Information processing system

Abstract

To provide an information processing system capable of efficiently arranging a plurality of objects on a loading table.SOLUTION: An information processing system comprises a control unit. The control unit is configured to: acquire object information including quantity information, size information and weight information on objects to be loaded onto a loading table; classify the objects to be loaded onto the loading table into one or more groups of objects whose differences in size and weight are equal to or less than thresholds; calculate an oligopolistic state of the objects; classify the loading table into one of a plurality of categories to which preliminarily prepared score parameter sets are made to correspond based upon the quantity information and oligopolistic state of the objects to determine the category of the loading table; and use the score parameter set made to correspond to the determined category to determine arrangement positions on the loading table for the objects to be loaded.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to an information processing system relating to palletizing. [Background technology]

[0002] For example, in a shipping base that is responsible for stocking and shipping multiple products such as household goods such as laundry detergent and fabric softener, multiple identical products are stored in one cardboard box and stored in a warehouse, and the corresponding cardboard box is shipped based on an order from the shipping destination. In this shipping process, the cardboard box that is taken out of the warehouse based on the order is transported from the warehouse to a loading location by a belt conveyor or the like, and then loaded onto a pallet at the loading location. In general, since there are a wide variety of products, the cardboard boxes that contain the products have different dimensions, weight, load-bearing values, etc. depending on the type of product. In addition, although the type and number of cardboard boxes (products) to be shipped are known in advance from the order from the shipping destination, the order in which the cardboard boxes are transported from the warehouse to the loading location is not determined, and there are cases in which the order is not known until just before loading onto the pallet. In such cases, it is common for a worker to instantly determine a stable arrangement while looking at the dimensions and type of products of the cardboard boxes that are transported in sequence, and load them on the pallet. This work requires skill, and efficient loading is difficult for workers who lack skill.

[0003] Patent Document 1 describes a method of virtually dividing a stowage area to set partial areas based on the size of the items, calculating virtual height data for each partial area weighted by the current height of items already loaded, and determining the stowage position of the items to be loaded based on the virtual height data. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2021-113110 A Summary of the Invention [Problem to be solved by the invention]

[0005] It is desirable to efficiently arrange a plurality of objects, such as cardboard boxes of different sizes, whose sequential transport order is unknown, on a loading platform such as a pallet.

[0006] An object of the present invention is to provide an information processing system capable of efficiently arranging a plurality of objects on a loading platform. [Means for solving the problem]

[0007] An information processing system according to one embodiment of the present invention is an information processing system for determining a placement position of an object when it is loaded onto a loading platform, the information processing system including a control unit, which acquires object information including number information, dimension information, and weight information of the objects to be loaded onto the loading platform, classifies the objects to be loaded onto the loading platform by using the object information to classify objects having a difference in dimension and weight equal to or less than a threshold into the same group to create one or more groups, calculates an object oligopoly state, which indicates the ratio of the number of objects belonging to the group to the total number of objects, classifies the loading platform into one of a plurality of categories to which a score parameter set representing a collection of weight coefficients of a plurality of loading evaluation functions prepared in advance is associated based on the object number information and the oligopoly state, determines a category of the loading platform, and determines a placement position of the object to be loaded onto the loading platform using the score parameter set associated with the determined category.

[0008] An information processing method according to one embodiment of the present invention is an information processing method executed by an information processing device, which obtains object information including number information, dimension information, and weight information of objects to be loaded onto a loading platform, and uses the object information to classify the objects to be loaded onto the loading platform into groups in which objects whose size and weight differences are below a threshold are grouped together to create one or more groups, and uses the object information to calculate an object oligopoly state, which is the ratio of the number of objects belonging to the group to the total number of objects, and based on the object number information and the oligopoly state, classifies the loading platform into one of a number of categories each associated with a score parameter set representing a collection of weighting coefficients for each of a number of loading evaluation functions prepared in advance, thereby determining the category of the loading platform, and using the score parameter set associated with the determined category, determines the placement position of the objects to be loaded on the loading platform.

[0009] A program according to one embodiment of the present invention causes an information processing device to execute the following steps: acquiring object information including information on the number, size, and weight of objects to be loaded onto a loading platform; using the object information to classify objects to be loaded onto the loading platform into one group, with objects whose size and weight differences are below a threshold, to create one or more groups; using the object information to calculate an object oligopoly state, i.e., the ratio of the number of objects belonging to the group to the total number of objects; classifying the loading platform into one of a plurality of categories, each of which is associated with a score parameter set representing a collection of weighting coefficients for each of a plurality of loading evaluation functions prepared in advance, based on the object number information and the oligopoly state, to determine the category of the loading platform; and using the score parameter set associated with the determined category to determine the placement position of the object to be loaded on the loading platform.

[0010] A recording medium according to one embodiment of the present invention is a computer-readable non-transitory recording medium having recorded thereon a program for causing an information processing device to execute the following steps: acquiring object information including number information, dimension information, and weight information of objects to be loaded onto a loading platform; using the object information to classify objects to be loaded onto the loading platform into one or more groups, with objects whose size and weight differences are below a threshold being grouped together; using the object information to calculate an object oligopoly state, i.e., the ratio of the number of objects belonging to the group to the total number of objects; classifying the loading platform into one of a plurality of categories, each of which is associated with a score parameter set representing a collection of weighting coefficients for a plurality of loading evaluation functions prepared in advance, based on the object number information and the oligopoly state, to determine the category of the loading platform; and determining the placement position of the objects to be loaded on the loading platform using the score parameter set associated with the determined category. Effect of the Invention

[0011] According to the present invention, a plurality of objects can be efficiently arranged on a loading platform. [Brief description of the drawings]

[0012] [Figure 1] This is a schematic diagram showing the shipping process in which packages (objects) transported in sequence from a product warehouse are stacked on a pallet (loading platform). [Diagram 2] 1 is a functional block diagram of an information processing system according to an embodiment of the present invention; [Diagram 3] 1 is a diagram showing a hardware configuration of an information processing device included in an information processing system according to an embodiment of the present invention. [Figure 4]FIG. 10 is a diagram explaining the basic search and wide search for extracting search cells, where (a) shows an example of a search cell extracted when one bale is placed on a pallet, and (b) shows an example of a search cell extracted when two bales are placed on a pallet. [Diagram 5] 13A shows a flow when the search method for the search cell is search method A, and FIG. 13B shows a flow when the search method is search method B. [Figure 6] (a) is a diagram explaining evaluation function x1, (b) is a diagram explaining evaluation function x2, (c) is a diagram explaining evaluation function x3, (d) is a diagram explaining evaluation function x4, and (e) to (h) are all diagrams explaining evaluation functions x5 and x6. [Figure 7] 13(a) and (b) are diagrams for explaining evaluation function x7, (c) to (e) are diagrams for explaining evaluation function x8, and (f) is a diagram for explaining evaluation function x9. [Figure 8] FIG. 13 is a diagram illustrating the evaluation function x10. [Figure 9] 13(a) to 13(c) are diagrams illustrating the evaluation function x11. [Figure 10] 13(a) and (b) are diagrams for explaining a method of creating divided areas used in evaluation using evaluation functions x12 to x14, and (b) is a diagram for explaining evaluation function x12. FIG. [Figure 11] 13(a) and 13(b) are diagrams illustrating the evaluation function x13. [Figure 12] FIG. 14 is a diagram for explaining the evaluation function x14. [Figure 13] 11 is a flow diagram of a parameter set generation process performed by a control unit of an information processing device constituting the information processing system. FIG. [Figure 14] 1 is a graph illustrating category classification according to the number of packages and HH index based on past loading performance data. [Figure 15] 13 is a graph illustrating an example of a method for selecting a score parameter set to be associated with each category. [Figure 16] 13 is a graph illustrating an example of a method for selecting a score parameter set to be associated with each category. [Figure 17]FIG. 4 is a flow diagram of a stowage process by the information processing system. [Figure 18] FIG. 31 is a top view of a group of packages in the shipping package information shown in FIG. 30 that have been stacked by the information processing system. [Figure 19] 18, where (a) is a side view of the packing group in FIG. 18 when viewed from the direction of arrow 19a, (b) is a side view of the packing group in FIG. 18 when viewed from the direction of arrow 19b, (c) is a side view of the packing group in FIG. 18 when viewed from the direction of arrow 19c, and (d) is a side view of the packing group in FIG. 18 when viewed from the direction of arrow 19d. [Figure 20] FIG. 18 is a flow diagram of a parameter set determination process in the flow of FIG. 17. [Figure 21] FIG. 18 is a flow diagram of the stowing process in the flow of FIG. 17. [Figure 22] FIG. 22 is a flow diagram of a search cell evaluation process (process A) in the flow of FIG. 21. [Figure 23] FIG. 22 is a flow diagram of the placement process (process B) in the flow of FIG. 21. [Figure 24] FIG. 22 is a flow diagram of the placement process (process C) in the flow of FIG. 21. [Diagram 25] FIG. 22 is a flow diagram of a transfer process (process D) in the flow of FIG. 21. [Figure 26] 13(a) and 13(b) are top views of a pallet illustrating how to determine the corner of a bale to be placed in a search cell. [Figure 27] 13(a) to 13(h) are diagrams illustrating arrangement patterns in which the bale is placed vertically and horizontally with each of the four corners of the bale positioned in the search cell. [Figure 28] FIG. 2 is a diagram for explaining XYZ coordinate axes set on a palette. [Figure 29] 13(a) to 13(c) are diagrams illustrating an example of constraint conditions when loading packages onto a pallet. [Diagram 30] This is the package information (shipping package information) of packages to be loaded onto a pallet. [Diagram 31] 13 is an example of work order information for bales transported from a product warehouse to a loading area. [Diagram 32] FIG. 13 is a diagram showing an example of calculating the HH index as an oligopoly state. [Diagram 33] This is a list of the contents of the evaluation functions. [Diagram 34] This is an excerpt of some of the score parameter set candidates, and is a list of score parameter set candidates with a calculated stowage success rate of 0.96 or more in a certain category. [Diagram 35] 13 is a list of the created simulation conditions. [Diagram 36] This figure explains the process for determining which package to transfer to the normal area when the package to be loaded cannot be placed on either the pallet or the temporary storage area, and the package in the temporary storage area can be placed in the normal area. [Figure 37] This figure explains the process of determining which packages to transfer preferentially from the temporary storage area to the normal area when all the packages to be loaded onto a pallet have been transported and placed in the temporary storage area. [Figure 38] 13(a) and 13(b) are diagrams for explaining a method of calculating the score of an evaluation function x1(svar). [Figure 39] FIG. 13 is a diagram illustrating an example of calculation of entropy in an oligopoly state. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] <Outline of the information processing system> Hereinafter, an information processing system 100 according to an embodiment of the present invention will be described with reference to the drawings. In the drawings, "stowage" is written as "stowage".

[0014] The information processing system 100 enables efficient stacking of one type of object or two or more types of objects with different dimensions that are transported sequentially onto a loading platform. In this embodiment, an example of shipping processing at a shipping base where multiple bales are stocked will be described. In this embodiment, the loading platform is a pallet PL, the objects placed on the loading platform are bales, and the temporary placement area is a temporary placement pallet TPL. In this embodiment, a flat pallet is used as the pallet. The bales as objects are typically cardboard boxes that contain multiple products of the same type. The dimensions, weight, load capacity, etc. of the bales vary depending on the type and number of products they contain. The bales are typically approximately rectangular parallelepipeds.

[0015] In this specification, the "bottom surface" of a bale (object) is the surface that contacts the horizontal surface when the bale is placed on the horizontal surface. Also, the "top surface" of a bale is the surface that is located above the bale when the bale is placed on the horizontal surface. If the bale is a roughly rectangular parallelepiped, the top surface is roughly parallel to the bottom surface.

[0016] As shown in Figure 28, pallet PL has a rectangular shape when viewed from above when laid flat. In pallet PL, X-axis and Y-axis are set parallel to each of two perpendicular sides, and Z-axis is set perpendicular to X-axis and Y-axis. Z-axis is parallel to the stacking direction when packages are stacked on pallet PL. In other words, pallet PL is viewed from above in the height direction Z as a plan view.

[0017] 28, with the stowage robot 9 positioned below the pallet PL as viewed from above, an i coordinate axis parallel to the Y axis and a j coordinate axis parallel to the X axis are set on the pallet PL, and in the ji coordinate, the back left corner of the pallet PL as viewed from the stowage robot 9 (the upper left corner of the pallet as viewed from the diagram) is set as the origin (0,0). The j coordinate value increases further to the right on the diagram, and the i coordinate value increases further down (from the back to the front as viewed from the stowage robot 9).

[0018] The shipping process at the shipping base involves unloading the package P, transporting the package P to a loading location, and loading the package P onto a pallet PL (hereinafter referred to as "loading process").

[0019] As shown in FIG. 1, at a product shipping base, a plurality of products are stored in a product warehouse S in a package state, and a plurality of packages P are delivered from the product warehouse S based on an order from a shipping destination (delivery of packages P). The delivered packages P are sorted for each shipping destination by a plurality of sorters 93 (931 to 93N), and then sequentially transported to a loading location 95 by a transport means such as a belt conveyor 92 (transportation to loading location of packages P). Note that, when there are many packages P to be shipped to one shipping destination, a plurality of sorters 93 may be used. Also, a sorter does not have to be provided, but providing a plurality of sorters 93 makes it possible to efficiently perform shipping work to a plurality of shipping destinations (loading onto pallets PL to be transported to each shipping destination). A plurality of packages P transported one by one by the belt conveyor 92 are loaded onto pallets PL arranged at the loading location 95 (loading process). In this embodiment, the loading process is performed by a loading robot 9.

[0020] In this embodiment, pallets PL and temporary placement pallets TPL are provided at the loading area 95. The temporary placement pallets TPL may be provided at a location other than the loading area 95, for example, adjacent to the loading area 95.

[0021] The temporary storage pallet TPL has a rectangular shape in a plan view. In this embodiment, the temporary storage pallet TPL serves as the temporary storage area 25, and more specifically, the entire loading surface of the temporary storage pallet TPL serves as the temporary storage area 25. The temporary storage area 25 is an area in which packages P are temporarily stored as necessary when loading the packages P onto the pallet PL. The temporary storage area 25 is provided in an area different from the pallet PL. In this embodiment, the position of the temporary storage pallet TPL at the loading location 95 is fixed. Note that the temporary storage area does not need to be provided on the temporary storage pallet TPL, and a predetermined area may be secured in an empty space to serve as the temporary storage area.

[0022] It is preferable to arrange the temporary placement pallet TPL and the pallet PL so that the distance between the stowage robot 9 and the pallet PL to be stowed is approximately the same as the distance between the stowage robot 9 and the temporary placement pallet TPL. This allows efficient transport of the packages P from the belt conveyor 92 to the temporary placement pallet TPL, and from the temporary placement pallet TPL to the pallet PL to be stowed. Note that, from the perspective of efficient transport of the packages P by the stowage robot 9, it is necessary to determine the temporary placement area 25 so that it is within the transportable range of the stowage robot 9.

[0023] The packages P are loaded onto the pallet PL and transported to the shipping destination. The pallet PL has a loading surface 20 on which the packages P are loaded. Information on the dimensions of the pallet PL and the height limit value when stacking the packages on the pallet PL is acquired in advance. Hereinafter, information on the pallet dimensions and the height limit value associated with each pallet is referred to as "pallet information (loading platform information)." Information on the dimensions of the pallet PL includes information on the X-axis dimension and the Y-axis dimension when the pallet PL is viewed in a plane. In this embodiment, the X-axis dimension of the pallet PL is 1100 mm, the Y-axis dimension is 1100 mm, and the height limit value is 1800 mm.

[0024] One or more pallets PL are prepared for one shipping destination. Stowage instruction information generated based on an order from the shipping destination is finalized prior to the stowage process (for example, two hours prior). The stowage instruction information includes package information (hereinafter sometimes referred to as "shipping package information") as object information to be shipped to the shipping destination. The shipping package information includes information on how many packages of which products are to be shipped (information on the number of each type of package to be shipped, dimensional information, weight information, etc.). Specifically, the shipping package information includes the package code assigned to each type of package P and the shipping quantity. The number of pallets PL required for one shipping destination is calculated based on the shipping package information and pallet information. If there are many packages to be shipped, multiple pallets PL are prepared taking into account the volume ratio and weight, and if one pallet PL is sufficient, one pallet PL is prepared.

[0025] As shown in Figure 1, when multiple pallets PL are prepared, loading of packages P onto the pallets PL is completed sequentially one pallet PL at a time. Shipping package information is also presented for each pallet. In other words, loading onto the next pallet PL does not begin until loading onto one pallet PL is completed, and packages are not loaded onto multiple pallets PL in parallel. The pallet PL onto which the packages P are to be loaded is prepared at a loading location 95 near the downstream of the belt conveyor 92. In the loading location 95, the position of the pallet PL onto which the packages P are to be loaded is referred to as the loading position 95a (see Figure 1).

[0026] When the loading process is started, first, the package P is loaded onto the first pallet 1st PL at the loading position 95a. When loading onto the first pallet 1st PL is completed, the first pallet 1st PL is moved to another location, and the second pallet 2nd PL is placed at the loading position 95a and loaded with the package P. In this manner, loading onto the pallet PL is completed one by one, and the packages are loaded. Furthermore, the work of moving the packages P onto the pallet PL or the temporary placement pallet TPL is performed one by one in the order in which the packages P are transported. In other words, the work of moving the nth (n is an integer of 1 or more) package P transported onto the pallet PL, which is the work of moving the package P onto the pallet PL, or the work of moving the package P onto the temporary placement pallet TPL, which is the work of temporarily placing the package P on the temporary placement pallet TPL, is not performed. The order in which the packages are transported is, in other words, the order in which the packages are moved. Note that there is no particular distinction between 1st PL, 2nd PL, and Last PL, and they are referred to as PL.

[0027] The order in which the packages are transported is unknown until immediately before they are loaded onto the pallet, and multiple types of packages are transported in a mixed order. Note that "transporting multiple types in a mixed order" includes cases where, for example, when two types of packages X and Y, three of each, are transported, the two types of objects are transported in a random order, such as X, Y, X, X, Y, Y, as well as cases where packages of the same type are grouped together and transported in succession, such as X, X, X, Y, Y, Y. In either case, the order in which the packages will be transported is unknown until immediately before they are loaded onto the pallet.

[0028] In the following description, an example will be given in which one pallet PL is required for one shipping destination.

[0029] In this embodiment, the shipping package information is finalized about two hours before the stowage process. The packages P are removed from the product warehouse S based on the shipping package information, but the order in which the packages P are transported to the stowage location 95 is not yet determined. The package identification sensor 8 (described later) provided near the sorter 93 detects the package codes attached to the packages P to be detected as they are transported in sequence, thereby identifying the type of package P. In this embodiment, the transport time of the packages from the sensing position of the package identification sensor 8 to the stowage position is about one minute, and the interval between the removal of the packages P is about three seconds. That is, the detection result of the package code attached to the leading (first) package P by the package identification sensor 8 can be obtained about one minute before the start of the stowage process, and the detection results of the package codes attached to the subsequent packages P can be obtained sequentially at intervals of about three seconds thereafter. In this way, by obtaining the package code information (package type information) attached to the packages P in sequence in the transport order, the package type information (package order information) of the transport order of the packages P that is successively updated can be obtained. In this embodiment, the control unit 3 of the information processing device 1 described later obtains the package code information of the package P to be loaded from the package sequence information generated by the automated warehouse control server 6, but it may also receive the information directly from the package identification sensor 8.

[0030] In this way, the transport order of the packages P is unknown until immediately before they are stowed onto the pallet PL. Note that in this embodiment, specific time values ​​have been given to explain that the shipping package information is determined well before (about two hours before) the start of the stowage process, whereas the transport order of the packages P is determined immediately before (about one minute or later: the point in time when the package identification sensor 8 detects the type of package P), but these values ​​are merely examples and are not limiting.

[0031] As shown in Fig. 30, the shipping package information includes information such as the package code, product name, package length (L), package width (W), package height (H), package number (number of packages of each type), and weight for each package. The shipping package information includes information on the number, dimensions, and weight of packages to be loaded onto a pallet PL. In this example, the packages to be loaded onto the pallet are six packages of diapers A and twelve packages of shampoo B. Fig. 31 is an example of work order information indicating the order in which each package included in the shipping package information shown in Fig. 30 was transported. In Fig. 31, "1" in "pack orientation" indicates that the arrangement pattern is horizontal, and "0" indicates that it is vertical.

[0032] Although details will be described later, in the information processing system 100, using the shipping package information (package number information, dimension information, and weight information), packages to be loaded onto a pallet to be loaded are classified into the same group if the difference in dimension and weight is less than a threshold value, and one or more groups are created. Packages that belong to the same group are called "packages of the same type," and packages that belong to different groups are called "packages of different types." "Packages of the same type" refers to "packages considered to be of the same size," and "packages of different types" refers to "packages not considered to be of the same size." Regarding "packages of the same type," in addition to packages of the same SKU, packages that are not the same SKU but are in the same group are considered to be the same type (same package). In the example of shipping package information shown in FIG. 30, a total of two groups are created: a group to which packages of diapers A belong and a group to which packages of shampoo belong.

[0033] Further, although details will be described later, the information processing system 100 performs three information processing steps (referred to as a first step, a second step, and a third step) roughly on the time axis. The first step is a process carried out in advance before actual stowage, and prepares category classification information and score parameter set information to be stored in the evaluation information DB 46 described later. The evaluation information DB 46 stores one or more score parameter sets (10 in this embodiment) associated with each category. Details of the first step will be described later with reference to Figs. 13 to 16 etc. The second step is related to the actual stowage process, and is performed after the shipped package information is finalized (in this embodiment, two hours before the actual stowage process of the packages onto the pallet). At this point, information on which packages will be stowed onto the pallet is available, but the order in which the packages will be transported is unknown. In the second step, based on the package information of the packages to be stowed onto the pallet, a category of the pallet to be stowed is determined using evaluation function information, category classification information, score parameter set information type information, etc. stored in the evaluation information DB 46. Furthermore, a score parameter set to be used when stowing the packages onto the pallet to be stowed is determined from multiple score parameter sets linked to the category. Details of the second step will be described later using the flow of FIG. 20, etc. The third step is a process performed when actually stacking the bales on the pallet. In the third step, the placement positions of the bales to be stacked are determined using the score parameter set determined in the second step, and the bales are stacked one by one. Details of the third step will be described later using the flows in Figures 21 to 25, etc.

[0034] <Overall configuration of information processing system> 2, the information processing system 100 of this embodiment includes an information processing device 1, an automated warehouse control server 6, a shipping management server 7, a package identification sensor 8, and a stowage robot 9. The information processing device 1 is a stowage server that performs processing related to stowage.

[0035] <Function block configuration of the shipping management server> The shipping management server 7 generates shipping instruction information and stowage instruction information based on an order (order data) from a shipping destination. The shipping management server 7 has a communication unit 70, a generation unit 71, and a storage unit 72. The storage unit 72 has a shipping instruction information DB 73, a stowage instruction information DB 74, and a warehouse information DB 75. DB means database.

[0036] The communication unit 70 receives orders (order data) from shipping destinations. The communication unit 70 is configured to be able to communicate with the automated warehouse control server 6, the information processing device 1, etc., via, for example, the Internet. The communication unit 70 transmits shipping instruction information to be stored in a shipping instruction information DB 73 to the automated warehouse control server 6. The communication unit 70 transmits stowage instruction information to be stored in a stowage instruction information DB 74 to the information processing device 1.

[0037] The shipping instruction information DB 73 stores, for each shipping destination, shipping instruction information generated by the generation unit 71 based on orders from the shipping destination. The shipping instruction information includes attribute information of the shipping destination such as the shipping destination name and shipping destination address, the bale code of the bale to be shipped to the shipping destination (bale to be shipped), the shipping quantity, the bale removal opening which is the position from which the bale P is removed in the product warehouse S, and other information.

[0038] The stowage instruction information DB 74 stores, for each shipping destination, the stowage instruction information generated by the generation unit 71 based on orders from the shipping destinations. The stowage instruction information includes information such as the number of pallets used for stowage, identification information for each pallet (hereinafter referred to as "pallet ID"), the pallet code of the pallet to be shipped to the shipping destination, and the shipping quantity.

[0039] The warehouse information DB75 stores information such as the type and number of bales stocked in the product warehouse S, and the location within the product warehouse S. The contents stored in the warehouse information DB75 are updated successively based on orders from the shipping destination, excluding the number of items to be shipped from the inventory currently in the product warehouse S (inventory allocation).

[0040] The generating unit 71 generates warehousing instruction information and stowage instruction information based on an order from a shipping destination. The generating unit 71 calculates the number of pallets included in the stowage instruction information based on the order from the shipping destination and the pallet information.

[0041] <Sensor for package identification> The package identification sensor 8 is provided near the sorter 93. The package identification sensor 8 is, for example, a code reader that reads a package code attached to the surface of the package P that is sorted by the sorter 93 and supplied to the belt conveyor 92. When the package P passes in front of the package identification sensor 8, the package code is read. A plurality of packages P pass in front of the package identification sensor 8 one by one in sequence. The type of package P that has passed in front of the package identification sensor 8 can be identified based on the detection result (reading result) of the package identification sensor 8. In other words, the order in which the packages P are conveyed can be identified. Note that the package identification sensor 8 is not limited to a code reader. For example, it may be a camera, and the type of package P may be identified using image information acquired by the camera. The detection result of the package identification sensor 8 is output to the automated warehouse control server 6.

[0042] <Function block configuration of the automated warehouse control server> An retrieval robot (not shown) is responsible for the task of retrieving the package P from the product warehouse S. The automated warehouse control server 6 controls the retrieval robot based on retrieval instruction information generated by the shipping management server 7, and retrieves the package P. The automated warehouse control server 6 also successively updates the package sequence information based on the detection results (package type information) of the package identification sensor 8.

[0043] As shown in FIG. 2, the automated warehouse control server 6 includes a communication unit 60, a shipping control unit 61, a sequence information generation unit 62, and a packing sequence information DB 63.

[0044] The communication unit 60 is configured to be able to communicate with the package identification sensor 8, the shipping management server 7, the information processing device 1, etc., for example, via an in-house network. The communication unit 60 receives the detection result of the package code output from the package identification sensor 8. The communication unit 60 transmits package sequence information to be stored in the package sequence information DB 63 to the information processing device 1. The communication unit 60 receives shipping instruction information from the shipping management server 7.

[0045] The delivery control unit 61 controls the delivery robot based on the delivery instruction information received from the shipping management server 7, and delivers the package P to be shipped from the product warehouse S.

[0046] The sequence information generation unit 62 generates updated packing sequence information based on the detection results successively output from the packing identification sensor 8, and stores it in the packing sequence information DB 63. The packing sequence information includes packing type information, which is the packing code detection result of the packing identification sensor 8, and packing movement work sequence information. In this embodiment, the packing sequence information is updated approximately every 3 seconds. The updated packing sequence information is successively transmitted to the information processing device 1.

[0047] <Stacking robot> In the example shown in FIG. 1, the stowage robot 9 is a vertical articulated robot having multiple links 97, multiple joints 96, and a two-jaw gripper arm 91 that serves as a gripping portion attached to the tip.

[0048] As shown in Fig. 2, the stowage robot 9 has a robot drive unit 90. The robot drive unit 90 has a motor and a drive circuit for driving the motor. The drive circuit drives the motor based on a robot control signal output from the information processing device 1. This controls the position and posture of the gripper arm 91 and the opening and closing of the two claws of the gripper arm 91. A package P is clamped and gripped by the two claws of the gripper arm 91, and is transported from the belt conveyor 92 onto the pallet PL or the temporary placement pallet TPL.

[0049] In this embodiment, the gripping unit of the stacking robot is a two-jaw gripper arm, but is not limited to this. Any gripping unit capable of gripping and transporting the package P may be selected as appropriate depending on the shape, material, size, etc. of the package to be gripped. For example, the gripping unit may be a gripper arm having three or more jaws. Also, for example, the gripping unit may be a suction pad. Note that suction pads are generally made of a flexible material and do not need to clamp the package, so they have the advantage of being less likely to damage the package. In this embodiment, a vertical articulated robot is used as an example, but a parallel link robot may be used as long as it is a robot capable of gripping and transporting the package P.

[0050] <Information processing device> <<Hardware configuration of information processing device>> As shown in FIG. 3, the information processing device (loading server) 1 includes a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, an input / output interface 15, and a bus 14 connecting these to each other.

[0051] The CPU 11 appropriately accesses the RAM 13 etc. as necessary, and performs various arithmetic processing while controlling all the blocks of the information processing device (loading server) 1. The ROM 12 is a non-volatile memory that stores firmware such as the OS program and various parameters to be executed by the CPU 11. The RAM 13 is used as a working area for the CPU 11, and temporarily stores the OS, various applications being executed, various data being processed, etc.

[0052] The input / output interface 15 is connected to a display unit 16, an operation reception unit 17, a storage unit 18, a communication unit 19, and the like.

[0053] The display unit 16 is a display device using, for example, an LCD (Liquid Crystal Display) or an OELD (Organic Electro Luminescence Display).

[0054] The operation reception unit 17 is, for example, a pointing device such as a mouse, a keyboard, a touch panel, or other input device. When the operation reception unit 17 is a touch panel, the touch panel can be integrated with the display unit 16.

[0055] The storage unit 18 is, for example, a non-volatile memory such as a hard disk drive (HDD), a flash memory (SSD; Solid State Drive), or other solid-state memory. The storage unit 18 stores the OS, various applications, various data, and the like.

[0056] The communication unit 19 is, for example, a NIC (Network Interface Card) for Ethernet or various modules for wireless communication such as wireless LAN, and is responsible for communication processing with the automated warehouse control server 6, the shipment management server 7, and the like.

[0057] Although not shown, the basic hardware configurations of the automated warehouse control server 6 and the shipping management server 7 are also substantially the same as the hardware configuration of the information processing device 1.

[0058] <<Function block configuration of information processing device>> As shown in FIG. 2, the information processing device 1 includes a communication unit 2, a control unit 3, and a storage unit 4.

[0059] [Communications Department] The communication unit 2 is configured to be able to communicate with the automated warehouse control server 6, the shipping management server 7, etc., for example, via an in-house network. The communication unit 2 corresponds to the communication unit 19 in the hardware configuration. The communication unit 2 receives stowage instruction information from the shipping management server 7. The communication unit 2 receives packing sequence information (including packing type information and packing movement work order information) from the automated warehouse control server 6. The communication unit 2 sequentially receives the packing sequence information from the automated warehouse control server 6 from approximately one minute before the start of the stowage process onto the pallet PL (approximately every three seconds in this embodiment). The communication unit 2 outputs a robot control signal generated by the control unit 3 to the stowage robot 9.

[0060] [Storage] The storage unit 4 includes a DB unit 40 and a program storage unit 49. The storage unit 4 corresponds to the storage unit 18 of the hardware configuration shown in FIG.

[0061] (DB Unit) The DB unit 40 has a package master DB 41, a stowage instruction information DB 43, a stowage information DB 44, a pallet information DB 45, an evaluation information DB 46, a stowage calculation parameter set information DB 47, and a search method information DB 48.

[0062] ((Packing Master DB)) The package master DB 41 stores, as package master information, information related to the package P for each type of package P. The types of packages P are classified according to the type and number of products contained in the package P.

[0063] The information related to the package P includes a package code given to the surface of the package P, the type of product contained in the package P (product name, etc.), the dimensions of the package P (package length (L), package width (W), package height (H)), the weight of the package P, the load limit (load capacity), etc. The package master DB 41 stores this information in association with one another.

[0064] The bale code is identification information for identifying different types of packages P, and is assigned differently for each type of package P. The bale code makes it possible to know the type of product contained in the package P. For example, the bale code is a one-dimensional code such as a barcode or a QR code (registered trademark), or a two-dimensional code.

[0065] The load limit (load capacity) is set in advance for each package P. The load limit indicates the maximum total weight of packages that can be placed on a package. The load limit can be set based on a load test, for example. Specifically, the load at which the package buckles is measured, and the load is multiplied by a predetermined safety factor to set the load limit value. The load limit may also be set based on actual stock data of packages in the product warehouse S at the shipping base.

[0066] The package master DB 41 may also have information about a regular pallet on which packages of the same stock keeping unit (SKU) are stacked. The information about the regular pallet includes the maximum number of packages that can be arranged on one pallet PL, the maximum number of packages P that can be arranged on one layer, the number of loaded layers, and the number of layers.

[0067] ((Loading Instructions Information DB)) The stowage instruction information DB 43 stores stowage instruction information received from the shipping management server 7. As described above, the stowage instruction information includes the number of pallets used for stowage, the pallet ID, shipping package information, etc. The shipping package information includes information such as the package code of the package to be shipped to the shipping destination and the shipping quantity. The control unit 3 can use the package code included in the stowage instruction information to refer to the package master information stored in the package master DB 41, and obtain information related to the dimensions, weight, loading weight limit, etc. of the package P to which the package code is assigned.

[0068] ((Loading Information DB)) The stowage information DB 44 stores stowage information. The stowage information includes information on the pallet PL (hereinafter, sometimes simply referred to as "pallet") to be stowed and on the packages P already stowed on the temporary storage pallet TPL. In detail, the stowage information DB 44 stores, for each pallet ID, position information on the pallet of the package P stowed on the pallet PL and identification information such as the package code of the package P, in association with each other. The stowage information DB 44 also stores, for the temporary storage pallet TPL, position information on the temporary storage pallet TPL of the package P temporarily placed on the temporary storage pallet TPL and identification information of the package P, in association with each other. From this information, the stacking status of the packages P in the XY plane of the pallet PL or the temporary storage pallet TPL when the pallet PL or the temporary storage pallet TPL is viewed in a plan view, and the stacking status of the packages P in the Z-axis direction (height direction) can be grasped.

[0069] The position of package P on the pallet or temporary placement pallet is expressed by the position coordinates of the center of package P in the X-axis, Y-axis, and Z-axis directions.

[0070] Hereinafter, the appearance of the entire pallet PL when viewed from above in the Z-axis direction is referred to as the surface 26, and the loading status of the packages P when viewed from above the pallet PL is referred to as the "surface status." The "surface status" includes information about the arrangement of the packages on the XY plane. The same applies to the temporary placement pallet TPL.

[0071] The stowage information includes information on the state of the surface of the pallet. Each time a package P is stowed on the pallet PL or the temporary storage pallet TPL, the state of the surface of the pallet changes, and the stowage information is successively updated. The stowage information DB 44 successively stores the updated stowage information of the pallet PL and the temporary storage pallet TPL.

[0072] ((Pallet information DB)) A pallet ID for identifying the pallet is assigned to the pallet used for stowage. The pallet information DB 45 stores pallet information (loading platform information) including information on the pallet dimensions (x-axis dimension, y-axis dimension) and height limit value for each pallet ID. The height limit value is the maximum allowable height of a group of packages on a pallet when the packages are stacked on the pallet, and in this embodiment, the height limit value is 1800 mm. The control unit 3 can obtain information on the dimensions and height limit value of the pallet PL by referring to the pallet information DB from the pallet ID included in the stowage instruction information stored in the stowage instruction information DB. All pallets PL used at a shipping base may be of the same type (flat pallets) and have the same dimensions, or the pallets may be of different types and / or dimensions.

[0073] ((Evaluation Information DB)) The evaluation information DB 46 stores evaluation function information, category classification information, and score parameter set information. Each piece of information will be described later. The category classification information and score parameter set information are calculated and stored in advance for each shipping destination.

[0074] The evaluation function information is information on an evaluation function used in the stowage calculation of package P. Stowage calculation is a calculation for determining the placement position of package P on pallet PL. The evaluation function is an evaluation index for evaluating stowage stability, space utilization rate, etc. In this embodiment, an example is given in which 14 evaluation functions (evaluation indexes) are used. Using multiple evaluation functions, stowage stability, space utilization rate, etc. are comprehensively evaluated to determine the placement position of package P to be stowed. Details of the evaluation function information will be described later, but the evaluation function information includes evaluation content information, positive / negative information of score parameters, information on values ​​to be multiplied for normalization, etc. for each evaluation function (see FIG. 33), and the score of each evaluation function is calculated using these.

[0075] Here, regardless of whether or not bales are arranged, the pallet viewed from above is virtually divided into cells of 10 mm square by lines in a lattice pattern (in this embodiment, a square lattice pattern). The lines include a lattice pattern consisting of a plurality of vertical lines extending in the X-axis direction and arranged at equal intervals in the Y-axis direction, and a plurality of horizontal lines perpendicular to the vertical lines, extending in the Y-axis direction and arranged at equal intervals in the X-axis direction, as well as lines forming a rectangular outline (periphery) of the pallet viewed from above. The lines forming the outline include two vertical lines and two horizontal lines corresponding to the four sides of the rectangle, and the vertical lines and horizontal lines are positioned perpendicular to each other. In addition, the lines forming the outline are positioned perpendicular to the vertical and horizontal lines that are perpendicular to each other. In other words, the vertical lines are positioned perpendicular to the horizontal lines or perpendicular to the horizontal lines, and the horizontal lines are positioned perpendicular to the vertical lines or perpendicular to the vertical lines. In this embodiment, the "cells" are used as the basis for determining the arrangement position of the bales, and the arrangement is considered so that the corners of the bales coincide with the cells, as will be described in detail later. The size of the cells is not limited to 10 mm square, but can be set appropriately depending on the dimensions of the pallet, the dimensions of the bale, etc.

[0076] Although details will be described later, the present invention involves extracting search cells, which are candidate cells for the placement position of the package to be stowed. The corners of the packages are placed in the search cells, and as described later, there are up to eight possible placement cases. For all search cells, all possible placement cases are scored and evaluated using an evaluation function. Then, the search cell with the highest score (the search cell with the highest evaluation) is determined as the placement position of the package to be stowed. In calculating the score of the search cell, a weighting coefficient (also referred to as a "score parameter") set for each evaluation function is used. Note that since there are multiple evaluation functions, in the following, the weighting coefficients set for each corresponding evaluation function are collectively referred to as a score parameter set.

[0077] The inventors discovered that there is an appropriate score parameter for the evaluation function used in calculating the score of a search cell to determine the placement position of a package on a pallet during the loading process, depending on the relationship between the number of packages to be loaded onto a pallet and the package oligopoly state, and proposed this invention.

[0078] The inventors have found that there are more appropriate score parameters and search methods for search cells even in classifications according to the type and number of bales to be shipped. Therefore, preferably, instead of preparing a single score parameter (score parameter set) for each relationship between the number of bales and the oligopoly state of the bales, multiple appropriate score parameters (score parameter sets) are prepared. Then, dummy data (referred to as "dummy transport order data" or "dummy transport pattern") for the transport order of the bales is created based on the number and type of bales to be loaded this time, and two types of search methods for the search cells are applied to each of the prepared score parameters (score parameter sets) to perform a loading simulation, and the search method for the search cells and score parameters (score parameter sets) to be used in the calculation of the current loading are selected. To avoid redundancy, hereinafter, the score parameters (score parameter sets) are also referred to as score parameter sets.

[0079] In the present invention, the category of the pallet to be loaded is determined by classifying the pallet to be loaded into one of a plurality of categories (four in this embodiment) prepared in advance according to the number of packages to be loaded and the oligopoly state of the packages. Then, the position of each package to be loaded on the pallet to be loaded is determined using the score parameter set associated with the category. As described above, it is preferable to prepare a plurality of score parameter sets for each category and select one from among them.

[0080] The "oligopoly state of packs (objects)" refers to the oligopoly state of packs, that is, the ratio of the number of packs belonging to each created group to the total number of packs. In this embodiment, the oligopoly state of packs is expressed using the HH index (Herfindahl-Hirschman index). A higher HH index value indicates a higher degree of oligopoly, and a lower HH index value indicates a lower degree of oligopoly. Note that the oligopoly state may be expressed not only by the HH index but also by entropy, as will be described later.

[0081] FIG. 32 is a diagram for explaining how to calculate the HH index. In FIG. 32, bales A to J belong to different groups. The HH index is the sum of the squared values ​​of the oligopoly rates of each bale. For example, as shown in FIG. 32, if there are 10 bales in total, of which 8 are bales A and 2 are bales B, the squared oligopoly rate of bales A (8 / 10) is calculated. 2 and the square of the oligopoly rate of bin B (2 / 10) 2 The sum of these two values, 0.68, is the HH index. A lower HH index value indicates that there are more types of bales. A higher HH index value indicates that there are fewer types of bales. In the example of shipping bales information shown in Figure 30, a group containing 6 bales of diapers A and a group containing 12 bales of shampoo are created, so (6 / 18) 2 And (12 / 18) 2 The sum of these is the HH index.

[0082] The category classification information is information related to the category classification of the pallets to be loaded. The category classification information includes value information of the number of packages and the HH index (representing the oligopoly state of packages) that define each of the multiple categories.

[0083] The score parameter set information is information related to score parameter sets of evaluation functions linked to each of the multiple categories. The score parameter set information is information that collects score parameters of each of multiple (14 in this embodiment) evaluation functions. The evaluation information DB 46 stores, as score parameter set information, one or more score parameter sets linked to each category. In this embodiment, an example is given in which 10 parameter sets are linked to each category, but the number of parameter sets may be one or more.

[0084] In this embodiment, the number of categories of pallets to be loaded that are prepared in advance is four. The four categories are referred to as the first category, the second category, the third category, and the fourth category (see FIG. 14). For example, the first category is a category in which the number of packages is less than the reference value and the HH index is equal to or greater than the reference value, i.e., a category in which the number of packages is small and the HH index is high. In FIG. 14, the vertical line 21 indicates the package number reference value, and the horizontal line 22 indicates the HH index reference value.

[0085] The category classification information and score parameter set information stored in the evaluation information DB46 are calculated in advance for each shipping destination, for example, based on the performance data. Specific methods for calculating the category classification information and score parameter set information will be described later. The performance data is stacking performance data for multiple pallets at the time of past shipment. The stacking performance data includes information on the type and number of packages stacked on one pallet, and work order information (transportation order information) when the packages were stacked on the pallet.

[0086] In the actual stowage process, the HH index of the pallet to be stowed is calculated using the shipping package information, and the category of the pallet to be stowed is determined based on the HH index and the number of packages.Then, using the parameter set linked to the determined category, the placement position of the package to be stowed on the pallet is determined, and the stowage process of the packages is performed.

[0087] The score parameter set information may be calculated and prepared so that it can be commonly used for a plurality of shipping destinations, and may be stored in the evaluation information DB 46, or may be calculated and stored for each shipping destination. It is more preferable that the parameter set information is calculated and stored for each shipping destination, and more suitable parameter set information can be prepared according to the tendency of the packages to be shipped for each shipping destination, and a plurality of packages can be efficiently arranged on a pallet. In this way, the parameter set information may differ for each shipping destination. In addition, if the type of packages to be shipped tends to differ depending on the season, the parameter set information may be prepared for each season. For example, refreshing products such as mint-containing body shampoo, body sheets, antiperspirants, etc. tend to be shipped in large numbers in summer and small in winter. In addition, masks, warmers, moisturizing body lotions, etc. tend to be shipped in large numbers in winter and small in summer. Since the number of shipments of so-called seasonal products tends to vary greatly depending on the season throughout the year, stable and efficient stacking is possible throughout the year by preparing parameter set information according to the season.

[0088] ((Loading calculation parameter set information DB)) The stowage calculation parameter set information DB 47 stores, for each pallet ID, a parameter set determined by a parameter set determination process in Fig. 20, which will be described later. Each time a pallet PL to be stowed changes, the parameter set used when performing the stowage calculation can be changed.

[0089] ((Search method information DB)) The search method information DB 48 stores information on search method A as a first search method and search method B as a second search method used to extract (search) a search cell. Search method A and search method B are each a combination of a basic search and a wide search, but the combination methods are different. The inventors have found that, depending on the type and number of packages to be shipped, there are cases where it is more preferable to extract search cells using search method A, and cases where it is more preferable to extract search cells using search method B. Therefore, as described below, it is preferable to extract search cells using two search methods, perform simulations for each, and adopt the more appropriate search method.

[0090] Search cells that are candidates for the placement positions of packages to be stowed are extracted (searched) by either search method A or search method B. Below, the basic search and wide search, search method A and search method B will be explained in that order.

[0091] The basic search and wide search will be explained with reference to Fig. 4. In the basic search and wide search, search cells are extracted based on the cells partitioned into 10 mm squares by the grid lines described above. In Fig. 4, the search cell C extracted in the basic search is shown with an upward slanting line to the right, and the search cell C extracted in the wide search is shown with a downward slanting double slanting line to the right.

[0092] Search cell C extracted by the basic search is a cell adjacent to a line whose height changes when viewed from above in the Y-axis direction and adjacent to a line whose height changes when viewed from above in the X-axis direction. The lines that form the outline of the pallet in top view correspond to lines whose height changes.

[0093] The search cell C extracted by the wide search is a cell that is adjacent to a line whose height changes when viewed from above in either the Y-axis or X-axis direction, and adjacent to a line whose height changes when viewed in the other direction, or an extension of that line.

[0094] As shown in Fig. 4(a) and (b), the search cells extracted in the basic search are also always extracted in the wide search. The package to be loaded is placed in the search cell with the highest score. Since the search cells are extracted based on the line where the height direction changes, when the package to be loaded is placed on the loading surface of a pallet on which no other packages are placed, placing the package based on the search cell makes it easy to place the package to be loaded so that the length (L) direction and / or width (W) direction sides of the package to be loaded are aligned with the outer shape (contour) of the pallet in a plan view. When placing the package to be loaded on a pallet on which other packages are already placed, placing the package based on the search cell makes it easy to place the package to be loaded adjacent to the other packages with almost no gaps in a plan view, and also makes it easy to place the package to be loaded so that the length (L) direction and / or width (W) direction sides of the package to be loaded are aligned with the length (L) direction and / or width (W) direction sides of the other packages in a plan view. This makes it easier to stack packages efficiently without wasting space (improved space utilization rate).

[0095] Both search method A and search method B are combinations of a basic search and a wide search. They will be explained with reference to Figs. 5(a) and (b).

[0096] As shown in Fig. 5(a), in search method A, the control unit 3 first performs a basic search (ST70) to extract a cell, and then performs a wide search (ST71) to extract a cell. Then, all cells extracted by the basic search and the wide search are adopted as search cells (ST72). In this way, in search method A, both the basic search and the wide search are performed.

[0097] As shown in FIG. 5(b), in search method B, the control unit 3 first performs a basic search (ST80), extracts a cell, and determines whether or not there is an extracted cell (ST81). If it is determined that there is an extracted cell (YES in ST81), the cell extracted in the basic search is adopted as the search cell (ST82). On the other hand, if it is determined that there is no extracted cell (NO in ST81), a wide search is performed (ST83), cells are extracted, and all extracted cells are adopted as the search cells (ST84). Basically, extracting the search cell in the basic search allows for more efficient stacking. In search method B, if the search cell can be extracted in the basic search, the wide search is not performed and the process ends.

[0098] Whether to adopt search method A or search method B is determined using the evaluation results using dummy conveying order data in the parameter set determination process in Fig. 20, which will be described later. Basically, search method B can result in a more efficient layout than search method A, but there are cases in which search method B does not necessarily result in an efficient layout, so by providing two methods, a more efficient layout can be achieved.

[0099] In addition, it would be ideal to evaluate all cells, but this would take a lot of processing time. In contrast, in this embodiment, a basic search and a wide search are combined to narrow down the cells (search cells) that are candidates for placement positions, thereby shortening the processing time required for cell evaluation.

[0100] Among the extracted search cells, a search cell that satisfies all of the following constraints: (1) space constraints, (2) weight constraints, (3) installation constraints, and (4) loading robot constraints can be the placement position for the package to be loaded. By setting such constraints, stable loading is possible.

[0101] (1) It does not extend beyond the placement surface 20 of the pallet PL (which in this embodiment has a size of 1100 mm×1100 mm) and does not exceed the specified loading height.

[0102] (2) Do not stack bales above the load capacity set for each bale.

[0103] (3) At least a certain percentage (e.g., 85% or more in this embodiment) of the bottom surface of each package is in contact with the placement surface of the pallet PL or other packages. For example, as shown in Fig. 29(a), when package P2 to be loaded is placed on package P1 that is placed on pallet PL, if less than a certain percentage (less than 85% in this embodiment) of the bottom surface of package P2 is in contact with package P1, which is another package, it is determined that the constraint condition is not observed.

[0104] (4) The stowage robot 9 does not stow packages in positions where it is physically unable to place them. For example, as shown in FIG. 29(b), when a package (P4) to be stowed is to be placed on package P3 on a pallet PL on which packages P1, P2, and P3 are already placed, the gripper arm 91 can place the package (P4) to be stowed without interfering with the packages already placed, and package P3 can be the placement position for package P4 to be stowed. On the other hand, as shown in FIG. 29(c), when a package (P3) to be stowed is to be placed on a pallet PL on which packages P1 and P2 are already placed, if an attempt is made to place the package (P3) to be stowed between package P1 and package P2, the gripper arm 91 will be interfered with by packages P1 and P2, and therefore the package (P3) to be stowed cannot be placed, and the space between package P1 and package P2 will not be the placement position for package P4 to be stowed.

[0105] (Program storage section) The program storage unit 49 stores programs for executing various processes performed by the control unit 3 of the information processing system 100. For example, the program storage unit 49 stores programs for causing the information processing device 1 to execute a stowage calculation process and a robot control signal generation process.

[0106] [Control Unit] The control unit 3 is composed of a CPU and the like. In the information processing device 1, the CPU 11 loads a program stored in the program storage unit 49 into the RAM 13 and executes it, thereby performing processes related to stacking packages P on a pallet PL and the like. The program may be stored in, for example, a non-transitory recording medium readable by a computer, and the program may be installed in the information processing device 1 using the recording medium. Alternatively, the program may be installed in the information processing device 1 via a global network or the like. Examples of recording media for supplying the program include hard disks, magnetic disks such as floppy (registered trademark) disks, optical disks such as DVD-ROMs, CD-ROMs, and CD-Rs, USB memories, memory cards, non-volatile memories such as ROMs, and the like.

[0107] The control unit 3 generates score parameter set information and category classification information to be stored in advance in the evaluation information DB 46 .

[0108] Furthermore, when stacking the packages to be stacked, the control unit 3 uses the evaluation information DB 46 to classify the pallets to be stacked into one of the first to fourth categories based on the shipping package information, and determines the category of the pallets. Furthermore, the control unit 3 creates a plurality of simulation conditions each consisting of a combination of a plurality of score parameter sets and a search method (search methods A and B) for the search cells, each associated with the category. Next, the control unit 3 creates dummy transport order data (also called "dummy transport patterns") with a plurality of different transport orders using the packages to be stacked on the pallets to be stacked, and performs a stacking calculation for each simulation condition using the dummy transport order data. The control unit 3 uses the stacking calculation result to determine one parameter set to be used during the stacking process from among the plurality of parameter sets, and also determines the search method for the search cells to be either search method A or B. The control unit 3 stores the parameter set determined for each pallet and information related to the search method for the search cells in the parameter set information DB 47 for stacking calculation. In this manner, the control unit 3 performs a process of determining the simulation conditions.

[0109] The control unit 3 also extracts a search cell using the board state information retrieved from the stowage information DB 44. The control unit 3 calculates a score for each search cell using the determined parameter set, and determines the search cell with the highest score as the placement position of the package to be stowed. The control unit 3 generates a robot control signal for controlling the drive of the stowage robot 9 based on the determined placement position information on the pallet of the package to be stowed. The control unit 3 outputs the generated robot control signal to the stowage robot 9. In this manner, the control unit 3 issues stowage instructions to the stowage robot 9.

[0110] Furthermore, the control unit 3 classifies packages to be loaded onto a single loading target pallet into the same group if the differences in size and weight are equal to or less than a threshold value, and creates one or more groups.

[0111] The thresholds for the dimensions can be set appropriately. As an example, in this embodiment, the threshold for the difference in dimensions is set to 1 mm. In other words, packages whose dimensions match in increments of 1 mm can be in the same group. The threshold is not limited to 1 mm. In calculations using the dimension values ​​other than for determining whether or not they are in the same group (for example, in evaluation function score calculations, stowage calculations, etc.), the dimensions are converted to centimeters and rounded up to 1 cm units.

[0112] In addition, the above "weight difference is below the threshold value" means that the weight difference within the same group for multiple packages, calculated using the following formula (A), is below the threshold value.

[0113] Weight difference within group = (Maximum - Minimum) / Maximum …(A)

[0114] In formula (A), the "maximum value" is the weight value of the heaviest bale among the multiple bales belonging to the same group, and the "minimum value" is the weight value of the lightest bale among the multiple bales belonging to the same group. The threshold value related to the weight can be set appropriately. As an example, in this embodiment, the threshold value is set to 0.2.

[0115] The reasons for setting standards for bales that are considered to be in the same group are as follows. That is, even if the SKUs are different, there is no problem in considering them as the same bales when stacking them, as long as the scent is different, for example. Also, bales' dimensions are measured manually, so there is some error. Also, bales' dimensions change due to environmental changes such as temperature, so there is some error.

[0116] <Evaluation function> In this embodiment, an example is given in which the score of a search cell is calculated using 14 evaluation functions. FIG. 33 is a list of the contents of each evaluation function. The 14 evaluation functions are indicated by x1 to x14, and the weighting coefficients of each evaluation function are indicated by a1 to a14. a1 to a14 (weights) are score parameters. In FIG. 33, specific numerical values ​​are entered as the evaluation function score parameters (weighting coefficients), but this is only an example, and the numerical values ​​of the score parameter set included in the parameter set determined in the parameter set determination process described later are entered. In addition, the value of ax calculated using the score parameters is also only an example, and the value of ax may differ depending on the score parameter set used. The above-mentioned "score parameter set" is a collection of score parameters (weighting coefficients) of the 14 evaluation functions x1 to x14. In FIG. 33, "x" indicates the score of each evaluation function without reflecting the score parameter (weighting coefficient). "ax" indicates the score of each evaluation function with the weighting coefficient and the positive or negative of x reflected in "x". The sum of the scores (ax) reflecting the weighting coefficients of each evaluation function becomes the score of the search cell. Regarding the positive or negative value of x, "1" indicates that the score parameter is positive, and "-1" indicates that the score parameter is negative. The "value to multiply for normalization" is set for the purpose of normalizing the score numerical value of each evaluation function to be within the range of 0 to 1.

[0117] The evaluation functions x1 to x9 relate to the evaluation of the search cell that is the placement position of the bale to be loaded when the bale to be loaded is placed. The evaluation functions x10 to x14 relate to the evaluation of the board state after the bale to be loaded is loaded.

[0118] The evaluation functions x1, x2, x4 to x8 relate to the evaluation of space utilization rate and stacking stability during stacking. The evaluation function x3 relates to the evaluation of stacking stability. The evaluation functions x10, x12 to x14 relate to the evaluation of the placement possibility of the next package to be stacked after the object to be stacked is placed. The evaluation functions x9 and x11 relate to the arm of the stacking robot.

[0119] In this embodiment, the number of evaluation functions is 14, but is not limited to this. It is preferable to use at least x1, x2, x4, x5, x6, x7, x8, and x10 as the evaluation functions, and these evaluation functions have high priority.

[0120] Although it is not essential to use x3, x12, x13, and x14, from the viewpoint of improving the stowage success rate, it is more preferable to further use these evaluation functions, and the priority of these evaluation functions is medium.

[0121] On the other hand, x9 and x11 are evaluation functions related to the stowage robot, and may not necessarily be used depending on the configuration of the stowage robot, and these evaluation functions have low priority.

[0122] The evaluation function x3 (medium priority) is an evaluation function related to stability, like the evaluation functions x1, x2, and x5 to x8 with higher priority. However, the evaluation functions x1, x2, and x5 to x8 have a higher priority than the evaluation function x3 because they are essential evaluation functions for realizing stable and efficient stacking that also relates to space utilization rate. Also, the evaluation function x3 (medium priority) is an evaluation function related to stability, like the evaluation function x4 with higher priority. Here, the evaluation function x4 is related to the stability of loading in the height direction, whereas the evaluation function x3 is related to the placement balance from the center of gravity in the in-plane direction perpendicular to the height direction, and from the viewpoint of the stability of stacking of packages, the evaluation function x4 has a higher priority than the evaluation function x3.

[0123] The evaluation functions x12 to x14 (medium priority) are evaluation functions related to the possibility of placing an object to be stacked next after placing a package to be stacked, similar to the evaluation function x10 with a high priority. All evaluation functions evaluate the board state after placing the package to be stacked. Details will be described later, but the evaluation function x10 is an evaluation based on the search cell, whereas the evaluation functions x12 to x14 are evaluation based on the divided areas obtained by dividing the board state using the information on the height direction of the board state. In the evaluation function x10, if no search cell is extracted, the stacking possibility becomes zero, so it is preferable to stack the objects in such a way that the number of search cells is as large as possible. In other words, the evaluation function x10 is based on the search cell extracted in the same way as when determining the placement position of the package to be stacked in actual stacking, and therefore has a higher priority than the evaluation functions x12 to x14.

[0124] Below, each evaluation function will be explained with reference to Figures 6 to 12. In each figure, the bales to be loaded are shown in parentheses as (P) or (Pm) (m is an integer equal to or greater than 1). In addition, in Figures 6(e), (f), and (g), bales of the same type are shown with the same dot density, and bales of different types are shown with different dot densities to distinguish them.

[0125] A further explanation of the evaluation function x1 (svar): Evaluation function x1 is an evaluation function related to the degree of contact between the packages to be loaded and other packages, and the parameter has a positive or negative value of +. The higher the score of evaluation function x1, the higher the space utilization rate during loading and the higher the loading success rate tends to be. In detail, evaluation function x1 relates to the area ratio where the side of the package to be loaded comes into contact with other packages. Here, "high space utilization rate" means that the loading space on the pallet is highly utilized and packages can be loaded with fewer gaps.

[0126] svar is calculated by the following formula (1). In formula (1), "a_cnt_sum" is calculated by the following formula (2). Here, the four sides of the bales to be loaded are referred to as the first side, the second side, the third side, and the fourth side. In formula (2), f1(ratio) is a normalized value of the ratio of the contact area of ​​the first side with other bales (contact area of ​​the first side with other bales / area of ​​the first side). Similarly, f2(ratio), f3(ratio), and f4(ratio) are normalized values ​​of the ratio of the contact area of ​​the second side, the third side, and the fourth side with other bales, respectively. As shown in formula (2), "a_cnt_sum" is calculated as the sum of f1(ratio), f2(ratio), f3(ratio), and f4(ratio).

[0127]

number

[0128]

number

[0129] f1(ratio), f2(ratio), f3(ratio), and f4(ratio) are calculated by the following formula (3). Due to the constraints of the arm of the loading robot, the bale to be loaded cannot be loaded with a gap of 0 cm in a location surrounded by other bales, so a maximum of two sides of the bale to be loaded will come into contact with other bales (see Fig. 6(a)). For example, if two of the four sides of the bale to be loaded are in contact with other bales, a_cnt_sum approaches 2, and the values ​​to be adopted for "a" and "th_area" in formula (3) are determined so that the numerator of formula (1) is approximately 200 when a_cnt_sum = 2 and approximately 400 when a_cnt_sum = 4. In other words, if two of the four sides of the bale to be loaded are in contact with other bales over the entire surface, the value of svar calculated by formula (1) will be approximately 1. Here, "15" was used for "a" and "0.35" was used for "th_area." In addition, in formula (3), "ratio" indicates the ratio of the contact area with other packages to the side surface before normalization. Figure 38(a) shows the relationship between f(ratio) (vertical axis) normalized by formula (3) and ratio (horizontal axis) before normalization. Figure 38(b) shows the relationship between svar (vertical axis) and a_cnt_sum (horizontal axis).

[0130]

number

[0131] A supplementary explanation of the evaluation function x2 (sflush). The evaluation function x2 is related to the height evaluation of the package and is an evaluation function related to the degree of contact between the top surface of the package to be loaded and the top surface of other packages at the same height, and the parameter is positive or negative. The higher the score of the evaluation function x2, the higher the space utilization rate during loading and the higher the loading success rate tends to be. sflush is calculated using the perimeter ratio of the package to be loaded that is in contact with other packages at the same height as the package to be loaded when the package to be loaded is placed. The "perimeter ratio" indicates the ratio of the contact length between the top surface of the package to be loaded and the top surface of other packages at the same height as the package to be loaded, relative to the total perimeter of the top surface of the package to be loaded when the package to be loaded is placed. The "perimeter ratio" will be explained using Figure 6(b). In Figure 6(b), P1 and P2 are assigned to the packages already loaded, and (P3) is assigned to the package to be loaded. These three packages are assumed to be cubes. As shown in Figure 6 (b), bale (P3) touches bale P1 but does not touch bale P2. The part of bale (P3) that touches bale P1 at the same height as bale P1 is the part shown in bold in the figure. The "perimeter ratio" is the ratio of the bold part to the perimeter of bale P3 to be stowed (the perimeter of the top surface of bale P3). In the example shown in the figure, since the bale is a cube, the perimeter ratio is 0.25. The value of the evaluation function x2 (sflush) is calculated using formula (4) below.

[0132]

number

[0133] In formula (4), "total length of adjacent sides / total length of four sides of bales to be loaded" indicates the "perimeter ratio." "Bales to be loaded" are the bales to be loaded. Furthermore, L indicates the length dimension of the bales to be loaded, and W indicates the width dimension of the bales. In the example shown in Figure 6(b), the score of the evaluation function x2(sflush) is 0.0625.

[0134] A further note on the evaluation function x3(p_ij). Evaluation function x3 is an evaluation function of the concentration of the center of gravity, and the parameter is positive or negative. Evaluation function x3 is related to stability during loading. The farther the center of gravity of the package to be loaded is from the center of gravity of the pallet, the less stable the loading is, and the smaller the score of evaluation function x3, the less stable the loading is. The score of evaluation function x3(p_ij) is calculated using the following formula (5).

[0135]

number

[0136] "d" in formula (5) is the distance d between the center of gravity of the package (P) to be loaded and the center of gravity of the pallet on the XY plane shown in Figure 6 (c). The number multiplied by d in formula (5) is for normalization.

[0137] A further explanation of the evaluation function x4(h_ij). The evaluation function x4 is an evaluation function for the height of the packages to be loaded, and the parameter is positive or negative. The evaluation function x4 is related to stability during loading. When the packages to be loaded are placed, the higher the height from the pallet loading surface 20, the less stable the loading will be, and the smaller the value of the evaluation function x4, the less stable the loading will be. The score of the evaluation function x4(h_ij) is calculated using the following formula (6).

[0138]

number

[0139] "h" in formula (6) indicates the height of the top surface of the package (P5) to be loaded as shown in Fig. 6(d) when the package is loaded. The value obtained by formula (6) indicates the ratio of the height of the top surface of the package to be loaded when the package is loaded to the height limit value (180 cm in this embodiment). In formula (6), the number multiplied by h is for normalization.

[0140] A further explanation of the evaluation function x5 (id_score_same). The evaluation function x5 is an efficiency evaluation function, and the parameter is positive or negative (+). The evaluation function x5 is related to the space utilization rate during loading and the stability during loading. The evaluation function x5 evaluates whether the package to be loaded is placed on a package of the same type as itself, and being placed on a package of the same type is advantageous for stable loading, and also tends to increase the space utilization rate during loading. The score of the evaluation function x5 is 0 or 1.

[0141] In the example shown in FIG. 6(e), the package to be stowed (P3) is placed on packages P1 and P2 of the same type, so the score of the evaluation function x5 (id_score_same) is 1. In the example shown in FIG. 6(f), the package to be stowed (P3) is placed across package P1 of a different type from itself and package P2 of the same type from itself, so the score of the evaluation function x5 (id_score_same) is 0. In the example shown in FIG. 6(g), the package to be stowed (P3) is placed on packages P1 and P2 of different types, so the score of the evaluation function x5 (id_score_same) is 0. In the example shown in FIG. 6(h), the package to be stowed (P3) is placed on a pallet PL, so the score of the evaluation function x5 (id_score_same) is 0.

[0142] A further explanation of the evaluation function x6 (id_score_diff). The evaluation function x6 is an efficiency evaluation function, and the parameter is positive or negative. The evaluation function x6 is related to the space utilization rate during stowage and the stability during stowage. The evaluation function x6 evaluates whether the package to be stowed will be placed on a package of a different type from itself; being placed on a package of a different type is disadvantageous for stable stowage, and also tends to result in a lower space utilization rate during stowage. The score of the evaluation function x6 is 0 or 1.

[0143] In the example shown in FIG. 6(e), the package to be stowed (P3) is not placed on packages P1 and P2 of different types, so the score of the evaluation function x6(id_score_diff) is 0. In the example shown in FIG. 6(f), the package to be stowed (P3) is placed across package P1 of a different type from itself and package P2 of the same type from itself, so the value of the evaluation function x6(id_score_diff) is 1. In the example shown in FIG. 6(g), the package to be stowed (P3) is placed on packages P1 and P2 of different types, so the score of the evaluation function x6(id_score_diff) is 1. In the example shown in FIG. 6(h), the package to be stowed (P3) is placed on a pallet PL, so the score of the evaluation function x6(id_score_diff) is 0.

[0144] The evaluation functions x5 and x6 are provided because the weights of the evaluation functions x5 and x6 are not the same, and therefore two variables are required.

[0145] A further explanation of the evaluation function x7 (area_cover). The evaluation function x7 is an evaluation function for the degree of contact of the bottom surface of the package, and the parameter is positive or negative. The evaluation function x7 is related to the space utilization rate during loading and the stability during loading. The score of the evaluation function x7 (area_cover) is calculated by squaring the area ratio of the contact area between the base supporting the package (the loading surface of a pallet or other packages) and the bottom surface of the package to be loaded to the bottom surface of the package to be loaded. The more the bottom surface of the package to be loaded is in contact with the base supporting the package, the more stable the loading and the higher the space utilization rate during loading tends to be. In the example shown in Figure 7(a), the bottom surface of the package to be loaded (P2) is in contact with package P1, which is the base supporting the package, over its entire surface, so the score of the evaluation function x7 is 1.0. 2 In the example shown in FIG. 7(b), half of the bottom surface of the package (P2) to be loaded is in contact with package P1, which serves as the base for supporting the package, so the score of the evaluation function x7 is 0.5. 2 It can be calculated using the following formula.

[0146] A further explanation of the evaluation function x8(under_box_num). Evaluation function x8 is an evaluation function related to the degree of stone placement (stability), and the parameter is positive or negative. The score of evaluation function x8(under_box_num) is an HH index calculated based on the ratio of the area of ​​the contact surface between the bottom surface of the package to be loaded when the package to be loaded is placed on top of a package already placed on the pallet (hereafter referred to as the "pack below"), to the bottom surface of the package to be loaded. The more packages the newly loaded package is placed on top of, the higher the evaluation. In the example shown in Figure 7(c), for example, if the bottom surface area is 100 cm 2 The bale (P3) to be loaded is in contact with the bale P1 below, so the score of the evaluation function x8 is (100 / 100). 2 In the example shown in FIG. 7(d), for example, the base area is 100 cm 2 The bales (P3) to be loaded in this example have half of their bottom surfaces in contact with the bales P1 below, and the other half of their bottom surfaces in contact with the bales P2 below, so the evaluation function x8 is (50 / 100) 2 +(50 / 100) 2 As shown in the example in Figure 7(e), when the package (P) to be loaded is the first package on the pallet PL (when the bottom surface of the package is in contact with the loading surface of the pallet), the score of the evaluation function x8 is 1.

[0147] A supplementary note on the evaluation function x9(ij). The evaluation function x9 is an evaluation function of the proximity to the arm (gripper arm) of the loading robot, and the parameter is positive or negative. The score of the evaluation function x9(ij) is calculated using the following formula (7).

[0148]

number

[0149] In formula (7), i indicates the i-coordinate value, and j indicates the j-coordinate value. As shown in Fig. 7(f), when the stowage robot 9 is positioned below the pallet PL as viewed from above, the more bales are placed below the pallet PL (toward the stowage robot 9), the more difficult it becomes to place the next bales due to arm constraints. As shown in formula (7) above, the j-coordinate value is multiplied by 0.5, but the i-coordinate value is not multiplied by 0.5 because the range in which the remaining bales can be placed is narrower when the bales are placed downward than when they are placed to the right in the figure.

[0150] A further explanation of the evaluation function x10(next_pos) is provided. The evaluation function x10 is an evaluation function related to the possibility of placement, and the parameter is positive or negative (+). The score of the evaluation function x10 indicates the expected value of how many places on the board surface packages that are transported after the package to be stowed can be placed after the package to be stowed. The score of the evaluation function x10(next_pos) is calculated using the following formula (8).

[0151] next_pos=(y1+y2+…yn) / n…(8)

[0152] In formula (8), y indicates the number of possible arrangement patterns for each package transported after the package to be stowed, and two arrangement patterns, vertical and horizontal, are assumed for each search cell on a search cell basis. n indicates the number of remaining packages. It is assumed that the probability of which package will be the next package to be transported among the remaining packages is equal. The diagram on the left side of FIG. 8 shows an example of a search cell extracted for the first remaining package on the board after package P to be stowed is placed, and the diagram on the right side shows an example of a search cell extracted for the nth remaining package. The arrangement patterns for each remaining package are calculated by multiplying the number of search cells by 2 (the two patterns, vertical and horizontal), and subtracting the number of arrangement patterns that do not satisfy the constraints. In the example shown in the diagram on the left side of FIG. 8, y1 (the number of possible arrangement patterns for the first remaining package) can be calculated using the formula 15 (the number of cells) × 2 (the two patterns, vertical and horizontal) - a (the number of arrangement patterns that do not satisfy the constraints). In the example shown on the right side of Figure 8, yn (the number of possible arrangement patterns for the nth remaining package) can be calculated using the formula 11 (number of cells) x 2 (2 patterns vertically and horizontally) - b (the number of arrangement patterns that do not satisfy the constraints).

[0153] The above-mentioned "vertical placement" refers to a case where, when viewed from above in the drawing, the long side (corresponding to the L dimension) of the rectangular top surface of the bale is oriented vertically (Y-axis direction) and the short side (corresponding to the W dimension) is oriented horizontally (X-axis direction). On the other hand, "horizontal placement" refers to a case where, when viewed from above, the long side (corresponding to the L dimension) of the rectangular top surface of the bale is oriented horizontally and the short side (corresponding to the W dimension) is oriented vertically. In either case, the bale is oriented so that the upper left corner is located in the search cell when viewed from above.

[0154] A supplementary explanation of the evaluation function x11(surface). The evaluation function x11 is related to the arm of the stowage robot, and is an evaluation function that reflects the area where the stowage robot cannot load, and the parameter is positive or negative. In Fig. 9(a) and (b), BL in BL1 and BL2 indicates a stowage block with one or more stowage. As shown in Fig. 9(a), when the top surface of the stowage block BL1 located on the back side is sufficiently lower than the height of the stowage block BL2 in front as seen from the stowage robot 9 on the board surface after the stowage target stowage is loaded, an area where the next stowage Pnext cannot be loaded will occur depending on the performance of the stowage robot 9. The score of the evaluation function x11 reflects this area where loading is not possible (also called "back shadow"). The score of the evaluation function x11(surface) is the value obtained by dividing the area of ​​the cell that satisfies the back shadow condition shown below by the pallet area.

[0155] Shadow condition: Height of front bale block > Height of loaded rear bale block + Height of representative bale + G(a)

[0156] In the example shown in FIG. 9(a), the "front packing block height" in the above-mentioned back shadow condition indicates the height a of packing block BL2. The "height of the loaded rear packing block" indicates the height c of packing block BL1. The "representative pack height" indicates one of the minimum, maximum, and average heights of the remaining packs, which is called the "height b of representative pack P'." The above G(a) indicates the difference that the stowage robot can tolerate, which is set according to the height a (cm) of the front packing block BL2. FIG. 9(c) is a graph showing the relationship between the height of the front packing block and the difference that the stowage robot can tolerate, and this graph is prepared and stored in advance.

[0157] 9(a), the unloadable area (hereinafter sometimes referred to as the "back shadow area") 50 is an area (shown by diagonal lines in the figure) sandwiched between two extension lines obtained by linearly extending lines connecting the two rear vertices of the top surface of the packing block BL1 that the stowage robot 9 hits first to the outline of the pallet PL in the XY plane as viewed from the stowage robot 9. When the back shadow condition is satisfied, the area of ​​the back shadow area 50 is the area of ​​the cell that satisfies the back shadow condition (condition of the unloadable area).

[0158] Since it is okay for the stowage robot to stop, if it is desired to increase the number of pallets that can be loaded, it is preferable to adopt the maximum value of the remaining bale heights as the "representative bale height." If the maximum value were adopted, the shadow area 50 would be set as the area in which bales can be loaded, even though the next bale (a bale that is not the maximum value) may not be loaded. This increases the success rate of the stowage calculation. On the other hand, when a bale that the stowage robot cannot place (a bale that is not the maximum value) arrives, the stowage robot may stop on-site and the bale may not be loaded according to the stowage calculation. Note that successful stowage refers to the case where all the bale targets for loading have been loaded onto the pallet, and unsuccessful stowage refers to the case where none of the bale can be loaded.

[0159] If it is desired to avoid stopping the stowage robot, it is preferable to use the minimum value of the remaining bale heights as the "representative bale height." If the minimum value is used, the shadow area 50 will be an area where bale stacking is not possible, even though the next bale (a bale that is not the minimum value) may be able to be stacked. This reduces the success rate of the stowage calculation, but makes it less likely that a robot error will occur, and on-site the stowage robot can stack bale stacks according to the stowage calculation without stopping.

[0160] A further note on the evaluation function x12(area_max). The evaluation function x12 is related to the possibility of placement, and the parameter is positive or negative (+). The score of the evaluation function x12(area_max) is calculated by dividing the board surface 26 after the packages to be loaded have been placed into sections based on differences in package height, and then dividing the board surface 26 into sections with the largest surface area of ​​the same height, multiplying the area value of the largest area by 1 / (110×110). The higher the score of the evaluation function x12, the higher the possibility of placing the next package.

[0161] FIG. 10(b) shows the board surface 26 shown in FIG. 10(a) divided into a plurality of divided areas. FIG. 10(a) shows the board surface state of a pallet PL on which the packing blocks BL1 to BL5 are arranged. An area on which the packing blocks BL1 to BL5 are not arranged is called an empty area. In FIGS. 10(a) and 10(b), "H" indicates the height of the packing block BL. "H" in an empty area on which no packing block is arranged is 0. In the example shown in FIG. 10(a), the board surface 26 has a first empty area 511 and a second empty area 512 on which no packing block is arranged, in addition to the areas on which the packing blocks BL1 to BL5 are arranged. The board surface 26 is divided according to the difference in the height of the packs, and is further divided by a rectangle that has the largest surface of the same height, so that a total of nine divided areas, the first to ninth divided areas, are formed on the board surface 26 as shown in FIG. 10(b). The first to fifth divided areas A1 to A5 are areas on which the packing blocks BL1 to BL5 are arranged, respectively. The sixth divided area A6 and the seventh divided area A7 are divided areas obtained by dividing the first empty area 511. The eighth divided area A8 and the ninth divided area A9 are divided areas obtained by dividing the second empty area 512. Evaluation functions x13 and x14, which use the concept of divided areas described later, will also be explained using the board 26 shown in Figures 10(a) and (b) as an example.

[0162] In the example shown in FIG. 10(b), the score of the evaluation function x12 is calculated by multiplying the area value of the sixth divided area A6 by 1 / (110×110).

[0163] A supplementary note about the evaluation function x13 (area_mean). The evaluation function x13 is related to the possibility of placement, and the parameter is positive or negative. The score of the evaluation function x13 is calculated using the following formula (9).

[0164]

number

[0165] For each remaining bale, each divided area is virtually filled with as many bales as possible, and the number of bales that can be placed in each divided area is added up to obtain an estimate of the number of bales that can be placed on board 26. The total area s of the bottom surfaces of the bales that can be placed on board 26 calculated for each remaining bale is added up and divided by the number of remaining bales n, and this value is multiplied by 1 / (110×110) to calculate the score of evaluation function x13.

[0166] With reference to Figures 11(a) and (b), an example of calculating the score of the evaluation function x13 (area_mean) when the remaining bales are bales P1 and P2 of the same type will be described. As shown in Figure 11(a), each of the first to ninth divided areas A1 to A9 is virtually laid out so that as many bales P1 as possible can be placed therein. The number of bales to be placed in each divided area is added up to provide an estimate of the number of bales that can be placed on the board surface 26. In the example shown in Figure 11(a), a total of 18 bales P1 can be placed on the pallet PL. Assuming the base area of ​​bales P1 is 500 cm 2 In this case, the total area s1 of package P1 is 9000 cm 2 Similarly, as shown in FIG. 11(b), if we spread out the parcels P2 in each divided area, a total of 18 parcels P2 can be placed on the pallet PL, and the value of s2 will be 9000 cm. 2 The score of the evaluation function x13 (area_mean) is calculated by multiplying (9000+9000) / 2 by 1 / (110×110) using formula (9).

[0167] A further note on evaluation function x14 (area_dead). Evaluation function x14 is related to placement possibility and evaluates the proportion of dead areas where none of the remaining packages can be placed, and the parameter is positive or negative -. The score of evaluation function x14 is calculated using formula (10) below. In formula (10), the "area of ​​the dead area" is the total area of ​​the divided areas where none of the remaining packages can be placed. The score of evaluation function x14 is calculated as the ratio of the area of ​​the dead area to the area of ​​the pallet.

[0168]

number

[0169] With reference to FIG. 12, a calculation example of the evaluation function x14(area_dead) in the case where the remaining packages are different types of packages P1 and P2 will be described. As shown in FIG. 12, the divided areas in which neither package P1 nor package P2 can be placed (divided areas other than the divided areas in which at least one of packages P1 and P2 can be placed) are the first divided area A1, the seventh divided area A7, and the ninth divided area A9. At least one of packages P1 and P2 can be placed in the other divided areas (the second to sixth divided areas, and the eighth divided area). In this case, the score of the evaluation function x14(area_dead) is calculated by multiplying the total area (dead area area) of the first divided area A1, the seventh divided area A7, and the ninth divided area A9 by 1 / (110×110) using formula (10). The evaluation function x14 only looks at whether or not one or more packages can be placed in each divided area, and the number of remaining packages n is not used in calculating the score of the evaluation function x14.

[0170] <Information processing method> When stacking the packages to be stacked on a pallet, the control unit 3 of the information processing device 1 extracts search cells that are candidate locations for the packages to be stacked, evaluates the search cells using an evaluation function, and determines the search cell with the highest evaluation (highest score) as the location for the packages to be stacked. The control unit 3 determines the parameters (coefficients) of the evaluation function used to evaluate the search cells based on the relationship between the number of packages to be stacked on the pallet and the oligopoly state of the packages. The location for the packages to be stacked is determined in the order in which the packages are transported, and instructions for stacking the packages on the pallet are preferably given to a stacking robot 9 for stacking work. By controlling the stacking robot 9, the packages can be automatically stacked on the pallet PL.

[0171] In detail, the control unit 3 classifies the pallet to be stowed into one of four categories: a first category with a small number of bales and a high HH index, a second category with a large number of bales and a high HH index, a third category with a small number of bales and a low HH index, and a fourth category with a large number of bales and a low HH index, based on bales information including information on the number and dimensions of bales to be stowed on the pallet. Each category is associated with one or more score parameter sets in advance, and after multiple simulation conditions consisting of combinations of the score parameter sets and search methods for search cells are created, the simulation conditions (including score parameter set information and information on the search method for search cells) to be used in the actual stowage process are determined (second process). Next, the search cell is evaluated based on the evaluation function, the score parameter set, and the search method for the search cell, and the placement position of the package to be stowed is determined (third process).

[0172] The number of bales and the standard value of the HH index used to classify the pallets to be loaded into one of the four categories above, as well as the parameter sets associated with each category, are calculated in advance using actual loading data from the past and stored (step 1).

[0173] In this way, the placement positions of the bales to be loaded are determined using a parameter set that is set based on the number of bales and the oligopoly state, allowing the bales to be loaded more reliably and stably, making possible efficient loading. Furthermore, by preparing multiple parameter sets based on the number of bales and the oligopoly state, and evaluating (simulating) the combination with the search method for the search cell, an appropriate parameter set and search method for the search cell can be selected, enabling more efficient loading.

[0174] The following describes in advance the process of generating category classification information and score parameter set information stored in the evaluation information DB 46, and the overall flow of information processing related to stacking packages on a pallet. In Fig. 13 and Fig. 20 to Fig. 25, which explain each processing flow, "ST" means "step." The information processing shown in Fig. 13 and Fig. 20 to Fig. 25 is processing that is mainly executed by the control unit 3.

[0175] [First step (generation of category classification information and score parameter set information)] The process of generating category classification information and score parameter set information will be described according to the flow in FIG. 13 and with reference to FIGS. 14 to 16 as appropriate.

[0176] The control unit 3 acquires stowage performance data for multiple pallets at the time of past shipment (ST1). In this embodiment, an example is given in which stowage performance data for 200 pallets is used. Note that the number of stowage performance data to be used is not limited as long as it is one or more, and from the viewpoint of realizing more reliable and efficient stowage, it is advisable to use stowage performance data for multiple pallets, preferably 100 pallets or more.

[0177] Next, the control unit 3 calculates the HH index and the number of packages loaded for each pallet based on the stacking record data (ST2). When calculating the HH index, the control unit 3 uses the stacking record data for each pallet to classify the packages to be stacked on the pallet into one or more groups, with packages whose dimensional and weight differences are below a threshold being grouped together, and calculates the oligopoly state of the packages, which is the proportion of the number of packages belonging to each group relative to the total number of packages, as the HH index. ST2 is repeated until calculations for all pallets (200 pallets in this embodiment) are completed.

[0178] Next, the control unit 3 uses the calculated number of loaded packages for 200 pallets and the HH index to calculate a package number standard value and an HH index standard value that are standards for classification into four categories (ST3).

[0179] In detail, as shown in Fig. 14, on a graph showing the number of packages on the horizontal axis and the HH index on the vertical axis, points for 200 pallets are plotted based on the calculated number of packages loaded and the HH index. As shown in Fig. 14, a vertical line 21 is set so that the number of plotted points on both sides of the vertical line 21 is approximately equal. The vertical line 21 indicates the package number standard value, and in the example shown in Fig. 14, the package number standard value is 35. As shown in Fig. 14, the HH index value is expressed as a round number in increments of 0.1, and a horizontal line 22 is set so that the number of plotted points on both sides of the horizontal line 22 is approximately equal. The horizontal line 22 indicates the HH index standard value, and in the example shown in Fig. 14, the HH index standard value is 0.2. In this way, the HH index standard value and the package number standard value can be set, and in the category classification shown in Fig. 14, approximately 50 pallets' worth of points are positioned in each category.

[0180] In this embodiment, the reference values ​​of the number of packages and the HH index, which are used to separate the categories, are calculated so that the number of points classified into the category below the reference value is equal to the number of points classified into the category above the reference value, but this is not limiting. For example, statistics such as the average value or the median value may be used. Values ​​may also be rounded off to a convenient number.

[0181] Next, the control unit 3 creates four categories, first to fourth, based on the calculated package number standard value and HH index standard value (ST4), and classifies each of the 200 pallets into one of the four categories. The control unit 3 also generates category classification information and stores it in the evaluation information DB 46 in association with the shipping destination. The category classification information is used when stacking packages at the time of shipping to the corresponding shipping destination. The category classification information includes information related to the package number and HH index that define each category. For example, referring to FIG. 14, the first category is a category in which pallets are placed on which packages with a package number less than the standard value (less than 35 packages) and an HH index equal to or greater than the standard value (0.2 or greater) are placed.

[0182] Next, the control unit 3 creates a plurality of score parameter set candidates, for example, 3000 to 20000, each of which is composed of scores of 14 evaluation functions, with different score values ​​(ST5). Here, it is assumed that 3000 candidates are created. From these 3000 candidates, for example, 10 score parameter sets (hereinafter, the description will be given assuming that 10 are selected) are selected for each category and linked to the category. In this embodiment, an example is given in which a score parameter set candidate common to categories is created when selecting a score parameter set to be linked to each category, but the score parameter set candidate may be changed for each category.

[0183] Next, for each category, for each pallet classified into the category (in this embodiment, 50 pallets are classified into one category), a stowage calculation is performed for the packages transported in the work order included in the stowage performance data of each pallet, using the score parameters (weighting coefficients) of each evaluation function included in the candidate score parameter set for each candidate score parameter set (ST6). In the stowage calculation, in determining the placement position of each package transported in the work order, after extracting a search cell, the score of each extracted search cell is calculated using the score parameters of each evaluation function included in the candidate score parameter set, and the placement position of the package to be stowed is determined as the search cell with the highest score. Then, the packages are stowed at the determined placement position. In the extraction of the search cell, first, a cell that is a search cell candidate is extracted by a basic search, and if a cell that is a search cell candidate is extracted by the basic search, the cell is set as the search cell. On the other hand, if a cell that is a search cell candidate is not extracted by the basic search, a wide search is further performed to extract a cell that is a search cell candidate, and the cell extracted by the wide search is set as the search cell. The loading calculations are carried out until all 50 pallets in one category have been calculated.

[0184] Next, in ST7, the control unit 3 uses the stowage calculation results to calculate the stowage success rate for each candidate score parameter set. Approximately 50 pallets are classified into each category, so the stowage calculation results for one category will yield 50 stowage calculation results for each candidate score parameter set. The stowage success rate is the ratio of successful stowage results to these 50 stowage calculation results. For example, if there are 40 successful stowage results for one candidate score parameter set, the stowage success rate is 0.8 (80%). The higher the stowage success rate, the more efficient the stowage.

[0185] Furthermore, in ST7, for each candidate score parameter set, the spatial occupancy rate when the packages transported in the work order included in the loading performance data of each pallet are loaded onto the pallet PL is calculated, and the median value is calculated. The spatial occupancy rate is calculated by dividing the total volume of the loaded packages by the product of the loading surface area (here, 1100 mm x 1100 mm) and the maximum height of the entire pallet when the packages are loaded. The higher the median spatial occupancy rate, the higher the loading stability.

[0186] Furthermore, in ST7, for each candidate score parameter set, the surface area of ​​the group of packages when the packages transported in the work order included in the stowage performance data of each pallet are stowed on the pallet PL is calculated. The surface area of ​​the group of packages is then used to calculate the surface area ratio, and the median value is calculated. The surface area ratio is calculated by dividing the surface area of ​​the group of packages by the sum of the surface areas of each package. The lower the median surface area ratio, the higher the space utilization efficiency and the higher the stowage stability.

[0187] The calculation process of the stacking success rate, the median of the space occupancy rate, and the median of the surface area ratio for each candidate score parameter set is repeated until the calculation process for all 3000 candidates of the score parameter set is completed.

[0188] Next, score parameter sets with a stowage success rate equal to or higher than a threshold are extracted (ST8). The threshold for the stowage success rate can be determined in consideration of the distribution of values ​​of the stowage success rate for each category. For example, if the threshold for the stowage success rate is set to a value that is too high, there may be no candidates for score parameter sets to be extracted based on the stowage success rate as a criterion. Therefore, it is preferable to set the threshold for the stowage success rate so that the number of candidates for score parameter sets to be extracted based on the stowage success rate is appropriately secured, while not deviating too far from the calculated maximum stowage success rate.

[0189] Next, using the calculated median space occupancy rate and median surface area rate, 10 optimal score parameter sets are extracted from the candidate score parameter sets extracted in ST8, and these are selected as the score parameter sets to be linked to the corresponding categories (ST9).

[0190] The processes of ST6 to ST9 are repeated until the process of all categories is completed. As a result, ten score parameter sets are associated with each of the four categories, and score parameter set information is generated. The control unit 3 stores the score parameter set information and category classification information in the evaluation information DB 46 in association with the shipping destination.

[0191] A method for selecting a set of 10 score parameters to be associated with categories using the median value of the space occupancy rate and the median value of the surface area rate of ST9 will be described with reference to FIGS. 15 and 16. FIG.

[0192] A specific procedure for selecting a score parameter set having a high median space occupancy rate and / or a low median surface area ratio will be described with reference to Figs. 15 and 16. In this embodiment, an example of creating 3000 score parameter set candidates is given, but in the description of selecting 10 score parameter sets from a plurality of candidates using Figs. 15 and 16, an example of selecting 10 score parameter sets from 300 score parameter set candidates is given for convenience. The graphs shown in Figs. 15(a) and 16 are plotted with the median space occupancy rate on the x-axis and the median surface area ratio on the y-axis, and the median space occupancy rate and the median surface area ratio of each calculated score parameter set candidate are plotted. A higher median space occupancy rate indicates higher stacking stability, and a lower median surface area ratio indicates higher stacking stability.

[0193] First, the 300 score parameter set candidates are sorted in descending order by the median of the space occupancy rate and ranked from 1 to 300. The 300 score parameter set candidates are also sorted in ascending order by the median of the surface area ratio and ranked from 1 to 300. Next, the sum of the rank of the median of the space occupancy rate and the rank of the median of the surface area ratio (called the "rank total") is calculated for each score parameter set. Then, 30 score parameter set candidates are selected in ascending order of the rank total value. FIG. 15(b) plots the ranked score parameter set candidates with the rank of the median of the space occupancy rate on the x-axis and the rank of the median of the surface area ratio on the y-axis. In FIG. 15(b), the 30 score parameter set candidates selected in descending order of the rank total value are indicated by open circles.

[0194] Next, 10 score parameter sets are randomly extracted from the selected 30 score parameter set candidates. In the graph shown in Fig. 16, the 30 score parameter sets selected in ascending order of rank total value are shown as open circles. Furthermore, the 10 score parameter sets randomly extracted from these 30 points are shown as open circles with a cross (x) superimposed on them.

[0195] In this manner, 10 score parameter sets to be associated with each category stored in the evaluation information DB 46 are determined. In this embodiment, 30 score parameter set candidates are selected in ascending order of rank total value. That is, 30 score parameter sets with high median space occupancy and low median surface area ratio are selected. Furthermore, from the selected 30, score parameter sets to be stored in the evaluation information DB 46 are randomly extracted. Here, as shown in FIG. 16, the selected 30 points are plotted in a high numerical range such as a median space occupancy of about 0.67 to 0.74 and in a low numerical range such as a median surface area ratio of about 0.37 to 0.49. The above "randomly extracted" means randomly extracted within this numerical range. If the score parameter sets are extracted with a bias, a combination of packs that is difficult to use may occur. In contrast to this, in the present embodiment, multiple score parameter sets to be linked to categories are selected from the different perspectives of space occupancy and surface area ratio, and score parameter sets to be stored in evaluation information DB46 are randomly extracted within numerical ranges suitable for stable stacking in terms of both space occupancy and surface area ratio, making it less likely that difficult packing combinations will occur and enabling stable stacking.

[0196] Fig. 34 shows an excerpt of some of the 3,000 score parameter set candidates, and is a list of score parameter set candidates with a stowage success rate of 0.96 or more extracted in ST8 based on the calculation result of the stowage success rate calculated in ST7 for one category. In Fig. 34, the dotted cells indicate score parameter sets selected in ST9 as score parameter sets to be linked to the corresponding category from among the score parameter set candidates with a stowage success rate of 0.96 or more.

[0197] In this manner, the category classification information and score parameter set information to be stored in advance in the evaluation information DB 46 can be calculated.

[0198] [Overall flow of information processing related to loading packages onto pallets (steps 2 and 3)] The process will be described according to the flow in Fig. 17. When the process of stacking onto a pallet PL starts, the control unit 3 executes a parameter set determination process (ST20), and then executes a stacking process (ST40). Although details will be described later, in the parameter set determination process, the category of the pallet to be stacked is determined, and a parameter set including a score parameter set and a search method for a search cell, which is determined from a score parameter set associated with the category, is determined as the parameter set to be used in stacking calculations when stacking packages onto the pallet to be stacked. The stacking process is executed using the determined parameter set.

[0199] When multiple types of packages that are transported sequentially are stacked on a pallet, a stacking process (ST40) is executed, and the packages in the shipping package information in Fig. 30 are stacked on the pallet to form a package group 10, as shown in Fig. 18 and Fig. 19. In Fig. 18 and Fig. 19, the diaper packages are labeled with the symbol A, the shampoo packages are labeled with the symbol B, and all the packages stacked on the pallet PL are labeled with a serial number.

[0200] [Second step (parameter set determination process)] The parameter set determination process will be described with reference to the flow in FIG. 20. When the parameter set determination process starts, the control unit 3 acquires shipping package information (ST21). Next, the control unit 3 uses the shipping package information to calculate the number of packages on the pallet to be loaded and the HH index (ST22). When calculating the HH index, the control unit 3 uses the shipping package information to classify the packages to be loaded on the pallet into one or more groups, with packages whose dimensional and weight differences are below a threshold being grouped together, and calculates the oligopoly state of the packages, which is the percentage of the number of packages belonging to each group relative to the total number of packages, as the HH index. In the case of the shipping package information shown in FIG. 30, the packages are classified into two groups, the number of packages is 18, and the HH index is approximately 0.51.

[0201] Next, the control unit 3 refers to the categorization information stored in the evaluation information DB 46, and determines the category of the pallet to be stowed based on the calculated number of packages and HH index (ST23). In this embodiment, the HH index standard value is 0.2, the number of packages standard value is 35, and classification into four categories is performed, so the pallet to be stowed is classified into the first category.

[0202] Next, the control unit 3 refers to the score parameter set information stored in the evaluation information DB 46 and extracts a plurality of score parameter set information (10 in this embodiment) associated with the category determined in ST23. In this embodiment, 10 types of parameter set information are combined with two types of search methods for search cells (method A and method B) to create 20 patterns of simulation conditions (ST24) as shown in FIG. 35. In FIG. 35, the simulation conditions are displayed so that the higher-ranked simulation conditions are at the top based on the calculation results of the stacking success rate, the median of the space occupancy rate, and the median of the surface area ratio, which will be described later, so the serial numbers of the 20 patterns of simulation conditions are not in numerical order. In FIG. 35, the columns of the parameter sets determined in ST31, which will be described later, are filled with dots.

[0203] Next, the control unit 3 uses the shipping package information to create a plurality of dummy transport sequence data (dummy transport patterns) (for example, 50 pieces) with different transport sequences (ST25).

[0204] Next, the control unit 3 performs a stowage calculation for each of the dummy conveying order data based on the score parameter set information and the search method of the search cell included in each simulation condition (ST26). ST26 is repeated until the stowage calculation process for all of the dummy conveying order data is completed.

[0205] Next, the control unit 3 calculates the stowage success rate, the median of the space occupancy rate, and the median of the surface area ratio as evaluation indexes for the multiple dummy transport sequence data based on the stowage calculation processing results calculated in ST26 for each simulation condition (ST27). ST27 is repeated until the calculation processing for all simulation conditions is completed.

[0206] Next, the control unit 3 sorts each parameter set in descending order based on the stowage success rate calculated in ST27 (ST28). Next, the control unit 3 sorts each parameter set in descending order based on the median of the space occupancy rate calculated in ST27 (ST29). Next, the control unit 3 sorts each parameter set in descending order based on the median of the space occupancy rate calculated in ST27 (ST30).

[0207] Next, a parameter set and a search method for search cells are determined based on the processing results of ST28 to ST30 (ST31). In detail, a parameter set constituting a simulation condition with the highest stowage success rate and a search method for search cells are adopted. If there are multiple simulation conditions with the highest stowage success rate, the simulation condition with the highest median space occupancy rate is adopted. Furthermore, if there are multiple simulation conditions with the highest median space occupancy rate, the simulation condition with the lowest median surface area is adopted. As shown in FIG. 35, in this embodiment, simulation condition No. 7 with the highest stowage success rate is adopted and is determined as the parameter set and search method for search cells to be used during stowage.

[0208] The control unit 3 associates the determined parameter set with the pallet to be stowed, and stores it in the stowage calculation parameter set information DB 47. In this embodiment, the parameter set is determined about two hours before the start of the stowage process.

[0209] [Third process (loading)] The stowing process will be described according to the flow of FIG.

[0210] When the stowage process (ST40) starts, the control unit 3 acquires the parameter set information determined in ST20 to be used when stowage of packages on the pallet to be stowed, which is stored in the stowage calculation parameter set information DB 47 (ST41). The control unit 3 also acquires shipping package information.

[0211] Next, the control unit 3 performs a process for determining the placement positions on the pallet or the temporary placement area for the packages to be loaded that are being transported in sequence, and loads the packages to be loaded in the order of transport. This will be described below.

[0212] The control unit 3 reads the surface state of the pallet to be loaded stored in the stowage information DB 44 (ST42). Next, the control unit 3 determines whether or not the package to be loaded satisfies the placement condition for placement in the temporary placement area (temporary placement area placement condition) (ST43).

[0213] A package that satisfies the "temporary placement area arrangement condition" refers to a package that satisfies the condition that it is relatively lighter and has a relatively smaller bottom surface compared to packages belonging to each group. The temporary placement area arrangement condition will be described in detail later.

[0214] When the control unit 3 determines that the package to be stowed satisfies the temporary placement area arrangement condition (YES in ST43), it proceeds to ST63. On the other hand, when the control unit 3 determines that the package to be stowed does not satisfy the temporary placement area arrangement condition (NO in ST43), it determines whether the search method for the search cell is method B or not (ST44).

[0215] If the control unit 3 determines that the search method of the search cell is method B (YES in ST44), it sets the search range to the basic search range (ST46) and proceeds to ST47. If the control unit 3 determines that the search method of the search cell is not method B (NO in ST44), it sets the search range to the basic search range and wide search range (basic search range + wide search range) (ST45) and proceeds to ST47.

[0216] In ST47, the control unit 3 extracts a search cell in the search range set in ST46 or ST45, that is, by method B or method A. Next, the control unit 3 executes a search cell evaluation process (referred to as process A) for the extracted search cell (ST48). Process A will be described with reference to FIG. 27 according to the flow of FIG. 22. FIGS. 27(a) and (e) are respectively horizontal and vertical placement patterns of the bale when the lower left corner of the bale is located in the search cell C. FIGS. 27(b) and (f) are respectively vertical and horizontal placement patterns of the bale when the lower right corner of the bale is located in the search cell C. FIGS. 27(c) and (g) are respectively horizontal and vertical placement patterns of the bale when the upper right corner of the bale is located in the search cell C. FIGS. 27(d) and (h) are respectively vertical and horizontal placement patterns of the bale when the upper left corner of the bale is located in the search cell C.

[0217] As shown in Fig. 22, in process A (search cell evaluation process) of ST48, the control unit 3 performs score calculation for search cells that satisfy the constraint conditions among the extracted search cells. In detail, for each search cell, as shown in Figs. 27(a) to (h), the control unit 3 determines whether the constraint conditions are satisfied in each of a total of eight layout patterns in which the four corners (corners of the package) of the rectangular upper surface of the package in top view, i.e., the upper left corner, the upper right corner, the lower left corner, and the lower right corner, are positioned in search cell C, and the package is placed vertically and horizontally (ST480). Next, score calculation is performed only for the layout patterns that are determined to satisfy the constraint conditions (ST481).

[0218] In this way, from the viewpoint of improving the quality of stacking, it is preferable to perform score calculation only when the constraint conditions are satisfied among the eight layout patterns, in other words, when the physical constraints are satisfied. However, this processing method is not limited to this. Below, other processing method examples will be described with reference to Figures 26 and 27.

[0219] As another processing method (search cell evaluation processing method), a step of calculating scores for the vertical and horizontal patterns may be performed after the step of determining which corner of the bale to be placed in the search cell. The step of determining the corner of the bale described above is a step of determining which corner of the bale to be loaded is to be placed in the search cell.

[0220] 26(a) and (b) are diagrams for explaining which corners of the packages to be stowed in ST480 are to be positioned in each of the search cells extracted in ST47. In the example shown in FIG. 26(a), search cells extracted in ST47, for example, in the basic search range and the wide search range, for a pallet PL on which two packages P1 and P2 are stacked, are shown. In FIG. 26(a) and (b), the search cells extracted in the basic search range are indicated by diagonal lines going up to the right, and the search cells extracted in the wide search range are indicated by double diagonal lines going down to the right. FIG. 26(b) is a diagram showing which corners of the packages are to be positioned in the search cells determined in the step of determining the package corners described above. In FIG. 26(b), the search cells marked with "D" indicate that the upper left corner of the package to be stowed is located. The search cells marked with "E" indicate that the upper right corner of the package is located. The search cells marked with "F" indicate that the lower left corner of the package is located. The search cell marked with a "G" indicates where the lower right corner of the bale is located.

[0221] In the step of determining the corners of the pallet PL described above, when viewed from above, looking from left to right horizontally and from top to bottom vertically, the search cell on the left side or above the line where there is a change in height is defined as the "search cell before the change occurs," and the search cell on the right side or below the line where there is a change in height is defined as the "search cell after the change occurs."

[0222] In the step of determining the corners of the bales described above, the control unit 3 determines the corner of the bales located in the search cell after a change has occurred in the horizontal direction and after a change has occurred in the vertical direction (search cell marked with "D" in Figure 26(b)) to be the upper left corner. The control unit 3 determines the corner of the bales located in the search cell before a change has occurred in the horizontal direction and after a change has occurred in the vertical direction (search cell marked with "E" in Figure 26(b)) to be the upper right corner. The control unit 3 determines the corner of the bales located in the search cell after a change has occurred in the horizontal direction and before a change has occurred in the vertical direction (search cell marked with "F" in Figure 26(b)) to be the lower left corner. The control unit 3 determines the corner of the bales located in the search cell before a change has occurred in the horizontal direction and before a change has occurred in the vertical direction (search cell marked with "G" in Figure 26(b)) to be the lower right corner.

[0223] Next, the control unit 3 creates a vertical arrangement pattern and a horizontal arrangement pattern for each search cell with the corners of the package determined in the step of determining the corners of the package described above positioned in the search cell, and calculates scores for each arrangement pattern using the score parameters of each evaluation function in the score parameter set information included in the parameter set information determined in ST20 (the score calculation step described above).

[0224] Such other processing methods may be adopted, which may save computational resources and reduce computation time.

[0225] When the processes in ST480 and ST481 are completed for all the search cells, the process proceeds to ST49 as shown in FIG.

[0226] In ST49, the control unit 3 determines whether the number of arrangement patterns in which the packages can be arranged (number of possible arrangement patterns) is 0. If the number of possible arrangement patterns is 0, this indicates that none of the arrangement patterns created in ST48 comply with the constraint conditions.

[0227] When the control unit 3 determines that the number of possible placement patterns is not 0 (NO to ST49), it determines the placement pattern of the search cell having the placement pattern with the highest score as the placement position of package P (determination of placement position information) (ST50). In other words, the placement position information on the pallet of the packages to be loaded is determined. This "placement position information" includes information on which search cell package P will be placed in, information on the orientation of package P (information on whether the top surface of the package is vertical or horizontal), and information on the corner of the top surface of package P located in the search cell (information on which of the four corners of the top surface is located in the search cell).

[0228] Next, the control unit 3 generates a robot control signal for controlling the driving of the stowage robot 9 based on the determined placement position information (ST51), outputs it to the stowage robot 9 (ST52), and proceeds to ST53. The stowage robot 9 operates based on the robot control signal, and places the packages to be stowed in the determined placement positions.

[0229] If the control unit 3 determines that the number of possible arrangement patterns is 0 (Yes in ST49), it determines whether the search method of the search cell is method B or not (ST56). If the control unit 3 determines that the search method of the search cell is not method B (NO in ST56), it proceeds to ST58. If the control unit 3 determines that the search method of the search cell is method B (YES in ST56), it determines whether a wide search has been performed (ST57). If the control unit 3 determines that a wide search has been performed (YES in ST57), it proceeds to ST58. If the control unit 3 determines that a wide search has not been performed (NO in ST57), it sets the search range to the wide search range (ST62) and proceeds to ST47.

[0230] In ST58, the control unit 3 determines whether or not the package P to be stowed can be placed in the temporary placement area. If the control unit 3 determines that the package P cannot be placed (NO in ST58), the control unit 3 proceeds to ST66. If the control unit 3 determines that the package P can be placed (YES in ST58), the control unit 3 sets the placement position of the package P to be stowed in the temporary placement area (ST59).

[0231] When the control unit 3 determines that the package P to be loaded satisfies the placement conditions for the temporary placement area (YES in ST43), it determines whether the package P to be loaded can be placed in the temporary placement area (ST63). When the control unit 3 determines that the package P to be loaded can be placed in the temporary placement area (YES in ST63), it sets the placement position of the package P to be loaded in the temporary placement area (ST59).

[0232] If the control unit 3 determines that the package to be loaded cannot be placed in the temporary storage area (NO in ST63), it determines whether or not the package to be loaded can be placed on the pallet (ST64). If the control unit 3 determines that the package can be placed (YES in ST64), it proceeds to ST44. If the control unit 3 determines that the package cannot be placed (NO in ST64), it reads the state of the temporary storage area (ST65) and proceeds to ST66.

[0233] In ST66, the control unit 3 determines whether the package placed in the temporary storage area can be transferred and placed on a pallet. If the control unit 3 determines that placement is not possible (NO in ST66), it ends the stowage process. If the control unit 3 determines that placement is possible (YES in ST66), it executes a transfer process (referred to as process C) to transfer the package placed in the temporary storage area from the temporary storage area to a pallet (ST67).

[0234] Process C (transfer process of a package from the temporary placement area to a pallet) will be described with reference to the flow in Figure 24. Using dimensional information about the packages in the temporary placement area, the control unit 3 sorts the packages in the temporary placement area in ascending order based on their width (W) dimension (ST670). Next, the control unit 3 sorts the packages in the temporary placement area in ascending order based on their length (L) dimension (ST671). Next, the control unit 3 sorts the packages in the temporary placement area in ascending order based on their height (H) dimension (ST672). Figure 36 shows an example of the sorting result.

[0235] Based on the sorting results of ST670 to ST672, the control unit 3 selects one of the top-most packages, determines that package as the package to be transferred to the pallet (ST673), and executes a placement process (referred to as process B) to place that package on the pallet (ST61). As a result, the package that was in the temporary storage area is placed (transferred) on the pallet. Process B (ST61) will be described later. Next, the process proceeds to ST68. In this way, in process C, the smaller package of the packages placed in the temporary storage area is determined as the package to be moved to the pallet.

[0236] In this way, when a package to be stowed cannot be placed on either the pallet or the temporary storage area, the package is transferred from the temporary storage area to the pallet. At this time, if a package with a large base area is selected as the package to be transferred to the pallet, it becomes more likely that the package to be stowed that has been transported can be placed on top of the transferred package. This makes it more difficult for a timing (trigger) to return the package from the temporary storage area to the pallet to occur, and reduces the opportunities to reduce the number of packages in the temporary storage area. In contrast, in this embodiment, by selecting a package with a small base area as the package to be transferred to the pallet, it is possible to increase the opportunities to reduce the number of packages in the temporary storage area, making it possible to free up as much space as possible in the temporary storage area, and making it easier to stow all of the packages to be stowed on the pallet.

[0237] Returning to Fig. 21, in ST68, the control unit 3 updates and stores the surface status of the temporary placement area and the pallet. Next, the control unit 3 determines whether or not the package to be stowed can be placed on the pallet (ST69). If the control unit 3 determines that the package can be placed (YES in ST69), it returns to ST42, but if it determines that the package cannot be placed (NO in ST69), it sets the placement position of the package P to be stowed in the temporary placement area (ST59).

[0238] Following ST59, the control unit 3 reads the state of the board in the temporary placement area (ST60). Next, the control unit 3 executes a process (process B) for placing the packages to be stacked in the temporary placement area (ST61). As a result, the packages to be stacked are placed in the temporary placement area. Process B (ST61) will be described later. Next, the process proceeds to ST53.

[0239] In ST53, the control unit 3 updates and stores the state of the pallet and the temporary storage area. When processing of all packages to be loaded onto the pallet is completed, the control unit 3 determines whether or not there are any packages in the temporary storage area (ST54). If the control unit 3 determines that there are no packages (No in ST54), it ends the processing, and if it determines that there are packages (Yes in ST54), it executes a process of transferring the packages in the temporary storage area to the pallet (referred to as process D) (ST55).

[0240] Process D will be described according to the flow of Fig. 25. The control unit 3 reads information on the temporary placement area and the surface status of the pallet, and all packages placed in the temporary placement area (ST550). Next, the control unit 3 sorts the packages in the temporary placement area in descending order based on the width (W) dimension (ST551). Next, the control unit 3 sorts the packages in the temporary placement area in descending order based on the length (L) dimension (ST552). Next, the control unit 3 sorts the packages in the temporary placement area in ascending order based on the height (H) dimension (ST553).

[0241] Next, the control unit 3 sorts the packages in the same group in the temporary placement area in descending order by the number of packages (ST554). Fig. 37 shows an example of the sorting result.

[0242] Next, the control unit 3 selects one of the bales at the top based on the sorting results of ST551 to ST554 (ST555). Next, the control unit 3 determines whether the search method for the search cell is method B or not (ST44). Note that in the temporary placement area, the bales are placed in a non-stacked state. In other words, all the bales placed in the temporary placement area are in the first tier.

[0243] If the control unit 3 determines that the search method of the search cell is method B (YES in ST44), it sets the search range to the basic search range (ST46) and proceeds to ST47. If the control unit 3 determines that the search method of the search cell is not method B (NO in ST44), it sets the search range to the basic search range and wide search range (basic search range + wide search range) (ST45) and proceeds to ST47.

[0244] In ST47, the control unit 3 extracts a search cell on the palette within the search range set in ST46 or ST45. Next, the control unit 3 executes a search cell evaluation process (process A) for the extracted search cell (ST48). Process A is as described above. Next, the process proceeds to ST49.

[0245] In ST49, the control unit 3 determines whether or not the number of arrangement patterns in which the packages can be arranged (number of possible arrangement patterns) is 0.

[0246] When the control unit 3 determines that the number of possible arrangement patterns is not 0 (NO to ST49), it determines the arrangement pattern of the search cell having the arrangement pattern with the highest score as the arrangement position of the package P (determination of arrangement position information) (ST50). Next, the control unit 3 generates a robot control signal for controlling the drive of the stowage robot 9 based on the determined arrangement position information (ST51), and outputs this to the stowage robot 9 (ST52). The stowage robot 9 operates based on the robot control signal, and transfers the top package selected in ST555 to a pallet. Next, the process proceeds to ST556.

[0247] When the control unit 3 determines that the number of possible arrangement patterns is 0 (Yes in ST49), it proceeds to ST556. The control unit 3 removes the topmost pack (ST556). That is, in the sorting results of ST551 to ST554, by removing the topmost pack, the pack that was located second becomes the topmost. When the steps from ST555 onwards are executed for this pack that was located second, and the pack that was located second is placed on the pallet, the state of the board changes, and the removed topmost pack may also become available for placement. For this reason, when the control unit 3 places the pack that was located second on the pallet, it performs the steps from ST555 onwards including the pack that was removed in ST556. That is, the removed topmost pack becomes the topmost pack. This is repeated for all packs that were in the temporary placement area until the processing from ST555 onwards is completed.

[0248] In this way, when transporting the packages temporarily stored in the temporary storage area to a pallet, packages with a relatively large base area are preferentially transported among the packages temporarily stored in the temporary storage area. This is because relatively small packages tend to be easier to stack, so packages with a relatively large base area that tend to be more difficult to stack are transported first, thereby increasing the success rate of stacking.

[0249] Process B (ST61) will be described according to the flow in Fig. 23. Process B relates to the arrangement of packages to be loaded on a pallet or a temporary placement area. As shown in Fig. 23, process B is almost the same as ST44 to ST52 in the flow in Fig. 21, except that the process ends when the number of possible arrangement patterns is 0 (YES in ST49).

[0250] ((Temporary Area Placement Conditions)) The temporary storage area placement conditions are conditions for preferentially placing a package P to be stowed in the temporary storage area 25 (temporary storage pallet TPL). The decision as to whether or not to preferentially transport a package P to be stowed to the temporary storage area 25 is made using information about the group.

[0251] Using information about multiple groups (two groups in the shipping packing information shown in FIG. 30), the control unit 3 determines groups that satisfy all three conditions (1) to (3) below as groups that satisfy the placement conditions for the temporary placement area. It then determines the other groups as groups that do not satisfy the temporary placement area placement conditions and are to be preferentially placed on the temporary placement pallet TPL. In the following description, the "short side" refers to the side in the width (W) direction, and the "long side" refers to the side in the length (L) direction.

[0252] (1) A group in which the maximum weight of parcels belonging to the same group is below the 75th percentile of all groups. Since the weights of multiple parcels belonging to the same group may differ, the maximum parcel weight within the group is used as the representative value. (2) A group in which the short sides of bales belonging to the same group are equal to or less than the 25th percentile of all groups. In this embodiment, bales whose dimensions match within 1 mm are in the same group, so the short side dimension value of the group and the short side dimension value of the bales belonging to the group are the same. (3) A group in which the long side of bales belonging to the same group is equal to or less than the 25th percentile of all groups. In this embodiment, bales whose dimensions match within 1 mm are in the same group, so the long side dimension value of the group and the long side dimension value of the group are the same.

[0253] A supplement to (2) and (3) above: "Groups whose short side (long side) is the 25th percentile or less of all groups" refers to the top 5 groups with the smallest numerical values ​​when the short side (long side) of the groups are sorted in ascending order, assuming that there are 20 total groups. Similarly, in (1) above, "groups whose maximum weight is the 75th percentile or less of all groups" refers to the top 15 groups with the smallest numerical values ​​when the maximum weight of the groups is sorted in ascending order, assuming that there are 20 total groups.

[0254] Bamboos that belong to a group that satisfies all of the above conditions (1) to (3) (a group that satisfies the temporary storage area arrangement conditions) are bamboos that are relatively light and have a relatively small bottom surface. By preferentially arranging such relatively light bamboos with small bottom surfaces in the temporary storage area, ultimately, pallet PL can be loaded with the relatively light bamboos with small bottom surfaces arranged at the top of the group of packages to be loaded onto pallet PL. This allows multiple bamboos to be arranged on pallet PL more efficiently, and also allows multiple bamboos to be arranged on pallet PL in a stable arrangement that is less likely to collapse.

[0255] As described above, the information processing system of the present invention uses past actual stowage data to prepare in advance a score parameter set according to the relationship between the number of bales to be stowed on a pallet and the oligopoly state of the bales, and calculates the score of a search cell for determining the placement position of the bales on the pallet using this score parameter set. With this configuration, multiple bales can be efficiently placed on a pallet.

[0256] <Modification> Although the embodiment of the present invention has been described above, the present invention is not limited to the above-described embodiment, and it goes without saying that various modifications can be made without departing from the spirit of the present invention.

[0257] [Variation 1] In the above embodiment, an example has been given in which the packs P are loaded onto the pallets PL by the loading robot 9, but the loading may be performed by a worker.

[0258] For example, the control unit 3 may generate image information showing the positions of the packages to be stowed on the pallet PL viewed from above, based on position information of the packages to be stowed, and display the image information on the display unit 16. The worker can place the packages to be stowed while referring to the image displayed on the display unit 16. This allows efficient stowage even if the worker lacks stowage skills.

[0259] Furthermore, the control unit 3 may generate audio information in addition to or instead of the image information based on the position information of the packages to be stowed, and output the audio information using an audio output unit such as a speaker. The audio information may include, for example, the position information of the packages to be stowed. In this way, stowage instructions may be given by audio information in addition to or instead of image information.

[0260] In addition, the control unit 3 may generate video information indicating the placement positions of the packages to be loaded, in addition to or instead of the above-mentioned image information, based on the placement position information of the packages to be loaded, and project the video information onto the pallet PL.

[0261] In this way, the placement position information of the packages to be stowed may be used to generate image information, audio information, video information for projection, etc., to assist the worker in the stowage work.

[0262] [Variation 2] In the above-described embodiment, a flat pallet is given as an example of a loading platform, but the present invention is not limited thereto. The loading platform may be, for example, a box pallet, a roll box pallet, a post pallet, a dolly, or the like, as long as it has a loading surface with a limited size. In addition, a part of an area in a room may be made to function as a loading platform, and the present invention may be applied to organizing objects. In addition, in the above-described embodiment, a roughly rectangular shaped package such as cardboard boxes is given as an example of an object, but the shape of the package is not limited to a roughly rectangular shaped package, and the present invention may be applied to packages of other shapes.

[0263] [Variation 3] In the above embodiment, the case where one pallet PL is required for one shipping destination has been described as an example, but multiple pallets PL may be required for one shipping destination. For example, when there are multiple pallets PL and packing information for stowage on the pallet is acquired for each pallet, the stowage processing according to the present invention may be performed for each pallet PL.

[0264] [Variation 4] In the above embodiment, an example was given in which the information processing system 100 is configured by a plurality of servers such as the automated warehouse control server 6, the shipping management server 7, and the information processing device (loading server) 1, but it may also be configured by, for example, a single server. Conversely, the control unit 3 does not need to be a single information processing device, and the operations of the control unit 3 may be performed by a plurality of information processing devices working together.

[0265] In the above embodiment, an example has been given in which the information processing device 1 has the control unit 3 and the memory unit 4, but the DB unit 40 included in the memory unit 4 may be configured as a separate server different from the information processing device 1. In addition, the DB unit 40 of the information processing device 1, the various DBs 73 to 75 of the shipping management server 7, and the packing sequence information DB 63 of the automated warehouse control server 6 may be configured as one server or multiple servers.

[0266] [Variation 5] In the above embodiment, four categories are prepared, and the pallets to be stowed are classified into one of the four categories. However, the number of categories is not limited to four. For example, the number of packages may be classified into three categories and the HH index (oligopoly state) into two categories, making a total of six categories, or the number of packages may be classified into three categories and the HH index (oligopoly state) into three categories, making a total of nine categories. The number of categories may be determined by the number of packages and the HH index, depending on the distribution of points when the number of packages and the HH index are plotted as indices, using past stowage performance data as shown in FIG. 14. By dividing the categories in this way, more efficient stowage is possible.

[0267] [Variation 6] In the above embodiment, an example is given in which the oligopoly state is expressed by the HH index, but it is not limited to the HH index, and entropy may be used. Entropy indicates the degree of disorder in pallet stacking. A method of calculating entropy will be described with reference to FIG. 39. In FIG. 39, PL01 to PL09 are pallets on which different types and numbers of bales are stacked. P1 to P4 indicate different types of bales. The number of transport patterns (transport sequence data) is calculated for each of pallets PL01 to PL09. For example, in PL09, all the bales are of the same type, so there is only one transport pattern. Entropy is expressed as the logarithm log(base e) when the number of transport patterns is N. e It can be found in N. [Explanation of symbols]

[0268] 3. Control section 100...Information processing systems P…Package (Object) PL…Pallet (loading platform) x1~x14…Evaluation function

Claims

1. An information processing system for determining the placement position of an object when it is loaded onto a loading platform, Object information, including the number of objects, dimensions, and weight of the objects to be loaded onto the loading platform, is acquired. Using the aforementioned object information, objects whose differences in dimensions and weight are below a threshold are classified as the same group to be loaded onto the loading platform, thereby creating one or more groups. Using the aforementioned object information, the oligopoly status of objects is calculated, which is the proportion of the total number of objects belonging to the group. Based on the number of objects and the oligopoly status, the loading platform is classified into one of several categories to which a score parameter set representing a collection of weight coefficients for each of several pre-prepared loading evaluation functions is associated, thereby determining the category of the loading platform. Using the score parameter set associated with the determined category, the placement position of the object to be stacked on the loading platform is determined. Equipped with a control unit Information processing system.

2. The aforementioned objects are transported one by one to the loading position on the loading platform. Multiple score parameter sets are provided for each of the aforementioned categories. The control unit, Using the object information, multiple dummy transport patterns of the objects with different transport orders are created. For each of the score parameter sets, a stacking calculation is performed using the score parameter set to load the objects to be transported in each dummy transport pattern onto the loading platform. For each of the score parameter sets, the stacking success rate is calculated using the results of the stacking calculation, which is the ratio of the number of dummy transport patterns that were successfully stacked to the number of dummy transport patterns created. For each of the score parameter sets, the space occupancy rate is calculated when the objects transported by the created dummy transport patterns are loaded onto the loading platform using the results of the stacking calculation. For each of the score parameter sets, the surface area of ​​the object group when the objects transported by the created dummy transport patterns are loaded onto the loading platform is calculated using the results of the loading calculation, and further, the surface area ratio is calculated, which represents the ratio of the surface area of ​​the object group to the sum of the surface areas of each object loaded onto the loading platform. Using the stacking success rate, the space occupancy rate, and the surface area ratio, a set of score parameters is determined from a plurality of score parameter sets to determine the placement position of the object to be stacked on the loading platform. The information processing system according to claim 1.

3. The evaluation function includes: an evaluation function relating to the degree of contact between the side of the object to be stacked and other objects; an evaluation function relating to the degree of contact between the top surface of the object to be stacked and the top surface of other objects at the same height as the top surface; an evaluation function relating to the height of the object to be stacked; an evaluation function relating to the fact that the stack of objects to be stacked is placed on an object of the same type as itself; an evaluation function relating to the fact that the stack of objects to be stacked is placed on an object of a different type than itself; an evaluation function calculated by squaring the area ratio of the contact portion between the base supporting the object to be stacked and the bottom surface of the object to be stacked with respect to the bottom surface of the object to be stacked; an evaluation function calculated using the HH index based on the area ratio of the contact surface between the object below and the object to be stacked with respect to the bottom surface of the object to be stacked when the object to be stacked is placed on top of an object already placed on the loading platform; and an evaluation function relating to the expected value of how many places can be placed where objects will be transported after the object to be stacked has been placed. The information processing system according to claim 1 or 2.

4. The aforementioned object is rectangular in shape, The control unit, In a top view, lines are virtually arranged in a grid pattern on the loading platform, and multiple cells are set up, separated by these lines. Based on the aforementioned line, multiple cells that can serve as candidate placement locations where the corners of the objects to be stacked overlap are extracted as search cells. Using the score parameter set used to determine the placement position of the object to be stacked on the loading platform, each of the search cells is evaluated, and the search cell with the highest evaluation is determined as the placement position of the object to be stacked. The information processing system according to claim 1 or 2.

5. In determining the placement position, the control unit evaluates the search cells extracted using each of the multiple search methods prepared in advance for extracting the search cells. The aforementioned multiple search methods are, The first search method uses search cells extracted by basic search and search cells extracted by wide search as evaluation targets, A second search method which, if a search cell can be extracted by the basic search, adopts that search cell as the target for evaluation, and if a search cell cannot be extracted by the basic search, adopts the search cell extracted by the wide search as the target for evaluation. Includes, The aforementioned line consists of vertical and horizontal lines that are perpendicular or perpendicular to each other. The search cells extracted by the basic search described above are those where no object is placed, or, when the loading platform on which the object is placed is viewed from above, adjacent to a line in which the height changes when viewed in a direction parallel to the vertical line, and adjacent to a line in which the height changes when viewed in a direction parallel to the horizontal line. The search cells extracted by the wide search are those where no object is placed, or, when the loading platform on which the object is placed is viewed from above, adjacent to a line in which the height changes when viewed in either the direction parallel to the vertical line or the direction parallel to the horizontal line, and adjacent to a line in which the height changes when viewed in the other direction, or an extension line of said line. The information processing system according to claim 4.

6. The object is placed in the loading platform or a temporary storage area located in an area different from the loading platform. The objects are placed on the loading platform or in the temporary storage area in the order of transport. The control unit, Using the object information, create multiple groups, If the object to be stacked satisfies the conditions that it is relatively light and has a relatively small base when compared to the objects belonging to each of the multiple groups, or if it does not satisfy the conditions and cannot be placed on the loading platform, the object to be stacked is temporarily placed in the temporary storage area. The information processing system according to claim 1 or 2.

7. The aforementioned oligopoly state can be expressed as the HH exponent or entropy. The information processing system according to claim 1 or 2.

8. An information processing method performed by an information processing device, Object information, including the number of objects to be loaded onto the loading platform, dimensions, and weight, is obtained. Using the aforementioned object information, objects whose differences in dimensions and weight are below a threshold are classified as the same group to be loaded onto the loading platform, thereby creating one or more groups. Using the aforementioned object information, the oligopoly status of objects is calculated, which is the proportion of the total number of objects belonging to the group. Based on the number of objects and the oligopoly status, the loading platform is classified into one of several categories to which a score parameter set representing a collection of weight coefficients for each of several pre-prepared loading evaluation functions is associated, thereby determining the category of the loading platform. Using the score parameter set associated with the determined category, the placement position of the object to be stacked on the loading platform is determined. Information processing methods.

9. In an information processing device, A step of acquiring object information including the number of objects to be loaded onto the loading platform, dimensions, and weight. Using the aforementioned object information, the steps include classifying objects to be loaded onto the loading platform by grouping objects whose differences in dimensions and weight are below a threshold into the same group, thereby creating one or more groups; Using the aforementioned object information, the step of calculating the oligopoly status of objects, which is the proportion of the total number of objects belonging to the group, Based on the number of objects and the oligopoly status, the loading platform is classified into one of several categories to which a set of score parameters, each representing a collection of weight coefficients for a pre-prepared evaluation function for multiple loading methods, is associated, thereby determining the category of the loading platform. The steps include determining the placement position of the object to be stacked on the loading platform using the score parameter set associated with the determined category, and A program that executes the command.

10. In an information processing device, A step of acquiring object information including the number of objects to be loaded onto the loading platform, dimensions, and weight. Using the aforementioned object information, the steps include classifying objects to be loaded onto the loading platform by grouping objects whose differences in dimensions and weight are below a threshold into the same group, thereby creating one or more groups; Using the aforementioned object information, the step of calculating the oligopoly status of objects, which is the proportion of the total number of objects belonging to the group, Based on the number of objects and the oligopoly status, the loading platform is classified into one of several categories to which a set of score parameters, each representing a collection of weight coefficients for a pre-prepared evaluation function for multiple loading methods, is associated, thereby determining the category of the loading platform. The steps include determining the placement position of the object to be stacked on the loading platform using the score parameter set associated with the determined category, and A computer-readable, non-temporary storage medium that contains a program to execute a program.

Images