ES2949257T3 - Packaging Robot - Google Patents

Abstract

A product packaging robot is provided comprising a plurality of heads for picking up product items from a plurality of pick-up locations, and moving the product items and dropping them off at a plurality of delivery locations. The plurality of heads move together in a first direction on a common rail. The robot may also comprise a range imaging camera and a controller to dynamically detect the location of delivery points. (Automatic translation with Google Translate, without legal value)

Description

DESCRIPTION

Packaging Robot

Technology field

The present invention generally relates to a packaging robot.

Background

Traditional balers for packaging products are manually operated and involve labor-intensive procedures. Product packaging work tends to be repetitive in nature, and may place personnel packaging the product in close proximity to dangerous machinery. When balers are automated, benefits can be provided such as safety by removing personnel from the proximity of machinery, increasing packaging speed, and increasing product yield since automated packaging can operate continuously. However, packaging automation commonly involves complex machinery that is also very expensive.

Document EP1803665 describes a collection head for transferring products supported on a basic structure on which a plurality of collection devices for collecting products are capable of moving and independently varying the separation between the collection heads.

WO 2017/127292 describes an apparatus for picking up articles supported on a bar and a plurality of pick-up heads that move independently along the bar and therefore vary the spacing between the pick-up heads.

Document EP0727355 describes an apparatus for picking up articles supported on a bar and a plurality of pick-up heads that can vary the pitch spacing between the pick-up heads.

It is an object of the present invention to provide an improved packaging robot, or at least to provide a useful option.

Summary of the invention

The present invention consists of a robot for packaging products that comprises;

a plurality of pick locations for the product, the pick locations arranged in an array, an exit area for the product, and

a plurality of heads, arranged in an array, for picking, holding and releasing an item of product, wherein the product packaging robot further comprises a controller for controlling movement of the heads,

wherein the plurality of heads move together as a set in a first direction between the array of pickup locations and the output zone,

wherein each head is independently movable in a second direction, one direction being orthogonal to the first direction, to vary the spacing of the plurality of heads in the second direction, wherein each of the plurality of pick locations further comprises a orientation module for presenting a product item in a desired pick orientation with respect to at least one feature of the product, each orientation module having

at least one orientation detection camera for detecting and determining the location of said at least one feature, and

an orientator configured to manipulate said product item in a desired pick orientation.

According to the invention, the orientation module is configured to indicate to the controller, when the feature is detected by the orientation detection camera, that a product item is almost ready to be picked up and the controller is configured to make a wait decision. or not waiting for an item to be almost ready before moving the heads to the output area.

According to a further aspect, each head is capable of unequally varying the spacing between the plurality of heads in the second direction.

According to the further aspect, the orientation module may indicate to the controller when a product item is in a desired pick orientation and ready to be picked.

According to a further aspect, the orientator multiaxially rotates the product article until the feature is detected by the orientation detection camera, after which the location of the feature with respect to the product and the feature with respect to which it is determined. the product must be oriented is selected from: a) a physical anatomical characteristic,

b) a portion more visually similar to a desired color,

c) a desired pattern, or

d) a desired texture.

According to a further aspect, the orientation of the product is determined through the at least one orientation detection camera with respect to a plurality of features, one of said features being a primary feature with respect to which the product is placed in the desired pickup orientation.

According to an additional aspect, the counselor understands:

a pair of laterally spaced controlled drive belts configured to support an item of product,

a boundary that at least partially surrounds the product item to restrict translational movements of the product,

wherein the plurality of drive belts may drive at different speeds to multiaxially rotate the product article.

According to a further aspect, the output area is provided with one or more receptacles for receiving the provided product, the one or more receptacles defining one or more desired locations for placing items of product, and the desired locations for placing items of product are the exit points for the product. Receptacles are trays or boxes, or baskets, or bags.

The product packaging robot further comprises a destination detection video camera to detect the location of the exit points, within the exit zone.

According to a further aspect, the target detection camera is an interval imaging camera selected from one of:

a) a structured light chamber,

b) an RGBD camera,

c) a stereo camera pair, or

d) a time-of-flight camera

According to a further aspect, the number of heads is less than the number of pick locations for the product. According to a further aspect, the movement of the heads in the second direction to vary the spacing of the heads is such that it allows the heads to each move between multiple pick-up locations and multiple exit points.

According to a further aspect, packaging a product item involves:

a) position the head array in the first direction so that at least one head is aligned in the first direction with a product item at a pick location,

b) independently varying the spacing of one or more heads in the second direction to align the at least one head with the product article in the second direction,

c) pick up the product item,

d) transporting the item of product in the first direction to the exit area so that the item of product is aligned in the first direction with a desired exit point, and

e) release the product item to the desired exit point.

According to a further aspect, before releasing the product items, the spacing of one or more heads is varied in the second direction.

Additionally or alternatively, packaging a product item also involves a rotation of the desired head about an axis parallel to the third direction to rotate the product item to a desired orientation. According to a further aspect, after receiving an almost ready signal, the controller expects a product item to be handled into the desired pick orientation depending on:

a) the number of heads that have picked up and hold an item of product,

b) the number of pick-up points that have a product item ready for pick-up, and

c) the number of pickup points that have a product item almost ready for pickup.

According to a further aspect, after receiving an almost ready signal, the controller ignores the signal and releases the product items currently held by the heads when:

a) all heads are already holding a product item, or

b) the number of items ready for picking is greater than or equal to the number of heads that are not holding an item of product.

According to a further aspect, the controller bases the decision to wait for an almost ready item of product to be ready on an estimated comparison of the time until the item of product is oriented in the desired pick orientation and is ready to be picked and the time required to release the product items currently held by the heads and return to the pick location.

The term “comprising” as used in this description and the claims means “consisting at least in part of”. When interpreting each statement in this specification and claims that includes the term “comprising”, other characteristics other than those above may also be present. Related terms such as “comprises” and “comprises” should be interpreted in the same way.

The invention consists of the above and also contemplates constructions covered by the wording of the appended claims, of which the following provides only examples.

Brief description of the drawings

Figure 1 is a perspective view of the product packaging robot showing the product accumulator, the receptacle conveyor and the product packaging area with products packed in some receptacles. Figure 2 is a side view of the product packaging robot. product of Figure 1,

Figure 3 is a partial view of the product packaging robot of Figure 2, showing only the product packaging area with the pick locations, the multi-head robot, and the output area with receptacles therein. Figure 4 is a view of the accumulator for the product, which feeds each of the orientation modules, showing the separators to channel the product into individual rows,

Figure 5 is a view of the accumulator and product orientation unit, showing individual orientation modules, some with products located in product feeds or pick locations, Figure 6 is a view in one module orientation, which shows the product feeding, pick location, orientation and the orientation detection camera that is directed towards the pick location,

Figure 7 is an end view of the multi-head and orientation robot modules, showing the individual heads separated between different pick locations, with heads that have picked up an item of product, in the process of picking up an item of product. product and without a product item, Figure 8 shows the assembly of Figure 7 showing a different separation of the individual heads where all the heads are all grouped towards the leftmost pick locations,

Figure 9 is a perspective view of the product packaging area showing the multi-head robot and its frame, the drive assemblies for the multi-head robot and the individual heads, the target detection camera located above the area outlet and a number of receptacles that are packed with products,

Figure 10 is a side view of the multi-head robot and product orientation unit, showing the multi-head robot located over the picking locations, one head having picked up a product item, and another head in the picking procedure. a product item,

Figure 11 shows the assembly of Figure 9 at a different angle,

Figure 12 shows a receptacle or tray partially filled with products, the contoured pockets of the tray visible at the unfilled exit points, and where the product items in the tray are all aligned with their stalks facing parallel directions,

Figure 13a shows a view of a product item that is an apple with the apple stem visible,

Figure 13b shows another view of the product article of Figure 13a, where the apple calyx is visible.

Detailed description

Referring to Figures 1 to 2, a product packaging robot 1 is shown. The product packaging robot 1 has an accumulator 2 for producing items 3 and a product packaging area 4. Next, the product enters the packaging area 4 through the accumulator 2, it is then packaged in the packaging area 4. Within the packaging area, the product passes from a number of pick locations 14, to an exit zone 5 where it is packaged onto the receptacles 7. The receptacles preferably have a number of desired locations for placing the product 3, which are the exit points 6 for the product items 3.

The receptacles can be located in the packaging area by a receptacle conveyor 8 that passes below the accumulator 2 and the product packaging area 4. The product that is packed in the receptacles 7 then leaves the packaging area on the conveyor 8. .

The product enters the product packaging robot 1 through the accumulator 2. The products can be fed to the accumulator 2 manually or by an external power source, such as a conveyor (not shown) or any other suitable means. The accumulator has an accumulator conveyor 9 that moves the product 3 to a number of separators 10 that are spaced apart to create channels between them. The width of the channels is such that it allows the passage of only one product at a time, forming the product in individual rows as seen in Figure 4.

The accumulator conveyor 9 may also preferably include a device for spreading the product substantially uniformly in the separators 10.

For example, with reference to Figure 4, a plurality of brushes 40 may be provided to guide or distribute the product as it passes along the accumulator conveyor 9.

The brushes 40, for example, may be attached to an endless belt so that they propel the product through the conveyor 9 as the belt rotates. To achieve substantially uniform distribution of product in each lane of the separator, the endless belt can rotate in both directions. For example, the brushes can move in one direction for a predetermined period of time and then in the opposite direction.

It has been found that this procedure can achieve an acceptable improvement in leveling the product flowing into each lane of the separators 10.

From the accumulator 2, the product is passed to the product orientation unit 11, as shown in Figures 5 and 6. The product orientation unit 11 has several orientation modules 12 corresponding to the number of product rows formed by the separators 10. Each orientation module 12 has a product feed 13 to provide items of product one at a time from the accumulator to a plurality of pick locations 14 where the product must be reoriented.

Each orientation module 12 includes at least one orientation detection video camera 15 directed toward its pick location 14. To change the orientation of the product article, an orientation 16 is provided at each pickup location 14. The orientation comprises a mechanism to multiaxially rotate a product item.

A boundary 18 is provided around at least a portion of the pick location 14 to restrict translational movement of the product article during its rotation. In the preferred form, the multiaxial rotation of the rotation means 16 is provided with a pair of separate belts 19 on which the products rest. The belts 19 can be driven at different speeds and in different directions, to multiaxially rotate the product article in any desired orientation.

The product packaging area 4 is provided with a multi-head robot 20 to transfer the product between the orientation modules 12 and the trays or receptacles 7.

As shown in Figures 7-11, the multi-head robot 20 has an array of individual heads 21, each for picking, transporting and dropping an item of product.

As particularly shown in Figures 7-11, the head array 21 is preferably arranged in a straight line. Alternatively, the heads 21 may be arranged in a stepped array, or in an irregular array.

The individual heads 21 of the multi-head robot 20 are fixed and travel as a unit between the product orientation unit 11 and the receptacles 7, along the XX axis as seen in Figure 1. Each head 21 can move independently in the remaining YY and ZZ directions, and may preferably be capable of rotating in the W direction around the ZZ axis.

As seen in Figure 9, the multi-head robot 20 is driven along a frame 22 in the XX direction by the multi-head drive assembly 23, which has a series of motors 24 and belt drives 25. The multi-head robot 20 then has a series of head drive assemblies 26, corresponding to the number of heads 21 of the multi-head robot, to drive each of the heads individually in the YY lateral direction.

The head drive assemblies 26 comprise a motor 27 and a belt 37 that extends through the frame 22. Each motor 27 and belt drive 37 engage one of the heads 21.

Each head 21 is also provided with a rotary drive 28 to change the orientation of the head around the ZZ axis, providing rotation in the W direction as shown in Figure 1.

At the end of each head 21 there is a product holder 29 for picking up, holding and releasing the product items 3. The product holder may have means for gripping a product item, or as in the preferred configuration the product holder is in the shape of a suction cup 30.

Vacuum is applied to the suction cup 30 through pneumatic line 31. The distal end of the suction cup 30 is a surface formed to allow a seal to be formed with the surface of a product article. When a suction cup 30 is located against a product article 3, and vacuum is provided to the suction cup 30, the product article can be held by the suction cup in order to be transported for packaging. Upon arrival at a desired exit point, the vacuum may be interrupted, or positive pressure may be provided to the suction cup to release the product article 3. Each suction cup 30 may be provided with a foldable pad portion 32 to allow the suction cup expand or contract along the ZZ axis of the head 21 when different pressures are provided to the suction cup 29. It is envisioned that the physical characteristics of the suction cups 30 can be adjusted to accommodate various types of products being packaged.

The receptacles 7 in which product items 3 are to be packed in the product packaging area 4 are supplied by the receptacle conveyor 8 running below the accumulator 2 and the multi-head robot 20. The receptacles 7 may be in the form of trays as shown in Figure 12, which have a series of depressions or receptacles 34 for receiving items of product.

Alternatively, the trays may have substantially flat internal surfaces, without depressions for receiving products. Furthermore, alternatively, the receptacles 7 may be in the form of boxes with upwardly extending sides and a substantially flat base, or baskets, or bags.

A target detection video camera 33 is provided to the product packaging robot. As shown in Figure 9, the target detection camera 33 may be located on the multi-head robot 20, to move with the head assembly 21. Alternatively, the target detection camera may be fixed on the frame 22 or another component so that it remains stationary while the multi-head robot moves.

Preferably, the destination detection chamber 33 is located such that it is substantially over the exit area 5 and the receptacles 7 that are to be packaged with products.

The target detection camera 33 may be any type of camera capable of detecting the contours of the tray or receptacle 7, such as a time-lapse imaging camera. Suitable time-lapse imaging cameras include cameras commonly known as structured light cameras, RGBD (red, green, blue and depth) cameras. blue and depth), stereo camera pairs and time-of-flight cameras. Such a time-lapse imaging camera captures distance information of an observed scene. Similar to a digital color camera that records a representation of a scene on a grid of discrete colors, known as pixels, time-lapse cameras record distances from surfaces in the scene that are for each pixel position.

A color camera can be combined with the time-lapse imaging camera to capture both color and distance information.

In order to control the functions of the product packaging robot which includes the XX direction movement of the multi-head robot 20, the YY and ZZ direction movements and the W direction rotation of each of the heads 21, the movement of the receptacle conveyor 8, the accumulator conveyor 9 and the product supplies 13, a controller 35 is provided. The controller 35 also receives signals from the orientation detection camera 15 of each orientation module 12 and the destination detection camera 33. The controller 35 processes these signals, moves and tracks the locations of the heads 21, and controls when the product items are picked from the pick locations and released at the output points 6.

By having the heads 21 of the product packaging robot moving together like the multi-head robot 20 in the XX direction, and then independently in the remaining directions, a system is provided with a plurality of heads, each for packaging a product item, but which naturally lends itself to avoiding crossings or collisions between the heads.

The product to be sorted enters the product packaging robot 1 via accumulator 2, where it is directed to one of a series of individual lanes and is fed to each of the orientation modules 12. The product is then fed 13 passes these items one at a time to the pick locations 14 of the orientation modules 12. The purpose of the orientation modules is to rotate each product item into a desired pick orientation.

The rotational orientation of each product item is determined by the orientation detecting camera 15 with respect to one or more features 36 of the product 3. The orientator 16 at each pick location 14 multiaxially rotates a product item until the Orientation detection 15 indicates that the product is in the desired pick orientation with respect to the physical feature 36.

As shown in Figure 13, the desired picking orientation is based on a feature 36 of the product, such as an anatomical feature, for example, the stem and/or calyx of an apple. Alternatively, or in addition, the feature may be based on a desired color, texture, pattern, size, perimeter, or some other definable characteristic of the product.

As the product item is rotated multiaxially by the orientator 16, the orientation detection camera 15 passes a signal to the controller 35 where the appropriate vision software locates a feature and calculates the orientation of the product.

The orientator 16 rotates the product item 3 until the orientation detection camera 15 locates the feature. Controller 24 then identifies the location of the feature with respect to the product geometry. Once the feature or a feature has been placed, the orienteer 16 is actuated by the controller 35 to place the product in a desired pick orientation. In this configuration, a product item can be quickly placed in the desired pick orientation once the feature has been placed.

In some configurations, the controller 35 may continue to monitor the feature and provide updated instructions to the orienteer 16 as the orienteer places the product in the desired pick orientation. As the product can vary in size, shape and texture, resulting in inconsistent product rotation characteristics, this feedback through continuous monitoring of the characteristic by the controller 35 can ensure that the orenter 16 accurately rotates the product at the right time. desired pickup orientation.

Alternatively, quickly placing the product in the desired pick orientation can be further achieved if the desired pick orientation only requires that the feature be anywhere within a plane (e.g., a plane in the XX and YY). In this configuration, the orientator 16 orients the product item to the desired pick orientation within an XX and YY plane by rotations about these two directions. To orient all products to the same final orientation in the receptacles, a rotation W about the z axis Z may be provided by each of the heads 21 when the product item is picked up. The required turnover of each product item is calculated by controller 35.

In an example where the product is apples and the feature is the stem 36, the planes can be placed in the picking orientation so that the stems are flat along the XX-Yy plane. The orientation of the feature around the ZZ direction is monitored by the controller 35, and the rotation required to achieve a constant predetermined output orientation is calculated by the controller 35. When each apple is picked up by a head, a rotation is then provided W to place the apple in the desired orientation in the receptacles, for example, so they all face a parallel direction.

In this configuration where the product is only oriented by the orienter in one plane, the time required for an item of product to be ready for pickup can be reduced further, resulting in a potential increase in product throughput through the robot. .

Alternatively, once the feature has been located by the orientation detection camera, the controller 35 may adjust the orientator 16 to manipulate the product until the located feature is aligned with a centroid of the product profile, as seen from the orientation detection camera. When the product is an apple, this feature alignment within the centroid can approximate an orientation along a theoretical peduncle-calyx axis. The orienteer may be coupled to rotate the product item to place it in the desired pick orientation, as described above.

By taking the additional step of aligning the feature with the centroid of the product profile, the controller 35 can more accurately identify the location of the feature, thus allowing for more precise placement in the desired pick orientation and as a result a more accurate final orientation. required in the receptacle.

However, although this stage can increase the accuracy of final product orientation, it can also result in a longer time taken before a product item is ready to be picked and a decrease in product throughput throughput. of the product packaging robot.

Accordingly, it may be desirable to reduce the time required for a product item to be ready for pickup so as not to use the additional step of aligning the feature with a centroid of the product profile. When this additional step is not taken, the controller 35 can be configured with software to approximate the location of the feature when it is first detected by the orientation detection camera.

In an alternative configuration there may be more than one feature, such as the stem and calyx of an apple, each of which the controller 35 may be programmed to identify. In the example where the product item is an apple, the controller can identify between the stem and the calyx of the apple.

When there is more than one feature, and the features are predictable anatomical markers located in a common relationship with each other, such as the peduncle and calyx of an apple, the desired picking orientation may be set with respect to one of the features, e.g. , the peduncle of an apple. As the controller 35 can identify which feature locates, for example, the stem or calyx of the apple, the orienteer can rotate through a different angle for each feature to arrive at the same desired picking orientation. For example, if a calyx was identified, the orienteer would rotate through 180 degrees or so to still place the product item in the desired pick orientation relative to the stem.

Alternatively, controller 35 can instruct a rotation of approximately 180 degrees, then adjust the alignment of the stalk with the orientation sensing camera before rotating the apple through the predetermined angle to achieve the desired picking orientation.

When the product must be oriented according to a representative color, texture, or pattern, the controller 35 can be programmed so that the feature is the desired representative color, texture, or pattern. In the case of these forms of features, the controller 35 may view all, or substantially all, of the surface of the product during its multiaxial rotation by the rotation means 16 before determining the closest portion matches the desired feature. The product item may then be oriented in the desired pick orientation with respect to this portion. In this configuration, the desired pick orientation may be such that the representative portion of each product item may be oriented upward, parallel to the ZZ direction so that the feature is clearly visible when the product is packaged. For example, it may be preferred to package apples with the “redder” portions facing up.

The product can be targeted based on a physical anatomical feature and a representative color, texture, or pattern. In this case, the controller 35 can monitor the location of both features, instruct the orienteer to rotate the product to the desired pick orientation so that the product is oriented with respect to both features. For example, you may prefer to package apples with the stem-calyx axis positioned in the XX-YY plane and with the “redder” portion of the apple surface facing upward. In this configuration, the controller would ensure that the apples are located both so that the peduncle-calyx axis is positioned in the XX-YY plane, and that the representative colored portion is oriented upward.

By orienting the product to a desired pick orientation with respect to a feature, it may be possible to create a greater degree of uniformity in the appearance of the packaged product. This can have an aesthetically pleasing effect and can increase the potential sales value of the product. Additionally, to produce non-uniform shapes, such as pears, orienting according to the non-uniform portions of the shape can increase the efficiency with which product 3 can be packaged in receptacles 7. As transportation costs for production increase to As the volume to be transported increases, more compact product packaging can reduce the transportation cost.

While the product items may first be oriented according to a feature, such as a stalk and/or calyx position, the product may also be oriented to present a surface compatible with the suction cup 30 or the product holder 29. Depending on the type of product, orientation to present a suction-compatible or grip-compatible surface may be provided during initial orientation of the product, or may be provided by a set amount of additional rotation once the desired feature has been identified by the motion detection camera. 15. Orientation For example, when apples are packed by the product packaging robot, the peduncle-calyx axis can be identified first by the orientation detection camera, then the planes rotate through an additional angle so that the axis of the pedunculocalyx is approximately parallel to the XX-YY plane. In the case of apples, the side portions that face upward in this orientation are generally similar in shape, with a continuous surface that may be more compatible with a suction cup. For other products where such side portions would not likely present a suction-compatible surface, a more suction-compatible portion may be established as a desired feature by the controller 35, the product surfaces analyzed by the orientation detection camera, and the product also oriented according to a portion suitable for suction.

As the peduncle and calyx are generally at opposite ends of the apple axis, this will orient the peduncle in the desired picking orientation. Targeting based on detected features other than the desired feature can be applied to any other type of product that has features that are in a common location with respect to the desired feature. While a peduncle and calyx can be used for orientation, other physical characteristics such as the bend or fold in the flesh of an apricot. In the example of an apricot, the detection of the fold or fold and its direction can be used by an orientation module 12 to more quickly orient the upper or lower portion of the apricot towards the orientation detection camera 15.

By having the orienteer orient the product item only in the XX-YY plane, the speed with which the product items are ready to be picked can be increased. This has the advantage of increasing product throughput through the guidance modules and therefore potentially increasing the throughput of the product packaging robot.

In addition to the characteristics analyzed to produce the orientation, they may optionally be oriented by the orientator 16 with respect to a physical size of the product. The controller 35 may determine from the orientation detection camera while the product is rotated the dimensions of the product, or the dimensions of the product when oriented based on other characteristics as described above. The controller 35 can then direct the heads 21 to pack the product into receptacles so that products of similar size are packed together. To package multiple product sizes at the same time, multiple receptacles 7 may be provided in the exit zone 5. In this configuration, an additional product size sensor may be provided above the pick locations in the orientation modules to provide more information. precise about the size of the product to the controller 35.

When a product item is in its desired pick orientation, this orientation is detected by the controller 35 and a head 21 can be assigned to pick the product item. Additionally, the controller 35 may also detect when a desired product feature or primary feature has been identified and/or aligned with the orientation detection camera. This would indicate to controller 35 that the product is almost ready to be picked up (for example, but not rotated to a desired picking orientation). This advance warning of availability can be used by the controller 35 to prioritize the movements of the heads 21 in order to increase the efficiency of packaging by the product packaging robot, and decrease the time that an item of product is waiting in the desired orientation position ready for pickup.

If all of the heads 21 have already picked up an item of product, or if there are a sufficient number of items ready to fill all of the heads, then the controller 35 may ignore any almost ready signal. Otherwise, the controller 35 can estimate from the almost ready signals which pick point will have an item of ready product nearby, and place a head 21 over that pick point.

When deciding whether to wait for an item of product to be ready after an almost ready signal, the controller 35 may decide based on a comparison between the estimated time to be ready and the time required to release the product from other heads and return to the pickup location. The controller may also base this decision on either or both of

a) the number of product items currently held by the heads 21,

b) the number of pickup points that have a product item ready for pickup, and

c) the number of pickup points that have a product item almost ready for pickup.

When making decisions about picking and releasing products, the controller 35 can be programmed to operate the product packaging robot to increase the speed at which products are packed and the efficiency of the robot in terms of the utilization of the heads 21. and 12 orientation modules.

In one configuration, the number of orientation modules 9 and pick locations 14 may be greater than or equal to the number of heads 21.

In a preferred configuration, the number of pick locations 14 is greater than the number of heads 21. For example, the number of pick locations may be 2-3 times greater than the number of heads. When the number of pick locations is greater than the number of heads, the heads 21 can then move laterally parallel to the YY axis between a plurality of different orientation modules 12 and pick locations 14 depending on where the product is to be picked.

In the preferred form, the product packaging robot has approximately twice the number of orientation modules and therefore product pick locations 14 as heads 21 in the multi-head robot 20. However, it is expected that the ratio of orientation modules and heads on the robot can be customized to fit the time commonly taken to place a product in the desired pick orientation. This time may vary between types of products. Having more pick locations allows the packaging robot to operate more efficiently, making it more likely that all pick heads will be able to immediately pick products that are ready and waiting in a desired pick orientation.

The movements of the heads 21 parallel to the YY axis between the pick locations of each orientation module 12 must depend on which pick locations have a product item ready, or almost ready, for pick, which heads are already holding an item of product and the distance from an available head to the pick location.

While packaging is occurring, the heads 21 of the multi-head robot may be in any desired position along an axis parallel to the YY axis, however, the heads 21 may not overlap each other. This means that not all heads 21 can service all pick locations. As shown in Figure 8, all of the heads 21 may be oriented at one end to one side of the product packaging area 4. Although in the configuration shown, the heads 21 are spaced apart by the distance between each pick location 14, It contemplates that the separation of the pick locations and the heads could be such as to accommodate the heads that move closer than the separation of the pick locations. This will facilitate the picking of products from the pick locations to each side of the product packaging area when the heads to that side were already carrying an item of product.

To pick a ready product item 3 from a pick location 14, the controller 35 must direct the multi-head drive unit 23 to align a desired pick head 21 of the multi-head direction XX over the relevant pick location. The head, or several heads, can also be directed by the controller to vary their spacing and moved in the YY direction to position the desired pick head over the relevant pick location. An extension of the head in the ZZ direction then brings the suction cup 30 into contact with the product article 3, and vacuum is applied. Because product items within a certain type, such as apples, may vary in size, the dimension of the product in the ZZ direction may be detected by the product orientation camera 15 or other sensors in the orientation modules 12. By communicating the sensed height of the product item to the controller 35, the relevant pickup head can be extended such that the suction cup 30 can be placed at the correct ZZ location to contact the product, but not extend to forced contact. with the product that may damage it.

When determining the desired pick head for a ready-to-produce item at a pick location, controller 35 may consider:

a) the status of the other heads, e.g. e.g., if they are already carrying a product item or are in the process of picking up a product item,

b) the lateral movement in the YY direction, and the number of heads required to articulate in the YY direction, for any head to align with the desired pick location,

c) the status and relative locations of other pickup locations, e.g. e.g., if other locations next to or near it have a product item ready,

d) the location of the currently unfilled exit points on the receptacles and the relative distances from the pick location with the product ready to the unfilled exit points.

The packaging product controller 35 may prioritize any of these factors to minimize or optimize to decrease the time that product items are ready and waiting at any pick location, decrease the total movement of the heads in the XX or YY directions, or otherwise increase the performance of the product packaging robot.

By considering any of the above factors, controller 35 may prioritize some pickup locations over other pickup locations. Additionally, controller 35 may cause product items to be picked up while other heads in the head array have already picked up a product item and are waiting to move to the output zone.

A foldable pad portion 32 is provided between the suction cup 30 and the head 21 to allow sealing of the suction cup and prevent damage to product items of different sizes. The folding pad may be used on the heads in conjunction with product size information provided by the orientation module sensors, or may be used in an embodiment where this information is not provided. The foldable pads 32 have a concertina-like structure, and the interior of the pad experiences vacuum or pressure that is provided by the air inlet 31 to the suction cup 30. Under positive pressure, the foldable pad 32 expands and lengthens Under vacuum, the folding pad 32 contracts and shortens the length. Furthermore, under a compression force, the folding pad 32 will deform and contract with low resistance. These features of the folding pad 32 decrease the possibility of a product item being damaged when picked up or released.

When a head 21 extends downward in the ZZ direction so that the suction cup 30 contacts the product 3, a degree of excessive extension of the head 21 may be possible without the rigid portions of the head coming into contact with the product and damage it, as the folding pad 32 will deform to compensate for excessive extension. After providing vacuum, the foldable pad will contract and shorten in length. When the product item is lowered and positioned over the desired exit point and the vacuum is released, the folding pad will expand, allowing the product to gently fall any remaining distance from where it was placed at the exit point. In the event that the product is located by the head 21 near or at the exit point 6, when the vacuum is released and the pad would normally expand to its neutral position, the folding pad 32 remains deformed with low resistance and does not cause damage to the product.

The configuration of the collapsible pad 32 and suction cup 30 may be such that when the vacuum is released, the deformable pad expands from its contracted position before the product is released from the suction cup. In this configuration, the product item is gently lowered onto the exit point, then released by the suction cup, preventing the product from falling onto the exit point, resulting in any possible damage.

Between when an item of product is picked up by a head and when it is released at the desired exit point, the product can be rotated into a desired exit orientation. This rotation in the output orientation is provided by a rotation of the head parallel to the ZZ axis, in the W direction as shown in Figure 1. This rotation is about an axis orthogonal to the XX and YY directions, and therefore will not change the rotational orientation of the product around these directions. This rotation may be provided in place of, or in addition to, any rotation about a direction ZZ provided by the orientator 16 of the orientation module 12. For example, the product may be presented on a tray having the same direction, as shown in Figure 12 where the peduncles are all aligned in parallel.

By providing the desired rotation of the product in the ZZ axis by the head 21 instead of the orienter 16, the time of a product item in the orientation module can be decreased. Additionally, the overall time to package an item of product can be decreased as rotation through the header can be provided in the transit time when the product moves from the pick location to the exit point.

In the example where a product item is classified as having two characteristics along a centroid axis such as the peduncle and calyx of an apple, the orientation module can orient the apple in the desired picking orientation with respect to the peduncle or calyx. In this example, the desired picking orientation would be an orientation where the peduncle-calyx axis lies in the XX-YY plane. The orientation module can then indicate to the controller 35 that the apple feature has been arranged with respect to, and the orientation of the peduncle-calyx axis with respect to the XX or YY directions. The controller 35 can then operate the head 21 that picks up the product article through a rotation to align the peduncle or calyx in the desired orientation, compensating for the difference in the orientation of the peduncle-calyx axis in the XX or YY direction. in collection.

For example, some apples will have their stem located and aligned in a particular direction by the orienteer, while others will have the calyx identified and aligned. Therefore, to achieve the desired final orientation of the planes with the peduncle or calyx in one direction, the controller can operate the head 21 to rotate some apples by 180 degrees or so. This will cause all apples to eventually orient themselves with their stems in one direction, regardless of whether they were oriented for picking with respect to the stem or the calyx.

The desired final orientation can be varied by a user through the controller 35. The controller 35 can be configured so that the final orientation entered by the user can be in three dimensions, and the controller can calculate the desired and sequential pick orientation and the required rotation of the heads to achieve the desired final orientation.

For example, the desired output orientation for a product item may be between 20 and 70 degrees between the XX and YY directions, or the desired output orientation may be approximately 45 degrees with respect to the XX or YY directions, or otherwise depending on the angle of the receptacles, so that the desired features are approximately 45 degrees to the sides of the receptacle.

From its vantage point over the exit zone 5, the target detection camera 33 can scan the contours of the surface below where the trays or receptacles 7 are located. The trays or receptacles will have a number of desired exit points 6 for product 3.

In trays where the desired exit points correspond to depressions or pockets 34 in the tray or box, the pockets 34 and therefore the desired exit points can be determined by scanning the contours of the trays or receptacles 7.

When trays without depressions or pockets are used, the target detection camera 33 can determine the ends of the tray, or the location of the side walls if present.

Additionally, outlines can be used to indicate whether a pocket is empty or contains a product item. Information from the target detection camera about the availability and location of exit points is communicated to the multi-head robot to determine where products are released. According to this exit point detection method using the destination detection chamber 33, a variety of trays of different sizes, types and configurations can be used during a single packaging operation.

For example, this may allow continuous sorting even when trays or receptacles, which may be placed by human or machine on the receptacle conveyor, are placed out of alignment with each other. Additionally, a continuous sorting operation may be permitted between batches of receptacle types or sizes. More of the output area can also be allowed to be filled with trays if smaller size trays are used, as the trays can still be filled even if they are placed parallel to each other, for example in the cross or YY direction. .

When the product is packaged in receptacles that are in the form of baskets, boxes or cartons that have multiple layers of products packed therein, the information provided by the destination detection camera may allow the controller 35 to consider the heights and product distributions in receptacles in making other packaging decisions. This can result in products, particularly products of different sizes or shapes, that can be sorted more efficiently in a given space in the receptacle. This can decrease the total storage space required for the packaged product and decrease the costs associated with shipping the product.

When packaging products from the orientation modules 12 to the trays or receptacles 7, the controller 35 of the product packaging robot prioritizes filling the free exit points that are furthest from the orientation modules. This acts to decrease the distance traveled and the time spent by the multi-head robot 20 traveling in the XX direction. The product packaging robot also prioritizes picking products as soon as they are ready to be picked up. This may mean that the multi-head robot may move over the pick locations 14 when some heads already have products on them if an item of product has just become ready.

Additionally, controller 35 may decide to move over a pick location while already holding product in other heads when it determines from an orientation detection camera that an item of product is almost ready to be picked. As described above, this almost ready signal may be provided when one of multiple desired features, not being the primary feature, is detected or when the primary feature is facing the orientation detection camera 15 and only a small rotation is required. to place the item in its desired pick orientation. When moving and waiting for guidance that a product item is ready, you can act to decrease the amount of time each orientation module is standing waiting with a product item ready to be picked up, and to increase the amount of time each targeting module is actively targeting a product item. Each of these priorities act to increase the throughput of the product that can be packed by the product packaging robot.

The receptacles 7 pass through the exit zone 5 on the receptacle conveyor 8. Depending on the configuration, the controller 35 can operate the receptacle conveyor 8 to move in a staggered manner, moving a predetermined distance in the direction XX, for For example, the distance may correspond to the separation between the exit points 11 and stopping until the pockets are loaded once the exit points are filled with product. Alternatively, the controller may operate the receptacle conveyor 8 so that the conveyor moves continuously, at a variable speed as necessary, or at a constant speed. Alternatively, the controller may control the speed and/or direction of the conveyor 8 to adapt.

To coordinate the multiple moving parts of the product packaging robot, the controller 35 can be configured to translate the coordinate systems of each moving component to a universal coordinate system. This would allow, for example, the location of available exit points to be continuously determined by information from the target detection camera 33 that was located attached to the multi-head robot as it moved in the XX direction. By translating the moving coordinate space of the target detection camera, the controller 35 can continuously determine the real space location of the available pockets. This translation procedure applies similarly to the other moving parts of the packaging robot, including the receptacle conveyor 8 and the individual heads 21.

A plurality of pickup locations are provided in the primary embodiment.

In the primary embodiment, the product packaging robot can package any type of product. Alternatively, the product packaging robot can package multiple types of products at the same time.

The product packaging robot may be provided with a single controller that operates all functions of the packaging robot and makes packaging-related decisions, or multiple distributed controllers may be provided that operate each aspect of the packaging robot function. and/or make decisions related to packaging.

In general terms, the product can be any type of agricultural or natural products, including fruits or vegetables. More specifically, the product may be any type of fruit or vegetable, specifically including, but not limited to, apples, peas, kiwis, melons, seeded fruits, tomatoes and peppers.

By providing a robot that has multiple heads, the product packaging robot can package an increased number of product items. By having the plurality of robot heads 21 movable independently in some directions but fixed so that they move together in one direction, a greater number of robots can be provided without the associated concerns of collisions between the multiple heads. Consequently, the product packaging robot is capable of providing a large number of heads in a compact space. By increasing the compactness of the heads, a large volume of products can be packaged while requiring a reduced total surface area of the product packaging robot. A packaging robot with a reduced footprint means that less packhouse space can be required to house the robot, and the number of product items packed in a given space can be increased.

The foregoing description of the invention includes preferred forms thereof. Modifications can be made thereto without departing from the scope of the invention as defined in the attached claims.

Claims (15)

1. A robot (1) for product packaging that comprises; a plurality of pick locations (14) for the product (3), the pick locations (14) arranged in an array, an exit area (5) for the product (3), and a plurality of heads (21), arranged in an array, each head for picking up, holding and releasing an item of product (3), wherein the product packaging robot (1) also comprises a controller (35) to control the movement of the heads (21), wherein the plurality of heads (21) move together as a set in a first direction between the array of collection locations (14) and the exit area (5), wherein each head (21) can move independently in a second direction, one direction being orthogonal to the first direction, to vary the separation of the plurality of heads (21) in the second direction, and wherein each of the plurality of pick locations (14) further comprises an orientation module (12) for presenting a product item (3) in a desired pick orientation with respect to at least one feature of the product, each having orientation module (12): at least one orientation detection camera (15) for detecting and determining the location of said at least one feature, and an orientator configured to manipulate said product article (3) in a desired picking orientation, characterized in that : The orientation module (12) is configured to indicate to the controller (35) when the feature is detected by the orientation detection camera (15) that a product item is almost ready to be picked, and the controller (35) is configured to make a decision to wait or not wait for an almost ready item before moving the heads (21) to the drop zone (5). 2. The robot of claim 1, wherein each head (21) is capable of unequally varying the spacing between the plurality of heads in the second direction. 3. The robot of claim 1 or claim 2, wherein the orientation module (12) can indicate to the controller (35) when a product item (3) is in a desired pick orientation and is ready to be picked . 4. The robot of claim 1, wherein the orientator multiaxially rotates the product article (3) until the feature is detected by the orientation detection camera (15), after which the location of the feature is determined. with respect to the product, and where the characteristic with respect to which the product must be oriented is selected from: to. a physical anatomical characteristic, b. a portion more visually similar to a desired color, c. a desired pattern, or d. a desired texture. 5. The packaging robot of any one of claims 1 to 4, wherein the orientation of the product (3) is determined through the at least one orientation detection camera (15) with respect to a plurality of characteristics, one of said features being a primary feature with respect to which the product is placed in the desired pick orientation. 6. The packaging robot of claims 1 to 5, wherein the orientator comprises: a pair of laterally spaced controlled drive belts (19), configured to hold a product item, a boundary (18) that at least partially surrounds the product article (3) to restrict translational movements of the product, wherein the plurality of drive belts (19) can drive at different speeds to multiaxially rotate the article (3). 7. The packaging robot of claims 1 to 6, wherein the exit area is provided with one or more receptacles for receiving the provided product, the one or more receptacles defining one or more desired locations for placing items of product, and The desired locations for placing product items are the exit points for the product, where the receptacles are trays or boxes, which have contoured pockets that are the exit points (6), and wherein the product packaging robot (1) further comprises a target detection camera for detecting one or more of the location, size and orientation of the contoured pockets. 8. The product packaging robot of claim 7, wherein the target detection camera is a time-lapse imaging camera selected from one of: to. a structured light chamber, b. an RGBD camera, c. a pair of stereo camera, or d. a time of flight camera 9. The packaging robot of claims 1 to 8, wherein the number of heads (21) is less than the number of pick locations (14) for the product. 10. The product packaging robot of claims 1 to 9, wherein the movement of the heads (21) in the second direction to vary the spacing of the heads is such that it allows the heads to each move between multiple locations. collection (14) and multiple exit points. 11. The packaging robot of claims 1 to 10, wherein packaging an item of product involves: a. positioning the head array (21) in the first direction so that at least one head is aligned in the first direction with a product item (3) at a pick location (14), b. independently varying the spacing of one or more heads (21) in the second direction to align, in the second direction, at least one head with the product article (3), c. pick up the product item (3), d. transporting the product item (3) in the first direction to the exit zone (5) so that the product article (3) is aligned in the first direction with a desired exit point (6), and and. release the product item (3) at the desired exit point (6). 12. The packaging robot of claim 11, wherein: before releasing the product items (3), the spacing of one or more heads (21) is varied in the second direction; I Packaging a product item (3) also involves a rotation of the desired head about an axis parallel to the third direction to rotate the product item (3) to a desired release orientation. 13. The packaging robot of claim 12, wherein, after receiving an almost ready signal, the controller (35) waits for a product item (3) to be manipulated into the desired pick orientation depending on: to. the number of heads that have picked up and hold an item of product (3), b. the number of pickup points that have a product item (3) ready for pickup, and c. the number of pickup points that have a product item (3) almost ready to be picked up. 14. The packaging robot of claim 13, wherein, after receiving an almost ready signal, the controller (35) ignores the signal and releases the product items (3) currently held by the heads (21) when: to. all heads (21) are already holding a product item (3), or b. The number of items ready for pick is greater than or equal to the number of heads (21) that do not hold a product item (3). 15. The product packaging robot of claims 13 or 14, wherein the controller (35) bases the decision to wait for an almost ready product item to be ready on an estimated comparison of the time until the product item ( 3) is oriented in the desired pick orientation and is ready to be picked and the time required to release the product items (3) currently held by the heads (21) and return to the pick location (14).