Robot teaching device and robot teaching method

Abstract

To provide a robot teaching device and a robot teaching method that enable a robot to execute a series of work at high work efficiency.SOLUTION: A feature amount obtaining part 11 obtains feature amount data 110 showing degrees of influence of operation of a worker 20 on objects 21a and 21b to be worked. A focus period discriminating part 12 discriminates, as focus periods 120a-120d, periods of time during which the worker 20 performs operation which is indispensable to the series of work, of periods of time during which operation data 100 is obtained, on the basis of the feature amount data 110. A teaching data generating part 13 generates teaching data 130 that is inputted to robots 30a and 30b so that the operation of the worker 20 in the focus periods 120a-120d is reproduced and so that operation of the worker 20 in a period other than the focus periods 120a-120d is not reproduced, on the basis of the operation data 100.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a robot teaching device and a robot teaching method. [Background technology]

[0002] The introduction of robots is progressing with the aim of reducing the labor required and automating various on-site tasks. In order to have a robot perform a desired task, it is necessary to teach it. To reduce the man-hours required for this teaching work, many technologies have been proposed that use devices such as cameras to capture the movements of a teacher, such as by imitating others, and then convert them into robot movements.

[0003] For example, Patent Document 1 provides a robot teaching device that detects the movements of a worker's fingers from an image acquired by an image input device and creates a control program for a robot that reproduces the work content. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 6038417 Summary of the Invention [Problem to be solved by the invention]

[0005] However, when the robot teaching device disclosed in Patent Document 1 tries to make the robot execute the worker's movements exactly, it also reproduces unnecessary movements such as the worker's hand tremors, hesitation, and mistakes, which reduces the robot's work efficiency. Also, when estimating the worker's movements and replanning the movements to make it easier for the robot to perform the work, it becomes difficult to completely reproduce the worker's intentions, especially in skilled work or work that requires precision.

[0006] The present invention has been made in consideration of the above-mentioned problems, and its purpose is to provide a robot teaching device and a robot teaching method that are capable of making a robot perform the series of tasks with high work efficiency. [Means for solving the problem]

[0007] In order to achieve the above object, the present invention provides a robot teaching device that teaches a robot a series of tasks to be performed by a worker on a work object, comprising: a motion acquisition unit that acquires motion data indicating the position and posture of the worker or the work object; a feature acquisition unit that acquires feature data that indicates the degree of influence that the worker's motion has on the work object; a key period determination unit that, based on the feature data, determines, within the period in which the motion data was acquired, a period in which the worker is performing motions that are essential to the series of tasks as a key period; and a teaching data generation unit that generates teaching data to be input to the robot based on the motion data so that the worker's motions during the key period are reproduced and the worker's motions outside the key period are not reproduced.

[0008] Furthermore, the present invention provides a robot teaching method for teaching a robot a series of tasks to be performed by a worker on a work object, comprising the steps of: acquiring motion data indicating the position and posture of the worker or the work object; acquiring feature data indicating the degree of influence of the worker's motion on the work object; determining, based on the feature data, a period during which the worker is performing motions essential to the series of tasks within the period during which the motion data was acquired as a key period; and generating teaching data to be input to the robot based on the motion data so that the worker's motions during the key period are reproduced and the worker's motions outside the key period are not reproduced. [Effects of the Invention]

[0009] According to the present invention, it is possible to have a robot perform the series of tasks with high work efficiency. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a diagram showing an overview of a robot teaching device according to a first embodiment of the present invention. [Figure 2] FIG. 4 is a diagram showing how an instructor's action is acquired in the first embodiment of the present invention. [Figure 3] FIG. 2 is a diagram showing how a robot in the first embodiment of the present invention is made to perform a teaching task. [Figure 4] FIG. 3 is a diagram showing a part of the action data acquired by the action acquisition unit in the first example of the present invention. [Figure 5] FIG. 2 is a diagram showing a glove to be worn by an instructor in the first embodiment of the present invention. [Figure 6] FIG. 3 is a diagram showing an example of a key period determination method in the first embodiment of the present invention. [Figure 7] FIG. 3 is a diagram showing an example of processing in a teaching data generating unit in the first embodiment of the present invention. [Figure 8] FIG. 3 is a diagram showing a modified example of the robot teaching device in the first embodiment of the present invention. [Figure 9] FIG. 10 is a diagram showing a glove to be worn by an instructor in a second embodiment of the present invention. [Figure 10] FIG. 10 is a diagram showing an example of a method for determining a key period in a second embodiment of the present invention. [Figure 11] FIG. 10 is a diagram illustrating a configuration of a key period determination unit in a third embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. In each drawing, the same elements are designated by the same reference numerals, and duplicated explanations will be omitted as appropriate. [Example]

[0012] 1 shows an overview of a robot teaching device 1 according to a first embodiment of the present invention. The robot teaching device 1 is a device in which an operator 20, acting as an instructor, demonstrates a task involving handling work objects 21a and 21b, thereby having a robot 30 perform the same task. The robot teaching device 1 is comprised of a computer with a processing function, an input / output interface for inputting and outputting signals to and from external devices, and the functions of each part are realized by executing a program stored in a storage device such as a hard disk drive.

[0013] The motion acquisition unit 10 acquires the motion of the instructor 20 or the measurement values ​​of the positions and postures of the work objects 21a, 21b as motion data 100 and outputs it to the teaching data generation unit 13. The feature acquisition unit 11 acquires feature data 110 for identifying key motions (described later) from the manner of instruction by the instructor 20 and outputs it to the key period determination unit 12. The key period determination unit 12 determines a period in which a key motion is being performed as a key period based on the feature data 110 and outputs the determination result (key period data 120) to the teaching data generation unit 13. The teaching data generation unit 13 creates teaching data 130 from the motion data 100 so that only the motions in the key periods are reproduced and that, for motions in other periods, the robots 30a, 30b follow optimal paths, and inputs this teaching data to the robots 30a, 30b.

[0014] A key motion refers to a motion that is essential for performing a series of taught tasks. For example, in the case of a stamping task, robots 30a and 30b can perform the stamping task by reproducing the hand movements of instructor 20 in such actions as holding the stamp, pressing the stamp onto the vermilion ink pad, pressing the stamp onto the paper, and placing the stamp. However, it is not necessary to reproduce the hand movements of instructor 20 in such actions as moving the stamp and checking the stamping position. The robot teaching device 1 in this embodiment extracts only the motions essential for the series of tasks from the imitation teaching and teaches them to robots 30a and 30b, thereby improving the work efficiency of robots 30a and 30b.

[0015] 2 shows how the actions of instructor 20 are acquired. It is assumed that instructor 20 grasps work objects 21a and 21b in front of work table 22 and performs a predetermined task. A structure 23 is placed on work table 22, and it is assumed that instructor 20 will either move to avoid interference between grasped work objects 21a and 21b and structure 23, or perform some kind of contact action such as pressing or rubbing.

[0016] The motion of the instructor 20 is acquired by a motion acquisition device 101. The motion acquisition device 101 is, for example, a camera, a positioning sensor, or a motion capture system. The motion data 100 acquired by the motion acquisition device 101 is six-dimensional time-series data of the position (x, y, z) and the attitude (roll, pitch, yaw) as viewed from a work table coordinate system 24 defined on the work table 22.

[0017] FIG. 3 shows how robots 30a and 30b are made to perform a teaching operation. The work environment, including the work object 21, work table 22, and structures 23a and 23b, is assumed to be the same as that at the time of teaching in FIG. 2. The positional relationship between the robots 30a and 30b and the work table coordinate system 24 is assumed to be known by being fixed to a uniquely determined position in advance or by being recognized by a camera or the like. When the teaching data 130 is executed, the joint angles of the robots 30a and 30b are calculated so that the positions and postures of the hands of the robots 30a and 30b match those of the teaching data 130, and the teaching operation is reproduced by executing these in chronological order.

[0018] According to the above configuration, the motion data 100 during the teaching work is acquired by the motion acquisition unit 10, and the robots 30a, 30b are caused to execute the motion data 100, thereby making it possible to make the robots 30a, 30b reproduce the teaching work.

[0019] 4 shows a portion of the six-dimensional motion data 100 acquired by the motion acquisition unit 10. Here, the horizontal axis is the motion data time axis 1000, and the vertical axis is the motion data displacement axis 1001, which indicates the position change on the x-axis. This data may contain unnecessary motions such as trembling or mistakes by the instructor 20, but it is considered difficult to distinguish these from the motion data 100 alone.

[0020] FIG. 5 shows a glove 111 worn by the instructor 20. A characteristic of the key motions assumed in this embodiment, such as grasping and pressing, is a change in the contact force between the work objects 21a, 21b and the instructor's 20's hand. Therefore, the contact force between the work objects 21a, 21b and the instructor's 20's hand is measured as a value (feature amount) indicating the degree of influence of the worker's 20's motion on the work objects 21a, 21b, and a period during which the contact force (feature amount) changes is extracted as a period during which the key motion is being performed (key period). In this embodiment, the change in the contact force (feature amount) is acquired as feature amount data 110 via the glove 111 worn by the instructor 20.

[0021] A plurality of pressure-sensitive sensors 1110a-1110g are arranged on the glove 111, and when the action acquisition unit 10 acquires the action data 100, the measured values ​​of the pressure-sensitive sensors 1110a-1110g are simultaneously acquired as feature data 110. Therefore, the feature data 110 in this embodiment is six-dimensional time-series data. In the example of FIG. 5, a plurality of pressure-sensitive sensors 1110a-1110g are arranged on the pads of each finger and on the palm, but the positions and number of sensors are not important.

[0022] FIG. 6 shows an example of a method for determining a key period.

[0023] 6(a) is a graph showing feature amount data 110 acquired by one of the pressure sensors 1110a to 1110g on the glove 111. Here, the horizontal axis is a feature amount data time axis 1100 on the same scale as the action data time axis 1000, and the vertical axis is a feature amount data axis 1101.

[0024] 6(b) shows a method for determining a key period. First, feature amount data 110 is differentiated with respect to time to obtain feature amount data time differential absolute value data 112. Here, the horizontal axis is feature amount data time axis 1100, and the vertical axis is feature amount data time differential absolute value axis 1120. Next, a threshold value 121 for determining a key period is set, and periods in which the feature amount data time differential absolute value exceeds the threshold value 121 are determined to be key periods 120a to 120d.

[0025] According to the above configuration, it is possible to extract a period in which a change occurs in the contact force between the work objects 21a and 21b and the instructor's hand.

[0026] FIG. 7 shows an example of processing in the teaching data generating unit 13.

[0027] 7(a) shows key periods 120a to 120d of the motion data 100. In order for the robots 30a and 30b to achieve the purpose of a series of tasks, it is necessary to reproduce the motion data 100 as it is in the teaching data 130 without changing the displacement and time scale of the motion data 100 during the key periods 120a to 120d, but it is not necessary to reproduce other periods.

[0028] 7(b) shows an example of teaching data 130 in which paths are created so that movements other than those in the key periods 120a to 120d are linear in the hand space. Here, the horizontal axis is a teaching data time axis 1300 on the same scale as the movement data time axis 1000, and the vertical axis is a teaching data displacement axis 1301.

[0029] The teaching data 130 for periods other than the key periods 120a to 120d is generated so as to result in the movement of the shortest path taking into account the displacement from the final value of the immediately preceding key period to the initial value of the immediately following key period, the range of motion of the robots 30a and 30b, the maximum allowable speed, and the maximum allowable acceleration. Furthermore, to reproduce the movements of the instructor 20 during the key periods 120a to 120d, the time axes of all six-dimensional elements of the movement data 100 during the key periods 120a to 120d must be aligned in the teaching data 130. Therefore, even if the displacement of a certain one-dimensional element from a key period to the next key period is zero, the value of the other one-dimensional element will be maintained until the other one-dimensional element reaches its initial value for the next key period.

[0030] The teaching data 130 for periods other than the key periods 120a to 120d may be linear in the joint space of the robots 30a and 30b, or may be a path that does not cause interference between the structures 23a and 23b, the work objects 21a and 21b, and the hands of the robots 30a and 30b.

[0031] Fig. 8 shows a modified example of the robot teaching device 1 in this embodiment. In Fig. 8, the robot teaching device 1 further includes a key period selection unit 14. The key period selection unit 14 displays key periods 120a to 120d determined by the key period determination unit 12 and the operation data 100, and also provides a means for the instructor 20 or other user to select a key period to be reproduced as teaching data from among the key periods 120a to 120d determined by the key period determination unit 12.

[0032] (summary) In this embodiment, a robot teaching device 1 that teaches robots 30a, 30b a series of tasks that a worker 20 performs on work objects 21a, 21b includes a motion acquisition unit 10 that acquires motion data 100 that indicates the position and posture of the worker 20 or the work objects 21a, 21b in the series of tasks, a feature acquisition unit 11 that acquires feature data 110 that indicates the degree of influence that the motion of the worker 20 has on the work objects 21a, 21b in the series of tasks, and a feature data acquisition unit 12 that acquires feature data 110 that indicates the degree of influence that the motion of the worker 20 has on the work objects 21a, 21b in the series of tasks. The robots 30a and 30b are provided with a key period determination unit 12 that determines, based on the data 110, periods during which movements essential to the series of tasks are being performed among the periods during which the movement data 100 was acquired, as key periods 120a to 120d, and a teaching data generation unit 13 that generates teaching data 130 to be input to the robots 30a and 30b based on the movement data 100 so that the movements of the worker 20 during the key periods 120a to 120d are reproduced and movements of the worker 20 outside the key periods 120a to 120d are not reproduced.

[0033] Furthermore, a robot teaching method for teaching robots 30a, 30b a series of tasks to be performed by a worker 20 on work objects 21a, 21b includes the steps of acquiring motion data 100 indicating the position and posture of the worker 20 or the work objects 21a, 21b, acquiring feature data 110 indicating the degree of influence of the motion of the worker 20 on the work objects 21a, 21b, determining, based on the feature data 110, periods during which the worker 20 performs motions essential to the series of tasks as key periods 120a to 120d within the period during which the motion data 100 was acquired, and generating teaching data 130 to be input to the robots 30a, 30b based on the motion data 100 so that the motions of the worker 20 during the key periods 120a to 120d are reproduced and motions of the worker 20 outside the key periods 120a to 120d are not reproduced.

[0034] According to the present embodiment configured as described above, the teaching data 130 for the robots 30a, 30b is generated so that, among the periods during which the motion data 100 of the worker 20 or the work objects 21a, 21b was acquired, the motion data 100 for the key periods 120a to 120d, during which the motion of the worker 20 had a high degree of influence on the work objects 21a, 21b, is reproduced, and the motion data 100 for the other periods is not reproduced. In this way, teaching data is generated that reproduces the motions of the worker that are essential for a series of tasks and does not reproduce other unnecessary motions, making it possible for the robots 30a, 30b to perform the series of tasks with high work efficiency.

[0035] The robot teaching device 1 in this embodiment also includes a glove 111 worn by the worker 20 and pressure sensors 1110a to 1110g attached to the glove 111, and the feature acquisition unit 11 acquires the measurement values ​​of the pressure sensors 1110a to 1110g as feature data 110. This makes it possible to extract key periods 120a to 120d based on the contact force between the work targets 21a, 21b and the hands of the worker 20.

[0036] Furthermore, the key period determination unit 12 in this embodiment determines, within the period during which the motion data 100 is acquired, periods during which the absolute time-differentiated values ​​of the measurement values ​​of the pressure sensors 1110a to 1110g are equal to or greater than a predetermined threshold value 121 as key periods 120a to 120d. This makes it possible to extract, as key periods 120a to 120d, periods during which a change occurs in the contact force between the work objects 21a, 21b and the hands of the worker 20.

[0037] Furthermore, the teaching data generator 13 in this embodiment generates teaching data 130 other than the key periods 120a to 120d so that the time axes of the elements of the operation data 100 during the key periods 120a to 120d match in the teaching data 130, and so that the robots 30a and 30b perform operations along the shortest path under the limitations of their range of motion, maximum allowable speed, and maximum allowable acceleration. This minimizes unnecessary movements in a series of tasks, thereby maximizing work efficiency.

[0038] Furthermore, the robot teaching device 1 in this embodiment includes a key period selection unit 14 that allows the worker 20 or other users to select whether or not to reproduce the operation data 100 for the period determined to be a key period by the key period determination unit 12 in the teaching data 130. This prevents the operation data 100 for the key period that the worker 20 or other users have determined not to need to be reproduced from the teaching data 130, thereby further improving the work efficiency of the robots 30a, 30b. [Example]

[0039] The configuration of the key period determination unit 12 in the second embodiment of the present invention will be described with reference to Figures 9 and 10. However, configurations not shown in Figures 9 and 10 are the same as those in the first embodiment.

[0040] Fig. 9 shows a glove 111A to be worn by the instructor 20 in this embodiment. A plurality of pressure sensors 1110a-1110g and bending sensors 1111a-1111e that measure bending angles of the fingers are arranged on the glove 111A. In Fig. 9, as an example, a plurality of pressure sensors 1110a-1110g are arranged on the pads of each finger and on the palm, and bending sensors 1111a-1111e are arranged on all fingers, but the positions and number of sensors are not important.

[0041] 10 shows an example of a method for determining a key period in this embodiment. Here, the horizontal axis represents a feature amount data time axis 1100A, and the vertical axis represents a feature amount data axis 1101A. In this embodiment, periods in which feature amount data 110A acquired by one of bending sensors 1111a to 1111e of glove 111A exceeds threshold value 121A are determined to be key periods 120Aa to 120Ac.

[0042] As described above, the key periods 120Aa to 120Ac determined by the bending sensors 1111a to 1111e and the key periods 120a to 120d determined by the pressure sensors 1110a to 1110g are integrated to obtain the final key period data 120.

[0043] (summary) The feature amount acquiring unit 11 in this embodiment acquires, as feature amount data 110, measurement values ​​of a plurality of types of sensors 1110a to 1110g and 1111a to 1111e that measure different physical amounts.

[0044] According to the present embodiment configured as described above, by acquiring feature amount data 110 using a plurality of types of sensors 1110a to 1110g, 1111a to 1111e that measure different physical quantities, it is possible to improve the accuracy of key period data 120. For example, in a situation where push buttons are provided on work objects 21a and 21b and the behavior of work objects 21a and 21b varies depending on the amount of depression of the buttons, this configuration is effective when instructor 20 selects a key period using key period selection unit 14 shown in FIG. [Example]

[0045] The configuration of the key period determination unit 12 in the third embodiment of the present invention will be described with reference to Fig. 11. However, the configuration not shown in Fig. 11 is the same as that in the first or second embodiment.

[0046] When multiple types of sensors (shown in FIG. 9) are used as in the second embodiment (shown in FIG. 9), the increase in the number of sensors increases the number of key periods determined by the key period determination unit 12. This makes it difficult for the instructor 20 or other users to select key periods to be reproduced as teaching data using the key period selection unit 14 (shown in FIG. 8) in the modified example of the first embodiment.

[0047] Therefore, as shown in Fig. 11, the key period determination unit 12 in this embodiment is configured to use a mathematical model 122 such as a neural network to output, for a plurality of sensor inputs, a probability 1220 that each time point in the operation data 100 falls within the key period to the key period selection unit 14 (shown in Fig. 8). Note that the mathematical model 122 for calculating the probability 1220 is not limited to a neural network.

[0048] (summary) The robot teaching device 1 in this embodiment is equipped with a key period selection unit 14 that allows the worker 20 or other user to select whether or not to reproduce each point in the period during which the operation data 100 was acquired in the teaching data 130, and the key period determination unit 12 calculates the probability that each point in the period during which the operation data 100 was acquired falls within the key period based on the measurement values ​​of multiple types of sensors 1110a to 1110g, 1111a to 1111e, and outputs the calculated probability to the key period selection unit 14. When allowing the worker 20 or other user to select whether or not to reproduce each point in the operation data 100 in the teaching data 130, the key period selection unit 14 displays the probability that each point in the period during which the operation data 100 was acquired falls within the key period.

[0049] According to the present embodiment configured as described above, by acquiring feature data 110 using multiple types of sensors, even if a large number of key periods are identified by the key period identification unit 12, it becomes easy for the instructor 20 or other users to select key periods to be reproduced in the teaching data 130 using the key period selection unit 14 (shown in FIG. 8).

[0050] Although the embodiments of the present invention have been described in detail above, the present invention is not limited to the above-described embodiments and includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to those including all of the described configurations. Furthermore, it is possible to add part of the configuration of one embodiment to the configuration of another embodiment, or to delete part of the configuration of one embodiment or replace it with part of another embodiment. [Explanation of symbols]

[0051] 1...robot teaching device, 10...motion acquisition unit, 11...feature acquisition unit, 12...key period determination unit, 13...teaching data generation unit, 14...key period selection unit, 20...worker (teacher), 21a, 21b...work object, 22...workbench, 23a, 23b...structure, 24...workbench coordinate system, 30a, 30b...robot, 100...motion data, 101...motion acquisition device, 110, 110A...feature data, 111, 111A...glove, 112...feature data time differential absolute value data, 120...key period data, 120a to 120d... Priority period, 120Aa to 120Ac...priority period, 121, 121A...threshold, 122...mathematical model, 130...teaching data, 1000...motion data time axis, 1001...motion data displacement axis, 1100, 1100A...feature data time axis, 1101, 1101A...feature data axis, 1110a to 1110g...pressure sensor, 1111a to 1111e...bending sensor, 1120...feature data time differential absolute value axis, 1210...motion acquisition unit, 1220...probability, 1300...teaching data time axis, 1301...teaching data displacement axis.

Claims

1. A robot teaching device that teaches a robot a series of operations that a worker performs on a work object, a motion acquisition unit that acquires motion data indicating a position and a posture of the worker or the work object; a feature amount acquiring unit that acquires feature amount data indicating the degree of influence that the worker's action has on the work object; a key period determination unit that determines, based on the feature amount data, a period during which the worker performs an action essential to the series of works, among the period during which the action data was acquired, as a key period; a teaching data generation unit that generates teaching data to be input to the robot based on the motion data so that the motion of the worker during the key period is reproduced and the motion of the worker outside the key period is not reproduced. A robot teaching device characterized by:

2. The robot teaching device according to claim 1, a glove to be worn by the worker; a pressure sensor attached to the glove, The feature amount acquisition unit acquires the measurement value of the pressure sensor as the feature amount data. A robot teaching device characterized by:

3. The robot teaching device according to claim 2, The key period determination unit determines, from among the periods during which the operation data is acquired, a period during which a time-differentiated absolute value of the measurement value of the pressure sensor is equal to or greater than a predetermined threshold, as the key period. A robot teaching device characterized by:

4. The robot teaching device according to claim 1, The teaching data generation unit generates the teaching data other than the key period so that the time axes of the elements of the operation data during the key period coincide in the teaching data, and so that the robot operates along the shortest path under the limitations of the range of motion, maximum allowable speed, and maximum allowable acceleration. A robot teaching device characterized by:

5. The robot teaching device according to claim 1, a key period selection unit that allows the worker or another user to select whether or not to reproduce the operation data of the period determined to be the key period by the key period determination unit in the teaching data; A robot teaching device characterized by:

6. The robot teaching device according to claim 1, a plurality of types of sensors for measuring different physical quantities as the feature amount data; The feature acquisition unit acquires measurement values ​​of the plurality of types of sensors as the feature data. A robot teaching device characterized by:

7. The robot teaching device according to claim 6, a key period selection unit that allows the operator or another user to select whether or not to reproduce each point in time during the period in which the operation data was acquired in the teaching data; the priority period determination unit calculates a probability that each time point in the period during which the operation data was acquired is included in the priority period based on the measurement values ​​of the plurality of types of sensors, and outputs the calculated probability to the priority period selection unit; The key period selection unit displays a probability that each time point in the period during which the operation data was acquired is included in the key period when the worker or the other user selects whether or not to reproduce each time point in the operation data in the teaching data. A robot teaching device characterized by:

8. A robot teaching method for teaching a robot a series of operations to be performed by a worker on a work object, comprising: acquiring motion data indicating the position and posture of the worker or the work object; acquiring feature data indicating the degree of influence of the worker's action on the work object; a step of determining, based on the feature amount data, a period during which the worker performs an action essential to the series of works as a key period within the period during which the action data was acquired; and generating teaching data to be input to the robot based on the motion data so that the motions of the worker during the key period are reproduced and the motions of the worker outside the key period are not reproduced. A robot teaching method comprising:

Images