Demand prediction device, demand prediction method, and demand prediction program

Abstract

A future inventory prediction unit (22) estimates a future inventory quantity at a supply destination of an article that is subject to prediction under the assumption that, at the supply destination, inventory increases due to orders being placed in accordance with an order placement condition and order quantity indicated by order placement model information (61), and inventory decreases in accordance with a transition of a past inventory quantity indicated by past inventory information (62). A demand prediction unit (23) predicts that demand for the order quantity indicated by the order placement model information (61) occurs at a supplier at the point in time when the order placement condition indicated by the order placement model information (61) is satisfied in the future inventory quantity estimated by the future inventory estimation unit (22).

Description

Demand forecasting device, demand forecasting method, and demand forecasting program

[0001] The present disclosure relates to technology for forecasting demand.

[0002] In supply chains in industries such as manufacturing or agriculture, a series of commercial transactions, from the procurement of raw materials to the sale of final products, are carried out by multiple businesses. In these supply chains, methods for predicting demand for products are being considered with the aim of reducing inventory management costs. Here, forecasting demand means predicting future order quantities from product suppliers. Hereinafter, the term "demand forecasting" will be used to refer to forecasting demand.

[0003] In general demand forecasting techniques, predictions are made by learning time-series data such as past order records. Patent Literature 1 describes a technique for improving the accuracy of demand forecasting by learning past demand generation trends from order records and adjusting the demand forecasting method.

[0004] Japanese Patent Application Laid-Open No. 2022-101879

[0005] The method of making a forecast by estimating long-term trends in demand, such as seasonality or periodicity, from order receipt data, as described in Patent Literature 1, is unable to make a demand forecast that follows small daily fluctuations in demand. The present disclosure aims to make it possible to make a demand forecast that follows small daily fluctuations in demand.

[0006] The demand forecasting device according to the present disclosure includes a future inventory estimation unit that estimates future inventory levels at the supply destination of the forecast target item, assuming that inventory will increase as orders are placed in accordance with the ordering conditions and order amounts indicated in the order model information, and that inventory will decrease in accordance with the trends in past inventory levels indicated by the past inventory information, and a demand forecasting unit that predicts that demand for the order amount indicated by the order model information will arise from the supplier at the time when the future inventory level estimated by the future inventory estimation unit satisfies the ordering conditions indicated by the order model information.

[0007] In this disclosure, future inventory levels are estimated using past inventory level trends, and demand is predicted from the future inventory levels. This makes it possible to perform demand forecasts that track small daily demand fluctuations by taking into account recent inventory level fluctuations.

[0008] 1 is an explanatory diagram of a target of demand forecasting according to the first embodiment. A configuration diagram of a demand forecasting system 100 according to the first embodiment. A configuration diagram of a demand forecasting device 10 according to the first embodiment. A flowchart of input reception processing according to the first embodiment. An explanatory diagram of order model information 61 according to the first embodiment. An explanatory diagram of past inventory information 62 according to the first embodiment. A flowchart of demand forecasting processing according to the first embodiment. An explanatory diagram of future inventory estimation processing according to the first embodiment. An explanatory diagram of a demand forecast result 64 according to the first embodiment. A configuration diagram of a demand forecasting device 10 according to a first modification. A configuration diagram of a demand forecasting device 10 according to the second embodiment. An explanatory diagram of commercial distribution structure information 65 according to the second embodiment. A flowchart of demand forecasting processing according to the second embodiment. An explanatory diagram of future inventory estimation processing according to the second embodiment. An explanatory diagram of a target of demand forecasting according to the second embodiment. A configuration diagram of a demand forecasting device 10 according to the third embodiment. An explanatory diagram of logistics information 66 according to the third embodiment. A flowchart of demand forecasting processing according to the third embodiment. An explanatory diagram of past inventory estimation processing according to the third embodiment. A configuration diagram of a demand forecasting device 10 according to the fourth embodiment. An explanatory diagram of parts configuration information 68 according to the fourth embodiment.

[0009] Embodiment 1. In embodiment 1, demand forecasting is performed independently for each business operator and each product. Therefore, in embodiment 1, as shown in FIG. 1 , the forecast target item, which is the item that is the subject of demand forecasting, is "Part A." An example will be described in which demand forecasting is performed for "Part A" that will arise from "Business A," the supply destination of "Part A," to "Business X," the supplier of "Part A." Here, when demand forecasting is performed for "Business B" or "Part B," it is possible to perform demand forecasting using the same processing as when demand forecasting for "Part A" that will arise from "Business A" to "Business X."

[0010] In the first embodiment, demand forecasting can be performed for each unit period, which is an arbitrary time unit such as a day or a week. In the following, an example will be described in which the unit period is one day and demand forecasting is performed on a daily basis.

[0011] ***Description of Configuration*** The configuration of the demand forecasting system 100 according to the first embodiment will be described with reference to Fig. 2. The demand forecasting system 100 includes a demand forecasting device 10, an input terminal 51, and an output terminal 52.

[0012] The demand forecasting device 10 is a computer such as a server. The demand forecasting device 10 receives order model information 61 and past inventory information 62 from an input terminal 51, performs demand forecasting, and outputs a demand forecast result 64 to an output terminal 52.

[0013] The input terminal 51 is a terminal for operating the demand forecasting device 10. Specific examples of the input terminal 51 include a PC, a tablet, or a smartphone. PC stands for personal computer. The demand forecasting device 10 may also serve as the input terminal 51. There may also be multiple input terminals 51.

[0014] The output terminal 52 is a terminal for outputting the processing contents of the demand forecasting device 10 to an inventory planner. Specific examples of the output terminal 52 include a PC, a tablet, a smartphone, etc. The input terminal 51 may also serve as the output terminal 52. There may also be multiple output terminals 52.

[0015] The configuration of the demand prediction device 10 according to the first embodiment will be described with reference to Fig. 3. The demand prediction device 10 is a computer. The demand prediction device 10 includes hardware such as a processor 11, a memory 12, a storage 13, and a communication interface 14. The processor 11 is connected to other hardware via signal lines and controls the other hardware.

[0016] The processor 11 is an IC that performs processing. IC stands for Integrated Circuit. Specific examples of the processor 11 include a CPU, a DSP, and a GPU. CPU stands for Central Processing Unit. DSP stands for Digital Signal Processor. GPU stands for Graphics Processing Unit.

[0017] The memory 12 is a storage device that temporarily stores data. Specific examples of the memory 12 include SRAM and DRAM. SRAM stands for Static Random Access Memory. DRAM stands for Dynamic Random Access Memory.

[0018] The storage 13 is a storage device that stores data. A specific example of the storage 13 is an HDD. HDD is an abbreviation for Hard Disk Drive. The storage 13 may also be a portable recording medium such as an SD (registered trademark) memory card, CompactFlash (registered trademark), NAND flash, a flexible disk, an optical disk, a compact disk, a Blu-ray (registered trademark) disk, or a DVD. SD is an abbreviation for Secure Digital. DVD is an abbreviation for Digital Versatile Disk.

[0019] The communication interface 14 is an interface for communicating with external devices. Specific examples of the communication interface 14 include Ethernet (registered trademark), USB, and HDMI (registered trademark) ports. USB stands for Universal Serial Bus. HDMI stands for High-Definition Multimedia Interface.

[0020] The demand prediction device 10 includes, as functional components, a reception unit 21, a future inventory estimation unit 22, a demand prediction unit 23, and an output unit 24. The functions of each functional component of the demand prediction device 10 are realized by software. The storage 13 stores a program that realizes the function of each functional component of the demand prediction device 10. This program is read into the memory 12 by the processor 11 and executed by the processor 11. In this way, the function of each functional component of the demand prediction device 10 is realized.

[0021] In the process of operation of the demand forecasting device 10, the storage 13 stores order model information 61, past inventory information 62, future inventory information 63, and demand forecast results 64.

[0022] 3 shows only one processor 11. However, there may be a plurality of processors 11, and the plurality of processors 11 may cooperate to execute programs that realize the respective functions.

[0023] ***Description of Operation*** The operation of the demand prediction device 10 according to the first embodiment will be described with reference to Figures 4 to 9. The operation procedure of the demand prediction device 10 according to the first embodiment corresponds to the demand prediction method according to the first embodiment. Furthermore, the program that realizes the operation of the demand prediction device 10 according to the first embodiment corresponds to the demand prediction program according to the first embodiment.

[0024] The operation of the demand prediction device 10 includes processing at the time of input reception and processing at the time of demand prediction.

[0025] The input reception process according to the first embodiment will be described with reference to Fig. 4. (Step S101: Reception process) The reception unit 21 receives input of order model information 61 and past inventory information 62 from the input terminal 51. However, the order model information 61 is only required the first time the input is received, and the reception unit 21 does not necessarily need to receive the order model information 61 every time the input reception process is executed.

[0026] Referring to FIG. 5 , the ordering model information 61 according to the first embodiment will be described. The ordering model information 61 includes one or more pairs of a business operator name, an ordering model, ordering conditions, an order quantity, and a procurement lead time. The ordering conditions are conditions for placing an order, and an order is placed when the ordering conditions are met. The order quantity is the amount to be ordered when the ordering conditions are met. The procurement lead time is the period from when a product order is placed until the product arrives and is managed as inventory. For example, assume that business operator A's ordering model is the reorder point method. In this case, information corresponding to the reorder point method includes an order point indicating the ordering conditions and the order quantity. Therefore, a set of information including business operator A, the reorder point method, the reorder point and order quantity, and the procurement lead time is set in the ordering model information 61. The reorder point method is an ordering method in which an order is placed when inventory falls below a certain threshold known as the reorder point.

[0027] The past stock information 62 according to the first embodiment will be described with reference to Fig. 6. The past stock information 62 includes one or more pairs of a date, a business name, a product name, and an inventory amount.

[0028] (Step S102: Storage process) The reception unit 21 writes the ordering model information 61 and the past inventory information 62 to the storage 13. However, the ordering model information 61 is recorded only when it is received in step S101. Here, the ordering model information 61 and the past inventory information 62 are written to the storage 13. However, the ordering model information 61 and the past inventory information 62 may be written to a storage device provided outside the demand prediction device 10, instead of the storage 13.

[0029] Demand forecasting processing according to the first embodiment will be described with reference to FIG. 7. (Step S201: Reading Process) The future inventory estimation unit 22 reads information about business operator A, which is a supplier of part A, the product being forecast, from the ordering model information 61 stored in the storage 13. Specifically, the future inventory estimation unit 22 reads the ordering model, ordering point, order quantity, and procurement lead time of business operator A. The future inventory estimation unit 22 also reads inventory information about part A at business operator A, which is the supplier, from the past inventory information 62 stored in the storage 13. Specifically, the future inventory estimation unit 22 reads the date and inventory quantity for the most recent reference period for part A at business operator A.

[0030] (Step S202: Future inventory estimation process) The future inventory estimation unit 22 estimates the future inventory amount at business operator A, which is the supply destination, based on the ordering model information 61 and past inventory information 62 read out in step S201. At this time, the future inventory estimation unit 22 estimates the future inventory amount by assuming that business operator A will increase its inventory as a result of orders being placed in accordance with the ordering conditions and order amounts indicated in the ordering model information 61, and that its inventory will decrease in accordance with the trends in the past inventory amounts indicated in the past inventory information 62. In the case of business operator A, the ordering conditions are the ordering point.

[0031] A specific description will be given with reference to FIG. 8 . Based on the order model information 61 and past inventory information 62 read in step S201, the future inventory estimation unit 22 calculates the future inventory change for each day during the prediction period starting from the day after the day to which the processing execution time falls. The future inventory estimation unit 22 then writes the daily future inventory change to the storage 13 as future inventory information 63. The future inventory estimation unit 22 sets each day during the prediction period starting from the day after the day to which the processing execution time falls as the target day, starting from the oldest day. The future inventory estimation unit 22 calculates the inventory for the target day by adding the inventory increase for the target day to the inventory for the day before the target day and subtracting the inventory decrease for the target day. In this way, the future inventory estimation unit 22 estimates the inventory for each day during the prediction period. If the order conditions are met on the day before the target day by the procurement lead time, the future inventory estimation unit 22 determines that the order quantity is the inventory increase for the target day. Here, since the ordering model is the reorder point method, if the inventory quantity on the day prior to the procurement lead time is equal to or less than the reorder point indicated by the ordering model information 61, the inventory quantity will increase by an amount equal to the order quantity indicated by the ordering model information 61. If the ordering conditions are not met on the day prior to the target day by the procurement lead time, the future inventory estimation unit 22 sets the inventory quantity to be increased on the target day to 0. The future inventory estimation unit 22 calculates a statistical value of the inventory decrease per day from the trend in inventory quantity for the most recent reference period read in step S201. The future inventory estimation unit 22 sets the statistical value to the inventory quantity that will decrease on the target day. Values ​​such as the average value and median value are used as the statistical value.

[0032] (Step S203: Demand forecasting process) The demand forecasting unit 23 calculates a demand forecast value for each day during the forecasting period. Then, the demand forecasting unit 23 writes the demand forecast value for each day as a demand forecast result 64 to the storage 13. As shown in FIG. 9 , the demand forecast result 64 includes a set of a date, a business name, a product name, and a demand amount. Specifically, the demand forecasting unit 23 predicts that, at the time when the future inventory amount estimated in step S202 satisfies the order conditions indicated by the order model information 61, demand for the order amount indicated by the order model information 61 will occur with business operator X, the supplier.

[0033] (Step S204: Output Process) The output unit 24 outputs the demand forecast result 64 to the output terminal 52.

[0034] ***Effects of First Embodiment*** As described above, the demand forecasting device 10 according to the first embodiment estimates future inventory levels using past trends in inventory levels and forecasts demand from the future inventory levels. This makes it possible to perform demand forecasts that track small daily fluctuations in demand by taking into account recent fluctuations in inventory levels. In other words, the demand forecasting device 10 does not perform demand forecasts using the order history of business operator X, but performs demand forecasts based on the daily inventory information of business operator A, the supply destination. This makes it possible to detect demand fluctuations before they are reflected in the order history and track small daily fluctuations in demand. In other words, by taking into account recent fluctuations in inventory levels, highly accurate demand forecasts are possible without waiting for the accumulation of order history, even when the market environment fluctuates significantly.

[0035] ***Other Configurations*** <Modification 1> In the first embodiment, each functional component is realized by software. However, in Modification 1, each functional component may be realized by hardware. The differences between Modification 1 and the first embodiment will be described below.

[0036] The configuration of the demand prediction device 10 according to Modification 1 will be described with reference to Fig. 10. When each functional component is realized by hardware, the demand prediction device 10 includes an electronic circuit 15 instead of the processor 11, memory 12, and storage 13. The electronic circuit 15 is a dedicated circuit that realizes the functions of each functional component, the memory 12, and the storage 13.

[0037] The electronic circuit 15 may be a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, a logic IC, a GA, an ASIC, or an FPGA. GA stands for Gate Array. ASIC stands for Application Specific Integrated Circuit. FPGA stands for Field-Programmable Gate Array. Each functional component may be realized by a single electronic circuit 15, or each functional component may be distributed across multiple electronic circuits 15.

[0038] <Modification 2> As a modification 2, some of the functional components may be realized by hardware, and other functional components may be realized by software.

[0039] The processor 11, the memory 12, the storage 13, and the electronic circuit 15 are collectively referred to as a processing circuit. In other words, the functions of the respective functional components are realized by the processing circuit.

[0040] Embodiment 2. Embodiment 2 differs from embodiment 1 in that the inventory reduction amount is determined based on the results of demand forecasts for downstream businesses in the flow of the forecast target product between businesses. In embodiment 2, this difference will be explained, and explanations of the same points will be omitted.

[0041] ***Description of Configuration*** The configuration of the demand forecasting device 10 according to the second embodiment will be described with reference to Fig. 11. The demand forecasting device 10 differs from Fig. 3 in that commercial distribution structure information 65 is stored in the storage 13.

[0042] ***Description of Operation*** The input reception processing according to the second embodiment will be described with reference to Figure 4. (Step S101: Reception processing) The reception unit 21 receives input of commercial distribution structure information 65 in addition to order model information 61 and past inventory information 62 from the input terminal 51. However, the order model information 61 is only necessary the first time the input is received, and the reception unit 21 does not necessarily need to receive the order model information 61 every time the input reception processing is executed.

[0043] Referring to FIG. 12 , commercial flow structure information 65 according to the second embodiment will be described. The commercial flow structure information 65 indicates the flow between businesses of part A, which is the product to be predicted. The commercial flow structure information 65 includes the name of a business, the name of an upstream business, and the name of a downstream business. The upstream business name indicates the business upstream of the business indicated by the business name. The downstream business name indicates the business downstream of the business indicated by the business name. The upstream business is a supplier that supplies part A to the business indicated by the business name. The downstream business is a destination to which the business indicated by the business name supplies part A.

[0044] (Step S102: Storage Process) The reception unit 21 writes the commercial distribution structure information 65 in addition to the order model information 61 and the past inventory information 62 to the storage 13. However, the order model information 61 is recorded only when it is received in step S101.

[0045] Demand forecasting processing according to the second embodiment will be described with reference to FIG. 13. (Step S211: Downstream Search Processing) The future inventory estimation unit 22 references the commercial distribution structure information 65 stored in the storage 13 to search for businesses downstream of business operator A, who is the supply destination of part A, the product being forecast. Specifically, the future inventory estimation unit 22 identifies businesses indicated by the downstream business names of business operator A in the commercial distribution structure information 65. The future inventory estimation unit 22 identifies businesses indicated by the downstream business names of each identified business operator. This process is repeated until there are no businesses indicated by downstream business names. This searches for all businesses downstream of business operator A. In the case of the commercial distribution structure information 65 shown in FIG. 12, business operators C and D are identified as businesses downstream of business operator A. Business operator F is identified as a business downstream of business operator C. Since there are no businesses downstream of business operator D, the search is terminated. Since there are no businesses downstream of business operator F, the search is terminated.

[0046] (Step S212: Reading process) The future stock estimation unit 22 reads out the ordering model information 61 and the past stock information 62 for each of the businesses searched for in step S211 (here, businesses C, D, and F) in addition to business A. Specifically, the future stock estimation unit 22 sets business A and each of the businesses searched for in step S211 as target businesses. The future stock estimation unit 22 reads out the information for the target business from the ordering model information 61. In addition, the future stock estimation unit 22 reads out the information on the inventory quantity for part A at the target business from the past stock information 62.

[0047] The processes of steps S213 and S214 are executed by selecting target businesses in order from downstream businesses.

[0048] (Step S213: Future Inventory Estimation Process) The future inventory estimation unit 22 estimates the future inventory amount at the target business operator based on the ordering model information 61 and the past inventory information 62 read in step S212. Specifically, as in the first embodiment, the future inventory estimation unit 22 sets each day in the prediction period starting from the day following the day to which the process is executed as the target day, in order from earliest to latest. The future inventory estimation unit 22 calculates the inventory amount on the target day by adding the inventory amount that will increase on the target day to the inventory amount on the day before the target day and subtracting the inventory amount that will decrease on the target day. However, in the second embodiment, if the target business operator does not have a downstream business operator, the future inventory estimation unit 22 determines the statistical value to be the inventory amount that will decrease on the target day, as in the first embodiment. On the other hand, if the target business operator has a downstream business operator, the future inventory estimation unit 22 determines the demand of the immediately downstream business operator to be the inventory amount that will decrease on the target day, as shown in FIG. 14. In FIG. 14, the ordering model is the reorder point method. If there are multiple businesses immediately downstream, the future stock estimation unit 22 determines that the total demand of the multiple businesses is the inventory amount that will decrease on the target day.

[0049] A specific example will be described with reference to Fig. 15. Here, the target business operator is selected from downstream businesses. Therefore, business operator F or business operator D is selected first as the target business operator. After processing for business operator F is completed, business operator C is selected as the target business operator. After processing for business operators C and D is completed, business operator A is selected as the target business operator.

[0050] For businesses F and D, there are no downstream businesses. Therefore, the daily inventory amount is calculated in the same way as in embodiment 1. For businesses C and A, there are downstream businesses. Therefore, the daily inventory amount is calculated by assuming that the demand of the businesses immediately downstream is the inventory amount that will decrease on the target day. When calculating future inventory for business C, the daily inventory decrease amount for business C is determined based on the results of demand forecasts for orders placed by business F to business C. When calculating future inventory for business A, the daily inventory decrease amount for business A is determined based on the results of demand forecasts for orders placed by business C and business D to business A.

[0051] (Step S214: Demand forecasting processing) The demand forecasting unit 23 predicts that when the future inventory amount estimated in step S213 satisfies the ordering conditions indicated by the ordering model information 61, demand for the order amount indicated by the ordering model information 61 will arise for the business operator immediately upstream.

[0052] The process in step S215 is the same as the process in step S204 in FIG.

[0053] ***Effects of Embodiment 2*** As described above, the demand forecasting device 10 according to Embodiment 2 determines the inventory reduction amount based on the results of demand forecasting for downstream businesses in the flow of the forecast target product between businesses. This makes it possible to accurately estimate the future inventory amount of the supply destination business. As a result, highly accurate demand forecasting becomes possible.

[0054] ***Other Configurations*** <Modification 3> For businesses downstream of business operator A, demand may be forecast using a known method other than the methods described in the first and second embodiments. When forecasting demand from business operator A to business operator X, the demand from the downstream businesses to business operator A, predicted by some method, may be used as the inventory reduction amount for business operator A.

[0055] Embodiment 3. Embodiment 3 differs from embodiment 2 in that the inventory quantity on the day to which the processing is to be performed is estimated by referencing logistics information 66 that indicates the transitions that occur in the logistics process for the prediction target item. In embodiment 3, this difference will be explained, and explanations of the same points will be omitted. Here, a case where a function is added to embodiment 1 will be explained. However, it is also possible to add a function to embodiment 2.

[0056] ***Description of Configuration*** The configuration of the demand forecasting device 10 according to the third embodiment will be described with reference to FIG. 16. The demand forecasting device 10 differs from that shown in FIG. 3 in that it includes a past stock estimation unit 25 as a functional component. The function of the past stock estimation unit 25, like the other functional components, is realized by software or hardware. The demand forecasting device 10 also differs from that shown in FIG. 3 in that logistics information 66 and past stock information 67 are stored in the storage 13.

[0057] ***Description of Operation*** The input reception process according to the third embodiment will be described with reference to Fig. 4. (Step S101: Reception process) The reception unit 21 receives input of the order model information 61 and the logistics information 66 from the input terminal 51. However, the order model information 61 is only required the first time the order model information 61 is received, and the reception unit 21 does not necessarily need to receive the order model information 61 every time the input reception process is executed.

[0058] With reference to FIG. 17 , the logistics information 66 according to the third embodiment will be described. The logistics information 66 indicates the transitions that occur in the logistics process for the prediction target product. The logistics information 66 includes the date, the business name, the product name, the event name, and the quantity. Here, the event name is the name of the event. The event is a product transition that occurs in the logistics process, such as the production, shipping, and arrival of the product. The quantity is the number of products that transition in the event.

[0059] (Step S102: Storage Process) The reception unit 21 writes the order model information 61 and the logistics information 66 into the storage 13. However, the order model information 61 is recorded only when it is received in step S101.

[0060] 18 , the demand forecasting process according to the third embodiment will be described. (Step S221: Reading process) The future stock estimation unit 22 reads out the information of the business operator A from the order model information 61 stored in the storage 13. The future stock estimation unit 22 also reads out the logistics information 66 stored in the storage 13.

[0061] (Step S222: Past Stock Estimation Process) The past stock estimation unit 25 refers to the logistics information 66 and estimates the inventory amount of part A, which is the prediction target item at business operator A, the supply destination, for each day of the most recent reference period up to the day to which the processing is executed. The past stock estimation unit 25 writes the estimated daily inventory amount to the storage 13 as past stock information 67. A specific description will be given with reference to FIG. 19 . The past stock estimation unit 25 sets each day of the most recent reference period up to the day to which the processing is executed as the target day, starting from the oldest. The past stock estimation unit 25 sets an initial value to the inventory amount on the day before the oldest day in the most recent reference period. The past stock estimation unit 25 calculates the inventory amount on the target day by adding the inventory amount that increases on the target day indicated by the logistics information 66 to the inventory amount on the day before the target day and subtracting the inventory amount that decreases on the target day indicated by the logistics information 66 from the inventory amount on the day before the target day. This estimates the inventory amount on the day to which the processing is executed. The inventory quantity to be increased on the target day is the quantity of part A arriving at business operator A on the target day according to the logistics information 66. The inventory quantity to be decreased on the target day is the quantity of part A shipped from business operator A on the target day according to the logistics information 66.

[0062] (Step S223: Future inventory estimation process) The future inventory estimation unit 22 uses the inventory amount indicated by the past inventory information 67 estimated in step S222 as the past inventory amount indicated by the past inventory information 62 to estimate the future inventory amount at operator A in the same manner as in embodiment 1.

[0063] The processes in steps S224 and S225 are the same as those in steps S203 and S204 in FIG.

[0064] ***Effects of Embodiment 3*** As described above, the demand forecasting device 10 according to embodiment 3 estimates the inventory amount on the day on which the processing execution time falls by referring to the logistics information 66, and thereby estimates the future inventory amount. This makes it possible to estimate the future inventory amount of business operator A even when it is not possible to obtain information on the inventory amount on the day on which business operator A's processing execution time falls. As a result, it is possible to perform demand forecasting even when it is not possible to obtain information on the inventory amount on the day on which business operator A's processing execution time falls.

[0065] Embodiment 4. Embodiment 4 differs from embodiment 3 in that the inventory quantity on the day to which the processing is performed is estimated using transitions that occur in the logistics process for a used product, which is a product produced using the prediction target product as a part. In embodiment 4, this difference will be explained, and explanations of the same points will be omitted.

[0066] ***Description of Configuration*** The configuration of the demand prediction device 10 according to the fourth embodiment will be described with reference to Fig. 20. The demand prediction device 10 differs from Fig. 16 in that part configuration information 68 is stored in the storage 13.

[0067] ***Description of Operation*** The input reception process according to the fourth embodiment will be described with reference to Fig. 4. (Step S101: Reception process) The reception unit 21 receives input of order model information 61, logistics information 66, and part configuration information 68 from the input terminal 51. However, the order model information 61 is only required the first time the input is received, and the reception unit 21 does not necessarily need to receive the order model information 61 every time the input reception process is executed.

[0068] 21 , the component configuration information 68 according to the fourth embodiment will be described. The component configuration information 68 indicates the quantity of component A, which is a prediction target item used in a product using the component. The component configuration information 68 includes a product name, a component name, and a quantity. The product name is the name of the product using the component. The component name is the name of a component, such as component A, used in the product using the component. The quantity is the number of components used in the product using the component.

[0069] (Step S102: Storage Process) The reception unit 21 writes the order model information 61, the logistics information 66, and the part configuration information 68 into the storage 13. However, the order model information 61 is recorded only when it is received in step S101.

[0070] 18 , the demand forecasting process according to the fourth embodiment will be described. (Step S221: Reading process) The future stock estimation unit 22 reads out the information of the business operator A from the order model information 61 stored in the storage 13. The future stock estimation unit 22 also reads out the logistics information 66 and the part configuration information 68 stored in the storage 13.

[0071] (Step S222: Past Stock Estimation Processing) As in the third embodiment, the past stock estimation unit 25 sets each day of the most recent reference period up to the day to which the processing execution time belongs as the target day, starting from the oldest, and estimates the inventory amount for the target day. At this time, the past stock estimation unit 25 refers to the logistics information 66 to determine whether or not business operator A is producing a product using part A as a component. If business operator A is not producing a product using part A as a component, the past stock estimation unit 25 estimates the inventory amount for the target day as in the third embodiment. If business operator A is producing a product using part A as a component, the past stock estimation unit 25 identifies the inventory amount that will decrease on the target day as follows and estimates the inventory amount for the target day. Except for the method of identifying the inventory amount that will decrease on the target day, the process is the same as in the third embodiment. Instead of obtaining the quantity of part A to be shipped from the logistics information 66, the past stock estimation unit 25 calculates the production and shipping amounts of products using part A that are produced based on part A. The past inventory estimation unit 25 identifies the amount of part A used in the production of the product using the part A or the amount of part A shipped in a form incorporated into the product using the part A from the part configuration information 68, and regards this amount as the inventory reduction amount.

[0072] A specific example will be described with reference to FIG. 21 . When calculating the shipping volume of part A, the past inventory estimation unit 25 first refers to the logistics information 66 to determine whether business operator A is producing a product that uses part A as a component, such as product A or product B. For example, if business operator A is producing product B, the past inventory estimation unit 25 identifies the production or shipping volume of product B on the target day from the logistics information 66. Then, the past inventory estimation unit 25 identifies the quantity of part A used in product B from the parts configuration information 68, and calculates the total amount of part A used in the production or shipping of product B. The past inventory estimation unit 25 regards the calculated total amount as the inventory amount that will decrease on the target day.

[0073] ***Effects of Embodiment 4*** As described above, the demand forecasting device 10 according to embodiment 4 estimates the inventory quantity on the day to which the processing execution time falls, using the transitions that occur in the logistics process for used products, which are products produced using the forecast target product as a component. This makes it possible to appropriately estimate the inventory quantity of forecast target products whose tracking is lost in the logistics information 66. In other words, even if tracking of the forecast target product is lost in the logistics information 66 due to, for example, a change in shape during the distribution process, it becomes possible to appropriately estimate the inventory quantity.

[0074] In addition, the word "unit" in the above description may be read as a "circuit," "step," "procedure," "process," or "processing circuit."

[0075] The embodiments and modifications of the present disclosure have been described above. Some of these embodiments and modifications may be combined and implemented. Furthermore, one or more of them may be implemented partially. Note that the present disclosure is not limited to the above embodiments and modifications, and various modifications are possible as needed.

[0076] 10 Demand forecasting device, 11 Processor, 12 Memory, 13 Storage, 14 Communication interface, 15 Electronic circuit, 21 Reception unit, 22 Future inventory estimation unit, 23 Demand forecasting unit, 24 Output unit, 25 Past inventory estimation unit, 51 Input terminal, 52 Output terminal, 61 Order model information, 62 Past inventory information, 63 Future inventory information, 64 Demand forecast results, 65 Distribution flow structure information, 66 Logistics information, 67 Past inventory information, 68 Parts configuration information.

Claims

1. A demand forecasting device that forecasts demand that will occur for each unit period, a future inventory estimation unit which estimates a future inventory amount at a destination of a prediction target product, assuming that inventory will increase as a result of orders being placed in accordance with the ordering conditions and order amount indicated by the ordering model information, and inventory will decrease in accordance with the trend in past inventory amount indicated by the past inventory information, and which estimates the inventory amount for each unit period during the prediction period, starting from the unit period next to the unit period to which the processing is executed, in order from the oldest to the newest, by adding the inventory amount that will increase in the target unit period to the inventory amount in the unit period preceding the target unit period and subtracting the inventory amount that will decrease in the target unit period to calculate the inventory amount in the target unit period; a demand forecasting unit that forecasts that a demand for an order amount indicated by the ordering model information will occur at a supplier at a time point when the ordering condition indicated by the ordering model information is satisfied in the future inventory amount estimated by the future inventory estimation unit; Equipped with The future inventory estimation unit calculates a statistical value of the inventory reduction amount per unit period from the trend in inventory amount indicated by the past inventory information, and regards the statistical value as the inventory amount that will be reduced in the target unit period.

2. A demand forecasting device that forecasts demand that will occur for each unit period, a future inventory estimation unit which estimates a future inventory amount at a destination of a prediction target product, assuming that inventory will increase as a result of orders being placed in accordance with the ordering conditions and order amount indicated by the ordering model information, and inventory will decrease in accordance with the trend in past inventory amount indicated by the past inventory information, and which estimates the inventory amount for each unit period during the prediction period, starting from the unit period next to the unit period to which the processing is executed, in order from the oldest to the newest, by adding the inventory amount that will increase in the target unit period to the inventory amount in the unit period preceding the target unit period and subtracting the inventory amount that will decrease in the target unit period to calculate the inventory amount in the target unit period; a demand forecasting unit that forecasts that a demand for an order amount indicated by the ordering model information will occur at a supplier at a time point when the ordering condition indicated by the ordering model information is satisfied in the future inventory amount estimated by the future inventory estimation unit; a past inventory estimation unit that estimates an inventory amount of the supply destination for each unit period of a past reference period up to the unit period to which the process execution time belongs, by referring to logistics information indicating a transition occurring in a logistics process for the prediction target product at the supply destination; Equipped with The future inventory estimation unit is a demand forecasting device that estimates a future inventory amount by using the inventory amount estimated by the past inventory estimation unit as the past inventory amount indicated by the past inventory information.

3. The future inventory estimation unit determines that the order quantity is an inventory quantity to be increased in the target unit period if the order condition is satisfied in a unit period that is a procurement lead time before the target unit period. The demand forecasting device according to claim 1 or 2.

4. The demand forecasting device forecasts demand that will occur for the supply source, which is an upstream business of the target business, in order from a downstream business in a flow between businesses of the prediction target product, and The future inventory estimation unit determines, when the target business operator has no downstream business operators, that the statistical value is the inventory amount that decreases in the target unit period, and when the target business operator has downstream business operators, that the demand of the downstream business operators is the inventory amount that decreases in the target unit period. The demand forecasting device according to claim 1 .

5. The past inventory estimation unit sets an initial value to the inventory amount in the unit period before the reference period of the unit period to which the processing execution time belongs, and estimates the inventory amount in the unit period to which the processing execution time belongs by calculating the inventory amount in the target unit period by adding the inventory amount that increases in the target unit period indicated by the logistics information to the inventory amount in the unit period before the target unit period indicated by the logistics information and subtracting the inventory amount that decreases in the target unit period indicated by the logistics information, for the unit period from the next unit period before the reference period to the unit period to which the processing execution time belongs. The demand forecasting device according to claim 2.

6. The logistics information indicates a transition occurring in a logistics process for a used product, which is a product produced by using the prediction target product as a part, The past stock estimation unit refers to part configuration information indicating the number of the prediction target items used in the use product, and takes into consideration the usage amount of the prediction target items calculated from the production volume of the use product, to identify the inventory amount to be reduced in a unit period of the target. The demand forecasting device according to claim 5.

7. A demand forecasting method for forecasting demand occurring in each unit period, comprising: the computer estimates future inventory levels at the destination of the prediction target item, assuming that inventory will increase as orders are placed in accordance with the ordering conditions and order levels indicated in the ordering model information, and that inventory will decrease in accordance with the trends in past inventory levels indicated in the past inventory information, and estimates inventory levels for each unit period during the prediction period, starting from the unit period next to the unit period to which the processing is executed, beginning with the oldest, by adding the inventory level that will increase in the target unit period to the inventory level in the unit period preceding the target unit period and subtracting the inventory level that will decrease in the target unit period to the inventory level in the unit period preceding the target unit period; The computer predicts that a demand for the order amount indicated by the ordering model information will occur at a supplier at a time when the estimated future inventory amount satisfies the ordering condition indicated by the ordering model information, and A demand forecasting method in which a computer calculates a statistical value of the inventory reduction amount per unit period from the trend in inventory amount indicated by the past inventory information, and regards the statistical value as the inventory reduction amount in the target unit period.

8. A demand forecasting method for forecasting demand occurring in each unit period, comprising: the computer estimates future inventory levels at the destination of the prediction target item, assuming that inventory will increase as orders are placed in accordance with the ordering conditions and order levels indicated in the ordering model information, and that inventory will decrease in accordance with the trends in past inventory levels indicated in the past inventory information, and estimates inventory levels for each unit period during the prediction period, starting from the unit period next to the unit period to which the processing is executed, beginning with the oldest, by adding the inventory level that will increase in the target unit period to the inventory level in the unit period preceding the target unit period and subtracting the inventory level that will decrease in the target unit period to the inventory level in the unit period preceding the target unit period; The computer predicts that a demand for the order amount indicated by the ordering model information will occur at a supplier at a time when the estimated future inventory amount satisfies the ordering condition indicated by the ordering model information, and a computer refers to logistics information indicating a transition occurring in a logistics process for the prediction target product at the supply destination, and estimates an inventory amount at the supply destination for each unit period of a past reference period up to the unit period to which the processing execution time belongs; A demand forecasting method in which a computer estimates future inventory amounts using an estimated inventory amount as the past inventory amount indicated by the past inventory information.

9. A demand forecasting program for forecasting demand occurring in each unit period, a future inventory estimation process for estimating a future inventory amount at a destination of a prediction target product, on the assumption that inventory will increase as a result of orders being placed in accordance with the ordering conditions and order amounts indicated by the ordering model information, and inventory will decrease in accordance with the trend in past inventory amounts indicated by the past inventory information, the future inventory estimation process being for each unit period during a prediction period starting from the unit period next to the unit period to which the process is executed, beginning from the oldest, and estimating the inventory amount for each unit period during the prediction period by adding the inventory amount that will increase in the target unit period to the inventory amount in the unit period preceding the target unit period and subtracting the inventory amount that will decrease in the target unit period to the inventory amount in the unit period preceding the target unit period; a demand forecasting process for forecasting that a demand for an order amount indicated by the ordering model information will occur at a supplier at a time point when the ordering condition indicated by the ordering model information is satisfied in the future inventory amount estimated by the future inventory estimation process; The computer functions as a demand forecasting device that performs the following: In the future inventory estimation process, a demand forecasting program calculates a statistical value of the inventory reduction amount per unit period from the trend in inventory amount indicated by the past inventory information, and regards the statistical value as the inventory amount that will decrease in the target unit period.

10. A demand forecasting program for forecasting demand occurring in each unit period, a future inventory estimation process for estimating a future inventory amount at a destination of a prediction target product, on the assumption that inventory will increase as a result of orders being placed in accordance with the ordering conditions and order amounts indicated by the ordering model information, and inventory will decrease in accordance with the trend in past inventory amounts indicated by the past inventory information, the future inventory estimation process being for each unit period during a prediction period starting from the unit period next to the unit period to which the process is executed, beginning from the oldest, and estimating the inventory amount for each unit period during the prediction period by adding the inventory amount that will increase in the target unit period to the inventory amount in the unit period preceding the target unit period and subtracting the inventory amount that will decrease in the target unit period to the inventory amount in the unit period preceding the target unit period; a demand forecasting process for forecasting that a demand for an order amount indicated by the ordering model information will occur at a supplier at a time point when the ordering condition indicated by the ordering model information is satisfied in the future inventory amount estimated by the future inventory estimation process; a past inventory estimation process for estimating an inventory amount of the supply destination for each unit period of a past reference period up to the unit period to which the process execution time belongs, by referring to logistics information indicating a transition occurring in the process of logistics for the prediction target product at the supply destination; The computer functions as a demand forecasting device that performs the following: A demand forecasting program in which, in the future inventory estimation process, the inventory amount estimated by the past inventory estimation process is regarded as the past inventory amount indicated by the past inventory information, and a future inventory amount is estimated.