JP2020186082A - Washing machine identification system - Google Patents

Abstract

To provide a washing machine identification system capable of classifying a washing machine arranged in an installation site of a container or the like into a fully automatic type washing machine and a drum type washing machine.SOLUTION: A washing machine identification system 10 comprises: an imaging device 1; a robot 4 having a hand capable of holding a washing machine 5; an identification unit 12 for classifying the washing machines 5 into a fully automatic type washing machine 5a and a drum type washing machine 5b; and a control unit 13 for controlling the robot 4 so as to convey the fully automatic type washing machine 5a from an installation site 6 to a first position 7 and convey the drum type washing machine 5b from the installation site 6 to a second position 8. The imaging device 1 picks up images of the plurality of washing machines 5 arranged in the installation site 6 with their side faces facing upward from above. The identification unit 12 classifies the washing machine 5 into the fully automatic type washing machine 5a and the drum type washing machine 5b based on the teaching by an operator and/or image information acquired by the imaging device 1.SELECTED DRAWING: Figure 1

Description

The present invention relates to a washing machine identification system that classifies washing machines by type.

At the recycling plant, washing machines are being dismantled in order to recycle them. The washing machine includes a housing, a washing tub, a cover, an operation panel, a motor, and the like. The washing machine housing, washing tub, and cover are recycled into synthetic resin. The operation panel and motor of the washing machine are recycled into metal.

Washing machines are classified into fully automatic washing machines, drum-type washing machines, two-tank washing machines, dryers and the like. Since the washing machines have similar structures for each type, if the washing machines can be classified for each type, the efficiency of the dismantling work can be improved. As a conventional method for identifying a washing machine, a method of classifying a washing machine into a fully automatic washing machine (one-tank type) and a two-tank type washing machine is disclosed (see Patent Document 1). The invention of Patent Document 1 classifies a washing machine into a fully automatic washing machine and a two-tank washing machine based on the ratio of the width dimension and the depth dimension of the front surface of the washing machine.

Japanese Unexamined Patent Publication No. 11-304458

In recent years, the demand for drum-type washing machines has increased. Most of the washing machines brought into the recycling plant are fully automatic washing machines, but the number of drum-type washing machines is increasing. A fully automatic washing machine and a drum-type washing machine are carried into the recycling plant in a state of being loaded in a container or the like. The structure of the fully automatic washing machine and the drum-type washing machine are different, and the dismantling method is also different. In order to improve the efficiency of dismantling work, it is desirable to classify washing machines into fully automatic washing machines and drum type washing machines.

Therefore, in the present invention, the washing machines arranged at the installation location of the container or the like are classified into at least a fully automatic washing machine and a drum type washing machine, and the fully automatic washing machine is transported to the first position to be a drum type washing machine. It is an object of the present invention to provide a washing machine identification system capable of transporting a washing machine to a second position.

In order to solve the above problems, one aspect of the present invention includes an imaging device that images a plurality of washing machines arranged at an installation location with the side facing upward from above, and a hand that can hold the washing machines. The robot, the recognition unit that generates the position information and the posture information of the washing machine based on the image information acquired by the imaging device, and / or the image information acquired by the imaging device based on the instruction by the operator. Based on the identification unit that classifies the washing machine into at least a fully automatic washing machine and a drum type washing machine, the position information and posture information input from the recognition unit, and the classification information input from the identification unit. , A washing machine identification including a control unit that controls the robot so that the fully automatic washing machine is transported from the installation location to the first position and the drum-type washing machine is transported from the installation location to the second position. It is a system.

According to the present invention, the washing machines arranged at the installation location are classified into at least a fully automatic washing machine and a drum type washing machine, the fully automatic washing machine is transported to the first position, and the drum type washing machine is the first. It can be transported to two positions. In addition, since the image pickup device images the side shape of the washing machine, it is possible to easily distinguish between the fully automatic washing machine and the drum type washing machine, and the classification accuracy of the fully automatic washing machine and the drum type washing machine is improved. Can be made to.

It is a figure which shows the structure of the washing machine identification system of one Embodiment of this invention. It is a side view of a container. It is an image showing the washing machine inside the container seen from above. FIG. 4A is an external perspective view of the fully automatic washing machine, and FIG. 4B is an external perspective view of the drum type washing machine. FIG. 5A is a perspective view showing a hand adsorbed on the side surface of the washing machine, and FIG. 5B is a bottom view of the hand. FIG. 6A is a diagram showing a first conveyor and a fully automatic washing machine seen from the transport direction, and FIG. 6B is a diagram showing a second conveyor and a drum type washing machine seen from the transport direction. .. It is a figure which shows the neural network. It is a figure which shows the hardware configuration of the upper control unit.

Hereinafter, the washing machine identification system according to the embodiment of the present invention will be described in detail with reference to the accompanying drawings. However, the washing machine identification system of the present invention can be embodied in various forms, and is not limited to the embodiments described in the present specification. The present embodiment is provided with the intention of allowing those skilled in the art to fully understand the scope of the invention by making sufficient disclosure of the specification.
(Configuration of washing machine identification system)

FIG. 1 shows the configuration of the washing machine identification system 10 according to the embodiment of the present invention. The washing machine identification system 10 includes a camera 1 as an image pickup device, an upper control device 2, a robot controller 3, and a robot 4.

Reference numeral 6 denotes a container in which the washing machine 5 is loaded, and is a place where the washing machine 5 is installed. The washing machine 5 is loaded into the container 6 with the side surface facing up. The washing machine 5 includes at least a fully automatic washing machine 5a and a drum type washing machine 5b. Reference numeral 7 denotes a first conveyor on which the fully automatic washing machine 5a is conveyed by the robot 4, which is the first position. Reference numeral 8 denotes a second conveyor on which the drum-type washing machine 5b is conveyed by the robot 4. FIG. 1 shows a plan view of the robot 4, the container 6, the first conveyor 7, and the second conveyor 8.

The camera 1 is arranged above the container 6 (see the side view of FIG. 2). The camera 1 takes an image of the plurality of washing machines 5 loaded in the container 6 from above, and acquires three-dimensional image information of the plurality of washing machines 5. FIG. 3 shows the washing machine 5 inside the container 6 as viewed from above. The image of FIG. 3 corresponds to the image information captured by the camera 1. The image of FIG. 3 is matched with the operating coordinates of the robot 4 (XYZ coordinates in the figure).

The host controller 2 performs processing related to the operation of the robot 4 based on the image information acquired by the camera 1, and outputs an operation command to the robot controller 3. The upper control device 2 includes a recognition unit 11, an identification unit 12, an operation planning unit (control unit) 13, a display device 14, and an input device 15.

The recognition unit 11 performs image processing on the image information acquired by the camera 1 and individually recognizes each of the plurality of washing machines 5. Then, the recognition unit 11 generates the position information and the posture information of each washing machine 5, and outputs the position information and the posture information to the motion planning unit 13. The position information is the XYZ coordinates of the center of the side surface of each washing machine 5, and the posture information is the inclination around the X-axis, the Y-axis, and the Z-axis of each washing machine 5. Further, the recognition unit 11 generates shape information on the side surface of each washing machine 5 and outputs the shape information to the identification unit 12.

The identification unit 12 classifies the washing machine 5 into at least a fully automatic washing machine 5a and a drum type washing machine 5b based on the instruction by the operator, and outputs the classification information to the operation planning unit 13. As shown in FIG. 3, an image of the side surface shape of the washing machine 5 is displayed on the display device 14. The operator uses an input device 15 such as a touch panel and a mouse while referring to the image displayed on the display device 14, and the washing machine 5 is a fully automatic washing machine 5a or a drum type on the image of the display device 14. We will teach you how to classify the washing machine 5b. When the operator teaches the classification, the teaching result is displayed on the display device 14. For example, as shown in FIG. 3, one ellipse 21 is displayed on the fully automatic washing machine 5a, and two ellipses 22 are displayed on the drum washing machine 5b. The operator also teaches the bottom of the washing machine 5 on the image of the display device 14 using the input device 15. When the operator teaches the bottom, the display device 14 displays the teaching result. For example, as shown in FIG. 3, ellipses 21 and 22 are displayed on the bottom of the washing machine 5.

As shown in FIG. 1, the identification unit 12 classifies the washing machine 5 into a fully automatic washing machine 5a and a drum type washing machine 5b based on the shape information of the washing machine 5 input from the recognition unit 11. The information is output to the operation planning unit 13. Specifically, the identification unit 12 inputs the shape information of the washing machine 5 into the learned learning machine that has undergone machine learning, so that the washing machine 5 becomes a fully automatic washing machine 5a and a drum-type washing machine 5b. Classify (details will be described later).

FIG. 4A shows an external perspective view of the fully automatic washing machine 5a, and FIG. 4B shows an external perspective view of the drum type washing machine 5b. The lid 23 of the fully automatic washing machine 5a is attached to the upper surface 5a1. The lid 24 of the drum-type washing machine 5b is attached to the front surface 5b2. In the drum type washing machine 5b, a curved inclined surface is formed from the upper surface 5b1 to the front surface 5b2. The fully automatic washing machine 5a and the drum-type washing machine 5b have similar shapes of the upper surface 5a1, 5b1, the front surface 5a2, 5b2, the back surface 5a3, 5b3, and the bottom surface 5a5, 5b5, but the shapes of the side surfaces 5a4, 5b4 are different. .. The classification accuracy can be improved by the operator classifying the fully automatic washing machine 5a and the drum type washing machine 5b with reference to the side surface shape of the washing machine 5 displayed on the display device 14. Further, the identification unit 12 classifies the fully automatic washing machine 5a and the drum type washing machine 5b based on the shape information of the side surface of the washing machine 5, so that the classification accuracy can be improved.

In addition to the fully automatic washing machine 5a and the drum type washing machine 5b, the container 6 may be loaded with a two-tank type washing machine and a dryer. These may be classified into a fully automatic washing machine 5a and conveyed to the first conveyor 7, or these may be classified separately from the fully automatic washing machine 5a and the drum type washing machine 5b, and the third conveyor and the like (not shown) may be classified. May be transported to.

As shown in FIG. 1, the motion planning unit 13 plans the operation to be performed by the robot 4 based on the position information and the posture information input from the recognition unit 11 and the classification information input from the identification unit 12. The operation command generated thereby is output to the robot controller 3. The motion planning unit 13 controls the robot 4 so as to transport the fully automatic washing machine 5a from the container 6 to the first conveyor 7 and the drum-type washing machine 5b from the container 6 to the second conveyor 8.

The recognition unit 11, the identification unit 12, and the operation planning unit 13 of the upper control device 2 may be realized by the CPU executing the program described later, or may be realized by the ASIC, FPGA, or other dedicated integrated circuit. ..

The robot controller 3 generates a drive signal to be output to the motors of each axis of the robot 4 based on the operation command input from the operation planning unit 13 of the host controller 2.

The robot 4 is an articulated robot. As shown in FIG. 5A, a hand 31 capable of holding the washing machine 5 is attached to the tip of the arm 4a of the robot 4. The hand 31 sucks the side surface of the washing machine 5. An inspection cover 26 is detachably attached to the back surface of the washing machine 5. As described above, the lids 23 and 24 that can be opened and closed are attached to the upper surface 5a1 of the fully automatic washing machine 5a and the front surface 5b2 of the drum-type washing machine 5b. Therefore, even if the back surface, the upper surface, and the front surface of the washing machine 5 are attracted, the cover 26 and the lids 23 and 24 are only removed, and the washing machine 5 cannot be lifted. By sucking the side surface of the washing machine 5, both the fully automatic washing machine 5a and the drum type washing machine 5b can be sucked and lifted by the hand 31.

The hand 31 includes an intermediary portion 33 connected to the vacuum source via a tube 32, a metal main body portion 34, and a sponge pad 35 attached to the main body portion 34. The main body 34 is formed in the shape of a quadrangular plate. As shown in FIG. 5B, a large number of evacuation holes 34a communicating with the vacuum source are formed on the lower surface of the main body 34. A sponge pad 35 is attached to the lower surface of the main body 34 by adhesion or the like. The sponge pad 35 also has a quadrangular plate shape. The sponge pad 35 is formed with a large number of suction holes 35a corresponding to a large number of evacuation holes 34a in the main body 34. The washing machine 5 can be sucked by making the vacuum pulling hole 34a of the main body 34 and the suction hole 35a of the sponge pad 35 negative pressure by the vacuum source. By attaching the sponge pad 35 to the main body 34, as shown in FIG. 5A, even if the side surface of the washing machine 5 is uneven, the side surface of the washing machine 5 can be sucked and lifted.

As shown in FIG. 1, the motion planning unit 13 is a robot so that the bottom of the fully automatic washing machine 5a conveyed to the first conveyor 7 faces the same direction based on the bottom information input from the identification unit 12. 4 is controlled. FIG. 6A shows the first conveyor 7 and the fully automatic washing machine 5a as viewed from the transport direction. When disassembling the fully automatic washing machine 5a, first, the top cover and the operation panel of the fully automatic washing machine 5a are removed on the first conveyor 7. As shown in FIG. 1, by aligning the directions of the bottoms of the fully automatic washing machine 5a, one worker can efficiently remove the top cover and the operation panel in an assembly line. After removing the top cover and operation panel, remove the washing tub and motor of the fully automatic washing machine 5a. The washing tub, the motor, etc. may be removed while the fully automatic washing machine 5a is placed on the first conveyor 7, or may be removed from the first conveyor 7.

As shown in FIG. 1, the motion planning unit 13 uses the robot 4 so that the bottom of the drum-type washing machine 5b conveyed to the second conveyor 8 faces the same direction based on the bottom information input from the identification unit 12. To control. FIG. 6B shows the second conveyor 8 as viewed from the transport direction. The drum-type washing machine 5b is unloaded from the second conveyor 8 by the worker. The dismantling of the drum-type washing machine 5b is performed by a dedicated worker.
(Classification processing by the identification unit)

As described above, the identification unit 12 inputs the shape information of the washing machine 5 into the learned learning machine that has undergone machine learning, so that the washing machine 5 becomes a fully automatic washing machine 5a and a drum type washing machine 5b. Classify. The learner comprises a neural network. In the neural network, for example, as shown in FIG. 7, the washing machine 5 is a fully automatic washing machine 5a and a drum type washing machine based on the image information of a two-dimensional binary pixel string input as the shape information of the washing machine 5. It is designed to output whether it is classified into a machine 5b, a two-tank washing machine 5c, or a dryer 5d.

The learner needs to machine-learn the correspondence between the input of the shape information of the washing machine 5 and the output of the classification before the execution mode in which the classification is actually performed. Machine learning is performed by adjusting the weighting coefficient of the edge that connects each node in the neural network. Specifically, teacher data using the shape information of the washing machine 5 as input data and the classification of the fully automatic washing machine 5a, drum type washing machine 5b, two-tank washing machine 5c, and dryer 5d by the operator as output data. Adjusts the weighting coefficient. As a result, the learner can learn the feature amount representing the correlation between the shape information of the washing machine 5 and the classification.

When the learning of the learner is completed, the execution mode becomes executable. In this execution mode, by inputting the shape information of the side surface of the washing machine 5 into the learned learning machine, the washing machine 5 can be made into a fully automatic washing machine 5a, a drum type washing machine 5b, a two-tank type washing machine 5c, and a drying machine. Classify into machine 5d. The learner is not limited to the neural network, but may be a Bayesian network or a support vector machine.
(Hardware configuration of host controller)

The hardware configuration of the host controller 2 will be described. As shown in FIG. 8, the host control device 2 includes a CPU 41, a ROM 42, a RAM 43, an input device 15, a display device 14, and a recording device 44. These are connected to each other so that signals can be transmitted via a bus 45 or an interface (not shown).

The program is recorded in the ROM 42, the RAM 43, and the recording device 44. When the CPU 41 executes the program, the processing of the recognition unit 11, the identification unit 12, and the operation planning unit 13 is executed.

The display device 14 is a monitor or the like. The CPU 41 causes the display device 14 to display an image of the washing machine 5 in the container 6. The input device 15 is a touch panel, a mouse, a keyboard, or the like. The CPU 41 executes the processing of the identification unit 12 based on the signal input from the input device 15.

The present invention is not limited to the above embodiment, and can be changed to other embodiments without changing the gist of the present invention.

In the above embodiment, the fully automatic washing machine loaded in the container is conveyed to the first conveyor, and the drum type washing machine loaded in the container is conveyed to the second conveyor, but the fully automatic washing machine loaded in the container is conveyed. The type washing machine may be transported to a container dedicated to the fully automatic washing machine, and the drum type washing machine loaded in the container may be transported to the container dedicated to the drum type washing machine.

In the above embodiment, the identification unit classifies the washing machine into a fully automatic washing machine and a drum type washing machine based on the instruction from the operator and the image information acquired by the camera, but the instruction from the operator or the camera Based on either one of the acquired image information, the identification unit may classify the washing machine into a fully automatic washing machine and a drum type washing machine.

1 ... Camera (imaging device)
4 ... Robot 6 ... Container (installation location)
5 ... Washing machine 5a ... Fully automatic washing machine 5a4 ... Side of fully automatic washing machine 5b ... Drum type washing machine 5b4 ... Side of drum type washing machine 7 ... First conveyor (first position)
8 ... 2nd conveyor (2nd position)
10 ... Washing machine identification system 11 ... Recognition unit 12 ... Identification unit 13 ... Operation planning unit (control unit)
14 ... Display device 15 ... Input device 31 ... Hand 34 ... Main body 34a ... Vacuum drawing hole 35 ... Sponge pad 35a ... Suction hole

Claims (6)

An imaging device that captures images of multiple washing machines placed at the installation location with the sides facing up from above,
A robot with a hand that can hold the washing machine,
A recognition unit that generates position information and posture information of the washing machine based on the image information acquired by the image pickup device.
An identification unit that classifies the washing machine into at least a fully automatic washing machine and a drum type washing machine based on the instruction by the operator and / or the image information acquired by the imaging device.
Based on the position information and attitude information input from the recognition unit and the classification information input from the identification unit, the fully automatic washing machine is conveyed from the installation location to the first position, and the drum type washing machine is moved. A washing machine identification system including a control unit that controls the robot so as to transport the robot from the installation location to the second position. The washing machine identification system
A display device that displays an image captured by the image pickup device, and
The first aspect of the invention is characterized in that the image displayed on the display device includes at least an input device for receiving instructions from an operator for classifying the washing machine into a fully automatic washing machine and a drum type washing machine. Described washing machine identification system. The identification unit according to claim 1 or 2, wherein the identification unit classifies the washing machine into at least a fully automatic washing machine and a drum type washing machine by inputting the image information into the learned learning machine. Washing machine identification system. The identification unit identifies the bottom of the fully automatic washing machine and
The control unit is characterized in that the robot is controlled so that the bottom of the fully automatic washing machine transported to the first position faces the same direction based on the bottom information input from the identification unit. The washing machine identification system according to any one of claims 1 to 3. The washing machine identification system according to any one of claims 1 to 4, wherein the hand sucks the side surface of the washing machine. The hand
A metal body with a large number of evacuation holes that communicate with the vacuum source,
The washing according to claim 5, further comprising a sponge pad attached to the main body portion, having a large number of suction holes corresponding to the large number of evacuation holes, and capable of contacting the side surface of the washing machine. Machine identification system.

Images