Method for preparing multi-parcels

Abstract

The invention relates to a computer-implemented method (100) for preparing multi-parcels, the method comprising the following steps: - obtaining (102) at least one group of items to be shipped; - obtaining (105) at least one list of virtual packages; - selecting (106, 107, 108, 109), from the list of virtual packages, a virtual package as a function of at least one dimension of at least one item of the or each group of items to be shipped; - opening (110) at least one instance of the virtual package; - for each item of the group or of at least one of the groups, item by item: - arranging (111, 112, 113, 114, 115) the item in the or one of the open instances as a function of the remaining capacity of the instance, this remaining capacity corresponding to the remaining maximum supported weight of the package, and as a function of the remaining available spaces in the instance; - updating (116) the remaining capacity of the instance in which the item is arranged, and updating the remaining available spaces in the instance; - forming (134) theoretical parcels, each theoretical parcel comprising the items arranged in one and the same instance.

Description

Multi-package preparation process

[0001] The invention relates to the field of parcel delivery, in particular order preparation in a warehouse.

[0002] In the current state of the art, in a warehouse, the order picker has a list of items to be shipped to a single recipient. The items on this list, intended to be grouped into a single package, form a logistics unit of measure (LOM). The list of items to be shipped is therefore designed to be transformed into one or more LOMs for shipment. A "logistics unit of measure" is generally called a "package" when it is associated with an order being picked or picked. The logistics specialist or order picker also has the machines or packaging materials needed to pack these LOMs for shipment.

[0003] For a given logistics unit, the packaging choice is generally made visually by the order picker based on their estimate of the volume of the products comprising the unit and their estimate of the packaging volume. However, faced with time constraints and a limited selection of container types, many choices are suboptimal. Poorly arranged within the packaging, the items occupy too small an overall volume. Consequently, the packaging for logistics units, chosen in an overly bulky manner, results in an average empty space that is too large. This average empty space leads to, among other drawbacks, overconsumption of packaging, overconsumption of storage space, significant material waste, associated greenhouse gas emissions, and major economic losses for those involved in the delivery sector.

[0004] A method for optimally arranging items within a logistics unit is already known, according to documents FR3141547A1, FR3141453A1, FR3141452A1, and FR3141546A1 filed on behalf of the applicant. These documents also describe a method for assigning packaging to select the optimal packaging for a given logistics unit, based on the items to be packaged.

[0005] However, prior to these optimal arrangement and allocation steps, it is often necessary to divide the list of items to be shipped into several logistics load units. Yet, the grouping of items in this list is done arbitrarily, for example, by taking items in the order they appear in an order. Therefore, determining the items that make up each logistics load unit is not optimized. In particular, these groupings do not minimize empty space within the logistics load units.

[0006] The invention aims in particular to group items, from an order of items to be shipped, in an optimal way for their future packaging, in order to reduce the empty space in the packages finally shipped.

[0007] To this end, the invention relates to a method for preparing multi-packages, implemented by computer and comprising the following steps: - obtaining at least one group of items to be shipped; - obtaining at least one list of virtual packages; - selecting, from the list of virtual packages, a virtual package based on at least one dimension of at least one item from the group or groups of items to be shipped; - opening at least one instance of the virtual package; - for each item in the group or at least one of the groups, item by item: arranging the item in the open instance or instances based on the remaining capacity of the instance, this remaining capacity corresponding to the maximum weight the package can support, and based on the remaining available spaces in the instance; updating the remaining capacity of the instance in which the item is arranged, and updating the remaining available spaces in the instance;- formation of theoretical packages, each theoretical package comprising the items arranged in the same instance.;

[0008] Thus, items are grouped into one or more instances of a virtual package. The selection of the virtual package is not arbitrary; it is based on at least one dimension of an item in the group. As a result, the choice of packaging in which the items are virtually arranged is optimized. In particular, the empty space in the packages generated from the item list is reduced on average thanks to these steps, and this reduction is tailored to each order. Taking into account the maximum remaining capacity during each arrangement prevents excessively heavy items from being grouped together. Arranging items based on available space optimizes the layout and therefore minimizes the empty space in each instance. The result of these steps is the reorganization of items into packaging instances in which the items are distributed and arranged in an optimized manner, at least with regard to the generated empty space.The term "empty space ratio" refers to one way, among others, of quantifying the empty space in a container containing items, by relating the available volume in the container to the volume occupied by the items within that container. Reducing the empty space ratio therefore reduces the amount of empty space generated.

[0009] The theoretical packages can then be assigned packaging in order to form the logistical load units.

[0010] Other optional features follow, taken alone or in combination.

[0011] Preferably, the process includes, when the open instance or none of the open instances has sufficient remaining capacity and / or available space to accommodate the item, opening a new identical instance of the virtual packaging to arrange the item in it.

[0012] Thus, a new, identical instance is opened to arrange the subsequent items from the group(s). Choosing identical instances, that is, representing the same virtual packaging previously selected based on the items, facilitates the optimized grouping of items and therefore reduces the amount of empty space generated at the end of the process.

[0013] Advantageously, the process includes the following steps for selecting the virtual packaging: - obtaining a list of packagings; - for each packaging in the list, obtaining the dimensions of the packaging and the capacity of the packaging, this capacity corresponding to the maximum weight supported by the packaging; - choosing the packagings from the list having a capacity greater than the heaviest item in the group(s) of items and less than a predetermined capacity value, having a volume greater than the largest item in the group(s) of items and less than the cumulative volume of the items in the group(s) of items, and having dimensions greater than the largest dimensions of the items in the group(s) of items; - among the chosen packagings, sorting the packagings according to one or more predetermined criteria; - among the sorted packagings, selecting the first packaging.

[0014] Thus, the virtual packaging is sufficiently robust and large to contain any item from any group, especially the largest one, but it cannot contain all the items from all groups. It is therefore specifically chosen to distribute the items across at least two instances.

[0015] Preferably, the process includes, for arranging at least one of the articles of the or at least one of the groups, traversing, in order of opening, the instances already open and, when an open instance has sufficient capacity and remaining available spaces, the virtual arranging of the article in the instance.

[0016] Therefore, a new instance is not opened for each group of items. Instead, each existing open instance containing items from one or more groups can be rearranged with items from one or more other groups. The input groups are thus reorganized differently within the instances. By proceeding in this way, we ensure that the empty rates of each instance, and consequently of the packages subsequently formed, are minimized.

[0017] Advantageously, the process includes the following preliminary steps: - obtaining a list of items to be shipped; - dividing the items into several groups of items, each or at least some of the groups comprising only items having in common at least one predetermined characteristic.

[0018] Thus, starting from a pre-existing list of items, typically from a single order destined for the same recipient, the items are grouped into different categories based on specific criteria. Indeed, recipients and order pickers may prefer to handle items grouped by common characteristics. These groups are then further consolidated into instances, while still respecting the groupings based on criteria. In fact, based on predetermined parameters, it may be decided that items with common characteristics should not be mixed with other items within the instances. In this case, for each new group of items to be arranged that includes common characteristics, a new instance is opened to begin the arrangement of the items in that group, even if other instances are already open and may have sufficient space and capacity.

[0019] Preferably, one or at least one of the characteristics common to articles in at least one of the same group is a common article model.

[0020] Thus, items are grouped upstream, within item groups, by model, for example, by item reference. It is indeed useful for an order picker to handle identical items and arrange them in common packaging or at least packaging located close to each other. Items of the same model but different colors can also be grouped together according to a predetermined setting. Alternatively, it can be configured that only completely identical items are grouped together.

[0021] Preferably, the process includes, for virtually arranging the articles of a group having the same article model in one or more open instances virtually already containing articles of one or more other models, the following steps: - identification, among the open instances, of one of the instances capable of containing all the articles of the group according to at least the remaining capacity of the instance and the remaining available spaces in the instance, - formation of a package including the articles respectively arranged in this instance.

[0022] Thus, if an already open instance can contain all identical articles, they are arranged in that instance.

[0023] Advantageously, the process includes, for virtually arranging the items of a group having the same item model in one or more open instances virtually already containing items of one or more other models, the following steps: - obtaining the number of items of the group, having the same model, to be arranged in the already open instances; - obtaining a maximum number of instances allowed to contain these items having the same item model with items of other models; - for each of the already open packaging instances, determining the remaining capacity of the instance, this remaining capacity corresponding to the maximum weight supported remaining by the packaging, determining the dividend from the division of the remaining capacity by the weight of an item to be arranged, and determining the modulo from this division;- for a value k starting at 2 and, at most, ending at the maximum number of allowed instances: formation of all combinations, called k-tuples, of already open packaging instances, each k-tuple comprising k instances; for each k-tuple, determination of the sum of the dividends of the instances of the k-tuple, and determination of the sum of the modulos of the instances of the k-tuple; selection of the k-tuples whose sum of dividends is greater than or equal to the number of items to be arranged; sorting of the selected k-tuples by increasing sum of the modulos;for each k-tuple in sort order, attempt to virtually arrange the items in the instances of the k-tuple, and, when all the items can be virtually arranged in these instances, formation of theoretical packages comprising the items respectively arranged in these instances of this k-tuple, or, when at least one of the items cannot be arranged in one of the k-tuples, increment k and reiteration of the previous steps with the new value of k.;

[0024] These steps are implemented particularly when a user authorizes the "unbundling" of items of the same model to place them in already open instances that already contain items of a different model. This minimizes the empty space rate of the instances and therefore of future logistics workloads. These steps also minimize the number of open instances, and thus the number of theoretical packages formed, and consequently, the number of packages. This, in turn, reduces the shipping costs of the items. The calculations performed take advantage of the fact that, since the items are identical, their weights and dimensions are identical.

[0025] Preferably, the process includes, before the virtual arrangement attempt, a verification step requiring that the group's items be arranged in k-tuple instances opened only consecutively.

[0026] Therefore, it is required that identical items, even if placed in separate containers with different items, not be distributed across open containers too far apart. Indeed, by creating consecutive theoretical packages containing these identical items, real logistics load units are generated, allowing the order picker to quickly, and even simultaneously, arrange all identical items in the designated packaging.

[0027] Advantageously, the process includes, when the open instances cannot contain all the identical items, opening a new identical instance of the virtual packaging to virtually arrange, item by item, as many items of the same model as possible.

[0028] Thus, if the already open instances cannot contain the articles of the same model according to the stated parameters, a new instance is opened, and the articles of the same model are arranged, article by article, within that instance. For the remaining articles that cannot be arranged in this new instance, due to insufficient capacity or available space, the previous steps of identifying an already open instance to contain these remaining articles are repeated, and, failing that, these remaining articles are distributed across the k-tuples of already open instances.

[0029] Advantageously, to achieve a virtual arrangement of an item in an instance, the items and the instance having respective virtual rectangular prism shapes, the following steps are implemented: • determination of the dimensions and coordinates of several remaining empty spaces of the instance, distinct from each other and having respective virtual rectangular prism shapes; • determination of possible positionings of the item in each empty space; • determination of respective distances between a predetermined vertex of the instance and the respective vertices of the item in each possible positioning; • based on the determined distances, selection of one of the possible positionings as the virtual arrangement of the item in the instance.

[0030] Thus, items are placed in an instance based on the maximum available space remaining within that instance. This minimizes the instance's empty space rate. Specifically, these virtual arrangement steps are repeated for each item, one after the other.

[0031] Preferably, the process includes a step of sorting the articles of one or each group according to at least one predetermined criterion chosen from the volume of each article, the weight of each article, and the density of each article, so as to form an ordered sequence of the articles of the group to be arranged virtually in the instance or instances.

[0032] Thus, the arrangement order of the items in a group depends on the characteristics of the items and is obtained simply, in order to minimize the rate of empty space generated at the end of the process while limiting calculation times.

[0033] Advantageously, the process includes a step of sorting the groups according to at least one predetermined criterion chosen from the cumulative volume of the articles in each group, the cumulative weight of the articles in each, and the cumulative density of the articles in each group, so as to form an ordered sequence of the groups whose articles are arranged virtually in the instances.

[0034] Thus, the groups are ordered to determine which items within each group should be arranged sequentially. The order in which the groups are arranged depends on the characteristics of the items in each group and is determined simply, in order to minimize the amount of empty space generated at the end of the process while also limiting computation time.

[0035] Preferably, the sequence of article groups is ordered so that the group(s) containing articles with the same pattern are placed before the group(s) of articles containing articles that do not have a common pattern.

[0036] Thus, the "outside groupings" articles are arranged virtually, at the end of the process, in the instances already containing the articles intended to be grouped.

[0037] Preferably, one or at least one of the characteristics common to the items in a group is chosen from the following: - at least certain dimensions of the items; - a level of fragility of the items; - a need for a specific type of packaging for the items, for example single-wall corrugated packaging, double-wall corrugated packaging, or a plastic bag; - the orientation of the items; - the cost of the items; - a level of weight of the items; - the fact that the items are pre-packaged; - a maximum packaging rate not to be exceeded for the items; and - a choice of a shipping preparation area for the items.

[0038] Thus, items are grouped together before their virtual arrangement, based on similarities, structural criteria, or other shared constraints. Processing items together that require, for example, the same type of packaging or have identical dimensions allows for a better response to business and organizational constraints. For instance, fragile items are grouped together for manual preparation because they require specific cushioning to minimize the risk of breakage.

[0039] Advantageously, the process then includes, for each theoretical package formed, a step of assigning a package carried out in the following way: - selection of available packages; - for each package selected, determination of a prioritization value of the package, the value depending on at least the volume of the package so that a package less voluminous than another package, is, all other things being equal, priority over this other package; - sorting of the packages according to their respective prioritization values ​​so as to form a sorted list of packages; - choice of one of the packages from the list according to the sorting; - verification of the suitability between the chosen package and a theoretical package to be packed, the verification relating at least to the weight of the theoretical package in relation to a maximum weight supported by the package.

[0040] Thus, these packaging assignment steps allow the selection of packaging, whose dimensions may be different from those of the previously chosen virtual packaging instance, adapted to the items in the instance, in order to minimize the empty space in the packaging.

[0041] Preferably, for each chosen package, the steps of the virtual arrangement, as stated above, of the items of the theoretical package in the packaging are implemented in order to determine the arrangement plan of the items in the packaging.

[0042] Thus, since the chosen packaging may have different dimensions than the previously selected virtual packaging instance, the arrangement may differ from that implemented in the instance. In other words, the item arrangement is reset to optimize the packaging's empty space. When several packages can be verified as suitable for the theoretical package, virtual arrangements are created for each of them. In this case, at the end of the process, the package with the lowest generated empty space is selected.

[0043] The invention also relates to a computer program comprising instructions which, when the program is executed by a computer, lead the computer to implement the steps of the process stated above.

[0044] The invention also relates to a computer-readable recording medium comprising instructions which, when executed by a computer, lead the computer to carry out the steps of the process described above.

[0045] The data processing device includes a processor configured to implement the steps of the process described above. Brief description of the figures

[0046] The invention will be better understood upon reading the following description, given solely by way of example and made with reference to the accompanying drawings in which:

[0047] is a diagram of computer resources enabling the implementation of the process of the invention;

[0048] is a flowchart of a process for preparing the invention according to an implementation method;

[0049] is an example diagram of a list of items to be shipped;

[0050] is a pattern for forming groups of articles from the list of the;

[0051] is a diagram of a list of virtual packages;

[0052] is a diagram of the result of a step in the process of the;

[0053] is a diagram of the result of a step in the process of the;

[0054] is a diagram of the result of a step in the process of the;

[0055] is a diagram of the result of a step in the process of the. Detailed description

[0056] Figure 1 shows the computer means for implementing process 100, which will be described later. These computer means 1, conventional for a person skilled in the art, include a data processing device 2, a data storage device 3 containing a computer program 4 and a database 5, as well as interaction means 6 for a potential user, such as a keyboard and mouse. All of these means 1 together form a conventional computer, but they can also be distributed across any remote hardware. The program 4 comprises instructions which, when executed by device 2 or any computer, cause them to carry out the steps of process 100 described below.Consequently, the medium 3 is any recording medium readable by the device 2 or by any computer, comprising instructions which, when executed by the device 2 or a computer, cause them to implement the steps of the process 100 described below. It also follows that the data processing device 2 includes a processor configured to implement the steps of the process 100.

[0057] The interaction means 6 allow a user to configure the process 100, in particular by assigning chosen values ​​to certain thresholds or variables, or by selecting certain characteristics rather than others, in accordance with the thresholds, characteristics, and other configurable parameters that will be clear to a person skilled in the art in the processes described below. Whether the steps are described indefinitely, particularly with the use of the pronoun "one," or whether computer means are explicitly mentioned, the virtual arrangement, comparison, selection, assignment, and other computational steps described, as well as, in general, all steps that will be clear to a person skilled in the art as being computer-based, are entirely implemented by the means 1, unless human intervention, for example by a user of the processes, is explicitly mentioned.

[0058] In the following, "article", which could also be called "product", will refer to any object, virtual or real, intended to be integrated into real or virtual packaging.

[0059] The term "packaging" refers to any object, virtual or real, intended to house one or more items.

[0060] Unless explicitly stated otherwise, both types of objects—articles and packaging—represent virtual shapes of perfect rectangular prisms. Naturally, the "real" articles and packaging corresponding to these virtual elements do not necessarily have the shape of a rectangular prism, much less a perfect one.

[0061] It will be clear to those skilled in the art that they are manipulated virtually, that is, solely through the lens of their data, by means 1, as described below. Thus, they are represented digitally by their three dimensions, and if necessary by their coordinates, in particular those of the items within a virtual packaging instance, in the form of structured data in database 5. It is through this data that the computer means 1 process these elements in the process 100 described below. The manner in which these dimensions are obtained and the manner in which this data is organized are not the subject of this request.

[0062] Similarly, database 5 contains a multitude of variables relating to specific characteristics of the items, such as weight, whether an item is considered "fragile" or not, the type of packaging required, and all parameters as clearly indicated in the description. The database also contains the necessary packaging characteristics. The method by which this data is obtained and organized into databases will not be described here and is not the subject of this application.

[0063] A "packaging instance," or simply an "instance," refers to a necessarily virtual copy of a package. In other words, it is a virtual copy corresponding to a previously selected packaging model. When an instance is "opened," no actual action is performed on the package; rather, a virtual package corresponding to a previously chosen packaging model is viewed in order to virtually arrange one or more items within it. An "open" instance is therefore a virtual copy of a package that has been opened previously and may already contain one or more virtual items that have been virtually arranged within it.

[0064] "Virtual arrangement" refers to identifying how to place a virtual item within a packaging instance. The result of the virtual arrangement includes, among other things, the coordinates of the virtual item arranged within the instance.

[0065] A "theoretical package" refers to a packaging instance and the items virtually arranged within it. A theoretical package is therefore a virtual object. When a real package is assigned to a theoretical package—that is, when a real package is chosen to hold the items of the theoretical package—the theoretical package becomes a "logistics unit of load."

[0066] The term "logistics load unit" refers to a set of items, including one or more commercial products, intended to be grouped together and arranged in a specific common package for delivery. This expression differs from the term "parcel" in that a parcel is often associated with a state or status of the logistics load unit. In particular, a parcel is generally considered a logistics load unit when it bears a delivery label, is in transit, or has arrived at its destination. The "logistics load unit" is distinct from any status related to a delivery. As with the other items described, it will be clear to those skilled in the art that the term "logistics load unit" regularly refers to a logistics load unit handled solely in computerized form by means of [1].

[0067] The term "user" refers to any individual seeking to obtain the results of the processes described below. This may include, in particular, individuals working in a delivery warehouse, such as order pickers, buyers, or logistics personnel. It may also refer to the recipient of an order for goods.

[0068] We will now describe a multi-package preparation method, implemented by means 1 and schematically represented in Figure 1. We will describe it by illustrating it with an example. The steps described are naturally not limited to this example, particularly to the number, types, and dimensions of the illustrated items and packages. Furthermore, the illustrations are schematic and do not represent the precise dimensions and positioning of the objects shown. In particular, the items and packages are represented in two dimensions, whereas the invention manipulates them in three dimensions.

[0069] Step 101 involves obtaining a list of items 8 through 12 to be shipped. Items with the same numerical reference are identical. Thus, items 8 share the same item model. Item 9 shares another identical item model. This list originates from a user order, containing items destined for the same recipient. Alternatively, it could be a list of multiple orders grouped together for various purposes. In all cases, it consists of items to be shipped that must be processed as a single unit.

[0070] Step 102 involves dividing the items into several groups, 13, 14, and 15, as illustrated in Figure 1. In our example, some of these groups include only items sharing at least one predetermined characteristic. Thus, in our example, items 8 and 9 are grouped into groups 13 and 14, respectively. In other words, the items are grouped "by item reference." Therefore, within each of these groups, 13 and 14, the items in the group are identical. Alternatively, it can be configured that the items grouped by reference differ from each other by a color, or another characteristic that does not affect the item's shape, dimensions, or weight.

[0071] Group 15 therefore includes unique items, or "out-of-group" items, meaning those not grouped by reference with other items. In our example, these items have no common characteristics.

[0072] In another embodiment, the items in the resulting list are grouped according to another criterion. For example, one or more of the characteristics common to the items in a group is chosen from the following: - at least certain dimensions of the items; - a level of fragility of the items; - a need for a specific type of packaging for the items, for example single-wall corrugated cardboard, double-wall corrugated cardboard, or a plastic bag; - the orientation of the items; - the cost of the items; - a level of weight of the items; - whether the items are pre-packaged; - a maximum packaging rate not to be exceeded for the items; and - a choice of a shipping preparation area for the items.

[0073] These methods of implementation can be combined. Thus, for example, the items in a list obtained can be grouped by specific type of packaging and then, within these groups, be grouped by item references to form subgroups, which are subsequently treated like the other groups.

[0074] In our example, all items require the same type of packaging, and none of the other characteristics mentioned are relevant. If different packaging types were specified, the following steps could be implemented independently for each type. In other words, the multi-pack preparation process would then only apply to items of the same packaging type, and it would be repeated for items of a different packaging type. Conversely, it could be stipulated that different packaging types would be affected.

[0075] In an unillustrated variant, the list of items can be divided into groups of items that have no common characteristics, for example, when each item in the list is unique. In other words, the groups to be processed could all correspond to group 15, all made up of items of different models. In this case, they are all processed according to the steps described below for group 15.

[0076] The result of step 102 is, in our example, obtaining groups of articles 13 to 15 to be shipped, all these groups being intended for the same recipient, or in any case part of a common list 7 of articles to be processed together.

[0077] Step 103 involves sorting the groups based on at least one predetermined criterion chosen from among the cumulative volume of the items in each group, the cumulative weight of the items in each group, and the cumulative density of the items in each group, in order to form an ordered sequence of groups whose items will be virtually arranged in virtual packaging instances as described below. The density of an item is its weight per unit volume. The result of step 103 is therefore an ordered sequence of groups based on one or more of these criteria. In the example, they are considered to be ordered by cumulative volume, resulting in a sequence corresponding to the numerical references: 13, 14, then 15.

[0078] In one variation, the sequence of article groups is further ordered so that the group(s) containing articles with the same pattern are placed before the group(s) containing articles without a common pattern. This criterion is combined with the previous ones but can take precedence: the groups are ordered first according to these groupings, then according to the other criteria. In our example, the result would be the same: groups 13 and 14, containing articles with common patterns respectively, are placed before group 15.

[0079] Step 104, for each group, involves sorting the items in one or both groups according to at least one predetermined criterion chosen from among the volume of each item, the weight of each item, and the density of each item, in order to form an ordered sequence of the group's items to be virtually arranged in the instance(s). The result of this step is therefore, for each group, an ordered sequence of item arrangement. In our example, sorting the items in groups 13 and 14 is irrelevant since the items are identical. It may even be unnecessary. However, sorting is important for group 15. For group 15, sorting by volume results in a sequence starting with item 10, then 12, then 11.

[0080] The result of these steps 103 and 104 is that, in our example, we will first arrange the articles of group 13, then those of group 14, then those of group 15, starting with article 10, then continuing with article 12 and article 11.

[0081] Steps 105 to 109 aim to select a virtual package, which can be called a "pilot package", instances of which will be used later to house the items of each group.

[0082] Step 105 involves obtaining a list of 20 packages, numbered 21 to 24. These virtual packages correspond to actual packages already in stock at a logistics provider. Each of these packages can have different characteristics, such as different dimensions or capacities. The "capacity" refers to the maximum weight of items the package can support.

[0083] If the items were grouped by packaging type in step 102, then this step, and subsequent steps, only concern one packaging type. In other words, for a different packaging type, a new step 105 is performed to obtain a list of packages corresponding to this new packaging type. Alternatively, it may be possible to consider different packaging types. The packaging list then contains packages of potentially different types.

[0084] Alternatively, they correspond to packaging references available to the logistics provider. Conversely, they may be packaging previously selected by the logistics provider or another user.

[0085] Step 106 is, for each package in list 20, obtaining the package dimensions and package capacity.

[0086] Step 107 involves selecting packaging from the list that has a capacity greater than the heaviest item in the item groups and less than a predetermined capacity value. The packaging must also have a volume greater than the largest item in the item groups and less than the combined volume of the items in the item groups, and dimensions greater than the largest dimensions of the items in the item groups.

[0087] In our example, the largest items are the 8 items of group 13. The choice therefore falls, in the example, on the packages which successfully pass the criteria mentioned with regard to these 8 items. It follows, in our example, that only the packages 21 and 23, which are able to accommodate at least one of these 8 items, but not able to accommodate all the items of all groups 13, 14 and 15, and which have a capacity less than a predetermined capacity, are selected.

[0088] The predetermined capacity value is either set by the user or fixed by default. Its purpose is to prevent the selection of packaging with a capacity far exceeding that of typical actual packaging, which will be assigned packaging later. This maximum capacity can be the combined weight of the items in the group.

[0089] Furthermore, choosing packaging that is not suitable for all items in the group aims to force the distribution of items across at least two packages. Without this criterion, packaging large enough to hold all items could be chosen, and the subsequent steps of distributing items across multiple instances would be unnecessary.

[0090] Step 108 involves sorting the selected packages, in this case packages 21 and 23, according to one or more predetermined criteria. In one embodiment, the first criterion is the volume of the packages, with sorting performed in descending order of volume. For packages of equal volume, a second criterion is considered: the sum of the three dimensions of each package, with sorting performed in ascending order of sum. Other criteria could be selected alternatively or in addition to sorting the packages. In particular, if several different types of packaging are in the list, the packages can be sorted by type; for example, cardboard packaging might be preferred to plastic packaging. The result of step 108 is an ordered sequence of virtual packages. In our example, they are considered to be sorted by volume, resulting in a sequence that conforms to the numerical references: package 21 followed by package 23.

[0091] Step 109 involves selecting the first virtual package from among the sorted packages; in our example, package 21. The result of steps 105 through 109 is therefore the selection of a virtual package, which we can call the "pilot virtual package," and which will be used to arrange the items from all the groups. Thus, in our example, it is based on this virtual package 21 that list 7 of the items from all the groups to be shipped is created.

[0092] Alternatively, the choice of this pilot packaging can be defined manually, for example by the logistics manager, or via other criteria. It should be noted that this virtual packaging does not necessarily correspond to the actual packaging that will subsequently be assigned to each theoretical package created. These virtual packaging instances aim to create theoretical packages by grouping items potentially differently and arranging them within the instances.

[0093] Step 110 is the opening of a first instance 31 of the virtual packaging.

[0094] Step 111 is the virtual arrangement of a first article 8 of group 11 to a predetermined position of the first open instance 31.

[0095] To virtually arrange the remaining items of group 11 in instance 31, the steps in accordance with the arrangement process described in documents FR3141547A1, FR3141453A1, FR3141452A1 and FR3141546A1 are carried out. We therefore represent below only the main steps for a given article: a step 112 of determining the dimensions and coordinates of several remaining empty spaces of instance 31, distinct from each other and having respective virtual rectangular prism shapes; a step 113 of determining possible positionings of the remaining article in each empty space; a step 114 of determining respective distances between a predetermined vertex of instance 31 and the respective vertices of the remaining article in each possible positioning; depending on the distances determined, a step 115 of selecting one of the possible positionings as the virtual arrangement of the remaining article in instance 31.

[0096] Step 116 is the update of the remaining capacity of instance 31 in which the item is arranged, this remaining capacity corresponding to the maximum weight supported remaining by the package, and the update of remaining available spaces in the instance.

[0097] Steps 112 through 116 of the arrangement process are then repeated successively for each of the remaining 8 items in group 13 to be virtually arranged within the instance, as long as sufficient capacity and available space remain, in order to determine an arrangement plan for all items in the instance. The result of a virtual arrangement of an item within an instance includes the coordinates of the item within the instance where it is arranged.

[0098] Throughout the description that follows, when we talk about arranging an item, we will be referring to steps 112 to 115 for arranging that item in an instance or in a package.

[0099] In step 117, when the open instance, in our example instance 31, does not have sufficient remaining capacity and / or available space to accommodate the item, a new identical instance 32 of the virtual packaging 21 is opened to accommodate the remaining item 8. The result of this step is illustrated in the figure.

[0100] From this stage onwards, when we have at least two open instances, in our example instances 31 and 32, the choice of the instance in which to virtually arrange each article then depends on the group to which it belongs and on predetermined choices of a user of the process.

[0101] Thus, for groups 13 and 14 where articles are grouped by article reference, the steps implemented depend on a predetermined parameter to allow or not to "break" these groups in order to arrange their respective articles in already open instances, to minimize the respective empty rates of the instances.

[0102] If this permission is not enabled, to begin arranging articles from each of these groups—in our example, to begin arranging articles from group 14—a new instance is opened. This way, articles grouped by reference are not mixed with other articles, even if they might be spread across multiple instances containing only identical articles. This variant is not shown.

[0103] Conversely, if it is permitted to "break" these groups to distribute these articles into already open instances, steps 118 to 130 described below are carried out.

[0104] Step 118 is the attempt to identify, among the open instances, one capable of holding all the articles in the group, based at least on the instance's remaining capacity and available space. Alternatively, the remaining volume of each instance is determined beforehand.

[0105] Thus, in the order the instances were opened, starting with the first opened instance, the remaining volume of the instance is checked, and then, if sufficient, the remaining capacity, by comparing these values ​​to the cumulative capacities of the items. If they are sufficient, the available space is checked to ensure it is compatible with the items. These checks could be performed in a different order. To verify the available space, a virtual arrangement of all items, conforming to steps 112 to 116, is attempted within the instance. If all items can be arranged within the instance, it is identified. Once an instance is identified, step 119 is implemented, which is the creation of a package containing the items arranged within that instance.

[0106] In our example, no instance is identified at step 118. The articles will therefore necessarily be distributed across several instances, in accordance with the following steps.

[0107] Step 120 involves obtaining the number of articles in the group, sharing the same model, to be arranged in the already open instances. In our example, this involves the five identical articles 9 from group 14.

[0108] Step 121 involves determining the maximum number of instances allowed to contain items of the same type along with items of other types. For example, an order picker or other user may have previously indicated that they do not want their identical items distributed across more than six mixed-item packages. The maximum number of instances may vary.

[0109] Step 122 is, for each of the already opened packaging instances, the determination of the remaining capacity of the instance, this remaining capacity corresponding to the maximum weight supported remaining by the packaging.

[0110] Step 123 is, for each of these instances, the determination of the dividend resulting from the division of the remaining capacity by the weight of an item to be arranged.

[0111] Step 124 is, for each of these instances, the determination of the modulo resulting from this division.

[0112] Steps 125 to 131 are performed for a value k starting at 2 and, at most, ending at the maximum number of allowed instances obtained in step 119.

[0113] Step 125 is the formation of all combinations, called k-tuples, of already open packaging instances, each k-tuple comprising k instances.

[0114] Step 126 is, for each k-tuple, the determination of the sum of the dividends of the instances of the k-tuple.

[0115] Step 127 is the determination of the sum of the modulos of the instances of the k-tuple.

[0116] Step 128 is the selection of k-tuples whose sum of dividends is greater than or equal to the number of items to be arranged.

[0117] Step 129 is the sorting of the selected k-tuples by increasing sum of the modulos.

[0118] Step 130 is, for each k-tuple in sort order, the attempt to virtually arrange the items in the instances of the k-tuple.

[0119] At step 131, when all the items can be arranged virtually in these instances, theoretical packages are formed comprising the items respectively arranged in these instances of this k-tuple.

[0120] Conversely, when at least one of the items cannot be arranged in one of the k-tuples, step 132 is the incrementation of k. We then repeat steps 125 to 131 with the new value of k.

[0121] Alternatively, before the virtual arrangement attempt 130, a verification step requires that the group's items be arranged in k-tuple instances opened only consecutively. Thus, identical items cannot be distributed across non-consecutive instances.

[0122] In our example, the result of these steps is illustrated at the, with the mixing, in the two open instances, of the identical articles 8 and 9.

[0123] In an unillustrated embodiment, when the open instances cannot contain all the identical items, a step is performed to open a new identical instance of the virtual package and virtually arrange, item by item, as many items of the same model as possible within it. In other words, when, for k reaching the maximum number of instances, none of the k-tuples can contain all the items of the same model, a new identical instance of the virtual package is opened, and the virtual arrangement steps are performed for as many items of the same model as possible. A new theoretical package is thus formed, containing the maximum number of items of the same model possible.For the remaining article(s) of the same model that cannot be arranged in this new instance due to insufficient capacity or available space, the previous steps 118 and 119 of identifying an already open instance to contain these remaining articles are repeated. If no instance is identified, steps 120 through 132 of distributing these remaining articles across the k-tuples of already open instances are performed. If necessary, a new instance is opened. These steps are repeated until all articles of the same model are arranged in instances.

[0124] In step 133, for items from a group that does not contain identical items (in our example, group 15) and that are compatible, from the point of view of all other possible grouping criteria, with the already open instances, the process is performed, in order of opening, through the already open instances. When an open instance has sufficient capacity and available space, the virtual arrangement of the item is performed within the instance according to the virtual arrangement process described previously. Thus, for our group 15, which has non-identical items and therefore does not need to be delivered grouped, the items are placed, one by one, as efficiently as possible in the already open instances according to steps 112 to 116, in order to minimize the empty spaces in these instances. In our example, the result is illustrated in Figure 1.

[0125] Following all these steps, all items from all groups have been arranged into instances. Thus, step 134 is the formation of theoretical packages, each theoretical package containing the items arranged in the same instance. In our example, the input list 7 results in the creation of two theoretical packages, each containing all the items from list 7 arranged in instances 31 and 32, respectively.

[0126] To create logistics load units, a series of steps are then performed to assign packaging to each theoretical package. As mentioned above, the packaging may differ from the dimensions of the previously selected instance. These assignment steps are described in documents FR3141547A1, FR3141453A1, FR3141452A1, and FR3141546A1. Only the main steps are shown below.

[0127] Thus, step 135 is the selection of available packaging. This packaging corresponds, for example, to actual packaging available in stock or to references that the logistics provider can have available.

[0128] Step 136 is, for each selected package, the determination of a prioritization value for the package, the value depending on at least the volume of the package so that a package less voluminous than another package, is, all other things being equal, given priority over that other package.

[0129] Step 137 is sorting the packages according to their respective prioritization values ​​in order to form a sorted list of packages.

[0130] Step 138 is choosing one of the packages from the list based on the sorting.

[0131] Step 139 is the verification of the suitability between the chosen packaging and a theoretical package to be packed, the verification relating at least to the weight of the theoretical package in relation to a maximum weight supported by the packaging.

[0132] For each package deemed suitable, a virtual arrangement step 138 is implemented again, according to the arrangement process described above, of the items from the theoretical package into the packaging. Thus, although an arrangement plan for the items in the instance is already available, the same steps are repeated to determine the arrangement plan for the same items, but this time in a chosen package that may be different from the instance.

[0133] In step 141, for each suitable package, the void ratio generated by the layout plan is then determined.

[0134] In step 142, the chosen packaging is the one for which the determined void ratio is the lowest.

[0135] Following these steps, a physical package is assigned to the theoretically formed package, thus producing a logistics load unit. This unit includes the virtually arranged items and the assigned packaging. The order picker can then prepare the actual logistics load unit, which includes the actual items corresponding to the virtually arranged items, as well as the actual packaging. In our example shown, the logistics load units are packaged with the respective packagings 33 and 34.

[0136] The result of these steps is the allocation of packages 33 and 34 but also the arrangement plans of the items in each of these packages, that is to say the coordinates of the items.

[0137] It should be noted that, during the packaging allocation process, the other criteria described in documents FR3141547A1, FR3141453A1, FR3141452A1, and FR3141546A1 may be taken into account for selecting and then sorting packaging. These include dimensions, to select only packaging suitable for the items, as well as criteria related to durability, price, and greenhouse gas emissions. Suitability verification is also based on the criteria described, such as fragility, price, and weight.

[0138] The invention is not limited to the embodiments shown and other embodiments will be obvious to a person skilled in the art. List of references

[0139] 1: Computer resources 2: Data processing device 3: Data storage medium 4: Computer program 5: Database 6: Means of interaction 7: List of items 8: Identical items 9: Identical items 10: Single item 11: Single item 12: Single item 13: Group of identical items 14: Group of identical items 15: Group of unique items 20: List of virtual packages 21: Package 22: Package 23: Package 24: Package 31: Package instance 32: Package instance 33: Affected package 34: Affected package 100: Parcel preparation process

Claims

A method (100) for preparing multi-packages, implemented by computer (1) and comprising the following steps: - obtaining (102) at least one group (13, 14, 15) of items (8, 9, 10, 11, 12) to be shipped; - obtaining (105) at least one list (20) of virtual packages (21, 22, 23, 24); - selecting (106, 107, 108, 109), from the list (20) of virtual packages, a virtual package based on at least one dimension of at least one item from each group of items to be shipped; - opening (110) at least one instance (31, 32) of the virtual package; - for each item in group (13, 14, 15) or in at least one of the groups, item (8, 9, 10, 11, 12) per item: arrangement (111, 112, 113, 114, 115) of the item in the or one of the open instances (31, 32) according to the remaining capacity of the instance, this remaining capacity corresponding to the maximum weight supported remaining by the packaging, and according to available spaces remaining in the instance;update (116) of the remaining capacity of the instance (31, 32) in which the item (8, 9, 10, 11, 12) is arranged, and update of the remaining available spaces in the instance;- formation (132) of theoretical packages, each theoretical package comprising the items (8, 9, 10, 11, 12) arranged in the same instance (31, 32).; Method (100) according to the preceding claim, comprising, when the open instance (31) or none of the open instances has the remaining capacity and / or the remaining available spaces sufficient to accommodate the article (8, 9, 10, 11, 12), the opening of a new identical instance (32) of the virtual packaging (21) to arrange the article therein. A method (100) according to any one of the preceding claims, comprising, for selecting the virtual packaging (21) according to the group or groups of articles (13, 14, 15) to be shipped, the following steps: - for each packaging (21, 22, 23, 24) in the list (20) of virtual packagings, obtaining (106) the dimensions of the packaging and the capacity of the packaging, this capacity corresponding to the maximum weight supported by the packaging; - choosing (107) the packagings from the list (20) having a capacity greater than the heaviest article in the group or groups of articles and less than a predetermined capacity value, having a volume greater than the largest article in the group or groups of articles and less than the cumulative volume of the articles in the group or groups of articles, and having dimensions greater than the largest dimensions of the articles in the group or groups of articles;- among the selected packages, sorting (108) of the packages (21, 22, 23, 24) according to one or more predetermined criteria.- among the sorted packages, selection (109) of the first package.; Method (100) according to any one of the preceding claims, comprising, for arranging at least one of the articles (8, 9, 10, 11, 12) of the or at least one of the groups (13, 14, 15), traversing (131), in order of opening, the instances (31, 32) already open and, when an open instance has sufficient capacity and remaining available spaces, the virtual arrangement (111, 112, 113, 114, 115) of the article in the instance. A method (100) according to any one of the preceding claims, comprising the following preliminary steps: - obtaining (101) a list of articles to be shipped; - dividing (102) the articles into several groups (13, 14, 15) of articles, each or at least some of the groups (13, 14) comprising only articles (8, 9) having in common at least one predetermined characteristic. Method (100) according to at least the preceding claim, wherein the or at least one of the common features of articles (8, 9) of at least one of the same group (13, 14) is a common article model. Method (100) according to at least the preceding claim, comprising, for virtually arranging the articles (8, 9) of a group (13, 14) having the same article model in one or more open instances virtually already comprising articles of one or more other models, the following steps:- obtaining (120) the number of articles of the group, having the same model, to be arranged in the already open instances;- obtaining (121) a maximum number of instances allowed to contain these articles having the same article model with articles of other models;- for each of the already open packaging instances, determining (122) the remaining capacity of the instance, this remaining capacity corresponding to the maximum weight supported remaining by the packaging, determining (123) the dividend resulting from the division of the remaining capacity by the weight of an article to be arranged, and determining (124) the modulo resulting from this division;- for a value k starting at 2 and, at most, ending at the maximum number of allowed instances: formation (125) of all combinations, called k-tuples, of already open packaging instances, each k-tuple comprising k instances; for each k-tuple, determination (126) of the sum of the dividends of the instances of the k-tuple, and determination (127) of the sum of the modulos of the instances of the k-tuple; selection (128) of the k-tuples whose sum of dividends is greater than or equal to the number of items to be arranged; sorting (129) of the selected k-tuples by increasing sum of the modulos;for each k-tuple in sort order, attempt (130) to virtually arrange the items in the instances of the k-tuple, and, when all the items can be virtually arranged in these instances, formation (131) of theoretical packages comprising the items respectively arranged in these instances of this k-tuple, or, when at least one of the items cannot be arranged in one of the k-tuples, increment (132) of k and reiteration of the previous steps with the new value of k.; Method (100) according to the preceding claim, comprising, prior to the virtual arrangement attempt (130), a verification step requiring that the group items be arranged in k-tuple instances opened only consecutively. A method (100) according to any one of the preceding claims, wherein, in order to achieve a virtual arrangement of an article (8, 9, 10, 11, 12) in an instance (31, 32), the articles and the instance having respective virtual rectangular prism shapes, the following steps are carried out: determination (112) of the dimensions and coordinates of several remaining empty spaces of the instance, distinct from one another and having respective virtual rectangular prism shapes; determination (113) of possible positionings of the article in each empty space; determination (114) of respective distances between a predetermined vertex of the instance and the respective vertices of the article in each possible positioning; based on the distances determined, selection (115) of one of the possible positionings as the virtual arrangement of the article in the instance. Method (100) according to any one of the preceding claims, comprising a sorting step (104) of the articles (8, 9, 10, 11, 12) of one or each group according to at least one predetermined criterion chosen from the volume of each article, the weight of each article, and the density of each article, so as to form an ordered sequence of the articles of the group to be arranged virtually in the instance or instances. Method (100) according to any one of the preceding claims, comprising a sorting step (103) of the groups (13, 14, 15) according to at least one predetermined criterion chosen from the cumulative volume of the items in each group, the cumulative weight of the items in each, and the cumulative density of the items in each group, so as to form an ordered sequence of the groups whose items are arranged virtually in the instances. A method (100) according to at least claim 5, wherein one or at least one of the characteristics common to the articles (8, 9) of a group is chosen from the following: - at least certain dimensions of the articles; - a level of fragility of the articles; - a need for a specific type of packaging for the articles, for example single-wall corrugated packaging, double-wall corrugated packaging, or a plastic bag; - the orientation of the articles; - the cost of the articles; - a level of weight of the articles; - the fact that the articles are pre-packaged; - a maximum packaging rate not to be exceeded for the articles; and - a choice of a shipping preparation area for the articles. A method (100) according to any one of the preceding claims, comprising, subsequently, for each theoretical package formed, a step of assigning a package made in the following manner: - selection (135) of available packages; - for each package selected, determination (136) of a prioritization value of the package, the value depending on at least the volume of the package so that a package less voluminous than another package, is, all other things being equal, prioritized over that other package; - sorting (137) of the packages according to their respective prioritization values ​​so as to form a sorted list of packages; - choice (138) of one of the packages from the list according to the sorting; - verification (139) of the suitability between the chosen package and a theoretical package to be packed, the verification relating at least to the weight of the theoretical package in relation to a maximum weight supported by the package. Computer program (4) comprising instructions which, when the program is executed by a computer, cause the computer to carry out the steps of the process (100) according to any one of the preceding claims. Computer-readable recording medium (3) comprising instructions which, when executed by a computer, cause the computer to carry out the steps of the process (100) according to any one of claims 1 to 13. Data processing device (2) comprising a processor configured to implement the steps of the process (100) according to any one of claims 1 to 13.

Images