Anomaly detection device when transferring egg cartons

Abstract

A device (1) for detecting abnormalities when transferring an egg pack comprises: a transfer device (3); an imaging device (4); and a control device (20). The transfer device (3) applies a vacuum force to the top surface of a lid of the egg pack to pick up and transfer the egg pack and then places the egg pack at a predetermined position of a container for transport. The imaging device (4) images a region including the predetermined position after the egg pack has been placed. The control device (20) detects abnormalities in the region on the basis of the captured image. The control device (20) instructs, on the basis of the detection result, the transfer device (3) to carry out a subsequent operation.

Description

[Technical Field]

[0001] The present disclosure relates to an abnormality detection device during egg carton transfer. [Background technology]

[0002] For example, Patent Document 1 discloses a transfer device for transferring egg cartons. This transfer device picks up and moves the egg cartons by adsorbing the top surface of the lid. The transfer device then places the egg cartons in a predetermined position in a transport container such as a cardboard box.

[0003] Multiple egg cartons are stacked vertically inside a cardboard box, but individual egg cartons are very unstable. For example, an egg carton consists of a body that accommodates the bottom of an egg and a lid that accommodates the top of an egg. The body has an uneven shape that conforms to the shape of the egg, and the lid also has a shape unique to egg cartons. Furthermore, the connection between the body and the lid is provided with an outward-protruding flange. Egg cartons are also designed to allow consumers to easily open the lid. When an egg carton is lifted by adhering to the top surface of the lid, the lid and the body may partially or completely separate in some cases. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2018-177325 Summary of the Invention [Problem to be solved by the invention]

[0005] As described above, egg cartons have multiple factors that can make them unstable, making automated handling difficult. For example, even if a robot completes the operation of transferring egg cartons into a transport container, the egg cartons may not actually be in the desired state, such as not being placed in the designated location, being tilted, having their lids removed, or having eggs exposed outside the carton. If egg cartons are not in the desired state inside the transport container, this can have a significant impact on the egg cartons that are loaded subsequently.

[0006] The present disclosure aims to reduce the significant impact on subsequently loaded egg cartons even if egg cartons that are not in a desired condition occur within a transport container. [Means for solving the problem]

[0007] The device for detecting abnormalities during egg carton transfer of the present disclosure includes a transfer device and a photographing device. The transfer device suctions the top surface of the egg carton lid, lifts and moves the egg carton, and places the egg carton in a predetermined position in a transport container. The photographing device photographs an area including the predetermined position after the egg carton is placed. The control device detects abnormalities within the area based on the photographed image. The control device instructs the transfer device to perform the next operation based on the detection result. [Effects of the Invention]

[0008] According to the present disclosure, even if an egg carton that is not in the desired state occurs in a transport container, it can be reduced from having a significant impact on egg cartons that are loaded subsequently.

[0009] The above and other objects, features, aspects and advantages of the present disclosure will become apparent from the following detailed description of the disclosure taken in conjunction with the accompanying drawings. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a side view schematically illustrating a pack assembly production system according to an embodiment of the present disclosure. [Figure 2] FIG. 2 is a side view schematically showing the transfer device according to the embodiment. [Figure 3] FIG. 2 is a plan view schematically showing a part of the pack assembly production system according to the embodiment. [Figure 4] FIG. 2 is a block diagram showing an abnormality detection device during egg pack transfer according to the embodiment. [Figure 5] 10 is a flowchart showing an example of the operation of a pack assembly production system. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, a pack assembly production system 2 according to this embodiment will be described with reference to FIGS.

[0012] First, the pack assembly will be described. The pack assembly is produced, for example, at a GP center (Grading and Packing Center). The pack assembly is loaded onto a truck or the like and transported to a store where consumers can purchase it.

[0013] The pack assembly is a state in which a plurality of egg packs P are packed in a transport container 50. In the transport container 50, a plurality of egg packs P are arranged on one plane and stacked vertically. The transport container 50 is a dolly, a cardboard box, a plastic container, a pallet, or the like. The transport container 50 of this embodiment is a mobile rack. This mobile rack is a dolly designed specifically for egg packs, is foldable, and has a plurality of shelves 51. Egg packs P are stacked in multiple layers on each shelf 51.

[0014] Egg cartons P are stacked in the transport container 50 so that the body P1 of the top egg carton P and the lid P2 of the bottom egg carton P are in contact. The top and bottom egg cartons may be arranged in the same or different ways. The egg cartons P have an uneven shape in the portion that accommodates the bottom of the eggs, and their position may change within the transport container 50 after being stacked. In other words, egg cartons P are stacked in a very unstable state when another egg carton P is placed on top of another egg carton P. For example, the body P1 of the top egg carton P may rest on the inclined surface of the lid P2 of the bottom egg carton P, or the lid P2 of the bottom egg carton may be depressed downward due to the weight of the top egg carton P. Furthermore, the distance between adjacent egg cartons P may change due to the provision of a deformable flange P3, which will be described later.

[0015] An egg carton P is a container in which multiple eggs are packed and sealed. Containers for egg cartons P are well known in the art and include various types made of different materials or shapes. The container comprises a main body P1 and a lid P2. The main body P1 accommodates the underside of the eggs and has a concave-convex shape that conforms to the eggs, sloping inward downward as a whole. The lid P2 accommodates the upper side of the eggs and has a concave-convex shape that conforms to the eggs. The lid P2 may also have a concave-convex shape that conforms to the eggs. The lid P2 has protrusions on its top surface. The protrusions are provided, for example, at each corner of the top surface to prevent the egg cartons P stacked on top from shifting.

[0016] The egg carton P is made of, for example, a transparent synthetic resin. The egg carton P is made to a thickness that allows it to deform under external force, and is provided with multiple ribs. When photographing such an egg carton P, it is difficult to capture its outline. In addition, the egg carton P has a glossy surface, and the unevenness of the ribs causes glare when photographed. Therefore, a special photographing method or image processing method is required to photograph the egg carton P.

[0017] The egg carton P has an outwardly protruding flange P3. One side of the flange P3 is connected to the main body P1 and the lid P2 by integral molding. One side of the flange P3 may be connected with a perforation. The other side of the flange P3 is connected using tape (not shown). The other side of the flange P3 may be connected by thermal welding. The main body P1 and the lid P2 are connected in a manner that allows them to be easily separated by consumers.

[0018] Egg cartons P come in a variety of sizes and shapes, but the drawing shows an egg carton P containing 10 eggs, with a flat top surface of the lid P2.

[0019] Production system 2 for producing pack assemblies is mainly composed of transfer device 3, photography device 4, and control device 20. Control device 20 has a detection unit 5 and a control unit 6. Transfer device 3, photography device 4, and control device 20 constitute an abnormality detection device 1 for egg pack transfer according to this embodiment.

[0020] In this embodiment, the abnormality detection device 1 photographs the production process of pack assemblies produced at a GP center or the like, inspects the image data obtained by this photography according to a predetermined model, and outputs the inspection results. This model is obtained, for example, by machine learning, and is stored in a storage device 7. The production system 2 determines whether or not there is an abnormality in the pack assembly based on the inspection results, and automatically detects whether or not the transfer device 3 is allowed to perform the next operation.

[0021] The production system 2 also includes a conveying section 10 for egg cartons, a waiting section 11 for egg cartons, a conveying section 12 for containers, and a waiting section 13 for containers.

[0022] The egg carton conveying section 10 is equipped with a sealing machine (not shown) on the upstream side. The sealing machine is used to connect the other side of the flange P3 and is called, for example, a tape sealer. The sealing machine applies tape to thousands to tens of thousands of egg cartons P per hour. Eggs, which are natural objects, do not have a uniform size or shape in egg cartons P. Therefore, depending on the combination of eggs placed in the container, the degree to which the lid P2 closes or the degree to which the other side of the flange P3 is connected (for example, how the tape fastens) may not be uniform.

[0023] The transfer device 3 sucks the upper surface of the lid P2 of the egg pack P. The transfer device 3 lifts and moves the egg pack P. The transfer device 3 places the egg pack P in a predetermined position in the transport container 50. The transfer device 3 includes, for example, a robot head 31 and a robot arm 32.

[0024] Robot head 31 has a holding unit 33. Holding unit 33 holds a plurality of egg cartons P. Holding unit 33 is composed of a plurality of suction members 34. Each suction member 34 suctions the top side of an egg carton P with a lid P2 closed. Each suction member 34 is connected to a vacuum mechanism (not shown). Each suction member 34 has a bellows structure that is expandable and contractible in the vertical direction.

[0025] Robot head 31 lifts and moves multiple egg cartons P at a time. In this embodiment, holding unit 33 holds egg cartons P only on the top surface of lid P2. Therefore, when egg cartons P are lifted by holding unit 33, the weight of the eggs places a large load on the connected flange P3. Furthermore, lateral shaking occurs as robot head 31 moves, making the holding state of holding unit 33 unstable.

[0026] The robot head 31 releases the holding state of the holding unit 33 within the transport container 50. The egg carton P is placed directly on the bottom of the transport container 50 or on top of an egg carton P that has previously been placed in the transport container 50.

[0027] The robot arm 32 is capable of changing the position and posture of the robot head 31. The robot arm 32 moves the robot head 31 between the egg carton standby section 11 and the transport container 50. The robot arm 32 may be, for example, a conventional articulated robot that can be operated arbitrarily, such as a six-axis vertical articulated robot.

[0028] The camera 4 captures an image of an area including a predetermined position. The camera 4 captures an image after the egg packs P are placed in the transport container 50. The predetermined position is the position where the transfer device 3 places the egg packs P in the transport container 50. The predetermined position for each transfer operation is stored, for example, in the memory of the control unit 6. The area includes the space around the transport container 50. The camera 4 captures an image every time the transfer device 3 places an egg pack P in a predetermined position. In this embodiment, the camera 4 is provided on the transfer device 3 and attached in a position (not shown) that does not interfere with the transfer operation. The camera 4 may capture an image from the top side of the egg packs P, from the side, or from an oblique direction relative to the egg packs P. The camera 4 is preferably movable vertically according to the height at which the egg packs P are stacked, and is attached, for example, to a robot arm 32.

[0029] The detection unit 5 detects abnormalities within the area based on the captured image. The detection unit 5 inspects the image data of the egg packs P captured by the imaging device 4 against predetermined items and outputs the results (for example, whether or not there is an abnormality in the loading of the egg packs). The inspection by the detection unit 5 is performed, for example, according to a predetermined learning model obtained by machine learning. Examples of abnormalities include egg packs P not being placed in a predetermined location, egg packs P being tilted, lids P2 of egg packs P being detached, eggs being exposed outside the packs, etc. The control device 20 is configured using a computer.

[0030] Based on the image data, the detection unit 5 determines the state of the egg packs P after transfer, specifically, whether the transfer was successful or unsuccessful. If the transfer has failed, for example, the following first to fourth cases may occur.

[0031] The first case is when no egg cartons P are placed in the predetermined locations. The detection unit 5 can determine the first case from, for example, the number of egg cartons P included in the image.

[0032] The second case is when the egg carton P is tilted. The detection unit 5 can determine the second case from, for example, the orientation of the egg carton P included in the image.

[0033] The third case is when the lid P2 of the egg pack P is removed. The detection unit 5 can determine this is the third case, for example, when the lid P2 cannot be detected or the label attached to the lid P2 cannot be detected.

[0034] The fourth case is when the eggs are exposed outside the pack. This can occur, for example, when the lid P2 opens during transfer and the eggs fly out of the pack. The detection unit 5 can determine this fourth case by detecting eggs outside the transport container 50. In this way, the state of the egg packs P stacked in multiple layers in the transport container 50 can be determined based on the image data.

[0035] The learning model is trained in advance using training data. Image data is acquired for a state in which the egg cartons P have been successfully transferred, i.e., the state of a large number of egg cartons P that have been determined to be properly loaded in the transport container 50, and training data is created. The training data includes information representing the normal state of the egg cartons P at each stage of loading into the transport container 50 and image data representing photographs of the egg cartons P loaded in the transport container 50. Image data of the egg cartons P included in the training data is input to each node in the input layer, and training is performed by adjusting the operation parameters of each node so that each node in the output layer outputs a score that matches the state indicated by the information included in the training data. For example, during training, an error is calculated using an error function with the output value of each node in the output layer and the output value expected from the training data as variables, and the operation parameters of each node are adjusted using the error backpropagation method to minimize the error.

[0036] The storage device 7 stores learning models obtained through machine learning. The learning model is a trained model for performing processing to determine the state of egg cartons based on image data. The learning model uses a neural network with an input layer, a middle layer, and an output layer. Image data is input to the input layer. For example, pixel values ​​for each pixel included in the image are input. The output layer outputs the state of the egg cartons. The storage device 7 has learning models trained from image data stored and accumulated for each type of container 50 and each type of egg carton P. That is, it has learning models corresponding to each product item. The storage device 7 also has learning models trained from image data stored and accumulated for each type of operation of the transfer device 3. That is, it has learning models corresponding to each process, and corresponding to various states of the carton assembly, including intermediate stages in producing a finished carton assembly. Note that there does not need to be a one-to-one correspondence between the learning models and each process; one learning model may correspond to multiple processes.

[0037] The control device 20 is configured by a dedicated computer or a general-purpose computer, including a CPU (Central Processing Unit), internal memory, an input / output interface, an AD (Analog-to-Digital) converter, etc. The functions of the control device 20 are fulfilled by the CPU and its peripheral devices working together in accordance with the programs stored in the internal memory.

[0038] Based on the detection result, the control unit 6 instructs the transfer device 3 to perform the next operation. Specifically, based on the detection result that is determined to be abnormal, the control unit 6 outputs an abnormality signal. The control unit 6 stops the transfer device 3 from starting the next operation.

[0039] The control unit 6 also controls the transfer device 3. Specifically, the control unit 6 controls the holding / release operation of the egg pack P by the holding unit 33. The control unit 6 controls the movement of the robot head 31 equipped with the holding unit 33 and the robot arm 32 that can change the position and posture of the robot head 31. The control unit 6 moves the egg pack P held by the holding unit 33 onto the shelf 51 of the container 50. When moving the egg pack P to the space below the shelf 51, the control unit 6 positions the tip of the robot arm 32 at a position shifted toward one end in the longitudinal direction from the center of gravity of the egg pack P, and holds the egg pack P by the holding unit 33.

[0040] Egg pack conveying section 10 conveys egg packs P and supplies them to egg pack waiting section 11. A packaging device (not shown) that places eggs in packs is arranged upstream of egg pack conveying section 10.

[0041] The waiting section 11 for egg packs is provided downstream of the conveying section 10 for egg packs, and is provided with a stopper 15 that limits the movement of the egg packs P so that the transfer device 3 can hold the egg packs P.

[0042] The container conveying section 12 conveys the container 50 and supplies it to the container waiting section 13. The standby section 13 for the containers is provided midway along the transport section 12 for the containers, and a stopper 16 is provided to limit the movement of the containers 50 so that the transfer device 3 can stably load the egg packs P.

[0043] 5 is a flowchart showing an example of a pack assembly production method and an abnormality detection process based on image data, which are executed by egg pack transfer abnormality detection device 1. First, when control unit 6 outputs a signal to transfer device 3 to start the transfer operation, transfer device 3 lifts egg pack P from egg pack standby section 11 and transfers it to a predetermined position in container 50 (step S1).

[0044] When the series of transfer operations is completed (step S2), the detection unit 5 performs a setting process to set up the device to detect abnormalities in egg cartons. The detection unit 5 may perform this setting process before step S1 or S2, or after step S5, which will be described later. In the setting process, the detection unit 5 acquires information about the process of the transfer device 3 from the control unit 6 (step S3), and selects a learning model corresponding to the process for the product item (step S4). The storage device 7 contains multiple learning models with different parameters, and selects the learning model to be used in step S6 from these learning models.

[0045] When the image capturing device 4 receives a signal indicating that at least a series of transfer operations has been completed, it captures an image of the egg packs P in the state after they have been transferred to the container 50 (step S5). The detection unit 5 performs a determination process to determine the state of the egg packs based on the image data. The determination process is a process of making a determination using a learning model (step S6). For example, the probability that the egg packs P belong to each category, either successful or unsuccessful, of being transferred to the container 50 is obtained as an estimation result, and the success or failure of the transfer is identified based on the estimation result. The determination result is output to the control unit 6 (step S7).

[0046] In step S8, if the transfer is successful, a normal signal is output to control unit 6 and the process proceeds to transferring the next egg pack P, but if the transfer is unsuccessful, the process proceeds to step S9. In step S9, control unit 6 outputs an error signal and outputs an abnormality signal to transfer device 3, prohibiting the start of the transfer operation of the next egg pack P.

[0047] As described above, the egg pack transfer abnormality detection device 1 according to this embodiment includes a transfer device 3, a photographing device 4, and a control device 20. The transfer device 3 suctions the top surface of the lid P2 of the egg pack P, lifts and moves it, and places the egg pack P in a predetermined position in the transport container 50. The photographing device 4 photographs an area including the predetermined position after the egg pack P is placed. The detection unit 5 of the control device 20 detects abnormalities within the area based on the photographed image. The control unit 6 of the control device 20 instructs the transfer device 3 to perform the next operation based on the detection result. As a result, the production system 2 can automatically produce a pack assembly in which egg packs P are stacked in the transport container 50 in a desired state. Furthermore, the abnormality detection device 1 can reduce the significant impact on subsequent egg packs P, even if an egg pack P that is not in the desired state is found in the transport container.

[0048] Furthermore, photographing device 4 photographs egg packs P stacked in multiple tiers in transport container 50, with egg packs P arranged in different ways on the upper and lower tiers. Photography device 4 photographs egg packs P each time transfer device 3 places egg packs P in transport container 50.

[0049] The detection unit 5 detects at least one of the following: that the egg pack P is not placed in a predetermined position, that the egg pack P is tilted, that the lid P2 of the egg pack P is removed, and that the eggs are exposed outside the pack. This makes it possible to deal with problems specific to egg packs P that are structured so that consumers can easily open the lid P2.

[0050] Regarding "whether the lid P2 of the egg carton P is detached," it is possible to install a separate inspection device downstream of the sealing machine to inspect whether the flange P3 of the egg carton P is securely connected. However, with a system like this embodiment, it is possible to detect whether the lid P2 of the egg carton P is detached at low cost while ensuring the processing speed of the egg cartons P. Furthermore, detecting whether the lid P2 of the egg carton P is detached can be performed together with inspection of another item.

[0051] The area includes the space around the transport container 50. Therefore, it is possible to detect egg cartons P that have fallen outside the container 50 when the egg cartons P are transferred.

[0052] The camera 4 takes a picture every time the transfer device 3 places an egg pack P in a predetermined position. To complete the pack assembly, the transfer device 3 must operate multiple times, but if an abnormality occurs, the operation can be temporarily suspended. This allows recovery before a serious accident occurs, such as eggs breaking in the egg pack P due to accumulated abnormalities.

[0053] The present disclosure is not limited to the above-described embodiments. In the above-described embodiment, an example using machine learning has been described, but the state of the egg carton P may be detected using general image processing without using machine learning.

[0054] The image capturing device 4 may capture only the vicinity of a predetermined position, or may capture a wide range including the vicinity of the predetermined position. The image capturing device 4 may be a fixed camera, or may be one whose height or direction can be changed depending on the position of the image to be captured. The image capturing device 4 may be installed in multiple locations.

[0055] The photographing device 4 may photograph after the transfer device 3 has performed multiple operations. For example, it is conceivable that the photographing device 4 may photograph after the transfer device 3 has transferred one layer of containers 50 or one tier of egg packs P on a shelf 51.

[0056] The abnormalities in egg packs P detected by detection unit 5 are not limited to those exemplified in the above embodiment. Furthermore, detection unit 5 may be located in a place separate from transfer device 3, and may be provided, for example, on a cloud.

[0057] Container 50 is not limited to the movable rack shown in the figure. Furthermore, the pack assembly may be in any state after a plurality of egg packs P have been stored in container 50. For example, this includes a state in which some of the shelves 51 do not have egg packs P placed on them, or a state in which the lid P2 of a cardboard box is not closed.

[0058] The transfer device 3 is not limited to a so-called six-axis vertical articulated robot. The transfer device 3 can be any robot or device that can move the robot head 31 along the X, Y, and Z axes.

[0059] Holding unit 33 holds, for example, three egg cartons P adjacent in the horizontal direction. Holding unit 33 may hold one or two egg cartons P, or may hold four or more egg cartons P simultaneously. Each egg carton P is held by a plurality of (for example, six) suction members 34. Note that the number and arrangement of suction members 34 may be any number. Furthermore, the arrangement of egg cartons P in container 50 is not limited to that shown in the figure.

[0060] The control device 20 of the present disclosure may be configured by a computer that is physically separate from the control device for controlling the transfer device 3.

[0061] The embodiments disclosed herein are examples and are not intended to be limiting. The scope of the present disclosure is defined by the scope of the claims, not the scope described above, and is intended to include all modifications within the meaning and scope equivalent to the scope of the claims. [Industrial Applicability]

[0062] The present disclosure can be used when transferring egg cartons. [Explanation of symbols]

[0063] 1...Anomaly detection device 3…Transfer device 4...Photographing equipment 5...Detection unit 6...Control section 20...Control device 50...Transport container P...egg carton P2…Lid

Claims

1. a transfer device that sucks the top surface of the lid of the egg carton, lifts and moves the egg carton, and places the egg carton in a predetermined position in a container for transportation; an imaging device that images an area including the predetermined position after the egg carton is placed; a control device; The control device Based on the captured image, at least one of the following abnormalities is detected within the area: a lid of the egg carton is removed; and an egg is exposed outside the carton. An abnormality detection device for egg carton transfer that instructs the transfer device to perform the next operation based on the detection results.

2. 2. The device for detecting abnormalities during egg carton transfer according to claim 1, wherein the control device further detects at least one of the following: that the egg cartons are not placed in a predetermined position; and that the egg cartons are tilted.

3. The device for detecting abnormalities during transfer of egg cartons according to claim 1 or 2, wherein the region includes a space around the transport container.

4. 4. The device for detecting abnormalities during egg pack transfer according to claim 1, wherein the photographing device photographs each time the transfer device places an egg pack at the predetermined position.

5. If the detection result indicates that there is no abnormality in the area, the next operation is an operation of continuing the processing by the transfer device, An abnormality detection device for transferring egg packs as described in any one of claims 1 to 4, wherein if the detection result indicates that there is an abnormality within the area, the next action is to stop processing by the transfer device.

6. The device for detecting abnormalities during egg carton transfer according to any one of claims 1 to 5, wherein the photographing device is provided in the transfer device.

Images