EP3702050A1 - Sorting of objects by parallel sorters with separate crossed passes - Google Patents

Abstract

Method of sorting objects to be prepared on distribution rounds using at least a first and a second sorters (T1, T2), in which said objects are distributed (St10) between said sorters, and (St20) is carried out a first sorting pass (P1) of the objects in parallel on the first sorter and the second sorter, we recover (St30) the first groups (G1) of objects at the first sorting outlets (S1) of the first and of the second sorters , which is brought to the input (E) of the first sorter (T1), we recover (St30) of the second groups (G2) of objects at the second sorting outlets (S2) of the first sorter and of the second sorter, which one brings in input (E) of the second sorter (T2), then one carries out (St40) successive sorting passes (P2, P3) in parallel on the first sorter and the second sorter until completion of the preparation of the objects on tour.

Description

Technical area

The invention relates to the field of methods for sorting objects to be prepared on distribution rounds using at least two sorters.

Prior art

In order to receive postal items from various sources of supply and distribute them to a large number of recipients, postal organizations first collect the items received and then route them to sorting centers according to their geographic areas. destination, where they are arranged in distribution sequences.

Within these sorting centers, the objects to be distributed undergo a phase during which the objects are separated into groups, then a so-called “sequencing” phase in which each group is treated individually on a dedicated sorter to arrange the objects into groups. sequence, typically so as to prepare distribution tours such as letter carrier tours.

In practice, a sorter performs several successive sorting passes on a given group of objects, each group being sorted a first time, recovered from the sorter exits, then brought back to the input of this same sorter and re-sorted by the latter, repeated operation. as many times as necessary to complete the “sequencing” operation.

In the current context of a gradual decrease in mail items by post, the main factor limiting sorting systems is no longer the quantity of items to be processed but the number of distribution points to which these items can be addressed.

The number of distribution points that can be supported by a sorter, that is, the number of distribution points for which a single sorter can arrange objects in a sequence, is limited by the two factors which are the number of sorter outputs and the number of sorting passes operated by this sorter on a given batch of objects.

It is known that the maximum number D _max of distribution points that can be supported by a sorter for preparation in sequence is equal to D _max = S ^P where S is the number of outlets of the sorter and P the number of sorting passes. .

We can refer to the patent US 7,170,024 which describes a method of sorting items in sequence, including sorters operating in parallel.

There are two conventional solutions for increasing the number of distribution points that can be supported by a sorter: increasing the number S of outlets or increasing the number P of sorting passes.

The first solution increases the number of outlets, but also the footprint of the sorting facilities and their cost.

The second solution considerably increases the duration of the consignment sequencing operations, thus reducing operational throughput and increasing costs.

There is also the solution of implementing virtual systems on a single sorter, but this solution is also not satisfactory because it is limited by the operational throughput of the sorter and / or requires a particular design thereof to allow an increase. throughput, the single sorter also becoming a critical element in sorting operations, any defect of which implies a loss of throughput for all of these operations.

Disclosure of the invention

The object of the invention is to provide a method for sorting objects to be prepared on tours, making it possible to substantially increase the number of distribution points supported without this being done at the cost of an increase in the number of outputs or the length of preparation for the tours.

• répartir lesdits objets entre lesdits premier et deuxième trieurs ;
• réaliser une première passe de tri des objets répartis sur le premier trieur et le deuxième trieur, cette première passe de tri se déroulant simultanément sur le premier trieur et le deuxième trieur ;
• récupérer des premiers groupes d'objets à des premières sorties de tri du premier trieur et à des premières sorties de tri du deuxième trieur, et amener ces premiers groupes d'objets en entrée du premier trieur ; et
• récupérer des deuxièmes groupes d'objets à des deuxièmes sorties de tri du premier trieur et à des deuxièmes sorties de tri du deuxième trieur, et amener ces deuxièmes groupes d'objets en entrée du deuxième trieur; puis
• réaliser des passes de tri successives en parallèle sur le premier trieur et le deuxième trieur sur les groupes d'objets amenés en entrée de ces trieurs, chaque lot d'objets formé à une sortie d'un trieur à l'issue d'une de ces passes de tri étant récupéré et renvoyé en entrée du même trieur pour y être trié à nouveau, jusqu'à achèvement de la préparation en des objets en tournées.

To this end, the invention relates to a method for sorting objects to be prepared on distribution rounds in several sorting passes using at least a first sorter and a second sorter each comprising sorting outlets, comprising the steps of:

• distributing said objects between said first and second sorters;
• carrying out a first sorting pass for the objects distributed over the first sorter and the second sorter, this first sorting pass taking place simultaneously on the first sorter and the second sorter;
• recovering first groups of objects from first sorting outlets of the first sorter and from first sorting outlets of the second sorter, and bringing these first groups of objects to the input of the first sorter; and
• recovering second groups of objects from second sorting outlets of the first sorter and from second sorting outlets of the second sorter, and bringing these second groups of objects to the input of the second sorter; then
• carry out successive sorting passes in parallel on the first sorter and the second sorter on the groups of objects brought to the inlet of these sorters, each batch of objects formed at an outlet of a sorter at the end of one of these sorting passes being recovered and sent back to the input of the same sorter to be sorted there again, until the completion of the preparation in objects on tours.

The sorting method according to the invention is thus based on the crossing of the flow of objects between two sorters at the end of a first pass, crossing followed by successive sorting passes according to a conventional method during which a set of objects is processed in several passes by the same sorter.

A first advantage of such a crossing is to increase the number of distribution points processed by two sorters, with as reference a single sorter comprising a number of outputs equal to the sum of the numbers of outputs of the two sorters, and this without increasing substantial footprint.

A second advantage is, that in addition to an increase in the number of distribution points supported, the processing time required is reduced compared to that of the single sorter, as detailed below.

• on peut préparer lesdites tournées de distribution au moyen de deux trieurs et trois passes de tri ;
• une allocation des sorties à des destinations peut suivre un procédé d'allocation dynamique ; et
• les objets peuvent être des objets postaux.

The sorting process according to the invention can have the following features:

• said distribution rounds can be prepared by means of two sorters and three sorting passes;
• allocation of outputs to destinations can follow a dynamic allocation process; and
• the objects can be postal objects.

Brief description of the drawings

• la figure 1 illustre les étapes d'un procédé de tri selon l'invention ;
• la figure 2 est un diagramme représentant les étapes de la figure 1 ; et
• la figure 3 représente une comparaison entre le procédé selon l'invention et un procédé classique.

The present invention will be better understood and other advantages will appear on reading the detailed description of the embodiment taken by way of non-limiting example and illustrated by the appended drawings, in which:

• the figure 1 illustrates the steps of a sorting process according to the invention;
• the figure 2 is a diagram representing the stages of figure 1 ; and
• the figure 3 represents a comparison between the method according to the invention and a conventional method.

Description of an embodiment of the method according to the invention

The figure 1 illustrates an embodiment of the method according to the invention, in the particular case where a first sorter T1 and a second sorter T2 are used in parallel over three sorting passes P1, P2 and P3 to prepare a batch L d 'objects in distribution rounds, the first sorter and the second sorter being identical and each comprising a set U1 of sorting outlets S1 and a set U2 of sorting outlets S2.

For purposes of illustration, we will consider that each of the two sorters T1 and T2 comprises an input E and a number of outputs equal to 100, the sorting outputs being denoted S00 to S99 and shared between the first outputs S1 sorting, denoted S00 to S49 here, and second sorting outputs S2, denoted S50 to S99 here.

The sorters are controlled by a C / C control unit comprising a computer memory Mem storing sorting plans which each establish correspondences between on the one hand the destination addresses (distribution points) of the objects to be prepared in distribution and these distribution routes and on the other hand between the distribution routes and the outlets of the sorters.

The control unit controls the sorters so that the sorting conforms to the sorting plans stored in the memory.

The sorters can be equipped with sensors intended to identify and track the objects to be sorted, and connected to the control system, according to conventional methods.

The succession of process steps is shown schematically on figure 2 , with a first sorting pass carried out by two sorters after distributing objects to be sorted into two batches each intended for one of the two sorters, then selectively crossing the objects between the two sorters and sorting them by successive sorting passes.

At a first distribution step St10, the objects forming the batch L of objects to be prepared in distribution rounds are distributed over the two sorters T1 and T2, manually or automatically, for example by bulk dumping then ginning on conveyors constituting a SysE input system according to conventional methods.

The objects can be parcels to be delivered or postal objects such as postal envelopes to be distributed to postal addresses, in which case an unstacker and / or a first sorting machine can be used as the SysE entry system.

Unless there is a priori information on the distribution of the objects between the distribution points, it is preferable to distribute these objects uniformly between the two sorters, so that these sorters have substantially the same workload.

At a step St20 of the first sorting pass P1 of the method, the objects distributed between the sorters undergo a first sorting pass in parallel by means of the first sorter and the second sorter, this first sorting pass taking place simultaneously on the first sorter and the second sorter for all distribution points.

For this first sorting pass, each distribution point is assigned exclusively either to one of the first outputs S1 or to one of the second outputs S2, both for the first sorter Tl and for the second sorter T2, according to a first sorting plan.

Thus, given distribution points are assigned to the output Snn of the first sorter T1 and to the output Snn of the second sorter T2, nn being an integer between 00 and 99 inclusive, so that objects intended for these distribution points will be sorted and will be grouped together at these two Snn outlets at the end of the first sorting pass.

More generally, in this mode of implementation, the groups of objects formed at the outputs of the first sorter are formed according to the same sorting criteria as the groups formed at the corresponding outputs of the second sorter, for example example by assigning the same distribution points to the same outlets of the two sorters.

This implementation mode can also be applied in the case where the outputs are logical destinations which then have to be allocated dynamically according to a conventional dynamic allocation principle.

At a crossing step St30 following the first sorting pass, the first groups G1 of objects are recovered at the first sorting outlets S1 of the first sorter T1 and at the first sorting outlets S1 of the second sorter T2, the outlets S00 to S49 here , and these first groups G1 of objects are brought at the input of the first sorter T1, and, in a similar manner, second groups G2 of objects are recovered at the second sorting outputs S2 of the first sorter T1 and at the second outputs S2 of the second sorter T2, the outputs S50 to S99 here, and these second groups G2 of objects are brought to the input of the second sorter T2.

In this way, the objects intended for the same distribution round which were separated between the two sorters T1 and T2 can be gathered on the same sorter, either the sorter T1 or the sorter T2 in our example, according to a second sorting plan which assigns distribution points exclusively to one or the other sorters for sorting passes following the first pass.

The distribution routes and the second sorting plan are preferably calculated so that, at the end of sorting, the objects constituting a given distribution route are all found at the outlet of only one of the two sorters, and physically separated from the objects of the other routes, for example by being placed in bins each dedicated to only one of the routes.

In fact, the first sorting pass results in the formation of 100 groups by the two sorters considered globally for a cumulative number of outlets of 200, 50 of these groups, recovered at the outlets S00 to S49 of the sorters, being brought to the input of the first sorter and the 50 other groups, recovered at the outputs S00 to S49 of the sorters, being brought to the input of the second sorter for the subsequent sorting passes P2 and P3, as illustrated by figure 1 .

To perform the crossing step, it is possible to handle the groups of objects at the outputs of the sorters manually, by means of conveyors and / or fully or partially automated mechanisms, or even by means of robot-shuttles, according to methods known to those skilled in the art.

At a step St40 of successive sorting passes following the first sorting pass P1 and the crossing, successive sorting passes are carried out in parallel on the first sorter and the second sorter on the groups of objects brought to the input of these sorters, each group of objects formed at an outlet from a sorter at the end of one of these sorting passes being recovered and returned to the inlet of the same sorter to be sorted again there, until the completion of the arrangement of the objects in sequence, according to the second sorting plan mentioned above.

Thanks to the first sorting pass and to the crossing step taking into account the second sorting plan, each sorter only has half of the distribution points and half of the objects to be sorted, against the totality of these points in the first pass, which doubles the separation power of each of the subsequent sorting passes with respect to the first pass and therefore increases the efficiency of each sorting pass accordingly.

In this mode of implementation of the method, two successive sorting passes P2 and P3 are carried out, a number generally sufficient to meet a wide range of practical needs in view of the number of outlets that conventional sorters have.

Indeed, the application of the formula D _max = S ^P seen above to the three sorting passes of this embodiment indicates that the maximum number of distribution points that can be processed by the two sorters of 100 outputs in parallel in 3 passes is 1,000,000.

It is important to note that this high number of distribution points that can be supported is obtained without significantly increasing the footprint, the cost of sorting facilities, or the preparation time of distribution routes compared to conventional solutions. , as explained in the detailed comparison below.

During a step St50 following the successive sorting passes, the groups of objects G _dist formed at the outputs of the sorters are recovered, optionally packaged, then dispatched to their destinations according to the distribution round to which they belong, according to methods conventional.

Comparative example

In order to evaluate the advantages of the method according to the invention, let us consider a comparison between the two sorters T1 and T2 used over three passes including the crossing step according to the embodiment described above and a single sorter T comprising an entry and as many outputs as the two sorters T1 and T2 considered globally, i.e. a number of outputs Ns when T1 and T2 each has N _s / 2 outputs, and on the assumption that the Sorters have, apart from their number of sorting outlets, identical technical characteristics.

soit $${\mathrm{D}}_{\max }\left(\mathrm{T}\mathrm{1},\mathrm{T}\mathrm{2},\mathrm{3}\mathrm{passes}\right)=\left({\mathrm{N}}_{\mathrm{S}}/\mathrm{8}\right)\times {{\mathrm{N}}_{\mathrm{S}}}^{\mathrm{2}}$$

à comparer au nombre de points de distribution D_max(T, 2 passes) pouvant être traités en 2 passes par le trieur unique T avec $${\mathrm{D}}_{\max }\left(\mathrm{T},\mathrm{2}\mathrm{passes}\right)={{\mathrm{N}}_{\mathrm{S}}}^{\mathrm{2}}$$

où nous constatons que $${\mathrm{D}}_{\max }\left(\mathrm{T}\mathrm{1},\mathrm{T}\mathrm{2},\mathrm{3}\mathrm{passes}\right)=\left({\mathrm{N}}_{\mathrm{S}}/\mathrm{8}\right)\times {\mathrm{D}}_{\max }\left(\mathrm{T},\mathrm{2}\mathrm{passes}\right)\mathrm{.}$$

The maximum number of distribution points that can be processed by the two sorters T1 and T2 according to the embodiment of the invention detailed above is then $${\mathrm{<figure-callout\; id="1"\; label="D\; max\; T"\; filenames="imgf0001.png"\; state="\{\{state\}\}">D\; </figure-callout>}}_{\max }\left(\mathrm{T}\right)\left(\mathrm{}\mathrm{1},\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="imgf0001.png"\; state="\{\{state\}\}">T</figure-callout>}\mathrm{2},\mathrm{3}\mathrm{passes}\right)=\mathrm{2}\times \left({\mathrm{NOT}}_{\mathrm{S}}/\mathrm{4}\right)\times \left({\mathrm{NOT}}_{\mathrm{S}}/\mathrm{2}\right)\times \left({\mathrm{NOT}}_{\mathrm{S}}/\mathrm{2}\right)$$

is $${\mathrm{<figure-callout\; id="1"\; label="D\; max\; T"\; filenames="imgf0001.png"\; state="\{\{state\}\}">D\; </figure-callout>}}_{\max }\left(\mathrm{T}\right)\left(\mathrm{}\mathrm{1},\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="imgf0001.png"\; state="\{\{state\}\}">T</figure-callout>}\mathrm{2},\mathrm{3}\mathrm{passes}\right)=\left({\mathrm{NOT}}_{\mathrm{S}}/\mathrm{8}\right)\times {{\mathrm{<figure-callout\; id="2"\; label="NOT\; S"\; filenames="imgf0001.png"\; state="\{\{state\}\}">NOT\; </figure-callout>}}_{\mathrm{S}}}^{\mathrm{2}}$$

to be compared with the number of distribution points D _max (T, 2 passes) that can be treated in 2 passes by the single sorter T with $${\mathrm{D}}_{\max }\left(\mathrm{T},\mathrm{2}\mathrm{passes}\right)={{\mathrm{<figure-callout\; id="2"\; label="NOT\; S"\; filenames="imgf0001.png"\; state="\{\{state\}\}">NOT\; </figure-callout>}}_{\mathrm{S}}}^{\mathrm{2}}$$

where we find that $${\mathrm{<figure-callout\; id="1"\; label="D\; max\; T"\; filenames="imgf0001.png"\; state="\{\{state\}\}">D\; </figure-callout>}}_{\max }\left(\mathrm{T}\right)\left(\mathrm{}\mathrm{1},\mathrm{<figure-callout\; id="2"\; label="T"\; filenames="imgf0001.png"\; state="\{\{state\}\}">T</figure-callout>}\mathrm{2},\mathrm{3}\mathrm{passes}\right)=\left({\mathrm{NOT}}_{\mathrm{S}}/\mathrm{8}\right)\times {\mathrm{D}}_{\max }\left(\mathrm{T},\mathrm{2}\mathrm{passes}\right)\mathrm{.}$$

Assuming that Ns is 200 as in the example of figure 1 , which is an order of magnitude commonly used in the field of postal sorting, then we find that the two sorters T1 and T2 can process 1,000,000 distribution points in 3 passes, 25 times more than the single T sorter in 2 passes.

• une durée t_nominal = N_O/D pour une passe de tri ;
• une durée t_parallèle = N_O/ (2xD) pour effectuer une passe de tri sur les deux trieurs T1 et T2 en parallèle ; et
• une durée t_demi-vol = (N_O/2)/D pour effectuer une passe de tri d'un demi-volume de N_O/2 objets.

The sorters T, T1 and T2 each considered have an operational throughput of D objects per hour, for a sorting pass of a volume of N _O objects, it is therefore necessary:

• a _nominal duration t = N _O / D for a sorting pass;
• a _parallel duration t = N _O / (2xD) to perform a sorting pass on the two sorters T1 and T2 in parallel; and
• a _half-flight duration t = (N _O / 2) / D to perform a sorting pass of half a volume of N _O / 2 objects.

à comparer avec la durée totale t_T de deux passes de tri sur le trieur unique T qui vaut 2xN_O/D, comme illustré par la figure 3.The total duration t _total of the three sorting passes on the two sorters T1 and T2 is then $${\mathrm{<figure-callout\; id="1"\; label="t\; T"\; filenames="imgf0001.png"\; state="\{\{state\}\}">t\; </figure-callout>}}_{\mathrm{T}}={\mathrm{t}}_{\mathrm{parallel}}+\mathrm{2}{\mathrm{xt}}_{\mathrm{half}-\mathrm{flight}}=\mathrm{1},\mathrm{5}{\mathrm{xN}}_{\mathrm{O}}/\mathrm{D}$$

to be compared with the total duration t _T of two sorting passes on the single sorter T which is equal to 2xN _O / D, as illustrated by figure 3 .

We note that the use of two sorters T1 and T2 in parallel in the method according to the invention makes it possible to process considerably more output points than a single sorter while completing the “sequencing” operation more quickly.

Strictly speaking, the additional time of the crossing step should be taken into account, but since most of the object processing time is taken up by the sorting passes, taking this additional time into account would not qualitatively change the conclusion of this comparison.

The first sorting pass and the crossing step make it possible to pre-sort all the objects on all the distribution points and to distribute pre-sorted groups of these objects so as to take them into account. load in parallel on sorters dedicated to subsets of distribution points, which increases the overall processing speed while increasing the number of distribution points that can be supported.

Of course, this conclusion is not limited to the example taken above of two sorters placed in parallel over three sorting passes.

The two sorters T1 and T2 are of the same design as the single sorter T and therefore do not overall occupy a floor area which is not significantly larger than the latter.

In addition, the method according to the invention does not require any particular adaptation of the pre-existing sorters in order to be implemented.

The above example is considered in the particular case of two identical sorters used for three sorting passes, for the purposes of explanation, but those skilled in the art will understand that the invention can be generalized to number sorters and in different characteristics, and for a different number of sorting passes, the main thing being to sort a set of objects distributed over several sorters working in parallel in successive passes, each sorter taking charge of all the points distribution of objects during the first sorting pass and only a subset of these distribution points for subsequent sorting passes, each sorter receiving during a crossing step immediately following the first sorting pass the objects sorted by the other sorters and intended for distribution points assigned, exclusively, to the outlets of the sorter considered for subsequent passes.

It is up to the practitioner to determine the number of sorters and the number of sorting passes to use, according to his particular needs and constraints.

Claims (4)

Method for sorting objects to be prepared on distribution rounds in several sorting passes, characterized in that it implements at least a first sorter (T1) and a second sorter (T2) each comprising sorting outlets (S1 , S2), and in that it comprises the steps of: - Distributing (St10) said objects between said first and second sorters; - Perform (St20) a first sorting pass (P1) of the objects distributed over the first sorter and the second sorter, this first sorting pass taking place simultaneously on the first sorter and the second sorter; - recovering (St30) of the first groups (Gl) of objects at the first sorting outlets (S1) of the first sorter and at the first sorting outlets (S1) of the second sorter, and bringing these first groups (G1) to input objects (E) of the first sorter (T1); and - recover (St30) second groups (G2) of objects at second sorting outlets (S2) of the first sorter and at second sorting outlets (S2) of the second sorter, and bringing these second groups (G2) to input objects (E) of the second sorter (T2); then - carry out (St40) successive sorting passes (P2, P3) in parallel on the first sorter and the second sorter on the groups of objects (G1, G2) brought to the input of these sorters, each batch of objects formed at an output from a sorter at the end of one of these sorting passes being recovered and sent back to the input of the same sorter to be sorted there again, until the completion of the preparation in objects on tours. The method of sorting objects according to Claim 1 or Claim 2, characterized in that the preparation of the objects in sequence is carried out by means of two sorters and three sorting passes. The object sorting method according to claim 1, characterized in that an allocation of the outputs to destinations follows a dynamic allocation method. The method of sorting objects according to claim 1, characterized in that the objects are postal objects.

Images