JPH06218682A - Robot for assembly - Google Patents

Abstract

PURPOSE:To provide an assembly robot which is provided with the image recognizing function which obviates the necessity of the shift operation of a TV camera directly over an assembly objective article before the shift of a chuck hand directly over the assembly objective article, shortens the operation time of the robot and carries out the monitoring and recognization for the assemly object, even during the assembly work by the robot. CONSTITUTION:The grasping center of a nipping-holding means 3 and the optical axis 7 of an image input means 1 are allowed to coincide, and the image information which shows the situation of the outside field of a robot is inputted into the image input means 1 through a through hole 2a farmed on the first assembly member 2 which is nipping-held by the nipping-holding means 3.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an industrial assembling robot, and more particularly to an assembling robot having a visual function of recognizing a position and a state of an object to be assembled from real-time image information.

[0002]

2. Description of the Related Art The industrial robots currently in practical use are structurally classified into three-dimensional articulated robots, horizontal articulated robots, vertical articulated robots, Cartesian coordinate robots and the like. The 3D articulated robot belongs to a high-grade robot that has an arm with a large degree of freedom in the 3D space, and is intended for complex operations in the 3D space such as arc welding and spray painting. Used in.

The horizontal articulated robot, which is also called a SCARA robot, has an arm having a plurality of rotatable joints in a horizontal plane and is highly functional for its price. Therefore, it is particularly suitable for assembling parts. There is. The vertical articulated robot is a robot having an arm having a plurality of rotatable joints in a vertical plane. Cartesian coordinate type robots are easy to teach and position and are used for automatic loading of electronic components.

On the other hand, with the development of computer technology and image processing technology, it has become possible for a robot to have a machine's eyes to recognize a work target and have a judgment ability. As a result, the robot has the ability to input information from the outside world, recognize it, understand it, and deal with problems that occur, rather than simply repeating a predetermined sequence of programs.

An example of a conventional horizontal articulated assembly robot with an image recognition device is shown in FIG. In the figure, an upper arm portion 62 is pivotally supported by a robot body 6 whose base is fixed to the floor surface by a shoulder joint 61 which is rotatable in a horizontal plane. A forearm portion 64 is pivotally supported at the tip of the upper arm portion 61 by an elbow joint 63 that is also rotatable in a horizontal plane.

An actuator 66 is pivotally supported at a position opposite to the elbow joint 63 from the center of the forearm 64 by a wrist joint 65 which is also rotatable in a horizontal plane.
The actuator unit 66 is provided with a chuck hand 13 that is vertically slidable.
It has a structure that can be opened and closed so that an object to be assembled can be clamped or released.

A television camera 11, which is an image input device, is installed in the foremost portion of the forearm portion 64, which is in front of the wrist joint 65, with the optical axis directed vertically downward, and the image of the outside world of the robot is imaged. The signal is input to the image processing apparatus.

As described above, the reason why the television camera is provided at the tip of the robot arm is that there is nothing obstructing the field of view of the television camera at the tip of the arm and the television camera can be easily attached. is there. Therefore, the optical axis of the television camera lens and the central position of the grip of the chuck hand 13 were separated by a distance (X) on the horizontal plane.

FIG. 6 shows the operation of such a horizontal articulated robot having a television camera at the tip of an arm when it is used for assembling parts in a uniaxial linear assembly in which a work is passed through a shaft. . Here, the uniaxial linear assembly relation means a structure in which several parts are arranged around one axis as shown in FIG. Such an assembly relationship is widely used for motors, bearings, automobile parts, and the like, and is the most general form of parts assembly.

The assembly sequence for passing the work 2 through the shaft 4 in FIG. 6 is shown in order from (A) to (G). (A) The TV camera 11 is moved right above the work 2 to be handled from the original position (home position), and the work 2 is photographed by the TV camera 11 and converted into an image signal.
Next, the image signal is taken into the image memory, and the image processing apparatus performs the feature extraction of the image, the comparison with the reference, and the determination to determine the presence / absence of the work, the front / back of the work, the quality of the work, etc. )).

(B) If there is no problem in this determination, the chuck hand 13 is moved right above the work 2. (C) Next, the chuck hand 13 is lowered to the work 2, the work 2 is grasped, and the chuck hand 13 is raised again (FIG. 7B).

(D) Next, the television camera 11 is moved right above the shaft 4. Next, take a picture of the shaft 4,
Image processing is performed to recognize the position of the shaft 4 (see FIG. 7).
(C)). (E) Next, the chuck hand 13 is moved right above the shaft 4.

(F) Next, the chuck hand 13 is lowered to insert the work 2 into the shaft 4. Next, the chuck hand 13 is opened, the work 2 is released, and the chuck hand 13 is raised (FIG. 7 (d)). (G) Next, the chuck hand 13 is returned to the original position. This completes one cycle of operation. In the above operation, the movement of the chuck hand 13 in the horizontal direction follows the route of (a) → (b) → (c) → (d) shown in FIG.

[0014]

However, in the conventional assembly robot with the image recognition device, the optical axis of the television camera and the grasping center of the chuck hand are not coaxial with each other, so that the chuck hand is moved before the movement. It is necessary to move the television camera directly above the object to be assembled, and the time loss due to this movement is an obstacle to improving the work efficiency of the robot.

Further, since the chuck hand is moved based on the recognition data after recognizing the assembly member, errors such as positional deviation are likely to occur, and the television camera leaves the object during the work. However, there was a problem that it could not be confirmed.

In view of the above problems, the main object of the present invention is to provide an assembly robot which eliminates the loss of time for moving the television camera prior to moving the chuck hand. Another object of the present invention is to provide an assembly robot having an image recognition function capable of monitoring and confirming an object to be assembled even during the assembly work of the robot.

[0017]

According to the present invention, the first
Holding means for holding the second assembly member, holding means for holding or placing the second assembly member, image input means for inputting the external environment of the robot as image information, and the image input means. Image recognition means for recognizing the external situation of the robot based on the image information.
In the assembling robot for assembling the assembling member and the second assembling member, the image input means is provided to the first assembling member sandwiched by the sandwiching means, with the image information indicating the external environment of the robot. The above problem is solved by allowing input through the through hole.

Further, in the present invention, the first assembly member can be pinched and moved so that the second assembly member is inserted into the through hole provided in the first assembly member. Further, in the present invention, the optical axis of the optical system of the image input means may be provided so as to face vertically downward.

[0019]

The television camera, which is an image input device for the robot to recognize the outside world, is provided coaxially with the grasping center of the chuck hand of the robot, so that the television camera can move at the same time as the movement of the chuck hand. There is no time lost to move the television camera directly above the assembly object prior to moving it directly above the assembly object. Also, since the TV camera is vertically above the grip center of the chuck hand, the TV camera and the chuck hand do not interfere with each other, and the assembly work by the chuck hand can be carried out through the through hole of the assembly member held by the chuck hand. You can monitor and confirm in time.

[0020]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows a positional relationship between a chuck hand 3 provided at the tip of an arm of a horizontal articulated robot and a television camera 1. An optical axis 7 of the television camera 1 is provided vertically downward, and a chuck hand 3 is provided so that the optical axis 7 coincides with a grasping center. As a result, while holding the workpiece 2 with the chuck hand 3,
An image of the shaft 4 placed on the mounting table 10 through the hole 2a provided in the work 2 can be used as an input image of the television camera 1. The range indicated by 5 in FIG.
It is a range in which the television camera 1 can capture an image through the hole 2a. In addition, from the robot body 6, the shoulder joint 61 and the upper arm 6
2, the structure up to the elbow joint 63 and the forearm 64 is the same as the conventional structure, and the description thereof is omitted.

FIG. 3 is a block diagram of the control circuit of the robot of this embodiment. The image of the outside world of the robot is converted into a video signal by the television camera 1 and input to the next A-D conversion circuit 11. The A-D conversion circuit 11 converts the video signal into a digital signal and stores it in the image memory 12 as digital image information. The image processing CPU 13 extracts image features from the image information stored in the image memory 12,
Through an image processing process such as pattern recognition and scene understanding, the front and back of the assembly target and the position of the assembly target are determined.

The robot control CPU 14 executes the control program by referring to the information from the image processing CPU 13, and controls the actuator 16 such as the arm and the chuck hand 3 via the servo mechanism 15. Further, information such as the angle of each joint is input from the sensor 17 to the servo mechanism 15, converted into a digital signal and sent to the robot control CPU 14.

Next, the operation procedure of the robot of this embodiment will be described with reference to FIG. The object to be assembled is the same work 2 and shaft 4 as in FIG. 6 which is an explanation of the conventional example. The assembly sequence in FIG. 2 is shown from (a) to (e). (A) The chuck hand 3 is moved from the original position (home position) to just above the work 2 to be handled, and the work is photographed by the TV camera 1 that moves simultaneously with the chuck hand 3 and converted into an image signal. Next, AD the image signal
It is converted and captured in the image memory, and by the image processing device,
Image feature extraction, image recognition, comparison with a reference, determination, etc. are performed to determine the presence or absence of a work, the front and back of a work, the quality of a work, and the like.

(B) If there is no problem in this determination, the chuck hand 3 is moved right above the work 2. Next, the chuck hand 3 is lowered, the work 2 is grasped, and the chuck hand 3 is raised (FIG. 2 (b)). (C) Next, the TV camera 1 is moved to a position right above the shaft 4. Next, the shaft 4 is photographed, image processing is performed, and the position of the shaft 4 is recognized. If necessary as a result of this recognition, the horizontal position of the chuck hand 3 is corrected.

(D) Next, the chuck hand 3 is lowered to insert the work 2 into the shaft 4. After that, the chuck hand 3 is opened, the work 2 is released, and the chuck hand 3 is raised (FIG. 2C). (E) Next, the chuck hand 3 is returned to the original position. This completes one cycle of operation. The horizontal movement of the chuck hand 3 at this time is as shown in FIG.
(E) → (f) shown in (a).

Comparing the operation of the robot according to the present embodiment shown in the above a to the operation of the conventional robot shown in the above A to G, the following operations are shown: A-A, R-C, H-E and N-. -F, ho-
Each of G corresponds. From these, it is clear that the movement of the television camera prior to the movement of the chuck hand 3, which is equivalent to B and D in the conventional example, is unnecessary. Further, since the shaft 4 can be seen through the hole of the work 2 held by the chuck hand 3, the work 2
Since the television camera 1 can pick up the image and recognize the image even during the assembling of the shaft 4 and the shaft 4, it is possible to deal with the abnormality during the assembling.

Although the preferred embodiment has been described above, this is not meant to limit the invention. For example, in the example,
Although the application of the present invention to a horizontal articulated robot has been shown, it is within the scope of the present invention to apply it to any other type of robot as well as a vertical articulated robot.

[0028]

As described above, according to the present invention,
With the chuck hand positioned directly above the assembly target,
The assembly target can be viewed with a TV camera. This eliminates the need to move the TV camera directly above the assembly object before moving the chuck hand onto the assembly object, unlike the conventional robot with visual function, which shortens the operation time of the robot. effective. Further, according to the present invention, there is provided an assembling robot having an image recognition function which allows the television camera to capture the image of the object to be assembled as an image input even during the assembling work of the robot and can monitor and confirm the assembling work. The effect is that you can.

[Brief description of drawings]

FIG. 1 is a side view of a TV camera mounting portion according to an embodiment.

FIG. 2 is an operation explanatory diagram of the embodiment.

FIG. 3 is a block diagram of an embodiment.

FIG. 4 is an explanatory diagram of a 1-axis linear assembly relationship of an assembly target.

FIG. 5: Side view of conventional robot

FIG. 6 is an operation explanatory diagram of a conventional example.

FIG. 7 is an operation explanatory diagram of a conventional example.

[Explanation of symbols]

 1 image input means (television camera) 2 first assembly member (workpiece) 3 clamping means (chuck hand) 4 second assembly member (shaft) 6 robot body 7 optical axis 10 mounting table

Claims (3)

[Claims] 1. A holding means for holding the first assembly member, a holding means for holding or placing the second assembly member, an image input means for inputting the external environment of the robot as image information, An assembly robot for assembling the first assembly member and the second assembly member, comprising image recognition means for recognizing an external situation of the robot based on image information input to the image input means, The assembly robot, wherein the image input means can input image information indicating a situation of the outside world of the robot through a through hole provided in the first assembly member sandwiched by the sandwiching means. 2. The assembling robot according to claim 1, wherein the first assembly member is pinched and moved so that the second assembly member is inserted into a through hole provided in the first assembly member. . 3. The assembly robot according to claim 1 or 2, wherein the optical axis of the optical system of the image input means is provided vertically downward.