WO2022157957A1 - 在庫費算出装置、及び、在庫費算出方法 - Google Patents
在庫費算出装置、及び、在庫費算出方法 Download PDFInfo
- Publication number
- WO2022157957A1 WO2022157957A1 PCT/JP2021/002381 JP2021002381W WO2022157957A1 WO 2022157957 A1 WO2022157957 A1 WO 2022157957A1 JP 2021002381 W JP2021002381 W JP 2021002381W WO 2022157957 A1 WO2022157957 A1 WO 2022157957A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- inventory
- cost
- data
- related information
- inventory cost
- Prior art date
Links
- 238000004364 calculation method Methods 0.000 title claims abstract description 199
- 238000010801 machine learning Methods 0.000 claims description 13
- 238000004519 manufacturing process Methods 0.000 claims description 12
- 238000013500 data storage Methods 0.000 description 31
- 238000010586 diagram Methods 0.000 description 28
- 230000002596 correlated effect Effects 0.000 description 18
- 230000006870 function Effects 0.000 description 16
- 238000000034 method Methods 0.000 description 12
- 238000012545 processing Methods 0.000 description 10
- 238000013528 artificial neural network Methods 0.000 description 8
- 230000000875 corresponding effect Effects 0.000 description 7
- 239000000463 material Substances 0.000 description 4
- 230000007423 decrease Effects 0.000 description 3
- 230000014759 maintenance of location Effects 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 230000002787 reinforcement Effects 0.000 description 3
- 230000006866 deterioration Effects 0.000 description 2
- 238000005265 energy consumption Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 230000029305 taxis Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000012384 transportation and delivery Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Definitions
- the present disclosure relates to an inventory cost calculation device and an inventory cost calculation method.
- Patent Document 1 for the purpose of realizing inventory management that reflects the company's sales strategy, each warehouse's policy (inventory reduction priority, service priority, etc.) and each distributor's policy (campaign implementation, inventory reduction implementation etc.) is used to calculate the cost incurred in the warehouse (warehouse cost).
- Patent Document 2 describes costs proportional to the period (depreciation cost, fixed cost, etc.), costs proportional to the number (inventory storage unit price, warehousing unit price, etc.), costs proportional to the number and period (personnel cost, equipment cost, etc.) cost, etc.) to calculate the warehouse cost.
- logistics bases in the supply chain of the manufacturing industry include bases managed by the manufacturer (e.g. factories, distributors) and bases not managed by the manufacturer (e.g. distributors). included.
- the manufacturer calculates inventory costs based on information for calculating inventory costs (inventory-related information) provided by each base.
- sites not managed by the company may not disclose all inventory-related information.
- Patent Document 1 which is based on the premise that inventory-related information can be obtained for each base, it is impossible to calculate inventory costs for bases that do not disclose all of the inventory-related information. I had a problem.
- the manufacturer tends to pass on the inventory to the bases not managed by the manufacturer in an attempt to reduce the inventory costs generated at the bases managed by the manufacturer. tends towards As a result, there will be excess inventory at bases that are not managed by the company, which will eventually lead to a decline in product selling prices.
- the inventory cost does not include the cost that correlates with the storage period of the inventory. Therefore, there is a problem that the actual inventory cost cannot be calculated because the inventory cost is not calculated taking into account the actual storage period, such as when the inventory is stagnant at the base.
- This disclosure solves the above-mentioned problems, and even if the information for calculating the inventory cost that occurs at each base in the supply chain of the manufacturing industry is not disclosed, it is possible to calculate the inventory cost. This is intended to suppress deterioration in accuracy.
- the inventory cost calculation device is an inventory cost calculation device that calculates the inventory cost of inventory items generated at each base including factories and sales agents in the supply chain of the manufacturing industry, and the inventory cost generated at the base an inventory-related information acquisition unit that acquires inventory-related information including actual data for calculating the inventory cost of the item and estimated data for estimating the inventory cost of the inventory item; An inventory cost calculation unit that calculates the inventory cost of the inventory based on the actual data and the estimated data included in the inventory-related information, and the ratio of the estimated data included in the inventory-related information acquired by the inventory-related information acquisition unit.
- an error estimating unit for estimating the level of error included in the inventory cost of the inventory item calculated by the inventory cost calculation unit, and the inventory cost of the inventory item calculated by the inventory cost calculation unit, and the error included in the inventory cost of the inventory item calculated based on the level of the error included in the inventory cost of the inventory item estimated by the error estimation unit, and the error calculated by the inventory cost calculation unit and an output unit for outputting an inventory cost of the inventory items.
- FIG. 1 is a configuration diagram of an inventory cost calculation device according to Embodiment 1;
- FIG. It is a figure which shows an example of the inventory related information disclosed with respect to the manufacturer. It is a figure which shows the function implement
- FIG. 10 is a diagram showing the relationship between an inventory cost error and a ratio of estimated data; It is a figure which shows the function implement
- FIG. 4 is a diagram showing inventory-related information of the factory 1; FIG. 4 is a diagram showing inventory-related information of the factory 1; FIG. 4 is a diagram showing inventory-related information of a factory 2; FIG. 4 is a diagram showing inventory-related information of a factory 2; FIG.
- FIG. 2 is a diagram showing inventory-related information of a sales agency 1;
- FIG. 2 is a diagram showing inventory-related information of a sales agency 1;
- FIG. 4 is a diagram showing inventory-related information of a sales agent 2;
- FIG. 4 is a diagram showing inventory-related information of a sales agent 2;
- FIG. 4 is a diagram showing inventory-related information of a sales agent 3;
- FIG. 4 is a diagram showing inventory-related information of a sales agent 3;
- FIG. It is a block diagram of the learning apparatus regarding an inventory cost calculation apparatus.
- FIG. 10 is a diagram showing functions implemented by a distributor inventory cost calculation unit according to the second embodiment;
- FIG. 10 is a flow chart of processing for acquiring an error level from a trained model;
- FIG. FIG. 11 is a configuration diagram of an inventory cost calculation device according to Embodiment 3;
- the inventory cost calculation device of the present disclosure calculates inventory costs for inventory items generated at each base including factories and distributors in the supply chain of the manufacturing industry.
- Bases in the supply chain of the manufacturing industry include bases managed by the manufacturer itself (eg factories, distributors) and bases of other companies not managed by the manufacturer (eg distributors).
- Distributors at their own bases and other companies' bases may disclose only a part of the information (inventory-related information) for calculating the inventory cost of inventory items to the manufacturer.
- FIG. 1 is a configuration diagram of an inventory cost calculation device 100 according to Embodiment 1.
- the inventory cost calculation device 100 includes a factory inventory data input unit 101 , a factory inventory data storage unit 102 , a sales agent inventory data input unit 103 , a sales agent inventory data storage unit 104 and a control unit 105 .
- the in-factory inventory data input unit 101 inputs data (time data, quantity data, cost unit price data) disclosed by the factory for calculating the inventory cost of inventory items generated in the factory. obtained through an undisclosed network. Then, the factory inventory data input unit 101 stores the acquired data in the factory inventory data storage unit 102 .
- the sales agent inventory data input unit 103 is data disclosed by the sales agent and used to calculate inventory costs for inventory items generated at the sales agent (seller type data, sales agent data). location data, inventory time data, quantity data, cost unit price data at the sales agency) are acquired via a network (not shown), for example. Then, the distributor inventory data input unit 103 stores the acquired data in the distributor inventory data storage unit 104 .
- Time data 106 , quantity data 107 , and unit cost data 108 are stored in the in-factory inventory data storage unit 102 .
- the time data 106, the quantity data 107, and the cost unit price data 108 are examples of inventory-related information for calculating inventory costs for inventory items generated at bases.
- the inventory-related information includes actual data for calculating inventory costs for inventory items generated in the factory.
- the actual data included in the inventory-related information includes, for example, data obtained from the in-factory inventory data input unit 101, that is, data disclosed by the factory.
- Examples of the time data 106 include inventory retention time and inventory manufacturing time.
- Examples of the quantity data 107 include material consumption, energy consumption, the number of workers, and the number of equipment used.
- Examples of unit cost data 108 include utility costs, manufacturing costs, inventory interest rates, depreciation costs, material costs, labor costs, and property taxes.
- the sales agent inventory data storage unit 104 stores sales agent type data 109, position data 110, time data 111, quantity data 112, unit cost data 113, and error level data 120.
- Distributor type data 109, position data 110, time data 111, quantity data 112, cost unit price data 113, and error level data 120 are inventory-related information for calculating inventory costs for inventory items generated at bases.
- the inventory-related information includes actual data for calculating the inventory cost of the inventory generated at the distributor and estimated data for estimating the inventory cost of the inventory.
- the actual data included in the inventory-related information includes data acquired from the distributor inventory data input unit 103, ie, data disclosed by the distributor.
- the estimated data included in the inventory-related information is based on the actual data for calculating the inventory costs of the inventory items generated by one distributor, and the inventory costs of the inventory items generated by another distributor. contains data for estimating As an example of the error level data 120, there is data indicating the level of error that occurs when calculating the inventory cost of the inventory that occurs at the sales agency based on the actual data and estimated data included in the inventory-related information. mentioned. The level of error correlates with the number of actual data and the number of estimated data included in the inventory related information.
- Examples of the distributor type data 109 include company sites managed by the manufacturer (eg, group companies, companies with capital ties, etc.), sites not managed by the manufacturer (eg, independent companies, companies with no capital relationship), capital, scale, number of stores, etc.
- An example of location data 110 is the location of a sales agent.
- Examples of the time data 111 include an inventory retention time and a loading/unloading work time.
- Examples of the quantity data 112 include the number of goods received, the number of goods discharged, the number of equipment used, and the number of workers.
- Examples of the cost unit price data 113 include labor costs, utility costs, inventory interest rates, property taxes, depreciation costs, and utility costs.
- the information contained in the sales agency type data 109, the location data 110, the time data 111, the quantity data 112, and the cost unit price data 113 are examples of base information.
- An in-factory inventory cost calculation unit 114 in the control unit 105 uses the data in the in-factory inventory data storage unit 102 to calculate an inventory cost for inventory items generated in the factory.
- a sales agent inventory cost calculation unit 115 in the control unit 105 uses the data in the sales agent inventory data storage unit 104 to calculate inventory costs for inventory items generated in the sales agent.
- the distributor inventory cost calculation unit 115 in the control unit 105 uses the data in the distributor inventory data storage unit 104 to estimate the error level included in the inventory costs in the distributor.
- the total inventory cost calculation unit 116 in the control unit 105 calculates the total Calculate the inventory cost of In addition, the total inventory cost calculation unit 116 in the control unit 105 calculates the inventory cost in the sales agent calculated by the sales agent inventory cost calculation unit 115 and the error in the sales agent calculated by the sales agent inventory cost calculation unit 115. From the level of , calculate the error included in the inventory cost in the distributor. That is, the total inventory cost calculation unit 116 is an example of an output unit.
- the estimation unit 119 in the control unit 105 obtains inventory-related information that has not been disclosed by a sales agent other than a certain sales agent based on actual data included in the inventory-related information of a certain sales agent. Outputs the estimated data obtained by estimating the actual data. That is, the estimating unit 119 in the control unit 105 calculates the inventory cost of the inventory items generated at another sales agent based on the actual data for calculating the inventory cost of the inventory items generated at the sales agent. Output guess data for guessing.
- FIG. 2 is a diagram showing an example of inventory-related information disclosed to manufacturers.
- the sales agency of its own base discloses the actual data of inventory-related information corresponding to the number of goods received, the number of goods discharged, the retention time, and the labor cost to the manufacturer.
- the sales agent at the other company's base disclosed the actual data of the inventory-related information corresponding to the number of goods received to the manufacturer.
- the data disclosed to the manufacturer is reflected in the in-factory inventory data storage unit 102 or the distributor inventory data storage unit 104 of the inventory cost calculation device 100 .
- distributors often disclose only part of the actual data of inventory-related information to the manufacturer, and rarely disclose all the actual data to the manufacturer.
- the amount of inventory-related information disclosed to the manufacturer also differs depending on whether the sales agent has its own base or another company's base.
- the amount of inventory-related information to be disclosed to manufacturers is larger for sales agents based in their own bases than for sales agents based at other companies' bases.
- the estimation unit 119 of the inventory cost calculation device 100 based on the actual data included in the inventory-related information of a certain sales agent, is disclosed by a sales agent other than a certain sales agent. Estimated data that is estimated from the actual data of the inventory-related information that has not been stored is output. The estimation unit 119 generates estimation data, for example, based on actual data of inventory-related information disclosed by sales agents with similar base information.
- the base information includes at least one of position information including the location of the sales agent, the size of the sales agent, and the number of items shipped to the sales agent.
- the base information may include the number of stores of the distributor, the capital, the number of workers, the number of equipment used, and the like.
- sales agent base information for example, sales agents located in areas with high land prices tend to have higher unit costs, while sales agents located in areas with low land prices
- sales agents located in areas with low land prices One can speculate that the located distributors tend to have lower unit prices for each cost. This increases the accuracy of inferring the inferred data and reduces the level of inventory cost error.
- FIG. 3 is a diagram showing functions realized by the sales agent inventory cost calculation unit 115.
- the inventory-related information acquisition unit 1151 obtains actual data for calculating the inventory cost of inventory items generated at the sales agent from the sales agent inventory data storage unit 104, and to estimate the inventory cost of a certain inventory item. Get inventory-related information, including inferred data for
- the inventory cost calculation unit 1152 calculates the inventory cost of the inventory item based on the actual data and the estimated data included in the inventory related information acquired by the inventory related information acquisition unit 1151 and outputs the calculated inventory cost to the overall inventory cost calculation unit 116 .
- the inventory cost calculation unit 1152 calculates the amount of inventory generated at the sales agent based on the estimated data of the inventory-related information even when the actual data of the inventory-related information is not disclosed by the sales agent. Inventory costs can be calculated.
- the error estimation unit 1153 outputs information (error level) indicating the accuracy of the inventory cost calculated by the inventory cost calculation unit 1152 .
- the error estimation unit 1153 calculates the inventory of the inventory items calculated by the inventory cost calculation unit 1152 according to the ratio of the number of estimated data included in the inventory-related information acquired by the inventory-related information acquisition unit 1151.
- the error level included in the cost is estimated and output to the total inventory cost calculation unit 116 .
- the error estimator 1153 refers to the error level data 120 to estimate the error level according to the ratio of the number of estimated data items included in the inventory-related information.
- the error level is indicated by "ratio (%)".
- the mode of the error level is not limited to this. At least, the error level may be information indicating the degree of error such as "high", “medium", or "low”.
- the error level data 120 includes data indicating the relationship between the level (percentage) of inventory cost error and the percentage of the number of estimated data, as shown in FIG. In pattern 4 of FIG. 4, the level of inventory cost error is set to 0 (%) of the inventory cost when the ratio of the number of estimated data is 0 (%) or more and less than 25 (%). Also, when the ratio of the number of estimated data is 25(%) or more and less than 50(%), the level of inventory cost error is set to 20(%) of the inventory cost. Further, when the ratio of the number of estimated data is 50(%) or more and less than 75(%), the level of inventory cost error is set to 30(%) of the inventory cost.
- the level of inventory cost error is set to 40 (%) of the inventory cost.
- the error estimation unit 1153 estimates that the level (percentage) of the inventory cost error corresponding to the ratio of the number of estimated data is 40 (%).
- the error estimation unit 1153 estimates that the level (percentage) of the inventory cost error corresponding to the percentage of the number of estimated data is 0 (%).
- the relationship between the inventory cost error level (ratio) and the estimated data number ratio is not limited to a stepwise function pattern such as pattern 4 in FIG. 4, and various patterns are possible. For example, there are linear function patterns such as pattern 1, upward convex function patterns such as pattern 2, downward convex function patterns such as pattern 3, and the like.
- the relationship between the level (percentage) of inventory cost error and the percentage of the number of estimated data is not limited to the exemplified pattern. should be larger.
- the relationship between the inventory cost error level (ratio) and the ratio of the number of estimated data may be set by the user's operation of the inventory cost calculation device 100, or set when the inventory cost calculation device 100 is introduced. may be Furthermore, the relationship between the level (percentage) of inventory cost error and the percentage of the number of actual data may be set for each distributor.
- FIG. 5 is a diagram showing functions realized by the in-factory inventory cost calculation unit 114.
- the inventory-related information acquisition unit 1141 acquires inventory-related information from the in-factory inventory data storage unit 102 for calculating an inventory cost for inventory items generated in the factory.
- the inventory-related information acquisition unit 1141 is an example of an inventory-related information acquisition unit that acquires inventory-related information for calculating the inventory cost of inventory items generated at the base.
- the inventory cost calculation unit 1142 calculates the inventory cost of the inventory items based on the inventory related information acquired by the inventory related information acquisition unit 1141 .
- the inventory cost calculation unit 1142 separately calculates the inventory cost that correlates with time and the inventory cost that does not correlate with time.
- the inventory cost calculation unit 1142 outputs the calculated inventory cost of the inventory item to the overall inventory cost calculation unit 116 .
- the in-factory inventory cost calculator 114 calculates and outputs the inventory costs generated in each factory.
- the function of the in-factory inventory cost calculation unit 114 may be realized by the function of the sales agency inventory cost calculation unit 115 .
- the function of the inventory-related information acquisition unit 1141 of the in-factory inventory cost calculation unit 114 may be realized by the function of the inventory-related information acquisition unit 1151 of the distributor inventory cost calculation unit 115 .
- the function of the inventory cost calculation unit 1142 of the in-factory inventory cost calculation unit 114 may be realized by the function of the inventory cost calculation unit 1152 of the distributor inventory cost calculation unit 115 .
- FIG. 6 is a diagram showing an example of hardware constituting the inventory cost calculation device 100.
- the CPU 401 executes the programs stored in the main memory, etc., so that the in-factory inventory cost calculation unit 114, the estimation unit 119, the distributor inventory cost calculation unit 115, and the overall inventory cost calculation unit shown in FIG. Each function of 116 is realized.
- the main memory 402 is, for example, a non-volatile memory, and stores various programs executed by the CPU 401 .
- An interface 404 is an interface for inputting/outputting data from an external device. The interface 404, for example, acquires information transmitted from factories and distributors via a network.
- the storage unit 405 is, for example, an HDD, and has the in-factory inventory data storage unit 102 and the distributor inventory data storage unit 104 . Various data processed by the CPU 401 are stored. Storage unit 405 transfers the stored data to CPU 401 .
- the in-factory inventory cost calculation unit 114 calculates the inventory costs generated in the factory by dividing them into costs that are correlated with time and costs that are not correlated with time. Costs correlated with time are calculated using the data table shown in FIG. FIG. 7 is an example of a table for calculating costs that are correlated with time.
- the inventory-related information acquisition unit 1141 acquires time data and cost unit price data for each inventory ID from the in-factory inventory data storage unit 102 . Examples of types of time data are residence time and production time. The value of the time data is a value in units representing time (seconds, minutes, hours, etc.). Examples of types of cost unit data are utility costs, manufacturing costs, inventory interest rates, and labor costs.
- the value of the cost unit price data is a value in a unit representing an amount of money per hour (an amount of money per second (yen), an amount of money per minute (US dollar), etc.).
- the inventory cost calculation unit 1142 calculates the inventory cost for each inventory ID by multiplying the value of the time data acquired by the inventory related information acquisition unit 1141 and the value of the unit cost data for each inventory ID.
- Costs that have no correlation with time are calculated using the data table shown in FIG. FIG. 8 is an example of a table for calculating costs that have no correlation with time.
- the inventory-related information acquisition unit 1141 acquires quantity data and cost unit price data from the in-factory inventory data storage unit 102 .
- Examples of types of quantity data are material consumption, energy consumption and number of workers.
- Quantity data values are values in units (pieces, people, liters, kilowatts, grams, etc.) that represent numbers, amounts, and weights.
- Examples of types of cost unit data are material costs, utility costs, and labor costs.
- the value of the cost unit price data is a value of a unit representing a price per quantity (price per unit (yen), price per kilowatt (US dollar), etc.).
- the inventory cost calculation unit 1142 calculates the inventory cost by multiplying the value of the quantity data acquired by the inventory related information acquisition unit 1141 by the value of the unit cost data.
- the method of calculating the inventory costs incurred by distributors will be described in Figure 2, divided into the items " ⁇ : Information disclosed to manufacturers" and "X: Information not disclosed to manufacturers”. .
- the distributor inventory cost calculation unit 115 calculates the inventory costs incurred by the distributor by dividing them into costs that are correlated with time and costs that are not correlated with time. Costs that are correlated with time are calculated using the data table shown in FIG. FIG. 9 is an example of a table for calculating costs that are correlated with time.
- the inventory-related information acquisition unit 1151 acquires time data and cost unit price data from the distributor inventory data storage unit 104 . An example of a type of time data is dwell time.
- the item "actual data/estimated data” contains data "actual data” when the type of time data is information disclosed to the manufacturer.
- the value of the time data is a value in units representing time (seconds, minutes, hours, etc.). Examples of types of cost unit data include labor costs, utility costs, and inventory interest rates.
- the item "actual data/estimated data” contains data "actual data” when the type of unit cost data is disclosed to the manufacturer.
- the value of the cost unit price data is a value in a unit representing an amount of money per hour (an amount of money per second (yen), an amount of money per minute (US dollar), etc.).
- the inventory cost calculation unit 1152 calculates the inventory cost by multiplying the value of the time data acquired by the inventory related information acquisition unit 1151 by the value of the cost unit price data.
- Costs that are not correlated with time are calculated using the data table shown in FIG. FIG. 10 is an example of a table for calculating costs that have no correlation with time.
- the inventory-related information acquisition unit 1151 acquires quantity data and cost unit price data from the distributor inventory data storage unit 104 .
- types of quantity data include the number of goods received and the number of goods shipped.
- the item "actual data/estimated data” contains data "actual data” when the type of quantity data is disclosed to the manufacturer.
- Quantity data values are values in units (pieces, people, liters, kilowatts, grams, etc.) that represent numbers, amounts, and weights.
- Examples of types of cost unit data are labor costs, utility costs, and inventory costs.
- the item "actual data/estimated data” contains data "actual data” when the type of unit cost data is disclosed to the manufacturer.
- the value of the cost unit price data is a value of a unit representing a price per quantity (price per unit (yen), price per kilowatt (US dollar), etc.).
- the inventory cost calculation unit 1152 calculates the inventory cost by multiplying the value of the quantity data acquired by the inventory related information acquisition unit 1151 by the value of the unit cost data.
- FIG. 11 is a flow chart showing the processing of the in-factory inventory cost calculator 114 .
- FIG. 12 is a flow chart showing the processing of the distributor inventory cost calculation unit 115.
- FIG. 13 is a flow chart showing the processing of the total inventory cost calculation unit 116. As shown in FIG. 11
- step S1 the in-factory inventory cost calculator 114 first selects a factory to be calculated from among a plurality of factories.
- step S2 the in-factory inventory cost calculation unit 114 selects an inventory ID to be first calculated from among the inventory IDs of a plurality of inventory items stored in the factory selected in step S2.
- step S3 the in-factory inventory cost calculation unit 114 (inventory related information acquisition unit 1141) obtains actual data (time data, cost unit price data) for calculating the inventory cost of the inventory item with the inventory ID selected in step S2. It acquires the inventory-related information including inventory information from the in-factory inventory data storage unit 102 .
- step S4 the in-factory inventory cost calculation unit 114 (inventory cost calculation unit 1142) calculates inventory items for each inventory ID based on the actual data (time data, cost unit price data) included in the inventory-related information acquired in step S3. Calculate the inventory cost of In step S5, the in-factory inventory cost calculation unit 114 determines whether or not there is an inventory ID for which inventory cost calculation has not been performed. , S2, S3, and S4. If No, go to step S6. In step S6, the in-factory inventory cost calculator 114 adds up the inventory costs of the respective inventory IDs to calculate costs that are correlated with time. As described above, steps S1 to S6 are the cost calculation flow that has a correlation with time.
- step S7 the in-factory inventory cost calculation unit 114 (inventory-related information acquisition unit 1141) obtains inventory-related information including actual data (quantity data, cost unit price data) for calculating the inventory cost of inventory items. Obtained from the in-factory inventory data storage unit 102 .
- step S8 the in-factory inventory cost calculation unit 114 (inventory cost calculation unit 1142) calculates the inventory cost based on the actual data (quantity data, cost unit price data) included in the inventory-related information acquired in step S7.
- step S9 the in-factory inventory cost calculator 114 adds up the inventory costs calculated in step S8 to calculate costs that are not correlated with time.
- the above steps S7 to S9 are the cost calculation flow that has no correlation with time.
- step S10 the in-factory inventory cost calculation unit 114 determines whether or not there is a factory for which inventory cost calculation has not been performed. Perform up to S9. If No, go to step S11. In step S11, the in-factory inventory cost calculator 114 outputs the inventory cost of each factory, and ends the process.
- step S101 the sales agent inventory cost calculator 115 first selects a sales agent to be calculated from among a plurality of sales agents.
- step S102 the distributor inventory cost calculation unit 115 (inventory-related information acquisition unit 1151) calculates the actual data for calculating the inventory cost of the inventory of the distributor selected in step S101 and the inventory cost of the inventory. Inventory-related information including estimation data (time data, cost unit price data) for estimating is obtained from the distributor inventory data storage unit 104 .
- step S103 the distributor inventory cost calculation unit 115 (inventory cost calculation unit 1152) calculates inventory items based on actual data and estimated data (time data, cost unit price data) included in the inventory-related information acquired in step S102. Calculate the inventory cost of In step S104, the distributor inventory cost calculation unit 115 adds up the inventory costs calculated in step S103 to calculate costs that are correlated with time.
- the above steps S102 to S104 are the calculation flow of the cost having a correlation with time.
- step S105 the distributor inventory cost calculation unit 115 (inventory-related information acquisition unit 1151) collects the actual data and the inventory cost for calculating the inventory cost of the inventory of the distributor selected in step S101. Inventory-related information including estimated data (quantity data, cost unit price data) for estimating is obtained from the distributor inventory data storage unit 104 .
- step S106 the distributor inventory cost calculation unit 115 (inventory cost calculation unit 1152) calculates inventory items based on the actual data and estimated data (quantity data, cost unit price data) included in the inventory-related information acquired in step S105. Calculate the inventory cost of
- step S107 the distributor inventory cost calculation unit 115 adds up the inventory costs calculated in step S106 to calculate costs that are correlated with time. The steps from step S105 to step S107 are the cost calculation flow that has no correlation with time.
- the sales agent inventory cost calculation unit 115 causes the inventory cost calculation unit 1152 to Estimate the level of error included in the calculated inventory cost of inventory. Specifically, the error estimating unit 1153 calculates the error level of the inventory cost from the number of items that are “actual data” and “estimated data” in the inventory-related information acquired in steps S102 and S105. Guess. The level of inventory cost error is estimated from the ratio of the number of estimated data to the total number of actual data and estimated data.
- step S109 the sales agent inventory cost calculation unit 115 determines whether or not there is a sales agent for which inventory cost calculation has not been performed. is performed from step S101 to step S108. If No, go to step S110. In step S110, the sales agent inventory cost calculation unit 115 outputs the level of error between the inventory cost of each sales agent and the inventory cost of each sales agent, and ends the process.
- step S ⁇ b>201 the total inventory cost calculation unit 116 acquires the inventory cost of each factory calculated by the inventory cost calculation unit 1142 of the intra-factory inventory cost calculation unit 114 .
- step S202 the total inventory cost calculation unit 116 calculates the inventory cost of each sales agent calculated by the inventory cost calculation unit 1152 of the sales agent inventory cost calculation unit 115 and the error of the sales agent inventory cost calculation unit 115.
- the error level of the inventory cost of each sales agent estimated by the estimation unit 1153 is acquired.
- step S203 the total inventory cost calculation unit 116 calculates the inventory cost of each sales agent based on the inventory cost of each sales agent acquired in step S202 and the level of error in the inventory cost of each sales agent. calculated as the amount of error.
- step S204 the total inventory cost calculation unit 116 calculates the inventory cost calculated by the inventory cost calculation unit 1142 of the in-factory inventory cost calculation unit 114 acquired in step S201, and the distributor inventory acquired in step S202.
- the inventory cost calculated by the inventory cost calculator 1152 of the cost calculator 115 is output. That is, the inventory cost of each factory and the inventory cost of each distributor are output.
- step S205 the total inventory cost calculation unit 116 outputs the error in the inventory cost of each sales agent calculated in step S203.
- FIG. 14 and 15 are diagrams showing inventory-related information disclosed by the factory 1.
- FIG. 16 and 17 are diagrams showing inventory-related information disclosed by the factory 2.
- the in-factory inventory cost calculation unit 114 selects the factory 1 in step S1 of FIG.
- the in-factory inventory cost calculator 114 selects the inventory ID 101 in step S2.
- the in-factory inventory cost calculator 114 acquires three rows of data in which the "inventory ID" in FIG. 14 is "101".
- step S5 the in-factory inventory cost calculation unit 114 returns to step S1 because there is an inventory ID "102" for which inventory cost calculation has not been performed.
- step S2 the in-factory inventory cost calculation unit 114 selects the inventory ID 102.
- step S3 the in-factory inventory cost calculation unit 114 acquires two rows of data in which the "inventory ID" in FIG. 14 is "102".
- step S5 the in-factory inventory cost calculation unit 114 does not have an inventory ID for which inventory cost calculation has not yet been performed, so the process proceeds to step S6.
- step S7 the in-factory inventory cost calculator 114 acquires one row of data in FIG.
- step S9 the in-factory inventory cost calculator 114 calculates the cost that has no correlation with time as 1000 (yen). (In this example, there is only one line of data acquired in step S7, so the value calculated in step S8 is used as the value in step S9. If there are multiple lines of data acquired in step S7, those values are They are added together in step S9.)
- step S10 the in-factory inventory cost calculation unit 114 returns to step S1 because there is a factory "factory 2" for which inventory cost calculation has not been performed.
- Factory 2 is selected in step S1, and steps S2 to S9 are processed in the same manner as in the case of factory 1.
- FIG. 18 and 19 are diagrams showing inventory-related information of the distributor 1.
- FIG. 20 and 21 are diagrams showing inventory-related information of the distributor 2.
- FIG. 22 and 23 are diagrams showing inventory-related information of the distributor 3.
- FIG. Figures 18 and 19 show the data for sales agent 1 (own base: sales agent of a group company of the factory), and Figures 20 and 21 show the data for sales agent 2 (other company's base: independent sales agent). 22 and 23 are used for the data of sales agent 3 (base of other company: independent sales agent).
- the sales agent inventory cost calculation unit 115 selects the sales agent 1 in step S101 of FIG.
- step S102 the distributor inventory cost calculator 115 acquires two rows of data in FIG.
- step S105 the distributor inventory cost calculation unit 115 acquires two rows of data in FIG.
- step S108 the distributor inventory cost calculation unit 115 (error estimation unit 1153) estimates the level (percentage) of inventory cost error corresponding to the percentage of the number of estimated data.
- the ratio of the number of estimated data is 0 (%).
- step S109 the distributor inventory cost calculation unit 115 returns to step S101 because there are distributors 2 and 3 that have not yet calculated the inventory cost.
- step S102 the distributor inventory cost calculator 115 acquires two rows of data in FIG.
- step S105 the distributor inventory cost calculator 115 acquires two rows of data in FIG.
- step S108 the distributor inventory cost calculation unit 115 (error estimation unit 1153) estimates the level (percentage) of the inventory cost error corresponding to the percentage of the number of estimated data.
- step S109 the sales agent inventory cost calculation unit 115 returns to step S101 because the sales agent 3 exists as a sales agent for which inventory cost calculation has not been performed.
- step S201 the total inventory cost calculation unit 116 acquires 1791 (yen) as the inventory cost of the factory 1 and 1502 (yen) as the inventory cost of the factory 2.
- step S202 the total inventory cost calculation unit 116 calculates 1800 (yen) as the inventory cost of sales agent 1, 0 (%) as the level of inventory cost error, 1055 (yen) as the inventory cost of sales agent 2, 40(%) is obtained as the level of inventory cost error, 1860 (yen) as the inventory cost of sales agent 3, and 40(%) as the level of error in the inventory cost.
- the inventory costs of the inventory items generated by the sales agent are calculated based on the estimated data of the inventory-related information. can do. As a result, it is possible to suppress deterioration in the calculation accuracy of the inventory cost. Further, according to the present disclosure, it is possible to output information (error level) indicating the accuracy of the inventory cost calculated by the inventory cost calculation unit 1152 . This makes it possible to grasp the level of error included in the calculated inventory costs.
- the sales agent are included in the inventory cost calculation, and the difference in the amount of information disclosed to the factory due to the difference in the type of sales agent is also taken into consideration. It is possible to calculate the cost of inventory. As a result, it is possible to prevent excess or shortage of inventory at the distributor.
- the error estimating unit 1153 determines the inventory cost from the ratio of the number of estimated data to the total number of actual data and estimated data included in the inventory-related information of the distributor. We estimated the level of error involved.
- the level of error included in the distributor's inventory costs is estimated using a learned model obtained by machine-learning the distributor's inventory-related information. Specifically, the error estimation unit 1153 determines the ratio of estimated data included in the distributor's inventory-related information acquired by the inventory-related information acquisition unit 1151, the error level included in the distributor's inventory costs, and, using a trained model generated by machine learning of the combination of location information about the distributor, to estimate the level of error included in the distributor's inventory costs.
- the base information includes, for example, at least one of location information including the location of the sales agent, the size of the sales agent, and the number of items shipped to the sales agent. Also, the base information may include the number of stores of the distributor, the capital, the number of workers, the number of equipment used, and the like. In this way, by estimating the level of error included in the sales agent's inventory costs using a machine-learned model that includes the sales agent's base information, for example, with a certain sales agent It is possible to infer the level of error of a distributor based on the level of error disclosed by another nearby distributor. This increases the accuracy of estimating the level of error and reduces the error in inventory costs.
- FIG. 24 is a configuration diagram of a learning device 200 related to the inventory cost calculation device 100.
- a trained model is generated by the learning device 200 .
- the learning device 200 includes a data acquisition unit 201 , a model generation unit 202 , and a learned model storage unit 203 .
- the learning device 200 may be connected to the inventory cost calculation device 100 via a network and may be a separate device from the inventory cost calculation device 100 .
- the learning device 200 may be incorporated in the inventory cost calculation device 100 .
- learning device 200 may reside on a cloud server.
- the data acquisition unit 201 obtains from the distributor inventory data storage unit 104 of the inventory cost calculation device 100 the ratio of estimated data included in the distributor's inventory-related information and the error included in the distributor's inventory inventory cost. level and base information on sales agents as learning data.
- the model generation unit 202 calculates the ratio of estimated data included in the distributor's inventory-related information output from the data acquisition unit 201, the level of error included in the inventory cost of the distributor's inventory, and the sales distributor's Learning the level of error included in the inventory costs of the sales agent according to the learning data created based on the combination of the location information about the store.
- the ratio of guess data included in the distributor's inventory-related information, the level of error included in the inventory cost of the distributor's inventory, and the location information on the distributor are combined to generate a trained model that infers the level of error contained in the inventory costs of
- the training data is the percentage of estimated data included in the distributor's inventory-related information, the error level included in the inventory cost of the distributor's inventory, and the base information on the distributor. It is associated data.
- supervised learning refers to a method of inferring a result from an input by giving a set of input and result data to a learning device to learn features in the learning data.
- a neural network consists of an input layer made up of multiple neurons, an intermediate layer (hidden layer) made up of multiple neurons, and an output layer made up of multiple neurons.
- the intermediate layer may be one layer, or two or more layers.
- the neural network is obtained by the data acquisition unit 201, the ratio of the estimated data included in the distributor's inventory-related information, the level of error included in the inventory cost of the distributor's inventory, and , the level of error included in the inventory costs of the sales agents is learned by so-called supervised learning according to learning data created based on a combination of base information on the sales agents.
- the neural network inputs the ratio of estimated data included in the distributor's inventory-related information and the base information about the distributor in the input layer, and the result output from the output layer is the distributor's inventory. It learns by adjusting the weights W1 and W2 so as to approach the level of error included in the item's inventory cost.
- the model generation unit 202 generates and outputs a learned model by executing the learning as described above.
- the learned model storage unit 203 stores the learned model output from the model generation unit 202 .
- FIG. 26 is a flow chart relating to learning processing of the learning device 200 .
- the data acquisition unit 201 obtains the ratio of estimated data included in the distributor's inventory-related information from the distributor's inventory data storage unit 104 of the inventory cost calculation device 100, and the inventory cost of the distributor's inventory. Get the level of error involved and location information about the distributor. In addition, the ratio of estimated data included in the distributor's inventory-related information, the level of error included in the inventory cost of the distributor's inventory, and the base information regarding the distributor were to be obtained at the same time. , each data may be input in association with each other, and each data may be obtained at different timings.
- the model generation unit 202 determines the ratio of estimated data included in the distributor's inventory-related information acquired by the data acquisition unit 201, the error level included in the inventory cost of the distributor's inventory, And, according to the learning data created based on the combination of the base information about the sales agent, by so-called supervised learning, learn the level of error included in the inventory cost of the sales agent, and generate a trained model. .
- the learned model storage unit 203 stores the learned model generated by the model generation unit 202.
- FIG. 27 is a diagram showing functions realized by distributor inventory cost calculation unit 115 according to the second embodiment. Only points different from the first embodiment will be described.
- the inventory-related information acquisition unit 1151 obtains actual data for calculating the inventory cost of inventory items generated at the sales agent from the sales agent inventory data storage unit 104, and to estimate the inventory cost of a certain inventory item. Get inventory-related information, including inferred data for In addition, the inventory-related information acquisition unit 1151 acquires from the sales agent inventory data storage unit 104 the ratio of estimated data included in the sales agent's inventory-related information and the base information regarding the sales agent.
- the error estimating unit 1153 infers the level of error included in the sales agent's inventory costs obtained using the trained model. In other words, by inputting the ratio of estimated data included in the inventory-related information of sales agents acquired by the inventory-related information acquisition unit 1151 into this learned model, and the location information related to sales agents, The level of error included in the inferred distributor inventory cost can be output.
- the learned model learned by the model generation unit 202 is used to output the level of error included in the inventory costs of the sales agent. and output the level of error included in the distributor's inventory costs using this learned model.
- the inventory-related information acquisition unit 1151 acquires the ratio of estimated data included in the inventory-related information of the sales agent and the base information of the sales agent.
- step c2 the error estimating unit 1153 inputs the ratio of estimated data included in the inventory-related information of the sales agent and the base information regarding the sales agent to the learned model stored in the learned model storage unit 203. do.
- the error estimator 1153 acquires the level of error included in the inventory costs of the distributor obtained from the learned model. Further, the error estimation unit 1153 outputs the acquired error level to the total inventory cost calculation unit 116 .
- the total inventory cost calculation unit 116 calculates the error in the inventory cost of the sales agent as a monetary amount using the error level included in the output inventory cost of the sales agent. Specifically, the total inventory cost calculation unit 116 multiplies the inventory cost of the distributor by a ratio corresponding to the level of error in the inventory cost of the distributor, thereby calculating the error in the inventory cost of each distributor. Calculate as an amount. As a result, it is possible to calculate the error in the inventory cost of the distributor as a monetary value, taking into consideration not only the ratio of the estimated data but also the base information.
- the error estimating unit 1153 calculates the ratio of estimated data included in the distributor's inventory-related information, the level of error included in the distributor's inventory costs, and the location of the distributor.
- a trained model generated by combinatorial machine learning was used to estimate the level of error included in the distributor's inventory costs.
- the error estimation unit 1153 is a learned model generated by machine learning of a combination of the ratio of estimated data included in the distributor's inventory-related information and the level of error included in the distributor's inventory costs. may be used to estimate the level of error included in the distributor's inventory costs.
- the learning device 200 generates a learned model through machine learning of a combination of the ratio of estimated data included in the distributor's inventory-related information and the level of error included in the distributor's inventory costs. Just do it.
- supervised learning is applied to the learning algorithm used by model generation unit 202
- the present invention is not limited to this.
- reinforcement learning unsupervised learning, semi-supervised learning, and the like as learning algorithms.
- model generation unit 202 As a learning algorithm used in the model generation unit 202, deep learning that learns to extract the feature amount itself can be used, and other known methods such as genetic programming, functional logic programming, support vector Machine learning may be performed according to a machine or the like.
- the error in the inventory cost calculated is assumed to be "A (eg, 40 (%))"
- the actual data 4 Assume that the error in the inventory cost calculated when all three are assumed data is "B (for example, 80 (%))”.
- a trained model may be generated by machine learning in which "A” is used as correct data and learning is performed so as to increase the calculation accuracy of "B".
- the level of error included in the inventory costs of a sales agent using a machine-learned machine-learned model that includes the base information of the sales agent, for example, a certain sales It is possible to infer the level of error of a distributor based on the level of error disclosed by another distributor that is located close to the distributor. This increases the accuracy of estimating the level of error and reduces the error in inventory costs.
- the estimating unit 119 obtains estimated data obtained by estimating actual data of inventory-related information not disclosed by a certain sales agent, which is included in inventory-related information disclosed by another sales agent. Generated based on real data.
- the estimating unit 119 performs learning generated by machine learning of a combination of actual data included in inventory-related information disclosed by a certain sales agent and location information about the certain sales agent. Presumed data obtained by estimating actual data of inventory-related information that is not disclosed by another sales agent may be output using the completed model. That is, the estimating unit 119 is a learned model generated by machine learning of a combination of actual data for calculating the inventory cost of inventory items generated at a certain sales agent and location information related to the certain sales agent. may be used to output estimated data for estimating inventory costs incurred at different distributors.
- a mode of generating a trained model by machine learning will be described with reference to FIG. 23 .
- a trained model is generated by the learning device 200 .
- the learning device 200 includes a data acquisition unit 201 , a model generation unit 202 , and a learned model storage unit 203 .
- the learning device 200 may be connected to the inventory cost calculation device 100 via a network and may be a separate device from the inventory cost calculation device 100 .
- the learning device 200 may be incorporated in the inventory cost calculation device 100 .
- learning device 200 may reside on a cloud server.
- the data acquisition unit 201 retrieves a combination of the actual data included in the inventory-related information of a certain sales agent from the sales agent inventory data storage unit 104 of the inventory cost calculation device 100 and the location information related to the certain sales agent. Acquire as training data.
- the model generation unit 202 is a learning model created based on a combination of the actual data included in the inventory-related information of a certain sales agent output from the data acquisition unit 201 and the location information related to the certain sales agent. According to the data, it learns inferred data that infers the actual data of the inventory-related information of another sales agent. In other words, inference data obtained by estimating the actual data of the optimal inventory-related information of another sales agent from the combination of the actual data included in the inventory-related information of the sales agent and the base information of a certain sales agent. Generate a trained model that
- the learning data is data in which the actual data included in the inventory-related information of a certain sales agent and the base information related to the certain sales agent are associated with each other.
- supervised learning refers to a method of inferring a result from an input by giving a set of input and result data to a learning device to learn features in the learning data.
- the actual data included in the inventory-related information of a certain sales agent and the base information of a certain sales agent are input to the input layer, and the result output from the output layer is the sales agent's Learning is performed by adjusting the weights W1 and W2 so as to approximate the actual data.
- the model generation unit 202 generates and outputs a learned model by executing the learning as described above.
- the learned model storage unit 203 stores the learned model output from the model generation unit 202 .
- FIG. 29 is a configuration diagram of the inventory cost calculation device 100 according to the third embodiment. Only points different from the first embodiment will be described.
- the estimation unit 119 acquires from the sales agent inventory data storage unit 104 the actual data included in the sales agent's inventory-related information and the base information regarding a certain sales agent.
- the estimation unit 119 infers estimated data obtained by estimating the actual data of the inventory-related information of the distributor obtained using the learned model. In other words, by inputting the actual data included in the inventory-related information of the sales agent and the base information about a certain sales agent into this trained model, the inventory of the sales agent inferred from the input information Estimated data obtained by estimating the actual data of the related information can be output.
- the trained model trained by the model generation unit 202 is used to output estimated data obtained by estimating the actual data of the inventory-related information of the distributor.
- a model may be acquired, and estimated data obtained by estimating the actual data of the sales agent's inventory-related information using this learned model may be output.
- the inventory of a certain sales agent is calculated. It is possible to generate estimated data (for example, the number of deliveries) included in the related information. Further, according to the present embodiment, by generating the estimated data using the learned model, it is possible to generate the estimated data closer to the actual data. Therefore, according to the present embodiment, even if information for calculating the inventory costs generated at each site is not disclosed, it is possible to suppress a decrease in accuracy in calculating the inventory costs.
- 100 inventory cost calculation device 101 factory inventory data input unit, 102 factory inventory data storage unit, 103 sales agent inventory data input unit, 104 sales agent inventory data storage unit, 105 control unit, 106 time data, 107 quantity data, 108 unit cost data, 109 distributor type data, 110 location data, 111 time data, 112 quantity data, 113 unit cost data, 114 factory inventory cost calculation unit, 115 sales agent inventory cost calculation unit, 116 overall Inventory cost calculation unit 119 Estimation unit 1151 Inventory related information acquisition unit 1152 Inventory cost calculation unit 1153 Error estimation unit 1141 Inventory related information acquisition unit 1142 Inventory cost calculation unit
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)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
Description
図1は実施の形態1に係る在庫費算出装置100の構成図である。
在庫費算出装置100は、工場内在庫データ入力部101、工場内在庫データ記憶部102、販売代理店在庫データ入力部103、販売代理店在庫データ記憶部104、制御部105により構成される。
ここで、例えば、自社拠点の販売代理店は、入庫数、出庫数、滞留時間、及び、労務費に対応した在庫関連情報の実データを製造業者に対して開示したものとする。また、他社拠点の販売代理店は、入庫数に対応した在庫関連情報の実データを製造業者に対して開示したものとする。製造業者に対して開示されたデータは、在庫費算出装置100の工場内在庫データ記憶部102、又は、販売代理店在庫データ記憶部104に反映される。図2に示すように、販売代理店は、在庫関連情報の一部の実データのみを製造業者に対して開示することが多く、全ての実データを製造業者に対して開示することが少ない。また、販売代理店は、自社拠点又は他社拠点という種類によっても、製造業者に対して開示する在庫関連情報の情報量が異なる。一般的に、自社拠点の販売代理店の方が、他社拠点の販売代理店よりも、製造業者に対して開示する在庫関連情報の情報量が多い。
在庫関連情報取得部1151は、販売代理店在庫データ記憶部104から、販売代理店で発生する在庫品の在庫費を算出するための実データ、及び、或る在庫品の在庫費を推測するための推測データを含む在庫関連情報を取得する。
在庫関連情報取得部1141は、工場内在庫データ記憶部102から、工場で発生する在庫品の在庫費を算出するための在庫関連情報を取得する。即ち、在庫関連情報取得部1141は、拠点で発生する在庫品の在庫費を算出するための在庫関連情報を取得する在庫関連情報取得部の一例である。在庫費算出部1142は、在庫関連情報取得部1141で取得された在庫関連情報に基づいて在庫品の在庫費を算出する。また、在庫費算出部1142は、時間に相関する在庫費、及び、時間に相関しない在庫費に分けて算出する。また、在庫費算出部1142は、算出された在庫品の在庫費を全体在庫費算出部116に出力する。なお、工場が複数存在する場合、工場内在庫費算出部114は、各工場で発生する在庫費を算出して、出力する。ところで、工場内在庫費算出部114の機能は、販売代理店在庫費算出部115の機能で実現してもよい。具体的には、工場内在庫費算出部114の在庫関連情報取得部1141の機能は、販売代理店在庫費算出部115の在庫関連情報取得部1151の機能で実現してもよい。また、工場内在庫費算出部114の在庫費算出部1142の機能は、販売代理店在庫費算出部115の在庫費算出部1152の機能で実現してもよい。
CPU401は、主記憶に格納されたプログラムなどを実行することにより、図1に示す、工場内在庫費算出部114、推測部119、販売代理店在庫費算出部115、及び、全体在庫費算出部116という各機能を実現する。主記憶402は、例えば、不揮発性のメモリであり、CPU401に実行される各種プログラムを記憶する。インターフェース404は、外部の装置とデータを入出力するためのインターフェースである。インターフェース404は、例えば、ネットワークを介して、工場や販売代理店から送信された情報を取得する。記憶部405は、例えば、HDDであり、工場内在庫データ記憶部102及び販売代理店在庫データ記憶部104を有する。CPU401に処理される各種データを記憶する。記憶部405は、記憶されたデータをCPU401へと転送する。
ステップS1において、工場内在庫費算出部114は、複数ある工場の中から最初に算出対象とする工場を選択する。ステップS2において、工場内在庫費算出部114は、ステップS2で選択された工場に保管された複数ある在庫品の在庫IDの中から最初に算出対象とする在庫IDを選択する。ステップS3において、工場内在庫費算出部114(在庫関連情報取得部1141)は、ステップS2で選択した在庫IDの在庫品の在庫費を算出するための実データ(時間データ、費用単価データ)を含む在庫関連情報を工場内在庫データ記憶部102から取得する。ステップS4において、工場内在庫費算出部114(在庫費算出部1142)は、ステップS3で取得した在庫関連情報に含まれる実データ(時間データ、費用単価データ)に基づいて各在庫IDの在庫品の在庫費を算出する。ステップS5において、工場内在庫費算出部114は、在庫費算出未実施の在庫IDがあるかどうかを判定し、Yesの場合はステップS1に戻って、在庫費算出未実施の在庫IDについてステップS1、S2、S3、S4を行う。Noの場合はステップS6に進む。ステップS6において、工場内在庫費算出部114は、各在庫IDの在庫費を足し合わせて、時間と相関関係を有する費用を算出する。以上、ステップS1からステップS6までが、時間と相関関係を有する費用の算出フローである。
ステップS101において、販売代理店在庫費算出部115は、複数ある販売代理店の中から最初に算出対象とする販売代理店を選択する。ステップS102において、販売代理店在庫費算出部115(在庫関連情報取得部1151)は、ステップS101で選択された販売代理店の在庫品の在庫費を算出するための実データ及び在庫品の在庫費を推測するための推測データ(時間データ、費用単価データ)を含む在庫関連情報を販売代理店在庫データ記憶部104から取得する。ステップS103において、販売代理店在庫費算出部115(在庫費算出部1152)は、ステップS102で取得した在庫関連情報に含まれる実データ及び推測データ(時間データ、費用単価データ)に基づいて在庫品の在庫費を算出する。ステップS104において、販売代理店在庫費算出部115は、ステップS103で算出した在庫費を足し合わせて、時間と相関関係を有する費用を算出する。以上ステップS102からステップS104までが、時間と相関関係を有する費用の算出フローである。
ステップS201において、全体在庫費算出部116は、工場内在庫費算出部114の在庫費算出部1142で算出された各工場の在庫費を取得する。
図14、15は、工場1から開示された在庫関連情報を示す図である。図16、17は、工場2から開示された在庫関連情報を示す図である。工場1のデータは図14と図15、工場2のデータは図16と図17を用いることとする。図11のステップS1において、工場内在庫費算出部114は、工場1を選択したとする。ステップS2において、工場内在庫費算出部114は、在庫ID101を選択したとする。ステップS3において、工場内在庫費算出部114は、図14の「在庫ID」が「101」であるデータ3行を取得する。ステップS4において、工場内在庫費算出部114は、1行目のデータから3(時間)×5(円/日)=15(円)、2行目のデータから70(秒)×10(円/秒)=700(円)、3行目のデータから2(日)×10(円/日)=20(円)と計算する。ステップS5において、工場内在庫費算出部114は、在庫費算出未実施の在庫ID「102」が存在するため、ステップS1に戻る。
図18、19は、販売代理店1の在庫関連情報を示す図である。図20、21は、販売代理店2の在庫関連情報を示す図である。図22、23は、販売代理店3の在庫関連情報を示す図である。販売代理店1(自社拠点:工場のグループ会社の販売代理店)のデータは図18と図19、販売代理店2(他社拠点:独立系の販売代理店)のデータは図20と図21、販売代理店3(他社拠点:独立系の販売代理店)のデータは図22と図23を用いることとする。図12のステップS101において、販売代理店在庫費算出部115は、販売代理店1を選択したとする。ステップS102において、販売代理店在庫費算出部115は、図18のデータ2行を取得する。ステップS103において、販売代理店在庫費算出部115は、1行目のデータから20(時間)×20(円/時間)=400(円)、20(時間)×5(円/時間)=100(円)と計算する。ステップS104において、販売代理店在庫費算出部115は、ステップS103で算出した400(円)と100(円)を足し合わせて、時間と相関関係を有する費用は400(円)+100(円)=500(円)と算出する。
ステップS201において、全体在庫費算出部116は、工場1の在庫費として1791(円)、工場2の在庫費として1502(円)を取得する。ステップS202において、全体在庫費算出部116は、販売代理店1の在庫費として1800(円)、在庫費の誤差のレベルとして0(%)、販売代理店2の在庫費として1055(円)、在庫費の誤差のレベルとして40(%)、販売代理店3の在庫費として1860(円)、在庫費の誤差のレベルとして40(%)を取得する。ステップS203において、全体在庫費算出部116は、販売代理店1の在庫費の誤差は1800(円)×0(%)=0(円)、販売代理店2の在庫費の誤差は1055(円)×40(%)=422(円)、販売代理店3の在庫費の誤差は1860(円)×40(%)=744(円)と算出する。ステップS204において、全体在庫費算出部116は、各工場の在庫費と、各販売代理店の在庫費を足し合わせて、1791(円)+1502(円)+1800(円)+1055(円)+1860(円)=8008(円)と出力する。ステップS205において、全体在庫費算出部116は、各販売代理店の在庫費の誤差を足し合わせて、0(円)+422(円)+744(円)=1166(円)と出力する。なお、全体在庫費算出部116は、各工場の在庫費、各販売代理店の在庫費、及び、各販売代理店の在庫費の誤差を夫々足し合わせずに出力してもよい。
以上の実施の形態1では、誤差推測部1153は、販売代理店の在庫関連情報に含まれる実データの数と推測データの数の合計数の内、推測データの数が占める割合から在庫費に含まれた誤差のレベルを推測した。実施の形態2では、販売代理店の在庫費に含まれた誤差のレベルを、販売代理店の在庫関連情報を機械学習した学習済モデルを用いて推測する。具体的には、誤差推測部1153は、在庫関連情報取得部1151で取得された販売代理店の在庫関連情報に含まれる推測データの割合、販売代理店の在庫費に含まれた誤差のレベル、及び、販売代理店に関する拠点情報の組み合わせの機械学習によって生成された学習済みモデルを用いて、販売代理店の在庫費に含まれた誤差のレベルを推測する。拠点情報は、例えば、販売代理店の立地を含む位置情報、販売代理店の規模、販売代理店に対して在庫品を出庫した数のうちの少なくとも一つを含む。また、拠点情報は、販売代理店の店舗数、資本金、作業者数、使用設備台数などを含んでいてもよい。このように、販売代理店の拠点情報を含めて機械学習した学習済モデルを用いて、販売代理店の在庫費に含まれた誤差のレベルを推測することにより、例えば、或る販売代理店と立地の近い別の販売代理店から開示された誤差のレベルに基づいて、或る販売代理店の誤差のレベルを推測することが可能となる。これにより、誤差のレベルの推測の精度が高まり、在庫費の誤差を低減することができる。
図24は、在庫費算出装置100に関する学習装置200の構成図である。
学習済みモデルは、学習装置200により生成される。学習装置200は、データ取得部201、モデル生成部202、及び、学習済モデル記憶部203を備える。学習装置200は、例えば、ネットワークを介して在庫費算出装置100に接続され、この在庫費算出装置100とは別個の装置であってもよい。また、学習装置200は、在庫費算出装置100に内蔵されていてもよい。さらに、学習装置200は、クラウドサーバ上に存在していてもよい。
図27は実施の形態2に係る販売代理店在庫費算出部115で実現される機能を示す図である。実施の形態1と相違する点のみ説明する。
実施の形態1において、推測部119は、或る販売代理店から開示されていない在庫関連情報の実データを推測した推測データを、別の販売代理店から開示された在庫関連情報に含まれた実データに基づいて生成した。実施の形態3において、推測部119は、或る販売代理店から開示された在庫関連情報に含まれた実データ、及び、或る販売代理店に関する拠点情報の組み合わせの機械学習によって生成された学習済みモデルを用いて、別の販売代理店から開示されていない在庫関連情報の実データを推測した推測データを出力してもよい。即ち、推測部119は、或る販売代理店で発生する在庫品の在庫費を算出するための実データ、及び、或る販売代理店に関する拠点情報の組み合わせの機械学習によって生成された学習済みモデルを用いて、別の販売代理店で発生する在庫品の在庫費を推定するための推定データを出力してもよい。
図29は実施の形態3に係る在庫費算出装置100の構成図である。実施の形態1と相違する点のみ説明する。
Claims (7)
- 製造業のサプライチェーンにおける工場及び販売代理店を含む各拠点で発生する在庫品の在庫費を算出する在庫費算出装置であって、
拠点で発生する在庫品の在庫費を算出するための実データ及び当該在庫品の在庫費を推測するための推測データを含む在庫関連情報を取得する在庫関連情報取得部と、
前記在庫関連情報取得部で取得された前記在庫関連情報に含まれる実データ及び推測データに基づいて前記在庫品の在庫費を算出する在庫費算出部と、
前記在庫関連情報取得部で取得された前記在庫関連情報に含まれる推測データの割合に応じて、前記在庫費算出部で算出された前記在庫品の在庫費に含まれた誤差のレベルを推測する誤差推測部と、
前記在庫費算出部で算出された前記在庫品の在庫費、及び、前記誤差推測部で推測された前記在庫品の在庫費に含まれた誤差のレベルに基づいて算出された前記在庫品の在庫費に含まれた誤差と、前記在庫費算出部で算出された前記在庫品の在庫費とを出力する出力部
を備えることを特徴とする在庫費算出装置。 - 前記誤差推測部は、
前記在庫関連情報に含まれる推測データの割合、及び、前記在庫品の在庫費に含まれた誤差のレベルの組み合わせの機械学習によって生成された学習済みモデルを用いて、前記販売代理店で発生する前記在庫品の在庫費に含まれた誤差のレベルを推測する
ことを特徴とする請求項1に記載の在庫費算出装置。 - 前記誤差推測部は、
前記在庫関連情報に含まれる推測データの割合、前記在庫品の在庫費に含まれた誤差のレベル、及び、前記販売代理店に関する拠点情報の組み合わせの機械学習によって生成された学習済みモデルを用いて、前記販売代理店で発生する前記在庫品の在庫費に含まれた誤差のレベルを推測する
ことを特徴とする請求項1に記載の在庫費算出装置。 - 或る販売代理店で発生する在庫品の在庫費を算出するための実データに基づいて、別の販売代理店で発生する在庫品の在庫費を推測するための推測データを出力する推定部を備え、
前記推定部は、
前記或る販売代理店で発生する在庫品の在庫費を算出するための実データ、及び、当該或る販売代理店に関する拠点情報の組み合わせの機械学習によって生成された学習済みモデルを用いて、前記別の販売代理店で発生する在庫品の在庫費を推定するための推定データを出力する
ことを特徴とする請求項1~3の何れか1項に記載の在庫費算出装置。 - 前記拠点情報は、
前記販売代理店の位置情報、前記販売代理店の規模、前記販売代理店の設備、前記販売代理店に対して在庫品を出庫した数のうちの少なくとも一つを含むこと
を特徴とする請求項3又は4に記載の在庫費算出装置。 - 前記在庫費算出部は、
時間と相関関係を有する在庫費、及び、時間と相関関係を有さない在庫費に分けて算出する
ことを特徴とする請求項1~5の何れか1項に記載された在庫費算出装置。 - 製造業のサプライチェーンにおける工場及び販売代理店を含む各拠点で発生する在庫品の在庫費を算出する在庫費算出方法であって、
拠点で発生する在庫品の在庫費を算出するための実データ及び当該在庫品の在庫費を推測するための推測データを含む在庫関連情報を取得する在庫関連情報取得ステップと、
前記在庫関連情報取得ステップで取得された前記在庫関連情報に含まれる実データ及び推測データに基づいて前記在庫品の在庫費を算出する在庫費算出ステップと、
前記在庫関連情報取得ステップで取得された前記在庫関連情報に含まれる推測データの割合に応じて、前記在庫費算出ステップで算出された前記在庫品の在庫費に含まれた誤差のレベルを推測する誤差推測ステップと、
前記在庫費算出ステップで算出された前記在庫品の在庫費、及び、前記誤差推測ステップで推測された前記在庫品の在庫費に含まれた誤差のレベルに基づいて算出された前記在庫品の在庫費に含まれた誤差と、前記在庫費算出ステップで算出された前記在庫品の在庫費とを出力する出力ステップ
を備えることを特徴とする在庫費算出方法。
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202180089699.1A CN116685990A (zh) | 2021-01-25 | 2021-01-25 | 库存费计算装置和库存费计算方法 |
PCT/JP2021/002381 WO2022157957A1 (ja) | 2021-01-25 | 2021-01-25 | 在庫費算出装置、及び、在庫費算出方法 |
JP2022553065A JP7237250B2 (ja) | 2021-01-25 | 2021-01-25 | 在庫費算出装置、及び、在庫費算出方法 |
TW110126586A TW202230236A (zh) | 2021-01-25 | 2021-07-20 | 庫存成本計算裝置以及庫存成本計算方法 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2021/002381 WO2022157957A1 (ja) | 2021-01-25 | 2021-01-25 | 在庫費算出装置、及び、在庫費算出方法 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022157957A1 true WO2022157957A1 (ja) | 2022-07-28 |
Family
ID=82548625
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2021/002381 WO2022157957A1 (ja) | 2021-01-25 | 2021-01-25 | 在庫費算出装置、及び、在庫費算出方法 |
Country Status (4)
Country | Link |
---|---|
JP (1) | JP7237250B2 (ja) |
CN (1) | CN116685990A (ja) |
TW (1) | TW202230236A (ja) |
WO (1) | WO2022157957A1 (ja) |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2009104408A (ja) * | 2007-10-23 | 2009-05-14 | Ntt Data Corp | 統合需要予測装置、統合需要予測方法、及び統合需要予測プログラム |
JP2017072870A (ja) * | 2015-10-05 | 2017-04-13 | 三菱電機インフォメーションシステムズ株式会社 | 損益予測装置及び損益予測プログラム |
JP2018020885A (ja) * | 2016-08-03 | 2018-02-08 | 富士通株式会社 | 輸配送計画プログラム、輸配送計画方法及び輸配送計画装置 |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016113844A1 (ja) | 2015-01-13 | 2016-07-21 | 株式会社日立製作所 | サプライチェーン設計装置、サプライチェーン設計方法およびプログラム |
-
2021
- 2021-01-25 JP JP2022553065A patent/JP7237250B2/ja active Active
- 2021-01-25 CN CN202180089699.1A patent/CN116685990A/zh active Pending
- 2021-01-25 WO PCT/JP2021/002381 patent/WO2022157957A1/ja active Application Filing
- 2021-07-20 TW TW110126586A patent/TW202230236A/zh unknown
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2009104408A (ja) * | 2007-10-23 | 2009-05-14 | Ntt Data Corp | 統合需要予測装置、統合需要予測方法、及び統合需要予測プログラム |
JP2017072870A (ja) * | 2015-10-05 | 2017-04-13 | 三菱電機インフォメーションシステムズ株式会社 | 損益予測装置及び損益予測プログラム |
JP2018020885A (ja) * | 2016-08-03 | 2018-02-08 | 富士通株式会社 | 輸配送計画プログラム、輸配送計画方法及び輸配送計画装置 |
Also Published As
Publication number | Publication date |
---|---|
JP7237250B2 (ja) | 2023-03-10 |
JPWO2022157957A1 (ja) | 2022-07-28 |
TW202230236A (zh) | 2022-08-01 |
CN116685990A (zh) | 2023-09-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Zhang | Procurement mechanism design in a two-echelon inventory system with price-sensitive demand | |
Chen et al. | Supply chain coordination with multiple shipments: The optimal inventory subsidizing contracts | |
Khalili-Damghani et al. | A fuzzy bi-objective mixed-integer programming method for solving supply chain network design problems under ambiguous and vague conditions | |
Yang et al. | Coordinating inventory control and pricing strategies under batch ordering | |
Avci | Lateral transshipment and expedited shipping in disruption recovery: A mean-CVaR approach | |
Birim et al. | Evaluating vendor managed inventory systems: how incentives can benefit supply chain partners | |
Saha et al. | Multi-item fuzzy-stochastic supply chain models for long-term contracts with a profit sharing scheme | |
Solgi et al. | Resilient supplier selection in complex products and their subsystem supply chains under uncertainty and risk disruption: A case study for satellite components | |
Pourakbar et al. | A joint economic lot-size model for an integrated supply network using genetic algorithm | |
Eksler et al. | An EOQ model for multiple products with varying degrees of substitutability | |
JP7237250B2 (ja) | 在庫費算出装置、及び、在庫費算出方法 | |
Priyan et al. | Economic design of multi-echelon inventory system with variable lead time and service level constraint in a fuzzy cost environment | |
Hachaïchi et al. | A policy gradient based reinforcement learning method for supply chain management | |
Routroy et al. | Study of one way product substitution on supply chain performance | |
CN111768157A (zh) | 动态物料计算方法及其系统 | |
Routroy et al. | Inventory planning for a multi-echelon supply chain | |
Koren et al. | Inventory management for stockout-based substitutable products under centralised and competitive settings | |
Li | The dynamic behaviors of a supply chain with stock-dependent demand considering competition and deteriorating items | |
Herrin | How to calculate safety stocks for highly seasonal products | |
Ghose et al. | Dynamic pricing in electronic commerce using neural network | |
Choi | Supply chain information sharing and the impact of information errors | |
Cheng | Dynamic pricing decision for perishable goods: a Q-learning approach | |
Teng et al. | Linking tactical and operational decision-making to strengthen textile/apparel supply chains | |
Eksler et al. | An EOQ Model for multiple products with varying degrees of substitutability. Economics Discussion Papers, No 2018-77 | |
Goonatilake et al. | Essentials of Novel Inventory Management Systems |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
ENP | Entry into the national phase |
Ref document number: 2022553065 Country of ref document: JP Kind code of ref document: A |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21921065 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 202180089699.1 Country of ref document: CN |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 21921065 Country of ref document: EP Kind code of ref document: A1 |