JP7538677B2 - Support system, support program, and control method - Google Patents

Description

The present invention relates to a support system that assists in creating product replenishment plans for vending machines, and a support program and control method for the support system.

Patent Document 1 describes a delivery request system that is applied to a logistics form in which products to be replenished are delivered from a logistics base via a delivery base to each vending machine at multiple delivery destinations using a specified container. The delivery request system includes a server that provides delivery request information specifying the type and quantity of products to be delivered to the delivery base to the system of a logistics company that is responsible for logistics to the delivery base. The server then determines, for each delivery destination, the type and quantity of products to be stored in each container in response to the replenishment information indicating the replenishment quantity of products for each delivery destination. The server also generates delivery request information that specifies the determined type and quantity of products for each container in association with delivery destination information, and transmits the delivery request information to the logistics company's system.

Patent No. 6641438

A vending machine has multiple storage rows that store products (e.g., tobacco, beverages, etc.). These storage rows are generally called columns, and each column usually stores one type of product. The system described in Patent Document 1 does not take into consideration the storage of new products that are not yet stored in the vending machine in columns. Therefore, when storing new products that are not yet stored, it is difficult to include the optimal product lineup and replenishment quantities in the replenishment plan.

The support system according to one aspect of the present invention is a support system that calculates the replenishment quantity of a product based on the predicted demand for the product contained in each column of a vending machine having multiple columns, and supports the creation of a replenishment plan for the product based on the calculated replenishment quantity, and is equipped with an acquisition means for acquiring recommendation information including product identification information that is selected from a group of products including uncontained products not contained in the vending machine and that identifies recommended products recommended to be contained in the vending machine and predicted demand quantity information for each recommended product, and a calculation means for calculating the replenishment quantity of the recommended products to be contained in the vending machine based on the acquired recommendation information.

Another aspect of the present invention is an assistance program for an assistance system including a computer that calculates a replenishment quantity for a product to be stored in each column of a vending machine having multiple columns based on predicted demand for the product, and assists in creating a replenishment plan for the product based on the calculated replenishment quantity, and causes the computer to function as an acquisition means for acquiring recommendation information including product identification information that is selected from a group of products including unstored products not stored in the vending machine and identifies recommended products that are recommended to be stored in the vending machine and predicted demand quantity information for each recommended product, and a calculation means for calculating the replenishment quantity of the recommended products to be stored in the vending machine based on the acquired recommendation information.

A control method according to one aspect of the present invention is a control method for a support system that calculates a replenishment quantity of a product based on a predicted demand for the product contained in each column of a vending machine having multiple columns, and supports the creation of a replenishment plan for the product based on the calculated replenishment quantity, and obtains recommendation information including product identification information that identifies recommended products selected from a group of products including uncontained products not contained in the vending machine and that are recommended to be contained in the vending machine, and predicted demand quantity information for each recommended product, and calculates the replenishment quantity of the recommended products to be contained in the vending machine based on the obtained recommendation information.

This allows you to create a replenishment plan that takes into account the optimal product lineup and replenishment quantities, even when stocking new products that are not yet stocked.

FIG. 1 is a schematic configuration diagram for explaining a support system. FIG. 1 is a schematic block diagram of an assistance system. Flowchart for planning. FIG. 13 is a schematic diagram showing an example of a record output from a prediction server. FIG. 13 is a schematic diagram showing an example of a record output from a recommended server. FIG. 13 is a schematic diagram showing an example of a record output to a prediction server. FIG. 13 is a schematic diagram showing an example of a record output to a linkage server. FIG. 1 is a schematic diagram illustrating an example of a record showing a replenishment plan. FIG. 13 is a schematic diagram illustrating an example of a record showing a replenishment plan including operation hours.

Below, exemplary embodiments for carrying out the present invention will be described in detail with reference to the drawings. However, the dimensions, materials, shapes, and relative positions of components described in the following embodiments can be set arbitrarily and can be changed according to the configuration of the device to which the present invention is applied or various conditions. Furthermore, unless otherwise specified, the scope of the present invention is not limited to the embodiments specifically described below.

With reference to FIG. 1, an example of a logistics form to which a support system 100 according to an embodiment of the invention is applied will be described. In the logistics form illustrated in FIG. 1, products are delivered from a shipping base 1 via a logistics base 2 to a distribution base 3 in various locations. Note that the products may be any item that can be sold in a vending machine 5 and are not limited to food and beverages, but an example in which the products are beverages will be described below. As an example, the shipping base 1 includes a beverage manufacturing factory 1A and a manufacturer's storage warehouse 1B. The beverages are delivered to the logistics base 2 in a state where they are packaged by type in packaging bodies PK such as cartons. Freight vehicles 4 or the like may be used as appropriate to deliver the packaging bodies PK.

The distribution base 3 has a designated area of responsibility, and products are distributed to vending machines 5 installed within the area of responsibility as delivery destinations DT. A distribution route is set within the area of responsibility. In FIG. 1, two distribution routes RT1 and RT2 (hereinafter sometimes referred to as distribution routes RT) are set within one area of responsibility. However, it is sufficient that at least one distribution route RT is set for one area of responsibility. When multiple distribution routes RT are set for one area of responsibility, the vending machines 5 present within the area of responsibility are appropriately assigned to one of the distribution routes RT.

A worker at the distribution base 3 loads beverages onto a delivery vehicle 6 and follows the distribution route RT to visit the vending machines 5 in sequence to replenish the vending machines 5 with beverages. The area covered is set to an extent that the vending machines 5 can be visited within a day or less. The logistics base 2 is provided to cover an area that is larger than the area covered by the distribution base 3. Although FIG. 1 shows only one logistics base 2, multiple logistics bases 2 may be provided to cover a specific area each. In this case, the multiple logistics bases 2 may be provided hierarchically so that the covered areas are successively narrowed. Alternatively, beverages may be relayed between multiple logistics bases 2 with different covered areas.

Containers, as an example of cargo objects, are used for delivery from the logistics base 2 to the distribution base 3, and for delivery from the distribution base 3 to the vending machine 5 at the delivery destination DT. As an example, the container is a foldable rectangular container whose size and shape are standardized. There may be multiple standards for containers, in which case a container with a standard suitable for the delivery of beverages may be appropriately selected. There are certain restrictions on container handling. As an example, each container has a limit set on the weight it can accommodate. The restrictions on handling are not limited to weight, and the total volume or length of the beverages may be set instead of or in addition to weight.

Picking of beverages into containers is performed at the logistics base 2 for each vending machine 5. The number of containers used for one vending machine 5 varies depending on the quantity of beverages to be delivered to the vending machine 5 (i.e., the number of bottles, cans, or bottles). In other words, multiple containers may be used for one vending machine 5. The containers in which beverages have been loaded at the logistics base 2 are delivered to the delivery base 3 in charge of the delivery destination DT to which the beverages in the containers should be delivered. The containers delivered to the delivery base 3 are delivered to the delivery destination DT in sequence according to the delivery schedule for the delivery destination DT of the delivery base 3. Then, the worker refills the vending machine 5 with the beverages contained in the container. In this way, at the delivery base 3, it is not necessary to open the package PK and perform picking work for each vending machine 5. At the logistics base 2, which is located upstream of the delivery route from the delivery base 3, i.e., closer to the shipping base 1, the containers are distinguished for each delivery destination DT and picking work is performed. However, depending on the facilities and location of the logistics base 2, the picking work may be performed at the delivery base 3, or a worker may pick the goods from the delivery vehicle 6 at the delivery destination DT.

Figure 1 shows an example where one vending machine 5 is installed at one delivery destination DT. However, the delivery destination DT is the unit where the work of loading and unloading products from the delivery vehicle 6 and refilling is performed, and one delivery destination DT may include multiple vending machines 5. The vending machines 5 do not necessarily need to be able to sell all types of beverages that can be delivered from the logistics base 2, but only need to be able to sell at least some types of beverages. The number of delivery destinations DT included in one delivery route RT may also be changed as appropriate. When beverages are delivered sequentially between multiple logistics bases 2 arranged in a hierarchical structure, or when beverages are delivered while being relayed between multiple logistics bases 2 with different areas of responsibility, picking into a container is performed at one of the logistics bases 2.

On the upstream side where picking takes place, delivery is made by package PK. On the downstream side, delivery is made by container. Delivery from the delivery base 3 is made periodically, for example, once a day. However, it is not necessary to set all delivery destinations DT as delivery targets for each delivery. Each time a delivery is made, a delivery destination DT within the delivery route RT may be selected as appropriate. The delivery route RT may be changed as appropriate according to delivery circumstances. Furthermore, the delivery destinations DT within one delivery route RT may be shared by multiple delivery vehicles 6.

Delivery from shipping base 1 to logistics base 2, and delivery from logistics base 2 to distribution base 3, are operations that cover a relatively wide area. On the other hand, delivery from distribution base 3 to destination DT is an operation within a relatively small area. In addition, delivery to destination DT is an operation that includes operational management such as replenishing beverages, collecting sales proceeds, and maintaining vending machines 5. For this reason, operations that cover a wide area are handled by, for example, a logistics company that has built a nationwide logistics network. And operations within a small area are handled by an operator that operates and manages each vending machine 5.

[Support system]
The support system 100 predicts the refill quantity of a product based on the predicted demand for the product contained in each column of a vending machine 5 having multiple columns, and supports the creation of a product refill plan based on the predicted refill quantity. For example, the predicted demand is a predicted potential sales quantity. For this purpose, the support system 100 includes a demand prediction system 10, an optimization system 20, and a linkage system 30. The demand prediction system 10 includes a prediction server 11, and selects multiple vending machines 5 that require refilling, and predicts the refill quantity as demand from the sales record of each vending machine 5. As an example, the prediction server 11 predicts a delivery date and a refill quantity so that costs are minimized, taking into account costs incurred in relation to refilling (e.g., delivery costs and refill costs) and losses due to out-of-stock items. Then, the prediction server 11 outputs the refill quantity of the product for each column of the vending machine.

As an example, the prediction server 11 calculates the replenishment quantity by referring to the sales quantity as the forecasted demand predicted based on the performance data. The performance data includes the quantity of beverages sold at each vending machine 5, or the theoretical inventory calculated from the quantity of beverages replenished to each vending machine 5. The performance data may also include various data collected by the linking server 31 from the vending machines 5, or a replenishment plan created by the linking server 31 based on the collected data. Furthermore, the performance data may include data entered by an operator.

The optimization system 20 includes a recommendation server 21, which predicts sales volume as potential demand for each product and recommends a column configuration as an assortment, i.e., optimal products to be accommodated in each of a number of columns. As an example, the recommendation server 21 predicts sales volume based on the past sales performance of all vending machines 5. That is, the recommendation server 21 predicts sales volume for each product in a product group that includes unaccommodated products not accommodated in each vending machine 5. The recommendation server 21 then outputs, for each vending machine 5, the products to be accommodated in each column and the predicted sales volume of the products. Alternatively, the products to be accommodated in each column and the predicted sales volume of the products may be determined by an operator.

The prediction server 11 and the recommendation server 21 may also predict the sales volume using a learning model constructed based on learning data including the past sales performance of all vending machines 5 and associated data related to the sales performance. As an example, the sales performance is the quantity sold for each product in a specific period, and the associated data is linked to the quantity. For example, the associated data is the year, month, date, time, day of the week, public holiday, holiday, weather, and temperature (as an example, the average temperature, maximum temperature, and minimum temperature in a specific period). The learning model may be constructed by the prediction server 11 and the recommendation server 21, or may be a learning model constructed in advance.

The linking system 30 links the output results of the demand forecasting system 10 and the optimization system 20. That is, the linking system 30 includes a linking server 31, and determines whether or not to reflect the column configuration recommended by the optimization system 20 for the vending machines 5 selected by the demand forecasting system 10 as the target for replenishment. For example, the demand forecasting system 10 calculates the replenishment quantity for 100 vending machines 5 that are part of the 500 vending machines 5 and require replenishment based on the sales history of each. The optimization system 20 also predicts the sales volume of each of the 100 vending machines 5 that require replenishment based on the sales history of the 500 vending machines 5. The linking system 30 then reflects the proposal by the optimization system 20 for 5 machines that are part of the 100 vending machines 5 that require replenishment. For the remaining 95 vending machines 5 of the 100 that require replenishment, the forecast result of the demand forecasting system 10 is adopted. The linking server 31 then creates and outputs a replenishment plan including the replenishment quantity for each product after the product lineup is changed to the optimal lineup. Furthermore, the linkage system 30 may add the work time required for replenishing each vending machine 5 to the replenishment plan and output it.

On the other hand, when the worker at the distribution base 3 decides on the column configuration and places an order for products, the order for products is placed with the logistics company at each distribution base 3 based on the performance data collected from the vending machines 5. The logistics company distributes the products from the logistics base 2 to the distribution base 3. Then, the worker at the distribution base 3 loads the products into the freight car 4 and transports them to each vending machine 5, taking into consideration the number of vending machines 5 to be visited for replenishment and sales forecasts. This type of delivery work is performed based on the experience and intuition of the worker. As a result, products that should be stored at the distribution base 3 are out of stock and inventory loss occurs. In response to this, the support system 100 according to this embodiment can propose the optimal product lineup for each vending machine 5 using performance data. This makes it possible to reduce product shortages and inventory loss, thereby improving sales and reducing costs.

The support system 100 may further include a delivery optimization system (not shown). The delivery optimization system creates and outputs a visit plan based on the replenishment plan obtained from the linkage system 30. As an example, the delivery optimization system creates a visit plan that determines the order and operating route for visiting each vending machine 5, taking into consideration the addresses of the vending machines 5 to be visited for replenishment, the work time required to replenish each vending machine 5, the location of each vending machine 5, and workers who are aware of the installation location of each vending machine 5, as well as visit constraints such as designated work time periods for each vending machine 5.

The prediction server 11, the recommendation server 21, and the linking server 31 are configured as a single logical server device by combining server units as multiple computers. However, the prediction server 11, the recommendation server 21, and the linking server 31 may be configured as a single server unit. Alternatively, the prediction server 11, the recommendation server 21, and the linking server 31 may be configured logically using cloud computing. At least two of the prediction server 11, the recommendation server 21, and the linking server 31 can be provided in a single server. For example, the linking server 31 can function as the prediction server 11 or the recommendation server 21. Furthermore, the prediction server 11 can function as the recommendation server 21.

[Delivery procedure]
The delivery operation may be repeated at an appropriate cycle, but as an example, it is repeated once a day. First, in the linkage system 30, performance data is collected as sales performance from the vending machines 5 in various locations. The performance data is collected via the Internet NT, or collected by a worker who visits the vending machines 5. The performance data is then used to calculate the replenishment quantity of beverages to be replenished in each vending machine 5. For example, the performance data is collected at an execution time determined in association with the date and time of delivery from the distribution base 3, such as the night of the day before the day delivery from the distribution base 3 is to be made. In addition, the performance data may be collected directly by the prediction server 11 or the recommendation server 21 without going through the linkage system 30.

The demand forecasting system 10 determines the replenishment timing, such as the replenishment date, for the vending machines 5 that should be replenished based on the performance data of each vending machine 5, and calculates the replenishment quantity of beverages. The linkage system 30 obtains recommendation information, described later, from the optimization system 20, and provides request data for requesting delivery to a logistics company according to the recommendation information and the prediction results of the demand forecasting system 10. Furthermore, at the logistics base 2, the picking contents for the containers of each vending machine 5 are determined based on the request data provided by the linkage system 30, and beverages are picked according to the determination results. However, depending on the facilities and location of the logistics base 2, the picking work may be performed by an operator from a delivery vehicle 6 at the distribution base 3 or the delivery destination DT. In this case, the request data is provided to the distribution base 3 from the linkage system 30.

The linkage system 30 also creates a replenishment plan for each delivery route RT corresponding to the distribution base 3 based on the information collected from the vending machines 5. The replenishment plan includes information identifying the vending machines 5 to be replenished, information identifying the beverages to be replenished, and the replenishment quantities of the beverages. The linkage system 30 then creates request data as an example of delivery request information to the logistics company based on the replenishment plan. The linkage system 30 transmits the created request data to a logistics system managed by the logistics company or to the distribution base 3.

As an example, the request data includes information specifying the contents of the picking request corresponding to each vending machine 5. Specifically, the information specifying the contents of the request includes the product code, the specified quantity, the container delivery destination, the doorstep number, the doorstep serial number, and the route number, etc., all linked to a vending machine code unique to each delivery destination DT. The specified quantity is the number of beverages to be picked into one container. The container delivery destination is the address or name of the delivery base 3, but may also be a unique base code for each delivery base 3.

The front door number is information for distinguishing between different delivery destinations (delivery destinations DT). As an example, the front door number is set to a 5-digit consecutive number for each vending machine code, starting from the first record in the request data. The front door serial number is a consecutive number set in sequence starting from 1 for each container corresponding to the same front door number. If the front door number is different, the front door serial number is also set in sequence starting from 1. The route number is information indicating whether the delivery route RT is the same or different and the number of delivery destinations DT included in the same delivery route RT. For example, the first two digits of the route number are set to a unique number for each delivery route RT corresponding to the delivery destination DT. The last two digits of the route number are set to a consecutive number starting from 1 for each delivery destination DT that shares a common delivery route RT. Therefore, by referring to the last two digits of the route number, the number of delivery destinations DT included in the same delivery route can be known.

The created request data is appropriately converted into a format suitable for data communication and sent to a logistics server of the logistics system. In the logistics system, beverages are picked according to the request data acquired by the logistics server. The request data specifies the product code and quantity of beverages to be placed in each container. Therefore, by carrying out the picking work according to the specifications of the request data, beverages to be replenished in each vending machine 5 are placed in containers separately for each vending machine 5.

When the logistics system acquires the request data, it creates a slip based on the request data. Specifically, the logistics system creates a slip for each container, which is then printed or output by an operator at the logistics base 2 as appropriate. Each slip contains information specifying the contents of the picking work, and information specifying the distribution base 3 and vending machine 5 to which the container should be delivered. One container corresponds to one vending machine 5. Next, the logistics company performs picking work to place beverages stored at the logistics base 2 into containers according to the instructions on the slip. Once the picking work corresponding to the request data is completed, the logistics company delivers each container containing the beverages to the distribution base 3.

At the distribution base 3, containers delivered from the logistics base 2 are temporarily stored. Then, when preparation for delivery begins, workers at the distribution base 3 load each container onto the delivery vehicle 6. The workers then use the delivery vehicle 6 to deliver the beverages to the delivery destinations DT, and sequentially replenish the beverages into the vending machines 5 at each delivery destination DT. At this time, since the containers are differentiated for each vending machine 5, the workers only need to unload the containers that correspond to the vending machines 5 at each delivery destination DT. This makes the replenishment work at the delivery destinations DT more efficient. However, depending on the facilities and location of the logistics base 2, the picking work may be performed at the distribution base 3, or the workers may pick the beverages from the delivery vehicle 6 at the delivery destination DT.

To execute the above-mentioned operation procedure, the linking server 31 accesses the delivery destination database, the product database, and the performance database to read or write various data. Each database can be provided in the linking system 30, but may also be provided outside the linking system 30.

The delivery destination database contains various information related to the delivery destination. As an example, the delivery destination database holds a group of records in which various information related to the delivery destination, such as the name of the delivery destination, the delivery route, and the delivery schedule, is recorded in association with a unique vending machine code for each delivery destination. The delivery route is information that specifies which delivery route RT of which delivery base 3 the delivery destination is included in. The delivery schedule is information that specifies the time of delivery to the delivery destination. For example, the delivery schedule includes information that specifies an irregular delivery date, or daily or a specific day of the week each week, or a delivery time period, etc. Furthermore, the delivery destination database may include information indicating details of the vending machine 5 installed at the delivery destination, such as information indicating the model and number of columns of the vending machine 5.

The product database holds a group of records in which various information about beverages is recorded in association with a unique product code for each beverage. Information about beverages includes product attributes such as product name, volume, and weight. The performance database contains performance data that is acquired by the linkage system 30 at any time.

[Support system configuration]
Next, the support system 100 will be described with reference to Fig. 2. As described above, the support system 100 includes a demand forecasting system 10, an optimization system 20, and a linkage system 30. The demand forecasting system 10 includes a prediction server 11, and may further include a client terminal (not shown) that inputs or outputs various data. In addition, the optimization system 20 includes a recommendation server 21, and the linkage system 30 includes a linkage server 31.

The prediction server 11 has a prediction control unit 12, the recommendation server 21 has a recommendation control unit 22, and the linking server 31 has a linking control unit 32. Each control unit is configured as a computer that combines a processor that executes various calculation processes and operation controls according to a specified program, internal memory required for the operation of the processor, and other peripheral devices. Furthermore, each control unit works together to function as the computer of the support system 100.

The processor is, for example, a CPU (Central Processing Unit) or an MPU (Micro-Processing Unit), and controls the entire device and various processes based on a program stored in the memory as a storage unit. The memory includes a RAM (Random Access Memory), which is a system work memory for the processor to operate, as well as storage devices such as a ROM (Read Only Memory), HDD (Hard Disc Drive), and SSD (Solid State Drive) that store programs and system software. In this embodiment, the CPU can perform various calculations, control, and judgment processing operations according to the program stored in the ROM or HDD. In addition, an operation unit including a keyboard or various switches for inputting predetermined commands and data is connected to each control unit via a wired or wireless connection. A display unit for displaying the input state, setting state, measurement results, and various information of the device is connected to each control unit via a wired or wireless connection. The control unit can also perform control according to a program stored in a portable recording medium such as a CD (Compact Disc), DVD (Digital Versatile Disc), CF (Compact Flash) card, or USB (Universal Serial Bus) memory, or in an external storage medium such as a server on the Internet.

Each of the prediction server 11, the recommendation server 21, and the linking server 31 has a storage unit (not shown) as a computer-readable non-transitory recording medium. The storage unit of the prediction server 11 stores the prediction program PG1, the storage unit of the recommendation server 21 stores the recommendation program PG2, and the storage unit of the linking server 31 stores the linking program PG3. The prediction program PG1, the recommendation program PG2, and the linking program PG3 work together to function as a support program. To this end, the prediction program PG1 causes the computer to function as the prediction unit 13, the selection unit 14, and the output unit 15. The recommendation program PG2 also causes the computer to function as the demand prediction unit 23 and the recommendation unit 24. The linking program PG3 also causes the computer to function as the acquisition unit 33, the judgment unit 34, the plan creation unit 35, and the linking unit 36. In other words, each unit is a logical device realized by a combination of computer hardware and software. Furthermore, the prediction control unit 12 and the recommendation control unit 22 may each have an acquisition unit that acquires performance data, etc. In this case, the respective acquisition units are operated by the prediction program PG1 or the recommendation program PG2.

The acquisition unit 33, which serves as an acquisition means, acquires recommendation information from the recommendation server 21, including product identification information that identifies the recommended products recommended to be stocked in the vending machine 5 and predicted demand quantity information for each recommended product. That is, the acquisition unit 33 acquires the recommendation information generated by the demand forecasting unit 23 from the recommendation server 21. The recommended products are selected from a group of products that includes unstocked products that are not stocked in the vending machine 5. The acquisition unit 33 then stores the acquired recommendation information in a storage unit (not shown). As an example, the product identification information is the product name, or a product code that is unique for each product. Furthermore, the demand quantity information is the number or weight as the predicted sales quantity, and may also include the volume of the column corresponding to the sales quantity.

The judgment unit 34, which serves as a judgment means, judges whether or not the placement of the recommended product will result in a replacement of the products contained in the vending machine 5 with the recommended product. The replacement may include replacement with an unstored product, as well as replacement with a product of the same type as the product already stored. For example, when a product of the same type is replaced, the number of columns that store the product will increase.

The plan creation unit 35, which serves as a plan creation means, creates a replenishment plan corresponding to the replenishment quantities of recommended products. As an example, the replenishment plan includes product codes and replenishment quantities linked to the column numbers of each column of each vending machine 5. The plan creation unit 35 also acquires the replenishment quantities calculated by the selection unit 14 from the prediction server 11. Furthermore, the plan creation unit 35 may create a replenishment plan by adding information about the work time required for replenishment. As an example, the work time information is included in the replenishment plan while being linked to the vending machine code of each vending machine.

The collaboration unit 36, which serves as a collaboration means, causes the selection unit 14 to calculate the replenishment quantity of the recommended product. As an example, the collaboration unit 36 causes the selection unit 14 to calculate the replenishment quantity of the recommended product by sending a calculation command together with the recommendation information to the prediction server 11. The collaboration unit 36 also acquires a replenishment target group that is a target for replenishing beverages from the prediction server 11 and transfers it to the recommendation server 21. The collaboration unit 36 then causes the demand forecasting unit 23 to generate recommendation information for the replenishment target group. That is, the collaboration unit 36 causes the demand forecasting unit 23 to predict the sales quantity of the recommended product for the replenishment target group. As an example, the collaboration unit 36 causes the demand forecasting unit 23 to predict the sales quantity by sending a prediction command together with the performance data to the recommendation server 21. Specifically, when the determination unit 34 determines that a replacement will occur, the collaboration unit 36 transfers the recommendation information acquired by the acquisition unit 33 to the prediction server 11 together with the calculation command.

The linking unit 36 also selects a reflection target from the group of replenishment targets on which the recommended information is to be reflected. The linking unit 36 then causes the selection unit 14 to calculate the recommended products and product replenishment quantities contained in the multiple columns of the reflection target. For example, the linking unit 36 selects the reflection target according to a predetermined selection condition described below. The linking unit 36 then passes the recommendation information, including information (e.g., a vending machine code) that identifies the vending machine 5 that constitutes the reflection target, to the prediction server 11.

The selection unit 14, which serves as a calculation means, acquires the recommendation information acquired by the acquisition unit 33 from the linking server 31. The selection unit 14 then calculates the replenishment quantities of the recommended product and other products to be stored in at least one vending machine 5 based on the recommendation information. The selection unit 14 repeats the calculations until the calculations for the multiple target vending machines 5 are completed. Specifically, the selection unit 14 receives recommendation information and a calculation command from the linking server 31 when a replacement of a product with a recommended product occurs. The selection unit 14 then calculates the replenishment quantities of the recommended product based on the recommendation information. Furthermore, the selection unit 14 may determine a column configuration indicating which product is to be stored in which column.

The prediction unit 13 also predicts the sales volume of each vending machine 5. As an example, the prediction unit 13 predicts the sales volume of each product for each vending machine 5 using a learning model constructed based on learning data including the sales history of products in multiple vending machines 5 and associated data related to the sales history.

The selection unit 14, which serves as a selection means, selects a replenishment target group, which is a plurality of vending machines 5 to be replenished, from among the vending machine group consisting of all vending machines 5, based on the sales volume predicted by the prediction unit 13, taking into consideration the cost required for replenishment (delivery cost), and calculates the required number of replenishments. For example, in a vending machine 5 with a large daily sales volume and a large impact on profits due to lost sales opportunities caused by stockouts, profits can be improved by preventing stockouts. On the other hand, in a vending machine 5 with a small daily sales volume and a small impact on profits due to lost sales opportunities, profits can be improved by reducing delivery costs and replenishment costs. In other words, if a vending machine 5 with a small impact is replenished many times, the increased cost due to replenishment exceeds the profit from sales. Therefore, the selection unit 14 determines, for example, a delivery date as the delivery timing at which the total cost obtained by adding up the delivery cost, replenishment cost, and loss cost as the amount of loss due to lost sales opportunities is minimized. The selection unit 14 then calculates the replenishment quantity required until the next delivery date after replenishment.

The output unit 15 outputs the replenishment quantity calculated by the selection unit 14 to the linking server 31. Specifically, the output unit 15 transmits to the linking server 31 the vending machine code that identifies each vending machine 5 included in the replenishment target group, the changed column configuration of the product, and the replenishment quantity of the product. When the selection unit 14 calculates the replenishment quantity, the output unit 15 may transmit the replenishment quantity to the linking server 31. Alternatively, when the output unit 15 receives a command to transmit the replenishment quantity from the linking server 31, the output unit 15 may transmit the replenishment quantity to the linking server 31. Furthermore, when the selection unit 14 requests a column configuration, the output unit 15 outputs the column configuration in addition to the replenishment quantity to the linking server 31.

The demand forecasting unit 23 as a generating means generates the recommendation information based on the sales record of the product group that can be stored in the vending machine 5. That is, the recommendation unit 24 generates the recommendation information based on the sales record of the product group including the products stored in the vending machine 5 and the unstored products not stored in the vending machine 5. As an example, the demand forecasting unit 23 generates the recommendation information including the demand quantity information using a learning model constructed based on the learning data including the sales record of the products in multiple vending machines 5 and the related data related to the sales record. Then, the demand forecasting unit 23 stores the recommendation information in a storage unit (not shown). Furthermore, the demand forecasting unit 23 may further predict an optimal column configuration indicating which recommended products and other products will be stored in which columns. In this case, the demand forecasting unit 23 stores the recommendation information including the column configuration in the storage unit.

Furthermore, the demand forecasting unit 23 may obtain information identifying the vending machines 5 that constitute the replenishment target group from the linking server 31, and generate recommendation information targeting the replenishment target group. The demand forecasting unit 23 may generate recommendation information to include forecast information indicating a change in sales performance when the recommendation information is reflected. Selection conditions for the linking unit 36 to select the above-mentioned reflection target may be set in association with the forecast information. As an example, the forecast information is the amount of profit taking into account the increase or decrease in sales, replenishment costs, and changes in delivery costs, or the increase or decrease rate of profit, and the increase or decrease in coverage rate. Furthermore, the demand forecasting unit 23 may predict the optimal column configuration for each vending machine 5 based on the predicted sales volume. As an example, the demand forecasting unit 23 predicts a column configuration to improve profits or improve coverage rate.

The recommendation unit 24 transmits the recommendation information generated by the demand forecasting unit 23 to the linking server 31. Specifically, the recommendation unit 24 outputs product codes linked to column numbers as a column configuration, and outputs the recommendation reason and predicted sales quantity for each product code, and outputs the monetary change in profit for each vending machine 5 linked to the vending machine code. In addition, when the demand forecasting unit 23 generates recommendation information, the recommendation unit 24 may transmit the recommendation information to the linking server 31. Alternatively, when the recommendation unit 24 receives a command to transmit recommendation information from the linking server 31, it may transmit the recommendation information to the linking server 31.

[Plan creation flow]
The creation of a product replenishment plan will be described with reference to FIG. 3. The selection unit 14 of the prediction server 11 selects a replenishment target group consisting of multiple vending machines 5 that require delivery for replenishment (S301). Then, the selection unit 14 obtains a column configuration after the change of the products to be put into the replenishment target group, and calculates the replenishment quantity of the products to be put in (S302). At this time, the selection unit 14 obtains a column configuration based on the prediction of the sales quantity of each vending machine 5 by the prediction unit 13, and calculates the replenishment quantity. Therefore, in the obtained column configuration, even if the products contained in the vending machine 5 run out, products that are not contained therein will not be added. As an example, the selection unit 14 obtains a delivery timing and a column configuration and calculates the replenishment quantity so that the costs of delivery, replenishment, and collection are minimized. For example, the selection unit 14 obtains a delivery timing and a column configuration and calculates the replenishment quantity so that the sum of the loss due to the product being sold out and the cost required to visit the vending machine 5 and replenish the products is minimized. As an example, the loss due to a product selling out is the difference between the predicted sales amount if the product does not sell out and the sales amount if the product sells out, and decreases as the replenishment interval becomes shorter. Also, the cost required for replenishment (e.g., amount) increases as the replenishment interval becomes shorter. Therefore, when a graph is drawn with the replenishment interval on the horizontal axis and the amount on the vertical axis, the point where the lines indicating the loss due to the product selling out and the cost required for replenishment intersect is the desired replenishment interval. Then, the delivery timing can be determined based on the desired replenishment interval.

Then, the output unit 15 of the prediction server 11 outputs the replenishment target group, the delivery timing, the changed column configuration of the product, and the replenishment quantity of the product. As an example, as shown in FIG. 4, the output unit 15 outputs to the linking server 31 a record in which a vending machine code identifying the replenishment target group is linked to a column number, and a product code and product name corresponding to the column number. The record includes the replenishment quantity of the product linked to the product code corresponding to the column number. Furthermore, when the column configuration is changed, the product code linked to each column number is shown as the changed column configuration. Furthermore, the record may include the delivery timing of each vending machine 5. Note that in FIG. 4 to FIG. 9, the product code and product name before the change are included in the record, but both can be omitted.

In the example of FIG. 4, there is no change in the column configuration for the vending machine 5 (hereinafter also referred to as vending machine MC1) identified by the vending machine code MC1. On the other hand, in the vending machine 5 (hereinafter also referred to as vending machine MC2) identified by the vending machine code MC2, the column into which "X Water" (hereinafter also referred to as product PC5) identified by the product code PC5 is inserted has been changed from column 2 to column 3. Furthermore, the column into which "X Juice" (hereinafter also referred to as product PC6) identified by the product code PC6 is inserted has been changed from column 3 to column 4. Furthermore, the column into which "X Energy" (hereinafter also referred to as product PC7) identified by the product code PC7 is inserted has been changed from column 4 to column 2.

In addition, in vending machine 5 (hereinafter also referred to as vending machine MC3) identified by vending machine code MC3, the column into which "X Black Tea" (hereinafter also referred to as product PC2) identified by product code PC2 is dispensed has been changed from column 4 to column 5. Furthermore, instead of "X Coffee" (hereinafter also referred to as product PC1) identified by product code PC1, "X Green Tea" (hereinafter also referred to as product PC4) identified by product code PC4 is dispensed into column 4. The deliverer collects product PC1 that is replaced with product PC4.

Then, the collaboration unit 36 of the collaboration server 31 acquires the replenishment target group output by the prediction server 11 and passes it to the recommendation server 21. As an example, the collaboration unit 36 transmits multiple vending machine codes that identify the replenishment target group to the recommendation server 21 along with a prediction command. Then, the demand prediction unit 23 of the recommendation server 21 predicts the sales volume for each of the replenishment target groups in a specified period from among the product groups that include the products that have been added to each of the replenishment target groups and the uncontained products that have not been added (S303), and generates recommendation information. For example, the specified period is four weeks. Alternatively, the specified period may be one week, two weeks, or three weeks. Note that the recommendation information may be created by an operator.

The demand forecasting unit 23 also predicts the optimal column configuration for each vending machine 5 based on the predicted sales volume. Furthermore, the demand forecasting unit 23 may predict the optimal column configuration for each vending machine 5 taking into account the target coverage rate of the product in addition to the predicted sales volume. The recommendation unit 24 of the recommendation server 21 then outputs the column configuration, the recommendation reason, the sales volume, and the price change. As an example, the recommendation unit 24 outputs a record that links the vending machine code with the column number, the product code and product name corresponding to the column number, the sales volume, and the price change, as shown in FIG. 5. If there is a recommended product, the record includes the recommendation reason linked to the product code. The sales volume is not limited to an integer, and in the example of FIG. 5, the sales volumes of 1.5 bottles and 3.5 bottles are included in the record. Note that as a result of the prediction and optimization by the demand forecasting unit 23, there may be cases where the products contained at the time of prediction are maintained and not changed.

In the example of FIG. 5, it is recommended that vending machine MC1 replace contained products with uncontained products. Specifically, it is recommended that products PC7 and PC6 be contained as uncontained products. In other words, it is recommended that product PC5 and "X Cider" identified by product code PC3 (hereinafter also referred to as product PC3) be replaced with uncontained products. Furthermore, in the record, increasing profits is listed as a reason for recommending the product replacement. Note that the recommendation reason included in the record may be an identification code or identification number assigned to each recommendation reason, or the like.

The record also includes information indicating the change in amount that will increase or decrease if changes are made in accordance with the recommendation, as a change in amount associated with an increase or decrease in profit. Specifically, the record shown in FIG. 5 includes, for vending machine MC1, an amount of 1,000 yen that will increase in response to increased profits. Furthermore, for vending machine MC2, the record includes an amount of 100 yen that will increase in response to increased profits. Also, for vending machine MC3, the record includes an amount of 50 yen that will decrease due to product replacement.

It is also recommended that vending machines MC2 and MC3 be replaced with products that are not yet contained. Specifically, it is recommended that vending machine MC2 be replaced with product PC1 and contained in vending machine MC2. It is also recommended that vending machine MC3 be replaced with product PC5 and contained in vending machine MC3. The reason for recommending the replacement of vending machine MC3 is to achieve the target coverage rate. The coverage rate is calculated for each product, and can be obtained by dividing the total number of vending machines 5 by the total number of vending machines 5 that contain that product and multiplying the result by 100.

The linking server 31 obtains a record including the product code of the recommended product, etc., as recommendation information from the recommendation server 21. The linking unit 36 then selects a vending machine 5 to which the recommendation information is to be reflected (S304). Specifically, the judgment unit 34 of the linking server 31 judges whether or not replacement of an uncontained product is recommended for the vending machine 5 to be judged (S305). If replacement is not recommended, the linking unit 36 does not reflect the recommendation information and does not treat the vending machine 5 that is the subject of the judgment process as a target for reflection (S306).

On the other hand, if replacement is recommended, the linking unit 36 selects the reflection target according to the selection condition. Specifically, the linking unit 36 determines whether the reason for recommendation is profit improvement or target achievement of the coverage rate (S307). Then, if the reason for recommendation is target achievement, the linking unit 36 selects the reflection target according to a predetermined amount set as a selection condition in association with the increase or decrease in profit. Specifically, the linking unit 36 determines whether the increase amount is equal to or greater than a predetermined amount (S309). As an example, the predetermined amount is 1,000 yen. If the increase amount is equal to or greater than the predetermined amount, the linking unit 36 selects the reflection target according to the selection condition. In addition, if the increase amount is less than the predetermined amount, the linking unit 36 does not reflect the recommendation information (S306). In addition, the selection condition in S309 may be any index related to profit improvement, such as an improvement rate.

When there is a reflection target, the linking unit 36 passes the vending machine code of the reflection target to the prediction server 11 as information identifying the reflection target. The linking unit 36 also passes the product codes of each column linked to the vending machine code and the sales quantity corresponding to the product code to the prediction server 11. Here, the sales quantity of the stored products that are not replaced may also change due to the purchase of a newly introduced recommended product. Therefore, the linking unit 36 passes the sales quantities of all products in the reflection target, not just the recommended products, to the prediction server 11. As an example, the linking unit 36 outputs a record in which the column number, the product code and product name corresponding to the column number are linked to the vending machine code of the reflection target, as shown in FIG. 6. Furthermore, the record includes the column configuration and sales quantity of the products in the reflection target.

The record in FIG. 6 includes the column configuration and sales quantities of products in vending machines MC1 and MC3 as the reflection targets. That is, vending machine MC2 is not included in the reflection targets because the amount of increase is less than a predetermined amount. The selection unit 14 then recalculates the replenishment quantities after replacement based on the sales quantities (S310). At this time, the selection unit 14 calculates the replenishment quantities of recommended products and also recalculates the replenishment quantities of contained products that are not replaced. Alternatively, the linking unit 36 may not include sales quantities other than the recommended products in the reflection information. The selection unit 14 may also calculate only the replenishment quantities of recommended products. In these cases, the replenishment quantities of contained products that have already been calculated are not changed.

The output unit 15 outputs the column configuration of the reflected product after the change and the refill quantity of the product including the recommended product. Then, the planning unit 35 of the linkage server 31 acquires information such as the refill quantity after the change from the output unit 15. As an example, the output unit 15 outputs a record in which the column number, the product code corresponding to the column number, and the product name are linked to the vending machine code of the reflected product, as shown in FIG. 7. Furthermore, the record includes the column configuration and refill quantity of the product in the reflected product. In the example of FIG. 7, the refill quantity of product PC1 has decreased from 10 to 8, and the refill quantity of product PC4 has decreased from 8 to 7, compared to FIG. 4. Furthermore, the refill quantity of product PC2 has increased from 5 to 6. Furthermore, products PC7 and PC6 have been added as recommended products, and the refill quantities are 15 and 8, respectively. Furthermore, for the vending machine MC3, the refill quantities of products PC4, PC3, and PC2 have increased or decreased, and product PC5 has been added as a recommended product.

The plan creation unit 35 combines the acquired data to create a replenishment plan including the replenishment quantities of the recommended product and the replenishment quantities of other products (S311), and the process ends. As an example, as shown in FIG. 8, the plan creation unit 35 creates a record that links a vending machine code with a column number and a product code and product name corresponding to the column number. Furthermore, the record includes the column configuration and replenishment quantity of each product. This makes it possible to create an optimal replenishment plan by combining the calculation results of the replenishment quantity based on the past sales performance of each vending machine 5 with the forecast results of demand based on the past sales performance of all vending machines 5. Furthermore, by using the recommended information, a replenishment plan can be created using the forecast results of demand for unstocked products for which there is no sales performance.

Furthermore, the plan creation unit 35 may calculate the work time required for replenishing and include it in the record to create a replenishment plan. As an example, the plan creation unit 35 creates a replenishment plan as shown in FIG. 9. In FIG. 9, the work time of the vending machine MC1 is 776 seconds, including a removal time of 300 seconds, a new input time of 300 seconds, a replenishment time of 44 seconds, and a loading time of 132 seconds. The work time of the vending machine MC2 is 336 seconds, including a removal time of 120 seconds, a new input time of 120 seconds, a replenishment time of 24 seconds, and a loading time of 72 seconds. The work time of the vending machine MC3 is 236 seconds, including a removal time of 60 seconds, a new input time of 60 seconds, a replenishment time of 29 seconds, and a loading time of 87 seconds. For example, the plan creation unit 35 calculates the removal time and the new input time as taking 60 seconds per column. The plan creation unit 35 also calculates the replenishment time as taking 1 second per bottle, and the loading time as taking 3 seconds per bottle.

Then, the plan creation unit 35 sends the created replenishment plan to the delivery server of the delivery optimization system. The delivery server creates a visit plan based on the acquired replenishment plan. The delivery server also sends the visit plan to each base, or creates a form that prints the visit plan. Alternatively, the visit plan may be created manually. In this case, the worker stores the created visit plan in the memory unit of the delivery server.

The link server 31 may also refer to the delivery schedule in the delivery destination database and pass on to the delivery optimization system the delivery destination DT to be selected as the next delivery destination from among the delivery destinations DT within the delivery route RT.

According to the support system 100 of this embodiment, an optimal replenishment plan can be created by combining the calculation results of the replenishment quantity based on the past sales performance of each vending machine 5 with the demand forecast results based on the past sales performance of all vending machines 5. Furthermore, by using the recommendation information, a replenishment plan can be created using the demand forecast results for unstocked products with no sales performance.

The present invention has been described above with reference to each embodiment, but the present invention is not limited to the above-mentioned embodiments. Inventions that have been modified without going against the present invention, and inventions equivalent to the present invention, are also included in the present invention. Furthermore, each embodiment and each modified form can be appropriately combined without going against the present invention.

For example, each functional unit can be provided in either the prediction server 11, the recommendation server 21, or the linking server 31. As an example, at least one of the prediction unit 13 and the selection unit 14 may be provided in the recommendation server 21 or the linking server 31. In addition, the demand prediction unit 23 may be provided in the prediction server 11 or the linking server 31.

5: Vending machine 10: Demand forecasting system 12: Forecasting control unit 13: Forecasting unit 14: Selection unit 20: Optimization system 22: Recommendation control unit 23: Demand forecasting unit 30: Linkage system 32: Linkage control unit 33: Acquisition unit 34: Judgment unit 35: Plan creation unit 36: Linkage unit PG1: Forecasting program PG2: Recommendation program PG3: Linkage program

Claims (11)

1. A support system for calculating a replenishment quantity of a product stored in each column of a vending machine having a plurality of columns based on a predicted demand for the product, and supporting creation of a replenishment plan for the product based on the calculated replenishment quantity,
an acquisition means for acquiring recommendation information including product specification information that specifies recommended products that are selected from a group of products including unstored products that are not stored in the vending machine and that are recommended to be stored in the vending machine, and predicted demand quantity information for each recommended product;
A calculation means for calculating a replenishment quantity of the recommended product to be accommodated in the vending machine based on the acquired recommendation information ;
A selection means for selecting a replenishment target group to be replenished from a vending machine group is provided,
The calculation means calculates the replenishment quantities of the products to be stored in each column of the selected replenishment target group, and calculates the replenishment quantities of the recommended products and products to be selected from the replenishment target group and stored in the multiple columns of the reflection target in which the recommendation information is reflected, in the support system. The support system according to claim 1, further comprising a plan creation means for creating the replenishment plan corresponding to the replenishment quantity. a determination unit for determining whether or not the recommended product will be replaced with the product contained in the vending machine by containing the recommended product;
The support system according to claim 2 , wherein the calculation means calculates the replenishment quantity based on the recommendation information when the replacement occurs. a generating unit that generates the recommendation information based on a sales record of the product group,
The assistance system according to claim 1 , wherein the acquisition unit acquires the generated recommendation information. The support system according to claim 4, wherein the generating means generates the recommendation information using a learning model constructed based on learning data including product sales records of multiple vending machines and associated data related to the records. The support system according to claim 4 or 5, further comprising: a linking means for selecting the reflection target from the group of replenishment targets, and causing the calculation means to calculate the replenishment quantities of the recommended products and the products. The support system according to claim 6, wherein the linking means selects the reflection target according to a predetermined selection condition. The generating means generates the recommendation information so as to include prediction information indicating a change in sales performance of the vending machine when the recommendation information is reflected,
The assistance system according to claim 7 , wherein the selection condition is set in association with the prediction information. The support system according to claim 2, wherein the plan creation means creates the replenishment plan by adding information about the work time required for replenishment. 1. An assistance program for an assistance system including a computer, the assistance program comprising: a computer that calculates a replenishment quantity of a product based on a predicted demand for the product in each column of a vending machine having a plurality of columns; and assists in creating a replenishment plan for the product based on the calculated replenishment quantity, the computer comprising:
an acquisition means for acquiring recommendation information including product specification information that specifies recommended products that are selected from a group of products including unstored products that are not stored in the vending machine and that are recommended to be stored in the vending machine, and predicted demand quantity information for each recommended product;
A calculation means for calculating a replenishment quantity of the recommended product to be accommodated in the vending machine based on the acquired recommendation information ;
functioning as a selection means for selecting a replenishment target group from a vending machine group ;
The calculation means calculates the replenishment quantities of the products to be contained in each column of the selected replenishment target group, and calculates the replenishment quantities of the recommended products and products to be contained in the multiple columns of the reflection target selected from the replenishment target group and in which the recommendation information is reflected, in an assistance program. 1. A control method for a support system for calculating a replenishment quantity of a product stored in each column of a vending machine having a plurality of columns based on a predicted demand for the product, and supporting creation of a replenishment plan for the product based on the calculated replenishment quantity, comprising:
Select a replenishment target group from the vending machine group,
obtaining recommendation information including product identification information that identifies recommended products that are selected from a group of products including unstored products that are not stored in the vending machine and that are recommended to be stored in the vending machine, and predicted demand quantity information for each recommended product;
Calculate the replenishment quantity of the product to be accommodated in each column of the selected replenishment target group;
A control method for calculating the replenishment quantities of the products to be contained in the multiple columns of the reflection target selected from the replenishment target group and in which the recommendation information is reflected, and calculating the replenishment quantities of the recommended products to be contained in the multiple columns of the reflection target based on the obtained recommendation information.

Images