, these causal structures often have causal relationships from activity indicators to management indicators and social issue indicators, and have a hierarchical tree structure. By hierarchically organizing this type of causal structure data 15b and expressing it as a KPI tree, it becomes easier to understand the causal relationships between a large number of indicators. In addition, in the example of Figure 5, indicators for multiple stakeholders, such as the operations department and maintenance department, are also listed.

Typically, social issue indicators are common to all stakeholders. However, management indicators and activity indicators differ for each stakeholder, and indicators may affect each other across stakeholders. By creating causal relationship structure data that includes multiple stakeholders, it is possible to understand the causal relationships between indicators in a more complex business structure. For example, it is possible to understand cases where there is a conflict of interest, where the indicators of one stakeholder improve while those of another stakeholder deteriorate. It is also possible to understand cases of synergy, where an improvement in the indicators of one stakeholder leads to an improvement in the indicators of another stakeholder.

In addition, on the right side of Figure 5, the business measure patterns (see Figure 4) included in the business measure list 15a are listed, and each business measure is connected to an activity indicator by an arrow. The arrow indicates the causal relationship with the indicator affected by the business measure. For example, in the case of the business measure pattern "update (replacement) of aging equipment," a solid arrow is connected between the indicator "equipment performance" and the indicator "failure frequency," because it is expected that updating the equipment will improve performance and reduce the frequency of failures. Since business measures involve changing part of business activities by modifying the current business design, the impact of business measures is generally often linked to activity indicators.

The causal relationships described above are generally constructed based on human experience and know-how. If experienced people define the causal relationships, a more accurate causal structure model can be constructed. On the other hand, when it comes to inexperienced business measures, there is a risk of mistakes or oversights in the causal relationships because there is no past track record. Causal relationships can also be automatically created from documents and internet information using natural language processing and machine learning.

Figures 6A to 6C show an example of the process of extracting business measure patterns when the user selects "operating rate" as the target indicator (see Figure 3). Note that in order to clearly show the steps of this process, information not necessary for explaining the extraction process (indicators, business measure patterns) is omitted from Figures 6A to 6C, but the causal relationship structure data 15b in each figure actually holds information equivalent to that in Figure 5.

Figure 6A shows the process in which the evaluation unit 13 sets the "operating rate", the indicator specified in step S1, as the target indicator.

Figure 6B shows the process in which the evaluation unit 13 starts from the target indicator "availability" and traces back the causal arrows, whether solid or dashed, to extract all other indicators that affect the "availability". This process extracts all indicators such as "operating time" and "downtime" that are upstream of the target indicator "availability".

Figure 6C shows the process in which the evaluation unit 13 extracts all business measure patterns linked to each indicator extracted in Figure 6B. This process extracts each business measure pattern, such as "update of aging equipment" and "remote monitoring and diagnosis", which are upstream of the target indicator "operation rate". As can be seen from a comparison of Figures 5 and 6C, Figure 6C also shows that business measure patterns such as "hiring new workers" and "training workers using digital knowledge" were not extracted as business measure patterns that affect the target indicator "operation rate".

In this way, in step S3, by tracing back the causal relationships in the causal relationship structure data 15b, it is possible to narrow down the business measures that may affect the target indicator.

A weighting factor can be set for each of the causal arrows mentioned above. By tracing the causal relationships backwards while taking these weighting factors into account, it is possible to evaluate the impact of the extracted business measures and rank the degree of impact of the extracted business measures. By prioritizing the extraction and evaluation of business measures with high impact rankings, it is possible to efficiently find the most effective business measures.

Note that while Figure 3 and Figures 6A to 6C show an example of a situation in which only one indicator is set as the target indicator, multiple indicators may be set as target indicators. In that case, the process corresponding to Figures 6A to 6C is performed for each target indicator to obtain the business measure patterns related to each target indicator, and then the business measure patterns common to all target indicators are identified, making it possible to extract business measure patterns that affect all target indicators.

<Step S4>
In step S4, the evaluation unit 13 evaluates the effect and influence on each index when the business measure pattern selected in step S3 is implemented, that is, the extent to which the index will change from the current state.

One example of an evaluation method in this step is agent simulation, which reproduces real-world operations. In agent simulation, asset data that corresponds to each asset in the simulation world is generated for each management unit, such as a real-world asset or an asset part, and the asset data is then operated and broken down autonomously to reproduce the behavior of the real-world asset. The data generated at this time is called an agent, and in the case of an asset, it is called an asset agent. Similarly, agents are generated in the simulation world for each maintenance worker, and are called maintenance worker agents. These perform actions such as waiting, moving, performing tasks, and resting in response to work instructions. In addition, agents corresponding to characters in the real world are generated, such as an operation agent that operates assets and requests troubleshooting when problems occur, and an allocation agent that assigns work to maintenance workers.

By using this type of agent simulation, complex business activities involving various real-world actors can be virtually reproduced in cyberspace, and activity indicators related to the business activities can be quantitatively predicted and evaluated from the behavior of each agent. To evaluate the effect and impact on indicators when a business measure is implemented, the situation in which the business measure is virtually implemented can be reproduced in a simulation, the indicators in that situation can be predicted and evaluated, and the difference can be compared with the indicators in the current business design. Note that the simulation method is not limited to agent simulation, and for example, a machine learning model generated from past performance data can be used.

<Step S5>
In step S5, the evaluation unit 13 stores the evaluation result 15c simulated in step S4 in the information storage unit 15. The output unit 14 outputs the evaluation result 15c stored in the information storage unit 15 to the output device 3 (liquid crystal display) and presents the evaluation result 15c to the user. An example of the presentation screen for the evaluation result will be described in Example 2.

According to the business policy evaluation device of this embodiment described above, it is possible to narrow down and evaluate suitable business policy measures that improve the target indicator from among the countless business policy options. Furthermore, according to the business policy evaluation device of this embodiment, it is possible to evaluate the effects of business policies by prioritizing business policies that have a large impact on the target indicator, and it is possible to quickly and efficiently select promising business policies. Furthermore, according to the business policy evaluation device of this embodiment, it is possible to narrow down rationally effective business policies based on the causal relationships between the indicators even in complex situations such as when there are multiple target indicators or when there are multiple stakeholders.

Next, a business policy evaluation device 1 according to a second embodiment of the present invention will be described with reference to Figs. 7 to 9. Note that a repeat explanation of the points in common with the first embodiment will be omitted, and only the differences will be described.

Figure 7 is a functional block diagram of the business policy evaluation device 1 according to the second embodiment. As can be seen from a comparison between Figures 1 and 7, the business policy evaluation device 1 of this embodiment differs from the business policy evaluation device 1 of the first embodiment in the following points. That is, first, a statistical evaluation unit 16 is added to the business policy evaluation device 1 of this embodiment. Second, in the business policy evaluation device 1 of this embodiment, the simulation performed by the evaluation unit 13 is a probabilistic event simulation. Third, in the business policy evaluation device 1 of this embodiment, the output of the statistical evaluation unit 16 is stored in the information storage unit 15 as a statistical evaluation result 15c'.

The processes executed by the business policy evaluation device 1 of this embodiment, which differ from the first embodiment in the above respects, will be described in sequence according to the flowchart in FIG. 8. Note that the description of the points in common with the flowchart in FIG. 2 will be omitted as appropriate.

The processing contents from step S1 to step S3 are the same as those in Example 1.

In step S4a, the evaluation unit 13 executes a stochastic event simulation multiple times. The stochastic event simulation executed in this step is, for example, in the case of a simulation of maintenance work, a simulation in which a failure probability of an asset or an asset component is set and failures occur randomly according to the failure probability. The failure probability is given by a distribution function such as the Weibull distribution. It is also possible to assign a distribution function to the work time of a worker or the lead time for sharing parts so that the work time or lead time varies randomly. By executing a stochastic event simulation multiple times that takes into account such randomly changing uncertain events, it becomes possible to evaluate the risk of uncertain events.

In step S4b, the statistical evaluation unit 16 receives the evaluation results of the random event simulation performed multiple times in step S4a, and statistically processes the evaluation results of the multiple indices. Examples of statistical processing include the mean, variance, standard deviation, confidence interval, etc.

In step S5, the statistical evaluation unit 16 stores the evaluation results of the indices statistically processed in step S4b as statistical evaluation results 15c' in the information storage unit 15. In addition, the output unit 14 outputs the statistical evaluation results 15c' stored in the information storage unit 15 to the output device 3 (liquid crystal display) and presents the evaluation results to the user.

Figure 9 is an example of a presentation screen for statistical evaluation results 15c' displayed on the output device 3 (liquid crystal display). The upper part of the screen in this figure shows the evaluation results of the target indicators in graph form. The area at the top of the screen allows the selection of the target indicators to be output to the graph; in this example, "green" has been selected as KPI1 of the target indicators, and "employee satisfaction" has been selected as KPI2. In this case, the statistical results when a specified business measure pattern is implemented are shown on a two-dimensional graph with KPI1 "green" on the horizontal axis and KPI2 "employee satisfaction" on the vertical axis.

In FIG. 9, in addition to the evaluation value of the target KPI when business measure A selected by the business measure extraction unit 12 is implemented, the results of the current business state are also plotted. This display method allows the user to intuitively recognize the difference between the two as the improvement effect due to business measure A. Note that in FIG. 9, the evaluation results of multiple probability event simulations are plotted small, and their average value is plotted large. This makes it possible to visualize results that take into account randomness due to uncertain events, understand the range in which the KPI may fluctuate, and evaluate risks.

In addition, the lower part of FIG. 9 displays the causal relationship structure data 15b (KPI tree) stored in the information storage unit 15, highlighting all indicators affected by business measure A (i.e., indicators downstream of business measure A). By displaying in this way, the ripple effect of business measure A on various indicators can be visually grasped. It also prevents the user from overlooking effects on indicators that were not anticipated.

Next, a business policy evaluation device 1 according to a third embodiment of the present invention will be described with reference to Figs. 10 and 11. Note that a repeat explanation of the points in common with the above-mentioned embodiments will be omitted, and only the differences will be described.

Figure 10 is a functional block diagram of the business policy evaluation device 1 according to the third embodiment. As can be seen from a comparison between Figures 1 and 10, the business policy evaluation device 1 of this embodiment is obtained by adding a causal relationship learning unit 17 to the first embodiment. The process executed by the business policy evaluation device 1 of this embodiment, which differs from the first embodiment in this respect, will be described using the flowchart of Figure 11.

The processing contents from step S1 to step S5 are the same as those in Example 1.

In step S6, the causal relationship learning unit 17 receives the causal relationship structure data 15b and the evaluation results 15c stored in the information storage unit 15, learns the causal relationships between the indicators from the evaluation results 15c, and if the learning results differ from the causal relationship structure data 15b, corrects and updates the causal relationships between the indicators in the causal relationship structure data 15b.

As explained in the first embodiment, the causal relationship structure data 15b used in the present invention is basically a causal relationship structure model constructed by humans based on human experience and know-how, and therefore does not reflect business measures that humans have no experience with, and there is a risk of errors or oversights in the causal relationships. Therefore, the causal relationship structure data 15b at the time of construction is often of low accuracy, and does not necessarily properly indicate the causal relationship structure between business measures and indicators, or the causal relationship structure between indicators.

Therefore, according to the third embodiment, the impact and effect on indicators of untried business measures can be virtually evaluated by business simulation, and by learning the results of the simulation evaluation, the causal structure can be reflected in the impact of untried business measures on each indicator, making it possible to improve errors and oversights in causal relationships.

The present invention is not limited to the above-described embodiments, but includes various modified examples. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those having all of the configurations described. It is also possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. It is also possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

1...Business policy evaluation device,
11...index designation unit,
12...Business policy extraction unit,
13...Evaluation unit,
14...output unit,
15... Information storage unit,
15a…Business policy list,
15b...causal relationship structure data,
15c...Evaluation results,
15c'...Statistical evaluation results;
16...Statistical Evaluation Section,
17...Causal relationship learning unit,
2. Input device;
3. Output device

Claims (8)

An index designation unit that designates a target index from a plurality of indexes prepared in advance;
a business measure extraction unit that extracts business measures related to the target indicator;
an evaluation unit that evaluates an influence of the extracted business measures on the target indicator;
an output unit that outputs an evaluation result of the evaluation unit;
A business policy evaluation device comprising: 2. The business policy evaluation device according to claim 1,
Further, the system has an information storage unit that stores causal relationship structure data that defines causal relationships between business measures and indicators and causal relationships between indicators,
The business policy evaluation device is characterized in that the business policy extraction unit extracts business policies related to the target indicators based on the causal relationship structure data. 3. The business policy evaluation device according to claim 2,
The business policy evaluation device, wherein in the causal relationship structure data, each indicator is classified into one of a social issue indicator, a management indicator, and an activity indicator. 3. The business policy evaluation device according to claim 2,
The business policy evaluation device, wherein each index in the causal relationship structure data is classified into an index of any one of a plurality of stakeholders. 3. The business policy evaluation device according to claim 2,
The business policy evaluation device further comprises a causal relationship learning unit that learns causal relationships between indexes based on the evaluation results of the evaluation unit and updates the causal relationship structure data. 3. The business policy evaluation device according to claim 2,
The business policy evaluation device is characterized in that the output unit outputs all of the affected indicators based on the causal relationship structure data. 2. The business policy evaluation device according to claim 1,
The business policy evaluation device is characterized in that the evaluation unit evaluates an index by executing a simulation including random probabilistic events a plurality of times, and performs statistical processing on the index evaluated a plurality of times. An index designation step of designating a target index from a plurality of indexes prepared in advance;
A business measure extraction step of extracting business measures related to the target indicators;
an evaluation step of evaluating an impact on the target indicator when the extracted business measure is implemented;
an output step of outputting an evaluation result of the evaluation step;
A business policy evaluation device comprising:

Images