Automated packing and processing system and method for shipping using object pose analysis

Abstract

An automated packing system for placing a plurality of objects into a shipping container is disclosed. The system includes a supply bin receiving conveyor for receiving a supply bin at a supply station, the supply bin receiving conveyor including a sensing unit for determining a range of advancement of the supply bin along the supply bin receiving conveyor in a conveyor direction; a detection system for detecting that a plurality of objects within the supply bin are aligned by an alignment system in response to a position of the supply bin on the receiving conveyor; an object selection system for selecting a selected object from the plurality of objects for placement into the shipping container; and a programmable motion device for grasping and retrieving the selected object from the plurality of objects at the supply station and placing the selected object into the shipping container in a selected orientation and attitude.

Description

[0001] priority

[0002] This application claims priority to U.S. Provisional Patent Application No. 63 / 107,302, filed October 29, 2020, and U.S. Provisional Patent Application No. 63 / 172,987, filed April 9, 2021, the disclosure of each of which is incorporated herein by reference in its entirety. Background Technology

[0003] The present invention relates generally to automated sorting and other processing systems, and more particularly to automated systems for packaging objects such as parcels, packages, products, and goods for transport.

[0004] A shipping center that packs and ships a limited range of goods, such as those from a source company that manufactures the goods, may only need to repeatedly accommodate the same limited range of goods using its systems and processes. On the other hand, a third-party shipping center that receives a variety of goods must utilize systems and processes capable of accommodating a wide range of goods.

[0005] For example, in an e-commerce order fulfillment center, workers pack units of goods into shipping containers such as boxes or plastic bags. One of the final steps in the order fulfillment center is packing one or more items into shipping containers. Order units destined for customers are typically packed manually at the packing station. Order fulfillment centers do this for a variety of reasons.

[0006] First, units need to be packed in the shipping materials. For most items, simply labeling the units and placing them in a mailbag is insufficient. Units need to be placed in boxes or bags to protect the items. Second, units are not typically stored in the shipping materials; they are not usually just ready to leave the roll-up door. These units need to be packed immediately upon receiving the order; otherwise, warehouse space utilization will be very low if items are packed only when they are being shipped.

[0007] Third, grouping multiple units destined for the same customer together reduces shipping costs. The marginal cost of adding a single unit to a box is typically many times lower than creating a new box for that single unit. Fourth, consolidating packing operations at the packing station improves warehouse efficiency. Packing staff don't walk around picking from shelves; they focus on packing as many units as possible per hour.

[0008] As the volume of goods and the number of destinations increase, each of these requirements becomes more challenging. Therefore, an automated system is needed for packing objects in preparation for shipment. Summary of the Invention

[0009] According to one aspect, the present invention provides an automated packaging system for placing multiple objects into a transport container. The system includes: a supply box receiving conveyor for receiving supply boxes at a supply station, the supply box receiving conveyor including a sensing unit for determining the range of travel of the supply box along the supply box receiving conveyor in a conveyor direction; a detection system for detecting, in response to the position of the supply box on the receiving conveyor, that multiple objects within the supply box are aligned by an alignment system; an object selection system for selecting a chosen object from the multiple objects and placing it into the transport container; and a programmable motion device for grasping and acquiring the selected object from the multiple objects at the supply station and placing the selected object into the transport container with a selected orientation and posture.

[0010] According to another aspect, the present invention provides an automated packaging system for placing multiple objects into a transport container. The automated packaging system includes: an object posture permission assessment system for generating posture permission data regarding whether any of the objects at a supply station is deformable and likely to change shape when placed in the transport container, or likely to move after being placed in the transport container; an object selection system for selecting a chosen object from the multiple objects to place into the transport container in response to the posture permission data; and a programmable motion device for grasping and acquiring the selected object from the multiple objects at the supply station and placing the selected object into the transport container with a selected orientation and posture in response to the posture permission data.

[0011] According to another aspect, the present invention provides an automated packing system for placing multiple objects into a destination box. The automated packing system includes: a supply box receiving conveyor for receiving supply boxes at a supply station; a supply detection system for detecting objects within the supply boxes; an object selection system for selecting one object from a plurality of objects and placing it into a transport container; a programmable motion device including an end effector for grasping and acquiring the selected object from the plurality of objects at the supply station; a destination box receiving conveyor for receiving a destination box at a destination station; a destination detection system for detecting the volume within the destination box; an attitude maintenance evaluation system for providing attitude maintenance evaluation data regarding the attitude and orientation of the object relative to the end effector; and an attitude adjustment system for adjusting at least the attitude or orientation of the object before placing the selected object into the destination box in response to the attitude maintenance evaluation data.

[0012] According to another aspect, the present invention provides a process for maintaining a vacuum cup of an end effector of a programmable motion device. The process includes determining that debris can be collected near the vacuum cup, and using the end effector of the programmable motion device to deposit the debris into a waste bin. Attached Figure Description

[0013] The specific embodiments can be further understood by referring to the accompanying drawings, in which:

[0014] Figures 1A to 1C A schematic diagram is shown of a gripper and object that are not centered at the target position;

[0015] Figures 2A to 2C An illustrative schematic diagram of a gripper and an object centered at a target position according to one aspect of the invention is shown;

[0016] Figure 3 An illustrative schematic diagram of a gripper and an object according to a data model is shown, based on one aspect of the invention.

[0017] Figure 4 An illustrative schematic diagram of a gripper and an object processed according to one aspect of the present invention is shown;

[0018] Figure 5 An illustrative schematic diagram of an analysis and control system according to one aspect of the present invention is shown;

[0019] Figure 6 An object processing system according to one aspect of the present invention is shown, wherein the input shipping container comprises a single SKU object;

[0020] Figure 7 It shows Figure 6 An illustrative diagram of a single SKU picking cell in the system;

[0021] Figure 8 It shows Figure 7 An illustrative floor plan of a single SKU picking cell;

[0022] Figure 9 It shows in Figure 7 A schematic diagram of the bidirectional deflector used in a single SKU picking cell;

[0023] Figure 10 It shows in Figure 7 An illustrative diagram of the weight-sensing conveyor section used in a single SKU picking cell;

[0024] Figure 11 It shows in Figure 7 An illustrative diagram of the programmable motion device used in a single SKU picking cell;

[0025] Figures 12A to 12C Illustrative schematic diagrams showing views of a single SKU box according to one aspect of the invention illustrate camera views. Figure 12A ), box volume scan ( Figure 12B ) and volume scan of the box after picking ( Figure 12C );

[0026] Figures 13A to 13C An illustrative schematic diagram of a grasping evaluation and detection system according to one aspect of the present invention is shown in the system.

[0027] Figure 14A and Figure 14B Illustrative diagrams are shown illustrating different handholding postures for grasping objects according to various aspects of the invention, showing the grasping of objects with additional force. Figure 14A ) and compensation capture ( Figure 14B );

[0028] Figures 15A to 15D An illustrative schematic diagram of an object movement (swinging) grasping detection system according to one aspect of the present invention is shown;

[0029] Figures 16A to 16D An illustrative schematic diagram of a portion of a box alignment system used in a system according to one aspect of the present invention is shown;

[0030] Figure 17A and Figure 17B An illustrative schematic diagram of a portion of a container position detection system used in a system according to one aspect of the present invention is shown;

[0031] Figure 18A and Figure 18B An illustrative schematic diagram of a container portion used in a packaging process in a system according to one aspect of the present invention is shown;

[0032] Figure 19A and Figure 19B An illustrative schematic diagram of a container packaging strategy in a system according to one aspect of the present invention is shown;

[0033] Figure 20 An illustrative schematic diagram of a portion of a packaging planning system used in a system according to one aspect of the present invention is shown;

[0034] Figure 21 An illustrative schematic diagram showing a portion of a packaged container illustrates container-object margins and object-object margins used in a system according to one aspect of the invention.

[0035] Figure 22An illustrative graphical representation of the relationship between margins and packing volume in a system according to an aspect of the invention is shown;

[0036] Figures 23A to 23E An illustrative schematic diagram of an end effector of a programmable motion device for placing an object into a container in a system, according to one aspect of the present invention, is shown.

[0037] Figure 24 An illustrative graphical representation of the detected weight change over time in a placement system according to one aspect of the invention is shown;

[0038] Figures 25A to 25C An illustrative schematic diagram of a process control system used in a system according to one aspect of the present invention is shown;

[0039] Figure 26 It shows Figure 6 An illustrative floor plan of the single SKU picking cell of the system;

[0040] Figure 27 It shows Figure 6 An illustrative side view of the system's single SKU picking cell;

[0041] Figure 28 It shows Figure 6 An illustrative rear view of the system's single SKU picking cell;

[0042] Figure 29 An object processing system according to one aspect of the present invention is shown, wherein the input shipping container comprises multi-SKU objects;

[0043] Figure 30 It shows Figure 29 An illustrative side view of a multi-SKU picking cell;

[0044] Figure 31 It shows Figure 29 An illustrative floor plan of a multi-SKU picking cell;

[0045] Figures 32A to 32C It shows Figure 28 An illustrative diagram of the feed analysis system in a multi-SKU picking cell;

[0046] Figures 33A to 33C Illustrative schematic diagrams showing views of a multi-SKU box according to one aspect of the invention are provided, illustrating camera views. Figure 33A ), box volume scan ( Figure 33B ) and volume scan of the box after picking ( Figure 33C );

[0047] Figures 34A to 34CAn illustrative schematic diagram of a sensing system in a container contents verification system according to one aspect of the present invention is shown;

[0048] Figures 35A to 35C An illustrative schematic diagram of a grabbing evaluation and detection system for use in a multi-SKU processing system according to one aspect of the present invention is shown;

[0049] Figure 36 It shows Figure 29 An illustrative side view of a multi-SKU picking cell;

[0050] Figure 37 It shows Figure 29 An illustrative back view of the multi-SKU picking cells;

[0051] Figure 38 An illustrative schematic diagram is shown of an end effector vacuum opening in an object handling system that is blocked by debris.

[0052] Figure 39 An illustrative schematic diagram shows an end effector vacuum opening in an object handling system that is completely blocked by debris.

[0053] Figure 40 An illustrative schematic diagram of a single SKU object handling cell, including a trash can as part of an object handling system in a system, is shown according to one aspect of the invention.

[0054] Figure 41 An aspect of the invention is shown. Figure 40 An illustrative schematic diagram of a system having an end effector for depositing debris into a bin within the system;

[0055] Figure 42 An illustrative schematic diagram is shown, according to one aspect of the invention, including a trash can as part of an object handling system within the system;

[0056] Figure 43 An aspect of the invention is shown. Figure 42 An illustrative schematic diagram of a system having an end effector for depositing debris into a bin within the system;

[0057] Figure 44 An illustrative schematic diagram of a vacuum source in a debris removal system according to one aspect of the present invention is shown.

[0058] Figure 45 An illustrative schematic diagram of a debris removal brush pad used in a debris removal system according to one aspect of the present invention is shown.

[0059] Figure 46 An illustrative schematic diagram of the process flow in a debris detection system according to one aspect of the present invention is shown; and

[0060] Figures 47A to 47C An illustrative schematic diagram of the process flow in a debris removal system according to one aspect of the present invention is shown.

[0061] The accompanying drawings are for illustrative purposes only. Detailed Implementation

[0062] According to various aspects, the present invention provides systems and methods for robotically packing and transporting containers (whether boxes or cardboard pallets) or holding one or more cargo units in preparation for transport of objects and other physical containers. The applicant has identified a need for a robotic system capable of picking units from an inventory system (such as inventory shipping boxes stored in AS / RS). A system is also needed capable of placing one or more units into transport containers (such as boxes, cardboard (which will later be shrunken)) or onto a conveyor for transport to a boxing or bagging machine. A robotic system is also needed capable of efficiently packing one or more units into boxes, for example by pre-specifying an optimal box size and then packing one or more units to leave as little air as possible in the box, thus avoiding large gaps.

[0063] In contrast, systems that assemble pallets using robots are different; at least in this case, they are provided as collections of cardboard boxes, typically filled with multiple units of some kind of goods, where the boxes can be easily stacked on top of each other. According to various aspects of the invention, a system is provided that allows individual units, rather than multiple units, to be packed, and importantly, these units may not be boxes; they are not like blocks that can be stacked arbitrarily. Instead, they may not be rigid (e.g., they may be in bags), or they may be rigid but not box-shaped. They may be cylindrical and therefore roll away, or their shape may not be suitable for stacking items on top of them. According to various aspects, the invention relates to the ability to pick units from a jumbled pile. According to other aspects, the invention relates to placing items in a targeted manner to prepare them for transport while minimizing internal voids and ensuring their integrity during transport. According to other aspects, the system can finish such boxes using any of a variety of automated box creation, box finishing systems, and automated bagging systems.

[0064] The applicant has further identified several challenges facing the automated packing system. These challenges include the need to compensate for the handheld posture of the clamped items, the need to compensate for errors and noise in the placement of items relative to other items, and the need to compensate for objects with low posture authority. These challenges also include the need to compensate for the fact that previously placed items may not be in their original positions while placing items, the need to pack without requiring a large amount of extra space around items, and the need to pack at high throughput.

[0065] Regarding hand-held posture Figure 1A An end effector 10 (e.g., a gripper) is shown centered on object 12, but the object is not centered at target position 14. Figure 1B An end effector 20 that is not centered on an object is shown, and the object is outside the target position 24. Figure 1C An end effector 30 is shown that is not centered on object 32, but the object is rotatably outside the target position 34. If the end effector is centered on the cardboard receiving surface, but the item is not centered on the receiving surface, the item may be located outside the range of the presented cardboard receiving surface, which would be problematic for attempting to pack (e.g., boxes) or shrink-wrap cardboard surfaces (e.g., if it is a shipping pallet). Figure 1A and Figure 2A The image shows an end effector 10 grasping an object 12 above a receiving surface 14. (Reference) Figure 2A The system will adjust the position of the end effector 10 to position the object above the receiving surface 14, so that the object will be placed at the desired placement position 16. Figure 2A As shown, position 16 is then precisely positioned on receiving surface 14. Figure 1B As shown, if the end effector 20 grasps an object 22 at a non-central position, then even when the end effector is centered above the receiving surface 24, the object 22 will be placed at a position 26 extending beyond the receiving surface 24, which is undesirable. (See reference) Figure 2B The system will adjust the position of the end effector 20 to position the object above the receiving surface 24, so that the object will be placed at the desired position 26. The gripper position and orientation have been selected to compensate for the handheld posture of the object being held. Similarly, if the end effector 30 grips... Figure 1C As shown, if the end effector is centered and oriented on the receiving surface 34, then the object 32 will be placed at a position 36 that is rotated beyond the receiving surface 34, which is also undesirable. (See reference) Figure 2CThe system will rotate the position of the end effector 30 to position the object above the receiving surface 34, so that the object will be placed at the desired position 36. Similarly, the gripper position and its orientation have been adjusted to compensate for the handheld posture of the held object. Even if the position and / or orientation of the end effector are known to provide a contact surface of the end effector parallel to the receiving surface, the position and / or orientation of the object may still need to be adjusted before placement.

[0066] It is also necessary to compensate for errors and noise in the placement of objects relative to other objects. These errors include errors in the known position of the object placed in a container (e.g., a box), errors in handheld pose estimation, robot localization errors, and errors due to unobserved passive degrees of freedom or compliance. For example, a flexible gripper holding an object may deflect and change the true handheld pose. It is also necessary to compensate for objects with low pose authority. Pose authority is the ability to place an object in a desired position and orientation. Some objects may be unsuitable for orientation, or the resulting pose may be unpredictable. For example, an object loosely packed in a bag may wrinkle or fold on its own, or it may swing / wobble uncontrollably before placement, making its resulting size and orientation unpredictable.

[0067] It is also necessary to compensate for the fact that previously placed items may not be in their original positions when placing items. Previously placed objects may have tipped over, fallen, or rolled. In other words, some objects may move after placement, and according to one aspect, the present invention relates to compensating for objects with low placement authority. Placement authority is the ability of an object to remain in its placement position and orientation.

[0068] Packing also needs to be done without requiring much extra space around the items. To prevent the held items from bumping into other items or transport containers, the robot can compensate by adding extra spacing around the held items. This extra spacing increases transport costs, and it is generally desirable to add as little margin as possible around the items. Additionally, high throughput packing is required, and the robot needs to determine where to place the object very quickly after confirming how it is held by the gripper.

[0069] According to one aspect of the invention, the system involves adding a scanner to the picking cell, the scanner identifying the position and orientation of the item as it is held by a gripper and en route to its placement location. The system images the item from multiple directions to identify its main axis, allowing the item to be aligned with other adjacent items in a cardboard box or order. Because of this capability, the system can begin with limited SKU coverage—primarily rigid boxed items—and then expand to clamshell items, followed by bagged items such as clothing.

[0070] refer to Figure 3 and Figure 4 One problem that sometimes arises is compensating for the reality model ( Figure 3 ) and reality itself ( Figure 4 Differences between them, for example, due to the product's non-rigid packaging and / or the location of the opening, may cause slight variations in the product's shape. Specifically, Figure 3 A model of product 40, held by the model of end effector 42, is schematically shown. The modeling is computational, and Figure 3 The purpose is to visually illustrate computational modeling. Position and orientation may have minor errors, and the actual dimensions of the object may differ from those in the database. Figure 4 An actual product 44 is shown being held by an actual end effector 46. As indicated at 48, the packing may change when the end effector lifts the packing (e.g., a box) from the top. Similarly, some packings are designed to lift the product from the bottom. As shown, when the end effector lifts the top surface (which may form part of the opening cover of the box) from the top, the top surface may be slightly lifted away from the rest of the box. This is not only undesirable for attitude control and packing, but this choice of gripping position can also significantly impair the system's ability to reliably pick and place items (e.g., if the top surface / cover breaks / tears from the rest of the box).

[0071] These differences affect the system's packaging performance. To minimize these differences to meet the requirements of certain applications, the system uses baseline handheld pose scanning and modeling, and records the degree to which the model reflects reality. For example, Figure 5 An analysis and control system is illustrated at 50, comprising a container detection unit 51 on a conveyor, a weight-sensing conveyor section 52, a weight-transmitting bidirectional conveyor section 53, a fixed-mount detection unit 54 adjacent to a programmable motion device 56, and a gripping planning detection unit 55 and a gripping analysis detection unit 57 associated with each programmable motion device 56. These detection units each provide information (e.g., via an intranet or the Internet) to one or more processing systems 100 that also communicate with a storage system 58. By accessing stored information about each item and by evaluating gripping parameters including gripping position and end effector placement, the system accesses, generates, and records data regarding object size, weight, packing, material, posture permissions, position permissions, gripping position, vacuum pressure level, and vacuum duration. For example, the system can determine one or more specific gripping positions and determine that the vacuum parameters are optimal for each object.

[0072] For example, Figure 6A system 110 is shown, comprising a pair of feed conveyors 112, 114, on which a single SKU inventory container (e.g., a shipping box) 116 is conveyed to a feed conveyor including a programmable motion device (such as an articulated arm 122). Figure 7 Further shown is a single SKU packing cell system 120. System 110 also includes a transport container conveyor 124 that provides transport containers 126 to the single SKU packing cell system 120. According to one aspect, the single SKU packing cell system picks individual units from inventory shipping boxes and places them in or on packages suitable for transport. A robot support structure 128 spans two inventory shipping box loops 112, 114 that feed inventory shipping boxes from a shipping box storage system (such as AS / RS) to the cell and extends above the transport container conveyor 124.

[0073] Detection unit 139 (e.g.) Figure 7 (Further shown) The movement and position of containers 116 and 126 on conveyors 112, 114, and 124 are monitored by detecting unique marks 135 and 127 on containers 116 and 126, respectively. Detection unit 139 on conveyors 112 and 114 detects mark 135 on container 116, while detection unit 141 on conveyor 124 detects mark 127 on container 126. Detection unit 138 on support structure 128 monitors the grasping and movement of objects by end effector 134 of programmable motion device 122. Detection unit 160 (e.g.) Figure 11 (As shown) Assists a programmable motion device in selecting and grasping objects. System 110 independently and intermittently moves containers 116 on conveyors 112, 114 to provide objects for processing at cell system 120, wherein objects are selectively placed in any of a variety of containers 126 arriving at a motion-controlled conveyor 124. Each conveyor 112, 114, 124 includes bidirectional deflectors 113, 115 (e.g., ...). Figure 9 and Figure 26 (as shown in more detail below) one or more sections, the bidirectional deflector deflects the container along an output direction orthogonal to the input direction. Certain bidirectional deflectors 113 and sections 117 of the conveyor 124 (e.g.) Figure 10 and Figure 26 (Further shown) includes rollers mounted on a torque sensor (e.g., for measuring weight), as discussed further below. Additionally, conveyor 124 includes a container alignment system 119 (in... Figure 10 and Figures 16A to 16D (As further shown in the diagram) for aligning containers when in proximity to a programmable motion device. The operation and control of the system, including the conveyor, detection unit, bidirectional deflector, container alignment system, and programmable motion device, are provided by one or more computer processing systems 100.

[0074] refer to Figure 7 System 110 can select one or both containers 132, 133 on conveyors 112, 114 for supplying objects to programmable motion device 122. The weight of each of containers 132, 133 can be determined independently (using reference below). Figure 9 (The torque sensor is discussed in more detail). A selected transport container 125 in container 126 on conveyor 124 is also provided close to the programmable motion device, and references... Figures 16A to 16D The selected object is pushed into the loading position on the conveyor 124 by the container alignment system 119. The selected object is then moved to the container 125 by a programmable motion device, determining its position and orientation within the container, and, after placement, detecting the weight of the container (discussed in more detail below) to confirm placement. A detection system 138 may be located around the area where the end effector 134 picks the object from the shipping container. The detection system 138 is positioned to capture all surfaces of the object, including its bottom. This allows the item to be scanned once it has been picked from the shipping container. This allows sufficient time to calculate the handheld posture and then perform motion planning to compensate for the handheld posture.

[0075] Figure 8 A top view of packing cell 120 is shown, illustrating the weight-sensing bidirectional deflector 113 of conveyors 112, 114 and the bidirectional deflector 115 of conveyor 124. Detection unit 138 may include depth sensors, such as depth cameras positioned around the station from multiple directions, to estimate the held object and its handheld posture. The feeding system may include multiple feed conveyors 112, 114 to provide multiple objects to programmable motion device 122. The weight-sensing bidirectional deflector 113 can determine not only the weight of the container but also its position on the rollers. Bidirectional deflectors 113, 115 may include a lateral directional belt 144. Figure 9 As shown in the diagram, the transverse directional belt can be raised to guide the container in a direction orthogonal to the direction in which the conveyor approaches the conveyor. Once detected and identified / confirmed, the object can be moved via a programmable motion device 122, such as an articulated arm, to a destination container 125 (e.g., a transport box) on the output conveyor 124.

[0076] The weight-sensing bidirectional deflector 113 includes a component mounted on the torque sensor 142 (e.g., Figure 9 As shown, on roller 140, the torque sensor is used to determine the weight and position of the container on the roller. Figure 9As further shown, the bidirectional deflectors 113 (and 115) include a belt 144 that, when engaged, can be lifted to a position that deflects the container. The belt can also be used to confirm the weight of the container (e.g., before and after removing an object) by lifting the container away from the weight-sensing roller and then placing the container back onto the weight-sensing roller.

[0077] By monitoring the output of each of the force sensor or torque sensor 142, the position of the container on the roller can be determined, and the roller can be advanced to bring the container to a specific position on the roller below the articulated arm 122. The destination conveyor 124 also includes a weight-sensing conveyor section 117, which includes a conveyor 150 mounted on the force sensor or torque sensor 152, as discussed above and Figure 10 This is shown in more detail below. Individual force sensors or torque sensors allow the system to determine the position of the container on the roller 150 mounted on the torque sensor 152. By monitoring the output of each of the force sensors or torque sensors 152, the position of the container on the roller can thus be determined, and the roller can be advanced to bring the container to a specific position on the roller below the hinge arm 122. The container alignment system 119 can then be engaged to position the container on the support 153, as described below. Figures 16A to 16D Let's discuss this in more detail.

[0078] The system includes a sensing system (e.g., 160) mounted above the container of the object to be processed, facing downwards, near the base of the articulated arm 122 having an end effector 134. (Reference) Figure 11 The perception system 160 may, for example (on its underside), include a camera, a depth sensor, and lights. It acquires a combination of 2D and 3D (depth) data. The depth sensor can provide depth information, which can be used in conjunction with camera image data to determine depth information about various objects in the view. The lights can be used to remove shadows and facilitate the recognition of object edges, and can be all on during use, or can be lit in a desired order to assist in object recognition. The system uses the image and various algorithms to generate a set of candidate grasping positions for objects in the box, as discussed in more detail below.

[0079] Figure 12A A view of container 132 from sensing system 160 is shown. The image view shows container 132 (e.g., on a conveyor), and container 132 contains objects 201, 202, 203, 204, 205. While in some systems the objects in each feeder may be non-homogeneous (multiple SKUs), in systems such as... Figure 12AIn the other systems shown, the objects may be homogeneous (a single SKU). The system will identify candidate gripping locations on one or more objects and may not attempt to also identify gripping locations on objects partially occluded by other objects. Candidate gripping locations can be indicated using a 3D model of the robot's end effector, placed in the location where the actual end effector will be used as the gripping location. For example, a gripping location may be considered good if it is close to the object's center of mass to provide greater stability during gripping and transport, and / or if it avoids locations on the object where a good vacuum seal may not be possible (such as lids, seams, etc.).

[0080] The sensing system 160 includes a sensing unit, a scanning and receiving unit, and an edge detection unit for capturing various characteristics of a selected object throughout the box. Similarly, Figure 12A The diagram shows a view from a capture system, which, according to an embodiment, may include a group or more similar objects 201, 202, 203, 204, 205. Figure 12B The scanned volume V is shown in the figure. 203 or density D 203 The difference is analyzed and compared with recorded data on items identified by identification tags provided by the detection system via the SKU sensing system, or recorded object data. Specifically, the scanned volume is compared with the volume of the identified SKU multiplied by the known number of objects in the box. After picking, the volume is scanned again ( Figure 12C To confirm the volume of the object being picked.

[0081] Figures 13A to 13C The grab evaluation detection unit 138 in the packaged cell 120 is shown (some components have been removed for clarity). Detection unit 138 includes an upper detection unit, which is as follows: Figure 13A The image points downwards towards the object 162 held by the end effector 134. The detection unit 138 further includes an intermediate detection unit, which is as follows: Figure 13B The object 162, which is roughly horizontally pointed to by the end effector 134, is shown; and the lower detection unit, as shown in the figure... Figure 13C The object 162 is shown pointing upwards and held by the end effector 134.

[0082] Once the handheld posture is captured by the detection unit 138 and sent to the robot application, the robot application then needs to send commands to the robot, including a time series of the joint angle, i.e., the robotic arm trajectory. The robotic arm trajectory needs to be selected to place the item in the desired location and orientation. For example, Figure 14AA graphical representation of an end effector 134 is shown at 170, which includes a flexible vacuum cup 162 deflected by a load (object 164). This load generates an unwanted force as shown at 166, the load including potential components in the x, y, and z directions. To compensate for the load (such as...) Figure 14B As shown, the mobile end effector 134 provides a reaction force as shown at 168, which is equal to and opposite to force 166, and also includes components in the x, y, and z directions.

[0083] The generation of the robotic arm trajectory should occur within the last 100 milliseconds of the item's journey to its placement location. If a compensated trajectory cannot be calculated in real time, the system pre-generates a trajectory roadmap of candidate poses. This roadmap is called a roadmap because it connects multiple trajectories linked by nodes representing common intermediate locations (like intersections on a street map). The roadmap includes hundreds or thousands of trajectories, each starting from a common intermediate location and ending at a densely sampled gripper position and orientation. A common intermediate location is selected such that a handheld pose estimate is available before the robot reaches it. The optimal next trajectory on the roadmap can then be selected before the robot reaches the end of its current trajectory. The robot then smoothly transitions to the trajectory that generates the desired item placement. If the item is rotated or translated along the xy-axis while in the hand, the placement trajectory can be simplified to a simple downward movement to avoid complex planning for conflicts with the contents of the transport container.

[0084] refer to Figures 15A to 15D The picking cell may include a gripping detection system that assesses whether an object is moving (e.g., swinging) when it is gripped. The detection system may include a swing detection system 220, which includes a plurality of sensing units 222 pointing to a detection area between the input shipping containers 132, 133 and the destination container 125 (e.g., a delivery box). Additional sensing systems (e.g., 138, 141) discussed above may also be used, but unit 222 specifically points to the area between the input shipping container and the destination container where the end effector 134 is programmed to stop. Figure 15A The image shows object 210 moving as it is grasped by vacuum cup 224, and... Figure 15B The end effector 134 is shown stopping in the analysis region. (Reference) Figure 15C Object 210 can continue to swing forward, and refer to... Figure 15DIt can even swing backward in the opposite direction. Although the use of the flexible vacuum cup 224 may result in slight movement, the system will set a threshold for detecting any movement (swinging) so that movement caused by the flexible vacuum cup is excluded from the object's swinging movement. For example, objects including flexible bags (e.g., polyethylene bags) will experience more noticeable swinging than rigid objects such as boxes.

[0085] The detection of this swaying movement of the object being processed is recorded and used in the packing process, for example, by preventing the item from being placed in a lower position within the container than other objects. Such swaying objects can be placed on top of other objects in the packing container (rather than below them), because objects placed on top of such flexible objects may move during placement, causing disruption and uncertainty in the packing process. The system can, for example, take multiple photographs at different times to determine the movement, and thus these methods can be used to detect whether an object held by the gripper is swaying. This information also affects placement, as the system now knows to allow the item to stop swaying before being placed into the shipping container and can be further adjusted to move the object more slowly. Similarly, this also affects packing strategies as discussed herein.

[0086] In addition, and refer to Figures 16A to 16D The destination container 125 on the conveyor 124 moves toward the programmable motion device 122 (e.g., Figure 16A (as shown), and stops when container 125 is on weight sensing conveyor section 117 (as shown). Figure 16B (As shown). The support rod 151 of the container alignment system 119 moves against the container 125 and pushes the container against the support rod 153 of the container alignment system 119 to the desired position on the conveyor 124 (e.g., as shown). Figure 16C (As shown). The articulated arm 122 can then handle one or more objects to be supplied to the container while the container is maintained in a known location / position between the support rod 151 and the support rail 153. Once complete, the support rod 151 releases the container, and rollers are engaged to further move the container along the conveyor 124 (as shown). Figure 16D (As shown) it is moved to another processing station. Similarly, other weight-sensing conveyors, as well as support rods and guide rail systems, can be used with the loading container 156 on the feed conveyor.

[0087] Figure 17A and Figure 17BA view of the upper sensing system (e.g., 160) of the transport container, viewed from above, is shown. Specifically, the system will know that the transport container 125 is positioned in the conveyor 124, and can have information about its general position on the conveyor due to the use of a weight-sensing conveyor roller 150 with sensor 152 and a box alignment system 119. The expected size of the box 125 is known (from stored modeling information) and its distance from the sensing system (again, 160) is taken into account. The system then attempts to map the known contours of the top of the container onto an image to align the system with the exact position of the container 125. In some embodiments, the stored modeling information can be combined with sensing information from a 3D depth sensor (also in sensing unit 160) to measure features of the container to determine its position in 3D. This registration information also helps to avoid the end effector contacting (impacting) the container during packing. Figure 17A As shown in 230, the known contour is not yet aligned. The system will perform internal (software) adjustments to realign it with the actual container 125, as follows: Figure 17B As shown. An attempt was made to align the top opening of the box to avoid issues related to the variable position of the cover.

[0088] As discussed above, the system includes one or more sensing units 139 located on or near the feed conveyor for identifying external markings on each of the containers 116, providing sensing data from which the contents of the container can be identified, and then determining the relative position of the container on conveyors 112, 114 and tracking the container's position. It is assumed that the containers of objects are marked with visually unique markings, such as barcodes (e.g., providing UPC codes), radio frequency identification (RFID) tags, or mailing tags, at one or more locations on their exterior, making them readily identifiable by a scanner for processing. The type of marking depends on the type of scanning system used but may include 1D or 2D barcode symbols. Various symbols or marking methods can be employed. It is assumed that the type of scanner used is compatible with the marking method. For example, marking can be used to encode identification marks (e.g., symbol strings), typically a string of letters and / or numbers, via barcodes, RFID tags, mailing tags, or other means. The symbol string uniquely associates the supplier's container with a specific set of homogeneous objects. Each of the single SKU feed containers (e.g., boxes or shipping containers) may include a mark to identify the box or shipping container, and a sensing unit that detects the mark can be positioned along the conveyor to determine the identity and location (e.g., box or shipping container) of each single SKU container.

[0089] The operation of the aforementioned system is coordinated with a central processing system 100, which communicates with articulated arms 122, sensing systems 138, 139, and 160, and conveyors 112, 114, and 124, as well as weight-sensing conveyor sections (e.g., wirelessly). The system determines the UPC associated with the supplier box and the outbound destination for each object based on a symbol string. The central control system 100 consists of one or more workstations or central processing units (CPUs). For example, the mapping between UPCs or mailing tags and outbound destinations is maintained by the central control system in a database called a manifest. The central control system maintains the manifest by communicating with a warehouse management system (WMS). The manifest provides an outbound destination for each inbound object.

[0090] The packing planner generates packing plans for orders (especially multi-item orders). Given the items in the order and their weight and dimensions, the planning system determines a packing plan that satisfies various constraints, such as placing smaller items on top of larger items, and packing objects with lower posture and / or position permissions later or last. These constraints are provided to an optimizer, which generates a sequence of items that should satisfy the constraints so that they can be automatically packed onto cardboard.

[0091] Figure 18A A container 125 (e.g., a packing box) into which objects are to be packed is shown, for example, reserving a larger area for packing larger objects and a smaller area for packing smaller objects. The system is aware of the position and orientation of the walls of box 125 (e.g., 212, 213, 215 as shown). Box 125 may, for example, receive objects 214 that the system knows have (or has determined to have) low positional permissions (the ability to remain in place upon placement). For example, such objects may include objects whose outer surfaces are typically cylindrical or spherical. The system can then place object 214 into the box biased towards one or both sides of the box (213, 215), thereby leaving more space 216 for other objects. Reference Figure 18B Then other objects 218 and 219 can be placed in the larger open space 216 of the box, thus placing the larger of the two new objects (218) at the bottom.

[0092] Figure 18A and Figure 18BSome potential ordering rules are illustrated, which will be incorporated into the packing planner. The selected set of rules will depend on the characteristics of the shipping container. The rule set may include, but is not limited to: placing the largest items first, placing smaller items on top of larger items (to generate the most item stack), placing the heaviest items first (to avoid crushing other items), placing known fragile items last (to avoid them being crushed), placing low-position-authority items at risk of rolling last (to avoid them rolling and obstructing the next picking), and placing non-rigid items with low placement authority last (so that subsequent placements will not tip over). Some rules may conflict with each other, requiring the use of context-sensitive parameters to resolve this. The system can experiment to adjust or learn the relative importance of the rules and determine which rules to prioritize. The system plans to generate the positions of all items in the order, the order in which they should arrive at the cells, and the size of the boxes the objects will fit into. These latter requests are routed to the Warehouse Management System (WMS) via a software interface.

[0093] To develop an effective and reliable plan, the packing planner will need SKU information, including weight, dimensions, and possibly other SKU attributes such as attitude and position permissions, like whether it is rolling or non-rigid. Additional information, such as raw data from dimensional measuring devices (e.g., the Cubascan system sold by Quantronix, Hamden, Connecticut), is also expected to improve packing performance. Part of the work involves developing mutually acceptable patterns while keeping in mind the cost of obtaining any information.

[0094] According to other aspects, the system also provides a container holding mechanism. For example, when a robot places an item into a box, the item may scratch the box wall. According to certain aspects of the invention, a mechanism may be needed to hold the container. The type and requirement of the holding mechanism will depend on the container. The system will also provide requirements analysis, such as the need to include a range of container sizes or potential types; the design, implementation, and testing of the holding mechanism; and electromechanical integration into single-pick and multi-pick cells (discussed in further detail below).

[0095] Depending on other aspects, the system can provide anomaly detection and handling procedures. While in most cases the combination of picking software and hardware will result in efficient and effective object picking and placement, real-world warehouse conditions occasionally necessitate the detection and mitigation of anomalies. Mitigation can be automated, such as selecting an alternative gripping posture, or may require manual intervention via a user interface or by routing the anomaly to a QA / QC station. The user interface provides detailed causes of the anomalies and allows the operator to indicate that the situation has been resolved and the robot can continue picking.

[0096] One anomaly that robotic picking solutions may encounter is multi-picking, where the robot inadvertently picks up multiple items. This can occur due to poor gripping positioning or packaging defects that accidentally bundle multiple items together. The picking cell addresses this issue with a high-precision scale mounted below the picking and destination locations. When the software detects that the weight of the held item significantly exceeds the expected weight of a single item, it compensates for multi-picking by instructing the robotic arm to return the item to the inbound container.

[0097] The system analyzes the effectiveness of the inspection mechanism and, if necessary, accelerates the development of additional inspection mechanisms, such as volume measurement of picked items via handheld gesture scanning. Another possible anomaly is that objects are incorrectly placed in the outbound container. An example of this is that the placed object is too large to fit inside the outbound container. The system will assess the frequency of this anomaly and, if necessary, develop sensing mechanisms to detect and mitigate the situation. This may include re-grabbing and manipulating items or marking outbound containers to be sent to the QA / QC station before heat shrinking and shipping.

[0098] Therefore, according to various aspects, the present invention provides systems and methods for automating the handling of objects at processing cells, regardless of whether the objects originate from single-SKU or multi-SKU supply boxes. For example, the system provides a handheld posture scanning and estimation process, through which sensors and computer processes estimate how the gripper will hold the items. The system also provides a handheld posture compensation placement planning system, through which a motion planning process compensates for how the gripper holds the items and performs this process at a sufficiently high speed to maintain high throughput. The system also provides a packing planning process, through which the SKU sequence is defined, the most suitable box size is determined, and the position or orientation of all items in the shipping container is specified, while taking into account ordering constraints, such as placing rigid items first. According to other aspects, the system provides a shipping container holding mechanism, through which the system can optionally hold the shipping container in place while placing items to mitigate the impact of collisions with the shipping container walls. According to other aspects, the system provides anomaly detectors and processing procedures that enable the sensing and application processes to detect when anomalies occur, such as an item being placed in the wrong position or orientation, and to take action to automatically correct it, or to signal for QA / QC checks.

[0099] During operation, a barcode scanner scans identification marks on inventory containers (e.g., shipping boxes) to locate SKUs before they arrive at the cell. The cell then initiates the construction and / or provision of appropriately sized shipping containers via WCS / WMS to ensure timely arrival and packaging of the corresponding SKUs. Inventory shipping boxes queue on inbound inventory conveyors and stop at right-angle belt transfer machines within robot reach. A series of appropriately sized empty shipping containers are then fed into the cell via belt conveyors in an order matching the SKU arrival sequence. As items are removed from the inventory shipping boxes, they are scanned, and the cell's orientation and orientation are determined for placement. By compensating for the cell's known position and orientation relative to the gripper, the cell is placed on and within the shipping container.

[0100] A scanner above the packing container detects cases where the placement of units is unsuitable for packing (i.e., outside the shipping container). In such cases, the units are either re-grabbed and repositioned, or directed to the Quality Assurance (QA) station. A barcode on the shipping container is scanned, or the shipping container is marked with a barcode to associate the units and shipping container with the corresponding order. During the placement of units from one inventory shipping container, the contents of a second inventory shipping container are scanned in preparation for the next pick. A conveyor belt transfers cardboard boxes with individual inventory units onto a conveyor belt that feeds them to a final packing machine (such as a shrink wrapper and a box or carton sealer). If a single pick from the inventory shipping container is no longer required, the container is transferred to the outbound conveyor. Otherwise, the inventory shipping container is held, and another pick is performed. The process is repeated when picking begins on the second inventory shipping container.

[0101] refer to Figure 19A and Figure 19B The system can allocate shipping containers given a set of contents and a planned packing schedule, as discussed below. For example, the system can evaluate... Figure 19A The first group of objects 226 shown should be packed into box 125', and as Figure 19B The group of larger objects 228 shown should be packed into a larger box 125". For example, the system, knowing the volume of each object and the number of different objects, can approximate the packing volume by adding the volumes of the individual objects and increasing the packing volume by 10%, 15%, or 20%, which depends at least in part on the allocated distances (margins) between the objects and between the objects and the container, as discussed further below.

[0102] refer to Figure 20Knowing a group of objects to be packaged together and the intended container, the system analyzes different combinations (arrangements) of the packaging order, orientation, and position of the objects within that group. For example, Figure 20 Container 240 is schematically shown, and the system calculates combinations of packing objects starting with a first object in a first position and orientation (as shown in 242) and combinations starting with a first object in a first position and second orientation (as shown in 244). Combinations starting with a second object in a first position and orientation are shown in 246, and combinations starting with a third object in a first position and orientation are shown in 248. Although not all permutations are shown in the illustration, the system can calculate packing arrangements for all combinations. Figure 20 The diagram at 260 illustrates another level of such combinations that has also been determined.

[0103] The offline system accepts packing requests and responds with a list of feasible packing plans. The system uses simulation to find a cell ordering that generates feasible packing plans. The online system uses the same underlying algorithms as the container height map and handheld posture data to iteratively determine the optimal placement for each next object. Placement planner routines determine the path the articulated arm must take to achieve the desired placement. Handheld posture routines determine how the gripper holds the cell. To place objects in the box and later efficiently pack other objects, the system knows the handheld posture of each object when it is grasped, as discussed herein. The placement planner performs searches in six-dimensional space, and these are all done offline to provide pre-computed paths. In online placement planning mode, the system reacts to the aforementioned placement at each step.

[0104] The container height map constructs a map of the box's contents to find the best placement for each subsequent object. The system is also capable of dynamically replanning in response to the state of the outbound container while it is being packed. This compensates for inaccuracies in the outbound box and placement, and mitigates rolling, displacement, or tipping of units within the outbound container. All packing is performed using a single end effector, as there is no second end effector to remove items. All objects are also placed only once, as it is desirable not to move objects that have already been placed. To safely pack objects and efficiently utilize volume, the system knows how the object is held by the gripper while being held (as discussed herein in terms of handheld posture). This provides not only information about the orientation of the object and the gripper but also information about the object's height.

[0105] The assumptions that may be used in the system include offline information (or possibly determined online measurements) such as: the object mass information is accurate to within 5%, the object size is accurate to within 5 mm, the object shape characteristics (e.g., the object is not spherical) are accurate, and the object is placed such that the largest face (surface) of the object is facing upwards.

[0106] The packing planning routine executes upon receiving an order and evaluates whether each order is compatible with a given box size (from smallest to largest). For each box, the system attempts to create a packing plan for the order and the box. The smallest box used in a successful packing plan is used. If a packing plan cannot be created for any box, the order is sent as an exception because it cannot be packed. As discussed above, the packing planner performs a search over all consecutive packing steps, considering all or almost all possible object arrangements, as well as all possible positions and orientations of the objects within the order. If multiple packing plans exist to fit the objects into the boxes, the plan with the smallest and largest object size is selected. The packing plan may maintain certain constraints, including minimum distances between objects, minimum distances between objects and container walls, maximum object height, stacking only stackable objects, and adherence to constraints on object size and properties.

[0107] Figure 21 The container 125 in which objects 270 and 272 are packed is shown. The distance between objects 270 and 272 (referred to herein as margin) is M. o-o As shown, the distance between the object and the inner wall of the adjacent container is M. o-c As shown. Reference Figure 22 The system employs a strategy of relaxation for certain parameters (such as margins). This involves dynamically adjusting object-to-object margins and object-to-container margins in conjunction with a packaging planning strategy. Figure 22 Figure 280 illustrates that the package volume is relatively small when the margin is large, and increases when the margin is small. The system described in this paper can start calculations with a large margin (conservatively) and iteratively recalculate with a smaller margin until it enters the inflection point region, as shown in Figure 282, where the benefit of increasing volume with decreasing margin is diminished. The margin in this inflection point region is preferred.

[0108] Depending on other aspects, the system provides compensation for the deflection of the gripper (e.g., a flexible vacuum cup gripper). Given a point on the rigid object (a point already detected by the system on the surface where the robot holds the object), the angle at which the system needs to hold the object to keep it horizontal for placement can be calculated. Because the torque applied by the vacuum cup is approximately linearly related to its deflection, the required offset angle can be calculated by finding the roots of a trigonometric equation.

[0109] Figures 23A to 23EA specific object placement process according to one aspect of the invention is illustrated, the specific object placement process involving placing an object into a destination container (such as a shipping box) without causing the object to fall (no falling impact force) and without pushing the object into the box (no compressive force). Specifically, Figure 23A The diagram illustrates how the end effector 134 holds an object 162, to be placed into a transport cassette 125, on a weight-sensing section 117 of an output conveyor 124 at a processing station. The output conveyor 124 includes conventional rollers and a weight-sensing section 117, which comprises rollers 150 mounted at either corresponding end of a torque sensor 152, as discussed above. The torque sensor 152 provides weight information data regarding the force applied to the rollers 150. Figure 23B As shown, when the transport container 125 moves onto the weight sensing section 117, the container is positioned below the end effector 134 and the object 162. The system can determine not only that the transport container 125 is positioned on the weight sensing section 117, but also its position on the roller 150, thus confirming its centering and the size of the container in the conveyor direction. The detected container weight is also used to confirm that the container, which was expected to appear at the station, has actually arrived.

[0110] Figure 23C The image shows object 162 being lowered into box 125 via a programmable motion device including end effector 134. It should be noted that the weight of box 125 will increase the weight of object 162. Figure 23D The end effector is shown leaving object 162 in box 125, and Figure 23E The diagram shows the removal of box 125 from the weight sensing conveyor section 117 of the processing station, thus removing any weight from the weight sensing conveyor section.

[0111] Figure 24 Figure 230 shows a graphical representation of the weight (in kilograms) detected on the weight-sensing conveyor section changing over time (in seconds). As shown, before the transport box 125 is received by the weight-sensing conveyor section 117, the weight detected at the weight-sensing conveyor section 117 (shown at 232) is zero kilograms. Between times t1 and t2, the box 125 moves onto the weight-sensing conveyor section 117, and the weight of the box 125 (e.g., approximately 200g) is detected until time t3, as shown at 234. Between times t3 and t4, an object is placed in the box, and the combined weight of the box and the object is detected (e.g., approximately 245g), as shown at 236. Between times t5 and t6, the box (with the object) moves away from the weight-sensing conveyor section, and the detected weight is confirmed to have returned to zero, as shown at 238.

[0112] Figures 25A to 25CA process control system according to one aspect of the invention is shown, which begins (step 1000) by moving a supply box on a supply station conveyor until the current supply box is on the supply trigger weighing roller (step 1002). The system identifies the supply box and obtains all known information about it (step 1004). Because the current supply box is located on the weighing conveyor section, the weighing conveyor can easily determine the current weight of the current supply box (step 1006). The sensing system then collects sensing data about the current supply box from above, and the processing system identifies as many (visible) objects as possible within the current supply box (step 1008).

[0113] The system then processes the sensing data and determines whether at least one object in the current supply bin has been identified (step 1012). If so, the system accesses the pose permission data, position permission data, and object weight data for each identified object (step 1014). If no object is identified in the current supply bin (step 1010), the system selects the topmost object (step 1014) and presents the selected topmost object to the sensing system as discussed above (step 1016). If the held object can be identified, the system proceeds to step 1012 as discussed above, and the system accesses the pose permission data, position permission data, and object weight data for the held object. If the held object cannot be identified, it is returned to the supply bin or moved to the exception bin (step 1018).

[0114] Having identified at least one object and obtained attitude permission data, position permission data, and object weight data (step 1012), the system then accesses destination box data regarding the packing status of the current destination box and the contents of the destination box (step 1020). The system then selects a chosen object, partly based on the attitude permission data and / or position permission data of each object, as well as the status and contents of the current destination box (step 1022). The system then grasps the chosen object from the current supply box and lifts it (step 1024), and then determines the new weight of the current supply box based on the weight data to confirm that the chosen object (with a known weight) has been removed from the supply box (step 1026). While the object is held by the end effector, the system can then use the sensing unit to determine attitude and orientation data regarding the held object and generate attitude holding evaluation data (step 1028).

[0115] Earlier or at this point, the destination station conveyor brings the current destination container to the weighing section of the conveyor and continues moving until it engages the destination trigger weighing roller (step 1030). Similarly, because the destination station conveyor comprises multiple weighing conveyors, the weight of the destination container is then determined (step 1032). The alignment system then engages to ensure the destination container is pushed against the destination support (step 1034). The destination sensing system then performs a volumetric scan of the destination container (step 1036), and then places the object in the destination container, partially based on attitude data (step 1038), followed by a weight measurement to confirm the object is in the destination container (step 1040). The system then performs a further volumetric scan of the destination container to confirm the object is correctly placed on the destination container (step 1042). The system then returns until its end (step 1044), at which point the process concludes (step 1046).

[0116] Figures 26 to 28 A system 110 is shown that includes a pair of feed conveyors 112, 114, on which a single SKU inventory container 116 is conveyed to a single SKU packing cell system 120 that includes a programmable motion articulated arm 122. Figure 26 and Figure 27 The top view and side view are shown respectively, and Figure 28 A rear perspective view is shown. System 110 also includes a transport container conveyor 164 that provides transport containers 166 to a single SKU packing cell system 120. According to one aspect, the single SKU packing cell system picks individual units from inventory shipping boxes and places them in or on packages suitable for transport. A robot support structure spans two parallel inventory shipping box loops 112, 114 that feed inventory shipping boxes from a shipping box storage system (such as AS / RS) to the cell. System 110 includes a weight-sensing conveyor (as discussed above) located below an articulated arm 122, and support rods and support rails as discussed above. Again, the weight-sensing conveyor and the support rod and rail system can be used with transport containers on the transport conveyor.

[0117] The system provides a system designed to package goods into shipping containers according to various aspects. On one hand, it involves packaging one or more units of a single Stock Inventory Unit (SKU) into one or more shipping containers, and on the other hand, it involves packaging multiple SKUs into one or more shipping containers, as discussed further below.

[0118] According to other aspects, a system involving picking multiple SKUs is provided. The system picks one order unit at a time from a pallet conveying individual units and packs them into a shipping container. Similar to single-SKU systems, multi-SKU systems interface with a packaging fabrication mechanism that feeds containers to the picking units. Inventory units are picked from shuttle pallets (depicted in yellow) that queue on a zero-pressure conveyor. A scanner scans the contents of the shuttle pallets, and another scanner scans the shipping container being packed. As with single-SKU systems, the scanners restore the unit's position and orientation for optimal placement within the container surrounding other units.

[0119] For example, Figure 29 A system 300 including a feed conveyor 302 is shown, on which a multi-SKU inventory shipping box 304 is conveyed to a multi-SKU packing cell system 300 including a programmable motion articulated arm 308. Figure 30 and Figure 31 The side view and top view are shown respectively, and Figure 36 and Figure 37 A rear side view and a rear perspective view of system 300 are shown. System 300 also includes a transport container conveyor 310 that provides transport containers 312 to a single SKU packing cell system 306. According to one aspect, the single SKU packing cell system picks individual units from inventory shipping boxes and places them in or on packaging suitable for transport. A robot support structure spans two parallel feed and transport container conveyors that feed the cells from inventory shipping boxes and transport containers from a shipping box storage system such as AS / RS.

[0120] System 306 includes a weight-sensing transmitter located below the articulated arm 308 (as referenced above). Figure 9 , Figure 10 and Figures 23A to 23E (As discussed above), and as referenced above. Figures 16A to 16D The support rods and support rails discussed. Similarly, the weight-sensing conveyor and the support rod and rail system can be used with transport containers on the transport conveyor. Furthermore, each of the multi-SKU feed containers (e.g., boxes or shipping containers) can include a mark identifying the box or shipping container, and a sensing unit that detects the mark can be positioned along the conveyor to determine the identity and location of each multi-SKU box or shipping container.

[0121] The system includes a sensing system (e.g., 320) mounted above the shipping container of the object to be processed, near the base of the articulated arm 308, facing downward toward the shipping container 304, such as... Figure 30As shown. The perception system 320 may, for example (on its underside), include a camera, a depth sensor, and lights. It also acquires a combination of 2D and 3D (depth) data. The depth sensor can provide depth information, which can be used in conjunction with camera image data to determine depth information about various objects in the view. The lights can be used to remove shadows and facilitate the recognition of object edges, and can be all on during use, or can be lit in a desired order to aid in object recognition. The system uses the image and various algorithms to generate a set of candidate grasping positions for objects in the box, as discussed in more detail below.

[0122] Figures 32A to 32C A feed container is shown moving along the feed conveyor 302 below the sensing unit 347 and the depth and edge detection system 343. When the feed container moves from the input side ( Figure 32A ), in sensing unit 347 and system 343 ( Figure 32B Below and away from sensing unit 347 and system 343 Figure 32C During movement, sensing unit 347 and system 343 use volume and depth sensing to determine or confirm the contents of each feed container. By using such sensing units and systems, the edges, volume, and density of the multi-SKU contents of the feed containers can be determined.

[0123] Figure 33A A view of the shipping container 304 from the sensing unit 320 is shown. The image view shows the shipping container 304 (e.g., on a conveyor), and the shipping container 304 contains objects 324, 325, 326, 327, and 328. Figures 33A to 33C In this system, the objects in each feed hopper are non-homogeneous (multiple SKUs). The system will identify candidate gripping positions on one or more objects and may not attempt to identify gripping positions for objects partially obscured by other objects. Candidate gripping positions can be indicated using a 3D model of the robot's end effector, placed in the location where the actual end effector will be used as the gripping position. For example, a gripping position may be considered good if it is close to the object's center of mass to provide greater stability during gripping and transport, and / or if it avoids locations on the object where a good vacuum seal may not be possible (such as lids, seams, etc.).

[0124] The sensing system 320 includes a sensing unit, a scanning and receiving unit, and an edge detection unit for capturing various characteristics of a selected object throughout the box. Similarly, Figure 33A The diagram shows a view from a capture system, which, according to one embodiment, may include a set of different objects 324, 325, 326, 327, 328. Figure 33B The scanned volume V is shown in the figure. 324 or density D324 The difference is analyzed and compared with recorded data on items identified by identification tags provided by the detection system via the SKU sensing system, or recorded object data. Specifically, the scanned volume is compared with the volume of the identified SKU multiplied by the known number of objects in the box. After picking, the volume is scanned again ( Figure 33C To confirm the volume of the object being picked.

[0125] Depending on other aspects, the scanning and receiving unit can also be used to determine the density of the collection of objects in the box, which is compared with a known density of the identified SKU multiplied by a known number of objects in the box, thereby determining the mass and volume of the objects. Volume data can be obtained, for example, using any of a LiDAR scanner, a pulse-time-of-flight camera, a continuous-wave time-of-flight camera, a structured light camera, or a passive stereo camera.

[0126] Figures 34A to 34C A transport container is shown moving along a transport conveyor 310 below a sensing unit 347 and a depth and edge detection system 345. When the transport container moves from the input side ( Figure 34A ), in sensing unit 347 and system 345 ( Figure 34B Below and away from sensing unit 347 and system 345 Figure 34C During movement, sensing unit 347 and system 345 use volume and depth sensing to determine or confirm the contents of each shipping container (e.g., if it is already partially filled). By using such sensing units and systems, the edges, volume, and density of the multi-SKU contents of the shipping containers can be determined.

[0127] According to some aspects, the system may additionally employ an edge detection sensor, which is used (again, in conjunction with processing system 350) to detect the edges of any object in the box, for example, using data on any of the following: intensity, shadow detection, or echo detection. The system can be used, for example, to determine any of the following: size, shape, and / or contour, to help confirm the number of objects in the box. In some aspects, the system can identify specific objects in the box and confirm their shape and size through such edge detection. Therefore, the above-described system can be used to confirm the number of objects packed into a container, and in some aspects, to initially confirm the number of objects in a shipping container.

[0128] Similarly, the operation of the above system is coordinated with a central control system 200, which also communicates (e.g., wirelessly) with the articulated arm 308, sensing system 320, and feed conveyor 302 and transport container conveyor 310. The system determines the UPC associated with the supplier box and the outbound destination for each object based on a symbol string. The central control system 400 consists of one or more workstations or central processing units (CPUs). For example, the mapping between UPCs or mailing tags and outbound destinations is maintained by the central control system in a database called a manifest. The central control system maintains the manifest by communicating with a warehouse management system (WMS). The manifest provides an outbound destination for each inbound object. Scanners are positioned above each inventory shipping box picking location to locate the pick point. Cells are coordinated via the WMS / WCS with transport container distribution mechanisms (such as automated boxing systems or cardboard strip distributors) that feed transport containers into the picking cells. Conveyors feed transport containers into the cells.

[0129] Figures 35A to 35C The grab evaluation detection unit 338 in the packaged cell 306 is shown (some components have been removed for clarity). Detection unit 338 includes an upper detection unit, which is as follows: Figure 35A The image points downwards towards the object 335 held by the end effector 334. The detection unit 338 further includes an intermediate detection unit, which is as follows: Figure 35B The object 335, which is generally horizontally pointed to by the end effector 334, is held by the end effector 334; and the lower detection unit, as shown in the figure... Figure 35C The object 335 is pointed upwards and held by the end effector 334.

[0130] Similarly, and as discussed above, to compensate for the way the robot holds the unit for placement, a scanner (not shown) scans the geometry of the unit while it is held by the robot to properly place the unit in the transport container. A third scanner above the transport container (placement area) scans the transport container after placement to provide feedback on the placement quality and initiate replacement if necessary.

[0131] Inventory shipping containers queue on the inbound inventory conveyor and stop at a right-angle belt transfer machine within robot reach. A series of appropriately sized empty transport containers are then fed into the cells via a belt conveyor in an order matching the SKU arrival sequence. As items are removed from the inventory shipping containers, they are scanned, and the cell's orientation and position are determined for placement. By compensating for the cell's known position and orientation relative to the gripper, the cell is placed on and within the transport container.

[0132] A scanner above the cardboard detects cases where the unit placement is unsuitable for packing (i.e., outside the shipping container). In such cases, the unit is either re-grabbed and repositioned or directed to the Quality Assurance (QA) station. A barcode on the shipping container is scanned, or the shipping container is marked with a barcode to associate the unit and shipping container with the corresponding order. During the placement of units from one inventory shipping container, the contents of a second inventory shipping container are scanned in preparation for the next pick. A conveyor belt transfers cardboard with individual inventory units onto a conveyor belt that feeds them to a final packing machine (such as a shrink wrapper and a box or carton sealer). If no further single pick from the inventory shipping container is required, the container is transferred to the outbound conveyor. Otherwise, the inventory shipping container is held, and another pick is performed. The process is repeated when picking begins on the second inventory shipping container.

[0133] In addition to the physical components, items need to be requested in the most suitable packing order before the unit arrives at the station. Orders to be shipped are initiated by querying the packing planner from the warehouse's WMS. The WMS provides a list of all SKUs in the order, along with their dimensions and weights. The packing planner determines the compatible SKU sequence for packing, which may include options for requesting multi-box orders for large unit quantities; or, alternatively, recommending manual packing.

[0134] The operational concept of a multi-SKU system is as follows: Items corresponding to an order arrive at cells in the shuttle pallet in the order specified by the packing planner. When a new order begins, a transport container of a pre-determined size by the packing planner is conveyed to the cell via a belt conveyor. As the unit is picked from the shuttle pallet, it is scanned by a posture scanner, which determines the unit's posture and orientation relative to the gripper. The known posture of the unit relative to the gripper is compensated for, and the unit is placed on or in the transport container with the position and orientation determined by the picking plan. After placing the unit, the picking plan is recalculated based on the actual position of the unit on or in the transport container. The packing plan positioning is then adjusted to compensate for the actual placement. Picking and placement are repeated, interspersed with recalculation of the packing plan, until all units have been picked and placed in the transport container.

[0135] Once all units have been placed, the full shipping container is transferred via belt conveyor to the final packing machine (such as a shrink wrapping machine and a cartoning or sealing machine). If at some point the actual placement no longer allows for the planned placement, but shrink wrapping is otherwise acceptable, the partially filled shipping container is transferred via belt conveyor to the final packing machine (such as a shrink wrapping machine and a cartoning or sealing machine). Any remaining unpacked units are returned to the mini-shuttle for later transport in another box. If items cannot be packed because the measured SKU size is larger than expected, these units are marked for QA. If at some point unit placement results in items being outside the cardboard area, the items are re-grabbed and repositioned.

[0136] Figure 36 and Figure 37 The diagram shows a rear side view and a rear perspective view of system 300, which includes a feed conveyor 302 on which multi-SKU inventory shipping boxes 304 are conveyed to a multi-SKU packing cell system 306 including a programmable motion articulated arm 308. System 300 also includes a transport container conveyor 310 that provides transport containers 312 to the multi-SKU packing cell system 306. According to one aspect, the single-SKU packing cell system picks individual units from the inventory shipping boxes and places them in or on packaging suitable for transport. A robot support structure spans two parallel conveyors 302, 310, which, under the control of one or more processing systems 200, feed transport containers to cells and inventory shipping boxes from a shipping box storage system (such as AS / RS).

[0137] System 300 includes a weight-sensing transmitter located below the articulated arm 308 (as referenced above). Figure 9 , Figure 10 and Figures 23A to 23E (As discussed above), and as referenced above. Figures 16A to 16D The discussed support rods and support rails. Similarly, the weight-sensing conveyor and the support rod and rail system can be used with both the input container on the feed conveyor and the transport container on the transport conveyor. The system may include multiple systems 306 positioned along a pair of feed conveyors and transport container conveyors.

[0138] The scanning and receiving unit can also be used to determine the density of the collection of objects in the box, which is compared with the known density of the identified SKU multiplied by the known number of objects in the box, thereby determining the mass and volume of the objects. Volume data can be obtained, for example, using any of a LiDAR scanner, a pulse-time-of-flight camera, a continuous-wave time-of-flight camera, a structured light camera, or a passive stereo camera.

[0139] According to other aspects, the system may additionally employ an edge detection sensor, which is used (again in conjunction with processing systems 100 and 200) to detect the edges of any object in the box, for example, using data on any of the following: intensity, shadow detection, or echo detection. The system can be used, for example, to determine any of the following: size, shape, and / or contour, to help confirm the number of objects in the box. In some aspects, the system can identify specific objects in the box and confirm their shape and size through such edge detection. Therefore, the above-described system can be used to confirm the number of objects in the box, and in some aspects, to initially estimate the number of objects (for a single SKU) in the box, and / or to confirm recorded data for any particular SKU.

[0140] During use, the end effector of any of the above systems may include either a pressure or airflow sensor within the end effector, and may record pressure and / or flow rate when no object is held and the vacuum is on, and when a known object is held under vacuum. Using this baseline information, the system can determine that the end effector needs cleaning, for example, debris that may have accumulated at the opening of the vacuum cup of the end effector. For example, Figure 38 An end effector 400 is shown, having a vacuum cup 402 communicating with a vacuum source via an internal channel 404. A sensor 406 is disposed within the channel 404 for measuring either pressure or airflow within the channel.

[0141] When debris 410 partially blocks the vacuum cup opening (such as...) Figure 38 (As shown) or debris 412 completely blocks the vacuum cup opening (as shown) Figure 39 As shown, the system may detect abnormal airflow or pressure readings during normal operation (e.g., between gripping or while gripping a known object). For example, if the system has recorded static values ​​for each of the pressure and airflow when the vacuum is on and no object is being gripped, a different set of static readings may indicate that the end effector opening needs cleaning. Similarly, if the system has recorded a set of known object gripping values ​​for each of the pressure and airflow when the vacuum is on and an object is being gripped, a different set of readings when gripping the same or similar object may indicate that the end effector opening needs cleaning. Depending on other aspects, debris may enter the vacuum cup against the internal mesh screen (again, causing partial or complete vacuum blockage).

[0142] Figure 40 The above reference is shown. Figures 6 to 28The system discussed includes a single SKU packing station 420 that receives a feed container 116 on feed conveyors 112, 114, and an articulated arm 122, and a transport container 126 on a transport container conveyor 124, as discussed above. The single SKU packing station 420 also includes a waste bin 430 into which the articulated arm can deposit (e.g., drop or blow) debris from a vacuum cup of an end effector 400. Figure 41 An end effector 400 is shown for an articulated arm 122 positioned above a trash can 430. The trash can 430 may include a transmitter 452 and a detector 454 along opposite top edges to detect whether any debris has fallen or been blown into the trash can 430.

[0143] Similarly, Figure 42 The above reference is shown. Figures 29 to 37 The system discussed includes a multi-SKU packing station 440 that receives a feed container 304 on a feed conveyor 302 and an articulated arm 308, and a transport container 312 on a transport container conveyor 310, as discussed above. The unit station 440 also includes a waste bin 430 into which the articulated arm can deposit (e.g., drop or blow) debris from a vacuum cup of an end effector 400. Figure 43 An end effector 400 of an articulated arm 308 positioned above a trash can 450 is shown. Similarly, a transmitter 452 and a detector 454 may be included along opposite top edges to detect whether any debris has fallen or been blown into the trash can 430.

[0144] When the vacuum cup 402 of the end effector 400 is positioned above the trash can (e.g., 430), any debris can be dislodged by turning off the vacuum. Further reference Figure 44 The vacuum source 414, connected to the vacuum cup 402 via vacuum hose 416 (in the example disclosed above), can be reversed (or switched) to cause a positive pressure to be delivered to the vacuum cup, thereby blowing any debris in the vacuum cup into the bin. When any debris enters the bin 430, a response signal from the transmitter / detector arrays 452, 454 is sent to the control system. Figure 45 A trash can according to another aspect is shown, the trash can including a brush pad 422, and a vacuum cup 402 of an end effector 400 can be dragged away from the trash can 430 along the brush pad to further facilitate the removal of debris from the vacuum cup. The use of the brush pad 422 is preferably combined with providing positive air pressure to inhibit debris from accumulating at the brush pad 422.

[0145] The process control system can determine during use the need to clean the vacuum cup of the end effector, for example, for debris that may have adhered to the vacuum cup due to vacuum, adhesive on the debris, or electrostatic charge. Specifically, and refer to... Figure 46 Upon startup (or when the system knows the vacuum cup is clean or contains any debris), the system can record (step 1100) the pressure and airflow at the end effector when the vacuum is activated and no object is being grasped. OV F OV The system can then record (step 1102) the pressure and airflow at the end effector as it grasps each of a plurality of different objects. KO1 F KO1 P KO2 F KO2 P KO3 F KO3 P KO4 F KO4 (etc.). Each of the different objects can be a commonly encountered object, such as a box weighing less than 1 lb, a box weighing 2 to 2.5 lb, a box weighing 10 to 12 lb, and a box weighing 20 to 22 lb. Record these values ​​and then compare them with the values ​​measured during use.

[0146] For example, when the system is not currently grasping an object, a vacuum is applied and pressure and airflow are detected (step 1104). If the detected pressure or airflow is not within the recorded open vacuum value (P... OV F OV If the pressure or airflow is within the range (e.g., ±5%), the system will set a clean vacuum cup indicator. Similarly, during use, when the system is currently gripping a known object typically encountered via vacuum, pressure and airflow are detected (step 1106). If the detected pressure or airflow is not within the known object value (P) recorded for that object... KO F KO If the value is within the range (e.g., ±5%), the system will set the clean vacuum cup indicator. The process continues during the use and operation of the articulated arm (step 1108) until the program ends (step 1110).

[0147] When the vacuum cup cleaning indicator has been set, the system can begin the vacuum cup cleaning routine (step 1200) between object handling operations by first moving the vacuum cup of the end effector past the trash can (step 1202), as shown. Figure 47AAs shown. The system can then shut off the vacuum (step 1204) and detect (step 1206) whether any debris has fallen into the trash can. If yes, the system terminates. If not, the system can switch to forced air (blower) by switching to the blower source or reversing the vacuum system 414 (step 1208) to provide forced positive pressure air to the vacuum cup 402 through the hose 416. The system can then detect (step 1210) whether any debris has fallen into the trash can. If yes, the system terminates. If not, the system can move the vacuum cup away from the trash can by dragging it along the brush pad 422 (optionally, engaging the blower) (step 1212), as... Figure 47B As shown. The system can then detect (step 1214) whether any debris has fallen into the trash can. If so, the system terminates. If not, the system can engage a blower to blow air along the length of the brush pad (step 1216) to remove any debris that may have accumulated on the brush pad 422. The system can then detect (step 1218) whether any debris has fallen into the trash can. If so, the system terminates. If not, the system can determine whether this is the first cleaning attempt for the event (step 1220), and if not, the system sets a service flag indicating that the end effector and vacuum cup require maintenance (step 1228).

[0148] If the system determines that this is the first cleaning attempt for the event (step 1220), the system can rotate the end effector 180 degrees along its longitudinal length (step 1222), effectively flipping the end effector so that the side previously opposite the brush pad now faces the brush pad. The system can then drag the now-flipped vacuum cup (optionally, in conjunction with a blower) along the brush pad 422, thereby removing it from the trash can (step 1224), as... Figure 47C As shown. The system can then detect (step 1226) whether any debris has fallen into the trash can. If yes, the system terminates. If not, the system can engage a blower to blow air along the length of the brush pad (step 1228) to remove any debris that may have collected on the brush pad 422. The system can then detect (step 1230) whether any debris has fallen into the trash can. If yes, the system terminates. If not, the system can set a flag indicating that the system's end effector needs maintenance (step 1232), and the vacuum cup cleaning routine terminates (step 1234). These processes and systems can be used with each of the single-SKU and multi-SKU systems disclosed above.

[0149] Similarly, the operation of the aforementioned system is coordinated with central control systems 100 and 200, which communicate with articulated arms, sensing systems, conveyors, alignment systems, and waste removal devices (e.g., wirelessly). The system determines the UPC associated with the supplier box and the outbound destination for each object based on a symbol string. Central control systems 100 and 200 consist of one or more workstations or central processing units (CPUs). For example, the mapping between UPCs or mailing tags and outbound destinations is maintained by the central control system in a database called a manifest. The central control system maintains the manifest by communicating with a warehouse management system (WMS). The manifest provides an outbound destination for each inbound object. Scanners are positioned above each inventory shipping box picking location to locate the grab point. Cells are coordinated via the WMS / WCS with shipping container distribution mechanisms (such as automated box-making systems or cardboard strip distributors) that feed shipping containers into the picking cells. Conveyors feed shipping containers into the cells.

[0150] Similarly, and as discussed above, to compensate for the way the robot holds the unit for placement, a scanner (not shown) scans the geometry of the unit while it is held by the robot to properly place the unit in the transport container. A third scanner above the transport container (placement area) scans the transport container after placement to provide feedback on the placement quality and initiate replacement if necessary.

[0151] Those skilled in the art will recognize that several modifications and changes can be made to the above-disclosed embodiments without departing from the spirit and scope of the invention.

Claims

1. An automated packaging system for placing multiple objects into a transport container, the automated packaging system comprising: A supply box receiving conveyor for receiving supply boxes at a supply station, the supply box receiving conveyor including a sensing device for determining the range of the supply box's movement along the conveyor in the conveyor direction; A detection system is configured to detect, in response to the position of the supply box on the receiving transmitter, that multiple objects within the supply box are aligned by an alignment system; An object posture permission assessment system is used to generate posture permission data regarding whether any of the plurality of objects in the supply box at the supply station may move after being placed in the transport container. An object selection system, the object selection system being used to select one object from the plurality of objects in response to the attitude permission data and place it into the transport container with a selected orientation and attitude; as well as A programmable motion device for grasping and acquiring a selected object from the plurality of objects at the supply station and placing the selected object into the transport container in a selected orientation and posture.

2. The automated packaging system according to claim 1, wherein the plurality of objects are homogeneous.

3. The automated packaging system according to claim 1, wherein the plurality of objects are heterogeneous.

4. The automated packaging system according to any one of claims 1 to 3, wherein the object posture permission assessment system further generates posture permission data regarding whether any of the plurality of objects at the supply station is deformable and whether its shape may change when placed in the transport container.

5. The automated packaging system of claim 1, further comprising a posture retention evaluation system for providing posture retention evaluation data regarding the posture and orientation of the selected object relative to the end effector when held by the end effector of the programmable motion device.

6. The automated packaging system of claim 5, further comprising a posture adaptation system for adapting the selected object to the posture of the end effector by adjusting the trajectory of the programmable motion device in response to the posture holding evaluation data to partially place the selected object into the transport container in a selected orientation and posture.

7. The automated packaging system of claim 1, further comprising a volume sensing system for providing volume data about the shipping container.

8. The automated packaging system of claim 7, wherein the volume sensing system determines whether the selected object has been moved after the selected object is placed in the transport container.

9. The automated packaging system of claim 1, wherein the transport container is a cardboard box.

10. The automated packaging system of claim 1, wherein the transport container is a transport pallet.

11. An automated packaging system for placing multiple objects into a transport container, the automated packaging system comprising: An object posture permission assessment system is used to generate posture permission data regarding whether any of the plurality of objects at the supply station is deformable and whether its shape may change when placed in the transport container. An object selection system, the object selection system being used to select a chosen object from the plurality of objects in response to the attitude permission data, and to place it into the transport container with a selected orientation and attitude; as well as A programmable motion device for grasping and acquiring a selected object from the plurality of objects at the supply station and placing the selected object into the transport container in the selected orientation and posture in response to the posture permission data.

12. The automated packaging system of claim 11, wherein the plurality of objects are homogeneous.

13. The automated packaging system of claim 11, wherein the plurality of objects are heterogeneous.

14. The automated packaging system according to any one of claims 11 to 13, further comprising a supply box receiving conveyor for receiving supply boxes containing the plurality of objects, the supply box receiving conveyor including a sensing device for determining the range of the supply box's advance along the supply box receiving conveyor in the conveyor direction.

15. The automated packaging system of claim 14, further comprising an alignment system for aligning the supply box with the support in a direction transverse to the conveyor direction.

16. The automated packaging system of claim 11, wherein the system further comprises a posture retention evaluation system for providing posture retention evaluation data about the posture and orientation of the selected object relative to the end effector when held by the end effector of the programmable motion device.

17. The automated packaging system of claim 16, further comprising a posture adaptation system for adapting the selected object to the posture of the end effector by adjusting the trajectory of the programmable motion device in response to the posture holding evaluation data to partially place the selected object into the transport container in a selected orientation and posture.

18. The automated packaging system of claim 11, further comprising a volume sensing system for providing volume data about the shipping container.

19. The automated packaging system of claim 18, wherein the volume sensing system determines whether the selected object has been moved after the selected object is placed in the transport container.

20. The automated packaging system of claim 11, wherein the transport container is a cardboard box.

21. The automated packaging system of claim 11, wherein the transport container is a transport pallet.

22. An automated packaging system for placing multiple objects into a transport container, the automated packaging system comprising: A supply box receiving conveyor, which is used to receive supply boxes at the supply station; A supply detection system, wherein the supply detection system is used to detect multiple objects within the supply box; An object selection system for selecting one object from the plurality of objects and placing it into the transport container; A programmable motion device, the programmable motion device including an end effector for grasping and acquiring a selected object from the plurality of objects at the supply station; A transport container receiving conveyor, which is used to receive destination containers at the destination station; A destination detection system, wherein the destination detection system is used to detect the volume inside the transport container; An attitude retention assessment system is used to provide attitude retention assessment data about the attitude and orientation of the selected object relative to the end effector when it is held by the end effector; An attitude adjustment system, which, in response to attitude maintenance evaluation data, adjusts at least the attitude or orientation of the selected object by adjusting the trajectory of the programmable motion device before placing the selected object into the transport container.

23. The automated packaging system of claim 22, wherein the plurality of objects are homogeneous.

24. The automated packaging system of claim 22, wherein the plurality of objects are heterogeneous.

25. The automated packaging system according to any one of claims 22 to 24, further comprising an object posture permission assessment system for generating posture permission data regarding whether any of the plurality of objects at the supply station is deformable and whether its shape may change when placed in the transport container.

26. The automated packaging system according to any one of claims 22 to 24, further comprising an object posture permission assessment system for generating additional posture permission data regarding whether any of the plurality of objects at the supply station may move after being placed in the transport container.

27. The automated packaging system of claim 22, wherein the supply box receiving conveyor includes a sensing device for determining the range of the supply box's movement along the supply box receiving conveyor in the conveyor direction.

28. The automated packaging system of claim 27, further comprising an alignment system for aligning the supply box with the support in a direction transverse to the conveyor direction.

29. The automated packaging system of claim 22, further comprising a volume sensing system for providing volume data about the transport container.

30. The automated packaging system of claim 29, wherein the volume sensing system determines whether the selected object has been moved after the selected object is placed in the transport container.

31. The automated packaging system of claim 22, wherein the transport container is a cardboard box.

32. The automated packaging system of claim 22, wherein the transport container is a transport pallet.

33. The automated packaging system of claim 22, further comprising a weight sensing device for determining whether the selected object is placed in the transport container.

34. The automated packaging system of claim 33, wherein the weight sensing device further determines whether the programmable motion device applies a compressive force to the transport container.

35. The automated packaging system of claim 33, wherein the weight sensing device is capable of determining the weight of the selected object that is significantly less than the weight of the transport container when it is empty.

36. The automated packaging system of claim 22, further comprising a debris cleaning detection system for determining whether the vacuum cup at the end effector of the programmable motion device needs cleaning.

37. The automated packaging system of claim 36, further comprising a debris removal system for removing debris from the vacuum cup in response to the debris cleaning detection system.

Images