Apparatus for gripping packages
Abstract
Method for accepting a plurality of parallelepiped (P) packages in a random way, said packages being of random size and shape although presenting a common and predetermined external longitudinal dimension, and for selecting one of said packages (P) in an accumulator (16) in order to place it in a different designated location, the improvement of said method comprising the following stages: a) accumulation of said packages (P) in an accumulator (16) in laterally justified alignment in the order in which they were received; and b) selecting and removing packets (P) from said accumulator (16) in a non-sequential order.

Term
Term ended
Projected expiry passed 21 November 2017, 8.8 years ago.
- Priority
- Filed
- Published
- Projected expiry
- Today
22 claims: 7 independent, 15 dependent
- 1ES 2 180 076 T3 REIVINDICACIONES 1. Máetodo para aceptar una pluralidad de paquetes paralelepipáedicos (P) de una manera aleatoria, siendo dichos paquetes de tamano y forma aleatorios aunque presentando una dimensián longitudinal externa comuán y predeterminada, y para seleccionar uno de dichos paquetes (P) en un acumulador (16) de cara a colocarlo en una ubicaciáon designada diferente, comprendiendo la mejora de dicho meátodo las siguientes etapas:a) acumulacioán de dichos paquetes (P) en un acumulador (16) en alineacioán justificada lateralmente en el orden en el que se recibieron;y b) selecciáon y retirada de paquetes (P) de dicho acumulador (16) en un orden no secuencial.
- 2Máetodo seguán la reivindicacioán 1, en el que en la etapa “a” dichos paquetes (P) se alinean en contacto lineal justificado lateralmente.
- 3Máetodo seguán la reivindicaciáon 1 oá reivindicacioán 2, en el que la selecciáon de la etapa “b” se basa en datos auxiliares referentes a dicha dimensiáon longitudinal externa comuán y predeterminada.
- 4Máetodo seguán cualquiera de las reivindicaciones 1 a 3, en el que la etapa “b” se consigue midiendo realmente la dimensiáon externa de cada paquete (P) despuáes de recibirlo.
- 5Máetodo seguán cualquier reivindicaciáon anterior, en el que la etapa “b” se consigue haciendo referencia a una lectura dimensional asignada a cada uno de dichos paquetes (P) antes de su recepcioán.
- 6Máetodo seguán cualquier reivindicaciáon anterior, en el que el máetodo incluye la etapa adicional de:c) rechazo de un paquete no seleccionado (P) desde el extremo situado máas adelante seguán la direcciáon de avance de dicho acumulador (16) despuáes de que se haya realizado un nuámero preseleccionado de selecciones en la etapa “b”.
- 7Máetodo seguán la reivindicacioán 3, en el que la etapa “b” comprende las siguientes etapas:d) determinacioán de una pluralidad de emplazamientos de esquinas paralelepipáedicas en una ubicaciáon designada de una pila, estando definido cada uno de dichos emplazamientos de esquinas paralelepipáedicas en su parte inferior por unas superficies directas de soporte dirigidas hacia arriba y capaces de soportar paquetes;e) definicioán de por lo menos una colocaciáon potencial de los paquetes dentro de por lo menos uno de entre dichos emplazamientos de las esquinas;y f) colocaciáon del paquete en dicha colocacioán potencial del paquete (P) seleccionada en la etapa “e”.
- 8Máetodo seguán la reivindicacioán 3, en el que la etapa “b” comprende las siguientes etapas:d) identificacioán de uno de dichos paquetes (P) en dicho acumulador (16);e) establecimiento de una pluralidad de colocaciones potenciales de paquetes en una ubicaciáon de la pila capaz de recibir dicho paquete identificado (P);y f) colocacioán de un paquete seleccionado (P) en dicha colocaciáon de entre dichas colocaciones potenciales de los paquetes.
- 9Máetodo seguán la reivindicacioán 8, en el que en la etapa “e”, el establecimiento de dichas posiciones de colocaciáon de paquetes incluye la definicioán de una colocaciáon de un paquete configurada de tal manera que, si se coloca, la superficie lateral del paquete seleccionado estaráa alineada y “enrasada” con el lámite exterior de la ubicaciáon designada de la pila.
- 10Máetodo seguán la reivindicaciáon 3 en el que la etapa “b” comprende las siguientes etapas:d) identificacioán de uno de dichos paquetes (P) en dicho acumulador (16) y de una posicioán de colocaciáon de un paquete en una ubicacioán de la pila capaz de recibir dicho paquete identificado (P);ES 2 180 076 T3 e) evaluacióon al menos parcial de la estabilidad de dicho paquete seleccionado (P) en dicha posicióon de colocacióon utilizando unos criterios de estabilidad preseleccionados, y f) concesioón de preferencia a la colocacióon de dicho paquete en dicha posicioón de colocacióon del paquete (P) basóandose en dichos criterios de estabilidad.
- 11Móetodo seguón la reivindicacioón 3 en el que la etapa “b” comprende las siguientes etapas:d) identificacióon de por lo menos una posicioón de colocacióon de un paquete en una ubicacióon de la pila capaz de recibir un paquete identificado (P);e) evaluacióon al menos parcial de la estabilidad de dicho paquete identificado no apilado (P) en dicha posicioón de colocacióon mediante la determinacióon del porcentaje de soporte que recibiróa la superficie inferior del paquete si se colocara en dicha posicióon, y aplicacioón de dicho porcentaje en relacioón con un umbral, y f) colocacióon de dicho paquete (P) en dicha posicioón de colocacióon del paquete uónicamente si se alcanza dicho umbral.
- 12Móetodo seguón la reivindicacióon 3 en el que la etapa “b” comprende las siguientes etapas:d) identificacióon de uno de dichos paquetes (P) y de una posicioón de colocacióon de un paquete en una ubicacióon de la pila capaz de recibir dicho paquete identificado (P);e) evaluacióon al menos parcial de la estabilidad de dicho paquete identificado no apilado (P) en dicha posicioón de colocacióon mediante la determinacioón del porcentaje de soporte que recibiróa la superficie inferior del paquete si se colocara en dicha posicioón, y aplicacióon de dicho porcentaje en relacioón con un umbral, y f) colocacióon de dicho paquete (P) en dicha posicioón de colocacióon del paquete uónicamente si se alcanza dicho umbral.
- 13Móetodo seguón la reivindicacióon 3, en el que la etapa “b” comprende las siguientes etapas:d) identificacióon de por lo menos una posicioón de colocacióon de un paquete en una ubicacióon de la pila capaz de recibir un paquete identificado (P);e) evaluacióon al menos parcial de la estabilidad de dicho paquete identificado no apilado (P) en dicha posicioón de colocacióon mediante la determinacióon de la superficie efectiva de soporte disponible para la superficie inferior del paquete si se coloca en dicha posicióon, definida dicha superficie efectiva de soporte por unos lómites de la superficie efectiva de soporte, y comparacióon de dicha superficie efectiva de soporte con una zona definida de centro de gravedad asignada a dicho paquete identificado no apilado (P), definióendose dicha zona estimada de centro de gravedad como una zona en la superficie inferior de dicho paquete a traveós de la cual es probable que pase el vector de fuerza del centro de gravedad de dicho paquete no apilado, y f) colocacióon de dicho paquete (P) en dicha posicioón de colocacióon del paquete uónicamente si dicha zona del centro de gravedad cae dentro de dichos lómites de la superficie efectiva de soporte.
- 14Móetodo seguón la reivindicacioón 13, en el que dichos lómites de la superficie efectiva de soporte determinados en la etapa “e” estóan definidos al menos parcialmente por un soporte soólido del paquete que se extiende por debajo de dicha superficie inferior del paquete a travóes de por lo menos una capa de paquetes situados debajo.
- 15Móetodo seguón la reivindicacioón 3, en el que la etapa ”b“ comprende las siguientes etapas:d) identificacióon de uno de dichos paquetes (P) en dicho acumulador (16) y de una posicioón de colocacióon de un paquete en una ubicacióon de la pila capaz de recibir dicho paquete identificado (P);e) evaluacióon al menos parcial de la estabilidad de dicho paquete identificado no apilado (P) en dicha posicioón de colocacióon mediante la determinacioón de la superficie efectiva de soporte disponible ES 2 180 076 T3 para la superficie inferior del paquete si se coloca en dicha posiciíon, definida dicha superficie efectiva de soporte por unos límites de la superficie efectiva de soporte, y comparaciíon de dicha superficie efectiva de soporte con una zona definida de centro de gravedad asignada a dicho paquete identificado no apilado (P), definiíendose dicha zona estimada de centro de gravedad como una zona en la superficie inferior de dicho paquete a traveís de la cual es probable que pase el vector de fuerza del centro de gravedad de dicho paquete no apilado, y f) concesiíon de preferencia a la colocaciíon (P) de dicho paquete en dicha posicioín de colocaciíon del paquete si dicha superficie efectiva de soporte se superpone suficientemente a dicha zona de soporte de gravedad.
- 16Míetodo seguín la reivindicaciíon 15, en el que dichos límites de la superficie efectiva de soporte determinados en la etapa “e” estían definidos al menos parcialmente por un soporte soílido del paquete que se extiende por debajo de dicha superficie inferior del paquete a travíes de por lo menos una capa de paquetes situados debajo.
- 17Míetodo seguín la reivindicacioín 3, en el que la etapa “b” comprende las siguientes etapas:d) identificacioín dentro de una ubicaciíon designada de la pila de una pluralidad de colocaciones de paquetes correspondientes a un paquete identificado (P);e) asignaciíon de un índice de calidad a cada una de dichas colocaciones de los paquetes utilizando unos criterios predeterminados de calidad de apilamiento;f) identificacioín de la colocaciíon del paquete que tiene asignado el índice de calidad mías alto en la etapa “e”;y g) apilamiento del paquete en dicha colocaciíon del paquete (P) identificada en la etapa “f”.
- 18Míetodo seguín la reivindicaciíon 17, en el que dicho índice de calidad incluye la utilizaciíon de un índice de estrechamiento progresivo, que penaliza las colocaciones de paquetes que dan como resultado una pila que se estrecha progresivamente.
- 19Míetodo seguín la reivindicacioín 17, en el que dicho índice de calidad incluye la utilizacioín de un índice de nivel, que da preferencia a las colocaciones de paquetes que dan como resultado paquetes apilados que tienen superficies superiores a aproximadamente el mismo nivel.
- 20Míetodo seguín la reivindicacioín 3, en el que la etapa “b” comprende las siguientes etapas:d) identificacioín de por lo menos un paquete (P) en dicho acumulador (16), e identificacioín tambiíen dentro de una ubicaciíon designada de la pila de una pluralidad de colocaciones de paquetes correspondientes a dicho paquete identificado (P);e) evaluaciíon de la estabilidad de dichas colocaciones de paquetes;f) asignacioín de un índice de calidad a cada una de dichas colocaciones estables de paquetes utilizando unos criterios predeterminados de calidad de apilamiento;g) identificacioín de la colocaciíon estable del paquete que tiene asignado el índice de calidad maís alto en la etapa “f”;y h) apilamiento del paquete (P) asociado a la colocaciíon del paquete, identificada en la etapa “g”, en dicha colocacioín del paquete, identificada en la etapa “g”.
- 21Míetodo seguín la reivindicacioín 20, en el que dicho índice de calidad incluye la utilizacioín de un índice de nivel, que da preferencia a las colocaciones de paquetes que dan como resultado paquetes apilados que tienen superficies superiores a aproximadamente el mismo nivel.
- 22Míetodo seguín la reivindicaciíon 1, en el que la etapa “b” incluye la valoraciíon de la estabilidad de un paquete seleccionado (P) si se coloca en una colocaciíon específica de paquetes apilados encima de uno o mías paquetes (P) colocados anteriormente que comprende las siguientes etapas:ES 2 180 076 T3 c) determinacioón del porcentaje de soporte directo del paquete que tendróla el paquete si se colocara en dicha colocacióon del paquete, y toma de una determinacioón de inestabilidad si no se cumple un umbral;d) determinacioón de la superficie efectiva de soporte disponible para la superficie inferior del paquete si se coloca en dicha colocacióon del paquete, definida dicha superficie efectiva de soporte por unos lólmites de la superficie efectiva de soporte, y comparacióon de dicha superficie efectiva de soporte con una zona definida de centro de gravedad asignada a dicho paquete identificado no apilado (P), definieóndose dicha zona estimada de centro de gravedad como una zona en la superficie inferior de dicho paquete a travóes de la cual es probable que pase el vector de fuerza del centro de gravedad de dicho paquete no apilado;e) comprobacióon del soporte lateral disponible para el paquete (P) si se coloca en dicha colocacioón del paquete;f) toma de una determinacióon de inestabilidad si no se cumplen unos umbrales predeterminados en cualquiera de las etapas “d” y “e”;g) comprobacioón de la distribucioón de la superficie de soporte en dicha superficie inferior de dicho paquete (P) si se coloca en dicha colocacióon del paquete, y toma de una determinacióon de estabilidad si se cumple un umbral predeterminado;y h) apilamiento del paquete (P) uónicamente si se toma una determinacióon de estabilidad para dicho paquete si se coloca en dicha colocacióon del paquete. NOTA INFORMATIVA: Conforme a la reserva del art. 167.2 del Convenio de Patentes Europeas (CPE) y a la Disposición Transitoria del RD 2424/1986, de 10 de octubre, relativo a la aplicación del Convenio de Patente Europea, las patentes europeas que designen a España y solicitadas antes del 7-10-1992, no producirán ningún efecto en España en la medida en que confieran proteccián a productos químicos y farmacáuticos como tales. Esta informacioán no prejuzga que la patente estáeonoincluáda en la mencionada reserva.
Independent claims22
630 paragraphs in 21 sections, as filed
IS 2 180 076 T3
DESCRIPTION
Procedure and apparatus for palletizing packages of irregular size and weight.
Technical sector
The present invention relates generally to the handling of packages, and relates more particularly to a method and apparatus for receiving packages of random size and stacking the packages in a stable configuration on a pallet or other suitable location, such that the complete pallet and its contents on it can be transported to a remote destination.
Background of the invention
In the prior art, it is generally known to obtain methods and apparatus for stacking individual packages (which may also be referred to as "boxes") into one or more groups, so that groups of packages can be transported. in the usual way to a remote location. Such prior art apparatuses tend to be grouped into "random" and "non-random" palletizing systems.
US Patent No. <sup>◦</sup> 5,1745,692 issued to Mazouz, entitled "Method and Apparatus for Randomly Arriving Mixed Size and Content Parcels", discloses a method and apparatus for stacking packages using a circular "carousel" type conveyor 2 that accepts packages and stacks on pallets such as reference 6. Certain "attribute factors" are used to select a package, for example toxicity, drop tests, crush ability, brittleness and content. It seems that an important part of Mazouz's description is the use of “voxels” that have a “unit length”. For modeling purposes the "largest common voxel" is determined.
Although prior art such as the Mazouz document includes advantages, there is always a need for improvements over such prior art that provide improved precision and efficiency of pallet stacking, which is provided by the present invention.
Document US 5501571 discloses an automated palletizing system that includes a queuing station that reorganizes the sequence of a limited number of boxes to improve a final pallet configuration, and is based on a "queuing module" A multi-segment conveyor (see Figures 3a and 3b), which includes a complex series of conveyor segments that reorder packages on their way to a final pickup location. It was noted that the arrangement of US 5501571 requires a large amount of space to house the conveyor apparatus along with multiple power sources and drive motors to ensure that all conveyors operate. If one section of the conveyor fails, then the operation of the entire queuing module is compromised until the section of the conveyor is repaired. It was further observed that in addition to being large and complicated, the queuing module with its multiple transporters would also be very noisy. In contrast the present invention uses a single accumulator conveyor on which the packages are held prior to sorting and subsequent stacking. The present invention is more mechanically simple, requires less space, and consequently, it will be more silent than the prior art system, shown in US 5501571.
Summary of the invention
The present invention overcomes the shortcomings of the prior art by providing a method and apparatus for measuring, accumulating, and palletizing packages, which provide improved stacking efficiency in a time-effective manner. In connection with the foregoing, a method is provided for accepting a plurality of parallelepipedic packets in a random manner, said packets being of random size and shape but presenting a common and predetermined external longitudinal dimension, and for selecting one of said packets in an accumulator of in order to place it in a different designated location, the improvement of said method comprising the following stages:
a) accumulation of said packages in an accumulator in laterally justified alignment in the order in which they were received; Y
b) selection and removal of packages from said accumulator in a non-sequential order.
Thus it is an object of the present invention to provide an improved method and apparatus.
ES 2 180 076 T3 for handling packages.
It is another object of the present invention to provide an improved method and apparatus for stacking packages in a stable manner.
It is another object of the present invention to provide an improved method and apparatus for stacking received packets in a random manner.
It is another object of the present invention to provide an improved method and apparatus for stacking packages that are adaptable for a wide range of package dimensions.
It is another object of the present invention to provide an improved method and apparatus for stacking packages that are efficient in their use of time and space.
It is another object of the present invention to provide an improved method and apparatus for identifying the "center position" of a package that is on an accumulator conveyor.
Other objects, features, and advantages of the present invention will become apparent upon reading the following detailed description of the preferred embodiment of the invention when considered in conjunction with the accompanying drawings and claims.
Brief description of the drawings
Figure 1 is an illustrative view of an apparatus 10 according to the present invention, including an accumulator conveyor, an input feed conveyor, a medication conveyor, a gripper, and a central processor. Similarly, pallets are shown that have packages stacked on them.
Figure 2 is a top plan view of a multi-pallet arrangement supplied by a conveyor belt.
Figure 3 is a flow chart illustrating the flow of packets and data in the apparatus according to the present invention.
Figure 4 is an interactive data path / decision path flow diagram illustrating a method according to the present invention.
Figure 5 is a top plan view illustrating a multi-pallet arrangement 50 having pallets on one side of an accumulator conveyor.
Figure 6 is an illustrative top plan view of a multiple pallet arrangement 60 with pallets on either side of the accumulator conveyor.
Figure 7 is a top plan view illustrating a multiple accumulator conveyor arrangement 70 following one embodiment of the present invention.
Figure 8 is a top plan view illustrating a second arrangement 80 of multiple accumulator conveyors.
FIG. 9 is a graphical illustrative view illustrating a pyramid stacking profile 90 implemented with the present invention.
Figure 10 is a three-dimensional graphical illustration of a corner 30 according to the present invention, positioned within a three-dimensional coordinate system that includes the mutually perpendicular axes X (horizontal), Y (vertical) and Z (horizontal).
Figure 11 is a view similar to that shown in Figure 11, except that the placement of a possible package "P1" has been considered, and several corners have been created.
Figure 12 is a view similar to Figures 11 and 10, except that another packet "P2" has been added, and additional corners have been added.
Figure 13 is a top plan view of a corner, and of the supporting surfaces considered
ES 2 180 076 T3 slots to be used in a corner.
Figure 14 is a side view illustrating the contiguous packet concept.
FIG. 15A is an illustrative view of a corner register 31, and the front, back, right, and left contiguous package lists and surfaces that are associated with each corner with the present invention.
Figure 15B is an illustrative view of two lists, a packet list (including a series of packet records) and a contiguous packet list. The ease of cross-referencing provided between these two lists is also illustrated.
Figure 16 is a top plan view (along the Y-axis), illustrating the provision of a model 160 with the present invention of various contiguous packet margins around the lateral periphery of the corners in accordance with the present invention.
Figure 17 is a flow chart illustrating the logic regarding updating the geometry model. As can be seen, after having filled the accumulator buffer, in step 172 a determination is made as to whether a valid placement (acceptable according to defined stacking rules) of a packet on the accumulator is available. If not, in step 174 the stacking process is completed (performed). If a valid placement of a package is available, step 173 is carried out, which consists of selecting the placement. Once the placement is done, in step 175 the corners are updated. Next at stage 176 the corners are merged. Next in step 177, the contiguous packs and surfaces are updated. Next in step 178 the accumulator is updated, whereupon the process is repeated.
Figure 18A is a view of the modeling technique that includes the use of corners according to the present invention, illustrating the creation of a corner on the right and on top of a package "A" that has just been placed. The view has been taken along the "Z" axis in the three-dimensional model.
Figure 18B is similar to Figure 18A, except that another package "B" has been placed in the top and right corners described in connection with Figure 18A.
Figure 19 is an illustration of a modeling technique according to the present invention, which consists of a "flush" and a "offset" to generate potential packet placements.
Figure 20 is an illustration of a model 200 following the modeling technique according to one aspect of the present invention, in which a pack, once positioned, provides a new support surface for the existing corners.
Figure 21 is a logic flow diagram 210 illustrating the operation of a basic flush decision according to the present invention, essentially including the steps of looking at the top, front, and right gaps between the package and the outer boundaries.
Figure 22 is a logic flow diagram 220 illustrating a Front Flush Decision according to the present invention, in which various considerations are made to decide whether to perform the "front flush", including considerations regarding the historical mean dimensions of the packages. as well as the dimensions of the cushioned packages.
Figure 23 is a logic flow diagram 230 illustrating a Right Flush Decision according to the present invention, in which various considerations are made to decide whether to perform the "right flush", including considerations regarding the historical mean dimensions of the packages. as well as the dimensions of the cushioned packages.
Figure 24 (af) illustrate a one-dimensional shift.
Figure 25 (ad) illustrates a two-dimensional shift.
Figure 26A-26E illustrates alternative selection sequence flow diagrams.
Figure 27 illustrates a box support relationship tree 270.
IS 2 180 076 T3
Figure 28 illustrates the 280 values placed in the relationship tree.
Figures 29A-29C illustrate concepts regarding corner fusion.
Figure 30 is a flow chart 300 illustrating various steps included in a Stability Check following the present invention, in which a center of gravity zone is defined for the package whose placement is being considered.
Figure 31 is an illustration 310 of a Center of Gravity Zone 311 within the "footprint" of a package having a bottom surface 312.
Figure 32 is a side plan view of a stack 320 of packages, illustrating the concept of an "effective support surface" 334 used with the present invention.
Figures 33A-33C are a series of top plan views of stacked packages illustrating characteristics of effective package support surfaces across multiple layers of stacking.
Figures 34A-34C are a series of top plan views of packages, similar to those of Figures 33A-33C, except that an approach rectangle is used as the effective support surface.
Figures 35A and 35B are lower plan illustrations of the lower surfaces of packages to be placed, Figure 35A illustrating the use of Package Corner Windows following the present invention, and Figure 35B illustrating the use of Package Edge Windows following the present invention.
Figure 36 illustrates the concept of using four bounding edges of a polygon.
Figure 37 illustrates the use of four additional vertices in the modeled configuration of Figure 36.
Figure 38 is an illustrative side plan view of a plurality of stacked packages, proposing the placement of an additional "Place Package" as shown.
Figure 39 is an illustrative top plan view of a plurality of stacked packages (some shown in dashed lines), illustrating the concept of using a Direct Support Surface Polygon in accordance with the present invention.
Figure 40 illustrates a model 400 showing the possibility that a center of gravity right angle may fall outside of a Direct Support Surface Polygon.
Figure 41 illustrates the concept of "lateral force".
Figure 42 illustrates a stack 420 of packages.
Figure 43 illustrates the lateral spread of weight within a group 430 of packages.
Figure 44 illustrates a stack 440 of packages.
Figure 45 is a representation of the data flow illustrating the advantages of statistics-based measurements.
Figure 46 is an illustration of the calculation of a Quality Index according to the present invention.
Figure 47 is a list of weighting factors used in the quality index according to the present invention.
Figures 48A and 48B are, respectively, front and top views of bundle groups 481, 482, respectively, illustrating vertical and side overlap, respectively.
Figures 49A-49C illustrate the calculation of the height count.
IS 2 180 076 T3
Figure 50 illustrates a relative height check using a packet group 500 model.
Figure 51 shows the concept of lateral overlap within a group 510 of packets.
Figures 52A-52C illustrate the concept of leveling a box to be positioned relative to an adjoining package.
Figure 53 illustrates the height adjustment of adjoining packages.
Figure 54 is a data stream illustration 540 showing the components used in the taper index calculation according to the present invention.
Figure 55 is an illustration 550 of the use of a Relative Height Threshold according to the present invention.
Figure 56 is a side plan view of a plurality of stacked packages evaluating a proposed "Package to Place" in relation to "Excessive Blocking". If D1 (the minimum horizontal blocked length at the bottom corner surfaces) is greater than one third of WA (the average width of all the boxes in the stack and the damper) then the tapering onyx will be [D1 | (WA | 3)].
Figure 57 is an illustrative side plan view of a plurality of stacked packages, a proposed "Package to Place" being evaluated on the basis of whether it blocked possible bottom placement.
Figures 58A-58B illustrate the concept of limotropic corner checking.
Figure 59 illustrates the testing of a thin box when the box to be placed is at a bend, adding a value of 2 to the other factors shown in Figure 54 before a threshold is applied if certain criteria are met.
Figure 60 illustrates the verification of a thin box when the box to be placed was not in a bend, obtaining as a result the sum of a value 2 (1- (D1 / WA)) in the equation shown in Figure 54 if certain criteria.
Figures 61A-61B illustrate the results of creating a gap in a stack 610 by a thin package.
Figures 62A-62C illustrate the advantageous concept of finding a "matching box" for a thin box.
Figures 63A-63B illustrate the amplification of a gap within the stack configurations 630, 631, respectively.
Figure 64 illustrates the calculation of a "ladder" portion of the taper onyx factor, which is added as shown in Figure 54 before applying a threshold.
Figure 65 illustrates the calculation of another value (D3 / WA / 3) that can be added to the “ladder” part of the tapering onyx factor that adds itself to the quality onyx as shown in Figure 46.
Figure 66 illustrates ladder stacking in a 660 stack.
Figure 67 illustrates the concept of stacking rules referring to the front boundary spaces in the model shown at 670.
Figure 68 illustrates the concept of stacking rules in relation to right bounding spaces in the model shown at 680.
Figure 69 illustrates another stacking rule regarding stacking limits, which if met will result in the addition of a value of 1.5 to the taper onyx calculation.
IS 2 180 076 T3
Figure 70 illustrates a hard right boundary within a stack 700.
Figure 71 illustrates a back corner at a boundary.
Figure 72 illustrates the rules that apply when an excessively wide gap is created when a package to be placed is not close to the front limit, in a configuration shown with reference 720.
Figures 73A-73C illustrate rear alignment concepts in stacks 730, 731, and 732, respectively.
Figures 74A-74C illustrate the definition of box group boundaries shown in dotted lines, in stacks 740, 741, and 742, respectively.
Figures 75, 76 illustrate, respectively, stacks 750, 760, illustrating the concept of lateral front corner engagement.
Figures 77, 78 illustrate stacks 770, 780, illustrating different potential locations to the right.
Figure 79 illustrates a stack 770 in which the box to be placed exceeds the side contiguous packages, which refer to the aspect ratio.
Figure 80 shows a stack 800 in which a box to be placed blocks the lower corners which are the corners located on boxes A and B.
Figures 81 and 82 show a gap with respect to adjacent boxes in stacks 810, 820 respectively.
Figure 83 shows a diagram illustrating the calculation of a penalty for a contiguous gap.
Figure 84 is a flow chart 490 illustrating a search for parameters based on mean volumetric efficiency in accordance with the present invention.
Figure 85 shows a graphical view of clamp apparatus 1000 embodying the present invention.
Figure 86 is a side view of the gripper apparatus of Figure 85.
Figure 87 is an exploded view of the gripper apparatus of Figure 85.
Figure 88 is a block diagram of a control system for operating the gripper apparatus of Figure 85.
Figure 89 is a flow chart illustrating the process of positioning the gripper apparatus adjacent to the package.
Figure 90 is an isolated view of element 1007.
Figures 91A-D schematically show the clamp's final approach to a package.
Figure 92 is an end view of a modified gripper apparatus 1100 in which the side clamping mechanism can be selectively moved over the level of the package when not needed.
Figures 93-96 are AG packages being accumulated on an accumulator conveyor.
Figure 97 is a block diagram showing the interaction of the software module.
Figure 98 is an illustrative view showing the relative sizes of a Moving Enlarged Rectangular Object, a Moving Rotating Object, and a Circumscribed Rectangular Object.
IS 2 180 076 T3
TO,
Figures 99A-99D illustrate the determination of positions along a placement path. Figure 100 shows a configuration of a laying path.
Figure 101 shows a pick-up path planning setup.
Figure 102 shows a calculation of the lift height of the shock absorber.
Figure 103 shows a collision check.
Figure 104 shows the weight passing from layer to layer within a stack that includes packages
B, C, D, E and F.
Figure 105 shows the spread of the weight of packages A, B and C within a stack.
Figure 106 shows an arrangement of tasks within a group 2000 of tasks.
Figure 107 shows a 2010 stack task execution flow diagram.
Figure 108 shows a 2020 planning task execution flow diagram.
Figure 109 shows a 2030 Path Task Execution Flowchart.
Figure 110 illustrates the definition of a Moving Combined Rectangular Object.
Figures 111, 112 and 113 are flow charts referring, respectively, to peripheral devices, task execution, printing and error handling.
Figure 114 illustrates a top plan view of a group 1140 of packages A, B, C, D, E and
F, which are being accumulated on an accumulator conveyor, the first accumulated package A being positioned against a reference end stop 1142, of known position, and all packages in lateral contact with a lateral alignment guide 1142 to align them laterally.
Figure 115 illustrates a top plan view of a group 1150 of packages A, C, D, E, F, and G that are being accumulated on top of an accumulator conveyor.
Figure 116 illustrates a top plan view of a group 1160 of packages A, C, E, F, and G being accumulated on an accumulator conveyor.
Figure 117 illustrates a top plan view of a group 1170 of packages A, B, C, D, E, F, and
G, which are being accumulated on an accumulator conveyor.
Detailed description of the preferred embodiment
A general outline of the detailed description is presented below.
<sup>45</sup> I. GENERAL OPERATION II. EXHIBITION<sup>or</sup> N MA<sup>or</sup>S DETAILED
A. RECEPTION<sup>or</sup>N WITH RANDOM <sup>50</sup> B. MEASURE<sup>or</sup> N
C. ACCUMULATION<sup>or</sup>N
D. PALETIZATION<sup>or</sup>N
1. Execution of Package Stacking Decision
A) Corner-Based Modeling
1) Model Configuration
a) Creation of Corners
b) Use of Surfaces in the Register of Corners
IS 2 180 076 T3
c) Establishment of Lists of Contiguous Packages
d) Establishment of Lists of Packages with Address
2) Updating the Model
a) Adding new corners
b) Updating of existing corners
c) Redundant corner fusion
d) Updating corner surfaces and adjoining package lists
B) Establishment of all possible placements
1) General Placement Options; Flush and Scroll
2) Front Flush
3) Side Flush
4) Displacement
C) Stability Check
1) Checking the percentage of the support surface
2) Calculation of the area of the center of gravity
3) Verification of the limits of the Effective Support Surface
4) Checking the side support
5) Checking the DS Surface Distance
6) Comparison of Zone CG with Polygon SD
D) Assessment of placement
1) General exposure
2) Progressive Narrowing Index Factor
Bottom Lock
Checking for thin packages Checking the ladder Checking the limits Checking the wide interstices Sum
Relative height threshold
3) Level Index Factor
FULLY LEVEL LEVEL SIDE BELOW
ABOVE BACK LOCK
4) Other Factors
Potential level packet count Grouping ratio of air-to-air fill of packets volume of packets dimensional coverage alignment
Surface bridging
TO<sup>or</sup> blocked area
Volume locked
Gap with adjoining packet
Age of packages
Potential field
Package weight
Distance to ceiling
5) Sum of the Quality Index
6) Determination and Tuning of Parameters
E) Execution of the Multiple Loop Decision Process
IS 2 180 076 T3
1) Variation of Loop Nesting
2) Shortcuts
two. Pickup (and Placement) of Packages with Gripper
a) Clamp configuration
b) Correction of errors
3. Effective Planning of the Placement and Pick-up Path
a) Modeling
b) Configuration of the Placement Path
c) Collection Path
d) General Passage Positions
e) Departure Height / Approach to Shock Absorber
f) Departure Height / Approach to Paláe
g) Search for Collision with a U<sup>to</sup>nico Paláe
h) Search for Collision with a U<sup>to</sup>Unique Package
i) Verification of Convexity and Inspection of Collision of Packages
j) Verification of the Adjacent Passage Position
k) Forward Check
l) Backward Check
m) Addiction of Lower Tolerance
Four. Placement
E. EFFECTIVENESS OF THE SYNCHRONIZATION<sup>to</sup>N
1. Robot Movement Planning
two. Timing arrangement
3. Planning a package in advance
Four. Multitask
F. CORRECTION<sup>to</sup>N OF ERRORS
III. CONCLUSION<sup>to</sup>N
I. General Operation
The present invention relates generally to palletizing packages of arbitrary size and weight. The invention contemplates the reception of individual packages where the distribution of the size and weight of the packages is purely random, each package being able to be different with respect to the rest of the packages Preferably, all the packages are rectangular in shape, and they will be made up of material. deformable such as corrugated cardboard or solid material such as laminated wood. By using a portal-type aerial robot to palletize such packages, the invented method of the arrangement will produce a simple mechanical configuration, which will improve the time of the package pick-up-place cycle and produce a better efficiency of pallet volume utilization.
A specific straight rectangular package such as those typically shipped via the post office is randomly delivered to the stacking apparatus according to the present invention along with a number of similar packages received randomly. These packages are measured, weighed, and accumulated arranged in a row on a buffer conveyor. Next, based on a predetermined set of stacking principles, a package is selected from the row of packages on the accumulator and stacked together with other packages on a pallet. If space permits, another randomly received package is measured and placed on the accumulator conveyor, and once again the stacking principles are invoked to select the "best" package to place on the pallet. This process continues until the destination pallet (s) is (are) full or no more packages can be stacked will follow the predetermined stacking guidelines.
To determine which package to select from the accumulator conveyor and to determine where to place it on the pallet, a model (which may be computer-based) is constructed of the packages already on the pallet. This is done using outside measurements of the packages taken before the packages were placed. This model includes the use of a plurality of “corners” (which are essentially straight rectangular spaces) that combine to compose the space that
ES 2 180 076 T3 remains on the pallet. Corners can overlap.
Once the corners are defined, it is then desirable to define a plurality of "potential packet placements" which may also be referred to as "candidate packet placements", or generically "packet placements". A specific package is selected for evaluation with a specific orientation within a specific corner. First, a check is made on whether the package will even fit in the corner with that orientation. If not, another orientation is selected. If the package does not fit the corner regardless of its orientation, another corner will be selected and the process will be repeated until a corner is located that accepts the package with some predetermined orientation, or until all orientation / corner combinations have been exhausted. possible for the given package. At that point a new package is selected, and the process repeats.
If a package / orientation combination is discovered that fits within the boundaries of a corner, and if the corner is large enough relative to the package, two or more different potential package placements can be generated within the corner by means of processes known as flush and displacement. As each of these potential packet placements is generated, it is evaluated following a stability check in an effort to find a stable potential packet placement. If a flush or offset within the corner is not possible, evaluate the individual package / orientation / corner combination and follow the stability check.
The stability check provides a determination of "stable" as opposed to "unstable" for the potential placements of the packages available within each specific package / orientation / corner combination. This analysis is partially based on the amount of actual and effective underside of support the package would have if it were actually placed in the package / orientation / corner combination being evaluated, and also evaluates the amount of lateral support that will be provided to the package. by other adjacent packages (already placed). Another part of the stability analysis concerns the support of corners and edges of the packages.
Once a specific potential placement of a package is identified as stable, this placement is then evaluated against a quality index, and a quality index value is calculated and stored for that specific package / orientation / corner combination.
In a similar way, other package / orientation / corner combinations are evaluated in relation to stability, and if they are stable, they are also evaluated according to the quality index. The package / orientation / corner combination that has the highest quality index is the one chosen to be actually “executed”, that is, in fact the chosen package is removed from the accumulator conveyor by means of the gripper and transferred by means of from the gripper to the chosen corner with the chosen orientation, in such a way that the best "package placement" has been selected.
During each "pick and place" cycle of the gripper, the gripper (which is considered in this case to have just placed a package) then moves along a "pick-up path" to pick up a selected package from the conveyor. accumulator, and then moves along a "placement path" to place the pack in its selected location. These "pick up" and "placement" paths (of the gripper and the pack, respectively) will be different for each cycle. For the purposes of temporal efficiency, each of the picking and placing paths is planned according to the present invention so that their distance is minimal. Following the present invention, these paths are limited to an arrangement in separate vertical planes, that is, the package (or gripper) will only move up or down or horizontally when moved towards its destination without any sideways movement. To plan such paths an assessment of potential obstructions (typically stacked packages) is performed between the start and end positions along the "pick up" or "place" paths. This evaluation includes a determination of obstructions (typically stacked packages) that intersect a vertical plane that intersect the ends of the roads, and the establishment of acceptable "passage positions" that are simply free (ie, by above) obstructions. Scanning processes are used to discard some of the passage positions so that a preferred pick-up (or placement) path is established that is convex in its longitudinal dimension. This situation results in collection and placement paths without interference that are close to the minimum distance between their ends, given the limitation of the vertical planes of the paths.
When the gripper picks up a package from among other packages within the "package line" on the accumulator conveyor, it is advantageous to know where the package is actually located within said package.
ES 2 180 076 T3 row of packages. In this way, a cumulative error correction analysis is performed to compensate for the difference between where the packet “should” be in the packet row (based on measurements made previously in the forward direction), and where the packet “should” be. which may actually be due to packet row compression or other factors. This error correction is carried out by comparing the actual length of the packet row (measured by a sensor) with the "nominal" length of the packet row (the mathematical sum of the packet lengths on the accumulator previously measured according to the address advance through the measurement station). Knowing the relative position of the "selected" packet within the row of accumulated packets, and knowing the total accumulated error between the real global length (measured) of the row of packets and the nominal (calculated) global length of the row of packets, a portion of the total accumulated error is applied to the selected package. Next, the clamp was sent to the point where it would be in the package if there was no deformation of the same, also applying the accumulated error correction to compensate for the deformation of the package or other real errors.
According to an embodiment of the present invention, a single processor is used to process information regarding many different tasks described above, such as the stacking task, the robot path planning task, the trajectory task, the device tasks. peripherals, a printing task and an error detection task. According to the present invention, these tasks are prioritized in such a way that the trajectory task has the highest priority, followed by the error task, the peripheral devices task, the planning task, the stacking task, and the stacking task. impressed. Whenever a high priority task is performed, CPU time is immediately shifted to a low priority task.
Referring now to Figure 1, a parcel palletizing apparatus 10 according to the present invention includes the following components: an input feeder conveyor 12, a measurement conveyor 14 (including a measurement arch 15), an accumulator conveyor 16, a robotic air carrier type packet gripper 17, and a system controller 18.
The input feeder conveyor 12 justifies and classifies the packages generically indicated with the reference P. The measurement conveyor 14 measures the size and weight of the incoming packages P. The accumulator conveyor 16 (which may be a roller type conveyor) accepts measured packages and places them in direct linear contact against an end stop S.
The air-type package gripper 17 (having the gripper elements generically indicated 19) picks up packages from the accumulator conveyor 16 one at a time and places them on a shovel according to an evaluation process described in detail below. In a preferred embodiment of the invention, only two orientations of the pack are used for placement on the pallet: the first orientation is the same as in the accumulator, and the other is rotated 90 degrees with respect to a vertical axis.
Referring now to Figure 3, in a preferred embodiment of the present invention, all components of the configuration shown in Figure 1 can be controlled by means of a controller 20 including a VME bus controller such as 22 which runs a real-time multitasking operating system. The control software for the input feeder conveyor 12, measurement conveyor 14, accumulator conveyor 16, air-type packet gripper 17, and stacking algorithms run on a processor based on a Motorola or other CPU. adequate. The processor (also referred to as a controller) communicates with multi-axis servo controllers such as those known in the art to control the input feed conveyor 12, the measurement station 14, the accumulator conveyor 16, and the gripper. 17 of air-type packages. It can also interface with an analog I / O card to collect data such as weight from load cells, which in one embodiment of the present invention may reside at the measurement station.
Referring now to Figure 1, the packages flow to palletizing apparatus 10 from input feeder conveyor 12. After passing through the measuring conveyor 14, the packages are then fed to the accumulator conveyor 16. The air-type package gripper 17 picks up a package from the accumulator conveyor 16 and places it on a holding pallet such as 11. . Within the system controller 18, the measured data of the packets is stored in a computer memory, where stacking algorithms search for the best placement schedule for the measured packets. Placement planning dictates which package should be picked up from the accumulator, where it should be placed on the pallet, and what orientation of the package was used.
IS 2 180 076 T3
Information regarding all stacked packages is stored in the computer memory. During the search, the controller 18 references a geometrical model of the pile, and develops a stacking schedule. After obtaining a stacking schedule, controller 18 then initiates a move schedule, which searches for a collision-free gripper and package path that has a minimum travel distance for each pick and place cycle. Using these paths, the controller will guide the robot through a series of packet stacking cycles resulting in a fully filled pallet.
A. Random Package Receipt
It is important to understand that the present invention is directed to a specific situation where packets that differ widely in size and shape are received in a purely random manner. This random supply may be provided by an external supply conveyor such as that known in the art (not shown), or it may also be provided by allowing the inlet feeder conveyor 12 shown in Figure 1 to be accessible in a manner that allows for manual entry of random packets as known in the art.
B. Measurement of Packages
After placing the packages on the input feeder conveyor 12, they are transported from there to a measuring conveyor 14 (See Figure 1), where the required information such as height, width, length and length can be obtained. weight of packages, unless such information is already known through a previous measurement.
After said measurements are taken by the measurement station 14, they are stored in the memory of the system, to be used as described later in this application. For example, the "common outer longitudinal dimension" is taken for all packages, which can be used to estimate the actual length and location of the packages when accumulated in linear contact on the accumulator conveyor.
C. Accumulation of Packages
Continuing with reference to Figure 1, the accumulator conveyor 16 (which may also be referred to as a "buffer" conveyor) is an apparatus that accumulates multiple packages in a location that is randomly accessible to the air-type package gripper 17. .
The accumulator conveyor 16 may be as known in the art. However, generally, the accumulator conveyor 16 acted as a buffer to allow the stacking algorithms to have multiple options when picking up a package. Size and weight information regarding packages on accumulator conveyor 16 is available for use in stacking decision processes. The package selected ultimately matched the one believed to produce the highest stacking volume efficiency while being stable in the stack immediately after placement and after complete palletizing. For this reason it can be understood that obtaining an available number of package options through the accumulator conveyor 16 is very important in achieving maximum pallet volume utilization efficiency.
As shown in Figure 2, the accumulator roller type conveyor 16 is divided into two parts: a pre-damping section 16A and a damping section 16B. The damping section 16B is within reach of the air-type pack clamp 17, while the previous damping section 16A is not.
Furthermore, as best shown in Figure 2, preferably all packages on accumulator conveyor 16 are laterally justified to one side of accumulator conveyor 16, such that all packages have one side aligned substantially along a common plane. . Also, preferably, the adjacent packets are in mutual or "linear" contact. Any package in the buffer section 16B can be picked up randomly from the top by means of the air-type package clamp 18. Since the width is known for each package on the accumulator conveyor 16, the center position of a selected package on the buffer 16B can be calculated as the sum of all the widths of the preceding packages, plus half the width of the selected package, an error correction can also be provided as described above.
ES 2 180 076 T3 detailed below.
As described above, the accumulator pre-buffer section 16A of FIG. 2 is beyond the reach of the package gripper 17, and contains one or more packages before the accumulator conveyor 16 in the direction of advance. In one embodiment of the invention, the package weight and geometric information in the pre-damper is also known, and is also used in making the stacking decision. This additional package information provides two benefits. The first advantage is that in selecting a pack that is currently in the damper, information regarding previously damped packets can be used to aid in stacking selection. For example, one pack in the pre-damper can be nearly identical to another in the damper, and when combined, both packs can be effectively stacked. In this situation, the decision to select the pack in the damper for stacking depends on the existence of the other pack in the previous damper. The second advantage of using a pre-buffer is that before the accumulator pack feeding is completed, the controlling computer can initiate the evaluation of a next pack selection for stacking.
It should be understood that, as shown in Figures 5 and 6, from an accumulator conveyor 16 mine of a pallet 11 can be stacked.
Simultaneous stacking of multiple pallets offers two advantages. The first advantage is the sorting ability. Multiple pallets could be transported to different destinations. Mixed incoming packages could be sorted and stacked on these different pallets, their destination being obtained from a barcode or dense code scan. Multi-pallet stacking achieves dual effects of sorting and stacking in one stage. The second advantage is high pallet utilization. When multiple pallets are assigned to the same destination, the stacking algorithms may have more placement options for a specific packet. These additional options will result in better pallet utilization and volume efficiency.
The multiple pallets should be positioned as close to the shock absorber as possible. The arrangement of multiple pallets could be varied, with pallets on one side of the accumulator as shown in Figure 5, or on both sides of the accumulator as shown in Figure 6.
Each grouping of multiple pallets can be arranged in multiple rows and columns. Figure 5 shows a single row of pallets on one side of the accumulator. Figure 6 shows a configuration in which there is a single row of pallets on both sides of the accumulator. The distance between the shock absorber and the pallet is an important determining factor for the time of the pick and place cycle. The shorter the distance, the faster the cycle would be, since the robot is required to travel less distance.
The location for placing the first pack on each pallet is chosen as far away as possible from the shock absorber and other pallets. From the initial pallet, each pallet would follow approximately the profile of a quarter of a pyramid. Figure 9 illustrates a typical stacking profile 90. This arrangement will reduce the number of obstacles in the pick-and-place path to place subsequent packages, so that in most cases the efficient robot would move in a straight line path to the target position.
Halfway through stacking the pallet, some packages may be difficult to fit into the semi-finished stack (packages that are too large, too long, etc.). These packages tend to remain at the non-power end of the accumulator. Such packages can be rejected, and the buffer can be refilled with new packages. This method could increase the efficiency of using the stacking volume of the pallets. The rejected packages could be recirculated back to the palletizing process, at which point the stack shape could be totally different and could allow the packages that have been recirculated to be successfully stacked. .
D. Palletizing
As described above, once the packages are on the accumulator conveyor 16, a decision is made based on which package should be picked up from the accumulator conveyor 16 and placed on a pallet, in a makeshift cart, or other suitable location. stacking. This "packet stacking decision" can be based on a predetermined set of rules established according to the above technique, or a predetermined set of rules established according to the present technique.
ES 2 180 076 T3 invention. Examples of predetermined sets of rules established in accordance with the prior art are set forth in US Patent No. 5,175,692 issued to Mazouz, hereinafter incorporated into this referenced application.
1. Execution of the Package Stacking Decision
As described above, a package stacking decision process can play an important role in a pallet loading system. Decide which package to choose from those grouped on the accumulator conveyor, which orientation of the package to use, and where the package will be placed in the stack.
A package stacking decision process according to the present invention can be considered to include five elements:
1) modeling that is based on “corner” based principles;
2) establishment of potential placement locations for all packages waiting in the accumulator;
3) verification of the stability of the potential placements of the packages; Y
4) evaluation of potential package placements based on a predetermined quality index; Y
5) selection of a package placement.
It should be understood that although much of the modeling performed in this case is carried out through computers, the actual stacking could be used to create the model without deviating from the present invention.
A) Corner-Based Modeling
1) Model configuration
Following a preferred embodiment of the present invention, special attention is paid to the "corners". All the unfilled free space on the pallet is divided into individual corners, which can be superimposed on each other. As described in detail elsewhere in this application, an evaluation is made as to which corner received which packet based on a predetermined set of comparative parameters.
As described in detail below, each corner register has an origin, three longitudinal dimensions (height, width, and length), the positions and dimensions of the support surfaces, and adjoining packages lists.
a) Creation of corners
Reference is now made to Figure 10. It should be understood that a pallet has a lower limit (in the XZ plane), a ceiling limit parallel to and separated above the lower limit, and four lateral limits. A corner 30 is calculated as a rectangular free space with all sides common to either a pallet boundary, or at least partially to existing package surfaces.
Reference is also made to Figure 11 below. Initially, without having placed any packages, all the space inside the pallet limits is a rectangular free space. This is truly the first corner and is indicated as Corner 1 in Figure 10. Once a package such as P1 in Figure 11 has been placed in Corner 1 (which is the only available corner), the original Corner 1 disappears. Then the free space that remains, from that corner, is subdivided into up to 6 partitions with respect to the package: below, above, at the back, at the front, to the left and to the right. These free space partitions, whenever they exist, will also be rectangular, and are referred to as successor corners.
In some cases there will not be six successor corners left after one corner has been used. For example, if a placed packet touches a boundary of the pallet, the space between the boundary and the packet is zero, and
ES 2 180 076 T3 the corresponding successor corner was canceled. In Figure 11, the P1 package is placed against the left, rear and lower limit of the pallet, with which only three new corners can be generated: one on the front (New Corner 2), one on the right (New Corner 3), and one above (New Corner 4). In this situation a complete set of corners (in this case 2, 3 and 4) occupies all the free space in the stack. It can be seen that each of these three corners shares volume and therefore was in "intersection" with each of the other two; For this reason, as can be seen later, it can be understood that a package once placed can be arranged within more than one corner.
b) Corners Support Surfaces; Bridging Capacity
Following an embodiment of section A, a corner always has at least one "support surface". Some corners, such as corresponding ones that have a shovel floor surface as a support surface, have a single support surface that covers the entire lower boundary of the corner. However, a corner may have multiple support surfaces provided by the upper surfaces of multiple packages, and thus it should be understood that the lower limit of a corner may not always include a support surface.
Referring now also to Figure 15A, in the memory of the controller processor, each corner 30 created according to the model is assigned a corner register 31. In the corner register 31, information on support surfaces is included, including all bridgeable surfaces (described in another section) within the limits of the corner. Two surfaces are bridgeable if they can be used in combination to support a package. According to an embodiment of the present invention, it is considered that two surfaces can support a package when the differences in their relative heights fall within a previously specified tolerance, for example, 1/2 inch, and when the distance between the two surfaces is within a specified distance, for example, half the average packet length of the accumulated packets.
Figure 13 illustrates this concept. As can be seen, there are five surfaces within the limits of the corner. Each surface is labeled in the figure with its height. It is accepted that "G" is less than the average length of the packets. Since all the relative heights of the adjoining packages fall within the bridging tolerance (1/2 inch), in this case these five surfaces "belong" to a corner, and were recorded in the corner register. In general, two contiguous surfaces, as long as they can be bridged either from left to right, or from the back to the front, and both are within the limits of a corner, would belong to the same corner.
As all bridgeable surfaces are contained in a corner, it can be understood that during a search for a package placement between corners, it is only necessary to search in one corner at the same time.
Information regarding the surfaces referred to above was contained within a corner register as shown in Figure 15A.
c) Establishment of Lists of Contiguous Packages
As shown in Figure 14, a corner may be surrounded by existing contiguous packages. There can be left and right contiguous packets, and front and rear contiguous packets. As described in detail elsewhere, these contiguous packages are referred to quite frequently in the search for near-optimal placement. For this reason, for the sake of greater efficiency of the calculation, each corner separately tracks such neighboring packages.
As shown in Figure 15A, a corner register 31, apart from containing support surface information, contains four contiguous packet lists: 1) a back contiguous packet list, 2) a left contiguous packet list, 3) a list of right contiguous packets, and 4) a list of front contiguous packets. Each element of a specific contiguous packet list contains a computer memory address with a reference to a corresponding packet record, stored in a separate packet list (see Figure 15B). The package list includes individual package records, each of which includes specific package information such as location, size, and weight. In this way, it is only necessary to include a single copy of a package record in the entire system, although the package record can be accessed very quickly whenever necessary by accessing the list of packages.
IS 2 180 076 T3
d) Establishment of Lists of Packages with Address
As shown in Figure 16, to determine, for a given corner, which packets qualify as "contiguous", four contiguous packet ranges are established: 1) rear contiguous packet range, 2) left contiguous packet range, 3) right contiguous packet range, and 4) front contiguous packet range. For each list of contiguous packages, only a package that is included in or intersects with the corresponding range of contiguous packages, and that is within the limits of the paláe (in the case of multiple paláes), will have the address of its registry. registered in the corresponding list of contiguous packages. As shown in Fig. 16, following an embodiment of the present invention, the width D of all the contiguous ranges of packages is chosen as the average width of the packages. which is updated every time a new pack enters the damper. This update provision is used to try to ensure that appropriate information is stored.
In a typical situation up to 200 corners are available for placement (typically when the stack is half full).
2) Model update
After having created the geometrical model for a paláe, it must be updated every time a package is placed on it.
Reference is made to Figure 17 below when a “valid placement” is found. During the stacking procedure, if the pallet is not full, the geometrical model will be updated after placing the package in question. The geometry model update includes adding new corners, updating existing corners, merging corners, and updating surfaces. Figure 17 is a flow chart illustrating the logic for updating the geometry model. As can be seen, after the accumulator buffer has been filled, in step 172 a determination is made as to whether a valid placement of a pack on the accumulator is available. If not, in step 174 the stacking process is completed (performed). If a valid placement is available, step 173 is carried out, which consists of selecting the placement. Once the placement is done, in step 175 the corners are updated. Next at step 176 the corners are merged. Next in step 177, the contiguous packs and surfaces are updated. Then in step 178 the accumulator is updated, whereupon the process is repeated.
a) Addition of New Corners
Reference is also made to Figure 11 below. Initially, without having placed any packages, all the space inside the pallet limits is a rectangular free space. This is truly the first corner and is indicated as Corner 1 in Figure 10. Once a package such as P1 in Figure 11 has been placed in Corner 1 (which is the only available corner), the original Corner 1 disappears. Then the free space that remains, from that corner, is subdivided into up to 6 partitions with respect to the package: below, above, at the back, at the front, to the left and to the right. These free space partitions, whenever they exist, will also be rectangular, and are referred to as successor corners.
In some cases, there will not be six successor corners left after one corner has been used. For example, if a placed package touches a border of the pallet, the space between the border and the package is zero, and the corresponding successor corner will be nullified. In Figure 11, the P1 package is placed against the left, rear and bottom edge of the pallet, with which only three new corners can be generated: one on the front (New Corner 2), one on the right (New Corner 3), and one above (New Corner 4). In this situation a complete set of corners (in this case 2, 3 and 4) occupies all the free space in the stack. It can be seen that each of these three corners shares volume and therefore is in "intersection" with each of the other two; For this reason, as can be seen later, it can be understood that a package once placed can be arranged within more than one corner.
b) Updating Corners at Intersection
The intersecting corners are updated after a pack is actually placed. As noted above, the corners can intersect each other. In the example of Figure 18A, the corner to the right of the observer of package A and the corner at the top above
ES 2 180 076 T3 package A is in intersection as shown in the double cross-section shaded area in Figure 18A. As shown in Figure 18B, if package B is placed to the right of package A, it can be seen that package B is taller than package A. Also, package B not only sits in the corner to the left. right of package A, but also intersects the corner at the top of package A. Thus, after placing a package in one corner, the "receiving" corner must be removed, and each intersecting corner must also be updated.
c) Redundant Corners Fusion
In order to guarantee a minimum consumption of the computer memory and a maximum calculation efficiency, after each cycle of stacking packages, only the “independent corners” will be kept by means of a process known as “fusion”. For a given corner, I will track only independent surfaces and independent contiguous packages. Said minimum set can be obtained in the following way. First, after each package is placed, all successor corners corresponding to the affected corners are generated, and the original corner is removed. With each successor corner, the original corner is guaranteed to inherit only those surfaces and information from neighboring packages that are significant to the successor corner. Third, all redundant corners are merged. In the merging process, all surfaces and contiguous packet information are checked so that no redundant members are recorded.
Reference is now made to Figures 29A-29C. Consider that there are two corners, labeled Corner 1 and Corner 2. Corner 1 has a C1FB floor boundary and Corner 2 has a C2FB floor boundary separated by a distance "d". Also consider that all the lateral limits of Corner 2 fall within or are the same as those corresponding to Corner 1, since the distance "d" falls within a predetermined threshold. For example, the C2SB side boundary of Corner 2 falls within Corner 1. The roof boundary of Corner 2 is the same at Corner 1. In that case, if Corner 1 and Corner 2 share two surfaces that are bridgeable (as the term was used with respect to the corresponding one shown in Figure 13), in that case Corner 2 will merge with Corner 1. That is, any information (including surfaces, contiguous packages) that Corner 2 has that has not yet been associated with Corner 1 will be merged with Corner 1, while any redundant information that Corner 2 has will be eliminated. Corner 2 itself (or indeed, its record).
d) Updating Lists of Surfaces and Contiguous Corner Packages
As can be understood, after a new package has been disposed in position, it can become a new contiguous member, or provide a new supporting surface for existing corners. To accommodate these possibilities, a search is performed until the necessary increase in areas and contiguous packages for each existing corner is completed.
Figure 20 illustrates an example of the need for upgrades. Before the package to be placed was in position, there were two corners 1, 2, associated with the upper surfaces of package A and package B, separately. The package to be placed generates a new surface not only for corner 1 (associated with the upper surface of package A), but also for corner 2 (associated with the upper surface of package B). If the increase of this surface is not carried out, each of these corners will erroneously lack a valid surface, said surface being in the previous example the upper surface of the package "to be placed".
B) Establishment of all Possible Placements
As noted above, potential packet placements are evaluated for stability and other characteristics. For this reason, it can be understood that there is a need to provide a means to generate said placements, so that their characteristics can be evaluated later.
1) General Placement Options
Following the present invention, a package can be placed in different ways or in different "locations" within a corner. These "placements" can be determined by actually placing the package on a pile and taking the resulting measurements from it, or alternatively it can be done with the aid of computers. In the same way, it should be indicated that all the calculations
ES 2 180 076 T3 regarding the placement of the packages depend on the actual physical measurements taken from the packages.
Reference is now made to Figure 19, which illustrates different potential pack mats within a corner that occupies the entire length and width of a pallet; in this way the limits of the corner are also the limits of the palóe. As shown in Figure 19, a package can be placed “flush” against a front boundary (“front flush”) of the pallet (and corner) as shown by the “A” and “B” positions of the packages; It can be “flush to the right” against the right edge of the pallet as shown by the “B” and “D” positions of the packages, or it can be moved to a location E somewhere in the middle zone with or without flush . Package “C” does not present any flush or offset. Position A presents only a frontal flush. Position B has a front and right flush. Position D only has a right flush.
It should be understood that the flush of the package is done to achieve an important part of the present invention, which is to cover as much space as possible within the limits of the pallet. Flushing the packets at least makes the outermost positions of the packets available for evaluation by means of the packet stacking decisions described elsewhere in this application.
It can be understood that the displacement is performed only after the decision to flush is completed. Also, the direction of travel within a corner depends on the location of the package placement after the flush decision is completed. For example, referring to Figure 19, if the placement of a package is flush in location "B" (front and right offset), offset will only be possible in the directions directed backward and to the left. If a package is not flush and remains in position “C”, only front and right movement is possible. From position "A", only a movement backwards and to the right is possible, and from position "D", a flush to the left and front is possible.
There are some criteria for trimming a package with the limit of one pallet. In most cases, to achieve a tight and solid stack, it is preferable to place the two-sided packages against adjacent packages or at the inner limits of the pallet. However, when a package when positioned is close to an outer boundary, the package can be completely flush with the front boundary, or with the side boundary, or both. In such cases, if the support surfaces located below do not reach the outer limit, the edge of a package can still be flush with the outer limit of said surfaces, protruding and being unsupported underneath part of the package. The advantage of such a flush is to maximize the horizontal span of the current layer of packages being stacked (the current stacking layer), which is the basis for the next stacking layer. Without a flush with the limits, it can be understood that a progressively narrowing pile can be created, which can be disadvantageous.
The flush of packets according to a specific direction with a boundary depends on the size of the boundary gap, which is the gap between the packet and the boundary of the pallet. If there is no gap (there is another packet between the packet and the edge of the pallet), no flush is performed at least in that direction.
Let's take the front flush as an example. Referring again to Figure 19, if the front gap 192 is large enough to allow the placement of another package, a package to be placed should not be flush with the front limit of the pallet, since that gap is still useful for this purpose. placement. However, if the gap 192 is small enough to be ignored, for example less than one sixth (1/6) of the average width of the packages, then the package to be placed should not necessarily be flush with the part frontal, since the leveling can present inconveniences. Whenever a pack is flush with the front, even in a small distance, the mutual lateral contact between the front pack and the adjoining rear packets tends to break, and the stability of the stack deteriorates. The present invention recognizes that a package to be placed should be flush with the front limit of the pallet only when there is a considerable but unusable gap.
Additionally, the frontal flush decision according to the present invention was subject to a restriction according to which an unstable tower is not generated in the front part. This is controlled by making use of a condition according to the invention (described in detail below) whereby the supporting package (s) located below are (are) in contact with subsequent contiguous packages or is (are) ) of a fairly large size by itself (s).
IS 2 180 076 T3
If the support pack (s) of a pack to be placed are in contact with its rear contiguous pack (s) or is / are of a fairly large size, although the front flush created a gap in the layer current being stacked, at least the previous layer is solid, so the stack is still stable. However, if the support package of the package to be placed is not in contact with its adjacent rear packages, and is not large enough, the package to be placed should not be flush with the front edge of the pallet. Otherwise the new gap created by the package to be placed will reduce the stability of the stack.
2) Flush with the Front Boundary of the Paló
As shown in Figure 21, a first basic decision is made as to whether the flush with the edge of the pallet was carried out in some way, or would remain at its origin as set forth in step 211. As shown In step 212 in Figure 21, if the package was at the ceiling edge of the pallet, no flush is performed, since it will not be necessary to place packages on top of the package.
If the package was not at the ceiling boundary, in step 213 the gap (if it exists) between the front part of the package to be placed and the front boundary is calculated. Based on the analysis of this gap (described later in relation to Figure 22), a decision is made in step 220 (Figure 21) on whether or not to make the package flush with the front boundary. This step 220 is detailed later with reference to Figure 22. If the decision is “S<sup>or</sup>I ", in step 214 (Figure 21) a frontal flush is executed. If no, step 214 is bypassed, and the decision regarding the front flush for that specific packet at that specific corner has come to an end.
3) Flush with the Lateral Boundary (or “to the Right”) of the Pole
A similar decision is then made as to whether to perform the lateral flush. Again using a gap calculated to the right (between the right edge of the pallet and the package to be placed) in step 215, in step 230 (Figure 21) a decision is made to "flush to the right". This step 230 is explained in detail later with reference to Figure 23. At this time the decision regarding the lateral leveling for that specific package in that specific corner has come to an end.
4) Displacement
As shown in Figure 21, the first decision is about whether to place a package at the original location of the corner or flush it with one of the outer limits of the corner (and therefore of the pallet). If the corresponding positioning is not stable, then a successive movement forward and / or sideways at the corner was tested. Once a stable placement is found, the decision-making moves towards the “Placement Comparison and Record” as described in detail later.
Shifting packages within a corner is practiced according to the present invention, and may be important in finding a stable placement. The movement of packages is limited to two dimensions in perpendicular directions: according to the rear / front dimension, and according to the right / left dimension. As the methods for these shifts are similar, the right shift situation was taken as an example, and therefore reference is made to Figures 24a-f.
For illustrative purposes, the displacement in only one dimension was first described with the aid of Figures 24a-f, the two-dimensional (preferred) displacement was subsequently described with the aid of Figures 25 ad. The group of Figures 24 af is a series of side plan views of a plurality of packages 240 including a position of a package to be placed (shown with an "X") that is "moved" along a direction according to the present invention.
For each surface (in this case the surface on the placed package), six different “offset” positions are possible:
Left edge overhang, Left edge alignment, Left edge overlap,
ES 2 180 076 T3 right edge overhang, right edge alignment, and right edge overlap.
"Left edge overhang" means that the left edge of the package to be placed was protruding from the package below as shown in Figure 24b.
"Left Edge Alignment" means that the left edge of the package to be placed aligns with the left edge of the package below as shown in Figure 24c.
"Left edge overlap" means that the left edge of the package to be placed overlaps the package underneath as shown in Figure 24f.
"Right edge overhang" means that the right edge of the package to be placed protrudes from the package below as shown in Figure 24e.
"Right Edge Alignment" means that the right edge of the package to be placed aligns with the right edge of the package underneath as shown in Figure 24d.
"Right edge overlap" means that the right edge of the package to be placed overlaps the package underneath as shown in Figure 24a.
The part that protrudes from the package is designated as proportional to the overall width of the package to be placed, although if desired it could be a constant value. In one embodiment of the present invention, it would be one fifth (1/5) of the width of the package to be placed, as shown in Figure 24.
The overlapping part of the package is also designated as proportional to the overall width of the package to be placed, although if desired it could be a constant value. In one embodiment of the present invention, it would be a quarter (1/4) of the width of the package to be placed, as shown in Figure 24.
According to an embodiment of the present invention, the movement sequence is determined according to the distance of the package to be placed from the left coverage limit. In other words, the offset is always performed in a direction corresponding to a specific dimension (the two-dimensional offset will be described shortly), and the next selected offset position is the next position to the right of the six positions described above. It can be understood that for different sizes of the boxes, the order of movement between the positions described above may not be the same. For example, if the box to be placed is wider than the box that provides the supporting surface below, the alignment of the right edge will come before the alignment of the left edge, and not after the alignment of the left edge as shown. in Figures 24 (c) and 24 (d).
It should also be understood that the displacement position of Figure 24 (a) was never really considered, since Figures 24 to f illustrate a shift to the right, and therefore the position of the package to be placed would be displaced from a location of Origin not shown (with the contact surface between the left side of the package to be placed and the right side of the uppermost box) to the right. However, Figure 24 (a) is provided to illustrate the relationship of the six travel positions for a specific support package such as A. Unless a stable location is found first, with the package to be placed and the box. B the same six shift positions shown between the package to be placed and box A were tested.
Referring below to Figures 25 ad, it can be seen that the present invention contemplates a two-dimensional shift. In the case of a two-dimensional displacement, (displacement to the right and also displacement towards the front), there are also the following cases of frontal displacement:
rear edge overhang, rear edge alignment,
ES 2 180 076 T3 trailing edge overlay, leading edge overhang, leading edge alignment, and leading edge overlay.
In this way it can be understood that there are 36 (six times six) possible combinations for each combination of package to be placed / support surface.
Figures 25 ad are a series of illustrative top plan views of a plurality of packages 250 including a position of a package to be placed (shown in bold outline), Figures 25b-d showing the position after being placed. "Displaced" according to the present invention. Figure 25a illustrates the position of the package in its original position, Figure 25b illustrates the package after being shifted to the right in such a way that its right edge overlaps the supporting package B, Figure 25c illustrates the package after be shifted to the front in such a way that its front edge overlaps the support pack C, and Figure 25b illustrates the pack after being shifted to the right in such a way that its right edge overlaps the support pack B, and also towards the front such that its front edge overlaps the support pack C. It should be understood that there are other intermediate offset positions not shown.
Again, the offset is done only in one direction for each dimension. The order of displacement depends on the dimensional characteristics of the package. Of all the possible placements, the smaller the sum of the distances between the box to be placed with respect to the back limit of the corner and the left limit of the corner, the earlier the placement was tested.
C) Stability Check
The "Stability Check" plays an important role in the random size packet stacking procedures described herein. If the stability check standard is too restrictive, then fewer packages can be stacked, resulting in a pallet full of gaps. On the contrary, if the stability check standard is not restrictive enough, then a stacked package may in fact become unstable and fall. The method according to the present invention tries to achieve a balance between these two points, while being efficient in terms of calculation speed.
The flow diagram of the Stability Checking process is as shown in Figure 30. In the flow diagram, a potential and specific location of a package was evaluated in relation to stability, being considered either "stable" or well "unstable". The complete process includes the following determinations: 1) determination of the percentage of support surface (step 301), 2) calculation of a center of gravity zone (step 304), 3) verification of the limits of an effective support surface ( step 305), 4) check of the lateral support (step 306), 5) check of the distribution of the support surface, and 6) check of a polygon of the support surface (step 324).
1) Checking the Percentage of the Support Surface
When in place, a stacked package can be supported by multiple support surfaces provided by packages located below. For a package to stack and become stable, a threshold of a sufficient supporting surface from below is necessary. According to a characteristic of the present invention, said threshold is used as part of the Stability Check. As shown in Figure 30, when the relationship between the total area of the direct support surfaces and that corresponding to the lower surface of a package to be placed was below a certain percentage, the package to be placed is considered unstable ( step 303). When the package is heavier, the required percentage becomes proportionally higher. Following one embodiment of the present invention, a percentage of about 65% can be accepted for a 70 pound package, and 50% for a package weighing up to 30 pounds. For any package that has an intermediate weight, a linear interpolation can be performed.
The area of the direct bearing surfaces is calculated as follows. First, a search is made of the maximum height of the support surfaces that fall within the limits or the "footprint" of the package to be placed. Second, an evaluation is carried out again
ES 2 180 076 T3 of all the support surfaces found, and direct support surfaces are considered only those whose heights are within a certain tolerance, for example 0.5 ", with respect to the maximum height. This situation is similar to the concept of "bridging" described above.
If the area of the direct support surfaces is less than the preselected threshold (step 302), the package is considered unstable (step 303), and the stability check ends. If the area exceeds the threshold, the stability check proceeds to its next stage.
2) Calculated from the Zone of the Center of Gravity
Referring now to Figure 30 and Figure 31, the next stage in the stability check concerns the use of a Center of Gravity Zone of a package. It can be estimated that the deviation of the vector of the force of gravity from the real center of gravity with respect to its geometric center passes through this Zone of Center of Gravity, which as shown with reference 310 in Figure 31 has the shape of a Rectangle embedded in semicircles on each short edge, centered on the geometric center of the bottom surface of the packages, and parallel to the bottom surface of the package. As shown in Figure 31, each dimension L and W of the surrounding rectangle can be expressed as a percentage of the corresponding dimension of the bottom surface of the package.
In the stacking of corrugated packages, the allowable protrusion beyond the supporting surface is sensitive to the weight of the package. If a heavy package protrudes too far beyond its support limit, the package will tend to lean over that edge and cause instability. Therefore, according to an embodiment of the present invention, the size of the Center of Gravity Zone increases proportionally to the weight of the package. The heavier the package is, the larger the Center of Gravity Zone. In one embodiment of the present invention, the size is accepted to be 50% of the dimension of the bottom surface of the package for a 70 pound package, and 30% of the dimension of the bottom surface of the package for a package whose weight is 20 pounds or less. As a corresponding percentage, a linear interpolation between 30% and 50% would be accepted for those packages whose weight limits are between 20 and 70 pounds. Using Figure 31 as an example, if a percentage of 50% is used, the L and W values of the center of gravity zone would be half the length and width values, respectively, of the bottom surface of the package.
Once the Center of Gravity Zone has been calculated for a package, it is stored and used in subsequent evaluations within the stability check.
3) Verification of the Limits of the Effective Surface of Sópórte
As shown in step 305 in FIG. 30, one embodiment of the present invention includes evaluating and using a Limits of the Effective Support Surface. As shown in Figure 32, an Effective Support Surface of a package is a support surface of the package that effectively presents solid support for the package underneath towards the surface of the pallet, and thus can withstand a greater force. downward thrust. Figure 32 illustrates an effective support surface 334 for a single stacked pack tower 320, which in the example shown is less than the Direct Support Surface 329 provided by pack 332 directly below the pack to be placed.
Reference is now made to Figures 33A-C. When a package 336 is supported by multiple packages 335 in the first layer, the limits of the effective support surface for the package to be placed would be the smallest convex polygon that circumscribes a group of rectangular support surfaces, for example, those shown in hatched area 337 in Figure 33A. Consider that this packet to be placed is used to support the next packet in the third layer, and that the next packet is also supported by other packets 336 in the second layer (Figure 33B). With this consideration, the effective support surface for the next packet 337 would be the smallest convex polygon that circumscribes a set of convex polygons of their support surfaces, as shown in the dark area hatched in Figures 33C.
As can be seen, multi-polygon merging requires a relatively long computation and can be time consuming to perform. To simplify the calculation, the limits of the effective support surface can be approximated by means of a rectangle circumscribing all the effective support surfaces located below. Following this approximation, the limits of the effective support surface shown in Figures 33A-C can be converted into their approximate expressions as shown in the corresponding Figures 34A-C. Follow a realization of the
ES 2 180 076 T3 present invention, the approximation is used as a value for the Limits of the Effective Support Surface (designated with reference 349).
In Figures 34A-C, hatched areas such as 347 are the approximate effective support surfaces for packages at 2<sup>to</sup> layer, and the dark shaded area 349 is the Effective Support Surface (approximate) for the next pack (not shown).
In step 305, after determining the Limits of the Effective Support Surface, they are compared with the Zone of Center of Gravity determined from step 304. If the Zone of Center of Gravity falls within the Effective Support Surface, step 305 results in a value of "S<sup>to</sup>I ", which is passed to step 308. As also indicated previously, if step 308 obtains a value of" S<sup>to</sup>I "from step 305 or to a value of S<sup>to</sup>I of step 306 (lateral support) the stability check will continue. Otherwise (two NOs) in step 311 the packet will be considered unstable.
It is recognized that the approximate rectangle may be much larger than the polygon of the effective support surface. For this reason, even if the center of gravity zone is within the limits of the rectangle, it may still be outside the limits of the polygon, and the package may not be stable. To confirm the stability of the package, the direct contact surface distribution will be further evaluated in the Stability Check as described below.
4) Checking the Side Support
As indicated above, even if the (approximate) Effective Support Surface Limits do not completely contain the Center of Gravity Zone of the package to be placed, the package can be stable if it has adequate lateral support.
A side support is a solid side contact that will prevent the package to be placed from falling out to that side. For a given package there are four possible left, right, rear and front side supports. Each check of the side supports is part of the stability check.
To determine the existence of a side contact, a search is made between the packets with immediate contiguity (contiguous packets). As shown by Figures 48A and 48B, if a contiguous (existing) package has sufficient lateral surface overlap with the package to be placed, the lateral contact is considered made. Following an embodiment of the present invention, as shown in Figure 48A, a sufficient lateral overlap is defined as a vertical overlap by part of the adjoining package over a certain fraction, such as 1/3, of the height dimension of a package to be placed, regardless of whether the package to be placed touches its lower edge or upper edge. Also, as shown in Figure 48D, the adjoining package should overlap a certain fraction, such as 1/3, of the side dimension of the package to be placed.
Reference is now made to Figure 41, which illustrates the potential placement of a package P to be placed, the placement of which is considered on top of previously placed packages A and B. As shown, package A provides a Lateral Force SF against package P to be placed, which may be insufficient to prevent package P from “falling out” of package B with respect to the Pivot Point.
PP.
The lateral support force provided by adjoining packages such as A is calculated as a product of the coefficient of friction and the cumulative lateral weight. While the coefficient of friction can be experimentally chosen with an approximate value of 0.2 for carton packages, the cumulative lateral weight requires careful calculation. Cumulative side weight includes the weight of all contact packages on that side, added to the full or partial weight of all packages supported on top of said side contact packages. The partial weight of a package will be considered if the package is also supported by other packages from below (accepting a uniform weight distribution). The side contact packages and the packages placed on them can be in contact with other packages on the opposite side, they can even support other packages on their top. Based on the same rule, the full or partial weights of all relevant packages will be included in the calculation of the cumulative lateral weight.
As an example reference is made to Figure 42. Suppose we consider the placement of a package P to be placed against the free vertical side of package A, partially on top of package H, and over the gap between packages G and H. The weight cumulative side of package A is the sum of the weight of
ES 2 180 076 T3 packets A, packet C (next to pack C), packet B (on top of pack C), pack E (next to pack C), and pack D (on top of pack E). The F and G packages, which do not have lateral overlap with the P package to be placed, do not provide any contribution to the cumulative lateral weight of the A package.
It can be understood that a package to be placed not only influences the cumulative lateral weight of its contiguous contact packages, and its contact packages, but also influences the cumulative lateral weight of its support packages, and its support packages. medium. For this reason an update is necessary. To improve calculation efficiency, all cumulative side weights are updated each time a new pack is placed on the stack. The update rule is as follows. The weight of the package to be placed was added to the cumulative lateral weight of all the contact packages. These packages can be in contact with other packages on the opposite side, in that case the weight will be added to the other packages as long as they are within the influence limits of the package to be placed, and so on. However, this can act like the propagation of a wave front across the surface of a lake. For each round of weight spread, there was an assigned threshold value. An already visited contact packet was marked with a number that is equal to the threshold value provided. When the nuomer has already reached the threshold value, any subsequent visits to the same packet in this round of propagations were stopped.
Once a lateral contact was found, a check was made on the balance of the force moments with respect to the potential pivot edges. First, the support bundles are searched downward until the pivot pivot edges meet, which correspond to the closest limit of the effective support surface. Second, the moment of the force against the pivot edges is calculated for all directly or indirectly supported packages. In this calculation, the center of gravity of the current package can be considered to be at the edge of the center of gravity zone that was close to the contact side. Third, the moment of the force against the contributing pivot edges from the lateral contact is calculated.
For a column of bundles there can be multiple lateral contacts, and it is necessary to be careful in calculating the lateral moment of force for these contacts. Consider that there are two side contacts for a packet column, one occurs in an upper layer and one in a lower layer. The weight of the pack that was in contact on the top layer and any additional weight on top should be subtracted from the cumulative lateral weight of the pack on the bottom layer.
If the moment resulting from the force is going to cause the package to rotate against the adjacent lateral package, then the package is not stable; otherwise a lateral support is considered established.
An example of such calculations is shown in Figure 38. As shown, Packages 2, 3, 4, and 5 in Figure 38 are already in place, and Package 1 is the "package to be placed" whose placement is being considered. The packages 3 and 4 in Figure 38 are located on top of a common support surface such as the pallet. Package 2 was on top of Package 2 with some part sticking out. Package 5 is on top of Package 4 with no part sticking out. Package 1 to be placed also protrudes with respect to Package 2, and is in lateral contact with Package 5 considering that the force due to said contact is concentrated at point L. The dotted lines show the limits of the Effective Support Surface of Package 1 to Place. Point D, arranged at the limit of the Effective Support Surface, is the point with respect to which the moment is calculated. The current calculation includes the weight of Packages 1 and 2. Package 1 is considered to have adequate lateral support from Package 5 if Package 5 does not slide laterally due to the lateral force exerted by Package 1 when attempting to pivot relative to point D.
5) Checking the Distribution of the Direct Support Surface
Reference is made to Figure 30 below. After the "CAE?" referred to above, in step 309, a process according to the invention includes a check of the distribution of a direct support surface of a package. The purpose of the Direct Support Surface Check is not only to check the stability of the current package, but also to ensure that the package to be placed has sufficient support underneath so that future packages can be stably stacked on top. The Direct Support Surface Check generically designated 312 (Figure 30) comprises two parts: a check of the support of the corners of the pack (step 312) and a check of the support of the edges of the pack (step 314), both explained with reference to Figure 35.
IS 2 180 076 T3
As shown in Figure 35A, following the check of the support of the corners of the package, on the lower surface of a package to be placed four ideontic windows 351 are established at the corners of the package outline. Each window 351 has a width that is 1/4, and a length that is 1/4 of the package outline 350. As long as there is part of any direct support surface that falls within these windows, the 352 magnified windows are used to determine the overlap condition. Each enlarged window has a width of 3/8, and a length of 3/8 of the "footprint" of the package. For each enlarged window, its overlap with all direct support surfaces is checked. If the minimum overlap dimension is above a threshold, then a solid support is considered to be established at the corner of the package. For a small package, the threshold can be related to the dimension of the window, for example 60% of the dimension of the enlarged window. If desired, the window sizes and threshold mentioned above can be varied for different ranges of stability.
If a package has support in the four corners (step 315), or in three corners and the relationship between the total area of the direct support surfaces and that corresponding to the lower surface of the package to be placed is large enough, for example , 70% (see step 317), the packet is considered stable (steps 316 and 321), and the Stability Check is complete. Otherwise, the package edge support check 314 is performed, with the help of edge support calculation from step 314.
The actual support check of the edges of the package comprises two checks; a solid edge support check and a float edge support check. As explained in detail below, at step 322, solid edge support and floating edge support were checked and evaluated as described in later paragraphs.
Following the verification of the solid support of the edges, for four edges of a package to be placed, it is considered that an edge has a solid support if (a) there is support in two windows 351 of the corners of the package (see Figure 35A) along the along this edge, or (b) there is lateral support along this edge, and the supporting surface below was sufficiently close to the edge. "Close enough" can be defined to mean that the closest supporting surface is within a certain gap, eg, about one third of the dimension of the supporting surface.
In the following, referring to Figure 35B, the check for the floating edge support is similar to the check for the corner support of the pack. As shown in Figure 35B, along each edge of the surface of the package to be placed, an edge window 356 is established, which is as long as the length of the edge, and is attached to 1/4 of the other. side of the surface dimension. If there are any support surfaces that fall within this window, then the process calculates the overlap length of an enlarged window for the support surfaces. The enlarged window 357 has a length equal to that of the edge, and in one embodiment has a width of 3/8 the other dimension of the surface. A support surface is considered to have sufficient overlap with the enlarged edge window if the minimum dimension of the overlap area was above a certain tolerance, eg 1.5 ". This way the overlap area will be at least 1.5 "by 1.5". For a small package, the threshold can be adjusted again in proportion to the size of the window. The overlap length is the maximum length spanned by the support surface. A package is considered to have edge floating support if the overlap length is above a certain tolerance, eg 1/4 of the length of the package edge.
A package to be placed has a front / back dimension, and a left / right dimension. The edge support checking step 322 corresponding to checking the stability of the package requires that the package not tip over in either dimension. In each dimension, a package should have a solid edge support, and at least one floating edge support on the opposite edge, or is unstable as designated by step 323. As an example, in the left / right dimension, if the left side of the package to be placed has a solid edge support, the right side should have at least one floating edge support, although it can also have a solid edge support. The same situation is true for the front / rear dimension.
6) Comparison of the Center of Gravity Zone with the Polygon of the Direct Support Surface
Next, referring again to Figure 30, after the edge support check step 322 has been completed, in Figure 39, a subsequent process according to
The invention includes a check that the center of gravity zone is completely within a polygon of the direct support surface (determined in step 325).
The package edge check described above does not guarantee that a package will not tip over on an arbitrary edge that is not parallel to the side surfaces of the package. To guarantee the latter, a circumscribed polygon (see the hatched polygon in Figure 39) can be built from the direct support surfaces (considered to be rectangular). If the center of gravity zone (See Figure 31) is completely within the polygon of the direct surface, in that case it can be considered that the package is stable, follow the analysis shown in Figure 30. Otherwise, in step 327 the package will be considered unstable, and the Stability Check is complete.
The construction of a polygon of the direct surface from rectangular surfaces is composed of two stages. The first stage scans each surface, and finds the front, back, left, and right bordering edges of the package. In the case shown in Figure 39, these edges are defined by the lines FB, BG, DH, and DE. If two edges occur on the same line, they will merge into one edge. After this stage, as shown in Figure 36, there could be a maximum of 8 vertices. If two edges are constructed from the surface of a package, in that case 2 vertices will be reduced to the point of intersection, as is the case of points B and D.
Continuing with the reference to Figure 39, it should be indicated that if in the "corner" defined by the two boxes of Level I a third additional box is placed such as the one shown with the dotted line, if the point "x" is extends outside the diagonal line HG, point X becomes an additional vertex.
The second stage scans again through each surface, and looks for 4 possible additional vertices that are beyond the convex polygon composed of the previous four edges. For example, the four corner points of each surface are labeled as the back left corner, the back right corner, the front left corner, and the front right corner. Correspondingly, as an example, the diagonal edge of the constructed polygon is labeled as the rear left edge, the rear right edge, the front left edge, and the front right edge. This is followed by a check of each corner point of the surface against the corresponding edge. If the corner point extends beyond the corresponding edge, that corner point will be considered as an additional vertex as indicated above. Theoretically, there could be multiple points beyond the same diagonal edge, but the probability is low. To reduce the heat load, only the first point is recorded. After this search and construction, in real practice it is almost impossible to obtain a resulting polygon with 12 different vertices, since in this case the package in question will have to be supported by at least 8 different packages. However, these four additional possible vertices are graphically represented in Figure 37.
Once the support surface polygon has been built, in step 325 a check is made as to whether the zone of center of gravity is within all the diagonal edges of the polygon and the result is a final stability determination of " stable "at step 327 or" unstable "at step 326. This ends the Stability Check.
D) Placement Evaluation
1. General exhibition
As described above, after having measured the packages, having defined the corners, and accepting that a plurality of potential and stable placement locations have been determined for the packages, an evaluation of each potential and stable placement location is performed. To find the “best”, follow some previously set parameters. This evaluation is performed using a “quality index”, referred to in Figures 45 and 46, which is assigned to the first location of a stable and potential package placement found for a specific package / orientation / combination. corner, follow the “location search” process shown in Figure 3. For the purposes of this application, such stable placements “found first” will be referred to as “indexed placements”. The “indexed” placement with the highest quality index is chosen as the “best” placement for all package / orientation / corner combinations and then the associated package is picked up from the accumulator conveyor and placed in the chosen location as described. in detail later. It will be possible to accumulate all the quality indices for all the indexed placements and compare all the indices at once. However, in order to obtain an efficiency of calculation following an embodiment of the present invention, it is
ES 2 180 076 T3 uses a “compare and register” stage 110 (Figure 3) to keep in memory the highest quality ondx found, until a higher quality ondx is found from the indexed placement. At this point, the previous stored quality ondx is replaced by the current highest quality ondx. After all indexed placement locations in a specific package / facing / corner combination have been evaluated, the indexed placement with the highest quality onyx will remain in memory. This procedure repeats itself until all package / orientation / corner combinations have been evaluated. Then, at that time, the setting with the highest ondx for all possible package / orientation / corner combinations will be in the memory of the computer. Then that specific indexed placement is chosen as the "best" placement for that specific package / orientation / corner combination, and the placement evaluation ends.
Referring to Figure 45, it can be noted that a significant feature provided by the present invention includes the use of statistically based dimensionless measurements and decisions based on the foregoing. Following the present invention, the invented stacking method tracks statistical measurements of all packages fed into and through the system, such as average length, width, height, area, volume of packages, and other characteristics. thereof. In addition, it keeps track of the corresponding measurements for the packages that are currently in the shock absorber.
In the evaluation of a package to be placed, the previous statistical measurements are used to compare the same measurements of the package to be placed. A measurement related to the dimensions of the package, such as the width of the package, the area of the package, becomes a dimensionless measurement such as the ratio of the width of the package to the average width of the packages, the ratio of the area of the package with respect to the average width of the packages, etc. Based on thresholds (either direct or multiple values) of these dimensionless measurements a resolution is calculated on whether, for example, the package is too small, too narrow, or uses too many surfaces.
Such statistically based measurement and evaluation techniques provide two advantages. First, the placement evaluation method used with them can handle different packet sizes, since the method as a whole does not depend on the length or width of a specific packet. Second, in deciding a current placement, the stacking method includes a consideration of its effect on potential future placements. Since very little packet data is available in the buffer, statistically based measurement techniques can provide a good estimate of future incoming packets.
In the following, referring to Figure 46, according to the present invention a variety of factors can be used to arrive at a quality onyx. These factors include Level Index, Progressive Narrowing Index, Potential Level Packet Count, Alignment, A Fill Ratio.<sup>or</sup>area, A<sup>or</sup>Package Area, Package Volume, Dimensional Coverage, Surface Bridging, A<sup>or</sup>Blocked Area, Blocked Volume, Gap with Contiguous Package, Package Age, Potential Field, Package Weight, and Ceiling Distance. As noted below, some of these factors themselves include multiple factors.
Each of the factors referred to above is multiplied by the corresponding weighting factors designated as W1 -W16 in Figure 46, the products of which are added together and provide the quality index referenced above. The weighting factors are shown in Figure 47.
two. Level Index Factor
One embodiment of the present invention includes the use of a level ondx. Bonuses and penalties associated with this tier onyx are used to encourage “shelf loading”. With rack loading, each package is preferably positioned snugly against existing packages, while the heights of the top surfaces follow a monotonously decreasing step profile. The highest package should preferably be placed on the inner boundary (rear and left).
The advantage of such a monotonously decreasing step profile is twofold. First, the free space to place the next package occupies the entire span up to the outer limit (front and right) of the palóe. There may be more opportunities to fit a package into space, and there is less likely a clearance problem for robot insertion of packages. Second, it has been observed that a profile of monotonically decreasing steps is a configu28
ES 2 180 076 T3 stable ratio. Each package tends to be made with a side support at the time of placement, and with a double side support (except at the limits) when the rest of the packages are placed.
A set of indices of individual levels can be set for each side. These indices are then combined into a resulting level index, which takes a value such as Fully Level, Side Level, Below, Above, Rear Block, or No Contiguous Packets.
Fully Level means that a package to be placed is level with the adjacent side packages as well as the adjacent rear packages.
Side Level means that a package to be placed was level with at least one side adjacent package.
Underneath means that a package to be placed was below either the adjacent side packages or the adjacent rear packages.
Overhead means that a package to be placed is above the adjacent side packages but does not block the adjacent rear packages.
Rear Blocking means that a packet to be placed blocks subsequent contiguous packets.
No Contiguous Packages means that there is no contiguous package around the corner.
As shown in Figure 47, the highest bonus is given to Fully Level situations (1600), the second bonus is for Level (1100). Under (-600) is given a slight penalty, and the penalty increases when it comes to Over (-2600) or Post Block (-4000).
According to the present invention a penalty of the type is used above a side contiguous packet. When a package is placed in the middle of a stack in such a way that it rises above the immediate side contiguous packages, it will intersect the free space at the side corners. In this case, valley-shaped corners are created on the sides. With a strict requirement of free space by the robot for inserting packages, it is possible that no packages can be inserted in these corners. In this way, interstices are created on the sides, and several isolated towers will grow vertically. Such towers are inherently unstable, as they lack lateral support. Even though later packages may be found to be inserted into such gaps, after filling the gaps some smaller gaps will remain due to free space requirements. For this reason, a heavy penalty is applied.
According to the present invention a subsequent contiguous packet blocking penalty is also used. For the same reason as that described in connection with the location above a side contiguous pack, a pack that blocks a rear contiguous pack is also an inferior placement. This placement is especially detrimental when the package is only allowed to have a front loading, in which any area blocked by the package to be placed is permanently unrecoverable.
As described above, a corner may have left contiguous, right contiguous, rear contiguous, and / or front contiguous packages. First, a set of level indices is calculated for each of the side contiguous packets. The calculation methods for these indexes are similar. Next, emphasis is placed on calculating the relative height of a package to be placed with respect to its adjoining packages on the left, as shown in Figure 52.
As shown in Figure 52, adjoining packages can be at varying distances from the package to be placed. A line is drawn at a distance, which in one of the embodiments is chosen as approximately a fraction of 0.7 of the mean width of the packages (in the statistical data), with respect to the left limit of the receiving corner, such that So those members that are in or intersecting with the line are considered as close contiguous packets, and those members on the far side of the line are considered as far contiguous packets. In the calculation, priority is given to the neighboring packet.
With all the members that have a lateral overlap (see Figure 51) with the package to be placed
ES 2 180 076 T3 a search is made for the maximum height. Those members that do not have any overlap are irrelevant. Those relevant packages may be at a higher level than the package to be placed. In the search for the maximum height, all contiguous packages that are level with the package to be placed were recorded. If the height of the package to be placed is greater than the maximum height, the level onyx will be provisionally labeled as Above. If it is approximately equal to the maximum height, the level onyx is labeled Level. If the height of the package to be placed is less than the maximum height, the level onyx was provisionally labeled as Below.
If the level onyx is Below, and there are also contiguous packages that are approximately the same height as the package, then an additional search is performed to verify that the package to be leveled is not blocked at the top. If so, the level onyx will be changed to Level.
If the level onyx for the package to be placed is Above, additional evaluations are performed. So far the check has been based on neighboring contiguous packets. The question arises as to whether the corners associated with the top surfaces of nearby contiguous packages are useful. If they cannot be used to place a package, and it is insisted that any placement to the right was Above and generates a heavy penalty, in that case a tower was probably produced in neighboring packages far away. That is not a desirable situation. For this reason, tests were performed on the value of the corners associated with the upper surfaces of the neighboring neighboring packages. If they cannot be used to place a package, in that case the package to be placed will be compared to neighboring distant packages. If the package to be placed is below the maximum height of the adjoining distant packages, in that case the level onyx was labeled Below.
A similar check is extended to packets next to the right and back. To test whether there is a useful corner on the left side of the back adjoining pack, or on the right side with respect to the back adjoining pack, or on the immediate front of the back adjoining pack, a search was made between all possible corners. After carrying out a screening of the coordinates, a corner in question can be selected. The size of the corner was then checked as well as the size of the corner surface. In addition, a temporary placement for the packages in the shock absorber was attempted. Whenever a package can be placed in the blocked area, in that case the package to be placed is labeled Post Lock.
3. Progressive Narrowing Index Factor
Next, referring to Figure 46, in evaluating the placement of a packet, a tapering onyx is used to oppose packet placements leading to a tapering stack. A placement believed to lead to a tapering pile was assessed a penalty, which is proportional to the tapering onyx as described in detail below. Depending on the nature of a problem, sometimes a tapering onyx is proportional to a certain measurement, other times it can be one or more strong penalty points. As shown in Figure 54, according to an embodiment of the present invention, the progressive narrowing onyx is formed by a sum of values referring to the following factors:
Bottom Lock Slim Package Ladder Limits
Wide Gap (See Figures 56-58) (See Figures 59-63) (See Figures 64-66) (See Figures 67-70) (See Figure 72)
In the determination of the progressive narrowing onyx, a check of the lower block is carried out according to an embodiment of the present invention. As shown in Figures 56-58, bottom lock occurs when a package to be placed placed inside one corner blocks another bottom corner on one side, causing it to be partially or completely blocked and becoming unusable due to the package to be placed. . This situation can occur anywhere in the stack, and it can occur in all directions. As shown in Figure 57, if it is most likely that it is not possible to place a package in a lower corner where the package whose placement was being considered should be placed, in that case the calculation "Progressive Narrowing Index 1 ”, which means that 1 was added to the sum of the Progressive Narrowing Index values that are being added in Figure 54. If “Rule 1” is also applied, as shown in Figure 56, the value D1 / (WA / 3) was also added with D1 being the minimum blocked horizontal length
ES 2 180 076 T3 on the lower corner surfaces, and WA being the average width of all the boxes in the stack and in the buffer (accumulator).
Referring now to Figures 59-63, in determining the taper index according to one embodiment of the present invention, a thin packet check is performed. As shown in Figures 61A-B an insulated thin pack can create an unusable narrow lip to place any future pack on top. A gap generated will impair not only the volumetric efficiency, but also the stability of the stack. Two towers separated by a gap are typically less stable than a solid pile.
In one embodiment of the present invention, a package is determined to be thin when its dimension in question is much less than the average width of the packages, for example, less than a fraction of 0.7 of the average width of the packages. A thin pack needs to be watched if it is not level with an adjoining pack, or if it becomes a separate island from a side adjoining pack. Even if a slim pack does not match the height of its side adjoining pack, or becomes a separate island from a side adjoining pack, it may still be possible to find a matching pack to place on the other side of the slim pack afterwards. having placed the slim pack.
Next, referring to Figures 62A-C, a search is made on the damper to see if any packages can be placed on the other side to match the height of the thin package. Said height matching can occur at the same level as the current corner (Figure 62A), at a level lower than the current corner (Figure 62B), or at a level higher than the current corner (Figure 62C), as long as said adjacent corners are close enough to the package to be placed. If such a match can be obtained, my chances are that the slim package will not create a gap in the future.
If a slim package cannot find any matching height on either side, and the corner has enough space that can be used to fit a large package, then the placement would be assigned a heavy penalty, such as shown in Figures 59 and 60.
The thin pack check can also be extended to the front / back dimension. With the above penalty in effect, thin packages tend to be placed either as a group, or in a smaller corner.
Following an embodiment of the present invention, a ladder check is also performed. A “ladder” situation is defined as described below with reference to Figure 64. We are evaluating the placement of a package to be placed within a corner that has its origin at 2. The lower corner “1” in the side has already been determined as unusable, and in addition the package to be placed will leave an unusable surface of width D1 on the side. This situation will be penalized if the unusable surface on the side is excessive, encouraging a different placement in the corner. It should be understood that the ladder can be produced not only on the left and right sides, but also on the front and back sides. Figures 64 and 65 show the stacking rules and their associated indices.
Following an embodiment of the present invention, a limit check is also performed. For the purposes of this description, reference is made to Figure 67, Figure 68 and Figure 69.
The purpose of a boundary check is to ensure that the free space remaining (to the right or front) after placing a package is small, or that it can be used to place another package that matches the same height of the package at place. Even if there is a package that can be placed on the right but is not level with the package to be placed, it is not desirable as it creates a narrow lip and a future package may not be able to be placed on top of it.
Referring to Figure 67 as an example it can be seen that the packages are being stacked from the back to the front. When the distance D2 from the front edge of a package to be placed is within a certain distance, such as the average width WA of the packages, from the front limit, a limit check will be triggered.
If the frontal space (between the front part of the package to be placed and the limit) is not negligible, in that case in the current shock absorber a heuristic judgment or a meticulous search will be carried out to check if there is any package that, with permissible orientations, it can be placed on the right and matches the same height of the package to be placed. Taking into account all these situations, if the
ES 2 180 076 T3 current placement generates a front corner with a narrow width, and the search cannot find any package to place on the right in a way that matches the height of the package to be placed, in that case a penalty point will be applied. the formula 1.1 x (D1 | WA).
A similar check applies to the right limit. Detailed calculations are as shown in Figures 67, 68 and 69.
Referring now to Figure 72, following an embodiment of the present invention, a "wide gap" check is also performed, which is part of the taper rate determination. An excessively wide gap will cause the formation of towers or a progressive narrowing. If certain criteria shown in Figure 72 are met, a value of Z (Min (D1, D2) / WA) or 1 can be added.
Following an embodiment of the present invention, a check is also made of the relative height (see Figure 50), which is the relative height of the lower surface of the corner with respect to the highest upper surface of the adjoining package.
The calculation referenced earlier in this section is not 100 percent deterministic; Usually the relative height check requires an answer as to whether a given corner is useful or not. Typically this is done by comparing the average size of the packages with the sizes of the corners in question, and the supporting surfaces. In addition, a packet check is performed on the damper, to see if there is any packet that can actually be placed in the corner. This calculation is inherently inaccurate; You are trying to decide if there is a future package that can be placed, but the information for that future package may not yet be available at that time.
Due to this inaccuracy, a relative height threshold is used (see Figure 55). The resulting taper index will be multiplied with the threshold value as shown in Figure 55. Relative height is the height of the adjoining package relative to the placement corner. As soon as said relative height is above a high threshold, for example twice the average height of the packages, in that case the aforementioned rate of taper will be eliminated. When the relative height is between a low threshold, for example one and a half times the average height of the packages, and the high threshold, the rate of taper will decrease linearly. This adds value to the robustness of the method.
4) Other Factors
Another factor used is a potential leveled packet count. The potential level pack count is the approximate number of packs that can be placed to match the same height as the pack to be placed. Two situations provide different types of calculation. The first situation occurs when the placement corner has enough area in it to accommodate other packages in addition to the package to be placed. If so, a search is made on the shock absorber. The search will return a candidate count of packages in the shock that share the same height (within a tolerance) as the package to be placed. The placement corner does not necessarily have to be able to hold as many packages. A limit count is then calculated as the ratio of the remaining area of the corner left after box placement and the mean area of the packages. Comparing the candidate count with the limit count, the lower number will be taken as the potential level packet count (see appendix). The second situation occurs when the package to be placed is quite close to an adjacent low or high corner, in that case the previous calculation will be based on the low or high corner in addition to the placement corner. The potential leveling count will be given a proportional bonus, which encourages the placement of multiple levels of packages. This is especially useful for corner selection.
To speed up the calculation, for each bundle in the damper, and each bundle orientation, a "matching height count" is precalculated to determine how many bundles in the damper share the same height as the current bundle. After picking up a pack from the damper and feeding a new pack into the damper, only the corresponding count will be updated.
Figure 49 illustrates the concept of such a "matching height count". In condition “A”, it is considered that there are four boxes, with box 1 presenting a height of 6 "and box 2 presenting a height of 8". Box 3 is 6 "high and Box 4 is 9" high. At this time the count of matching heights for each box is 1, 0, 1, and 0. Condition B shows Box 3 removed.
ES 2 180 076 T3 of the buffer conveyor, such that the count for the remaining three Boxes 1, 2, and 4 is 0, 0, and 0, since no box has another box with a "matching" height. After you have added Box 5 (which has a height of 8 ”) to the shock, the count of matching heights for Boxes 1, 2, 4, and 5 is 0, 1, 0, and 1.
Other factors are used, each of which includes the use of “bundling of packages”. These factors are the air fill ratio, package area, package volume, dimensional coverage ratio, and alignment.
When a package is placed in a certain corner, an assessment of the quality of a group of packages can be made. A package group is a set of shock absorber packages (including the package to be placed) that can be fitted in an area at the current corner, and which also coincide with the same height as the package to be placed. Referring now to Figures 74A-C, in the configuration "(a)", the geomometric boundaries (the shaded area in Figures 74A-C) of a group are bounded by the front with the front boundary of the pallet and by the right with the right edge of the package to be placed, or as shown in Figure 74 (b) with the right edge of the adjacent rear packages (level and taller than the package to be placed), whichever is longer. It can also include a right edge of the pallet when the packet is close enough to that edge, as shown in Figure 74 (c).
If the limits of the calculation are wider than the box to be placed, in that case a placement is made to the right, such as that shown in Figures 77 and 78. Furthermore, the right box can be placed in a lower corner than the current corner.
We also looked for the placement on the front of the box to be placed. The front corner may have a deviation towards the left side of the box to be placed, as shown in Figures 75 and 76. Also, the positioning may be limited by the provisional positioning to the right of the box to be placed, such as shown in Figures 77 and 78.
The Filling Ratio of A<sup>or</sup> Area is defined as the relationship between the total area of the grouped packages with respect to the area within the calculation limits mentioned above. A bonus for the Filling Ratio of A<sup>or</sup> Area will encourage maximum packing of packages in a limited area.
The A<sup>or</sup> Package Area is the total area of the package group. The A Bonus<sup>or</sup> Package Area is proportional to the ratio between the grouped package area and the average package area. Said bonus is applied in the following situations: (a) a corner was near a boundary of the pallet, and the bonus of the area would promote a tight fit with the boundary, (b) there are gaps between the surface of the corner and the boundary of the corner, and a large area of the packages encouraged coverage of those interstices, or or (c) when placing a first package on a new shelf.
Package Volume is the total volume of the package group. The bonus of the Volume of the Packages is proportional to the relationship between the Volume of the Group of Packages and the average volume of the packages. Following one embodiment of the invention, the Pack Volume bonus is applied to the initial pack of a new shelf. In this case the height is as important as the area, since if a shelf is too low, then few packages can be placed on the side without being above the current shelf. The Package Volume bonus will also apply when a package is near the ceiling of the pallet. This will encourage the top of the pack to be as close to the ceiling as possible.
The Dimensional Coverage Ratio is divided into 2 onxes: the Front Dimensional Coverage Ratio and the Side Dimensional Coverage Ratio. The Front Dimensional Coverage Ratio is defined as the ratio of the maximum dimensional span from the front to the back of a group of packages with respect to that corresponding to the corner support surface. The Lateral Dimensional Coverage Ratio is defined as the ratio between the maximum stretch of the lateral dimension of a group of packages with respect to that corresponding to the corner surface. Side Dimensional Coverage is applied when a corner was close (less than the average length of the packages) to a side boundary. The bonus for both dimensional coverage relationships is to encourage maximum dimensional fill to the limit.
In the calculation of a Dimensional Coverage Ratio, if along that dimension the package to be placed protrudes above a lower corner (See Figure 80), in that case the projecting part will be penalized with a negative term.
IS 2 180 076 T3
Furthermore, if, as shown in Figure 79, along the front dimension the package to be placed extends beyond all the front edges of its adjacent lateral packages that are taller than or level with the package to be placed, and the free space in the front of the adjoining lateral packages is useful, that part that extends further beyond will also be penalized with a negative term. The negative term is calculated as the ratio of the length that extends further than the average width of the packages.
An alignment factor is also used. With regard to the Alignment factor, it is desirable that a package to be placed can be aligned with its rear contiguous packages and its side contiguous packages. The alignment with the subsequent contiguous packets is taken as an example. Referring now to Figure 73, a package to be placed is considered to be aligned with the adjoining rear packages ("back alignment") if the right edge of the package to be placed is close to the right edge of a special rear adjoining package. The special rear contiguous package is the member furthest to the right of the rear contiguous packages that will be next to the package to be placed, and that will not be below the height of the package to be placed. In Figure 73 (a), the box to be placed is "rear-aligned" with the adjoining back Box A since its right edges are close enough (within a pre-set tolerance) to each other, and Box A it is the rightmost rear contiguous packet. In Figure 73 (b), it is still considered that the box to be placed is aligned at the rear with Box A; although Box A is no longer the rightmost box (it is now BoxB), it is not the next box even though it is an adjoining package. In case 73 (c), there is still rear alignment since Box B, although close enough and is an adjoining package, is still lower than the box to be placed.
A bonus will be proportional to the number of lineups. It can be seen that there is a maximum of 3 alignments, rear, left and right.
Another factor used is Surface Bridging (see Figure 33). A pack can bypass the front / back side or the left / right side, or both. The bridging bonus is proportional to the tower build count of the bridged surfaces. Each surface features a tower count. The tower count is 0 when a surface bridges multiple surfaces underneath. If the surface is supported only by one surface, then the tower count will be increased by one relative to the tower count of the supporting surface. As the tower grows, so does the tower count. The bonus for bridging will be proportional to the tower count. The higher the tower, the greater the bonus. This is because bridging becomes more important when the height of the tower increases.
The A is also used as considerations.<sup>to</sup> Locked Area and Locked Volume. A package to be placed can block a certain portable area and the volume of portable lower corners. The calculation of the Blocked Area and the Blocked Volume complies with the overlap rule. As shown in Figure 78, a package to be placed can not only block a lower corner, at the same time it can render the remaining area or volume on the side unusable. This area and volume can be produced on the back, left, right or front.
Usually a package to be placed will block part of the area or volume. The larger the surface area of the package, the more area or volume it will probably block. Taking this aspect into account, the package to be placed is assigned a discount on the blocked area and the blocked volume. The discount is proportional to the area and volume of the package to be placed. The proportional factor is chosen as one eighth of the values corresponding to the package to be placed.
The penalty does not apply directly to the blocked area or the blocked volume itself. The area (volume) blocked will be divided by the area (volume) mean, and a penalty will be based on the resulting ratio.
Gaps with neighboring packets are also taken into account. There can be two types of interstices. A gap is measured from a bundle to a corner boundary. This is called a corner gap, and may be due to shifting or flushing of the packages as described above. The other gap is measured between the package to be placed and neighboring packages, as shown in Figure 79. This is called the top gap. The future package to be placed on top will be affected by the upper gap. The upper gap is partially due to the height of the package to be placed. If there is such a possibility, it is desirable that both types of gaps with respect to the side and the rear are small. Therefore, a penalty will be applied
ES 2 180 076 T3 proportionally to the ratio of said gap with respect to the average width of the packages.
The age of the bundles in the bundle damper is also a factor. After each pack is placed, a pack that still remains in the shock will increase its age count by one. A small bonus is also applied proportional to the age count of each package. Usually the package that stays in the shock for a long time has an unusual shape (too long or too big). If for a certain placement, both an unusual shape package and a regular shape package are good candidates, in that case the unusual shape package should be selected first, since it will free up a space in the shock absorber, and in it it will be possible to add a more useful package.
If, as in one embodiment of the present invention, the packet buffer mechanism allows packet rejection, in that case a packet age threshold can be selected to filter old packets in order to reject them. It can be understood that older packages will be near the end of the accumulator.
The Potential Field is another factor. A penalty is applied that is proportional to the distance of the corner surface from the floor of the pallet, the distance from the rear boundary and the distance from the left boundary. Said penalty exerts a force of attraction towards the interior origin of the pallet. Along the vertical dimension, a heavy penalty will encourage a placement in a lower corner. Along the horizontal dimensions, the penalty is arranged in such a way as to encourage placement from the inside out, and so that it fills a short dimension first before a longer dimension. Such a potential field penalty will help attract packet placement so that a tightly nested stack is formed.
The Weight of the Packages is another consideration. To achieve a stable stack, it is desirable that the heaviest packages are placed near the bottom, and the light packages are placed near the top. Based on the statistics of the distribution of the package weights, a weight threshold can be selected at some point above the average weight of the packages, eg, 30 pounds. A height threshold is set at half the height of the pallet, for example, 2 feet for a pallet with a height of 4 feet. When the weight of a package is above the weight threshold, if the bottom of the package is below the height threshold, a bonus will be assigned, otherwise a penalty will be assigned. The bonus or penalty is proportional to the weight of the package, and is proportional to the distance between the bottom of the package and the height threshold. For a tall pallet, avoid placing excessively heavy packages near the top.
The Distance to the Ceiling is another factor that can be taken into account. When a package is close to the ceiling of the pallet, and there is no package that can be placed on top of the package to be placed, a penalty will be applied based on the distance between the top surface of the package with respect to the ceiling of the pallet. The penalty is produced to promote a placement that minimizes said gap that does not provide any service.
In summary, by referring back to Figure 46 below, the approximately sixteen factors shown in this figure can be used to obtain the Quality Index as described below.
5) Sum of the Quality Index
As described above, once it has been determined that the placement of a package is stable, its relative merit is assessed by means of the Quality Index, calculated as shown in Figure 46. As shown, the Quality Index is a weighted sum of the factors described above. The detailed weighting factors are as shown in Figure 47, where a positive value represents a bonus, and a negative value represents a penalty.
A quality index is calculated for each potential placement, and the actual placement is the one with the maximum index. Following an embodiment of the present invention, the typical values of the Quality Index are between -4000 and +1600.
6) Parameters
Making stacking decisions involves many parameters, such as the corresponding values35
ES 2 180 076 T3 regarding the stacking rules, or the weighting factors in the calculation of the quality index. The invented method uses a computer simulation to tune each parameter independently. Each parameter passes through a loopback as shown in Figure 84.
Following this process of the search loop, first an initial value of the paraometer is assigned. In addition, a file is provided containing a set of packet data corresponding to a "test group" of illustrative packets based on a selected historical sequence of packet data. For the paraometer in question, a number N (N can be chosen as 200) was performed from random order computer simulations of batteries. The corresponding mean volumetric efficiency was recorded together with the value of the paraometer. The parameter can then be varied within a previously selected range, and the package file will be reinitialized. N other computer simulations will be performed for the new value of the parameter. Finally, a relationship is obtained between a set of mean volumetric efficiencies of the pile with respect to a set of values of the paraometers. A topical parameter search diagram is shown in Figure 84. The best paraometer is selected as the one that produces the maximum mean volumetric efficiency.
E) Execution of the Multiple Loop Decision Process
This is the overall process that actually searches through all the placements and finds the best solution.
As described above, each pack in the damper, and the selected allowable orientations of each pack are given an opportunity to visit each corner in the stack. Then, at a given corner, the possible offset and flush of the boundaries are attempted to see if the package can have a stable placement in the corner. A stable placement was compared to a previously recorded placement. If the current placement is better than the recorded placement, the current placement will be recorded instead. At the end of the search loop, the recorded placement was selected as the best placement.
1) Varying of the Nesting of Loops
The selection sequence between allowable orientations, corners and bundles in Figure 3 can be modified as shown in Figures 26A-26E. In Figure 26A, the inner loop of the loop is applied to the corners, the middle loop is applied to the orientations, and the outer loop is applied to the packages. In Figure 26B, the loops are, respectively, orientations, bundles, and corners. In Figure 26C, the 3 loops are, respectively, bundles, orientations, and corners. In Figure 26D, the 3 loops are, respectively, corners, bundles, and orientations. In Figure 26E, the 3 loops are, respectively, bundles, corners, orientations.
2) Shortcuts
In the search, the allowable orientations are arranged in such a way that the orientations with the smallest height dimensions will come first, and those with the highest height dimensions will come last. Similarly, in the corner layout, the corners with the lowest surface height will come first, and those with the highest surface height will come last. As a package tends to be more stable when its shortest dimension is positioned as its height, and is in a lower corner, the search loop can end as long as a satisfactory placement is found, such as a box that is level with its packages. side contiguous.
In addition, when a packet displacement is carried out, the greater the displacement, the wider the gap in the limit will be, therefore the quality of the placement tends to deteriorate. For this reason, during one round of the placement search, as long as a satisfactory placement is found, subsequent searches do not need to make use of additional offsets.
2) Picking up the Package with the Gripper
As described above, with reference to Figure 1, according to the present invention, a gripper 17 is used to pick up and move packages such as P. This gripper may be such as those disclosed in the art or it can be as described in this subsection referring to Figures 85-92.
IS 2 180 076 T3
The clamp 17 shown in Figures 85-92 includes, generally, a hollow mounting column 1005 to the bottom of which a pair of suction cups 1007 is attached. Attached to the top of the column is a suction valve assembly 1010. A vacuum line 1011 extends from the valve 1010 to a vacuum pump (not shown). Another vacuum line (not shown) connects the valve 1010 with the suction cups 1007 through the hollow interior of the column 1005. The valve assembly 1010 was mounted on the end of a portal arm 1014. A pair of scissor actuators 1015 are functionally mounted on column 1005 so as to extend toward opposite sides of the column, and one of a pair of clamp plates 1018 is attached to each end of scissor actuators 1015. Referring to Figure 87, column 1005 also supports a lathe assembly 1020 for operating scissor actuators 1015. These auxiliary assemblies and their parts will be described in more detail below.
Figure 87 shows the components included in the support column 1005 in more detail, with the items detached to show additional detail. Suction cups 1007 are spaced apart and mounted on a crossbar 1025 attached to the bottom of column 1005. Each suction cup includes a rectangular rubber sleeve on a steel foot that defines a mesh-covered opening (not shown ) to prevent dirt from entering the vacuum system. In the space between the two suction cups 1007, a contact sensor 1027 or microswitch is positioned on the crossbar 1025. Also pivotally attached to the crossbar is an actuator arm 1028 that swings to activate the touch sensor 1027. when the arm 1028 is raised by touching a package during the downward movement of the gantry arm 1014. Sensor 1027 and arm 1028 extend below the level of suction cups 1007. In addition, a set screw 1029 allows the position of the arm 1028 to be adjusted so that the distance between the suction cups and the contacted surface at which the touch sensor 1027 is activated can be set at a predetermined distance, for example, medium. inch.
As shown in Figure 87, each lower end 1031 of each scissor actuator 1015 is pivotally connected via a hinge pin 1032 to an end bracket 1033 that was attached to one of the side clamp plates 1018. Further up on the side clamp plates a ground shaft 1035 was mounted vertically and was separated from the clamp plate by a pair of shaft mounts 1036 at the ends of shaft 1035. An end bar 1038 extending transversely includes a linear bearing 1040 across the center thereof, the bearing being fitted on shaft 1035 to allow vertical movement of rod 1038 along shaft 1035. At each end of the two end bars 1038, a hinge pin 1041 extends laterally and receives an upper end 1043 of each scissor actuator 1015. A lower center joint 1045 of each scissor actuator was pivotally mounted on a hinge pin 1047 extending from each end of crossbar 1025. An upper center joint 1050 was pivotally attached to a support bearing 1052 slidably mounted on a vertical ground shaft 1054 that was attached to column 1005 by shaft mounting blocks 1055 at each of its ends. Thus, it was noted that the tops of the scissor actuators 1015 are slidably movable up and down along the column 1005 and the side clamp plates 1018, and the inward and outward movement of the scissor actuators was synchronized with respect to the center, which is attached to the column.
Tension springs 1058 are stretched between each lower hinge pin 1032 and end bars 1038. These springs force the scissor actuators into an extended position, ie, a release position of the clamp side plates 1018. The side clamp plates are brought together to engage a package by lathe assembly 1020. Mounted on the side of the column 1005 was a servo motor 1060 of the type including a built-in encoder and a brake. Attached to the motor drive extender shaft 1060 is a spool 1062. A cable 1065 wound on the spool passes over a pulley 1063 and a pulley 1064 and is attached to one of the clamping side plates 1018. As motor 1060 rotates spool 1062 to pull cable 1065, the force on the side clamp plate bends the scissor actuators 1015 against the force of the springs 1058 until the clamp plates 1018 engage the sides. of the package. The inner surfaces of the clamping plates are fitted with 1066 abrasive strips to provide high friction between the clamping plates and the sides of the pack.
The clamp mechanism 17 is controlled by a control circuit 1070 shown in Figure 88. A controller 1071 may be a standalone processor, although preferably it is the same computer described above with reference to Figure 1. The controller receives input signals from the contact sensor 1027 and distance sensors 1068. In response to these signals, the controller sends control signals to the vacuum suction valve 1010 and to the reversible lathe motor 1020. When
ES 2 180 076 T3 the suction is activated, the suction cups 1007 grip a surface that is very close. As the lathe motor rotates, the cable 1065 either looses (opening the clamping side plates 1018) or coils (closing the clamping side plates) depending on the direction of rotation.
The controller causes the gripper to be positioned very close to the top surface of the package following a series of steps that are represented in a diagram in Figure 89. Initially, the gripper is positioned vertically over the package, and the distance with With respect to the top surface it is monitored at block 1080, while the pointer arm 1014 moves the gripper toward the package at block 1081. During this movement, at block 1082 the side tie plates 1018 are positioned wider apart than the package by approximately 1.5 inches, the dimensions of which have been determined. In particular, the nominal height of the package, measured by the sizing arc 14 described above, is set as a target position in the frame arm reference frame at block 1083. However, this height represents the peak point of the upper surface, and may not be the height of the part to which the suction cups 1007 are to be attached. For example, the measured height may be at one edge of the package, while the central portion of the top surface may be recessed a significant distance away. It is rare for a top surface to have a depression greater than half an inch.
As described above, the contact sensor is set so that it is activated at a predetermined distance from the top surface of said package, preferably half an inch. Figure 91A schematically shows the position of this sensor approaching the package. When the controller receives the contact signal, block 1084 stores the z1 position of the suction cups 1007 in the reference frame. Figure 91B shows the clamp position where the sensor is triggered. The controller continued to lower the gripper until the suction cups reached the target position z2 (measured package height), which is stored in block 1085. However, if in block 1086 it is determined that the touch sensor 1027 has not been triggered by the time the target position z2 is reached, in block 1087 the controller lowers the target an additional distance d equal to the distance default (preferably half an inch), and returns to block 1084. The controller continued to lower the clamp until the touch sensor was triggered. Figure 91C shows the suction cups 1007 at the target position z2.
When the touch sensor 1027 has tripped at position z2, the controller calculates in block 1088 the actual gap e between the suction cups and the top surface of the package:
e = d - (z2 - z1)
At block 1089, e is compared to a predetermined tolerance, preferably one eighth of an inch. If e is not less than the tolerance, the clamp is lowered a final distance increment equal to e. Next at block 1091, the side clamp plates 1018 are moved to a position where only about one eighth of an inch clearance is left on each side of the package. If in block 1089 e is less than the tolerance, the process goes directly to block 1091. In Figure 91D the final position of the suction cups 1007 is shown. Next at block 1092 the controller operates the vacuum valve 1010 so that it grabs the package by suction, and immediately afterwards at block 1093 the side clamp plates 1018 are made to approach each other a predetermined distance (preferably approximately three-quarters of an inch) to apply clamp pressure to the sides of the package. Alternatively, the lathe motor 1020 can be configured to apply a constant torque to cause the clamp plates to apply an approximately constant force to the package, proportional to the weight of the package.
Those skilled in the art will understand that the steps shown in Figure 89 can be carried out very quickly and, in some cases, at the same time. For example, movement of the side clamp plates 1018 inward can occur as the clamp is being lowered. In addition, the gate arm 1014 may begin to lift the package when the side clamp plates are moving to apply pressure.
After the package has been placed by the pole arm, as described elsewhere in this application, the winch motor 1020 is operated in a manner that extends the cable 1065 sufficiently to allow the springs 1058 Open the hold-down plates 1018 only about one eighth of an inch on each side beyond the measured package dimension. This prevents contiguous packages stacked on the pallet from being knocked over. Again, to eliminate delay, the pack release motion and the pole arm lift motion can be
ES 2 180 076 T3 produce at the same time.
The touch sensor 1027 also functions in a manner that monitors the presence of the package during the transfer by the gantry arm 1014. If for any reason the package is dropped, the sensor input signal will no longer be present. The controller noticed the absence of the signal and the gantry arm was guided to a stop.
Preferably the clamp mechanism 1000 is designed so that it has a very small footprint, eg, seven by seven inches, when the scissor actuator 1015 is fully retracted. The gripper can be used for side stacking as well as vertical stacking. From the above it will be observed that the gripper can lift packages of various sizes. For a parcel transport operation, the suction force and clamping side plates can be designed to handle parcels up to or above 32 by 32 inches and 150 pounds. The top surface of the packages need not be flat or even as long as the suction cups can be tightly closed enough to catch the package.
A modified gripper mechanism 1100 is shown in Figure 92. The main difference from the embodiment described above is the inclusion of a sliding mechanism 1102 that allows the clamping side plates 1018 to be pushed back on the suction cups 1007 when the suction cups can reliably lift a package without the aid. of the clamping plates. The advantage of this feature is that the side clamp plates will not interfere with adjacent packages in the stack of packages. In the embodiment shown in Figure 87, the upper center joint 1050 of the scissor actuator 1015 is still pivotally mounted on the backing bearing 1052, although the shaft mounts 1055 holding the ground shaft 1054 are attached to the top. from a 1104 carriage instead of column 1005. At the lower end of carriage 1104, a hinge pin 1105 is mounted to receive the lower center joint 1045 of scissor actuator 1015. Thus, expansion and contraction of the scissor actuator along the vertical axis takes place in relation to carriage 1104.
Carriage 1104 also includes a pair of bearings 1106 extending toward column 10 05. These bearings fit a vertical ground shaft 1108 that was secured to column 1005 by means of a pair of shaft mounts 1109. A solenoid actuator 1112 is mounted on column 1005 on carriage 1104, with its piston rod attached to the top of the carriage. As the solenoid bar is extended, carriage 1104 moves down along axis 1108, moving side clamp plates 1018 into the proper position (shown in dashed lines) to grab a package. When the solenoid bar is retracted, the clamp plates are raised to the position shown in Figure 92, outside the zone of activity of the suction cups 1007. Those skilled in the art will observe that the solenoid actuator could be replaced by a pneumaotic or hydraulic actuator.
Referring now to Figures 93-96, a feature according to the present invention is disclosed that allows compensation for errors to be obtained during the "pick-up" stage of the clamp. This characteristic allows to identify the "central position" of the package.
As shown in Figure 93, packages "A", "B", "C" and "D" are in linear contact on an accumulating conveyor beginning at an origin point O. The last package on the conveyor is on the path of a distance sensor S, which is located at an angle Θ with respect to the x-axis, which is the axis of travel of the conveyor and the axis along which, from now on, all distances are measured with with respect to the origin O. By detecting the distance "d" of packet D with respect to the sensor "S" and calculating the cosine of Θ, the component "x" of the distance "d" can be determined. Such information combined with the known dimensions of the accumulator conveyor (namely, the distance from the origin to the sensor S) allow the calculation of the True Line Length of the packages on the accumulator conveyor. The knowledge of the nominal lengths of the individual packages, measured by the sizing conveyor located above according to the direction of advance, allows the calculation of the Nominal Length of the Line of packages A, B, C and D. The difference between these two values is the Total Error of Line "e". As can be understood, such an error, if left untreated, can cause problems later in the stacking process.
This error is compared with thresholds, such as +1.2 inches and -1.5 inches, and if the error exceeds any of these thresholds, it is assigned the threshold value in question.
At this time, the system is ready for the removal of another package, and for this reason it is at the beginning of its cycle. Then, as shown in Figure 94, the lone can be removed
ES 2 180 076 T3 a package, such as "B", and another package "E" can be placed in linear contact with the other packages. The error "e" is then recalculated, and another packet (eg packet "D) is removed and replaced by a packet such as F to obtain the configuration shown in Figure 95.
As described above, after calculating each "e" error, the gripper then removes a package from the group of accumulated packages on the accumulator conveyor. It is desirable to know as precisely as possible the center (at least in the "x" direction) of the packages in order to achieve a placement of the collected packages that is consistent in positions located further apart according to the direction of advance. For this reason the error “e” is proportional to the individual packets in the lines as shown in Figure 96, where:
e = total error i = packet number (6 in Figure 96)
N = total no. Of packets (7 in Figure 96) q = individual estimate of packet error <sub>what</sub> = <sup>(i-1) e</sup><sup>what</sup> = TN-ry
As can be understood, the above calculation basically "spreads" the total error "e" proportionally among the packets located between the first and the last packets in the line. The individual packet error q is subtracted from the Normal Distance to the center of the packet F to provide the x-coordinate to be used by the gripper.
As an example, consider that, as shown in Figure 96, there are seven packages on the accumulator conveyor, and a gripper has selected the sixth package for removal. Consider that the total error e is 5 inches. In this way, the calculation referred to above was carried out as follows:
q = <sub>N</sub>'= = 4.17 inches
Another characteristic of the invention is the use of a "weighting" system to calculate the error "e" shown in Figure 93. As described in relation to Figures 93-95, during each operating cycle a total error "e". In the case of permanent rheometer operation, three consecutive previous errors can be used to calculate a "weighted" e 'error. This e 'is calculated using the following formula:
e '= 0.2 (e [t-2]) + 0.3 (e [t-1]) + 0.5 (e [t]) where e [t] = error corresponding to the error of the current cycle e [t-1] = error corresponding to the previous cycle e [t-2] = error corresponding to the cycle prior to the previous cycle
As can be seen, this formula assigns more weight to the most recent error, but at the same time provides some weight, although less, to the previous errors.
For the true first withdrawal operation immediately after startup, the errors e [t - 2] and e [t - 1] will not be available, and therefore in this case the error value e will not be weighted. For the second withdrawal operation immediately after startup the error e [t - 2] will not be available, but the weighted error e 'can be obtained by means of the formula:
0.4e [t-1] + 0.6e [t]
3) Effective Planning of the Placement and Collection Path
According to one aspect of the present invention, special attention is paid to the planning of the “placement path” which is the route that the robotic gripper takes when it moves a package taken from the accumulator to the palóe (s), (see In general, Figure 0700.4) as well as the planning of the “collection path”, which is the route that the gripper takes after it releases the package and will pick up a second package designated to be collected (see Figure 0700.5 in general). .
The near minimum distance path planning method is implemented according to the present invention with the use of the hardware described above (for example, the accumulator conveyor
ES 2 180 076 T3 and the clamp) in conjunction with a software module, such as one written in "C" language. This module communicates with other software modules as shown with reference 970 in Figure 97. As described in another section in this application, other stacking 971 algorithms according to the present description decide which packet to collect from the accumulator buffer. and where to put it on the pallet. The module 971 for planning the (almost) minimum distance path according to the present description determines a set of positions "passing" of the almost minimum distance (described in detail later), along which an approximation of a path of minimum distance. Module 972 "path planning" (see Figure 97) adapts curves using these step positions. The planned path is fed to commercial servo control boards 974 (such as a Galil 530 series) to control a gripper robot as described elsewhere in this application.
As indicated above, road planning includes both a “laying path” planning, which goes from the shock absorber pick-up position to the stacking position, and a “pick-up path” planning, which goes from the previous placed position to the accumulator pick-up position.
a) Rectangular Base Modeling that Accepts Twists
In the system according to the present invention, a rectangular pattern can be used to reach all the locations of the packages, whether they are packages standing or moving.
In relation to the stationary packages, as shown in Figure 2, in a system according to an embodiment of the present invention, there may be one or more pallets 11 for stacking, and an accumulator buffer conveyor such as 16. The edges of the pallets 11 and the accumulator conveyor 16 are fixed so that they are parallel or perpendicular to each other. Furthermore, for all packages on the accumulator buffer or on the pallets, their edges will be parallel or perpendicular to those corresponding to the pallets and the accumulator conveyor. In this way it can be seen that the stationary packages, whether they are located in the accumulator buffer or on the pallets, are easily modeled using discrete rectangular objects, which can be analyzed in relation to the intersection or relative separation as known in the art. or as shown in this application.
In relation to moving objects, as shown below in Figure 98, the modeling system according to the present invention includes the use of a Circumscribed Rectangular Object 982 to represent a rotating object 982 in motion.
As described above, the pack (and corresponding clamp 17) can be rotated from its original orientation on the accumulator conveyor to a second orientation on a pallet. Similarly, the gripper can rotate while empty when moved to a selected package for placement. During said rotation, the edges of the moving object are generally neither parallel nor perpendicular to those of the pallets, accumulator conveyor, or stationary packages. For this reason, a circumscribed rectangular object is constructed in the model that circumscribes the entire rotating object, be it the empty gripper or a specific caught package. For this reason, according to the present invention, the edges of the circumscribed rectangular object will be parallel or perpendicular to the limotropic edges of the poles.
Figure 98 shows an example of plane rotation in which the moving object rotates about a vertical axis. As can be understood, during the rotation of package 981, the size of a Circumscribed Rectangular Object 982 will vary for each position along the path of travel, as the angle between the rectangular shapes will change.
According to the present invention, use is also made of an Expanded Rectangular Object 983 in Motion. The rectangular object 981 in motion, whether the original package boundaries of a package undergoing a simple translation or the Circumscribed Rectangular Object of a package undergoing a rotation, was scaled up to the size of the Expanded Rectangular Object in Motion according to a tolerance of collision previously specified in each horizontal dimension. As noted above, the Moving Expanded Rectangular Object will also vary with position if turns occur.
Referring below also to Figure 110, according to the present invention further use is made of a Combined Rectangular Object in Motion for a segment of motion. The Moving Combined Rectangular Object can be roughly calculated as a rec41 object
ES 2 180 076 T3 minimum tangular that can be adapted to Expanded Rectangular Objects in Motion in the limiter positions of the motion segment.
In the following description, the size of the moving object is calculated based on either an enlarged rectangular moving object (when referring to a fixed position) or a combined rectangular moving object (when referring to a moving segment). .
Following the model referenced above, a control point can be used as a reference point to designate the position of a moving package or a returning gripper. Once the position of the control point, and the angle of rotation of the moving object have been specified, the location of any other point on the moving object is fully specified. Unless otherwise stated, in the following description, a position is referred to as the spatial location of this control point.
Reference is made generically below to Figures 99-103. During the laying path (see in particular Figure 100), a central point on the bottom surface of the moving package is chosen as the control point. During the collection path (see Figure 101), a central point on the lower surface of the gripper is chosen as a control point (such as a central point on the suction cup). In fact, any other suitable point on the moving object can be chosen as the control point. The invented methods are not limited to a specified choice of control point.
Following an embodiment of the present invention, the placement path step positions and the collection path path positions will be stored in two independent lists of data. For each list, the contained step positions will be ordered based on the ascending horizontal distance (from furthest near to furthest away) from their initial step position. The calculated horizontal distances will be recorded along each position. Each list of positions will be "convex carrier"; that is, the list will correspond to a convex path of step positions. During the tour, the checkpoint will follow a path of almost minimal distance. This path is arranged in a vertical plane, passing through a convex set of passage positions.
The turn, if it exists, will start at the starting position of step, and will end at the final position of step. The amount of rotation will be linearly proportional to the horizontal travel distance with respect to the initial step position. The calculated pitch angle of rotation will be added to the corresponding pitch position.
It should be understood that the aforementioned distribution that allows the angle of rotation to be linearly proportional to the horizontal travel distance is established to obtain greater comfort in the calculation. The present invention also contemplates that the angle of rotation can be linearly proportional to the distance of spatial travel.
b) Layout Path Configuration
Reference is now made to Figure 100, which illustrates a package being moved along a placement path. In the model according to the present invention, the positioning path is composed of a series of different positions, including a pick-up position 986, a starting position 987 for step, a position 988 for the exit passage of the shock absorber, an end position 989 passing through, and a discharge position 990, all of these positions referring to the control point of the moving package. Other step positions can be added later as described in detail in later paragraphs. On a full path of movement, the package to be placed will start at pick-up position 986, follow all waypoints, and finally reach unload position 990. Following an embodiment of the present invention, all step positions are kept in a vertical plane.
The pickup position 986 is the position where the package rests on the package buffer (also referred to as an accumulator), ready for stacking. The starting pitch position 987 is above the pickup position a distance known as the "starting height from the shock" which is variable as described later.
The "damper exit position" 988 corresponds to the horizontal limiter position at the instant the moving pack leaves the damper. The height of the damper exit position 988 is the same as the starting step position 987. If the exit position 103 of the
ES 2 180 076 T3 damper is above a constructed straight line that passes through the starting position 987 of pitch and the end position 989 of pitch (which is the case in Figure 100), then the exit position of the damper will be added to the ordered positions.
The “final step position” 989 was near and above the “discharge position” 990. The final step position 989 is different from the discharge position 990 horizontally by a previously specified separation, and vertically by a distance denominated "height of approach to the palóe" 992.
As described in detail below, intermediate passage positions are inserted to avoid collision with any already stacked packages.
c) Configuration of the Collection Path
As described above, a collection path as defined in the present invention is defined as the path in which the gripper moves while it is "empty", from a previous position of placement above a pallet to a position shock absorber pickup.
Reference is now made to Figure 101, which illustrates an empty gripper carried along a collection path. In the model according to the present invention, the collection path was made up of a series of different positions, including the original 993 “positioning position” in which the clamp released its last pack, an “initial step position” 994, "A damper entry passage position" 995, a "final passage position" 997, and a "pick-up position" 997, all positions being referenced, as described above, to the control point of the moving package. Later, other step positions can be added to avoid collision as described in detail in later paragraphs. On a full path of motion, the gripper began at the 993 drop position, followed all the pitch positions, and finally reached the 997 pickup position. Following an embodiment of the present invention; all passage positions are kept in a vertical plane.
As shown in Figure 101, after a package is placed, the starting position 994 is at a location that is arranged above the placement position at a distance known as the “starting height from the pallet” 998 .
As shown in Figure 101, the entry passage position 995 in the damper corresponds to a horizontal limotropic position in which the moving gripper enters the damper. The height of the damper entry position 995 is the same as the step end position 996. The final step position was at a location that is available above the pick-up position by a distance known as the “shock approach height” 999. If the entry position to the damper is above a constructed line that passes through the initial step position and the final step position (which is the case in Figure 101), then the entry position to the damper will be added. to the list of ordered positions.
d) General Passage Positions
As shown in Figure 1, a package P is positioned among other packages similarly positioned on top of a pallet 11. For the purposes of this description, as shown in Figures 99A-99D, the initial position 987 is shown. step of package P, and the final position 989 of step of the box. After having established these two positions 987 and 989, the method according to the present invention involves a determination of the preferred intermediate positions 987 and 989 of the path. In the first place, the processing computer determines all the packages that are in intersection with a straight path CR that begins in the first position 987 and ends in the final position 989, the path presenting a width equal to that corresponding to the package that was being moved. . This determination may include the concepts of pallet collision checking and packet collision checking as detailed below.
For the purposes of this description, packages X, Y and Z were considered to intersect with said straight path. Each package was then reviewed in relation to the possible creation of a pass-through position. For this description it was considered that packet C was deposited before packet B, which was deposited before packet A. Thus, according to an embodiment of the present invention, packet C was first checked. According to an embodiment of the present invention, two "passage positions" are established which, if added to the list of "passage positions", would allow the package to
IS 2 180 076 T3
P hover over the package being evaluated, so that the package will leave a space with respect to the stopped package that will be within a specified tolerance (the difference between the Expanded Rectangular Object 983 in Motion and the Circumscribed Rectangular Object 982). Next, following the present invention, an "Ordering of Positions" is carried out.
Following an embodiment of the present invention, all passing positions are established prior to the ordering of the positions. Next, each proposed step position is evaluated in relation to convexity.
e) Starting height from the shock absorber or approach height to the shock absorber
Referring now to Figures 100 and 101, the measurements of the starting height 991 from the shock absorber and the approach height 999 to the shock absorber (which dictate, respectively, the heights of the shock output position 988 and of the shock absorber. 995 input position on the damper) depend on the heights of the accumulated packages on the accumulator conveyor, and they are preferably fixed in such a way that there is no interference between the path of the moving object and any accumulated packets on the accumulator conveyor. In other words, these two distances define the minimum height at which the moving object will approach or depart from the accumulating conveyor.
As indicated above, the size of the moving object is calculated as corresponding to a combined rectangular object for a segment of movement between the limiting position of the shock absorber and the pick-up position. To check for such interference, a search is made of the maximum height of the upper surfaces of the packages in the accumulator buffer that are within the swept area of the moving object. Calculation methods are described with reference to Figures 102A-102E.
If the moving object is moving in a direction perpendicular to the accumulator, as shown in Figure 102A, then the approach height to the shock absorber is taken as the height of the collected package, and the starting height from the shock absorber is taken as the maximum height of the two contiguous packages. The latter can also comprise more contiguous packages, located before they will follow the direction of advance, in the accumulator, if these are overturned to fill the gap of the collected package immediately after it has been lifted. If said adjoining "gap filler" packages in said gap are taller than the collected package, then the maximum height of said packages will be taken as the lift height from the damper.
If the moving object moves diagonally from its position on the accumulator conveyor, as shown in Figures 102B to 102E, a search will be carried out in the accumulator to determine if the moving object will collide with the located packages. in the accumulator. If there is a collision, then the object is lifted in such a way as to avoid collision.
f) Starting height from the palóe or approach height to the palóe
Referring generically again to Figures 100 and 101, each of the starting height 994 from the pallet and the pallet approach height 992 is determined based on the possible collision with contiguous packages on the pallet. A check is made of each package on the pallet. The maximum height of the upper surfaces of the packages that may collide with the moving object will be taken as the starting height from the pallet or the approach height to the pallet.
g) Search for Collision with a U<sup>or</sup>nico Palóe
To search for a specific paláe in relation to a possible collision with a moving object, the entry position and exit position of the object are defined at the boundaries of the paláe. The entry position is a limiting position in which the moving object begins to have a horizontal intersection with the edge of the pallet, and the exit position is a limiting position in which the moving object has almost left the pallet. completely.
Generally, the moving object will have an entry position as well as an exit position at the limits of the pallet when a moving object intersects the limits of the pallet. There are two exceptions. For the collection path, the entry position does not exist for the “origin” paláe (the paláe that just received a new package). Similarly, for the placement path, the exit position does not exist for the “destination” paláe (the paláe that is about to receive
ES 2 180 076 T3 a pack). Except for the original pallet in the planning of the collection path, in general if an entry position for a pallet cannot be found, the moving object will not have a horizontal overlap with the pallet, and the pallet is prevented from colliding. with the moving object.
The heights of the moving object in the entry position and the exit position are calculated, if such positions exist. These heights can be obtained from a linear interpolation along a line that passes through the initial step position and the final step position. The lower height between these two boundary positions of the pallet will be taken as the height threshold (lower). If a boundary position is lost, in that case, instead of the comparison, the starting height or the approach height will be used. If the maximum height of the top surfaces of the packages stacked on the pallet is less than the threshold height, the pallet is prevented from colliding with the moving object. The maximum height of the top surfaces of packages stacked on a pallet is recorded whenever a new package is stacked on a pallet.
When a spin occurs, the two calculations above can involve iteration. Initially, the size of the moving object is taken as corresponding to the combined moving object for a movement path between the initial step position and the final step position. Using this information, a set of entry points and exit points is calculated. A smaller size can then be obtained by relying on the combined moving object for a movement path between the entry and exit positions. The new moving object can be used to achieve a more accurate threshold height.
If the boundaries of the pallet are at intersection with the moving object, then a search is carried out for each package in the pallet so that a check is made for a possible collision with the moving object whose size has been determined previously. For each package already placed on the pallet, there can also be an entry position and an exit position. If there is an entry position, in that case the entry position will be added to the list of positions on the path of the tour. An exception occurs in the case that a placement path is close to the destination position. If the collision position goes beyond the final step position along the path of the route, then the collision position may be false since the moving object expands with tolerance. In such a situation, the correct shape of the object (instead of the Expanded Rectangular Moving Object) will be used to perform the collision check.
For a given stacked package, if there is an entry position, there will usually be an exit position. When considering a pick-up path with the package in close proximity to the home location, even without an entry location, there may still be an exit location. When a starting position is found, the starting position will also be added to the list of course positions.
h) Purchase of Collision with Packages
The individual collision check is made up of two independent checks along two orthogonal dimensions. The two orthogonal dimensions are the front / back dimension and the left / right dimensions. It can be understood that if the moving object collides with the stacked package, the collision must occur either in the front / rear dimension, or in the left / right dimension (Figure 103B), but not both.
Reference is now made to Figures 103A-B. For collision along the front / back dimension, the moving object can be in two boundary positions as shown in Figure 103A. These boundary positions can be either the entry position or the exit position depending on the direction of travel. Based on the geometrical sizes of the package and the moving object, the coordinate along the left / right dimension for the control point of the moving object can be calculated. As shown in Figure 103, the z axis will be defined as the front / back direction, and the x axis along the left / right direction. z1 and z2 are the corresponding front and back z coordinates of the packet. The z coordinate of the control point (center of the bottom surface) of the moving object will be i (z1-w) when the moving object is on the bottom side, and z2 when the moving straight angle is on the top side. The corresponding x-coordinate of the control point can be solved from an equation of the straight line of the path of travel. Once the control point coordinates have been obtained, the controller can calculate the two boundary x-coordinates along the contact lines in the x direction. If there is any length of overlap along the horizontal direction between the moving rectangle and the packet, the collision can be considered to exist.
IS 2 180 076 T3
For collision along the left / right dimension, the moving box can be in two limotropic positions as shown in Figure 103B. Each limotropic position can be either an entry position or an exit position depending on the direction of travel. Based on the geometric sizes of the packages and the moving object, the horizontal coordinates of the control point of the moving object can be calculated. Using the same coordinate system as above, x1 and x2 are the corresponding x coordinate of the package. The x coordinate of the defined control point will be (x1-1 / 2) when the moving object is on the left side, and (x2 + 1/2) when the moving object is on the right side. The corresponding z coordinate of the control point can be solved from an equation of the straight line of the path of travel. Once the control point coordinates have been obtained, the controller can calculate the limotrophic z coordinates along the contact line. If there is any length of overlap along the vertical direction between the moving object and the package, then the collision exists.
i) Package Insertion without Collision
Initially, there are only two positions in the list, the initial step position and the final step position. The input position or the output position of the damper can be added as described above. Subsequently, for each insertion of a position, a consistency check was carried out to ensure that the ordered list maintains “convexity”. In other words, the upward convex nature of the path is maintained when new positions are added to the list.
j) Checking for Adjacent Packages
Whenever a position is entered, an adjacent pair of positions will be searched in the corresponding list, one being immediately before and one immediately after based on the distance of the horizontal travel.
If the new position was within a predefined tolerance for one of the existing positions in the pair in terms of horizontal travel distance, it was dealt with in two ways. If the new position is lower than the existing position, the new position will be discarded. Otherwise, the existing height was raised to the new height, and, as described later, a forward search and a reverse search would be performed.
If the position is outside the predefined tolerance, a line will be built that passes through the pair of existing positions. If the height of the new position is less than or equal to that lon, the new position will be discarded, otherwise the new position will be inserted, and a forward search and a backward search will be performed.
k) Search Forward
With the newly inserted position as the starting position, the next position is selected as the check position, and the position following the next position, if any, is selected as the reference position. After constructing a line that passes through the start position and the reference position, a check is made as to whether the check position is above the line, below the line, or in the line. If it was above the lone, the search ends. Otherwise, the check item is removed from the list. The reference position is then re-labeled as the check position, and the next position, if it exists, is re-labeled as the new reference position. A new testing cycle has started. This was repeated until convexity was met for all positions along the forward direction.
l) Search towards Atróas
With the newly inserted position as the start position, the previous position is selected as the check position, and the position that comes before the previous position, if it exists, is selected as the reference position. After constructing a line that passes through the starting position and the reference position, a check is made as to whether the check position was above the line, below the line, or in the line. If it was above the lone, the search ends. Otherwise, the check item will be removed from the list. The reference position is then re-labeled as the check position, and the position that comes before that position, if applicable.
ES 2 180 076 T3 that exists, is relabelled as the reference position, and the same cycle is started again. This was repeated until the convexity was met for all positions along the backward direction.
m) Vertical Tolerance
Each position in an ordered list was complemented with a tolerance. This is equivalent to lifting by a specified tolerance distance all segments of the convex polygon that join the positions in the list, such that the moving object has the least probability of colliding with stacked packages or packages in the accumulator buffer. This can also be achieved by widening the moving rectangle according to a collision tolerance on the lower surface.
E. Effectiveness of Synchronization
The software system includes the following components: robot motion planning software, robot path interpolation software, peripheral device control software, error handling and message display software, and stacking algorithms. The robot motion planning software schedules a pick and drop motion sequence based on a package stack schedule, and schedules each segment of the motion. The path interpolation software executes a segment of planned motion and interpolates a new position each cycle of the path to communicate to the servo boards. The peripheral device control software coordinates the movement of the feeding belt, the measuring station and the accumulator. The stacking algorithm is computer software that, as described above, generates a packet stack schedule. Stack planning dictates which packet to pick from the packet accumulator, where to place the packet on the pallet, and which packet orientation to use.
All components according to an embodiment of the invention are controlled by means of a controller of the palletizing system based on a VME bus. Following this realization, the main processor board used is a Motorola MVME167. The real-time multitasking operating system used is pSOS +. Control software for belt feeder, measuring station, accumulator, and robot, and stacking algorithms can all be run on this board. This main processor communicates with Galil servo boards to control the gantry robot, measurement station, feeder belt, and accumulator. It can also communicate with an I / O Matrix board to collect data from load cells, which reside at the measurement station.
1. Robot Movement Planning
During a pick and place motion cycle, the robot is controlled with maximum acceleration and deceleration throughout all motion segments. The motion profile contains a mixture of square wave acceleration / deceleration and sine wave acceleration / deceleration, both well known in the field of control. Given a maximum acceleration, the square wave profile can achieve faster motion than the sinusoidal profile. On the other hand, the sine wave profile caused less vibration than the square wave profile. During a pick and place cycle, the gantry followed a square wave profile all the time except for the last segment of the placement cycle. In that segment, there is a commutation to a sinusoidal profile. This method achieves a short cycle time, while preserving the precision of placement.
two. Synchronization arrangement
In the initial configuration of the system, in the first place the processes of the measurement station and the accumulator supply will be executed. After the accumulator has been filled, from now on, the processes of measuring the packages and feeding will be executed in parallel to the movement of the robot.
As soon as the robot picks up a pack from the accumulator and avoids any collision with the neighboring packs, the cycle of feeding the packs starts. The feeding of the packages continued until the accumulator was full; that is, the last packet will partially end at the measurement station.
According to the present invention, the accumulator feeding cycle time is kept less than or equal to half the robot pick-up and placement cycle time. When the robot returns
ES 2 180 076 T3 to pick up another pack, the controller must check if the accumulator has been powered before falling on that pack. If the accumulator has not been powered, the robot will remain idle, and the stacking cycle will be delayed. When a relatively large pack is removed from the accumulator, the gap will typically be filled with one or two packets. Occasionally, filling the gap requires three packets, if they are small; however, this situation is unlikely.
The next pack is then fed into the trailing edge of the induction tape. In a preferred embodiment, a photoelectric cell is installed in the induction tape which acts as a deceleration sensor. The location of the deceleration sensor is positioned such that when a package begins to decelerate from that point, it will stop at the trailing edge of the induction belt. The computer will monitor the deceleration sensor, and will stop the induction belt when a new package arrives. In this arrangement, as long as the measurement of a previous package has been completed, there will be a minimum route to move the next package over the measurement station.
3. Planning a Package in Advance
Following the present invention, the stacking decision making is executed in parallel to the robot movement and the measurement and feeding of the packages. Stacking decision making prepares a stack plan regarding which package to pick from the shock, where to place it on the pallet, as well as what orientation to use. As soon as the robot motion planner accepts the stack schedule, the stacking decision maker starts calculating for a next placement.
Stacking decision making uses a pre-packaged one-pack geometry model for the accumulator damper and pallet. Physically, the robot gripper has just finished placing the last pack, and has just started moving toward the damper to pick up a planned pack. However, in the One Package Ahead model, the planned package is considered to have been removed from the shock, and must have been placed on the pallet. The shifting of packet positions in the accumulator is also considered to have ended. Based on this one-stage model in advance, the next stack schedule is calculated.
In this way, when a robot finishes placing the current package, the selection of the next package has already been prepared, so that there is no time delay.
Four. Multitask
As can be seen, according to the present invention, a multitude of different tasks are being carried out, which can pose difficulties in a real-time operating system such as the one used. As can be understood, it can be disadvantageous if the system must "wait" for calculations to occur before allowing a collection path or a placement path to be traversed.
Following the present invention, prioritization of control tasks can effectively use computer time, thereby reducing or eliminating the delay for the cells in the stacking cycle.
Follow the present invention, as shown in Figure 106, there are 6 software tasks that can be executed on a processor: Stacking task, Planning task, Path task, Peripheral device task, Print task and Task. of errors.
The stacking task performs a planning of the stack of packages, including which package to pick from the damper, where to place it on the stack, and which orientation to use. This planning includes the stacking process described above. The planning task designs and executes the previously described sequence of movement of pick-up and placement of the gantry robot. The planning task can also be extended to the management of the user interface. The trajectories task manages the real-time interpolation of the motion segments as described in relation to the planning of the placement path and the pick-up path. Critical time-related maintenance obligations such as checking whether an E-STOP (emergency stop) button has been pressed are also implemented in this task. The peripheral devices task manages the control for the input feeder conveyor, the metering station and the accumulator. The print job sends all the messages to a monitor screen or an LED. The error task becomes active when the operator presses the E-STOP button, or if a hardware or software error occurs.
IS 2 180 076 T3
According to the present invention, an assignment of relative priorities of the tasks is carried out. CPU time is allocated based on the priority of the tasks: whenever a high priority task was ready to run, a low priority task that is running was interrupted, and that high priority task was allocated time. calculation. Each of the above tasks is assigned a different priority. The trajectory task has the highest priority, followed by the error task, the measurement station task, the planning task, the stacking task, and the printing task. As long as a specific high priority task has completed, the CPU time will immediately be shifted to a task that has a lower priority.
According to the present invention, a provision of the execution time of the tasks is provided. As shown in Layout 2000, Figure 106, a high priority task gave up computer CPU time whenever it was voluntarily idle. It remains roasted until it is explicitly activated by an external task. A high priority task also gave up computer CPU time when waiting for an event. It will remain roasted until it receives that event from another task.
As shown with reference 2010 in Figure 107, after the stacking task generates a schedule for picking up and placing packages in reference 2012, it “sent an event” to the 2013 planning task, allowing it to be set. in execution, and was left inactive in reference 2014 to give up computer time. It will resume its cycle when the schedule task has taken the schedule from the current stack.
As shown in Figure 108, the schedule task will not run unless it receives a packet stacking schedule from the stacking task. After receiving the schedule, the schedule task triggered the stacking task to extract a stacking schedule from the next package. The planning task then plans and executes the picking and placing movement sequences. Each stage in the sequences topically includes a planning of the movement segments. After planning a motion segment, the planning task will remain inactive until it is triggered by a trajectory task. Immediately before picking up an accumulator pack, the planning task waited for an event performed by the accumulator. Immediately after lifting a package, a scheduling task triggered the peripherals task to feed the next package.
As shown at 2030 in Figure 109, an interrupt clock triggered the trajectory task with a fixed interval, for example 32 milliseconds. Once activated, the path task will update the robot path positions if the current segment has not been completed, or will transition to the next planned segment. Based on the instruction (stored in the computer memory) of the planned task, the trajectory task can activate the planning task either when all the planned segments of the movement have been carried out or immediately after the transition to the next segment of the movement. movement. The trajectory task then executes various maintenance obligations, and becomes inactive.
As shown at 2040 in Figure 111, the peripheral devices task will initiate a motion program to feed and measure a package at 2041. The motion program runs on the servo board. During the execution of the program, most of the time the task of the peripheral devices is in the idle mode. When it reaches the chrome positions, it is activated by the servo board to exchange data with the servo board or to read data from the sensor at reference 2042. This process will be repeated until the accumulator is full. The peripheral devices task will then send an event to the scheduling task in reference 2044, and will remain inactive until it receives the event back from the scheduling task.
As shown in Figure 112, the print job continues to display messages at reference 2052 as long as one exists. This task runs only when the rest of the tasks are not running. If the system is going to include additional low priority tasks, the deactivation and activation method used above can be extended to the printing task.
As shown in Figure 113, normally the error task is not running. It is activated by an exceptional error to handle errors in reference 2062. After processing, it remains inactive in reference 2061.
IS 2 180 076 T3
F. Correction of Errors
As can be understood, when the packages are placed in linear contact, they may become somewhat compressed or deformed from their original dimensions that were previously detected according to the direction of advance. For this reason the actual length of the accumulated line (the actual end-to-end length of the accumulated line of packets) may differ from the sum of the "nominal lengths" of the packets measured above. For this reason, if the gripper only relied on the individual measurements above to go and pick up a specific packet, it could be “skewed” an error distance to a disadvantage if, for example, compression of the line has caused the length real of the accumulated lone is less than the sum of the nominal lengths. Reference is now made to Figures 114-117 to illustrate a method and apparatus according to the present invention for selecting individual packets from a row of accumulator packets, including error correction to compensate for the difference between packet lengths. before accumulation and their actual lengths while accumulated.
As shown in Figure 114, a group 1140 of packages "A", "B", "C" and "D" are in linear contact along their lengths on an accumulator conveyor, the first accumulated package A being positioned against a reference end stop 1142, of known location, and at least packages "A", "B", "C" and "D" in lateral contact with a lateral alignment guide 1142 to align them laterally. It should be understood that the term "length" is relative to the orientation in which the packages were placed when loaded into the accumulator. The last packet D on the conveyor is on path b of a distance sensor S, which was at a partial angle to the X-axis, which is the axis along which all distances are measured hereafter. By detecting the distance "d" from the sensor "S" and calculating the cosine of d, the "X" component of the distance "d" can be determined. This information combined with the dimensions of the accumulator conveyor allow the calculation of the Real Length of the Line of the packages on the accumulator conveyor. The knowledge of the nominal lengths of the individual packages, measured by the sizing conveyor (not shown) located above according to the direction of advance, allows the calculation of the Nominal Length of the Line of packages A, B, C and D. The difference between these two values is the Real Line Error, or total error “e”. This error is compared to thresholds, such as +1.2 inches and -1.5 inches. and if the error exceeds any of these thresholds, it is assigned the threshold value in question.
At this moment, the system is ready for the removal of a package, and for this reason it is at the beginning of its cycle. Next, as shown in Figure 115, a package, such as "B", is removed from the line, and another package "E" is placed in linear contact with the other packages. The error "e" is then recalculated and another packet (eg packet "D) is removed and replaced by a packet such as F to obtain the configuration shown in Figure 116.
As described above, after calculating each error "e", the gripper then removes a package from among the packages located on the accumulator conveyor. It is desirable to know as precisely as possible the center (at least in the "X" direction) of the packages to achieve a placement of the collected packages that results in consistent positions located beyond the direction of advance. For this reason the error “e” is proportional to the packets in the lines as shown in Figure 117, where:
e = total error (calculated according to Figure 114) i = packet number (# 6 in Figure 117)
N = total no. Of packets (a no. Of 7 in Figure 117) q = individual estimate of packet error <sub>α</sub> = <sup>(i-1) e</sup>' <sup>what</sup>= (N-1)
As can be understood, basically this calculation "spreads" proportionally the total error "e" between the packets located between the first and the last packets in the line, without actually measuring the location of the intermediate packets.
Another characteristic of the invention is the use of a "weighting" system to calculate a Weighted Error of the Real Line "e" '. As described in connection with Figures 114116, during each operating cycle a Real Line Error "e" is calculated. In the case of steady-state operation, three consecutive Real Line Errors can be used to calculate a "weighted" "e" 'error. This e 'is calculated using the following formula:
ES 2 180 076 T3 e '= 0.2 (e [t-2]) +0.3 (e [t-1]) +0.5 [(e [t]) where t = current cycle t- 1 = previous cycle t-2 = one cycle before the previous cycle
As can be seen, this formula assigns more weight to the most recent error, but at the same time provides some, albeit less weight, to earlier errors.
In summary, it can be seen that the above process described in relation to Figures 114-117 provides a method and apparatus according to the present invention for selecting individual packages from a line of accumulated packages, including error correction to compensate for the difference between the lengths of the packages before accumulation and their actual lengths while they are accumulated.
Contents21
85 sheets
Sheet 1 Sheet 2 Sheet 3 Sheet 4 Sheet 5 Sheet 6 Sheet 7 Sheet 8 Sheet 9 Sheet 10 Sheet 11 Sheet 12 Sheet 13 Sheet 14 Sheet 15 Sheet 16 Sheet 17 Sheet 18 Sheet 19 Sheet 20 Sheet 21 Sheet 22 Sheet 23 Sheet 24 Sheet 25 Sheet 26 Sheet 27 Sheet 28 Sheet 29 Sheet 30 Sheet 31 Sheet 32 Sheet 33 Sheet 34 Sheet 35 Sheet 36 Sheet 37 Sheet 38 Sheet 39 Sheet 40 Sheet 41 Sheet 42 Sheet 43 Sheet 44 Sheet 45 Sheet 46 Sheet 47 Sheet 48 Sheet 49 Sheet 50 Sheet 51 Sheet 52 Sheet 53 Sheet 54 Sheet 55 Sheet 56 Sheet 57 Sheet 58 Sheet 59 Sheet 60 Sheet 61 Sheet 62 Sheet 63 Sheet 64 Sheet 65 Sheet 66 Sheet 67 Sheet 68 Sheet 69 Sheet 70 Sheet 71 Sheet 72 Sheet 73 Sheet 74 Sheet 75 Sheet 76 Sheet 77 Sheet 78 Sheet 79 Sheet 80 Sheet 81 Sheet 82 Sheet 83 Sheet 84 Sheet 85
Every citation, both ways
| Document | Relation | Office | Cited during |
|---|---|---|---|
| CN112093133A | Cited by | China | Search report |
43 members in 9 offices
Priority claims5
| Document | Office | Kind | Date |
|---|---|---|---|
| 19960756675 | United States of America | – | |
| 75667596 | United States of America | A | |
| 75667596 | United States of America | A | |
| 97951601 | – | – | – |
| US19960756675 | – | – | – |
Members43
| Document | Office | Kind | |
|---|---|---|---|
| CA2272614A1 | Canada | A1 | |
| CA2485248A1 | Canada | A1 | |
| WO9823511A2 | World Intellectual Property Organization (WIPO) | A2 | |
| US5908283A | United States of America | A | |
| WO9823511A3 | World Intellectual Property Organization (WIPO) | A3 | |
| JP2000511146A | Japan | A | |
| EP1034124A2 | European Patent Office (EPO) | A2 | |
| US6120241A | United States of America | A | |
| US6286656B1 | United States of America | B1 | |
| US6290454B1 | United States of America | B1 | |
| EP1211203A2 | European Patent Office (EPO) | A2 | |
| EP1034124B1 | European Patent Office (EPO) | B1 | |
| AT219020T | Austria | T | |
| ATE219020T1 | Austria | T1 | |
| DE69713407D1 | Germany | D1 | |
| EP1211203A3 | European Patent Office (EPO) | A3 | |
| US2002106273A1 | United States of America | A1 | |
| DK1034124T3 | Denmark | T3 | |
| ES2180076T3This record | Spain | T3 | |
| DE69713407T2 | Germany | T2 | |
| JP2003335417A | Japan | A | |
| US6699007B2 | United States of America | B2 | |
| US2004165980A1 | United States of America | A1 | |
| EP1211203B1 | European Patent Office (EPO) | B1 | |
| AT276186T | Austria | T | |
| ATE276186T1 | Austria | T1 | |
| DE69730758D1 | Germany | D1 | |
| EP1489025A2 | European Patent Office (EPO) | A2 | |
| DK1211203T3 | Denmark | T3 | |
| CA2272614C | Canada | C | |
| ES2225394T3 | Spain | T3 | |
| EP1489025A3 | European Patent Office (EPO) | A3 | |
| DE69730758T2 | Germany | T2 | |
| JP3778574B2 | Japan | B2 | |
| EP1489025B1 | European Patent Office (EPO) | B1 | |
| AT343536T | Austria | T | |
| ATE343536T1 | Austria | T1 | |
| DE69736869D1 | Germany | D1 | |
| US7210894B2 | United States of America | B2 | |
| DE69736869T2 | Germany | T2 | |
| ES2278252T3 | Spain | T3 | |
| CA2485248C | Canada | C | |
| JP4081402B2 | Japan | B2 |
Numbers
- Publication
- 2180076
- Publication, DOCDB
- 2180076
- Publication, EPODOC
- ES2180076T
- Application
- 97951601
- Application, DOCDB
- 97951601
- Application, EPODOC
- ES19970951601T
Titles2
- Spanish
- PROCEDIMIENTO Y APARATO PARA PALETIZAR PAQUETES DE TAMAÑO Y PESO IRREGULARES.
- English
- PROCEDURE AND APPARATUS FOR PALLETIZING IRREGULAR SIZE AND WEIGHT PACKAGES.
Classification
- CPC, 6
- B65G47/90
- B65G47/905
- B65G47/91
- B65G61/00
- G06Q10/043
- Y10S414/116
- IPC, 7
- G01B21 16
- B65B35 44
- B65G47 90
- B65G47 91
- B65G57 03
- B65G61 00
- G06Q10 04