JP7544794B2 - Information processing device and information processing method - Google Patents

Description

The present invention relates to an information processing device and an information processing method.

In recent years, there has been active technological development to promote Digital Transformation (DX), and as part of this, importance is being placed on status monitoring of projects such as software development. Status monitoring can be divided into two categories: focused monitoring and wide-area monitoring. Focused monitoring is monitoring of projects by the management department. Wide-area monitoring is monitoring of all projects using specified tools. In status monitoring, first, projects that are judged to have a high possibility of failure or a large impact in the event of failure in wide-area monitoring based on data such as costs entered into the tool are targeted for focused monitoring. Next, with support from the management department, efforts are made to prevent deterioration or improve the projects that are the subject of focused monitoring.

Conventional wide-area monitoring would identify projects after problems, such as costs exceeding estimates, had become apparent. In light of the goal of leading a project to success, it is preferable to detect signs of deterioration in a project at an early stage, before problems have become apparent. However, with conventional wide-area monitoring, there was a problem in that it was difficult to detect signs of deterioration in a project at an early stage, before problems have become apparent.

Furthermore, in conventional wide-area monitoring, investigations are conducted to determine whether or not a project of concern should really be the subject of priority monitoring. Investigations include, for example, investigating the scale of the project and the amount of overruns in estimated costs. However, such investigations require a huge amount of human resources. For this reason, projects with large order amounts and other projects that would be greatly affected if the situation were to worsen are given priority, and monitoring of all projects, including small-scale projects, is not conducted. In other words, with conventional wide-area monitoring, there was the problem of the large human resources costs involved in determining which projects should be the subject of priority monitoring.

In addition, when monitoring a project, it is customary to obtain forecast values for a specified period of time for actual values of data such as costs. By using forecast values in conventional wide-area monitoring, there is room to improve the accuracy of early detection of signs of project deterioration.

Patent Document 1 discloses a next month's cost calculation device, a next month's cost calculation method, and a next month's cost calculation program that can contribute to understanding costs and profits in real time by referring to the latest data at that time and calculating costs as preliminary values before the current month's inventory data is finalized.

JP 2021-144687 A

In view of these circumstances, the present invention aims to detect early signs of deterioration in a project through wide-area monitoring, while reducing the human cost of deciding whether or not to make a project a target for priority monitoring.

The present invention solves the above problems,
a generation unit that generates a prediction model using explanatory variables constituting monitoring information, which is information for monitoring the status of a completed first project, as input information and using an estimated cost overrun of the first project as output information;
a prediction unit that inputs explanatory variables of a second project in progress into the prediction model and outputs a predicted value of an estimated cost overrun of the second project and explanatory variables that are the basis of the predicted value;
The generation unit is
generating the prediction model using as input information a forecast value at time B for an actual value at time A in a specified explanatory variable designated from the explanatory variables of the first project, and an expected value calculated using the actual value at time B;
The prediction unit is
The information processing device inputs explanatory variables of the second project, including a forecast value at a second time point for an actual value at a first time point and an expected value calculated using the actual value at the second time point in a specified explanatory variable specified from the explanatory variables of the second project , into the prediction model, and outputs a forecast value of an estimated cost overrun of the second project and the explanatory variables that are the basis for the forecast value .

The present invention also provides a method for producing a method for manufacturing a semiconductor device comprising the steps of:
An information processing device,
a first step of generating a prediction model in which explanatory variables constituting monitoring information, which is information for monitoring the status of a completed first project, are used as input information, and an estimated cost overrun of the first project is used as output information;
A second step is carried out in which explanatory variables of a second project in process are inputted into the prediction model, and a predicted value of an estimated cost overrun of the second project and explanatory variables on which the predicted value is based are outputted;
In the first step,
generating the prediction model using as input information a forecast value at time B for an actual value at time A in a specified explanatory variable designated from the explanatory variables of the first project, and an expected value calculated using the actual value at time B;
In the second step,
The information processing method inputs explanatory variables of the second project, including a forecast value at a second time point for an actual value at a first time point and an expected value obtained using the actual value at the second time point in a specified explanatory variable specified from the explanatory variables of the second project , into the prediction model, and outputs a forecast value of an estimated cost overrun of the second project and the explanatory variables that serve as the basis for the forecast value .

The present invention enables wide-area monitoring to detect early signs of deterioration in a project and reduce the human costs of deciding whether or not to target a project for priority monitoring.

FIG. 2 is a diagram illustrating an example of a functional configuration of an information processing device according to the present embodiment. 13 is a screen example of output information from a prediction unit. 4 is an example of a flowchart illustrating a process according to the present embodiment.

First Embodiment
[composition]
The information processing device 100 shown in FIG. 1 is a computer that detects signs of deterioration in a project in progress by wide-area monitoring. The information processing device 100 includes hardware such as an input unit, an output unit, a control unit, and a storage unit. For example, when the control unit is configured with a CPU (Central Processing Unit), information processing by a computer including the control unit is realized by program execution processing by the CPU. In addition, the storage unit included in the computer stores various programs for realizing the functions of the computer according to instructions from the CPU. This realizes collaboration between software and hardware. The program can be provided by recording it on a recording medium or via a network. The output unit may include a function of a display unit that displays on a screen.

A project is a task to achieve a specific goal. Projects can be classified into past projects that have been completed and projects that are in progress. In this embodiment, past projects are called "first projects" and projects that are in progress are called "second projects." In addition, when there is no need to distinguish between past projects and projects that are in progress, they are simply called "projects."

As shown in FIG. 1, the information processing device 100 includes a generation unit 1 and a prediction unit 2. The information processing device 100 also stores a first project DB 3 and a second project DB 4.

The generation unit 1 generates a prediction model for predicting the actual estimated cost of the second project by machine learning. The actual estimated cost is an estimated value of the cost that will occur by the end of the second project. Note that there are multiple types of cost, such as manufacturing cost and sales cost, but in this embodiment, the cost is described as an economic resource consumed to achieve a specific purpose measured in monetary terms, and includes manufacturing cost, sales cost, and the like.
The prediction unit 2 uses the prediction model generated by the generation unit 1 to predict the actual estimated costs of the second target project.
The first project DB 3 is a database that stores the monitoring information of the first project for each first project.
The second project DB 4 is a database that stores the monitoring information of the second project for each second project.

<Project monitoring information>
The project monitoring information is information for monitoring the status of the project. The first project monitoring information may be composed of, for example, production number information, explanatory variables, and objective variables.

The production number information is information for identifying the first project. For example, the production number information includes, but is not limited to, a production number, a progress rate (%), and a production number name.
The production number is an identifier of the first project and can be expressed, for example, as an alphanumeric string.
The progress rate is a parameter that quantitatively indicates the progress of the first project.
The project name is the name of the first project, and can be expressed, for example, in a conceivable word.

The explanatory variables are variables that express the state of the first project. There are multiple types of explanatory variables. Examples of explanatory variables include, but are not limited to, the start time of work, the end time of work, the actual cost, the forecast cost, the estimated cost, the person in charge, the work hours, and the actual man-hours. For example, Boruta can be used to extract explanatory variable candidates, and the optimal explanatory variable can be selected from them, but the method of selecting explanatory variables is not limited to this.
The work start time is the start time (year, month, and date) of the first project.
The work completion time is the completion time (date) of the first project.
The actual cost is the cost incurred from the start of the work to a specified time.
Projected costs are costs that are expected to occur from a specified time until the end of the work.
The estimated cost is the cost expected to be incurred from the start of the work to the end of the work.
The person in charge is the person (or persons) in charge of the first project.
The working time is the time each person in charge spent working on the first project from the start of the work to a specified time.
The actual man-hours are the man-hours completed from the start of work to a specified time out of the total man-hours constituting the first project.
The predetermined time is any time between the start time and the end time of the work.

The objective variable is a variable that depends on the explanatory variables. For example, the objective variable can be the final actual cost, which is the actual cost at the end of the first project, but is not limited to this.

(Progress rate)
The progress rate can be calculated, for example, based on a time standard. For example, if the period from when work on a project starts to when it ends is 30 days, and the target period is the 15th day from when work starts, then the progress rate is 50%. Since the first project is a past project that has been completed, the progress rate at the current time is 100%. Here, the explanatory variables of the first project can be values that change according to the progress rate. The monitoring information for the first project can be configured as a set of explanatory variables for each progress rate.

For example, the actual cost as an explanatory variable is a set of actual costs at all progress rates between 0% and 100%. The actual cost at a progress rate of X% is the cost incurred between the start of work and the time corresponding to the progress rate of X%. Similarly, the forecast cost as an explanatory variable is a set of forecast costs at all progress rates between 0% and 100%. The forecast cost at a progress rate of X% is the cost expected to be incurred between the time corresponding to the progress rate of X% and the time the work is completed. Note that there are explanatory variables that do not change depending on the progress rate, such as the start and end times of work, but it is preferable to treat such explanatory variables as constants that take the same value depending on the progress rate.

On the other hand, the monitoring information for the second project can be composed of, for example, production number information and explanatory variables.

The production number information is information for identifying the second project. For example, the production number information includes, but is not limited to, a production number, a progress rate (%), and a production number name.
The production number is an identifier of the second project and can be expressed, for example, as an alphanumeric string.
The progress rate is a parameter that quantitatively indicates the progress of the second project.
The project name is the name of the second project, and can be expressed, for example, in a conceivable word.

The explanatory variables are variables that express the state of the second project. There are multiple types of explanatory variables. Examples of explanatory variables include, but are not limited to, the start time of work, the end time of work, actual cost, forecast cost, estimated cost, person in charge, work time, and actual man-hours. For example, Boruta can be used to extract explanatory variable candidates, and the optimal explanatory variable can be selected from them, but the method of selecting explanatory variables is not limited to this.
The work start time is the start time (year, month, and date) of the second project.
The work completion time is the completion time (date) of the second project.
The actual cost is the cost incurred from the start of the work to a specified time.
Projected costs are costs that are expected to occur from a specified time until the end of the work.
The estimated cost is the cost expected to be incurred from the start of the work to the end of the work.
The person in charge is the person (or persons) in charge of the second project.
The working time is the time each person in charge spent working on the second project from the start of the work to a specified time.
The actual man-hours are the man-hours completed from the start of work to a specified time out of the total man-hours constituting the second project.
The predetermined time is the present time between the start time and the end time of the work.
As already explained, the actual estimated cost of the second project is the sum of the actual cost and the forecast cost of the second project.

The progress rate of the second project can be calculated as the ratio of the period from the start of work to the end of work to the period from the start of work to the present. The actual cost of a progress rate of X%, which corresponds to the present, is the cost incurred from the start of work to the present. The forecast cost of a progress rate of X%, which corresponds to the present, is the cost expected to be incurred from the present to the end of work. Note that there are explanatory variables that do not change depending on the progress rate, such as the start and end of work, but it is preferable to treat such explanatory variables as constants that take the same value depending on the progress rate.

<Prediction model>
(Training)
The generation unit 1 can generate a prediction model by combining, for example, multiple decision trees used in a random forest. A random forest is a machine learning algorithm, and is an ensemble learning algorithm that uses multiple decision trees and makes predictions by majority vote. The decision tree can be configured as a tree-like logic that combines, for example, judgment conditions using explanatory variables. The judgment conditions can be designed as appropriate, and can be, for example, whether or not the average daily working hours of a person in charge are 5 hours or more.

The generation unit 1 can train the prediction model using the monitoring information of the first project as training data. For example, for each first project, the explanatory variables from the monitoring information of the first project at a period equivalent to a progress rate of 50% can be used as the input of the prediction model. In addition, a value based on the objective variable of the monitoring information of the first project can be used as the output of the prediction model. For example, the estimated cost overrun value obtained by subtracting the estimated cost from the final actual cost of the first project can be used as the output of the prediction model. The generation unit 1 trains the prediction model using monitoring information of a predetermined number of first projects.

Here, the generating unit 1 prepares a plurality of prediction models, and the output of the prediction models can be a plurality of stages of output. For example, the generating unit 1 prepares a first prediction model and a second prediction model. The input of the first prediction model targets the monitoring information of all the first projects, and can be an explanatory variable of the monitoring information at a time when the progress rate is 50%. The output of the first prediction model can be whether or not there is an estimated cost overrun (greater than 0M (0 yen)) obtained by subtracting the estimated cost from the final actual cost of the first project. Next, the input of the second prediction model targets the monitoring information of the first project in which there is an estimated cost overrun in the output of the first prediction model, and can be an explanatory variable of the monitoring information at a time when the progress rate is 50%. The output of the second prediction model can be whether or not the estimated cost overrun is 1M (1 million yen) or more. As a result, the output of the prediction model can be classified into three values: estimated cost overrun is 0M or less, greater than 0M and less than 1M, and 1M or more.

(prediction)
The prediction unit 2 predicts the actual estimated cost of the second project to be predicted using the trained prediction model. For example, the prediction unit 2 inputs explanatory variables from the monitoring information of the second project, that is, at the present time, i.e., at a time corresponding to a predetermined progress rate (preferably 50% or more, but may be less than 50%), into the prediction model. Then, the prediction unit 2 can obtain a value based on the actual estimated cost as the output of the prediction model. For example, the prediction unit 2 can obtain a value of the estimated cost overrun, which is obtained by subtracting the estimated cost from the actual estimated cost.

When the forecasting model is the first forecasting model or the second forecasting model described above, the forecasting unit 2 inputs the current explanatory variables from the monitoring information of the second project to be forecasted into the first forecasting model. Then, the forecasting unit 2 can obtain a value indicating whether or not there has been an estimated cost overrun (greater than 0M (0 yen)) as the output of the first forecasting model. If there has been an estimated cost overrun, the forecasting unit 2 inputs the current explanatory variables from the monitoring information of the second project into the second forecasting model. Then, the forecasting unit 2 can obtain a value indicating whether or not the estimated cost overrun is 1M (1 million yen) or more as the output of the second forecasting model. As a result, the estimated cost overrun can be classified into three values: 0M or less, greater than 0M and less than 1M, and 1M or more.
Note that 1M and 0M are merely examples, and the value may be greater than 1M or greater than 0M, or may be less than 1M or less than 0M.

The prediction unit 2 can output information including the output of the prediction model. For example, the display unit of the information processing device 100 can display the output information of the prediction unit 2 as shown in FIG. 2 on a screen. As shown in FIG. 2, the output information of the prediction unit 2 can be in a tabular format with "prediction results", "production number information", and "explanatory variables and features of the prediction results" as columns, and the second project as rows.

The "prediction result" indicates the output content of the prediction model. The "prediction result" is composed of an "item number", a "predicted value", and a "confidence factor".
"Item number" is a line number assigned to each second project.
"Predicted value" is the result according to the three-value classification of estimated cost overrun when the forecast model is a combination of the first and second forecast models already explained. "1: Overrun by 1 million yen" corresponds to 1 million or more. "2: Overrun by less than 1 million yen" corresponds to greater than 0 million and less than 1 million. "3: No problem" corresponds to 0 million or less.
"Confidence" is the reliability of the prediction and is indicated by a value between 0% and 100%. For example, the confidence can be calculated using bagging, but is not limited to this.

The "production number information" is the same as the production number information of the monitoring information of the second project.
"Explanatory variables and features of prediction results" shows the explanatory variables that contribute to the "prediction results.""Explanatory variables and features of prediction results" consists of a "list of explanatory variables" and a "feature ranking."
The "list of explanatory variables" is the same as the explanatory variables in the monitoring information of the second project.
The "feature ranking" indicates the rank of explanatory variables that contribute to the "prediction result". The higher the rank, the greater the contribution rate of that explanatory variable to the output of the predicted value. For example, the contribution rate of each variable can be calculated using the Shapley Additive exPlanations (SHAP) algorithm, but is not limited to this.

According to the first embodiment, it is possible to extract second projects whose estimated cost exceeds the estimated cost by 1M or more. Therefore, it is possible to detect early signs of deterioration in the project through wide-area monitoring.

(Decision on whether to subject to intensive monitoring)
The management department, who has obtained the output information in Figure 2, judges whether or not to make the second project, which has an estimated cost overrun of 1M, a target for intensive monitoring. Conventionally, even if AI detected signs of deterioration in a project, the basis for the AI's prediction results was a black box. For this reason, it was not easy for the management department to track down the causes of the signs of deterioration in the project based on the AI's prediction results, and it required a large amount of human resources to judge whether or not to make it a target for intensive monitoring.

The "Feature Ranking" in Figure 2 can be said to provide the basis for why the estimated cost overrun was 1M or more. By referring to the "Feature Ranking," the management department can easily identify the explanatory variables that contributed significantly to the prediction that the estimated cost overrun was 1M or more. As a result, it becomes easier to decide whether or not to subject the second project to priority monitoring, and the human cost of deciding whether or not to subject it to priority monitoring can be reduced.

[process]
The process executed by the information processing device 100 is as shown in Fig. 3. That is, first, the generating unit 1 generates a prediction model (step S1). Next, the generating unit 1 trains the prediction model using monitoring information of the first project at a predetermined progress rate (step S2). Next, the predicting unit 2 uses the prediction model to output a predicted value of the estimated cost overrun of the target second project and explanatory variables contributing to the predicted value as prediction grounds (step S3). The management department judges, based on the prediction grounds, whether or not to target the second project showing signs of deterioration as a target for focused monitoring.

Second Embodiment
In the explanation of the second embodiment, differences from the first embodiment will be explained, and overlapping points will be omitted. In the first embodiment, the explanatory variables of the monitoring information of the first project, which is the training data, were explanatory variables at a time when the progress rate was equivalent to 50%. In the second embodiment, the time of the explanatory variables of the first project used in the training data is periodic.

For example, if the duration of the first project, i.e., the period from the start of work to the end of work, is several months, the period of the explanatory variables used in the training data, i.e., the training date (learning date), is set to the 25th of each month.
Therefore, if work on the first project of work number A starts on 4/15 and ends on 6/30, the training dates for the first project of work number A will be 4/25, 5/25, and 6/25. In other words, from the monitoring information for the first project of work number A, the explanatory variables on 4/25 (explanatory variables with a progress rate equivalent to 4/25), the explanatory variables on 5/25 (explanatory variables with a progress rate equivalent to 5/25), and the explanatory variables on 6/25 (explanatory variables with a progress rate equivalent to 6/25) for a total of three times will be input into the prediction model.
Also, if the work start date for the first project of work number B is 5/1 and the work end date is 9/15, the training dates for the first project of work number B will be 5/25, 6/25, 7/25, and 8/25. In other words, from the monitoring information for the first project of work number B, the explanatory variables for 5/25 (explanatory variables with a progress rate equivalent to 5/25), 6/25 (explanatory variables with a progress rate equivalent to 6/25), 7/25 (explanatory variables with a progress rate equivalent to 7/25), and 8/25 (explanatory variables with a progress rate equivalent to 8/25) for a total of four times will be input into the prediction model.
In addition, if the work start date for the first project of work number C is 6/1 and the work end date is 7/10, the training date for the first project of work number C will be 6/25. In other words, the explanatory variables for 6/25 (explanatory variables with a progress rate equivalent to 6/25) from the monitoring information for the first project of work number C are used as input for the prediction model for one session.

As a result, explanatory variables at multiple types of progress rates are input into the prediction model for most of the first project. Using the prediction model trained in this way, the prediction unit 2 predicts the actual estimated cost of the second project. In this case, even if the progress rate of the second project is low (e.g., about 30%) and explanatory variables at a period equivalent to the low progress rate are input into the prediction model, it has been confirmed that the confidence level of the predicted value output by the prediction unit 2 (see Figure 2) is sufficiently high.

According to the second embodiment, the training days are set to regular dates, and explanatory variables for a plurality of types of progress rates for the same first project can be input to the prediction model, thereby making it possible to forecast the actual estimated cost of the second project earlier.
In addition, by regularizing the training dates, the input to the forecasting model for all first projects can be systematized, simplifying the processing required for training.

<Detecting signs of deterioration using forecast values>
As the project progresses, actual values from the start of work to the present time are acquired for each explanatory variable constituting the monitoring information of the project. For example, the acquired actual values may be actual values for one day, or may be actual values for one month with the 25th of each month as a dividing line. In addition, a forecast value can be prepared for the acquired actual values. The forecast value is a planned value (or a predicted value) of the actual value at a certain point in the future. The forecast value can be determined by a person in a management department or the like based on the progress of the project, social conditions, etc., for example, but is not limited to this, and may be determined by AI, for example. For example, a forecast value for the next month (as of the 25th of the next month) can be determined for the actual value for the current month as of the 25th of the current month. In this embodiment (first and second embodiments), the forecast value can be used to improve the accuracy of detecting signs of deterioration.

As already explained, there are various types of explanatory variables, such as the start time of work, the end time of work, actual cost, forecast cost, estimated cost, person in charge, working hours, and actual man-hours. Actual values are obtained for each type of explanatory variable, and forecast values can technically be prepared for each type of explanatory variable. However, there are explanatory variables such as "start time of work" that cannot be planned after the project has started and cannot have forecast values. There are also explanatory variables such as "forecast cost" that are not practical for determining forecast values (in other words, it is necessary to forecast the forecast, and although it can be conceived, it is extremely difficult to find a use for it). Therefore, it is assumed that a human being can specify the type of explanatory variable for which a forecast value can be prepared. An explanatory variable specified as one for which a forecast value can be prepared is called a "specified explanatory variable". The information processing device 100 can determine whether the explanatory variable used in the deterioration prediction of this embodiment is a specified explanatory variable based on an input from the user. In this embodiment, we will continue the explanation assuming that "start time of work" and "estimated cost" are not designated explanatory variables, but this embodiment does not prevent "start time of work" and "estimated cost" from being technically treated as designated explanatory variables.

For example, suppose that the actual value for the previous month, processing cost, which is a specified explanatory variable for a certain project, is obtained as of the 25th of the previous month. At this time, a forecast value for the current month (one month ahead) is prepared as of the 25th of the previous month for the obtained actual value for the previous month, and the user can input the forecast value into the information processing device 100. After that, the actual value for the current month can be obtained as of the 25th of the current month (before closing), so that the forecast value for the current month can be compared with the actual value for the current month. The result of the comparison can be obtained as an estimated value that indicates an evaluation of the actual value against the forecast.

(Assumed value)
For example, the expected value can be calculated as the ratio of the actual value to the forecast value. This ratio is called the "expected rate." In the above example, the expected processing cost rate (for the current month) can be calculated as expected rate = (actual value for the current month) / (forecast value for the current month (prepared in the previous month)). The expected rate can be calculated not only for the current month, but also for the previous month, the month before last, etc. In other words, the expected processing cost rate (for the nth month) can be calculated as expected rate for the nth month = (actual value for the nth month) / (forecast value for the nth month (prepared in the (n-1)th month)). In addition, the expected rates can be calculated not only for processing costs, but also for other types of explanatory variables such as outsourcing costs and labor hours.

For example, the expected value can be found as the increase or decrease in the actual value relative to the forecast value. This increase or decrease is called the "expected deviation." In the above example, the expected deviation (for the current month) can be found as expected deviation = (actual value for the current month) - (forecast value for the current month (prepared in the previous month)). The expected deviation can also be found not only for the current month, but also going back to the previous month, the month before, etc. In other words, the expected deviation (for the nth month) can be found as expected deviation for the nth month = (actual value for the nth month) - (forecast value for the nth month (prepared in the (n-1)th month)). The expected deviation can also be found for other types of explanatory variables, not just processing costs, such as outsourcing costs and labor hours.

This embodiment is characterized in that the expected values obtained for the designated explanatory variables of the project are set as the explanatory variables of the project. More specifically, the explanatory variables of the project are the forecast values at a second time point (a time point in the future from the first time point. If the project is a second project, the latest second time point is the present time) for the actual values at a first time point (a point in the past during the project period) for the designated explanatory variables of the project, and the expected values obtained using the actual values at the second time point. Note that forecast values are prepared at appropriate times not only for the second project in progress but also for the first project that has been completed, so that expected values can be added to the explanatory variables for both the first project and the second project.

The applicant incorporated the expected value as an explanatory variable and performed the prediction of deterioration in wide-area monitoring described in the first embodiment. As a result, it was confirmed that the accuracy of early detection of signs of deterioration in the project in wide-area monitoring was improved. Specifically, the value of "confidence" in Figure 2 was generally improved compared to the case where the expected value was not incorporated as an explanatory variable. In addition, as described in the first embodiment, the expected value as an explanatory variable can be the basis for the estimated cost overrun of 1M or more. Figure 2 shows the expected value "expected processing cost rate (December 2022)" being presented in the "feature ranking" (symbol A).

(Expected value fluctuation)
Although the explanation has been given on how to obtain the expected value on a monthly basis, the expected value fluctuation value on a monthly basis can be obtained using the expected value on a monthly basis. The expected value fluctuation value on a monthly basis is a value that indicates the time change of the expected value for each month. When the expected value is an expected rate, the expected value fluctuation value becomes the expected rate fluctuation value. The expected rate fluctuation value for the current month can be obtained as follows: expected rate fluctuation value for the current month = {(expected rate for the current month) - (expected rate for the previous month)} - {(expected rate for the previous month) - (expected rate for the month before last)}. In addition, the expected rate fluctuation value can be obtained not only for the current month, but also going back to the previous month, the month before last, etc. In other words, the expected rate fluctuation value (for the nth month) can be obtained as follows: expected rate fluctuation value for the nth month = {(expected rate for the nth month) - (expected rate for the (n-1)th month)} - {(expected rate for the (n-1)th month) - (expected rate for the (n-2)th month)}. The expected rate fluctuation value can be obtained regardless of the type of explanatory variable (i.e., for processing costs, outsourcing costs, labor hours, etc.).

In addition, when the expected value is an expected deviation, the expected value fluctuation value is the expected deviation fluctuation value. The expected deviation fluctuation value for the current month can be calculated as follows: expected deviation fluctuation value for the current month = {(expected deviation for the current month) - (expected deviation for the previous month)} - {(expected deviation for the previous month) - (expected deviation for the month before last)}. The expected deviation fluctuation value can also be calculated not only for the current month, but also for the previous month, the month before last, etc. In other words, the expected deviation fluctuation value (for the nth month) can be calculated as follows: expected deviation fluctuation value for the nth month = {(expected deviation for the nth month) - (expected deviation for the (n-1)th month)} - {(expected deviation for the (n-1)th month) - (expected deviation for the (n-2)th month)}. The expected deviation fluctuation value can be calculated regardless of the type of explanatory variable (i.e., for processing costs, outsourcing costs, labor hours, etc.).

This embodiment is characterized in that the expected value variation value obtained for the specified explanatory variable of the project is set as the explanatory variable of the project. More specifically, the expected rate variation value obtained using the expected value at the third time point (a past point during the project period) and the expected value at the fourth time point (a time point in the future from the third time point. If the project is the second project, the latest second time point is the current time) is set as the explanatory variable of the project. Note that, like the expected value, the expected value variation value can be added to the explanatory variables for both the first project and the second project. Also, the expected value variation value (expected rate variation value, expected deviation variation value) is not limited to being calculated as the deviation of the deviation as in the above formula, but may be calculated using a ratio. That is, for example, the expected value fluctuation value for the nth month may be {(expected value for the nth month) - (expected value for the (n-1)th month)}/{(expected value for the (n-1)th month) - (expected value for the (n-2)th month)}, or the expected value fluctuation value for the nth month may be {(expected value for the nth month)/(expected value for the (n-1)th month)}/{(expected value for the (n-1)th month)/(expected value for the (n-2)th month)}.

The applicant incorporated the expected value variation value as an explanatory variable to perform the prediction of deterioration in wide-area monitoring described in the first embodiment. As a result, it was confirmed that, as with the expected value, the accuracy of early detection of signs of deterioration in the project in wide-area monitoring was improved. Furthermore, the expected value variation value as an explanatory variable can be the basis for the estimated cost overrun of 1M or more, as described in the first embodiment.

(Assumed cumulative value)
Furthermore, the expected value on a monthly basis can be used to calculate, for example, a cumulative expected value for three months (e.g., quarterly). The cumulative expected value for three months is the sum of the expected values (three) for each month included in the three months in question. If the expected value is an expected rate, the cumulative expected value for three months is the cumulative expected rate for three months. If the expected value is an expected deviation, the cumulative expected value for three months is the cumulative expected deviation for three months. The cumulative expected deviation can be calculated regardless of the type of explanatory variable (i.e., for processing costs, outsourcing costs, labor hours, etc.).

Specifically, the cumulative expected value for the current month (for three months) can be calculated as the cumulative expected value for three months = the expected value for the current month + the expected value for the previous month + the expected value for the month before last. The cumulative expected value for three months can also be calculated not only for the current month, but also going back to the previous month, the month before last, etc. In other words, the cumulative expected value for three months (for the nth month) can be calculated as the cumulative expected value for three months in the nth month = the expected value for the nth month + the expected value for the (n-1)th month + the expected value for the (n-2)th month.

The period for calculating the cumulative expected value is not limited to three months, but may be six months, twelve months, etc. Specifically, the cumulative expected value for m months (for the nth month) can be calculated as follows: cumulative expected value for m months = expected value for the nth month + expected value for the (n-1)th month + ... + expected value for (n-(m-1)) months.

This embodiment is characterized in that the cumulative expected value obtained for the specified explanatory variable of the project is set as the explanatory variable of the project. More specifically, the cumulative expected value obtained by adding up the expected values obtained from the fifth time point (a past point during the project period) to the sixth time point (a time point in the future than the fifth time point. If the project is the second project, the latest second time point is the current time) is set as the explanatory variable of the project. Note that, like the expected value and expected value variance, the cumulative expected value can be added to the explanatory variable for both the first project and the second project.

The applicant incorporated the estimated cumulative value as an explanatory variable to perform signs of deterioration in wide-area monitoring as described in the first embodiment. As a result, it was confirmed that, as with the estimated value, the accuracy of early detection of signs of deterioration in a project in wide-area monitoring was improved. In addition, the estimated value fluctuation value as an explanatory variable can be the reason why the estimated cost estimate exceeded the estimated cost by 1M or more, as described in the first embodiment. Compared to using a single estimated value, using the estimated cumulative value obtained by adding up multiple estimated values has the advantage of making it easier to identify the discrepancy between the forecast value (cumulative) and the actual value (cumulative), making it possible to more accurately capture signs of deterioration.

[Modification]
(a): In the first and second embodiments, the progress rate is calculated based on a time standard using the period of the project. However, for example, the progress rate may be calculated based on the degree of completion of the work to be done in the project.
(b): In the first embodiment, explanatory variables at a time point corresponding to a progress rate of 50% in the monitoring information of the first project are input to the prediction model for each first project. However, for example, explanatory variables at a time point corresponding to any same progress rate other than 50% may be input to the prediction model for each first project. Also, explanatory variables at a time point corresponding to a different progress rate for each first project may be input to the prediction model.
(c): In the second embodiment, by regularizing the training days, multiple types of progress rates are substantially selected for the same first project, and explanatory variables at the selected progress rates are input to the prediction model. However, for example, a user of the information processing device 100 may operate the input unit to select multiple types of arbitrary progress rates for the same first project, and input explanatory variables at the selected progress rates to the prediction model.
(d): In the present embodiment, the case where the expected value is calculated on a monthly basis has been described. However, the expected value may be calculated on a cycle other than monthly, for example, on a biweekly (every two weeks) basis, weekly basis, or daily basis.

(e) It is also possible to realize a technology that appropriately combines the various technologies described in this embodiment.
(f): The software described in this embodiment can be realized as hardware, and vice versa.
(g) Other modifications may be made to the hardware, software, flow charts, etc. as appropriate without departing from the spirit and scope of the present invention.

Reference Signs List 100 Information processing device 1 Generation unit 2 Prediction unit 3 First project DB
4. Second Project DB

Claims (4)

a generation unit that generates a prediction model using explanatory variables constituting monitoring information, which is information for monitoring the status of a completed first project, as input information and using an estimated cost overrun of the first project as output information;
a prediction unit that inputs explanatory variables of a second project in process into the prediction model and outputs a predicted value of an estimated cost overrun of the second project and explanatory variables that are the basis of the predicted value;
The generation unit is
generating the prediction model using as input information a forecast value at time B for an actual value at time A in a specified explanatory variable designated from the explanatory variables of the first project, and an expected value calculated using the actual value at time B;
The prediction unit is
an information processing device that inputs explanatory variables of the second project, including a forecast value at a second time point for an actual value at a first time point and an expected value calculated using the actual value at the second time point, for a specified explanatory variable specified from the explanatory variables of the second project, into the prediction model, and outputs a forecast value of an estimated cost overrun of the second project and the explanatory variables that are the basis for the forecast value . The generation unit is
generating the prediction model using, as the input information, an assumed rate fluctuation value calculated using the assumed value at time C and the assumed value at time D in a specified explanatory variable designated from the explanatory variables of the first project;
The prediction unit is
2. The information processing device according to claim 1, wherein explanatory variables of the second project , including an assumed rate fluctuation value calculated using the assumed value at a third time point and the assumed value at a fourth time point , are input to the forecast model, and a predicted value of an estimated cost overrun of the second project and the explanatory variables that are the basis for the predicted value are output . The generation unit is
generating the prediction model using as the input information a cumulative estimated value obtained by adding up the estimated values obtained from time point E to time point F for a specified explanatory variable specified from the explanatory variables of the first project;
The prediction unit is
3. The information processing device according to claim 1, further comprising: inputting explanatory variables of the second project , including a cumulative estimated value obtained by adding up the estimated values obtained from the fifth time point to the sixth time point, into the prediction model; and outputting a predicted value of an estimated cost overrun of the second project and the explanatory variables that are the basis for the predicted value . An information processing device,
a first step of generating a prediction model using explanatory variables constituting monitoring information, which is information for monitoring the status of a completed first project, as input information and using an estimated cost overrun of the first project as output information;
A second step is carried out in which explanatory variables of a second project in process are inputted into the prediction model, and a predicted value of an estimated cost overrun of the second project and explanatory variables on which the predicted value is based are outputted;
In the first step,
generating the prediction model using as input information a forecast value at time B for an actual value at time A in a specified explanatory variable designated from the explanatory variables of the first project, and an expected value calculated using the actual value at time B;
In the second step,
an estimated cost overrun for the second project and the explanatory variables serving as the basis for the estimated cost overrun are output as input to the prediction model, the estimated cost overrun being calculated based on the actual value at the first time point and the forecast value at the second time point, for a specified explanatory variable specified from the explanatory variables of the second project; and

Images