Method and system for progressive picking

Abstract

Storage and picking system (10) for progressively picking articles (22) according to new orders that define ordered articles (22) with respect to a respective ordered article type (46) including an associated quantity (44), the system (10) comprising: a picking warehouse (12) in which a plurality of different article types (46) forming an article assortment (24) for picking are stored, the picking warehouse (12) having a storage capacity that ensures a picking range of many days; a picking station (14) for conventional picking of articles (22) stored in the picking warehouse (12); a buffer warehouse (16) in which selected article types (46) are preferably stored in storage units (80) and which has a storage capacity that ensures a picking range of a few days, wherein the buffer warehouse (16) is filled with the selected article types (46), which essentially originate from the picking warehouse (12), according to a filling order that is cyclically re-determined based on data from old orders that represent already completed picking orders (40) from the past and that defines the selected article types (46) with their associated piece numbers (44), wherein the data of the old orders comprise ordered article types (46) including their associated piece numbers (44) and associated order times; a picking control (30); and at least one workstation (18) and a consolidation buffer device (20), wherein the workstation (18) is connected to the buffer storage (16) in terms of material flow technology and wherein the ordered articles (22) are taken and packed at the workstation (18), preferably from the storage units (80), or are taken and delivered to the consolidation buffer device; wherein the order-picking control (30) is configured to cyclically generate the filling order, the generation comprising the following steps: Selecting (S18) the article types (46) that were ordered frequently and in larger quantities in the past for the buffer stock (16) from the data of the old orders; Determining (S20) article type-specific buffer quantities for each of the selected article types (46); and Generating (S22) the filling order that defines the selected article types (46) with their associated buffer quantities.

Description

The present invention relates to a storage and picking system for the progressive picking of articles according to re-orders, which define ordered articles with respect to a respective ordered article type including an associated piece count, wherein data of old orders are taken into account and evaluated. The invention further relates to a method for progressive picking and to a method for cyclic filling of a buffer store. In the invention, the aim is in particular to arrange the articles in the storage region in a predictive (progressive) manner, so that increases in efficiency are possible.When marketed over the Internet (e-commerce), great fluctuations occur in the article locations (totality of all different article types) of the providers within short time intervals (days, weeks or months). Order structures (number of different article types and associated quantity / quantities per order) change in a short term, frequent and difficult to predict. During the Wei's weigt time, for example, more is ordered in terms of quantity and bandwidth than in summer during the Ferien time. Whereas a new appearance of a best-to-measure cell can initially be settled well, i.e. quickly and in large numbers, the turnover decreases with increasing time. After a certain time, the new appearance is no longer a bestseller. The market is then saturated. However, there may again be a new bestseller for this.Imaging these fluctuations by reorganizing (redistribution) the articles within the warehouse is almost impossible. The e-commerce dealers must design their storage and picking systems to the load peaks (e.g. Weikin's attention) described above, which leads to inefficient operation of the system at high overall investments during the remaining times of low load.Especially in the field of e-commerce, the providers are faced with a very high number of different article types or SKUs (Stock Keeping Units) which are ordered over the year. It can occur that on a single day 20% of the year requirement of a specific article type and the remaining 80% are partially implemented distributed over the remainder of the year.Further, there are large fluctuations from day to day in the ordered article types. Today, for example, the article "A" sells itself particularly well and no longer in the morning. For this, the next day, the article "B" sells particularly well. This means that steep ABC curves are common especially in e-commerce. The respective ABC day curves rise very steeply. It is problematic that the article-type-specific compositions of the 20% portion (A article) and the 80% portion (B and C articles) of the ABC curve change daily. These daily changes do not allow the corresponding ABC distribution to be mapped onto the storage and picking area (for example short paths for A articles and longer paths for B and C articles). The reorganization effort in a static article provision would be too great. However, the providers must be able to react flexibly to these varying requirements.Another problem is the varying ordering frequency seen over a year for a respective article type. This applies all the more if a specific article type is ordered only in small and unpredictable numbers. There may be days, for example, where a particular article type is ordered up to 40 times, but this article type is not ordered at all over 30% of the days. Here too, it is not possible to organize or distribute the article location accordingly in the storage area.Yet another problem is severe changes in load on the system during a day course. For example, there are temporal ordering peaks in the evening just after Feierabend because many potential customers then have the time to surf on the Internet and place orders, whereas there is no time for it during the normal working time. In order to counter this problem, the providers must again design their storage and picking systems to the possible maximum load. The systems must therefore be designed to be correspondingly large with regard to a picking performance in order to be able to compensate for these maximum loads, which are only of very short duration. During the remaining time, the system is only partially utilized, so that the cost benefit aspect is negative.Further, there are great variations in the numbers ordered and in the order structures over the course of one month. As already indicated above, during the Wehnacht store, for example, three times the quantity of articles is converted as at other times. During the Weigt store, the probability is also greater that the customers are simultaneously ordering a plurality of different articles, whereas in the remaining time often only a single article type (single line / one order line) is ordered. In order to compensate for these fluctuations, a high investment is again necessary due to a corresponding design of the system for the maximum load and a design for greatly varying application structures.The document DE 10 2013 103 869 A1 relates, according to its title, to a storage and picking system and a method for rearranging an article location.Document DE 10 2013 101 659 A1 relates, according to its title, to a method and an installation for providing, commissioning and / or packaging articles.It is therefore an object of the present invention to provide a storage and picking system and a method for progressive picking which solves in particular the problems occurring in e-commerce dealers due to varying order intensities and order structures.This object is achieved by a storage and picking system according to claim 1.The system of the invention is essentially characterized by an increase in performance. The increase in performance is achieved by a preliminary commissioning. A small part of the article orthosis, which is usually stored in the long-term store, is transferred into a buffer region arranged separately therefrom, namely preferably with a composition different daily. The initial composition of the articles transferred from the long-term store to the short-term buffer is basically determined from an analysis of old orders that have already been completed without employing the system and concept of the present invention. In addition, it is optionally also possible to take account of new orders which have already been dealt with but have not yet been processed. Due to the pre-commissioning, the travel times in the area of the conventional bearing are shortened by a displacement into the buffer bearing.In addition, the provider / system operators can more flexibly respond to continuously and rapidly changing requests.The buffer store is fitted with comparatively few article types in relatively small numbers. The articles for the buffer store are selected such that the buffer store is emptied by itself as a result of the ongoing consumption (picking), preferably within a very short time (picking range). The picking range of the buffer warehouse is thus very short, whereas the picking range of the picking warehouse is very large. Orders can be serviced from the picking warehouse for relatively long periods of time (days, weeks, months) without refilling or refilling the picking warehouse. The buffer store, on the other hand, should be refilled as often and quickly as possible. Because the buffer store almost completely emptied by itself, it is not necessary to return it to the picking store. This aspect also contributes to the increase in performance. Few articles are moved within the system.In the area of the workstation, batch-oriented work is preferably carried out. In a conventional system, when the conventional batch size (time in which orders for batch processing are collected) is one to two hours, the batch size is substantially increased in the invention.Preferably, the step of selecting the article types comprises: determining the different article types contained in the data of the old orders; for each of the article types contained in the data of the old orders, determining a temporal ordering frequency, preferably by quotienting from a number of ordering time units during which the respective article type was ordered and a number of ordering time units during which orders were possible; for each of the article types contained in the data of the old orders, determining a quantitative ordering frequency, preferably the total number of pieces; ordering the different article types according to temporal and / or quantitative ordering frequencies, preferably taking into account temporal and / or quantitative thresholds; and selecting the temporal and / or quantitative most frequently occurring article types.It is understood that if an article type has been frequently ordered in the recent past, the same article type will also be frequently ordered in the next days. Therefore, the time and quantity order frequencies are determined and then the article types having the highest order frequencies are selected for the buffer store. Threshold values can be taken into account in order to exclude point-like "outliers" during the evaluation.Preferably, the step of cyclically generating the filling order further comprises: determining a distribution of the article-specific buffer piece numbers to the storage units.In addition to the weight and the volume of the articles, the volumes of the storage units and possible (compartment) storage periods of the storage units can also be taken into account. Optionally, limit values for such parameters are also taken into account, in particular during the raising. In this way, it is prevented that, for example, when it is erected on entire storage units or entire compartments, too many articles of a specific article type are stored in the buffer store. If too many articles are stored in the buffer store, the probability increases that the buffer store does not drain automatically.If a specific article type remains in the buffer store beyond the actual picking range, an active re-storage of this article type from the buffer store into the picking store is possibly required. The re-storage requires time. Residual stocks remaining in the buffer store block the space for new article types that would actually have to be stored in the buffer store according to future filling orders. This is not desired. Therefore, the filling of the storage containers is optimized.The distribution of the article-type-specific buffer piece numbers to the storage units is therefore preferably determined taking into account article-type-specific volumes and weights and storage unit-specific dimensions. Optionally, corresponding limit values can in turn be taken into account.In a particular embodiment, the article-type-specific buffer piece numbers for the selected article types are determined by multiplying article-type-specific average placement numbers by the picking range of the buffer store.In particular, the buffer store has a fixed physical size which corresponds to a fixed number of (preferably uniform) storage units which the buffer store can accommodate to the maximum, wherein the fixed number of storage units is preferably determined by a further analysis of a plurality of data sets of old orders.The physical size of the buffer store is usually determined only once, namely during the initial planning of the system. Although the buffer bearing can be made scalable, it is desirable to operate the buffer bearing for as long as possible without structural changes. The operator / supplier should thus initially decide for a specific buffer size. The buffer size in turn should be small compared to the size of the picking warehouse, because the buffer warehouse only has a picking range of a few days. A picking range is understood to mean a time duration within which a number of orders can be supplied with articles from a warehouse without refilling the warehouse. It is recommended to calculate the respective buffer location whose population changes cyclically for many cycles (cycle duration is preferably one day) in order to be able to form an average value which satisfies all requirements. Of course, in the future, peaks may again and again occur which actually require a larger buffer bearing. However, it also occurs repeatedly that the buffer bearing is oversized compared to the current demand. A compromise is found here.In a further special embodiment, the picking controller is further configured to divide the re-orders into buffer-store-free, mixed and picking-store-free re-orders, wherein the buffer-store-free and mixed re-orders are processed at the work station and the picking-store-free re-orders are processed at the picking station. This is again expressed as path time optimization. Articles that are frequently needed are preferably served from the buffer, which holds A articles in particular. The remaining articles are picked from the conventional warehouse.It is furthermore advantageous if the buffer store has a storage unit storage device, in particular a conveyor carousel, on which the storage units, which are preferably stacked one above the other, circulate endlessly, or a horizontally circulating carousel rack.In addition, a packing station can be provided.Preferably, a fill order generation cycle is one (work) day.It is furthermore advantageous if the consolidation buffer device is a shelf shelf shelf, a continuous shelf, a pick-to-bucket arrangement or an endlessly revolving tray sorter and has a picking guidance system.It is also advantageous if the picking controller is further configured to carry out warehouse management, picking guidance and order management and to cause the buffer warehouse to be filled with the articles of the selected article types.The object is further achieved by a method according to claim 14.The concept of the invention is distinguished in that the orders are divided into three categories. The categories decide where or how a corresponding order is processed.Preferably, the method further comprises the steps of: generating a filling order according to the above-mentioned steps; and transferring articles, which are determined with respect to an article type and an associated quantity by the filling order, from the picking store into the buffer store.In particular, the method comprises the steps of: determining a current fill state of the buffer store before the buffer store is to be filled according to the fill order; comparing the current fill state with a desired fill state corresponding to the fill order; determining a difference between the current and the desired fill state; determining an adjusted fill order corresponding to the difference; transferring the articles determined by the adjusted fill order from the picking store into the buffer store; and filling and refilling the storage units with the articles.It is understood that the features mentioned above and those still to be explained below can be used not only in the respectively specified combination, but also in other combinations or alone, without departing from the scope of the present invention.Exemplary embodiments of the invention are illustrated in the drawings and are explained in more detail in the following description. The following are shown: FIG. 1 is a block diagram of a storage and picking system according to the invention; FIG. 2 shows a block diagram of an exemplary job or a job data set; FIG. 3 is a block diagram of an exemplary picking controller; FIG. 4 is a block diagram illustrating an article flow; FIG. 5 is a perspective view of a buffer store and of two work stations; FIG. 6 shows a perspective illustration of a workstation with a sorter as consolidation buffer device; FIG. 7 shows distribution curves of article type-specific ordering frequencies; FIG. 8 is a table showing the ordering frequencies of FIG. 7 and explaining a filling order for the buffer store; FIG. 9 shows a curve of a number of storage units required for the buffer store according to data from the past for the purpose of determining a physical size of the buffer store; FIG. 10 shows a course of a duration which is required for filling the buffer bearing; FIG. 11 is a flow chart for generating a filling order for the buffer bearing; FIG. 12 shows detailed steps for the diagram of FIG. 11 ; FIG. 13 shows a flow diagram of a picking method; and FIG. 14 shows a comparison of required working times.When vertical and horizontal orientations are mentioned below, it goes without saying that elements and features which are associated with these orientations can be interchanged with one another at any time by a corresponding rotation, and therefore such orientations are not to be understood as restrictive. Furthermore, identical parts and features are provided with identical reference numerals. The disclosures contained in the description can be applied analogously to identical parts and features with the same reference numerals. Position and orientation information (e.g. "top", "bottom", "lateral", "longitudinal", "horizontal", "vertical" and the like) are related to the immediately described figure. When the position or orientation is changed, these details are to be transferred analogously to the new position or orientationAs is customary in (intra) picking, in storage and picking systems (distribution systems, material handling systems, etc.), a longitudinal direction is denoted by "X", a transverse direction by "Z" and a height direction by "Y". The directions X, Y and Z preferably define a Cartesian coordinate system.FIG. 1 shows a block diagram of a storage and picking system 10 according to the invention.The storage and picking system 10 (also referred to briefly below as "system 10") has a picking store 12, at least one picking station 14, a buffer store 16 and at least one workstation 18 with a consolidation buffer or a consolidation buffer device 20. The picking store 12 is a conventional store which is configured to carry out a conventional picking, as will be described in more detail below. The picking store 12 has the function of a long-term store for articles 22 (not shown) and serves as a supply store for the buffer store 16. The picking store 12 can be implemented by way of example by shelf racks, an automated small-part store (AKL), continuous racks, carousel racks or the like. The buffer store 16 has the function of a temporary store for the articles 22, and only a very small portion of the article location 24, preferably the "best" A articles, is temporarily provided in the buffer store 16. The buffer store 16 is preferably dimensioned such that 1-10%, in particular 1.5-6%, of the article types 46 (FIG. 2 ) of the entire article orthosis 24 are also stored in the buffer store 16.An "article" is understood below to mean an inventory or picking unit within the storage and picking system 10. The picking unit, which is also referred to as storage goods, can comprise a storage loading aid and the article 22 itself. The picking unit can, however, also comprise only the article 22 if no storage loading aid is present. As (storage) loading aids, use is made, for example, of pallets, containers, cartons, shelves, (hanging) pockets or the like. An "article" is to be understood in particular as an article. The articles 22 are (smallest) units of the article location 24 that can be distinguished by article type 46 (see FIG. 2 ). A "container" is a general term for a unit that can be handled and can be moved manually or by means of a technical device (e.g. conveyor system, storage and retrieval device, load-receiving means, etc.). The terms "article", "container", "storage goods", "picking unit", "inventory unit" and "piece goods" are used equivalently below.To move the articles 22 within the system 10, various types of conveyors may be used (e.g., roller conveyors, belt conveyors, chain conveyors, overhead conveyors, belt conveyors, belt conveyors, etc.) not shown in FIG. 1.The terms "conveyor" and "conveyor" are used equivalently below. A conveyor system includes essentially all technical and organization devices (e.g., drives, sensors, switches, control elements, etc.) for moving or transporting the articles 22 and for directing material and article streams 26.The article streams 26 are indicated by arrows in FIG. 1. The articles 22 enter the system 10 via a goods entrance (WE) 28. The goods entrance 28 is connected to the picking store 12 in terms of material flow. The articles 22 are framed in the goods entrance 28 and subsequently stored in the picking store 12. If necessary, the articles 22 from the article entrance 28 can also be stored directly in the buffer store 16. However, the articles 22 in the buffer store come substantially from the picking store 12. a small part of the articles 22 is cyclically transferred from the picking store 12 with a constantly changing composition (article type and quantity) and stored in the buffer store 16 according to a filling order. The transferred articles 22 are then present both in the picking store 12 and in the buffer store 16. Normally, there is no return storage of the articles 22 from the buffer store 16 into the picking store 12. Only in exceptional cases are residual articles stored back from the buffer store 16 into the picking store 12, as is indicated by a dashed line arrow.In the picking store 12, the conventional picking takes place, preferably of bundles / or C-articles. A spatial arrangement / distribution of the articles 22 in the picking store 12 takes place, for example, depending on their access frequencies. The access frequency itself is usually categorized. The "access frequency" is generally understood to mean a number of requirements of an article type 46 per unit time. Alternatively, categorization or classification can also be carried out as a function of paragraph sets, envelope frequencies or other criteria. In this context, one also speaks of the ABC distribution already mentioned at the beginning. A sorting or arrangement of the article location 24 according to access frequencies can be expressed by a so-called Lorenz curve. The term "pareto" distribution is also used analogously. It is usual that the spatial arrangement and distribution of the articles 22 of the article location 24 is taken into account already during a planning phase of the picking warehouse 12 and is implemented accordingly later. Therefore, it is extremely difficult to situationally react to when the access frequencies of the articles 22 of the assortment 24 change frequently and / or briefly during system operation, as is the case in particular in the field of e-commerce.The terms "high speed rotators" (or "high speed rotators") and "low speed rotators" (or "low speed rotators") are understood below as meaning A articles or B or C articles of the article orthosis 24. A articles have high paragraph amounts, turnover rates or access frequencies, C articles have low levels. The boundaries between A, B and C articles are determined in each case. It is understood that the low speed rotators can also be extended by the middle rotators ("B-articles").E-commerce dealers often have a very large item location 24 (30,000 to 200,000 different item types 46 are not rare). A probability that an access frequency of one of the article types 46 changes is high because in the e-commerce the access frequencies of the articles change by customer behavior (hypes) or controlled by advertisements or special price actions. A further aspect is the purchase occasions (partly deliberately used) such as e.g. Weikin's law, eastern, mother day, father day, valentin day, ferry onset, school year onset or the like. During an offer phase, the frequency of access to one of the article types 46 which is currently being offered is generally higher than during a phase in which the same article type 46 is not in the special offer. Conventional storage and picking systems are not able to react sufficiently flexibly and quickly to access frequency fluctuations within the articles 22 of the assortment 24. In particular in the case of e-commerce applications, the following difficulties arise:Shipment must take place on the same day as an order is received, wherein a time at which an order is possible last can be very close to the actual delivery.An ABC structure changes frequently and frequently with respect to the article types involved.Generally, very diverse article locations can be present (for example from small-part modeaceses to large pieces of furniture or heavy machines such as washing machines).There are extreme peak loads (e.g. during the pre-sign time, before eastern or the like).Many deliveries are returned (e.g., because a shoe size does not fit). These return goods must be incorporated again into the system in order to be able to be delivered again later in another order.Customer desires are difficult to predict.A fast return-of-investment (ROI) is desired by the plant operator / supplier.The conventional picking of the articles 22 of the picking store 12 of FIG. 1 takes place at one or more picking stations 14. It is understood that the picking station(s) 14 can also be arranged remote from the picking station 12. It can be picked according to different principles.Nowadays, a distinction is roughly made between two different picking principles according to which work can be performed at the picking stations 14. It is picked either according to the "man-to-goods" principle or according to the "goods-to-man" principle. The present invention can be operated according to any of the principles, also in combination, supporting a picking person (not shown) as much as possible. The picking can be carried out manually or automatically.In conventional picking, picking is frequently carried out according to the principle "man-to-goods". For the purpose of picking, the picking person moves in a (decentralised) region of the warehouse 12, wherein provision units (e.g. storage containers, shelves, cartons, pallets, etc.) are stored and provided at access locations at stationary locations within the warehouse 12. In accordance with a picking order 40, the customer-specified articles 22 are assembled according to type 46 and number by collecting them. The picking person removes the desired article or articles 22 and places them in a collecting device (carriage, container, carton, etc.).Alternatively, in the case of the "goods-to-mann" principle, the articles 22 to be picked are transported to the picking person, so that the picking person, which will also be referred to below as a "picking person", has to run as little as possible or not at all for reasons of better ergonomy in order to carry out a picking operation (removal of the articles from a source and delivery to a destination). The articles 22 to be picked are transported within the system 10, in particular from and to the picking stations 14.Alternatively, storage containers 80 may be statically stored on racks and collecting devices for the orders may be dynamically guided past the storage containers in the rack. The collecting devices can be realized by application containers on a conveyor system or by trays of a tray conveyor, which are tilted into the application container at the end of the conveying path.In addition, there are a multiplicity of different picking guidance strategies which are referred to by terms such as, for example, "pick-to-belt", "pick-by-light", "put-to-light".The picking guidance strategy or picking guidance "pick-by-light" offers significant advantages over classic, manual picking techniques. In pick-by-light systems, each access station (source) has a signal lamp with a numerical or even alphanumeric display and with at least one acknowledgement key and possibly input or correction keys. When an order container into which the articles 22 are deposited arrives at a picking position, the signal lamp lights up at that access location (source) from which the articles 22 can be removed. The number to be drawn appears on the display. The removal is then confirmed by the acknowledge key and the inventory change can be fed back to a warehouse management system 62 (FIG. 2 ) in real time. Pick-by-light systems often operate on the principle "man-to-goods". In the case of a put-to-light system, the destination is displayed visually.Furthermore, in the picking warehouse 12, a document-free picking by means of "pick-by-voice" is possible. There, communication takes place between a picking controller 30 (data processing system) and the picking person by means of voice. Instead of printed order picking lists or data radio terminals (i.e. mobile data acquisition units (MDE), the order picking person usually works with a headset (earphone and microphone), which can be connected to a commercially available pocket PC, for example. The orders 40 (FIG. 2 ) are transmitted from the warehouse management system 62 to the picking person by means of radio, usually by means of WLAN / WiFi.Furthermore, a distinction is made during picking between order- and article-related picking on the basis of the selected strategy, wherein the items 22 themselves can be put together either serially, that is to say successively, or in parallel, that is to say simultaneously. During order-related picking, an order 40 is processed as a whole, i.e. all articles 22 of the order 40 are assembled serially. In the case of article-related picking, on the other hand, a plurality of orders 40 are processed in parallel by the picking person, said orders respectively composing or searching only a part of the orders 40, namely the article to be processed.Since the picking person actually no longer has to run in the "goods-to-mann" principle, because the piece goods to be picked are transported directly to him, the customer or picking orders are frequently processed in parallel, which is referred to as "batch picking". In "batch picking", a plurality of customer orders are combined in an article-oriented manner in order to have to remove as few SKUs (Stock Keeping Unit) as possible from one of the stores 12 or 16 and, after picking has taken place, to have to move them back into the store 12 or 16. In the article-oriented analysis of the orders 40, a group of customer orders containing all orders (order lines) pertaining to a particular article type 46 are looked at and then generate article-oriented transport commands for the SKUs. The SCU corresponding to that particular article type 46 is then swapped out and transported to station 14 or 18. At the station 14 or 18, the picking person grasps all articles 22 of the respective article type 40 and delivers them in a predefined number to correspondingly provided destination locations. This process is also referred to as "picking" regardless of whether the articles 22 are provided with or without LHM.Each destination or each destination location is assigned one of the picking orders 40, so that the delivery of the articles 22 takes place in order-oriented fashion. The destination locations are held in stock at the stations 14 or 18 until all article types 46 of the associated picking order 40 have been given to the destination location. In this context, one generally refers to a two-stage picking.The manner in which picking is performed in the system 10 may depend on many factors. One factor that may play a role is an average job structure 84 (FIG. 4 ). It makes a difference whether different articles 22 are to be picked in small quantities or whether the same articles 22 are to be picked in large quantities over and over again.Returning to FIG. 1, when the picking is completed at the workstation 18 and / or the picking station 14, the completely picked articles 22 are transferred into a goods exit (WA) or shipment 32, from where the articles 22 arrive at their travel to the customers. Optionally, one or more packing stations 34 can also be provided, where the completely picked articles 22 are packed and / or repackaged into shipping carriers (not shown). In FIG. 1, an exemplary packing station 34 is shown by a dashed line between a storage and picking area 34 and a packing and shipping area 36.FIG. 2 shows an exemplary picking order 40. The (picking) "order" 40 consists of one or more order positions, which are also referred to as order lines 42. Each of the order lines 42 indicates a respective quantity / quantity 44 of an article type 46 that a customer 48 has ordered. A old order 40 is an order 40 which has already been processed, i.e. finished and finished, in the past. A re-job 40 is a job 40 to be processed in the future.The orders 40 are present as data sets. Each of the data sets may include a header field 50, an (optional) priority field 52, and / or an item field 54. The header 50 may include, among other things, information about the customer 48 who has placed an order, a (customer) address or identification number (customer), and an order / order number. Priority field 50 contains information as to whether it is a normal job or a rapid job. A fast job is a job 40 with high (processing) priority, which is usually treated with priority over normal jobs 40. The article field 54 has the order lines 42. Each of the order lines 42 has at least details about the associated piece count 44 of an ordered article 22 and about its article type 46.A coordination of the processing of the orders 40 is performed by an order handling system or an order management 56 (FIG. 3 ), which is usually integrated into the picking controller 30, which may also have a goods management system 58, for example. The order processing is usually carried out in a computer-assisted manner by means of the order processing system 56.FIG. 3 shows a block diagram of the picking controller 30. the picking controller 30 can also have a (warehouse) place manager 62 and an information display integrated. The picking controller 30 is usually implemented by a data processing system which preferably operates in online operation for delay-free data transmission and data processing. The picking controller 30 can have a central or decentral structure. The picking controller 30 can be of modular construction and implement the following functions: the order manager 56, the implementation of the picking guidance strategies 60, the inventory management system 58, which in turn can comprise the warehouse manager 62, which in turn can contain the material flow 64 and the (warehouse) place manager 66, and / or an interface manager 68. Communication may be via lines 70 or wirelessly (arrow 70) as shown in FIG. 1.The order management 56 of FIG. 3 ensures that (picking) orders 40 arriving from customers 48 are distributed to the stations 14 and / or 18 for completion (processing). Factors such as load, load-rate distribution, (conveyor) path optimizations, and the like play a role at station 14. The stations 14 and 18 can receive complete orders 40, but also only partial orders, assigned for processing. Not all article types 46 of a picking order 40 need, however, be picked at the same station 14 or 18.FIG. 4 is a block diagram illustrating a flow of material in an example system 10, which may be structured like the system 10 of FIG. 1. For the sake of simplifying the illustration, not all elements of the system 10 of FIG. 1 are illustrated in FIG. 4. Thus, for example, the goods intake 28 is missing.In FIG. 4, the material flow or the article flow is indicated by arrows between the blocks involved. The source of all articles 22 within the system 10 represents the picking store 12. Within the corresponding block, an exemplary storage unit in the form of a (single) storage container 80 is shown. It will be appreciated that a plurality of storage containers 80 are deployed in the store 12. A storage unit is a unit that serves to store the articles 22. However, the bearing unit is also a standard arithmetic quantity to express the physical quantity of the bearings 12 and 16. Preferably, the same bearing units are used in the stores 12 and 16.The articles 22 are preferably stored in the storage containers 80 in a manner known per the art. This means that the storage containers 80 each store only articles 22 of a single article type 46. It will be appreciated that other storage loading tools may be used, such as shelves, bags, cartons, pallets, or the like. Shelves, shelf assemblies or other storage devices for providing the storage containers 80 are not shown for the sake of simplified illustration. The picking station 14 is arranged here within the region of the picking store 12. At the picking station 14, picking is carried out manually, for example, according to the man-to-goods principle, in that the picking persons travel or travel through the picking store 12 with picking trucks. The order picking carriages (not shown) are equipped, for example, with order containers (not shown). The picking persons remove the articles 22 from the storage containers 80 and discharge them into the order containers. The picking and delivery, i.e. the picking, takes place according to the picking orders 40, e.g. by means of pick-by-voice.Block 82 serves to clarify an application structure 84. application structure 84 shown in FIG. 4 has, by way of example, 3590 applications (A). These 3590 jobs are composed of 5342 job lines (AZ). The 3590 orders define 5891 Piece (STK) where a (single) piece is to be matched to an inventory unit or article 22. The factors F1 and F2 express relationships between the orders 40, order lines 42 and the numbers 44. The factor F1 is here about 1.49 and specifies an average number of job lines 42 per job 40. The factor F2 is here about 1.1 and specifies an average number of pieces 44 per order line 42. Both factors F1 and F2 are of the order of 1 and are therefore comparatively low. The order of magnitude of 1 indicates that the application structure 84 considered here originates from the region of the E commerce (B2C, business-to-customer). In B2B (business-to-business) applications, the factors F1 and F2 would be significantly larger.Four storage containers 80- 1 to 80- 4 are exemplarily shown in the buffer store 16 of FIG. 4. It will be appreciated that the (total) number of storage containers 80 in buffer store 16 is limited by the physical size of buffer store 12. The dimensioning of the buffer bearing 16 will be discussed in more detail later.The buffer storage 16 supplies the work station 18 with the storage containers 80- 1 to 80- 4. FIG. 4 shows a circuit of the storage containers 80- 1 to 80- 4, since the storage containers 80 can be guided through the region of the work station 18 several times a day, in particular if the orders 40 are processed block by block. FIG. 4 shows a snapshot in which the storage container 80- 1 has already passed the work station 18 and is stored back, while the storage containers 80- 2 to 80- 4 are still located in the region of the work station 18 and are processed. The storage containers 80- 2 and 80- 3 are located in the region of the consolidation buffer device 20 for processing multi-line orders 40.The consolidation buffer device 20 is implemented here, for example, as a shelf 86 with shelf compartments 88. Four of the six shelf compartments 88 are already loaded with articles 22 that can belong to different article types 46. The shelf compartments 88- 5 and 88- 6 are (still) empty. It will be appreciated that more or fewer shelf compartments 88 may be provided in the shelf 86. Each of the shelf compartments 88 is assigned one of the orders 40. The total number of shelf compartments 88 or destination (buffer) locations can be determined in advance, as will be explained in more detail below. Usually, 20 to 100 buffer sites are provided. In the area of the shelf 88 or the consolidation buffer 20, preferably "pick-by-light" and "put-to-light" guidance strategies are used.The storage container 80- 4 is likewise located in the region of the work station 18. the articles 22 are removed directly from the storage container 80- 4, not buffered, but rather are packaged directly, as will be explained in the following, because they are single-line orders 40.The aforementioned cycle serves for the processing of "buffer-store-clean" and "mixed" jobs 40. This corresponds to 40.2% of the application lines 42 or 39.3% of the piece numbers 44.66 % of all orders 40 represent "picking store-clean" and "mixed" orders 40. This is indicated by the article streams shown on the right in FIG. 4, which leave the picking store 12 downward in the direction of a packing station 34- 3 and a merging point 92.34 % of all orders 40 represent "buffer store clean" orders 40.14,2 % of all the jobs 40 represent "mixed" jobs. 14.2% of all orders 40 therefore require both articles 22 which are stored in the picking store 12 and articles 22 which are stored in the buffer store 16.51,8 % of all orders 40 represent "order-picking-store-clean" orders 40, i.e. they require articles 22 only from the store 12.29,1 % of all orders 40 have only a single order line 42. These single-line orders 40 are operated from the storage container 80- 4, which can later also be used for processing multi-line orders 40. In the example of FIG. 4, a total of 1043 orders are single-line, which require a total of 1124 pieces. This means that some of the single line jobs 40 require more than 1 piece. The articles 22 are removed from the storage container 80- 4 in order-oriented fashion, i.e. in the correct number of items, and are placed in corresponding shipping containers 90 (e.g. cardboard, collecting containers, pallet, etc.).The packing of the single-line orders 40 is effected in the region of the packing station 34- 1 in FIG. 4. There, three shipping containers 90 are shown by way of example, each of which is loaded with a different number 44 of articles 22. Although the packing station 34-1 is shown separate from the work station 18, it should be understood that the steps of removing, repackaging, and packing may all be performed at the work station 18. In other words, the packing station 34- 1 may be integrated into the work station 18. The same applies to the packing station 34- 2 which is used for finishing the orders 40 which have a plurality of order lines 42 with different article types 46, wherein all article types 46 are stored in the buffer store 16.4,9 % of all orders 40 are multi-row and are supplied exclusively from the store 16.3,9 % of all orders 40 are also pre-commissioned via consolidation buffer 20 in order to be brought together later with articles 22 from picking store 12 and with (single-line) articles 22. The articles 22 of the mixed orders 40 originating from the buffer store 16 and the picking store 12 are merged at a point 92 and are packed and made ready for shipping at the packing station 34- 3. It is understood that the packing station 34- 3 can again be integrated into the work station 18.The ready-picked and ready-to-ship packed single-line, buffer-store-clean orders 40 (29.1%) exit the system via the article exit 32, for example in packages 94. The mixed orders 40 (14.2%) and the order-picking-store-free orders 40 (51.8%) come from the third packing station 34- 3 (66%).It is understood that only a single packing station 34 may be provided, which may be arranged inside or outside the area of the work station 18. FIG. 4 shows only an exemplary configuration of the system 10.48,2 % of all orders can thus be served from the buffer store 16. Only 51.8% of all orders must still be picked conventionally from the picking store 12. In order to achieve such effectiveness or such loading of the buffer store 16 or relief of the picking store 12, the type of loading of the buffer store 16 (filling order) with the article types 46 and the physical size of the buffer store 16 must be selected very carefully. This applies all the more because the buffer store 16 has a much smaller capacity (article type diversity and number of pieces) than the picking store 12. The mounting of the buffer bearing 16 will be described in more detail with reference to FIGS. 7 and 8.FIG. 5 shows a perspective view of a part of a system 10. In particular, a buffer store 16 and, by way of example, two work stations 18- 1 and 18- 2 are shown, at each of which a picking person 130 picks manually from storage containers 80. The work stations 18- 1 and 18- 2 are locally adjacent to one another and are supplied with the storage containers 80 via a conveying system 132.The buffer store 16 is implemented in FIG. 5 by way of example as a conveyor cell 134, which has, for example, three conveying sections 136- 140, which are arranged substantially parallel to one another and are connected to one another at the end side in such a way that the storage containers 80 can circulate endlessly within the conveyor cell 134. The storage containers 80 are preferably stored in the buffer store 16 in the form of stacks 142. For reasons of a simplified illustration, no (storage) stacks 142 are shown in the region of the conveyor top unit 134. Only at an upstream end of the conveyor 132 are four stacks 142 shown by way of example, which are separated by a destacking device 144 in order to be fed individually to the picking persons 130. A stacking device 146 is provided at a downstream end of the conveyor 132 for restacking the separated storage containers 80 before being fed back into the conveyor gyro 134.If initial filling or refilling of the buffer store 16 from the picking store 12 (not shown) is automated, a conveyor system 148, for example, is provided, which can be connected to the conveyor system gyroscope 134 or the conveyor system 132 at any point and which has a feed section 150 and a discharge section 152. In the case of manual replenishment, the conveying system 148 is not required.In FIG. 5, each of the workstations 18- 1 and 18- 2 each has a table 154 for either directly repacking the articles 22 (not shown) from the storage containers 80 into a shipping container 90 or placing them into one of the shelves 86 (see FIG. 4 ) that represent the consolidation buffer devices 20. However, the articles 22 can also be directly packaged. In this case, the work stations 18- 1 and 18- 2 also simultaneously represent the corresponding packing stations 34 (cf. FIG. 4 ). Fully packaged packages 94 may be packaged onto target cargo carriers (e.g., pallets) 156, which are preferably provided in the area of workstations 18- 1 and 18- 2.FIG. 6 shows a perspective view of a part of a further workstation 18, which in turn has a conveying system 132 for providing storage containers 80. Consolidation buffer device 20 is implemented in the form of a sorter 158. The sorter 158 has a plurality of sorter trays 160 which are fixedly connected to an endlessly circulating conveying means (not shown). The picking person 130 can be shown via a monitor 162 which quantities are to be removed from the storage container 80 respectively available for processing. Furthermore, the picking person 130 can be displayed what of the article types 46 are. In addition, the picking person 130 can be displayed, for example, from which the container compartment the respective article type 46 can be removed. FIG. 6 shows whole storage containers 80 without compartment subdivision. Furthermore, detection devices (e.g. scanners, RFID readers, cameras, etc.) can be provided which monitor and register a removal of the articles 22 from the storage containers 80 and / or a delivery of the removed articles 22 to a respective tray 160. The picking person 130 preferably always sends only a single piece to a single one of the trays 160. The trays 160 thus loaded then move away from the picking person 130 and pass downstream destination locations, which are not shown here. Each of the destination locations is assigned to one of the orders 40 so that the articles 22 of a single order 40 can be collected there.As an alternative to the sorter 158, a pick-to-bucket arrangement can also be used. Exemplary pick-to-bucket arrangements are disclosed in patent applications DE 10 2004 014 378 A1 and DE 10 2006 057 266 A1.Referring now to FIGS. 7 and 8, selection of the article types 46 and determination of the associated quantities for the buffer store 16 will be explained.FIGS. 7A to 7C show various (article type) distributions over longer periods of time, wherein, by way of example, December 1, 2010 was selected as the relevant calculation day. The article type distributions of Figs. 7A to 7C were obtained from data of old orders to simulate and verify the concept of the invention computationally, i.e. by means of computers. Viewed over the year, the article orthosis 24 considered by way of example for this purpose has approximately 22,000 different article types 46. The graph of FIG. 7A shows five article types 46 ("SKU1-5") that have sold most frequently and most recently. These are article types numbered "436112", "445290", "318317", "465401", and "459282".In order to determine the article location 24 for the buffer store 16 for the 1.12.2010, the data of the old orders are preferably evaluated from the immediate past, which are within a time window 96 before the 1.12.2010. In the present case, a time window 96 having an exemplary size of 14 days was considered. The time window 96 is illustrated in FIG. 7A by two vertical auxiliary lines 98. The time window 96 extends from 17.11.2010 to 30.11.2010. The time window 96 thus covers two weeks with a total of 11 working days (2 x 5 days of the week+1 x days of the Samstag). Orders on 11 days were thus possible. Curves 100- 1 to 100- 5 illustrate the profile of the respective article type distributions. The curves 100 are partially interrupted. Interrupts are always shown when no ordering was made at all on a possible order day. The more continuous and continuous, or more continuous, the curves 100 are, the more frequently and more temporally stable the corresponding article types 46 have been ordered. The higher the amplitudes of the curves 100, the more quantities were ordered from the respective article types 46. FIG. 7A shows that the most common article types 46 have been ordered on the order of 30-60 pieces per day (within time window 96). Nevertheless, each of the curves 100 fluctuates greatly, so that mathematical prediction over a future curve profile (e.g. by extrapolation) is impossible.This strongly fluctuating behavior is also recognizable with less common article types, as is shown in FIGS. 7B and 7C. FIG. 7B illustrates the order amounts and order frequencies of the article types 46 occupying the ranking places 101 to 105. FIG. 7C illustrates the ranking locations 201 to 205. The daily order amounts in FIGS. 7B and 7C are significantly smaller than in FIG. 7A.The temporal size of the time window 96 may be varied. Preferably, 14 days are chosen. However, more or fewer days can also be selected. The time window 96 is preferably immediately before the relevant day for which the placement or the continuously cyclically changing article location for the buffer store 16 is determined. It is understood that a time interval is possible between the calculation day and the time window 96. Studies have shown that the greater this time interval, the more inefficient the buffer storage 16 is used. In the example of FIG. 4, 48.2% of all orders from the buffer store 16 could be served.The farther the time window 96 is from the calculation day, the smaller this percentage will fail.FIG. 8 shows a table 110 corresponding to the article type distributions of FIGS. 7A to 7C. The first column represents the calculation tag (1.12.2010) which is the same for all table rows. The second column of the table 110 represents the respective article type 46. The third column of the table 110 represents a total sum of the quantities ordered article-specific within the time window 96 or a total piece count. The fourth column of Table 110 represents a number of days on which each type of article was ordered. The fifth column of the table 110 represents an ordering probability. The sixth column of the table 110 represents a arithmetically determined value expressing a significance for selection of the respective article type to be transferred from the picking store 12 into the buffer store 16. The seventh column represents a buffer piece count and indicates how many pieces of the article type concerned are identified in the filling order. It is understood that by such a transfer usually not all pieces of the selected article type are transferred from the picking store 12 into the buffer store 16. Usually, only a small number of the selected article types are transferred from the picking store 12 into the buffer store 16, because the buffer store 16 has a very much smaller capacity and usually also a very much smaller picking range.The article types 46 of the table 110 of FIG. 8 (column 2) are sorted in descending order of their significance (column 6). The five article types 46 of FIG. 7A, which occur most frequently in terms of time and quantity in the time window 96, correspond to the first five rows of the table 110 of FIG. 8.The first two rows of the table 110 of FIG. 8 are considered in more detail below.The first row relates to the article type "445290". The article type "445290" was ordered for 11 days of 11 possible days in the period of 17.11.2010 to 30.11.2010. Therefore, the ordering probability (column 5) is 100%. In total, in the relevant period 215, pieces of the article type "445290" were ordered, as shown in the third column. The value "1720" (column 6) is determined from the product of the total number of pieces (215) with the order days (11) minus a threshold value (3) of days at which the corresponding article type must actually have been ordered within the time window 96 (215 x (11-3)=1320). This threshold value can be selected as desired. In the present example, the corresponding article type must therefore have been ordered over more than three days within the time window 96, in order to be taken into account in the evaluation or the creation of the table 110.The second row of Table 110 relates to the article type "436112" ordered 234 times in total on 10 out of 11 possible days so that the ordering probability is 91% (10 / 11=0.91). According to the above calculation method, this gives a valence of "1638" (234 x (10-3) = 1638).Thus, the first two lines of the table 110 express that although the article type "436112" has been ordered more frequently in terms of quantity than the article type "445290", the fact that the ordering probability for the article type "445290" is greater than the ordering probability for the article type "436112" is more important. An article type 46 required each day should in any case be present in the assortment of buffer stores 16.The number of article types 46 that can be selected for the assortment of buffer stores 16 depends, among other things, on the physical size of the articles 22 as well as on a physical size of the buffer store 16. The dimensioning of the buffer bearing 16 will be described in more detail below. In addition, the item types 46 with low ordering probabilities are entirely removed from the table 110 because otherwise high, but very unlikely, amounts of these buffer types 46 would be included in the buffer warehouse 16 (e.g., action items that are no longer in action or remaining amount sales, etc.). Furthermore, article types 46 can be excluded or limited that, due to their volume or weight, under or exceed a predetermined amount of storage to prevent inefficient storage.The (buffer) picking rates for the article types 46 selected for the buffer store 16 are dependent on the picking range of the buffer store 16. The buffer piece count of column 7 of table 110 corresponds to the product of an average order quantity (quotient of sum and actual days of the order) and the (configurable) picking range. In the table 110, the 7th column was calculated with a picking range of 2 days for the buffer store 16. The article type "445290" of the first column of table 110 is therefore incorporated by calculation at 39.1 pieces into the buffer store 16 for the calculation day 1.12.2010.In the sorting of the rows of the table 110, additional parameters, such as the volume, the dimension and the weight of the article type 46 concerned, can also be taken into account additionally and / or alternatively. Furthermore, threshold values can again be defined for these parameters. An article type 46 which is particularly heavy or large should rather not be transferred into the buffer store 16, because it ergonomically makes picking difficult and because the space in the buffer store 16 is greatly restricted in comparison to the space in the picking store 12.From the buffer piece numbers of column 7 of Table 10 of Figure 11, container compartment divisions can be determined based on the volume and weight of the article type 46 involved. In the present case, it is assumed by way of example that the storage containers 80 are all of the same size and are furthermore of divisible design. The storage container 80 can have, for example, the standard dimensions 500 x 600 x 400 and thus defines a maximum total volume. The containers 80 may optionally be divided into 2, 4, 6 or 8 compartments, for example. Small compartment divisions are chosen whenever the article types 46 to be stored have a smaller volume. Few to no compartment divisions are selected when the selected article types 46 have large volumes and / or weights.In addition, threshold values can be defined for the maximum number of storage containers 80 which are occupied by one and the same article type 46 in the buffer store 16. Furthermore, rules can be defined which allow primarily the calculated buffer piece count (see column 7 of table 110 of FIG. 8 ) to be increased to a maximum possible piece count for the selected container compartment. In order to prevent disproportionate volumes from being transferred to the buffer store 16 by backing up, further threshold values for maximum rounding-up can be defined. Thus, it has proven advantageous, for example, if not more than five times the calculated buffer piece count is transferred into the buffer store 16.The picking range of the buffer store 16 is preferably selected such that after the end of a cycle the articles 22 located in the buffer store 16 are almost used up. Preferably, the assortment for buffer store 16 is determined every day to the new one. The cycle is then one day. Optionally, the buffer store 16 can also be filled more often on the same day, wherein the cycle duration is then correspondingly shorter. The buffer store 16 is then refilled daily or refilled in such a way that the desired daily new assortment is obtained. Retrievals from the buffer store 16 into the picking store 12 are not desired and are carried out only in the exceptional case. For this purpose, for example, a maximum service life can be defined which determines how long one of the article types 46 is allowed to remain at a maximum in one of the storage containers 80 without the corresponding container 80 being refilled or refilled.Preferably, the buffer piece counts (determined by calculation) are incremented to full bins or entire bins 80. The "adjusted" buffer piece count may be indicated in another (eighth) column (not shown). The adjusted number of buffer pieces can also take account of remaining inventory of the corresponding article type from the preceding cycle (e.g. difference between the mathematically determined number of buffer pieces and the actual number of remaining pieces). The number of required container compartments or containers is determined in advance on the basis of the volume and weight of the corresponding article type 46.FIG. 9 shows a profile 112 of a (daily) required number of storage units in the buffer store 16 with an initial filling of the buffer store 16 on the 1.12.2010. FIG. 9 is provided to illustrate the determination of a storage capacity of the buffer store 16.As before, it is assumed that, by way of example, uniform storage containers 80 are used for storing the articles 22 in the buffer store 16. Furthermore, the curve 112 has been determined on the assumption that the buffer store 16 has the 500 most frequent article types 46 of the article location 24, which in this case comprises a total of approximately 22,000 different article types 46. It will be appreciated that more or less than 500 article types may be selected. The selection, among other things, starting from the size of the entire article location 24, the supplier-specific order structure 84, the selected picking range and the like.In order to insert the articles 22 of the table 110 of FIG. 8 into the buffer store 16 on 1.12.2010, approximately 450 (standard) storage volumes 80 are required. On 2.12.2010, the same procedure as described above is performed to determine the assortment of buffer stores 16 for the calculation day 2.12.2010. This assortment for the 2.12.2010 may be compared to the remaining amounts of the assortment of the 1.12.2010 to determine a makeup needed to have the articles 22 desired for the 2.12.2010 in the buffer store 16. The number of storage containers increases to approximately 600.In the same way, the numbers of containers required for days 3.12 to 31.12.2010 are determined, resulting in the course 112 in the month December 2010. The course 112 is partially interrupted. This is explained by days when the system 10 is out of service (e.g., suntag, etc.). The curve 112 of FIG. 9 shows an average number of approximately 600 storage containers which are required to fill the buffer store 16 in a satisfactory manner.It is understood that the absolute number of storage containers and the course 112 greatly depend on the general order structure 84 of the relevant supplier. The industry can also play a role in which the supplier is working. Depending on how many old order data are present, the time period (in FIG. 9, this is the month of December 2010) can be extended or shortened. It is recommended to determine the dimensioning of the buffer bearing 16 during periods when the system is subjected to peak loads such as the Bowden cable store.FIG. 10 shows a calculated filling duration for filling or refilling the container numbers of FIG. 9 daily. It can be seen that on the day of the initial filling, i.e. on the 1.12.2010, about 5 hours are required for filling all 450 storage containers 80. Thereafter, the filling duration drastically falls (factor 4-5).FIG. 11 shows a flow chart of a method 120 for determining and generating a filling order for the buffer store 16.In a first step S 10, data of old orders are collected. These data can be provided by the supplier and relate to picking orders 40 already completed in the past. The old orders provide information, among other things, about the general order structure 84 of the supplier.In a step S 12, the data from all old orders are analyzed according to contained article types 46. The article types 46 usually correspond to the order rows 42.In a step S 14, article-type-specific ordering frequencies can then be determined both in terms of a quantity / quantity (see column 3 of the table 110 of FIG. 8 ) and in terms of an ordering frequency or probability (see columns 4 and 5 of the table 110 of FIG. 8 ). First, whether the quantitative order frequency or the time order frequency (order probability) is determined is freely selectable. The determination can also be carried out in parallel.The determination of the article types and the corresponding numbers for the buffer store 16 generally does not have to be performed column by column, but can also be performed row by row, for example.In a step S 16, the old orders are evaluated article-type-specifically. For example, column 6 of table 110 of FIG. 8 is determined in the manner described therein. The evaluation step S16 also includes ordering the rows of the table 110 of FIG. 8 (ascending, descending, etc.). Steps S12 to S16 are performed for all the article types 46 included in the data of the old orders. In this case, the threshold values mentioned in connection with FIG. 8 can be additionally taken into account.In a step S 18, those of the article types 46 are selected for the buffer store 16, which have been ordered frequently and in larger quantities in the past, for example the 500 article types 46 with the highest weights (column 5 of FIG. 8 ). The number of article types 46, here 500, may vary from cycle to cycle, particularly depending on a fill level of the buffer store 16 and a assortment composition.In step S 18, optionally, it is also possible additionally to select article types 46 which were not contained in the analysis of step S 12 but have a suddenly increased ordering probability on account of a (day) up-to-date event. For example, if bad weather is predicted for the next day, umbrellas and / or tissue could be included in the assortment of buffer warehouse 16 by selecting the respective types of articles 46, even though umbrellas and tissue did not have significant relevance within the time window 96. In this case, the operator can choose to replace a preferably very small part of the assortment determined purely by the above-described algorithm with the umbrellas and / or tissue. Preferably, selected article types 46 with lower significance are exchanged. Similarly, articles of spontaneous advertising action could be performed to name another example.After it has been determined in step S 18 which of the article types 46 have been selected for filling the buffer store 16, the buffer piece numbers are determined in an article-type-specific manner in a step S 20 (see column 6 in FIG. 8, calculated buffer piece number). This step may also include the above-described raising to LHM bins or whole LHM resulting in a further column of the table 110 of Figure 8 not shown there (actual buffering number, bin and bin partitioning dependency).In a step S 22, the current filling order for the buffer store 16 is then generated. The filling order has in particular the information from the columns 2, 7 and / or 8 of the table 110 of FIG. 8. The filling order can be converted in the picking controller 30 into corresponding transport and repackaging orders (see warehouse management 62, material flow control 64 and place management 66 of FIG. 3 ).In a step S 24, the buffer bearing 16 is then correspondingly initially filled or refilled.The flow chart of FIG. 12 serves to clarify steps S 14 and S 16 of the flow chart 120 of FIG. 11, as indicated by a diversion "A" in FIG. 11.Step S14 of FIG. 11 can be divided into steps S14-1 and S14-2 shown in FIG. 12. In step S14-1, the time ordering frequencies are determined. This means in the example of FIG. 8 that it is determined on how many days a specific article type 46 has actually been ordered. This is usually done by quotienting the values "order frequency in days" and "total number of possible order days". In step S14-2, the quantitative ordering frequencies are determined, which is done in Figure 8 by summing all orders within the time window 96, as expressed by the third column of the table 110 of Figure 8.In step S16-1, the weights of column 6 of table 110 of FIG. 8 are determined, article type-specific. The article types are then evaluated in step S16-2 by ordering or sorting the values in an article-type-specific manner.FIG. 13 shows a flow diagram of a picking method 200. The picking method 200 illustrates the manner in which and at which stations the orders 40 that have not yet been processed are processed in a system 10 according to the invention.The actual picking is based on an order-oriented analysis of the new orders according to the respectively contained article types 46 (step S 30). The picking controller 30 thus analyzes the re-orders on the basis of the contained article types 46 on a order-oriented basis.In a step S 32, it is queried in a job-oriented manner whether the article types 46 contained in a respective job 40 are all capable of being picked with articles 22 which are stored in the buffer store 16. If all article types 46 of the respective order can be picked from the buffer store 16, this is a "buffer store-free" order 40. the order 40 is then classified as a buffer store-free order 40 (step S 34) and subsequently processed at the workstation or workstations 18 (step S 36).If the query of step S 32 reveals that not all article types 46 can be picked from the buffer store 16, a query is made in a step S 38 as to whether all contained article types 46 can be picked exclusively from the picking store 12. The contained article types 46 cannot be picked from the buffer store 12 if the quantity of the corresponding article type 46 present in the picking buffer 16 is not sufficient to completely process the corresponding order 40, or if the corresponding article type 46 is not present at all in the buffer store 16. If the query of step S 38 reveals that all contained article types 46 can be picked exclusively from the picking store 12, the corresponding order 40 is classified as a "picking store-clean" order 40 (step S 40) and subsequently processed at the picking station or stations 18 (step S 42).If, however, the query of step S 38 reveals that not all contained article types 46 can be picked exclusively from the picking store 12, the logical consequence of the queries of steps S 32 and S 38 is that it must be a "mixed" order 40. The corresponding classification is carried out in step S 44. The mixed order 40 is processed at the work station or stations 18 and the picking station or stations 14 (step 46).It is understood that as an alternative to the queries of steps S 32 and S 38 it could be queried (step not shown) whether the corresponding order 40 has both article types 46 which are stored exclusively in the picking store 12 and also article types 46 which are also stored in the buffer store 16.Furthermore, after classification of the order 40, it could be queried according to step S 34 ("buffer store-ready") whether it is a single-line order 40 or a multi-line order 40. The multi-line jobs 40 require processing including the consolidation buffer 20 (see FIG. 4 ). The single line jobs 40 may be handled directly, i.e., without time delay and intermediate buffer steps, i.e., repackaged, collected, and / or packaged.Some of the advantages of the invention are considered in more detail below.A great advantage is to be seen in the increase in the picking efficiency or performance (finished orders / time), which is associated with article types 46 which are particularly frequently required in the past by the pre-picking or the manner of filling (assortment) the buffer store 16.FIG. 14 shows a diagram in which a number of employees is plotted against the time of day, the constellation considered in FIG. 4 being used as the basis. In a conventional system in which the buffer store 16 is missing and the picking takes place exclusively from the picking store 12, 10 employees distributed over the entire day are required to handle the work volume (StdT, prior art). This is indicated in FIG. 14 by a dashed line. If, in such a conventional system, additional load peaks are taken into account, which usually occur after the night and in the evening, the dot-dash curve (prior art with peak) results. In this case, 8 employees are usually required in the morning to do the work occurring, and even 12 employees are required from about 12.30 hours. According to the invention, for equal requirements, less than 8 employees are constantly required in the case of uniform utilization (see curve "Erf"). In the case with an unequal distribution of the load, slightly more than 6 employees are required in the morning and approximately 9 employees are required in the afternoon ("First with Peak"). It can be clearly seen that with the invention fewer employees are required to do the same work effort. This means that the system 10 of the invention operates more efficiently.In the present invention, the work effort (picking and / or filling) is shifted into those time phases during which the system 10 is less used altogether, such as e.g. on the morning.The present invention therefore makes it possible to achieve performance increases of from 20 to 30% in comparison with conventional solutions. Personnel costs can be reduced or more orders 40 can be managed with constant personnel outlay. This increase is achieved above all in systems which pick up products on the goods-to-man principle. Investment in larger (conventional) installations is not required. Existing installations may be extended by the work station 18 and buffer store 16 to implement the invention.A further advantage is seen in the fact that, with an increasing conversion, power peaks no longer have to be covered for a short time by increased use of personnel. The personnel insert (We working when and where?) no longer has to be controlled. Thus, for example, the short-term setting of rental personnel is superfluous. This aspect is important because there are large differences in performance between stock personnel and rental personnel.LIST OF REFERENCE CHARACTERS10 Warehouse and picking system 12 Picking warehouse 14 Picking station 16 Buffer warehouse 18 Workstation 20 Consolidation buffer device 22 Article 24 Article location 26 Article streams 28 Goods entrance 30 Picking controller 32 Packing station 34 Warehouse and picking area 36 Packing and shipping area 40 (picking) order 42 Order position / order line 44 Quantity / quantity 46 Article type 48 Data set 50 Header field 52 Priority field 54 Article field 56 Order management 58 Goods management system 60 Picking control strategy 62 Warehouse management 64 Material flow 66 Space management 68 I / O management 70 Lines 80 Storage containers 82 Order block 84 Order structure F 1, F 2 Factors 86 Shelf 88 Shelf compartment 90 Shipping container 92 Block 94 Package 96 Time window 98 Auxiliary lines 100 Distribution 110 Table 112 Profile of container number 120 Method for determining a filling order 130 Picking person 132 Conveyor system 134 FT gyro 136- 140 Conveying paths 142 Stack 144 Stacking device 148 Conveyor system, optionally 150 Feed path 152 Discharge path 154 Table 156 Target load carrier 158 Sorter 160 Trays 162 Monitor

Claims

Storage and picking system (10) for the progressive picking of articles (22) according to re-orders defining ordered articles (22) with respect to a respective ordered article type (46) including an associated quantity (44), wherein the system (10) comprises: a picking warehouse (12) in which a plurality of different article types (46) forming an article location (24) for picking are stored, wherein the picking warehouse (12) has a storage capacity that ensures a picking range of many days; a picking station (14) for a conventional picking of articles (22) stored in the picking warehouse (12); a buffer store (16) in which selected article types (46) are preferably stored in storage units (80) and which has a storage capacity which ensures a picking range of a few days, wherein the buffer store (16) is filled with the selected article types (46) which come substantially from the picking store (12) according to a filling order which is determined cyclically anew on the basis of data of old orders which represent already completed picking orders (40) from the past and which defines the selected article types (46) with their associated numbers (44), wherein the data of the old orders comprise ordered article types (46) including their associated numbers (44) and associated ordering times; a picking controller (30); and at least one workstation (18) and a consolidation buffer device (20), wherein the workstation (18) is connected in terms of material flow to the buffer store (16) and wherein the ordered articles (22) are taken and packed or taken at the workstation (18), preferably from the storage units (80), and are discharged to the consolidation buffer device; wherein the picking controller (30) is configured for the cyclical generation of the filling order, wherein the generation comprises the following steps: selecting (S 18) for the buffer store (16) the article types (46), which have been ordered frequently and in larger quantities in the past, from the data of the old orders; determining (S 20) article-type-specific buffer piece numbers for each of the selected article types (46); and generating (S22) the filling job defining the selected article types (46) with their respective buffer piece numbers.The system according to claim 1, wherein the selecting (S18) the article types (46) comprises: determining (S12) the different article types (46) included in the data of the old orders; for each of the article types (46) included in the data of the old orders, determining (S14-1) a temporal ordering frequency, preferably by quotienting from a number of ordering time units during which the respective article type (46) was ordered and a number of ordering time units during which orders were possible; for each of the article types (46) included in the data of the old orders, determining (S14-2) a quantitative ordering frequency, preferably the total number of pieces; ordering (S16-2) the different article types (46) according to temporal and / or quantitative ordering frequencies, preferably taking into account temporal and / or quantitative threshold values; and selecting the most temporally and / or quantitatively occurring article types (46).The system of claim 1 or 2, wherein cyclically generating the fill order further comprises: determining a distribution of the article type specific buffer piece counts among the storage units.The system according to claim 3, wherein the distribution of the article type-specific buffer piece numbers to the storage units (80) is determined taking into account article type-specific volumes and weights and storage unit-specific dimensions.The system of any of claims 1 to 4, wherein cyclically generating the fill order further comprises: replacing a small portion of the selected article types with article types whose ordering probability has increased abruptly due to a current event.The system of any of claims 1 to 5, wherein the article type-specific buffer item counts for the selected article types (46) are determined by multiplying article type-specific average order item counts by the picking range of the buffer warehouse.The system according to any one of claims 1 to 6, wherein the buffer store (16) has a fixed physical size corresponding to a fixed number of (uniform) storage units (80) that the buffer store (16) can accommodate to the maximum, wherein the fixed number of storage units (80) is preferably determined by a further analysis of a plurality of data sets of old orders.The system according to any one of claims 1 to 7, wherein the picking controller (30) is further configured to divide the re-orders into buffer-store-free, mixed and picking-store-free re-orders (40), wherein the buffer-store-free and mixed re-orders (40) are processed at the workstation (18) and the picking-store-free re-orders (40) are processed at the picking station (14).The system according to any one of claims 1 to 8, wherein the buffer store (16) comprises a storage unit storage device, in particular a conveyor circle (134), on which the storage units (80), which are preferably stacked one above the other, rotate continuously or comprise a horizontally rotating carousel shelf.The system of any of claims 1 to 9, further comprising a packing station (34).The system of any one of claims 1 to 10, wherein a fill order generation cycle is one day.The system of any of claims 1 to 11, wherein the consolidation buffer device (20) is a shelf shelf shelf, a pass shelf, a pick-to-bucket assembly, or a revolving sorter (158) having a picking control system.The system according to any one of claims 1 to 12, wherein the picking controller (30) is further configured to perform a warehouse management (62), a picking guide (60) and a job management (58) and to cause the buffer warehouse (16) to be filled with the articles (22) of the selected article types (46).Method for progressive picking of articles (22) according to re-orders (40) in a storage and picking system (10) according to one of claims 1 to 13, comprising the following steps: analyzing a plurality of re-orders (40) according to article types (46) contained; dividing the analyzed re-orders (40) into: buffer store-free re-orders (40) containing exclusively article types (46) stored in the buffer store (16); mixed re-orders (40) containing article types (46) stored in the buffer store (16) and in the picking store (12); and picking store-free re-orders (40) containing article types (46) stored only in the picking store (12); at the workstation (18), order-oriented processing of the article types (46) which are stored in the buffer store (16) and which are contained in the buffer-clean and mixed re-orders (40) by removing, preferably from the corresponding storage units (80) for the purpose of directly subsequent packaging or by removing and order-oriented collecting in the consolidation buffer device (20); order-oriented removal of the article types (46) which are contained in the mixed re-orders and which are stored only in the order-picking store (12) and order-oriented merging with the corresponding article types (46) from the workstation (18); and processing of the order-picking-clean re-orders by removing the article types (46) from the order-picking store (12) for the purpose of directly subsequent packaging.Method for cyclically filling a buffer store in a storage and picking system according to one of claims 1 to 13, comprising the steps of: generating a filling order according to the steps of one of claims 1 to 4; and transferring articles, which are determined with respect to an article type and an associated piece count by the filling order, from the picking store into the buffer store.The method of claim 15, further comprising the steps of: determining a current fill state of the buffer store (16) before the buffer store (16) is to be filled according to the fill order; comparing the current fill state to a desired fill state corresponding to the fill order; determining a difference between the current and desired fill states; determining an adjusted fill order corresponding to the difference; transferring the articles (22) determined by the adjusted fill order from the picking store (12) to the buffer store (16); and filling and refilling the storage units (80) with the articles (22).

Images