WO2022157876A1 - Planning support device, program, and planning support method - Google Patents

Planning support device, program, and planning support method Download PDF

Info

Publication number
WO2022157876A1
WO2022157876A1 PCT/JP2021/001992 JP2021001992W WO2022157876A1 WO 2022157876 A1 WO2022157876 A1 WO 2022157876A1 JP 2021001992 W JP2021001992 W JP 2021001992W WO 2022157876 A1 WO2022157876 A1 WO 2022157876A1
Authority
WO
WIPO (PCT)
Prior art keywords
kpi
event
management
management information
prediction
Prior art date
Application number
PCT/JP2021/001992
Other languages
French (fr)
Japanese (ja)
Inventor
輝明 下田
知章 掛田
Original Assignee
株式会社日立製作所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社日立製作所 filed Critical 株式会社日立製作所
Priority to PCT/JP2021/001992 priority Critical patent/WO2022157876A1/en
Publication of WO2022157876A1 publication Critical patent/WO2022157876A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling

Definitions

  • the present invention relates to a plan support device, a program, and a plan support method that support planning.
  • KPI Key Performance Indicator
  • Patent Document 1 discloses a management information display method realized by executing a program stored in the memory of a computer equipped with a CPU, a memory, and a storage device. wherein profit and loss data for each product stored in the storage device is read into the memory, each profit and loss amount obtained by sorting the products in descending order of profit amount is displayed on a display device, and one of the products is When specified, the profit and loss data and the profit rate for each region of the specified product stored in the storage device are read into the memory, and each profit and loss amount obtained by sorting the regions in descending order of profit amount is stored in the profit rate. Also, a method of displaying management information is described, which is characterized by displaying on the display device together with the management information.
  • Patent Document 1 analyzes the sales, profit and loss, profit rate, and growth rate for each product using the management information display method described above, and identifies products that currently contribute to the overall profit and loss, and products that are promising for future profit acquisition. , it is stated that it has the effect of identifying products whose future earnings are expected to be low and accurately reflecting the sales strategy.
  • conventional plan management systems have functions to display planned items to be managed as various indicators, and to predict the impact of the current plan and display changes in various indicators. be able to.
  • the present invention has been made in view of such a background, and aims to provide a plan support device, a program, and a plan support method that support planning including predictions that are difficult to formulate.
  • the planning support apparatus includes an event reflecting unit that changes management items constituting management information related to one or more management targets according to an event that changes according to a predetermined rule; a prediction unit that predicts a prediction target item related to the management target based on the management information changed by the event; and a KPI calculation unit that calculates a KPI based on the management information and the prediction target item.
  • the prediction unit predicts the past management information and the past results of the prediction target item using a trained machine learning model as teacher data.
  • FIG. 4 is a data configuration diagram of a management information database according to the embodiment;
  • FIG. 4 is a data configuration diagram of an event information database according to the embodiment;
  • FIG. 4 is a data configuration diagram of a KPI information database according to the embodiment;
  • FIG. 4 is a data configuration diagram of a supplier database according to the embodiment;
  • FIG. 3 is a data configuration diagram of a parts database according to the embodiment;
  • FIG. 4 is a data configuration diagram of a manufacturing management database according to the embodiment;
  • 4 is a data configuration diagram of management information according to the embodiment;
  • FIG. 4 is a data configuration diagram of a KPI calculation result table according to the embodiment;
  • FIG. 4 is a data configuration diagram of a KPI threshold determination result table according to the embodiment; It is a display example of KPI of "all products" in all events according to the present embodiment. It is a display example of KPIs of events in which the targets of all KPIs according to the present embodiment are achieved. It is an example of a display which specified KPI which concerns on this embodiment. It is a display example of specifying a KPI with a condition according to the present embodiment.
  • 1 is a diagram showing a recording medium according to this embodiment; FIG.
  • the planning support device of the present embodiment relates to planning of parts procurement in a manufacturer that manufactures a product that is a large machine such as a railway vehicle, and calculates and displays KPIs when parts suppliers and delivery dates are changed. .
  • the user of the planning support device sets an event to change the ordering information (management information) including the part supplier, delivery date, order quantity, etc. due to the change in the manufacturing deadline of the product.
  • the planning support device predicts the delivery date of parts when an event (plan change) is executed, and calculates KPI based on the prediction result.
  • KPIs include the rate of coincidence between the delivery date and arrival date of ordered parts, the number of occurrences of rescheduling at the manufacturing site due to delays in the arrival of ordered parts, the total manufacturing work time, the number of manufacturing delay days, manufacturing costs, and the like.
  • the planning support device uses a machine learning model trained with learning data (teacher data) whose input is past order information and whose output (correct label) is the arrival date. Predict.
  • the user can predict (simulate) how the KPI will change if the procurement plan (draft) is changed by changing the ordering party and delivery date of parts due to the change in the manufacturing deadline of the product. become.
  • By finding events (changes in plans/plan proposals) that satisfy important KPIs users will be able to formulate optimized plans. Moreover, even when the plan is changed during execution of the plan, the user can make the optimum change.
  • FIG. 1 is a functional block diagram of a planning support device 100 according to this embodiment.
  • the planning support device 100 is a computer and includes a control section 110 , a storage section 120 and an input/output section 180 .
  • User interface devices such as a display, a keyboard, and a mouse are connected to the input/output unit 180 .
  • the storage unit 120 is composed of storage devices such as ROM (Read Only Memory), RAM (Random Access Memory) and SSD (Solid State Drive).
  • the storage unit 120 stores a management information database 210, an event information database 230, a KPI information database 240, a customer database 250, a parts database 260, a manufacturing management database 270, a program 121, and a prediction model 122.
  • the management information database 210, the event information database 230, the KPI information database 240, the supplier database 250, the parts database 260, and the manufacturing management database 270 are respectively represented by a management information DB (database), an event information DB, and a KPI information database.
  • DB client DB
  • parts DB and manufacturing control DB.
  • the program 121 includes a description of the procedure for planning support processing (see FIG. 8, which will be described later).
  • the prediction model 122 is a machine learning model used to predict the arrival date 218 contained in the management information database 210 (see FIG. 2 described later).
  • FIG. 2 is a data configuration diagram of the management information database 210 according to this embodiment.
  • the management information database 210 in this embodiment stores order information for parts.
  • the management information database 210 is, for example, tabular data, one row (record) indicates ordering information for parts, and includes an order number 211, an order request date 212, a product 213, a part 214, a part 215, and a customer 216. , due date 217, arrival date 218, valid flag 219 (shown as valid F in FIG. 2), and supplementary information 220 columns (attributes).
  • attributes such as order quantity and price may be included.
  • the order number 211 is the identification information of the order.
  • the order placement date 212 is the date the order was placed.
  • the product 213 is the name or identification information of the product in which the ordered parts are used.
  • a part 214 is a part of the product 213 in which the ordered part is used.
  • the part 215 is the name or identification information of the ordered part.
  • Supplier 216 is the name or identification information of the supplier.
  • the delivery date 217 is the delivery date of the part.
  • the arrival date 218 is the date when the ordered parts were delivered. In FIG. 2, none of the parts have been delivered and the arrival date 218 is blank.
  • a valid flag 219 indicates whether the order is valid or invalid.
  • the validity flag 219 is valid at the time of ordering, but becomes invalid when the order is canceled or when the production of the part is finished.
  • Supplementary information 220 is additional information related to the order or parts.
  • FIG. 3 is a data configuration diagram of the event information database 230 according to this embodiment.
  • the event information database 230 includes information on events indicating changes in parts procurement plans, such as changes in product manufacturing plans.
  • the event information database 230 is data in, for example, a table format, one row (record) indicates an event, and columns of identification information 231 (shown as # in FIG. 3), event content 232, and event reflection content 233 ( attributes).
  • the identification information 231 is identification information of the event.
  • the event content 232 indicates the content of the plan change. Examples of changes include postponing or bringing forward the deadline for completion of manufacturing of products, and discontinuing manufacturing itself. In addition, there are changes in parts suppliers and confirmation of part delivery dates.
  • the event reflection content 233 indicates the change content (rule indicating change procedure) of the attribute (management item) included in the management information database 210 (see FIG. 2) when the event is executed. Examples of changes (examples of rules) include postponing or bringing forward the delivery date 217 in line with the product manufacturing completion deadline, changing the order request date 212 and supplier 216 due to a change in the supplier, and supplementing the confirmation of the parts delivery date. For example, the information 220 is updated to "confirmed one week ago".
  • FIG. 4 is a data configuration diagram of the KPI information database 240 according to this embodiment.
  • the KPI information database 240 includes KPI information that is an index for evaluating a plan.
  • the KPI information database 240 is, for example, tabular data, one row (record) indicates a KPI, and columns of identification information 241 (described as # in FIG. 4), KPI content 242, and KPI threshold setting 243 ( attributes).
  • the identification information 241 is identification information of the KPI.
  • the KPI content 242 indicates the content (meaning) of the KPI. Examples of the contents include the rate of coincidence between the delivery date and the arrival date of the ordered parts, the total manufacturing work time, and the manufacturing cost.
  • the KPI threshold setting 243 indicates the target value of the KPI. The target value is not limited to the numerical value itself, and may be an increase/decrease value or an increase/decrease ratio compared to before the plan change.
  • FIG. 5 is a data configuration diagram of the customer database 250 according to this embodiment.
  • the supplier database 250 is a database containing information on suppliers of parts.
  • the supplier database 250 is, for example, data in tabular form, one row indicates a supplier, a supplier 251, the current number of orders placed 252, the current order amount 253, the cumulative number of orders placed in the past six months 254, Includes columns (attributes) for Cumulative Order Amount 255 for the past 6 months and Reliability 256 for the past 6 months.
  • the supplier 251 is the name or identification information of the parts supplier, and corresponds to the supplier 216 (see FIG. 2).
  • the current number of orders placed 252 and the current order amount 253 are the number of orders placed to the supplier 251 at the present time and the order amount, respectively.
  • the cumulative number of orders placed in the past 6 months 254 and the cumulative order amount 255 in the past 6 months are the number of orders placed and the amount of orders placed in the past 6 months, respectively.
  • the reliability of the past 6 months 256 is an index calculated from the rate of coincidence between delivery dates and arrival dates of parts in the past 6 months, the quality of parts delivered in the past 6 months, and the like.
  • FIG. 6 is a data configuration diagram of the parts database 260 according to this embodiment.
  • the parts database 260 is a database containing information on product parts.
  • the parts database 260 is, for example, data in tabular form, one row indicates a part, and columns of a part 261, a drawing 262, a design difficulty level 263, a manufacturing difficulty level 264, an estimated number of days for designing 265, and an estimated number of days to manufacture 266. (attributes).
  • the part 261 is the name or identification information of the part and corresponds to the part 215 (see FIG. 2).
  • Drawing 262 shows the drawing data of the part.
  • the design difficulty level 263 and the manufacturing difficulty level 264 indicate the difficulty level of designing the part and the difficulty level of manufacturing the part, respectively, and are indicated by levels from A to E.
  • the estimated design days 265 and estimated manufacturing days 266 indicate the number of days required for part design and the number of days required for part manufacturing, respectively.
  • FIG. 7 is a data configuration diagram of the manufacturing management database 270 according to this embodiment.
  • the manufacturing management database 270 contains information related to product manufacturing.
  • the manufacturing management database 270 is, for example, tabular data, one row indicates one product to be manufactured, and identification information 271 (shown as # in FIG. 7), product 272, orderer 273, manufacturing start date 274. contains the columns (attributes) of
  • the identification information 271 is identification information for each manufactured product.
  • Product 272 indicates the product to be manufactured and corresponds to product 213 (see FIG. 2).
  • the orderer 273 indicates the person who ordered the product 272 and is the delivery destination of the manufactured product 272 .
  • the manufacturing start date 274 indicates planned values and actual results for the manufacturing start date.
  • the manufacturing management database 270 may further include planned values and actual results such as manufacturing completion dates, shipping dates, periods of each process involved in manufacturing, personnel, and costs.
  • control unit 110 includes a CPU (Central Processing Unit) and includes a learning unit 111 , an event reflection unit 112 , a prediction unit 113 , a KPI calculation unit 114 and a result output unit 115 .
  • CPU Central Processing Unit
  • the learning unit 111 prepares learning data (teacher data) and trains and generates a prediction model 122, which is a machine learning model.
  • the learning data is the management information database 210 (see FIG. 2), which is order information for parts that have already been delivered. Specifically, data other than the arrival date 218 in the management information database 210 becomes the input data, and the arrival date 218 becomes the correct label (output data).
  • the input data may include the supplier database 250 (see FIG. 5) and the parts database 260 (see FIG. 6).
  • the prediction model 122 trained using such learning data and correct labels is machine learning used when predicting the arrival date 218 from information such as the parts 215, the order request date 212, the supplier 216, and the delivery date 217. Become a model.
  • the event reflection unit 112 changes the management information database 210 according to the selected event (see the event information database 230 shown in FIG. 3). Specifically, as will be described later, the management information 210A (see FIG. 9 described later), which is equivalent data to the management information database 210, is changed (see steps S12 and S13 in FIG. 8 described later). For example, when an event whose identification information 231 is “event 1” is selected, the event reflection unit 112 changes the delivery date 217 of the record whose product 213 is “product A” in the management information database 210 to 3 postpone for a month.
  • the prediction unit 113 predicts the arrival date 218 using the prediction model 122 .
  • the arrival date 218 is predicted using the prediction model 122 with the management information database 210 (see management information 210A shown in FIG. 9 to be described later) as input data.
  • the input data may include the supplier database 250 (see FIG. 5) and the parts database 260 (see FIG. 6).
  • the KPI calculation unit 114 calculates KPIs for individual products and all products according to the KPI contents 242 for the KPIs in the KPI information database 240 (see FIG. 4). For example, for the KPI whose identification information 241 is "KPI1", the KPI calculation unit 114 calculates the matching rate between the delivery date 217 and the arrival date 218 in the management information database 210 for individual products and all products. Further, for example, the KPI calculation unit 114 calculates the KPIs whose identification information 241 is “KPI3” to “KPI5” by referring to the manufacturing management database 270 (see FIG. 7). The KPI calculation unit 114 may also refer to the supplier database 250 and the parts database 260 .
  • the result output unit 115 outputs (displays) the KPI calculated by the KPI calculation unit 114 to a display connected to the input/output unit 180 (see FIG. 1).
  • the result output unit 115 does not always output all KPIs, and may output KPIs specified based on user's designation.
  • FIG. 8 is a flow chart of planning support processing according to the present embodiment.
  • control unit 110 starts processing to repeat steps S 12 to S 16 for each of the unchanged event and the event included in event information database 230 .
  • a no-change event is an event in which the management information database 210 (see FIG. 2) is not changed. Note that in the repeated processing of steps S12 to S16, events without change are processed first.
  • step S ⁇ b>12 the event reflecting unit 112 reads the content of the management information database 210 into a different area from the management information database 210 in the storage unit 120 .
  • this read data is referred to as management information 210A (see FIG. 9 described later).
  • step S13 the event reflecting unit 112 changes the management information 210A according to the event.
  • step S14 the prediction unit 113 uses the management information 210A as input data, predicts the arrival date using the prediction model 122, and stores the arrival date in the management information 210A.
  • FIG. 9 is a data configuration diagram of the management information 210A according to this embodiment.
  • the management information 210A has the same configuration as the management information database 210 (see FIG. 2), and the columns (attributes) are assigned the same reference numerals as those of the management information database 210.
  • FIG. 10 is a data configuration diagram of the KPI calculation result table 310 according to this embodiment.
  • the KPI calculation result table 310 stores KPIs calculated by the KPI calculator 114 .
  • the KPI calculation result table 310 is data in tabular form, for example, one row (record) indicates the KPI after the event is reflected, and includes columns (attributes) of a product 311, an event 312, and KPIs 313-317.
  • a product 311 indicates a product to be calculated, and includes “all products” which are all products in addition to individual products.
  • An event 312 indicates an event that caused the management information database 210 to be changed.
  • a record whose event 312 is "-" is a KPI in a no-change event and indicates a reference value.
  • Other records show KPIs based on the management information 210A (see FIG. 9) changed based on the event indicated by the event 312, and include increase/decrease values/ratios from the reference value.
  • KPIs 313-317 indicate calculated KPIs.
  • KPI1 of the KPI 313 of the record in which the product 311 is “all products” and the event 312 is “event 1” has a matching rate of 77% for the due date and the arrival date of ordered parts (see FIG. 4), which is the reference value. It shows that it increased by 7 from “KPI3" of KPI 315 indicates that the total manufacturing work time is 1100 hours, which is 110% of the reference value.
  • step S16 the KPI calculation unit 114 determines whether or not the calculated KPI satisfies the target value (see KPI threshold setting 243 shown in FIG. 4).
  • the determination result table 320 Stored in the determination result table 320 (see FIG. 11 described later).
  • FIG. 11 is a data configuration diagram of the KPI threshold determination result table 320 according to this embodiment.
  • the KPI threshold determination result table 320 stores whether or not the KPI determined by the KPI calculation unit 114 has achieved the target value.
  • the KPI threshold determination result table 320 is, for example, tabular data. One row (record) indicates whether or not the KPI target value has been achieved after the event is reflected. )including.
  • the product 321, event 322, and KPIs 323-327 are the same as the product 311, event 312, and KPIs 313-317 of the KPI calculation result table 310 (see FIG. 10). However, the KPIs 323 to 327 store whether or not the KPI has achieved the target value (True/False).
  • step S17 the result output unit 115 inquires of the user and acquires the KPIs, their conditions, and the products to be displayed on the display (not shown) connected to the input/output unit 180 (see FIG. 1). do.
  • step S18 the result output unit 115 displays the KPI acquired in step S17, its conditions, and the KPI corresponding to the product on the display.
  • a display example will be described below.
  • FIG. 12 is a display example 330 of KPIs for "all products" in all events according to this embodiment. More specifically, it is a display example 330 when "all products" are selected as display target products in step S17 and all KPIs are selected. In the display example 330, the vertical direction is for each event, and the horizontal direction is for each KPI. By referring to the display example 330, the user can comprehend the impact of plan changes (differences in plan content) on KPIs.
  • FIG. 13 is a display example 340 of KPIs of events in which the goals of all KPIs according to this embodiment are achieved. More specifically, it is a display example 340 when "all products" is selected as the display target product in step S17 and "target value attainment for all KPIs" is selected as the KPI condition.
  • the result output unit 115 selects a record in which all the KPIs 323 to 327 are "True” indicating that the target value has been achieved, among the records in which the product 321 is "all products” in the KPI threshold determination result table 320 (see FIG. 11). to explore.
  • the result output unit 115 acquires records in which the event 322 is “event 5” as a search result, and displays the KPI of “event 5”.
  • FIG. 14 is a display example 350 specifying KPIs according to this embodiment. More specifically, it is a display example 350 when "all products", “product A” and “product B” are designated for “KPI1” and “all products” is selected for "KPI3" in step S17. .
  • FIG. 15 is a display example 360 in which KPIs are specified with conditions according to this embodiment. Specifically, in step S17, "all products", “product A”, and “product B” are selected for “KPI1”, “all products” are selected for "KPI3”, and "KPI1 and KPI3 are This is a display example 360 when the condition "target value achieved" is selected.
  • the result output unit 115 refers to the KPI threshold determination result table 320 (see FIG. 11) to search for a record that satisfies the conditions, and displays the KPI of the event corresponding to the search result record.
  • step S19 the result output unit 115 acquires the user operation, and if the operation is to end (step S20 ⁇ YES), the planning support process is terminated. NO) Return to step S17.
  • the plan support device 100 supports formulation of a plan for procurement of product parts. More specifically, the planning support device 100 predicts the delivery date of parts when an event is executed to change order information including an order destination, delivery date, order quantity, etc. of parts set by the user. Next, the planning support apparatus 100 calculates and displays KPIs, which are evaluation indices for planning, such as the coincidence rate between the delivery date and the arrival date of ordered parts and the total manufacturing work time. The planning support apparatus 100 selects and displays events, KPIs, and products designated by the user, or displays only events that satisfy the target values of KPIs set by the user.
  • KPIs evaluation indices for planning
  • a prediction model 122 (see Fig. 1), which is a machine learning model that uses past results as teacher data, is used to predict the arrival date. Even if there are no empirical rules or well-known theories that can be used to predict arrival dates, machine learning technology can be used to make predictions based on past performance.
  • the user of the planning support device 100 sets changes (differences, corrections) from the basic plan as events in addition to the basic plan that serves as a reference, and how the KPI changes when the event is executed (the plan is changed). It becomes possible to predict (simulate) whether A target value (threshold) can be set for the KPI, and it becomes possible to try how to change the plan in order to meet the target. Even if the goals of all KPIs cannot be achieved, it is possible to search for a plan that satisfies the goals by focusing on the important KPIs, and the user can formulate an optimized plan.
  • Event ⁇ In the above-described embodiment, events are set so that when the completion of manufacturing is postponed/advanced, the delivery of parts is also postponed/advanced by the same period (see “Event 1" and "Event 2" in FIG. 3). .
  • the event reflection unit 112 may refer to a PERT (Program Evaluation and Review Technique) diagram of product manufacturing and change the management information so that the postponement/advance period differs for each part.
  • PERT Program Evaluation and Review Technique
  • parts are subject to management, and related to planning of parts procurement, but other plans may be used.
  • you can set KPIs and target values for various management targets, and test how to change the plan (which plan to adopt) to achieve the target values. become.
  • machine learning technology can be used to make predictions based on past performance and calculate KPIs.
  • the planning support device 100 can be used for product purchase planning.
  • the management information (management information database 210) includes the amount of merchandise purchased by merchandise, by store, and by day, the weather in the area where the store is located, the supplier, the purchase price, the purchase amount, and the like.
  • the event is forecasted weather for each store and for each day.
  • the target of prediction is the sales volume for each product, each store, and each day. KPIs include gross profit, the amount of unsold items, and whether there are shortages.
  • the planning support device 100 can be used for campaign planning for members.
  • This member is a subscriber of a cellular phone service or a mail-order member.
  • the management information includes the subscription period for each member, the number of times the service is used, the amount of service used, the amount of money used, and the like.
  • An event is a campaign such as a charge discount or privilege.
  • the target of prediction is the number of times the service is used, the amount of service used, the amount of money used, etc. after the event is executed.
  • KPIs include a retention rate of members, a usage amount, a usage amount, and the like.
  • the planning support device 100 can be used for maintenance planning of manufacturing machines.
  • the management information includes introduction time for each manufacturing machine, cumulative operating time, operating time since the last maintenance inspection, failure rate, and the like.
  • the event is the next maintenance inspection time or the maintenance inspection interval.
  • the prediction target is the predicted date of failure occurrence.
  • KPIs include production volume by product, interval between failures, operating rate by manufacturing machine, average operating rate, and the like.
  • the planning support device 100 can be used for production planning of greenhouse-grown agricultural products.
  • the management information includes the planting time for each house, weather data, fertilizer data, growth data, and the like.
  • the event is the amount of adjustment of the environment in the house, the amount of fertilizer, and the like.
  • the target of prediction is harvest time, harvest amount, and the like.
  • KPIs include the amount of harvest, the amount of sales, and the amount of waste.
  • the program 121 is stored in the planning support device 100, which is a computer.
  • the program 121 in the recording medium 910 may be read, loaded into the storage unit 120 and executed, or may be installed from the recording medium 910 and executed.
  • FIG. 16 is a diagram showing a recording medium 910 according to this embodiment.
  • the arrival date is predicted using machine learning technology, but instead of machine learning, other methods such as a linear regression model may be used for prediction.
  • the management information database 210 and other databases are stored in the planning support apparatus 100, information stored in another apparatus may be accessed.
  • KPIs to be displayed, their conditions, and products are acquired. (see step S17).
  • the KPIs to be displayed and the KPIs necessary for confirming the conditions may be calculated after acquiring the KPIs to be displayed, their conditions, and the products.
  • the result output unit 115 identifies an event that has achieved the target value of the KPI and displays the KPI corresponding to the event. Events may be arranged and displayed in order. For example, the result output unit 115 may display the KPIs so that the events are arranged in descending order of achievement. In addition, the result output unit 115 may display KPIs such that events are arranged in descending order of the degree of achievement of preset KPIs of high importance.
  • Prediction model (machine learning model) 210 management information database (management information) 210A Management information 211 Order number (management item) 212 Order request date (control item) 213 Products (control items) 214 parts (control items) 215 parts (control items) 216 Business Partners (Management Items) 217 Delivery date (control item) 218 Arrival date (forecast target item) 219 valid flag (control item) 220 Supplementary Information (Control Items) 230 Event information database 233 Event reflection content (rule) 240 KPI information database 242 KPI threshold setting (target value, threshold)

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Educational Administration (AREA)
  • Game Theory and Decision Science (AREA)
  • Development Economics (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

A planning support device (100) is provided with an event reflection unit (112) that changes management items constituting management information about one or more management targets in accordance with events to be changed under predetermined rules, a prediction unit (113) that predicts items to be predicted related to the management targets on the basis of the management information changed in accordance with the event, and a KPI calculation unit (114) that calculates KPIs on the basis of the management information and the items to be predicted. The prediction unit (113) performs prediction using a prediction model (122) trained by training data constituted by past management information and past records of the items to be predicted.

Description

計画支援装置、プログラムおよび計画支援方法Planning support device, program and planning support method
　本発明は、計画の立案を支援する計画支援装置、プログラムおよび計画支援方法に関する。 The present invention relates to a plan support device, a program, and a plan support method that support planning.
　事業を進めるにあたり経営者／管理者は、商品の開発、原材料・部品の調達、製造、販売、保守などを通じて、各種工程や原価、販売価格・数量などを計画する。また、計画を立案するときには、売上・利益・市場シェアなどの重要な指標（ＫＰＩ：Key Performance Indicator）を設定し、ＫＰＩが目標を達成するように検討を重ねる。 　In advancing the business, managers/managers plan various processes, costs, sales prices, quantities, etc. through product development, procurement of raw materials and parts, manufacturing, sales, maintenance, etc. Also, when formulating a plan, set important indicators (KPI: Key Performance Indicator) such as sales, profit, market share, etc., and review repeatedly so that the KPI will achieve the target.
　各種経営評価指標（ＫＰＩ）となる経営情報の表示方法として特許文献１には、ＣＰＵ、メモリ、記憶装置を備える計算機の前記メモリに格納されるプログラムを実行することによって実現される経営情報表示方法であって、 前記記憶装置に格納される製品別の損益データを前記メモリに読み込み、 利益額の大きい順に前記製品を並べ替えた各損益額を表示装置に表示し、 前記製品のうちの１つが指定されたとき、前記記憶装置に格納される指定された製品の地域別の損益データと利益率を前記メモリに読み込み、 利益額の大きい順に前記地域を並べ替えた各損益額を、前記利益率とともに前記表示装置に表示することを特徴とする経営情報表示方法が記載されている。 As a method of displaying management information that serves as various management performance indicators (KPI), Patent Document 1 discloses a management information display method realized by executing a program stored in the memory of a computer equipped with a CPU, a memory, and a storage device. wherein profit and loss data for each product stored in the storage device is read into the memory, each profit and loss amount obtained by sorting the products in descending order of profit amount is displayed on a display device, and one of the products is When specified, the profit and loss data and the profit rate for each region of the specified product stored in the storage device are read into the memory, and each profit and loss amount obtained by sorting the regions in descending order of profit amount is stored in the profit rate. Also, a method of displaying management information is described, which is characterized by displaying on the display device together with the management information.
特開２００６－１４６５２８号公報JP-A-2006-146528
　特許文献１には、上記した経営情報表示方法により、製品毎の売上高、損益、収益率、成長率を分析し、全体損益に現時点で貢献している製品、将来の収益獲得が有望な製品、将来の収益が低くなると予想される製品について特定し、販売戦略を的確に反映させる効果があると記載されている。この経営情報表示方法を含め、従来の計画管理システムでは、計画され管理対象となる項目を各種指標として表示する機能と、現在立案している計画の影響を予測して各種指標の変化を表示することができる。 Patent Document 1 analyzes the sales, profit and loss, profit rate, and growth rate for each product using the management information display method described above, and identifies products that currently contribute to the overall profit and loss, and products that are promising for future profit acquisition. , it is stated that it has the effect of identifying products whose future earnings are expected to be low and accurately reflecting the sales strategy. Including this management information display method, conventional plan management systems have functions to display planned items to be managed as various indicators, and to predict the impact of the current plan and display changes in various indicators. be able to.
　しかしながら、計画の影響の予測は近似曲線など経験則や公知の理論で定式化した計算方法を利用するものである。このため予測を定式化できない場合には適用することができないという問題がある。このような問題は、商品の開発から販売、保守といった事業／経営全体に限られるものではなく、調達、製造、販売などを含め様々な工程に係る計画の立案や変更、実行する際に発生する。 However, the prediction of the impact of the plan uses empirical rules such as approximate curves and calculation methods formulated by known theories. For this reason, there is a problem that it cannot be applied when the prediction cannot be formulated. Such problems are not limited to the entire business/management from product development to sales and maintenance, but occur when planning, changing, and executing plans related to various processes including procurement, manufacturing, and sales. .
　本発明は、このような背景を鑑みてなされたものであり、定式化が困難な予測を含めて計画策定を支援する計画支援装置、プログラムおよび計画支援方法を提供することを課題とする。 The present invention has been made in view of such a background, and aims to provide a plan support device, a program, and a plan support method that support planning including predictions that are difficult to formulate.
　上記した課題を解決するため、本発明に係る計画支援装置は、１つ以上の管理対象に係る管理情報を構成する管理項目を所定のルールに従って変更するイベントに応じて変更するイベント反映部と、前記イベントにより変更された前記管理情報に基づいて、前記管理対象に係る予測対象項目を予測する予測部と、前記管理情報および前記予測対象項目に基づいてＫＰＩを算出するＫＰＩ算出部と、を備え、前記予測部は、過去の前記管理情報および前記予測対象項目の過去の実績を教師データとして訓練された機械学習モデルを用いて予測する。 In order to solve the above-described problems, the planning support apparatus according to the present invention includes an event reflecting unit that changes management items constituting management information related to one or more management targets according to an event that changes according to a predetermined rule; a prediction unit that predicts a prediction target item related to the management target based on the management information changed by the event; and a KPI calculation unit that calculates a KPI based on the management information and the prediction target item. , the prediction unit predicts the past management information and the past results of the prediction target item using a trained machine learning model as teacher data.
　本発明によれば、定式化が困難な予測を含めて計画策定を支援する計画支援装置、プログラムおよび計画支援方法を提供することができる。上記した以外の課題、構成および効果は、以下の実施形態の説明により明らかにされる。 According to the present invention, it is possible to provide a planning support device, a program, and a planning support method that support planning including predictions that are difficult to formulate. Problems, configurations, and effects other than those described above will be clarified by the following description of the embodiments.
本実施形態に係る計画支援装置の機能ブロック図である。It is a functional block diagram of a plan support device concerning this embodiment. 本実施形態に係る管理情報データベースのデータ構成図である。4 is a data configuration diagram of a management information database according to the embodiment; FIG. 本実施形態に係るイベント情報データベースのデータ構成図である。4 is a data configuration diagram of an event information database according to the embodiment; FIG. 本実施形態に係るＫＰＩ情報データベースのデータ構成図である。4 is a data configuration diagram of a KPI information database according to the embodiment; FIG. 本実施形態に係る取引先データベースのデータ構成図である。4 is a data configuration diagram of a supplier database according to the embodiment; FIG. 本実施形態に係る部品データベースのデータ構成図である。3 is a data configuration diagram of a parts database according to the embodiment; FIG. 本実施形態に係る製造管理データベースのデータ構成図である。4 is a data configuration diagram of a manufacturing management database according to the embodiment; FIG. 本実施形態に係る計画支援処理のフローチャートである。It is a flow chart of plan support processing concerning this embodiment. 本実施形態に係る管理情報のデータ構成図である。4 is a data configuration diagram of management information according to the embodiment; FIG. 本実施形態に係るＫＰＩ算出結果テーブルのデータ構成図である。4 is a data configuration diagram of a KPI calculation result table according to the embodiment; FIG. 本実施形態に係るＫＰＩ閾値判定結果テーブルのデータ構成図である。FIG. 4 is a data configuration diagram of a KPI threshold determination result table according to the embodiment; 本実施形態に係る全てのイベントにおける「全製品」のＫＰＩの表示例である。It is a display example of KPI of "all products" in all events according to the present embodiment. 本実施形態に係る全てのＫＰＩについて目標を達成したイベントのＫＰＩの表示例である。It is a display example of KPIs of events in which the targets of all KPIs according to the present embodiment are achieved. 本実施形態に係るＫＰＩを指定した表示例である。It is an example of a display which specified KPI which concerns on this embodiment. 本実施形態に係る条件付きでＫＰＩを指定した表示例である。It is a display example of specifying a KPI with a condition according to the present embodiment. 本実施形態に係る記録媒体を示す図である。1 is a diagram showing a recording medium according to this embodiment; FIG.
≪計画支援装置の概要≫
　以下に、本発明を実施するための形態（実施形態）における計画支援装置について説明する。本実施形態の計画支援装置は、例えば鉄道車両のような大型機械である製品を製造するメーカにおける部品調達の計画に係り、部品の発注先や納期を変更した場合のＫＰＩを算出して表示する。 ≪Overview of Planning Support Device≫
Below, the plan support apparatus in the form (embodiment) for implementing this invention is demonstrated. The planning support device of the present embodiment relates to planning of parts procurement in a manufacturer that manufactures a product that is a large machine such as a railway vehicle, and calculates and displays KPIs when parts suppliers and delivery dates are changed. .
　詳しくは、計画支援装置の利用者は、製品の製造期限の変更にともなう、部品の発注先や納期、発注量などを含む発注情報（管理情報）を変更するイベントを設定する。計画支援装置は、イベント（計画変更）を実行した場合の部品の着荷日を予測し、予測結果に基づいてＫＰＩを算出する。ＫＰＩとしては、発注部品の納期と着荷日の一致率や、発注部品の着荷遅延による製造現場のリスケジュール発生件数、製造総作業時間、製造遅延日数、製造コストなどがある。 Specifically, the user of the planning support device sets an event to change the ordering information (management information) including the part supplier, delivery date, order quantity, etc. due to the change in the manufacturing deadline of the product. The planning support device predicts the delivery date of parts when an event (plan change) is executed, and calculates KPI based on the prediction result. KPIs include the rate of coincidence between the delivery date and arrival date of ordered parts, the number of occurrences of rescheduling at the manufacturing site due to delays in the arrival of ordered parts, the total manufacturing work time, the number of manufacturing delay days, manufacturing costs, and the like.
　予測の手法として、機械学習がある。詳しくは、計画支援装置は、入力が過去の発注情報であり、出力（正解ラベル）が着荷日である学習データ（教師データ）で訓練された機械学習モデルを用いて、発注情報から着荷日を予測する。
　利用者は、製品の製造期限の変更にともない部品の発注先や納期を変更するという調達計画（案）の変更について、どのような変更を行えばＫＰＩがどう変化するかを予想（シミュレーション）できるようになる。重要なＫＰＩを満たすようなイベント（計画／計画案の変更）を見つけ出すことで、利用者は最適化された計画を立案することができるようになる。また、計画実行中に計画を変更する場合であっても、利用者は最適な変更を行うことができるようになる。 As a prediction method, there is machine learning. Specifically, the planning support device uses a machine learning model trained with learning data (teacher data) whose input is past order information and whose output (correct label) is the arrival date. Predict.
The user can predict (simulate) how the KPI will change if the procurement plan (draft) is changed by changing the ordering party and delivery date of parts due to the change in the manufacturing deadline of the product. become. By finding events (changes in plans/plan proposals) that satisfy important KPIs, users will be able to formulate optimized plans. Moreover, even when the plan is changed during execution of the plan, the user can make the optimum change.
≪計画支援装置の構成≫
　図１は、本実施形態に係る計画支援装置１００の機能ブロック図である。計画支援装置１００はコンピュータであり、制御部１１０、記憶部１２０、および入出力部１８０を含んで構成される。入出力部１８０には、ディスプレイやキーボード、マウスなどのユーザインターフェイス機器が接続される。 ≪Configuration of planning support device≫
FIG. 1 is a functional block diagram of a planning support device 100 according to this embodiment. The planning support device 100 is a computer and includes a control section 110 , a storage section 120 and an input/output section 180 . User interface devices such as a display, a keyboard, and a mouse are connected to the input/output unit 180 .
　記憶部１２０は、ＲＯＭ（Read Only Memory）、ＲＡＭ（Random Access Memory）およびＳＳＤ（Solid State Drive）などの記憶機器から構成される。記憶部１２０には、管理情報データベース２１０や、イベント情報データベース２３０、ＫＰＩ情報データベース２４０、取引先データベース２５０、部品データベース２６０、製造管理データベース２７０、プログラム１２１、予測モデル１２２が記憶される。なお、図１では管理情報データベース２１０、イベント情報データベース２３０、ＫＰＩ情報データベース２４０、取引先データベース２５０、部品データベース２６０、および製造管理データベース２７０を、それぞれ管理情報ＤＢ（database）、イベント情報ＤＢ、ＫＰＩ情報ＤＢ、取引先ＤＢ、部品ＤＢ、および製造管理ＤＢと記載している。 The storage unit 120 is composed of storage devices such as ROM (Read Only Memory), RAM (Random Access Memory) and SSD (Solid State Drive). The storage unit 120 stores a management information database 210, an event information database 230, a KPI information database 240, a customer database 250, a parts database 260, a manufacturing management database 270, a program 121, and a prediction model 122. In FIG. 1, the management information database 210, the event information database 230, the KPI information database 240, the supplier database 250, the parts database 260, and the manufacturing management database 270 are respectively represented by a management information DB (database), an event information DB, and a KPI information database. DB, client DB, parts DB, and manufacturing control DB.
　プログラム１２１は、計画支援処理（後記する図８参照）の手順の記述を含む。予測モデル１２２は、管理情報データベース２１０（後記する図２参照）に含まれる着荷日２１８を予測する際に用いられる機械学習モデルである。 The program 121 includes a description of the procedure for planning support processing (see FIG. 8, which will be described later). The prediction model 122 is a machine learning model used to predict the arrival date 218 contained in the management information database 210 (see FIG. 2 described later).
≪計画支援装置の構成：管理情報データベース≫
　図２は、本実施形態に係る管理情報データベース２１０のデータ構成図である。本実施形態における管理情報データベース２１０には、部品の発注（注文）情報が格納される。管理情報データベース２１０は、例えば表形式のデータであって、１つの行（レコード）は部品の発注情報を示し、注文番号２１１、注文依頼日２１２、製品２１３、部位２１４、部品２１５、取引先２１６、納期２１７、着荷日２１８、有効フラグ２１９（図２では有効Ｆと記載）、および補足情報２２０の列（属性）を含む。他に、発注数量や価格などの属性を含んでもよい。 <<Configuration of Planning Support Device: Management Information Database>>
FIG. 2 is a data configuration diagram of the management information database 210 according to this embodiment. The management information database 210 in this embodiment stores order information for parts. The management information database 210 is, for example, tabular data, one row (record) indicates ordering information for parts, and includes an order number 211, an order request date 212, a product 213, a part 214, a part 215, and a customer 216. , due date 217, arrival date 218, valid flag 219 (shown as valid F in FIG. 2), and supplementary information 220 columns (attributes). In addition, attributes such as order quantity and price may be included.
　注文番号２１１は、発注の識別情報である。注文依頼日２１２は、発注した日付である。製品２１３は、発注した部品が使われる製品の名称ないしは識別情報である。部位２１４は、発注した部品が使われる製品２１３における部位である。部品２１５は、発注した部品の名称ないしは識別情報である。取引先２１６は、発注先の名称ないしは識別情報である。納期２１７は、部品の納期である。 The order number 211 is the identification information of the order. The order placement date 212 is the date the order was placed. The product 213 is the name or identification information of the product in which the ordered parts are used. A part 214 is a part of the product 213 in which the ordered part is used. The part 215 is the name or identification information of the ordered part. Supplier 216 is the name or identification information of the supplier. The delivery date 217 is the delivery date of the part.
　着荷日２１８は、発注した部品が納品された日付である。図２では、何れの部品も納品されておらず、着荷日２１８は空白である。有効フラグ２１９は、発注が有効か無効かを示す。発注時点では、有効フラグ２１９は有効であるが、例えば発注が取り消された場合や、部品が製造終了である場合には、無効となる。補足情報２２０は、発注ないしは部品に係る付加的な情報である。 The arrival date 218 is the date when the ordered parts were delivered. In FIG. 2, none of the parts have been delivered and the arrival date 218 is blank. A valid flag 219 indicates whether the order is valid or invalid. The validity flag 219 is valid at the time of ordering, but becomes invalid when the order is canceled or when the production of the part is finished. Supplementary information 220 is additional information related to the order or parts.
≪計画支援装置の構成：イベント情報データベース≫
　図３は、本実施形態に係るイベント情報データベース２３０のデータ構成図である。イベント情報データベース２３０は、製品製造計画の変更など部品調達計画の変更を示すイベントの情報を含む。イベント情報データベース２３０は、例えば表形式のデータであって、１つの行（レコード）はイベントを示し、識別情報２３１（図３では＃と記載）、イベント内容２３２、およびイベント反映内容２３３の列（属性）を含む。 ≪Configuration of planning support device: event information database≫
FIG. 3 is a data configuration diagram of the event information database 230 according to this embodiment. The event information database 230 includes information on events indicating changes in parts procurement plans, such as changes in product manufacturing plans. The event information database 230 is data in, for example, a table format, one row (record) indicates an event, and columns of identification information 231 (shown as # in FIG. 3), event content 232, and event reflection content 233 ( attributes).
　識別情報２３１は、イベントの識別情報である。
　イベント内容２３２は、計画変更の内容を示す。変更例としては、製品の製造完了期限の延期や前倒し、製造自体の中止などがある。他に、部品の発注先の変更や、部品納期の確認などがある。
　イベント反映内容２３３は、イベント実行時の管理情報データベース２１０（図２参照）に含まれる属性（管理項目）の変更内容（変更手続きを示すルール）を示す。変更例（ルールの例）としては、製品の製造完了期限に合わせての納期２１７の延期や前倒し、発注先の変更にともなう注文依頼日２１２や取引先２１６の変更、部品納期の確認にともなう補足情報２２０の「１週間前に確認済み」への更新などがある。 The identification information 231 is identification information of the event.
The event content 232 indicates the content of the plan change. Examples of changes include postponing or bringing forward the deadline for completion of manufacturing of products, and discontinuing manufacturing itself. In addition, there are changes in parts suppliers and confirmation of part delivery dates.
The event reflection content 233 indicates the change content (rule indicating change procedure) of the attribute (management item) included in the management information database 210 (see FIG. 2) when the event is executed. Examples of changes (examples of rules) include postponing or bringing forward the delivery date 217 in line with the product manufacturing completion deadline, changing the order request date 212 and supplier 216 due to a change in the supplier, and supplementing the confirmation of the parts delivery date. For example, the information 220 is updated to "confirmed one week ago".
≪計画支援装置の構成：ＫＰＩ情報データベース≫
　図４は、本実施形態に係るＫＰＩ情報データベース２４０のデータ構成図である。ＫＰＩ情報データベース２４０は、計画を評価する際の指標であるＫＰＩの情報を含む。ＫＰＩ情報データベース２４０は、例えば表形式のデータであって、１つの行（レコード）はＫＰＩを示し、識別情報２４１（図４では＃と記載）、ＫＰＩ内容２４２、およびＫＰＩ閾値設定２４３の列（属性）を含む。 ≪Configuration of planning support device: KPI information database≫
FIG. 4 is a data configuration diagram of the KPI information database 240 according to this embodiment. The KPI information database 240 includes KPI information that is an index for evaluating a plan. The KPI information database 240 is, for example, tabular data, one row (record) indicates a KPI, and columns of identification information 241 (described as # in FIG. 4), KPI content 242, and KPI threshold setting 243 ( attributes).
　識別情報２４１は、ＫＰＩの識別情報である。
　ＫＰＩ内容２４２は、ＫＰＩの内容（意味）を示す。内容の例として、発注部品の納期と着荷日の一致率、製造総作業時間、製造コストなどがある。
　ＫＰＩ閾値設定２４３は、ＫＰＩの目標値を示す。目標値は、数値そのものとは限らず、計画変更前と比較しての増減値や増減比であってもよい。 The identification information 241 is identification information of the KPI.
The KPI content 242 indicates the content (meaning) of the KPI. Examples of the contents include the rate of coincidence between the delivery date and the arrival date of the ordered parts, the total manufacturing work time, and the manufacturing cost.
The KPI threshold setting 243 indicates the target value of the KPI. The target value is not limited to the numerical value itself, and may be an increase/decrease value or an increase/decrease ratio compared to before the plan change.
≪計画支援装置の構成：取引先データベース≫
　図５は、本実施形態に係る取引先データベース２５０のデータ構成図である。取引先データベース２５０は、部品の取引先に係る情報を含むデータベースである。取引先データベース２５０は、例えば表形式のデータであって、１つの行は取引先を示し、取引先２５１、現在の発注数２５２、現在の発注金額２５３、過去６か月の累積発注数２５４、過去６か月の累積発注金額２５５、および過去６か月の信頼性２５６の列（属性）を含む。 ≪Configuration of planning support device: customer database≫
FIG. 5 is a data configuration diagram of the customer database 250 according to this embodiment. The supplier database 250 is a database containing information on suppliers of parts. The supplier database 250 is, for example, data in tabular form, one row indicates a supplier, a supplier 251, the current number of orders placed 252, the current order amount 253, the cumulative number of orders placed in the past six months 254, Includes columns (attributes) for Cumulative Order Amount 255 for the past 6 months and Reliability 256 for the past 6 months.
　取引先２５１は、部品の発注先の名称ないしは識別情報であって、取引先２１６（図２参照）と対応する。現在の発注数２５２、および現在の発注金額２５３は、それぞれ現時点における取引先２５１への発注件数、および発注金額である。過去６か月の累積発注数２５４、および過去６か月の累積発注金額２５５は、それぞれ過去６か月間における発注件数、および発注金額である。過去６か月の信頼性２５６は、過去６か月間における部品の納期と着荷日との一致率や、過去６か月間に納品された部品の品質などから算出された指標である。 The supplier 251 is the name or identification information of the parts supplier, and corresponds to the supplier 216 (see FIG. 2). The current number of orders placed 252 and the current order amount 253 are the number of orders placed to the supplier 251 at the present time and the order amount, respectively. The cumulative number of orders placed in the past 6 months 254 and the cumulative order amount 255 in the past 6 months are the number of orders placed and the amount of orders placed in the past 6 months, respectively. The reliability of the past 6 months 256 is an index calculated from the rate of coincidence between delivery dates and arrival dates of parts in the past 6 months, the quality of parts delivered in the past 6 months, and the like.
≪計画支援装置の構成：部品データベース≫
　図６は、本実施形態に係る部品データベース２６０のデータ構成図である。部品データベース２６０は、製品の部品に係る情報を含むデータベースである。部品データベース２６０は、例えば表形式のデータであって、１つの行は部品を示し、部品２６１、図面２６２、設計難易度２６３、製造難易度２６４、想定設計日数２６５、および想定製造日数２６６の列（属性）を含む。 ≪Configuration of planning support device: parts database≫
FIG. 6 is a data configuration diagram of the parts database 260 according to this embodiment. The parts database 260 is a database containing information on product parts. The parts database 260 is, for example, data in tabular form, one row indicates a part, and columns of a part 261, a drawing 262, a design difficulty level 263, a manufacturing difficulty level 264, an estimated number of days for designing 265, and an estimated number of days to manufacture 266. (attributes).
　部品２６１は、部品の名称ないしは識別情報であって、部品２１５（図２参照）に対応する。図面２６２は、部品の図面データを示す。設計難易度２６３および製造難易度２６４は、それぞれ、部品を設計する難易度および部品を製造する難易度を示し、Ａ～Ｅまでのレベルで示される。想定設計日数２６５および想定製造日数２６６は、それぞれ部品設計に要する日数および部品製造に要する日数を示す。 The part 261 is the name or identification information of the part and corresponds to the part 215 (see FIG. 2). Drawing 262 shows the drawing data of the part. The design difficulty level 263 and the manufacturing difficulty level 264 indicate the difficulty level of designing the part and the difficulty level of manufacturing the part, respectively, and are indicated by levels from A to E. The estimated design days 265 and estimated manufacturing days 266 indicate the number of days required for part design and the number of days required for part manufacturing, respectively.
≪計画支援装置の構成：製造管理データベース≫
　図７は、本実施形態に係る製造管理データベース２７０のデータ構成図である。製造管理データベース２７０は、製品の製造に係る情報を含む。製造管理データベース２７０は例えば表形式のデータであって、１つの行は製造される１つの製品を示し、識別情報２７１（図７では＃と記載）、製品２７２、注文主２７３、製造開始日２７４の列（属性）を含む。 ≪Configuration of Planning Support Device: Manufacturing Management Database≫
FIG. 7 is a data configuration diagram of the manufacturing management database 270 according to this embodiment. The manufacturing management database 270 contains information related to product manufacturing. The manufacturing management database 270 is, for example, tabular data, one row indicates one product to be manufactured, and identification information 271 (shown as # in FIG. 7), product 272, orderer 273, manufacturing start date 274. contains the columns (attributes) of
　識別情報２７１は、製造される個々の製品の識別情報である。製品２７２は、製造される製品を示し、製品２１３（図２参照）に対応する。注文主２７３は、製品２７２の発注者を示し、製造した製品２７２の納入先である。製造開始日２７４は、製造開始日の計画値と実績を示す。製造管理データベース２７０は、さらに、製造完了日、出荷日、製造に含まれる各工程の期間や人員、原価などの計画値と実績とを含んでもよい。 The identification information 271 is identification information for each manufactured product. Product 272 indicates the product to be manufactured and corresponds to product 213 (see FIG. 2). The orderer 273 indicates the person who ordered the product 272 and is the delivery destination of the manufactured product 272 . The manufacturing start date 274 indicates planned values and actual results for the manufacturing start date. The manufacturing management database 270 may further include planned values and actual results such as manufacturing completion dates, shipping dates, periods of each process involved in manufacturing, personnel, and costs.
≪計画支援装置の構成：制御部≫
　図１に戻って、制御部１１０は、ＣＰＵ（Central Processing Unit）を含んで構成され、学習部１１１、イベント反映部１１２、予測部１１３、ＫＰＩ算出部１１４、および結果出力部１１５を備える。 ≪Configuration of planning support device: control unit≫
Returning to FIG. 1 , the control unit 110 includes a CPU (Central Processing Unit) and includes a learning unit 111 , an event reflection unit 112 , a prediction unit 113 , a KPI calculation unit 114 and a result output unit 115 .
　学習部１１１は、学習データ（教師データ）を準備して、機械学習モデルである予測モデル１２２を訓練して、生成する。学習データは、納入済み部品の発注情報である管理情報データベース２１０（図２参照）である。詳しくは、管理情報データベース２１０のなかで着荷日２１８を除くデータが入力データとなり、着荷日２１８が正解ラベル（出力データ）となる。入力データとして取引先データベース２５０（図５参照）や部品データベース２６０（図６参照）を含めてもよい。このような学習データと正解ラベルとを用いて訓練された予測モデル１２２は、部品２１５や注文依頼日２１２、取引先２１６、納期２１７などの情報から着荷日２１８を予測する際に用いられる機械学習モデルとなる。 The learning unit 111 prepares learning data (teacher data) and trains and generates a prediction model 122, which is a machine learning model. The learning data is the management information database 210 (see FIG. 2), which is order information for parts that have already been delivered. Specifically, data other than the arrival date 218 in the management information database 210 becomes the input data, and the arrival date 218 becomes the correct label (output data). The input data may include the supplier database 250 (see FIG. 5) and the parts database 260 (see FIG. 6). The prediction model 122 trained using such learning data and correct labels is machine learning used when predicting the arrival date 218 from information such as the parts 215, the order request date 212, the supplier 216, and the delivery date 217. Become a model.
　イベント反映部１１２は、選択されたイベント（図３記載のイベント情報データベース２３０参照）に応じて管理情報データベース２１０を変更する。詳しくは、後記するように、管理情報データベース２１０と同等のデータである管理情報２１０Ａ（後記する図９参照）を変更する（後記する図８のステップＳ１２～Ｓ１３参照）。例えば、識別情報２３１が「イベント１」であるイベントが選択された場合には、イベント反映部１１２は、管理情報データベース２１０のなかで、製品２１３が「製品Ａ」であるレコードの納期２１７を３か月延期する。 The event reflection unit 112 changes the management information database 210 according to the selected event (see the event information database 230 shown in FIG. 3). Specifically, as will be described later, the management information 210A (see FIG. 9 described later), which is equivalent data to the management information database 210, is changed (see steps S12 and S13 in FIG. 8 described later). For example, when an event whose identification information 231 is “event 1” is selected, the event reflection unit 112 changes the delivery date 217 of the record whose product 213 is “product A” in the management information database 210 to 3 postpone for a month.
　予測部１１３は、予測モデル１２２を用いて着荷日２１８を予測する。詳しくは、管理情報データベース２１０（後記する図９記載の管理情報２１０Ａ参照）を入力データとして予測モデル１２２を用いて着荷日２１８を予測する。入力データとして取引先データベース２５０（図５参照）や部品データベース２６０（図６参照）を含めてもよい。 The prediction unit 113 predicts the arrival date 218 using the prediction model 122 . Specifically, the arrival date 218 is predicted using the prediction model 122 with the management information database 210 (see management information 210A shown in FIG. 9 to be described later) as input data. The input data may include the supplier database 250 (see FIG. 5) and the parts database 260 (see FIG. 6).
　ＫＰＩ算出部１１４は、ＫＰＩ情報データベース２４０（図４参照）にあるＫＰＩについて、ＫＰＩ内容２４２に従い個別製品および全製品に係るＫＰＩを算出する。例えば、識別情報２４１が「ＫＰＩ１」であるＫＰＩについては、ＫＰＩ算出部１１４は、管理情報データベース２１０の納期２１７と着荷日２１８との一致率を、個別製品および全製品について算出する。また、例えば、識別情報２４１が「ＫＰＩ３」～「ＫＰＩ５」のＫＰＩについては、ＫＰＩ算出部１１４は、製造管理データベース２７０（図７参照）を参照して算出する。ＫＰＩ算出部１１４は、他に取引先データベース２５０や部品データベース２６０を参照してもよい。 The KPI calculation unit 114 calculates KPIs for individual products and all products according to the KPI contents 242 for the KPIs in the KPI information database 240 (see FIG. 4). For example, for the KPI whose identification information 241 is "KPI1", the KPI calculation unit 114 calculates the matching rate between the delivery date 217 and the arrival date 218 in the management information database 210 for individual products and all products. Further, for example, the KPI calculation unit 114 calculates the KPIs whose identification information 241 is “KPI3” to “KPI5” by referring to the manufacturing management database 270 (see FIG. 7). The KPI calculation unit 114 may also refer to the supplier database 250 and the parts database 260 .
　結果出力部１１５は、ＫＰＩ算出部１１４が算出したＫＰＩを入出力部１８０（図１参照）に接続されたディスプレイに出力（表示）する。結果出力部１１５は、全てのＫＰＩを出力するとは限らず、利用者からの指定に基づいて特定されるＫＰＩを出力する場合もある。 The result output unit 115 outputs (displays) the KPI calculated by the KPI calculation unit 114 to a display connected to the input/output unit 180 (see FIG. 1). The result output unit 115 does not always output all KPIs, and may output KPIs specified based on user's designation.
≪計画支援処理≫
　図８は、本実施形態に係る計画支援処理のフローチャートである。
　ステップＳ１１において制御部１１０は、変更なしのイベント、およびイベント情報データベース２３０に含まれるイベントそれぞれについてステップＳ１２～Ｓ１６を繰り返す処理を開始する。変更なしのイベントとは、管理情報データベース２１０（図２参照）を変更しないというイベントである。なお、ステップＳ１２～Ｓ１６を繰り返し処理において、変更なしのイベントが最初に処理される。 ≪Planning support processing≫
FIG. 8 is a flow chart of planning support processing according to the present embodiment.
In step S 11 , control unit 110 starts processing to repeat steps S 12 to S 16 for each of the unchanged event and the event included in event information database 230 . A no-change event is an event in which the management information database 210 (see FIG. 2) is not changed. Note that in the repeated processing of steps S12 to S16, events without change are processed first.
　ステップＳ１２においてイベント反映部１１２は、管理情報データベース２１０の内容を記憶部１２０のなかで管理情報データベース２１０とは異なる領域に読み込む。以下、この読み込んだデータを管理情報２１０Ａ（後記する図９参照）と記す。
　ステップＳ１３においてイベント反映部１１２は、イベントに応じて管理情報２１０Ａを変更する。 In step S<b>12 , the event reflecting unit 112 reads the content of the management information database 210 into a different area from the management information database 210 in the storage unit 120 . Hereinafter, this read data is referred to as management information 210A (see FIG. 9 described later).
In step S13, the event reflecting unit 112 changes the management information 210A according to the event.
　ステップＳ１４において予測部１１３は、管理情報２１０Ａを入力データとし、予測モデル１２２を用いて着荷日を予測して、管理情報２１０Ａの着荷日に格納する。
　図９は、本実施形態に係る管理情報２１０Ａのデータ構成図である。管理情報２１０Ａは、管理情報データベース２１０（図２参照）と同等の構成であって、列（属性）には管理情報データベース２１０と同じ符号を付与している。管理情報データベース２１０とは異なり、着荷日２１８には、予測部１１３が予測した着荷日が格納されている。 In step S14, the prediction unit 113 uses the management information 210A as input data, predicts the arrival date using the prediction model 122, and stores the arrival date in the management information 210A.
FIG. 9 is a data configuration diagram of the management information 210A according to this embodiment. The management information 210A has the same configuration as the management information database 210 (see FIG. 2), and the columns (attributes) are assigned the same reference numerals as those of the management information database 210. FIG. Unlike the management information database 210, the date of arrival 218 stores the date of arrival predicted by the prediction unit 113. FIG.
　図８に戻って、ステップＳ１５においてＫＰＩ算出部１１４は、管理情報２１０Ａに基づいてＫＰＩ情報データベース２４０（図４参照）に含まれる各ＫＰＩを算出する。また、ＫＰＩ算出部１１４は変更なしのイベントにおけるＫＰＩの算出値である基準値（基準ＫＰＩとも記す）と比較して、増減値または比率を算出する。算出したＫＰＩおよび増減値／比率は、後記するＫＰＩ算出結果テーブル３１０（後記する図１０参照）に格納される。 Returning to FIG. 8, in step S15, the KPI calculator 114 calculates each KPI contained in the KPI information database 240 (see FIG. 4) based on the management information 210A. Also, the KPI calculation unit 114 calculates an increase/decrease value or a ratio by comparing with a reference value (also referred to as reference KPI) which is a calculated value of KPI in an event without change. The calculated KPI and increase/decrease value/ratio are stored in a KPI calculation result table 310 (see FIG. 10 described later).
　図１０は、本実施形態に係るＫＰＩ算出結果テーブル３１０のデータ構成図である。ＫＰＩ算出結果テーブル３１０には、ＫＰＩ算出部１１４が算出したＫＰＩが格納される。ＫＰＩ算出結果テーブル３１０は、例えば表形式のデータであって、１つの行（レコード）はイベント反映後のＫＰＩを示し、製品３１１、イベント３１２、ＫＰＩ３１３～３１７の列（属性）を含む。 FIG. 10 is a data configuration diagram of the KPI calculation result table 310 according to this embodiment. The KPI calculation result table 310 stores KPIs calculated by the KPI calculator 114 . The KPI calculation result table 310 is data in tabular form, for example, one row (record) indicates the KPI after the event is reflected, and includes columns (attributes) of a product 311, an event 312, and KPIs 313-317.
　製品３１１は、算出対象の製品を示し、個々の製品の他に全ての製品である「全製品」がある。
　イベント３１２は、管理情報データベース２１０の変更のもととなったイベントを示す。イベント３１２が「－」であるレコードは、変更なしイベントにおけるＫＰＩであり、基準値を示す。他のレコードは、イベント３１２に示されるイベントに基づいて変更された管理情報２１０Ａ（図９参照）に基づいたＫＰＩを示し、基準値との増減値／比率を含む。
　ＫＰＩ３１３～３１７は算出されたＫＰＩを示す。 A product 311 indicates a product to be calculated, and includes “all products” which are all products in addition to individual products.
An event 312 indicates an event that caused the management information database 210 to be changed. A record whose event 312 is "-" is a KPI in a no-change event and indicates a reference value. Other records show KPIs based on the management information 210A (see FIG. 9) changed based on the event indicated by the event 312, and include increase/decrease values/ratios from the reference value.
KPIs 313-317 indicate calculated KPIs.
　製品３１１が「全製品」でイベント３１２が「イベント１」であるレコードのＫＰＩ３１３の「ＫＰＩ１」は、発注部品の納期と着荷日の一致率（図４参照）が、７７％であり、基準値から７増加したことを示している。また、ＫＰＩ３１５の「ＫＰＩ３」は、製造総作業時間が、１１００時間であり、基準値の１１０％であることを示している。 "KPI1" of the KPI 313 of the record in which the product 311 is "all products" and the event 312 is "event 1" has a matching rate of 77% for the due date and the arrival date of ordered parts (see FIG. 4), which is the reference value. It shows that it increased by 7 from "KPI3" of KPI 315 indicates that the total manufacturing work time is 1100 hours, which is 110% of the reference value.
　図８に戻って、ステップＳ１６においてＫＰＩ算出部１１４は、算出したＫＰＩが目標値（図４記載のＫＰＩ閾値設定２４３参照）を満たしているか否かを判定して、判定結果を後記するＫＰＩ閾値判定結果テーブル３２０（後記する図１１参照）に格納する。
　図１１は、本実施形態に係るＫＰＩ閾値判定結果テーブル３２０のデータ構成図である。ＫＰＩ閾値判定結果テーブル３２０には、ＫＰＩ算出部１１４が判定したＫＰＩが目標値達成の当否が格納される。ＫＰＩ閾値判定結果テーブル３２０は、例えば表形式のデータであって、１つの行（レコード）はイベント反映後におけるＫＰＩ目標値達成の当否を示し、製品３２１、イベント３２２、ＫＰＩ３２３～３２７の列（属性）を含む。製品３２１、イベント３２２、ＫＰＩ３２３～３２７は、ＫＰＩ算出結果テーブル３１０（図１０参照）の製品３１１、イベント３１２、ＫＰＩ３１３～３１７とそれぞれ同様である。但し、ＫＰＩ３２３～３２７には、ＫＰＩが目標値達成の当否（True／False）が格納される。 Returning to FIG. 8, in step S16, the KPI calculation unit 114 determines whether or not the calculated KPI satisfies the target value (see KPI threshold setting 243 shown in FIG. 4). Stored in the determination result table 320 (see FIG. 11 described later).
FIG. 11 is a data configuration diagram of the KPI threshold determination result table 320 according to this embodiment. The KPI threshold determination result table 320 stores whether or not the KPI determined by the KPI calculation unit 114 has achieved the target value. The KPI threshold determination result table 320 is, for example, tabular data. One row (record) indicates whether or not the KPI target value has been achieved after the event is reflected. )including. The product 321, event 322, and KPIs 323-327 are the same as the product 311, event 312, and KPIs 313-317 of the KPI calculation result table 310 (see FIG. 10). However, the KPIs 323 to 327 store whether or not the KPI has achieved the target value (True/False).
　図８に戻って、ステップＳ１７において結果出力部１１５は、入出力部１８０（図１参照）に接続されるディスプレイ（不図示）に表示するＫＰＩやその条件、製品を、利用者に問い合わせて取得する。
　ステップＳ１８において結果出力部１１５は、ステップＳ１７で取得したＫＰＩやその条件、製品に応じたＫＰＩをディスプレイに表示する。以下、表示例を説明する。 Returning to FIG. 8, in step S17, the result output unit 115 inquires of the user and acquires the KPIs, their conditions, and the products to be displayed on the display (not shown) connected to the input/output unit 180 (see FIG. 1). do.
In step S18, the result output unit 115 displays the KPI acquired in step S17, its conditions, and the KPI corresponding to the product on the display. A display example will be described below.
　図１２は、本実施形態に係る全てのイベントにおける「全製品」のＫＰＩの表示例３３０である。詳しくは、ステップＳ１７において表示対象製品として「全製品」が選択され、さらに全てのＫＰＩが選択された場合の表示例３３０である。表示例３３０では、縦方向はイベント別に、横方向はＫＰＩ別に、「全製品」のＫＰＩの値および基準値との増減値／比率が表示されている。表示例３３０を参照することで、利用者は計画変更（計画内容の違い）によるＫＰＩへの影響を把握することができるようになる。 FIG. 12 is a display example 330 of KPIs for "all products" in all events according to this embodiment. More specifically, it is a display example 330 when "all products" are selected as display target products in step S17 and all KPIs are selected. In the display example 330, the vertical direction is for each event, and the horizontal direction is for each KPI. By referring to the display example 330, the user can comprehend the impact of plan changes (differences in plan content) on KPIs.
　図１３は、本実施形態に係る全てのＫＰＩについて目標を達成したイベントのＫＰＩの表示例３４０である。詳しくは、ステップＳ１７において表示対象製品として「全製品」が選択され、ＫＰＩの条件として「全てのＫＰＩについて目標値達成」が選択された場合の表示例３４０である。結果出力部１１５は、ＫＰＩ閾値判定結果テーブル３２０（図１１参照）の製品３２１が「全製品」であるレコードのなかで全てのＫＰＩ３２３～３２７が目標値達成を示す「True」となっているレコードを探索する。結果出力部１１５は、探索結果としてイベント３２２が「イベント５」であるレコードを探索結果として取得し、「イベント５」のＫＰＩを表示する。 FIG. 13 is a display example 340 of KPIs of events in which the goals of all KPIs according to this embodiment are achieved. More specifically, it is a display example 340 when "all products" is selected as the display target product in step S17 and "target value attainment for all KPIs" is selected as the KPI condition. The result output unit 115 selects a record in which all the KPIs 323 to 327 are "True" indicating that the target value has been achieved, among the records in which the product 321 is "all products" in the KPI threshold determination result table 320 (see FIG. 11). to explore. The result output unit 115 acquires records in which the event 322 is “event 5” as a search result, and displays the KPI of “event 5”.
　図１４は、本実施形態に係るＫＰＩを指定した表示例３５０である。詳しくは、ステップＳ１７において「ＫＰＩ１」については「全製品」と「製品Ａ」と「製品Ｂ」とが指定され、「ＫＰＩ３」については「全製品」が選択された場合の表示例３５０である。 FIG. 14 is a display example 350 specifying KPIs according to this embodiment. More specifically, it is a display example 350 when "all products", "product A" and "product B" are designated for "KPI1" and "all products" is selected for "KPI3" in step S17. .
　図１５は、本実施形態に係る条件付きでＫＰＩを指定した表示例３６０である。詳しくは、ステップＳ１７において「ＫＰＩ１」については「全製品」と「製品Ａ」と「製品Ｂ」とが選択され、「ＫＰＩ３」については「全製品」が選択され、さらに「ＫＰＩ１とＫＰＩ３とが目標値達成」という条件が選択された場合の表示例３６０である。結果出力部１１５は、ＫＰＩ閾値判定結果テーブル３２０（図１１参照）を参照して条件を満たすレコードを探索して、探索結果のレコードに対応するイベントのＫＰＩを表示する。 FIG. 15 is a display example 360 in which KPIs are specified with conditions according to this embodiment. Specifically, in step S17, "all products", "product A", and "product B" are selected for "KPI1", "all products" are selected for "KPI3", and "KPI1 and KPI3 are This is a display example 360 when the condition "target value achieved" is selected. The result output unit 115 refers to the KPI threshold determination result table 320 (see FIG. 11) to search for a record that satisfies the conditions, and displays the KPI of the event corresponding to the search result record.
　図８に戻って、ステップＳ１９において結果出力部１１５は、ユーザ操作を取得し、終了の操作であれば（ステップＳ２０→ＹＥＳ）計画支援処理を終了し、終了の操作でなければ（ステップＳ２０→ＮＯ）ステップＳ１７に戻る。 Returning to FIG. 8, in step S19, the result output unit 115 acquires the user operation, and if the operation is to end (step S20→YES), the planning support process is terminated. NO) Return to step S17.
≪計画支援装置の特徴≫
　計画支援装置１００は、製品の部品調達に係る計画の策定を支援する。詳しくは、計画支援装置１００は、利用者が設定した部品の発注先や納期、発注量などを含む発注情報を変更するイベントを実行した場合における部品の着荷日を予測する。次に、計画支援装置１００は、発注部品の納期と着荷日の一致率や製造総作業時間など、計画の評価指標となるＫＰＩを算出して表示する。計画支援装置１００は、利用者が指定したイベントやＫＰＩ、製品を選択して表示したり、利用者が設定したＫＰＩの目標値を満たすイベントに限定して表示したりする。 ≪Features of planning support equipment≫
The plan support device 100 supports formulation of a plan for procurement of product parts. More specifically, the planning support device 100 predicts the delivery date of parts when an event is executed to change order information including an order destination, delivery date, order quantity, etc. of parts set by the user. Next, the planning support apparatus 100 calculates and displays KPIs, which are evaluation indices for planning, such as the coincidence rate between the delivery date and the arrival date of ordered parts and the total manufacturing work time. The planning support apparatus 100 selects and displays events, KPIs, and products designated by the user, or displays only events that satisfy the target values of KPIs set by the user.
　着荷日の予測には、過去の実績を教師データとする機械学習モデルである予測モデル１２２（図１参照）が用いられる。着荷日の予測に利用可能な経験則や公知の理論など定式化された手法がない場合であっても、機械学習技術を用いることで過去の実績に応じた予測ができるようになる。 A prediction model 122 (see Fig. 1), which is a machine learning model that uses past results as teacher data, is used to predict the arrival date. Even if there are no empirical rules or well-known theories that can be used to predict arrival dates, machine learning technology can be used to make predictions based on past performance.
　計画支援装置１００の利用者は、基準となる基本計画の他に、基本計画からの変更（違い、修正）をイベントとして設定し、イベントを実行（計画を変更）したときのＫＰＩがどう変化するか予想（シミュレーション）できるようになる。ＫＰＩには目標値（閾値）を設定することができ、目標を満たすために、どのように計画を変更するか試すことができるようになる。全てのＫＰＩの目標達成ができない場合でも、重要なＫＰＩに絞って目標を満たす計画を探すことも可能となり、利用者は最適化された計画を立案することができるようになる。 The user of the planning support device 100 sets changes (differences, corrections) from the basic plan as events in addition to the basic plan that serves as a reference, and how the KPI changes when the event is executed (the plan is changed). It becomes possible to predict (simulate) whether A target value (threshold) can be set for the KPI, and it becomes possible to try how to change the plan in order to meet the target. Even if the goals of all KPIs cannot be achieved, it is possible to search for a plan that satisfies the goals by focusing on the important KPIs, and the user can formulate an optimized plan.
≪変形例：イベント≫
　上記した実施形態では、製造完了を延期／前倒しする場合に部品の納期も同じ期間だけ延期／前倒しするようにイベントを設定している（図３記載の「イベント１」や「イベント２」参照）。これに替わり、イベント反映部１１２は、製品製造のＰＥＲＴ（Program Evaluation and Review Technique）図を参照して、部品ごとに延期／前倒しする期間が異なるように管理情報を変更するようにしてもよい。 ≪Modification: Event≫
In the above-described embodiment, events are set so that when the completion of manufacturing is postponed/advanced, the delivery of parts is also postponed/advanced by the same period (see "Event 1" and "Event 2" in FIG. 3). . Alternatively, the event reflection unit 112 may refer to a PERT (Program Evaluation and Review Technique) diagram of product manufacturing and change the management information so that the postponement/advance period differs for each part.
≪変形例：計画対象≫
　上記した実施形態は、部品を管理対象とし、部品調達の計画に係るものであったが、他の計画であってもよい。以下の例に示すように、様々な管理対象について、ＫＰＩと目標値を設定して、目標値を達成するためにどのように計画を変更するか（どの計画を採用するか）試すことができるようになる。また、経験則や公知の理論など定式化された手法がない場合であっても、機械学習技術を用いることで過去の実績に応じた予測を行ってＫＰＩが算出できるようになる。 ≪Modification: Planning target≫
In the above-described embodiment, parts are subject to management, and related to planning of parts procurement, but other plans may be used. As shown in the example below, you can set KPIs and target values for various management targets, and test how to change the plan (which plan to adopt) to achieve the target values. become. Even if there is no formalized method such as empirical rules or known theories, machine learning technology can be used to make predictions based on past performance and calculate KPIs.
　計画支援装置１００は、商品仕入計画に用いることができる。管理情報（管理情報データベース２１０）は、商品別、店舗別、日別の商品の仕入れ量や店舗のある地域の天候、納入者、仕入れ値、仕入れ量などである。イベントは、店舗別、日別の予測天候である。予測対象は、商品別、店舗別、日別の販売量である。ＫＰＩとしては、粗利、売れ残り量、欠品の当否などがある。 The planning support device 100 can be used for product purchase planning. The management information (management information database 210) includes the amount of merchandise purchased by merchandise, by store, and by day, the weather in the area where the store is located, the supplier, the purchase price, the purchase amount, and the like. The event is forecasted weather for each store and for each day. The target of prediction is the sales volume for each product, each store, and each day. KPIs include gross profit, the amount of unsold items, and whether there are shortages.
　計画支援装置１００は、会員向けのキャンペーン計画に用いることができる。この会員とは、携帯電話サービスの加入者や、通信販売の会員である。管理情報は、会員別の加入期間、サービス利用回数や利用量、利用金額などである。イベントは、料金割引、特典付与などのキャンペーンである。予測対象は、イベント実行後のサービス利用回数や利用量、利用金額などである。ＫＰＩとしては、会員の定着率、利用量、利用金額などがある。 The planning support device 100 can be used for campaign planning for members. This member is a subscriber of a cellular phone service or a mail-order member. The management information includes the subscription period for each member, the number of times the service is used, the amount of service used, the amount of money used, and the like. An event is a campaign such as a charge discount or privilege. The target of prediction is the number of times the service is used, the amount of service used, the amount of money used, etc. after the event is executed. KPIs include a retention rate of members, a usage amount, a usage amount, and the like.
　計画支援装置１００は、製造機械の保守計画に用いることができる。管理情報は、製造機械別の導入時期、累計稼働時間、前回保守点検からの稼働時間、故障率などである。イベントは、次回保守点検時期ないしは保守点検間隔である。予測対象は、故障発生予測日である。ＫＰＩとしては、製品別の生産量、故障間隔、製造機械別稼働率や平均の稼働率などがある。 The planning support device 100 can be used for maintenance planning of manufacturing machines. The management information includes introduction time for each manufacturing machine, cumulative operating time, operating time since the last maintenance inspection, failure rate, and the like. The event is the next maintenance inspection time or the maintenance inspection interval. The prediction target is the predicted date of failure occurrence. KPIs include production volume by product, interval between failures, operating rate by manufacturing machine, average operating rate, and the like.
　計画支援装置１００は、ハウス生育の農産物の生産計画に用いることができる。管理情報は、ハウス別の植え付き時期、気象データ、肥料データ、生育データなどである。イベントは、ハウス内環境の調整量、肥料の量などである。予測対象は、収穫時期、収穫量などである。ＫＰＩとしては、収穫量、売上金額、廃棄量などがある。 The planning support device 100 can be used for production planning of greenhouse-grown agricultural products. The management information includes the planting time for each house, weather data, fertilizer data, growth data, and the like. The event is the amount of adjustment of the environment in the house, the amount of fertilizer, and the like. The target of prediction is harvest time, harvest amount, and the like. KPIs include the amount of harvest, the amount of sales, and the amount of waste.
≪プログラム≫
　上記した実施形態では、プログラム１２１は、コンピュータである計画支援装置１００に記憶される。記録媒体９１０にあるプログラム１２１が読み込まれて、記憶部１２０にロードされて実行されてもよいし、記録媒体９１０からインストールされて実行されてもよい。
　図１６は、本実施形態に係る記録媒体９１０を示す図である。プログラム１２１が格納された記録媒体９１０からコンピュータ９００に、プログラム１２１をインストールすることで、コンピュータが計画支援装置１００として機能することができるようになる。 ≪Program≫
In the embodiment described above, the program 121 is stored in the planning support device 100, which is a computer. The program 121 in the recording medium 910 may be read, loaded into the storage unit 120 and executed, or may be installed from the recording medium 910 and executed.
FIG. 16 is a diagram showing a recording medium 910 according to this embodiment. By installing the program 121 from the recording medium 910 storing the program 121 to the computer 900 , the computer can function as the planning support device 100 .
≪その他変形例≫
　以上、本発明のいくつかの実施形態について説明したが、これらの実施形態は、例示に過ぎず、本発明の技術的範囲を限定するものではない。例えば、上記した実施形態では、機械学習技術を用いて着荷日を予測しているが、機械学習の替わりに線形回帰モデルなど他の手法を用いて予測してもよい。また、管理情報データベース２１０他のデータベースは計画支援装置１００に記憶されるが、別の装置に記憶される情報をアクセスするようにしてもよい。 ≪Other Modifications≫
Although several embodiments of the present invention have been described above, these embodiments are merely examples and do not limit the technical scope of the present invention. For example, in the above-described embodiment, the arrival date is predicted using machine learning technology, but instead of machine learning, other methods such as a linear regression model may be used for prediction. Also, although the management information database 210 and other databases are stored in the planning support apparatus 100, information stored in another apparatus may be accessed.
　また、上記実施形態では、全てのＫＰＩを算出しＫＰＩ閾値（目標値）との比較をした（図８記載のステップＳ１５，Ｓ１６参照）後に、表示対象となるＫＰＩやその条件、製品を取得している（ステップＳ１７参照）。これに替えて、表示対象となるＫＰＩやその条件、製品を取得した後に、表示対象となるＫＰＩや条件確認に必要なＫＰＩを算出するようにしてもよい。 In the above embodiment, after all KPIs are calculated and compared with KPI thresholds (target values) (see steps S15 and S16 in FIG. 8), KPIs to be displayed, their conditions, and products are acquired. (see step S17). Alternatively, the KPIs to be displayed and the KPIs necessary for confirming the conditions may be calculated after acquiring the KPIs to be displayed, their conditions, and the products.
　図１３、図１５では、結果出力部１１５は、ＫＰＩの目標値を達成したイベントを特定して、当該イベントに対応するＫＰＩを表示しているが、達成度（増減値／比率）に応じた順序にイベントを並べて表示するようにしてもよい。結果出力部１１５は、例えば、達成度の高い順にイベントが並ぶようにＫＰＩを表示してもよい。また、結果出力部１１５は、予め設定された重要度の高いＫＰＩの達成度が高い順にイベントが並ぶようにＫＰＩを表示してもよい。 In FIGS. 13 and 15, the result output unit 115 identifies an event that has achieved the target value of the KPI and displays the KPI corresponding to the event. Events may be arranged and displayed in order. For example, the result output unit 115 may display the KPIs so that the events are arranged in descending order of achievement. In addition, the result output unit 115 may display KPIs such that events are arranged in descending order of the degree of achievement of preset KPIs of high importance.
　本発明はその他の様々な実施形態を取ることが可能であり、さらに、本発明の要旨を逸脱しない範囲で、省略や置換等種々の変更を行うことができる。これら実施形態やその変形は、本明細書等に記載された発明の範囲や要旨に含まれるとともに、特許請求の範囲に記載された発明とその均等の範囲に含まれる。 The present invention can take various other embodiments, and various modifications such as omissions and substitutions can be made without departing from the scope of the present invention. These embodiments and modifications thereof are included in the scope and gist of the invention described in this specification and the like, and are included in the scope of the invention described in the claims and equivalents thereof.
１００　計画支援装置
１１１　学習部
１１２　イベント反映部
１１３　予測部
１１４　ＫＰＩ算出部
１１５　結果出力部
１２２　予測モデル（機械学習モデル）
２１０　管理情報データベース（管理情報）
２１０Ａ　管理情報
２１１　注文番号（管理項目）
２１２　注文依頼日（管理項目）
２１３　製品（管理項目）
２１４　部位（管理項目）
２１５　部品（管理項目）
２１６　取引先（管理項目）
２１７　納期（管理項目）
２１８　着荷日（予測対象項目）
２１９　有効フラグ（管理項目）
２２０　補足情報（管理項目）
２３０　イベント情報データベース
２３３　イベント反映内容（ルール）
２４０　ＫＰＩ情報データベース
２４２　ＫＰＩ閾値設定（目標値、閾値） 100 Planning support device 111 Learning unit 112 Event reflection unit 113 Prediction unit 114 KPI calculation unit 115 Result output unit 122 Prediction model (machine learning model)
210 management information database (management information)
210A Management information 211 Order number (management item)
212 Order request date (control item)
213 Products (control items)
214 parts (control items)
215 parts (control items)
216 Business Partners (Management Items)
217 Delivery date (control item)
218 Arrival date (forecast target item)
219 valid flag (control item)
220 Supplementary Information (Control Items)
230 Event information database 233 Event reflection content (rule)
240 KPI information database 242 KPI threshold setting (target value, threshold)

Claims (7)

  1. 　１つ以上の管理対象に係る管理情報を構成する管理項目を所定のルールに従って変更するイベントに応じて変更するイベント反映部と、
    　前記イベントにより変更された前記管理情報に基づいて、前記管理対象に係る予測対象項目を予測する予測部と、
    　前記管理情報および前記予測対象項目に基づいてＫＰＩを算出するＫＰＩ算出部と、を備え、
    　前記予測部は、過去の前記管理情報および前記予測対象項目の過去の実績を教師データとして訓練された機械学習モデルを用いて予測する
    　ことを特徴とする計画支援装置。     an event reflecting unit that changes management items constituting management information related to one or more management targets in accordance with an event that changes according to a predetermined rule;
    a prediction unit that predicts a prediction target item related to the management target based on the management information changed by the event;
    a KPI calculation unit that calculates a KPI based on the management information and the prediction target item;
    A plan support apparatus, wherein the prediction unit predicts the past performance of the management information and the prediction target item using a machine learning model trained as teacher data.
  2. 　選択された前記管理対象に係る管理情報および予測対象項目に基づいて算出されたＫＰＩを表示装置に出力する結果出力部をさらに備える
    　ことを特徴とする請求項１に記載の計画支援装置。     The planning support apparatus according to claim 1, further comprising a result output unit that outputs a KPI calculated based on the management information and the prediction target item related to the selected management target to a display device.
  3. 　前記ＫＰＩには所定の閾値があり、
    　前記ＫＰＩ算出部は、算出されたＫＰＩが当該ＫＰＩの閾値を満たすか否かを判定し、
    　選択された前記ＫＰＩであり、前記所定の閾値を満たすＫＰＩの算出のもととなった管理情報に対応するイベントと、当該ＫＰＩとを表示装置に出力する結果出力部をさらに備える
    　ことを特徴とする請求項１に記載の計画支援装置。     The KPI has a predetermined threshold,
    The KPI calculation unit determines whether the calculated KPI satisfies the threshold of the KPI,
    The method further comprises a result output unit that outputs to a display device an event corresponding to management information that is the selected KPI and that is the basis of the calculation of the KPI that satisfies the predetermined threshold, and the KPI. The plan support device according to claim 1.
  4. 　前記ＫＰＩ算出部は、前記管理項目を変更しないイベントにおけるＫＰＩである基準ＫＰＩ、および前記ＫＰＩと前記基準ＫＰＩとの増減値／比率を算出する
    　ことを特徴とする請求項１に記載の計画支援装置。     The plan support apparatus according to claim 1, wherein the KPI calculation unit calculates a reference KPI, which is a KPI in an event in which the management item is not changed, and an increase/decrease value/ratio between the KPI and the reference KPI. .
  5. 　前記管理対象は、製品を構成する部品であって、
    　前記管理項目は、前記部品の発注日、発注先、納期を含み、
    　前記予測対象項目は、発注された前記部品の着荷日であり、
    　前記イベントは、前記納期を所定期間延長する変更、前記納期を所定期間前倒しする変更、前記発注先の変更、発注取消しの何れかを含む
    　ことを特徴とする請求項１に記載の計画支援装置。     The management target is a part that constitutes a product,
    The management items include order date, supplier, and delivery date of the parts,
    the item to be predicted is the arrival date of the ordered parts,
    2. The plan support apparatus according to claim 1, wherein the event includes any one of a change to extend the delivery date for a predetermined period, a change to move the delivery date forward for a predetermined period, a change of the supplier, and a cancellation of the order.
  6. 　コンピュータを、請求項１～５の何れか１項に記載の計画支援装置として機能させるためのプログラム。 A program for causing a computer to function as the planning support device according to any one of claims 1 to 5.
  7. 　計画支援装置が実行する計画支援方法であって、
    　前記計画支援装置は、
    　１つ以上の管理対象に係る管理情報を構成する管理項目を所定のルールに従って変更するイベントに応じて変更するステップと、
    　前記イベントにより変更された前記管理情報に基づいて、前記管理対象に係る予測対象項目を予測するステップと、
    　前記管理情報および前記予測対象項目に基づいてＫＰＩを算出するステップと、を実行し、
    　前記計画支援装置は、前記予測対象項目を予測するときに、過去の前記管理情報および前記予測対象項目の過去の実績を教師データとして訓練された機械学習モデルを用いて予測する
    　ことを特徴とする計画支援方法。     A plan support method executed by a plan support device,
    The planning support device is
    a step of changing management items constituting management information related to one or more management targets according to an event that changes according to a predetermined rule;
    predicting a prediction target item related to the management target based on the management information changed by the event;
    calculating a KPI based on the management information and the prediction target item;
    When predicting the prediction target item, the plan support device predicts the past management information and the past performance of the prediction target item using a machine learning model trained as teacher data. Planning assistance method.
PCT/JP2021/001992 2021-01-21 2021-01-21 Planning support device, program, and planning support method WO2022157876A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/JP2021/001992 WO2022157876A1 (en) 2021-01-21 2021-01-21 Planning support device, program, and planning support method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2021/001992 WO2022157876A1 (en) 2021-01-21 2021-01-21 Planning support device, program, and planning support method

Publications (1)

Publication Number Publication Date
WO2022157876A1 true WO2022157876A1 (en) 2022-07-28

Family

ID=82548599

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2021/001992 WO2022157876A1 (en) 2021-01-21 2021-01-21 Planning support device, program, and planning support method

Country Status (1)

Country Link
WO (1) WO2022157876A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115562948A (en) * 2022-12-05 2023-01-03 成都索贝数码科技股份有限公司 Large-scale parallelization multi-KPI prediction method and system

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2016118975A (en) * 2014-12-22 2016-06-30 株式会社日立製作所 Marketing measures optimizer, method and program
JP2017162044A (en) * 2016-03-08 2017-09-14 株式会社日立ソリューションズ東日本 Production planning device, production planning method and production planning program
JP2020119085A (en) * 2019-01-21 2020-08-06 株式会社日立製作所 Computer system and method of providing information useful to accomplish goal relating to object

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2016118975A (en) * 2014-12-22 2016-06-30 株式会社日立製作所 Marketing measures optimizer, method and program
JP2017162044A (en) * 2016-03-08 2017-09-14 株式会社日立ソリューションズ東日本 Production planning device, production planning method and production planning program
JP2020119085A (en) * 2019-01-21 2020-08-06 株式会社日立製作所 Computer system and method of providing information useful to accomplish goal relating to object

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115562948A (en) * 2022-12-05 2023-01-03 成都索贝数码科技股份有限公司 Large-scale parallelization multi-KPI prediction method and system

Similar Documents

Publication Publication Date Title
US7974851B2 (en) Method and system for price planning
US8165910B2 (en) Method and system for price planning
US20040199445A1 (en) Business activity management system
WO2003081492A1 (en) Business profit improvement support system
JP2004274956A (en) Method, device and program for making generation plan and power selling and buying plan
CN115526686A (en) E-commerce ERP system with profit trial calculation function and computer equipment
US20040098403A1 (en) Method and apparatus for displaying real-time status of product availability, orders, and sales revenue
JP2004021364A (en) Management intention decision support system
JP2007323680A (en) Management decision support system
Huang Information architecture for effective Workload Control: an insight from a successful implementation
Rad et al. Metrics for project management: Formalized approaches
WO2022157876A1 (en) Planning support device, program, and planning support method
JP2004094943A (en) Integrated management information management system
US20080052143A1 (en) Sales Plan Evaluation Support System
Batra et al. Right product, right time, right location: Quantifying the semiconductor supply chain
JP2001312535A (en) Support system for production and sales schedule
Masvidal Andreu Inventory drivers in a pharmaceutical supply chain
Dorfling Exploring the potential influence of inventory forecasting on profitability
Lin et al. Managing a sea of quality information at Ark‐La‐Tex aquatics
Li et al. A system dynamics approach of a laptop manufacturing supply chain operation simulation modelling using AnyLogic
JPH09179916A (en) Component inventory schedule planning support system
Santos Digitalization and Optimized Inventory Management of a General Warehouse
Chase Building A Holistic Supply Chain With Consumption-Based Forecasting & Planning.
Ortiz Improvement of Order-Processing Process at a Pharmacy
Βαρλάς Supply chain stochastic modeling: analysis, design and optimization of flows in supply networks

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21920989

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21920989

Country of ref document: EP

Kind code of ref document: A1