JP2021112809A - Production system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a production system capable of more multifaceted and efficient use of a robot and an automatic guided vehicle. SOLUTION: A machine tool 10 and a robot 35 having cameras 41a and 41b for capturing images and performing work on the machine tool 10 and a robot 35 mounted on the machine tool 10 are set for the machine tool 10. The automatic guided vehicle 45 via the position and the control device 50 for controlling the automatic guided vehicle 45 and the robot 35 are provided. The machine tool 10 has a holding table for holding the work machined by the machine tool 10, and the control device 50 operates at least one of the robot 35 and the automatic guided vehicle 45 to hold the work held on the holding table. The images of the above are captured by the cameras 41a and 41b from a plurality of imaging directions, and the three-dimensional shape of the work is recognized based on the obtained plurality of images. [Selection diagram] FIG. 5

Description

The present invention includes a machine tool that processes a workpiece, a robot that performs work on the machine tool, an automatic guided vehicle that carries a robot and passes through a work position set for the machine tool, and these robots and an unmanned vehicle. The present invention relates to a production system composed of a control device for controlling the operation of an automatic guided vehicle.

Conventionally, a production system called FMS (Flexible Manufacturing System) has been known. In this production system, for example, a plurality of machine tools are appropriately arranged in a factory, and an unmanned transport vehicle is used to transport workpieces and tools before processing to each machine tool, and the unmanned transport vehicle is also used. By using it, the workpieces processed in each machine tool and used tools are collected. And, by such a production system, unmanned automatic production is realized.

As the automatic guided vehicle, the automatic guided vehicle disclosed in Japanese Patent Application Laid-Open No. 2019-8359 is conventionally known. This unmanned carrier is a distance measuring device that rotates and drives a light projecting unit that emits projected light and outputs distance measurement data based on the light received by the projected light reflected by the object to be measured, and distance measurement data. A map creation unit that creates map information based on the map information and an obstacle sensor that detects an obstacle are provided, and the obstacle sensor is arranged in a measurement invalid area in a circular area including a measurement effective area of the distance measuring device.

According to this automatic guided vehicle, first, when the automatic guided vehicle is operated by manual operation and traveled in the building of the factory, the factory is based on the distance measurement data measured by the distance measuring device. The map creation unit creates planar map information that defines the inner contour of the building and the outer contour of the machine machine or other device arranged in the building.

In addition, the automatic guided vehicle identifies its own current position (self-position) in the building based on the distance measurement data measured by the distance measuring device and the map information created in advance by the map creating unit. Then, the automatic guided vehicle autonomously travels on a predetermined route along a predetermined route based on the map information and the self-position information in the building, and transports the above-mentioned workpieces and tools.

In recent years, as disclosed in Japanese Patent Application Laid-Open No. 2017-132002, a production system in which a robot equipped with a camera is mounted on the above-mentioned automatic guided vehicle has been proposed. In this production system, an automatic guided vehicle equipped with a robot moves to a work position set for the machine tool, and at the work position, the robot executes work such as attaching / detaching a work to the machine tool. NS.

In such a production system, one robot moved by an automatic guided vehicle can perform work such as attaching / detaching a work to a plurality of machine tools, so that the robot is fixed to the machine tool. Since the degree of freedom in the layout of the machine tool is increased as compared with the case where the machine tool is arranged in the state, the layout of the machine tool can be set to a layout capable of further improving the production efficiency. In addition, compared to the conventional production system in which the robots are arranged in a fixed state, one robot can work on more machine tools, so that the equipment cost can be reduced. ..

JP-A-2019-8359 Japanese Unexamined Patent Publication No. 2017-132002

By the way, in the above-mentioned production system, the automatic guided vehicle equipped with the robot travels in a wide range in the factory building where many machine tools are arranged. Such an automatic guided vehicle is set so that the mounted robot attaches / detaches a work to a machine tool, etc., but the robot is a high-performance equipment equipped with a camera or the like, and the robot and the automatic guided vehicle are transported. It would be convenient if these could be used for other purposes during the free time of the car, in other words, during the waiting time, and it would be beneficial because it would lead to the efficient use of the equipment.

The present invention has been made under such a background, and in a production system using an automatic guided vehicle equipped with a robot, the robot and the automatic guided vehicle can be used more multifacetedly and efficiently. The purpose is to provide a production system that has been made possible.

The present invention for solving the above problems
Machine tools that perform predetermined machining on workpieces and
A robot that has a camera that captures images and works on the machine tool,
An automatic guided vehicle equipped with the robot and passing through a work position set for the machine tool.
A production system including the automatic guided vehicle and a control device for controlling a robot.
The machine tool has a holding table for holding a workpiece machined by the machine tool.
The control device operates at least one of the robot and the automatic guided vehicle to capture an image of the work held on the holding table from a plurality of imaging directions by the camera, and based on the obtained plurality of images. The present invention relates to a production system configured to recognize a three-dimensional shape of the work.

According to the production system of this aspect (first aspect), the work held on the holding table is imaged by the camera under the control of the control device, and at that time, the robot and the automatic guided vehicle are imaged. By operating at least one of them, the work is imaged by the camera from a plurality of imaging directions. Then, the control device recognizes the three-dimensional shape of the work based on the plurality of images captured by the camera. Any known process can be applied to the process of recognizing the three-dimensional shape of the work from a plurality of images.

Further, the present invention includes a machine tool that performs predetermined processing on a work and a machine tool.
A robot that has a camera that captures images and works on the machine tool,
An automatic guided vehicle equipped with the robot and passing through a work position set for the machine tool.
A production system including the automatic guided vehicle and a control device for controlling a robot.
The machine tool includes a holding table for holding a workpiece machined by the machine tool, and a rotating mechanism for rotating the holding table.
The control device transmits an operation signal for rotating the rotation mechanism to the machine tool to rotate the holding table, and the camera captures images of the work from a plurality of imaging directions, and a plurality of obtained images are obtained. The present invention relates to a production system configured to recognize the three-dimensional shape of the work based on the above.

According to the production system of this aspect (second aspect), under the control of the control device, the work held on the holding table is imaged by the camera, and at that time, the rotation mechanism is rotated. By transmitting a signal to the machine tool and rotating the holding table, the work is imaged by the camera from a plurality of imaging directions. Then, the control device recognizes the three-dimensional shape of the work based on the plurality of images captured by the camera.

Further, the present invention includes a machine tool that performs predetermined processing on a work and a machine tool.
A robot that has a camera that captures images and works on the machine tool,
An automatic guided vehicle equipped with the robot and passing through a work position set for the machine tool.
A production system including the automatic guided vehicle and a control device for controlling a robot.
The machine tool includes a holding table for holding a workpiece machined by the machine tool, and a rotating mechanism for rotating the holding table.
The control device transmits an operation signal for rotating the rotation mechanism to the machine tool to rotate the holding table, and at the same time, operates at least one of the robot and the automatic guided vehicle to be held by the holding table. The present invention relates to a production system configured to capture an image of a work by the camera from a plurality of imaging directions and to recognize a three-dimensional shape of the work based on the obtained plurality of images.

According to the production system of this aspect (third aspect), the work held on the holding table is imaged by the camera under the control of the control device, and at that time, the rotation mechanism is rotated. By transmitting a signal to the machine tool to rotate the holding table and operating at least one of the robot and the automatic guided vehicle, the camera images the work from a plurality of imaging directions. Then, the control device recognizes the three-dimensional shape of the work based on the plurality of images captured by the camera.

In the production system of the first to third aspects, the control device recognizes the three-dimensional shape of the work as described above, so that, for example, the recognized work shape and the recognized work shape are registered (stored) in advance. The part number of the work is recognized by comparing with the work shape, and further, by collating the recognized part number of the work with the part number of the work currently scheduled to be machined by the machine tool, the holding table can be used. It is determined whether or not the held work is a work scheduled to be machined, and if the work held on the holding table is a work not scheduled to be machined, an alarm is output. Processing can be performed. The product number of the work to be machined by the machine tool can be obtained from the numerical control device, for example, via communication with the numerical control device of the machine tool.

In the holding table, the work machined by the machine tool is generally supplied and held by an automatic feeding device or an operator as appropriate, but sometimes due to a misunderstanding of the shape or the like, the shape of the holding table Another very similar work (work that is not scheduled to be processed) may be mixed. In such a case, when this unscheduled work is machined by a machine tool, there is an inconvenience that the tool and the work collide with each other to damage them, or even if they do not collide, the work becomes a defective product. According to the production system according to the present invention, it is possible to confirm whether or not the work held on the holding table is the work scheduled to be machined in the machine tool, so that the above-mentioned inconvenience occurs. It can be prevented in advance.

In addition, the dimensions of the workpiece after machining can be calculated by appropriately processing the captured image, and by calculating the dimensions of the workpiece in this way, it is confirmed whether or not the machining is properly performed. can do. Further, the control device recognizes the part number of the work, and when the part number of the work is scheduled to be machined, the operation program executed by itself is planned based on the part number of this work. It is possible to confirm whether or not the operation program has been executed, and if it is an unscheduled operation program, it is possible to stop the operation and execute a process such as outputting an alarm.

In each of the production systems of the first to third aspects described above, it is preferable that the plurality of imaging directions are orthogonal to each other.

As described above, in each production system according to the present invention, the three-dimensional shape of the work is recognized by the control device based on the plurality of images captured by the camera. Then, by recognizing the three-dimensional shape of the work in this way, for example, the part number of the work is recognized, and further, the work held on the holding table is scheduled to be machined based on the recognized part number of the work. It is possible to determine whether or not the work is being machined, and if the work is not scheduled to be machined, it is possible to perform processing such as outputting an alarm.

When an unplanned work is machined by a machine tool, there is an inconvenience that the tool and the work collide with each other and they are damaged, or even if they do not collide, the work becomes defective, but it is held by the holding table. By confirming whether or not the work is a work to be machined, it is possible to prevent such inconvenience from occurring.

In addition, the dimensions of the workpiece after machining can be calculated by appropriately processing the captured image, and by calculating the dimensions of the workpiece in this way, it is confirmed whether or not the machining is properly performed. can do. In addition, when the product number of the work is scheduled to be machined, the control device determines whether or not the operation program it is executing is the scheduled operation program based on the product number of this work. It can be confirmed, and if it is an unscheduled operation program, the operation can be stopped and processing such as outputting an alarm can be executed.

It is a top view which showed the production system which concerns on one Embodiment of this invention. It is a perspective view which showed the machine tool and the like which concerns on this Embodiment. It is a front view which showed the holding table, the pallet, etc. which concerns on this embodiment in a partial cross section. It is a perspective view which showed the automatic guided vehicle and the robot which concerns on this embodiment. It is a block diagram which showed the control apparatus and the like which concerns on this Embodiment. In this embodiment, it is explanatory drawing for demonstrating the process of recognizing the shape of a work. It is a perspective view which showed the holding table which concerns on other embodiment of this invention.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. As shown in FIGS. 1 and 5, the production system 1 of this example includes at least one machine tool 10, a work stocker 33, a robot 35, an automatic guided vehicle 45 equipped with the robot 35, and these robots. It is configured to include a control device 50 that controls 35 and an automatic guided vehicle 45.

As shown in FIGS. 1 and 2, the machine tool 10 of this example is a horizontal machining center, and is a portion corresponding to one side of a T-shaped bed 11 and a bed 11 in a plan view (the longitudinal direction in FIG. 1 is the lateral direction). The column 12 and the bed 11 are erected on the other side of the column 12 and the bed 11 so as to be reciprocating in the X-axis direction. A saddle 16 provided so as to be reciprocating in the Z-axis direction, two pallets Pa on which mounting jigs J are arranged, and a pallet changing device 20 provided at the other end of the bed 11. It also includes a pallet holding device 25, a numerical control device (not shown), and the like.

On the surface of the column 12 on the saddle 16 side, a spindle holding base 13 is arranged so as to be reciprocating in the Y-axis direction indicated by an arrow, and the spindle head 14 is held by the spindle holding base 13. Then, the spindle 15 is rotatably held at the center of the spindle at the spindle head 14, and the tool T is held at the spindle 15. Further, the saddle 16 holds the pallet Pa horizontally rotatable and detachable, and the prismatic mounting jig J is fixedly mounted on the upper surface of the pallet Pa. Work Ws are attached to each side surface of the attachment jig J.

Then, under the control of the numerical control device (not shown), the column 12, the spindle holding base 13, and the saddle 16 are rotated in the X-axis direction and the Y-axis, respectively, in a state where the spindle 15 and the tool T are rotated about the axis. By moving in the direction and the Z-axis direction, the tool T and the work W move relatively in the X-axis direction, the Y-axis direction, and the Z-axis direction, whereby the work W is machined by the tool T.

As shown in FIG. 3, the pallet holding device 25 horizontally rotates a holding base 27 that holds the pallet Pa detachably, a base 26 that holds the holding base 27 horizontally and rotatablely, and the holding base 27. It is composed of a rotation mechanism 28 and the like. Further, the rotation mechanism 28 is centered on a worm wheel 30 attached to the lower end of the holding base 27, a worm 29 that meshes with the worm wheel 30, a drive shaft 31 connected to the worm 29, and the drive shaft 31. The worm wheel 30 rotates horizontally by rotating the drive shaft 31 and the worm 29 about the axis by the drive motor 32, which causes the holding base 27 and the pallet Pa to rotate horizontally. ..

The pallet changing device 20 has an H-shape in a plan view and includes a swivel arm 21 provided so as to be horizontally rotatable, and the pallet Pa held on the holding table 27 is one of the comb-shaped arms 21b. , 21b. The pallet changing device 20 is in a state where the saddle 16 is moved to the swivel arm 21 side and the pallet Pa held by the saddle 16 is located between the comb-shaped arms 21a and 21a on the other side. By moving the arm 21 upward, that is, by lifting it up, both pallets Pa are lifted upward, and then the swivel arm 21 is horizontally rotated by 180 °, and then the swivel arm 21 is moved downward. The pallet Pa held in the holding table 27 is held in the saddle 16, and the pallet Pa held in the saddle 16 is held in the holding table 27. That is, the pallet Pa held in the saddle 16 and the pallet Pa held in the holding table 27 are exchanged.

In the machine tool 10, the robot 35 attaches / detaches the work W to / from the pallet Pa held on the holding table 27. That is, the state in which the AGV 35 is positioned at the working position P _1, processed workpiece W mounted on the mounting jig J is removed from the mounting jig J by the robot 35, then the AGV after the vehicle 35 has moved to the working position P _2, it is stored in the stocker 33 by the robot 35. Then, with the AGV 35 is positioned at the working position P _2, the stocker 33 to the stowed raw workpiece W is taken out by the robot 35, then, the AGV 35 is the working position P _{After moving to 1} , the raw work W is attached to the attachment jig J by the robot 35. The work stocker 33 has an accommodating portion for accommodating the work W, and both the unprocessed work and the processed work are accommodated in this accommodating portion.

As shown in FIGS. 1 and 4, the automatic guided vehicle 45 is equipped with the robot 35 mounted on a mounting surface 46 which is an upper surface thereof, and an operation panel 47 which can be carried by an operator. The operation panel 47 includes an input / output unit for inputting / outputting data, an operation unit for manually operating the automatic guided vehicle 45 and the robot 35, a display capable of displaying a screen, a speaker for outputting sound, and the like.

Further, the automatic guided vehicle 45 is provided with a sensor capable of recognizing its own position in the building of the factory (for example, a distance measurement sensor using a laser beam), and is built under the control of the control device 50. configured indoor so that the vehicle travels in trackless, in this example, via the working position P ₁ and P ₂ is set to at least the machine tool 10.

As shown in FIGS. 1 and 4, the robot 35 is an articulated robot having three arms, a first arm 36, a second arm 37, and a third arm 38, and is a tip of the third arm 38. A hand 39 as an end effector is attached to the portion, and two cameras 41a and 41b are attached via a support bar 40. The cameras 41a and 41b are composed of, for example, a CCD area image sensor.

The control device 50 is housed in the housing of the unmanned transport vehicle 45, and as shown in FIG. 5, the operation program storage unit 51, the moving position storage unit 52, the operation posture storage unit 53, and the map information storage unit 54. , Work shape storage unit 55, manual operation control unit 56, automatic operation control unit 57, map information generation unit 58, position recognition unit 59, work shape recognition unit 60, and input / output interface 61. The control device 50 is connected to the machine tool 10, the work stocker 33, the robot 35, the cameras 41a and 41b, the automatic guided vehicle 45, and the operation panel 47 via the input / output interface 61.

The control device 50 is composed of a computer including a CPU, RAM, ROM, etc., and includes the manual operation control unit 56, the automatic operation control unit 57, the map information generation unit 58, the position recognition unit 59, and the work shape recognition unit 60. The function of the output interface 61 is realized by a computer program, and a predetermined process is executed. Further, the operation program storage unit 51, the movement position storage unit 52, the operation posture storage unit 53, the map information storage unit 54, and the work shape storage unit 55 are composed of appropriate storage media such as RAM.

The manual operation control unit 56 is a functional unit that operates the automatic guided vehicle 45 and the robot 35 according to an operation signal input from the operation panel 47 by the operator. That is, the operator can manually operate the automatic guided vehicle 45 and the robot 35 by using the operation panel 47 under the control of the manual operation control unit 56.

The operation program storage unit 51 operates the automatic guided vehicle 45 when generating an automatic driving program for automatically driving the automatic guided vehicle 45 and the robot 35 at the time of production, and map information in a building to be described later. It is a functional part that stores a program for generating a map. The automatic operation program and the map generation program are, for example, input from the input / output unit provided on the operation panel 47 and stored in the operation program storage unit 51.

In addition, this automatic driving program includes a command code regarding a moving position, a moving speed, and a direction of the automatic guided vehicle 45 as a target position for the automatic guided vehicle 45 to move, and the operation of the robot 35 in which the robot 35 operates sequentially. Contains the directive code for. Then, in this example, with respect to the pallet Pa held on the holding table 27, the command code relating to the above-mentioned attachment / detachment operation for attaching / detaching the work W by the robot 35 is included, and the cameras 41a and 41b provided in the robot 35 are used. A command code relating to an imaging operation for capturing an image of the work W at indexing positions I ₂ , I ₃ , and I _{4 of the holding table 27 is included.} In addition, the map generation program includes a command code for running the automatic guided vehicle 45 on an untracked route throughout the building so that the map information generation unit 58 can generate map information.

The map information storage unit 54 is a functional unit that stores map information including layout information of the building on which the automatic guided vehicle 45 travels and the machine tools 10 and the like arranged in the building, and the map information is the map. It is generated by the information generation unit 58.

The map information generation unit 58 runs the automatic guided vehicle 45 according to the map generation program stored in the operation program storage unit 51 under the control of the automatic operation control unit 57 of the control device 50, which will be described in detail later. At that time, the spatial information in the building is acquired from the distance data detected by the sensor, and the plane shape of the machine tool 10 or the like arranged in the building is recognized, for example, the pre-registered machine tool. Based on the plane shape of the machine 10 or the like, the position, plane shape, etc. (arrangement information) of the specific machine tool 10 or the like arranged in the building are recognized. Then, the map information generation unit 58 stores the obtained spatial information and the arrangement information of the machine tool 10 and the like in the map information storage unit 54 as map information in the building.

The position recognition unit 59 is a functional unit that recognizes the position of the automatic guided vehicle 45 in the building based on the distance data detected by the sensor and the map information in the building stored in the map information storage unit 54. The operation of the automatic guided vehicle 45 is controlled by the automatic operation control unit 57 based on the position of the automatic guided vehicle 45 recognized by the position recognition unit 59.

The moving position storage unit 52 is a moving position as a specific target position for the automatic guided vehicle 45 to move, and is a functional unit that stores a specific moving position corresponding to a command code in the operation program. The moving position includes the work positions for the robot 35 to perform work on the machine tool 10 via the machine tool 10 described above, that is, the work positions P ₁ and P _{2 described above.} .. The movement position is determined, for example, by manually driving the automatic guided vehicle 45 with the operation panel 47 under the control of the manual operation control unit 56 to move the automatic guided vehicle 45 to each target position, and then recognizing the position. The position data recognized by the unit 59 is set by the operation of storing the position data in the moving position storage unit 52. This operation is called a so-called teaching operation.

The motion posture storage unit 53 is a posture (motion posture) of the robot 35 that changes sequentially when the robot 35 operates in a predetermined order, and is data related to the motion posture corresponding to the command code in the motion program. It is a functional part that memorizes. The data related to this operating posture is obtained when the robot 35 is manually operated to take each target posture by a teaching operation using the operation panel 47 under the control of the manual operation control unit 56. It is the rotation angle data of each joint (motor) of the robot 35 in each posture, and this rotation angle data is stored in the operation posture storage unit 53 as data related to the operation posture.

Then, in this example, the pallet Pa held on the holding table 27 is held by using the data of the operation posture in the attachment / detachment operation of attaching / detaching the work W by the robot 35 and the cameras 41a and 41b provided in the robot 35. The data of the operating posture in the imaging operation of imaging the work W at the indexing positions I ₂ , I ₃ , and I _{4 of the table 27 is stored in the operating posture storage unit 53.}

The automatic driving control unit 57 is a functional unit that uses either the automatic driving program or the map generation program stored in the operation program storage unit 51 to operate the automatic guided vehicle 45 and the robot 35 according to the program. .. At that time, the data stored in the moving position storage unit 52 and the operating posture storage unit 53 are used as needed.

The work shape storage unit 55 is a functional unit that stores data related to the shape of the work W, for example, dimensional data or three-dimensional model data of the work W, and a product number of the work W in association with each other. It is input from the input / output unit provided on the operation panel 47 and stored in the work shape storage unit 55.

The work shape recognition unit 60 executes a process of recognizing the shape of the work W based on the images captured by the cameras 41a and 41b. It should be noted that any known process for recognizing the three-dimensional shape of the work from a plurality of images can be applied to this shape recognition process. Hereinafter, an example thereof will be briefly described.

For example, when the shape recognition target workpiece W is in the indexing position I ₂ of holder 27, the camera 41a, an image captured by 41b is FIG. 6 (a), as the image shown in (b) .. Further, it is assumed that the cameras 41a and 41b image the target work W with their optical axes along the Z axis.

The cameras 41a and 41b include a plurality of photoelectric conversion elements arranged in two dimensions in multiple rows and multiple columns, digitize the voltage signal output from each photoelectric conversion element according to the light receiving intensity, and then shade the voltage signals. It is converted into a level value and output as two-dimensional shading image data having the same arrangement as the arrangement of the photoelectric conversion elements. Then, the work shape recognition unit 60 receives the two-dimensional shading image data output from the cameras 41a and 41b.

The work shape recognition unit 60 recognizes the shape of the target work W by processing such as generating three-dimensional model data of the target work W based on the received two-dimensional shading image data. Specifically, first, each received two-dimensional image data is binarized at a predetermined threshold value to extract each image of the target work W, and the extracted binarized image is scanned in the raster direction to obtain 2 An image (contour line image) related to the contour line of the digitized image is extracted (see FIGS. 6 (a) and 6 (b)).

Next, the coordinate position of the actual contour line of the target work W in the three-dimensional space is calculated by using the triangulation method based on the two contour line images extracted respectively. Here, a method of calculating the coordinate position of the actual contour line of the target work W in the three-dimensional space by the triangulation method will be briefly described.

For example, the two cameras 41a and 41b that capture the images shown in FIGS. 6A and 6B and the substance of the target work W have a positional relationship shown in FIG. 6C on the XZ plane. It is in. In FIG. 6C, Fa is the central surface of each lens of the cameras 41a and 41b, and Ga is each image plane. Further, a is the lens center point of the camera 41a, and b is the lens center point of the camera 41b. Ra is the distance between the lens center surface Fa and the image plane Ga, and La is the distance between the lens center points a and b of the two cameras 41a and 41b, which are precisely measured in advance. ing. c ₁ and d ₁ are axes parallel to the Z axis and pass through points a and b, respectively.

Now, a reference point and Aa in the entity of the target workpiece W, when the axis c _1, d ₁ sets the line connecting the point of intersection with the imaging surface Ga between the reference point Aa, and each line and the lens center surface Fa Cross at Aa ₁ and Aa _{2, respectively.} That is, the light received from the reference point Aa to the two cameras 41a and 41b _{intersects at the Aa 1} point and the Aa ₂ point on the central surface of each lens, and is polarized by each lens to form the Aa ₁ point and the Aa ₂ The image is formed on each imaging plane from the point in _{parallel with the axes c 1} and d _1. It should be noted that the _{points Aa 1} and Aa ₂ correspond to the portions indicated by the same signs in the contour line images shown in FIGS. 6 (a) and 6 (b), respectively.

Therefore, if the reference point Aa each camera 41a, the lens center point a is imaged on the imaging surface of the 41b, the distance X _1, X ₂ from b divided by triangulation, by the following equation 1 each The distance Za in the Z-axis direction from the cameras 41a and 41b to the actual body of the target work W can be calculated. The distances X ₁ and X ₂ can be easily calculated from the contour line images shown in FIGS. 6 (a) and 6 (b) captured and processed by the cameras 41a and 41b. However, c ₂ and d ₂ in the figure are axes parallel to the Y axis and pass through the lens center points a and b, respectively.

(Formula 1)
Za = Ra (La-X _1- X ₂ ) / (X ₁ + X ₂ )

Then, the dimensions of the target work W are calculated from the calculated distance Za between the cameras 41a and 41b and the target work W, and the relationship between the distance Za preset by calibration and the like and the magnification of the image. That is, the shape and dimensions can be recognized by calculating the coordinate positions of the vertices of the actual contour image of the target work W shown in FIG. 6A.

The work shape recognition unit 60 recognizes the shape and dimensions of the target work W by the above processing. Similarly, the work shape recognition unit 60 processes the images captured by the cameras 41a and 41b when the target work W is at the indexing positions I ₃ and I _{4 of the holding table 27, and in the same manner as described above.} , Recognize the shape and dimensions of the target work W in each imaging direction. Then, the work shape recognition unit 60 calculates the overall three-dimensional shape data of the target work W by synthesizing the shape data of the target work W in the obtained three orthogonal directions, and 3 of the target work W. Recognize dimensional shapes.

Next, the work shape recognition unit 60 searches the data stored in the work shape storage unit 55 based on the calculated three-dimensional shape data of the target work W, recognizes the product number of the target work W, and recognizes the product number of the target work W. By collating the recognized product number with the product number of the work W currently being machined obtained from the numerical control device (not shown) of the machine tool 10, the target work W is scheduled to be machined by the machine tool 10. Determine if it is a working work. Then, when the target work W is a work that is not scheduled to be machined in the machine tool 10, the work shape recognition unit 60 performs a process of outputting an alarm to the operation panel 47.

According to the production system 1 of this example having the above configuration, the automatic operation program stored in the operation program storage unit 51 is executed under the control of the automatic operation control unit 57 of the control device 50. The automatic guided vehicle 45 and the robot 35 operate according to this automatic driving program.

That is, the automatic guided vehicle 45 in accordance with an automatic operation program, the way in the machine tool 10 and the working position _P 1 which is set respectively stocker 33, _{P 2,} at each work position _P 1, _{P 2} , The robot 35 performs the above-mentioned work on the machine tool 10 and the work stocker 33.

Then, in this example, the rotation mechanism 28 is driven under the control of the numerical control device (not shown) of the machine tool 10, and the (recognition) target work W attached to the attachment jig J of the holding base 27 is placed. indexing each indexed position _I 2, _{I 3} and _{I 4,} each indexing position _I 2, _{I 3} and the target workpiece W that has been indexed to _{I 4,} the camera 41a by assume an appropriate imaging posture to the robot 35, Image is taken by 41b.

Then, the work shape recognition unit 60 processes the images captured by the cameras 41a and 41b from the three orthogonal directions as described above, calculates the overall three-dimensional shape data of the target work W, and obtains the data. The product number of the target work W is recognized by referring to the data stored in the work shape storage unit 55 based on the three-dimensional shape data, and the recognized product number and the numerical control device of the machine tool 10 (shown in the figure). By collating with the part number of the work W currently being machined obtained from (1), it is determined whether or not the target work W is a work scheduled to be machined by the machine tool 10. Then, when the target work W is a work that is not scheduled to be machined in the machine tool 10, the work shape recognition unit 60 performs a process of outputting an alarm to the operation panel 47.

Generally, the work W is appropriately put into the work stocker 33 by an automatic supply device or an operator, but sometimes due to a misunderstanding of the shape or the like, another work having a shape very similar (a work not scheduled to be machined). ) May be mixed. In such a case, when an unscheduled work is attached to the mounting jig J and machined by the machine tool 10, even if the tool T and the work W collide with each other and they are damaged or do not collide. The inconvenience that the work W becomes a defective product occurs. According to the production system 1 of this example, it is possible to confirm whether or not the work W attached to the mounting jig J on the holding table 27 is the work W scheduled to be machined by the machine tool 10. It is possible to prevent such inconvenience from occurring.

Further, if the processed work W is taken as the target work W by the cameras 41a and 41b, the obtained image is processed by the work shape recognition unit 60, and the dimensions of the processed work W are calculated. , It is possible to confirm whether or not the machining is properly performed in the machine tool 10.

Further, as a result of recognizing the product number of the work W, if the product number of the work W is scheduled to be machined, the control device 50 plans an operation program executed by itself based on this product number. It is possible to confirm whether or not the operation program has been executed, and if it is an unscheduled operation program, it is possible to stop the operation and output an alarm to the operation panel 47. ..

Although the specific embodiments of the present invention have been described above, the modes that can be adopted by the present invention are not limited to the above-mentioned modes.

For example, in the above example, when the target work W is imaged from three orthogonal directions by the cameras 41a and 41b, the operation of rotating the holding table 27 and the operation of causing the robot 35 to take an imaging posture are executed together. However, the image is not limited to this, and the image is taken by a combined operation of the operation of rotating the holding table 27 and the operation of moving the automatic guided vehicle 45 to an appropriate position while keeping the posture of the robot 35 constant. You can do it.

Alternatively, the target work W may be imaged from three orthogonal directions by keeping the rotation position of the holding table 27 constant and operating either the robot 35 or the automatic guided vehicle 45.

Further, in the above example, the target work W is imaged from three directions, but the present invention is not limited to this, and the target work W may be imaged from a plurality of directions of two or more directions. Calculates the three-dimensional shape data of the target work W based on the obtained image data.

Further, in the above example, the work shape storage unit 55 stores the shape data of the work W and the product number in association with each other, and the work shape recognition unit 60 calculates the three-dimensional shape data of the target work W. Based on this three-dimensional shape data, the work shape storage unit 55 is referred to to recognize the product number of the target work W, but the present invention is not limited to this, and the work shape storage unit 55 may have a plurality of directions. The work shape recognition unit 60 acquires image data from a plurality of directions of the target work W and acquires the acquired image, while being configured to correlate and store the model data of the work W and the product number viewed from the above. Based on the data, the work shape storage unit 55 is referred to, and for example, the model data of the work W in a plurality of directions is collated with the image data by a pattern matching method so that the product number of the target work W is recognized. You may.

Further, in the above example, a rotation mechanism 28 is provided, and the holding base 27 is configured to rotate horizontally by the rotation mechanism 28. However, the present invention is not limited to this, and FIG. 7 shows, instead of the rotation mechanism 28. As shown, for example, the rod-shaped member 34 is erected at four corners of the pallet Pa, the rod-shaped member 34 is grasped by the hand 39 of the robot 35, and the pallet Pa is rotated by the operation of the robot 35. Is also good.

Again, the description of the embodiments described above is exemplary in all respects and not restrictive. Modifications and changes can be made as appropriate for those skilled in the art. The scope of the present invention is shown not by the above-described embodiment but by the scope of claims. Further, the scope of the present invention includes modifications from the embodiment within the scope of the claims and within the scope of the claims.

1 Production system 10 Machine tool 11 Bed 12 Column 13 Spindle holder 14 Spindle head 15 Spindle 16 Saddle 20 Pallet exchange device 21 Swivel arm 25 Pallet holding device 26 Base 27 Holding base 28 Rotating mechanism 33 Workstocker 35 Robot 39 Hand 41a, 41b Camera 45 Automated guided vehicle 47 Operation panel 50 Control device 51 Operation program storage unit 52 Movement position storage unit 53 Operation posture storage unit 54 Map information storage unit 55 Work shape storage unit 56 Manual operation control unit 57 Automatic operation control unit 58 Map information Generation part 59 Position recognition part 60 Work shape recognition part Pa pallet J Mounting jig

Claims (4)

Machine tools that perform predetermined machining on workpieces and
A robot that has a camera that captures images and works on the machine tool,
An automatic guided vehicle equipped with the robot and passing through a work position set for the machine tool.
A production system including the automatic guided vehicle and a control device for controlling a robot.
The machine tool has a holding table for holding a workpiece machined by the machine tool.
The control device operates at least one of the robot and the automatic guided vehicle to capture an image of the work held on the holding table from a plurality of imaging directions by the camera, and based on the obtained plurality of images. , A production system characterized in that it is configured to recognize the three-dimensional shape of the work. Machine tools that perform predetermined machining on workpieces and
A robot that has a camera that captures images and works on the machine tool,
An automatic guided vehicle equipped with the robot and passing through a work position set for the machine tool.
A production system including the automatic guided vehicle and a control device for controlling a robot.
The machine tool includes a holding table for holding a workpiece machined by the machine tool, and a rotating mechanism for rotating the holding table.
The control device transmits an operation signal for rotating the rotation mechanism to the machine tool to rotate the holding table, and the camera captures images of the work from a plurality of imaging directions, and a plurality of obtained images are obtained. A production system characterized in that it is configured to recognize the three-dimensional shape of the work based on the above. Machine tools that perform predetermined machining on workpieces and
A robot that has a camera that captures images and works on the machine tool,
An automatic guided vehicle equipped with the robot and passing through a work position set for the machine tool.
A production system including the automatic guided vehicle and a control device for controlling a robot.
The machine tool includes a holding table for holding a workpiece machined by the machine tool, and a rotating mechanism for rotating the holding table.
The control device transmits an operation signal for rotating the rotation mechanism to the machine tool to rotate the holding table, and at the same time, operates at least one of the robot and the automatic guided vehicle to be held by the holding table. A production system characterized in that an image of a work is imaged by the camera from a plurality of imaging directions, and the three-dimensional shape of the work is recognized based on the obtained plurality of images. The production system according to any one of claims 1 to 3, wherein the plurality of imaging directions are orthogonal to each other.

Images