JP6205890B2 - Work position recognition apparatus, work position recognition method, and robot system - Google Patents

Description

  The present invention relates to a technique for recognizing the position of a moving workpiece.

  As a technique for recognizing the position of a moving workpiece, there is known a technique for detecting a conveyance speed of a conveyance device by a sensor such as an encoder installed in a conveyance device that conveys the workpiece, and the detected conveyance speed is transmitted to a robot control device. Feedback is used to synchronize the robot movement with the transport speed. According to this technique, even if the conveyance speed of the conveyance device fluctuates, the robot can follow the moving workpiece.

JP 2000-263475 A

  However, because of the problem of conveyance accuracy in the conveyance device, the workpiece position also fluctuates in the height direction and in the lateral direction with respect to the conveyance direction. Therefore, the robot can accurately follow the workpiece only by detecting the conveyance speed. It is difficult. On the other hand, it is conceivable to add an additional sensor to detect the fluctuation of the workpiece position in the height direction and in the lateral direction with respect to the conveyance direction. There was a risk of increased costs.

  Accordingly, the present invention provides a workpiece position recognition device, a workpiece position recognition method, and a robot system capable of improving the followability of a robot with respect to a moving workpiece while suppressing the number of additional sensors for detecting workpiece position fluctuations. The purpose is to provide.

  The workpiece position recognition device includes a first surface that is installed at a second predetermined position on a pallet on which a workpiece is placed at a first predetermined position, is conveyed along with the workpiece, and is inclined up and down in a lateral direction with respect to the conveying direction. The second surface having a lower vertical slope than the first surface is arranged so as to appear alternately in the transport direction when viewed from above, and is installed above the scale movement trajectory. A sensor that detects a distance to the surface and outputs a signal according to the distance, and a fluctuation amount calculation unit that calculates a fluctuation amount of the workpiece position based on the signal output from the sensor are provided.

  In addition to the workpiece position recognition device, the robot system outputs a command to the robot that picks up the workpiece on the pallet and to perform a predetermined operation on the robot based on data stored in advance. And a robot controller, and the robot controller corrects the instruction signal output to the robot based on the variation amount of the workpiece position calculated by the variation amount calculation unit.

  It is possible to improve the followability of the robot with respect to the moving workpiece while suppressing the number of sensors for detecting the fluctuation of the workpiece position.

It is the schematic which shows an example of the robot system which has a workpiece | work position recognition apparatus. It is a side view which shows an example of a robot. The details of a scale are shown, (a) is a perspective view of a pallet, and (b) is an enlarged view of the Q section in (a). FIG. 4 shows another example of the scale, (a) is a perspective view of the first plate, (b) is a perspective view of the second plate, and (c) is a perspective view of the entire pallet on which the scale of another example is installed. The case of being conveyed in a steady state will be described. (A) is a top view of the scale, (b) is a pallet moving speed change diagram, (c) is a pallet height change diagram, and (d) is a displacement sensor. It is an output signal waveform diagram. A case where the conveyance speed fluctuates will be described. (A) is a plan view of a scale, (b) is a pallet moving speed change chart, (c) is a pallet height change chart, and (d) is an output signal of a displacement sensor. It is a waveform diagram. The case where the pallet fluctuates in the horizontal direction will be described. (A) is a plan view of the scale, (b) is a pallet moving speed change diagram, (c) is a pallet height change diagram, and (d) is a displacement sensor. It is a wave form diagram which shows an output signal. The case where the pallet fluctuates in the height direction will be described. (A) is a plan view of the scale, (b) is a pallet moving speed change diagram, (c) is a pallet height change diagram, and (d) is a displacement sensor. FIG. It is a flowchart which shows the calculation process which calculates the moving speed of a workpiece | work. It is a wave form diagram which shows an example of the output signal of a displacement sensor. It is a flowchart which shows the arithmetic processing which calculates a workpiece | work horizontal direction variation | change_quantity. It is a wave form diagram which shows an example of the output signal of a displacement sensor. It is a flowchart which shows the arithmetic processing which calculates workpiece | work height direction variation | change_quantity. It is a wave form diagram which shows an example of the output signal of a displacement sensor. It is a flowchart which shows the correction | amendment process which correct | amends the instruction | indication signal to a robot. The reference position of the scale is shown, (a) is a plan view of the pallet, and (b) is a front view of the scale. It is explanatory drawing explaining the content of the correction process in a robot system. The structure which concerns on 2nd Embodiment is demonstrated, (a) is a top view of a pallet, (b) is a left view of a pallet, (c) is a right view of a pallet. It is a top view of the pallet explaining the composition concerning a 3rd embodiment. It is explanatory drawing explaining the structure which detects the attitude | position fluctuation | variation of a pallet.

Hereinafter, a first embodiment for carrying out the present invention will be described in detail with reference to the accompanying drawings.
FIG. 1 schematically shows an example of a robot system.

  The robot system 10 includes a robot 12 that performs assembly work for handling parts not shown and assembling the workpiece 100, a robot controller 14 that controls the operation of the robot 12, and a workpiece position recognition that recognizes the position of the workpiece 100 as a variation amount. Device 16.

  A workpiece transfer pallet (hereinafter simply referred to as “pallet”) 102 on which the workpiece 100 is placed at a first predetermined position is set in a fixed direction (at a set speed) by a transfer device 104 such as a belt conveyor or a roller conveyor. The workpiece 100 is sequentially conveyed in the direction of the white arrow in the drawing (that is, the X direction), whereby the workpiece 100 moves to the movable range of the robot 12.

  The robot 12 is, for example, a vertical articulated robot having six axes, and includes a robot hand 18 for handling parts to be assembled to the workpiece 100 placed on the pallet 102 as a work tool.

  As an example of the robot 12, FIG. 2 shows a base 12b that is fixedly fixed to the floor surface 12a and a rotation axis H1 that is orthogonal to the floor surface 12a via a first joint element 12c. A base end arm 12d connected to the base 12b, and a first intermediate arm 12f connected to the base end arm 12d so as to be rotatable around a rotation axis H2 orthogonal to the rotation axis H1 via a second joint element 12e; The second intermediate arm 12h connected to the first intermediate arm 12f rotatably around the rotation axis H3 orthogonal to the rotation axis H2 via the third joint element 12g, and the fourth joint element 12i, A third intermediate arm 12j connected to the second intermediate arm 12h so as to be rotatable around a rotation axis H4 orthogonal to the rotation axis H3, and a rotation axis H5 parallel to the rotation axis H4 via a fifth joint element 12k. Around A fourth intermediate arm 12l that is rotatably connected to the third intermediate arm 12j and a fourth intermediate arm 12l that is rotatable around a rotation axis H6 orthogonal to the rotation axis H5 via a sixth joint element 12m. The tip arm 12n is shown. For example, the robot hand 18 is detachably attached to the tip of the tip arm 12n.

  The robot 12 has a rotation axis H1 between the base 12b and the proximal arm 12d and a rotation axis H2 between the proximal arm 12d and the first intermediate arm 12f, and the first intermediate arm 12f and the first intermediate arm 12f. Between the second intermediate arm 12h and the third intermediate arm 12l, between the second intermediate arm 12h and the third intermediate arm 12l, between the second intermediate arm 12h and the third intermediate arm 12j. Six servo motors (not shown) are provided around the rotation axis H5 to generate relative rotation around the rotation axis H6 between the fourth intermediate arm 121 and the tip arm 12n. Each servo motor has an encoder (not shown) that detects its rotation angle.

  The robot controller 14 outputs an instruction signal to perform a predetermined operation on the robot 12 (servo motor) based on data stored in advance. Specifically, this instruction signal instructs the servo motor to rotate to a predetermined rotation angle. An angle signal indicating the rotation angle of the servo motor is fed back from the encoder to the robot controller 14 and controlled to rotate to a predetermined rotation angle.

  In the robot controller 14, data stored in advance for the robot 12 to perform a predetermined operation is created, for example, by robot teaching. In robot teaching, the position and orientation of the robot hand 18 when the robot 12 is caused to perform a predetermined operation on the stationary workpiece 100 are pointed as six parameters (X, Y, X, Rx, Ry, Rz). Each is recorded sequentially. Here, X, Y, and Z mean spatial coordinates in a three-dimensional orthogonal coordinate system, and Rx, Ry, and Rz mean rotation angles around the X, Y, and Z axes, respectively. It shall be. The position / orientation of the robot hand 18 thus recorded as the six parameters is built into the robot controller 14 or the like as data (hereinafter referred to as “teaching data”) for reproducing a predetermined operation of the robot 12. Stored in the storage means. When the movement path between the positions of the robot hand 18 specified by the two teaching data is, for example, a straight line, the movement of the robot 12 between the positions specified by the two teaching data is performed using linear interpolation or the like. Stipulate. If the movement route is not a straight line, curve interpolation, circular interpolation, or the like may be used.

As shown in FIG. 1, the workpiece position recognition device 16 includes a scale 20, a displacement sensor 22 (first sensor), and a variation calculation unit 24.
The scale 20 is combined with the displacement sensor 22 so that the amount of change in the conveyance direction (X direction in the drawing) of the workpiece 100 position, the amount of change in the height direction (Z direction in the drawing) of the workpiece 100 position, and the position of the workpiece 100 Is used to detect the amount of fluctuation in the horizontal direction (Y direction in the figure) with respect to the transport direction.

  As shown in FIG. 3A, the scale 20 is a plate-like body that extends from one end 20a to the other end 20b in the conveying direction of the pallet 102 (in the direction of the white arrow in the drawing, that is, the X direction). On the pallet 102 to be placed, the pallet 102 tilts up and down in the lateral direction (Y direction in the figure) (inclined at an angle θ with respect to the upper surface 102a of the pallet 102), and does not interfere with the placement of the workpiece 100 2 is installed at a predetermined position. In the scale 20, as shown in FIG. 3B, through slits 20c of a predetermined gap A that penetrates the plate-like body in the thickness direction and extends in the lateral direction are repeatedly formed at a predetermined interval D in the transport direction. Has been. The width of the through slit 20c, that is, the lateral dimension W of the through slit 20c, even when the pallet 102 transported by the transport device 104 changes most in the lateral direction, the laser light emitted from the displacement sensor 22 described later passes through. It is set so that it can pass through the slit 20c. Assuming that the width B of the main body of the scale 20 between the adjacent through slits 20c is equal to the predetermined gap A, the upper surface 20d of the scale 20 that is inclined up and down in the horizontal direction with respect to the transport direction, and the upper and lower inclinations than the upper surface 20d of the scale 20 When viewed from above, the upper surface 102a of the pallet 102 having a small size appears alternately at a predetermined interval A in the transport direction. Hereinafter, in order to facilitate understanding, the width B of the scale 20 main body between the adjacent through slits 20c is assumed to be the same dimension as the predetermined gap A (A = B).

  The scale 20 is not limited to the plate-like body as described above, and is installed at a second predetermined position on the pallet 102 and has a first surface that is inclined up and down in the lateral direction with respect to the transport direction, and the first surface. The second surface having a smaller inclination than the surface may be formed so as to appear alternately in the transport direction when viewed from above. For example, as shown in FIG. 4A, one of the trapezoidal legs is an inclined leg 26c inclined with respect to a perpendicular perpendicular to the bases 26a and 26b, and the other is orthogonal to the upper base 26a and the lower base 26b. A right trapezoidal first plate 26 (thickness = A) which is a vertical leg 26d, and a perpendicular leg which is orthogonal to the bases 28a and 28b rather than the inclined leg 26d of the first plate 26, as shown in FIG. 4B. As shown in FIG. 4 (c), the second plate 28 (thickness = A) having the right trapezoidal shape having the inclined legs 28d inclined close to 28c, and the bases 26a, 26b, 28a and 28b are all attached to the pallet 102 as shown in FIG. Even if the comb blade-shaped scale 30 formed by alternately stacking in the transport direction (X direction in the figure) is placed at the second predetermined position of the pallet 102 in a state of being perpendicular to the upper surface 102a of the pallet. Good. Further, the comb blade-like scale 30 in which the first plates 26 and the second plates 28 are alternately laminated may be formed by integral molding from the beginning. If one of the first plate 26 and the second plate 28 can be installed on the pallet 102 at predetermined intervals, the surface facing the upper side of the first plate 26 or the second plate 28 corresponds to the first surface, and the two first plates Since the upper surface 102a of the pallet 102 between 26 or between the two second plates 28 corresponds to the second surface, the other plate may be omitted.

  1 and 3 again, the displacement sensor 22 is installed above the trajectory of the scale 20 where the pallet 102 is transported and moved by the transport device 104, more specifically, above the trajectory of the through slit 20c. The distance to the upper surface 20d of the scale 20 or the upper surface 102a of the pallet 102 is detected, and a signal is output to the variation calculation unit 24 according to the distance.

  The displacement sensor 22 is, for example, a triangular distance measuring device configured by combining a light emitting element using a semiconductor laser or the like and a light receiving element using a CCD (Charge-coupled device) or a CMOS (Complementary metal-oxide-semiconductor). This is a laser displacement sensor of the type. When a laser-type displacement sensor is used as the displacement sensor 22, the semiconductor laser light is emitted toward the movement locus of the scale 20. The irradiated light is diffusely reflected on the upper surface 20d of the scale 20 or the upper surface 102a of the pallet 102 through the through slit 20c depending on the position of the scale 20 in the transport direction, and a part of the reflected light is a light receiving element. Although the focal point is focused on, the focal point position moves according to the distance from the light emitting element to the diffusely reflecting surface. Therefore, by detecting the focal point point, the upper surface 20d of the scale 20 or the upper surface 102a of the pallet 102 is detected. Detect distance. The displacement sensor 22 is not limited to the laser type, and may be a non-contact type such as an ultrasonic type, an LED type, or an eddy current type.

  Although the displacement sensor 22 is not limited, for example, when the distance to the measurement object is shorter, that is, when the distance to the upper surface 20d of the scale 20 is, a signal having a higher voltage level is output and the measurement object is output. It is configured to output a signal having a lower voltage level when the distance to the object is longer, that is, when the distance to the upper surface 102a of the pallet 102 is longer.

  A fluctuation amount calculation unit 24 having a built-in computer calculates the fluctuation amount of the workpiece 100 position based on the signal output from the displacement sensor 22 and recognizes the workpiece 100 position from the calculated fluctuation amount of the workpiece 100 position. The variation amount of the workpiece 100 position is output to the robot controller 14 that is communicably connected to the variation amount calculation unit 24.

Next, the profile of the signal output from the displacement sensor 22 will be described according to the variation mode of the pallet 102 with reference to FIGS.
FIG. 5 shows an example of a signal profile output from the displacement sensor 22 when the transport device 104 is transporting the pallet 102 in a steady state.

  Here, as shown in FIG. 5A, the steady state is that the locus (black solid line) of the laser light emitted from the displacement sensor 22 toward the scale 20 is parallel to the center line of the scale 20 or to this. 5 along a straight line, that is, there is no change in the position of the pallet 102 in the lateral direction, and the moving speed of the pallet 102 is constant and does not change as shown in FIG. As shown in FIG. 5C, the height position is also constant and does not fluctuate.

  In such a steady state, the laser beam emitted from the displacement sensor 22 starts diffuse reflection from the one end 20a to the upper surface 20d of the scale 20, and then diffusely reflects on the upper surface 102a of the pallet 102 through the through slit 20c. The laser beam repeats diffuse reflection alternately on the upper surface 20d of the scale 20 and the upper surface 102a of the pallet 102 every predetermined time up to the other end 20b. Further, the distance from the displacement sensor 22 to the upper surface 20d of the scale 20 and the distance from the displacement sensor 22 to the upper surface 102a of the pallet 102 are constant because the pallet 102 does not vary in the height direction and the lateral direction. Accordingly, the signal output from the displacement sensor 22 is, as shown in FIG. 5D, a constant first signal level L1 according to the distance to the upper surface 20d of the scale 20, and the upper surface 102a of the pallet 102 through the slit. According to the distance, a constant second signal level L2 and a waveform appearing alternately every predetermined time.

FIG. 6 shows an example of a signal profile output from the displacement sensor 22 when the moving speed of the pallet 102 conveyed by the conveying device 104 varies.
As shown in FIGS. 6A and 6C, the locus of the laser beam irradiated from the displacement sensor 22 toward the scale 20 (black solid line) and the height position of the pallet 102 are the same as in the steady state. When the moving speed of the pallet 102 gradually decreases as shown in FIG. 6B without changing, the signal output from the displacement sensor 22 has the waveform shown in FIG.

  That is, the laser light emitted from the displacement sensor 22 starts diffuse reflection from one end 20a to the upper surface 20d of the scale 20, then passes through the through slit 20c and diffusely reflects on the upper surface 102a of the pallet 102, and the other end. Up to 20 b, diffuse reflection is alternately repeated on the upper surface 20 d of the scale 20 and the upper surface 102 a of the pallet 102. Further, the distance from the displacement sensor 22 to the upper surface 20d of the scale 20 and the distance from the displacement sensor 22 to the upper surface 102a of the pallet 102 are constant because the pallet 102 does not vary in the height direction and the lateral direction. However, since the moving speed of the pallet 102 is gradually decreased, the time during which the laser beam is applied to the upper surface 20d of the scale 20 or the upper surface 102a of the pallet 102 gradually increases. Therefore, as shown in FIG. 6D, the signal output from the displacement sensor 22 has a constant first signal level L1 and a constant second signal level L2 alternately. The waveform gradually increases in duration.

FIG. 7 shows an example of a signal profile output from the displacement sensor 22 when the pallet 102 conveyed by the conveying device 104 fluctuates in the lateral direction.
As shown in FIGS. 7B and 7C, the moving speed of the pallet 102 and the height position of the pallet 102 do not vary as in the steady state, and the laser beam irradiated from the displacement sensor 22 to the scale 20 7 (black thick solid line) is a waveform that gradually moves from the higher side of the scale 20 to the lower side, the signal output from the displacement sensor 22 has the waveform shown in FIG.

  That is, the laser light emitted from the displacement sensor 22 starts diffuse reflection from the one end 20a to the upper surface 20d of the scale 20, then passes through the through slit 20c and is diffusely reflected from the upper surface 102a of the pallet 102 to the other end 20b. The upper surface 20d of the scale 20 and the upper surface 102a of the pallet 102 repeat the diffuse reflection alternately every predetermined time. Further, the distance from the displacement sensor 22 to the upper surface 102a of the pallet 102 does not change and is constant. However, the distance from the displacement sensor 22 to the upper surface 20d of the scale 20 gradually decreases. Therefore, as shown in FIG. 7 (d), the first signal level L1 where the voltage gradually decreases according to the distance to the upper surface 20d of the scale 20, and constant according to the distance to the upper surface 102a of the pallet 102. The second signal level L2 is a waveform that appears alternately every predetermined time.

FIG. 8 shows an example of a signal profile output from the displacement sensor 22 when the pallet 102 conveyed by the conveying device 104 fluctuates in the height direction.
As shown in FIGS. 8A and 8B, the trajectory of the laser light irradiated from the displacement sensor 22 toward the scale 20 (black solid line) and the movement speed of the pallet 102 fluctuate in the same manner as in the steady state. If the height of the pallet 102 is gradually reduced, the signal output from the displacement sensor 22 has the waveform shown in FIG.

  That is, the laser light emitted from the displacement sensor 22 starts diffuse reflection from the one end 20a to the upper surface 20d of the scale 20, then passes through the through slit 20c and is diffusely reflected from the upper surface 102a of the pallet 102 to the other end 20b. The upper surface 20d of the scale 20 and the upper surface 102a of the pallet 102 repeat the diffuse reflection alternately every predetermined time.

  However, the distance from the displacement sensor 22 to the upper surface 20d of the scale 20 and the distance from the displacement sensor 22 to the upper surface 102a of the pallet 102 are gradually increased. Therefore, as shown in FIG. 8D, the first signal level L1 where the voltage gradually decreases according to the distance to the upper surface 20d of the scale 20, and the voltage according to the distance to the upper surface 102a of the pallet 102. The second signal level L2 that gradually decreases is a waveform that appears alternately every predetermined time.

  The exemplary profile of each variation mode of the signal output from the displacement sensor 22 is as described above. However, the detection resolution and the data sampling period of each variation mode of the displacement sensor 22 are determined according to the performance specifications of the displacement sensor 22. When the sampling period is 0.33 ms and the repeatability is 1 μm, the conveying speed of the conveying device 104 is 20 mm per second, the predetermined gap A of the through slit 20 c is 0.1 mm, and the scale 20 is inclined. If the angle θ is 30 °, the result is as follows. That is, the detection resolution is 6.6 μm for the variation in the conveyance direction, 1.7 μm for the variation in the horizontal direction, and 1 μm for the variation in the height direction, and the data sampling period is 5 ms regardless of the variation mode. The predetermined gap A and the inclination angle θ of the scale 20 may be set based on a desired data sampling period, a desired detection resolution, and a performance specification of the displacement sensor 22. When the detection resolution is improved, the sampling period of the displacement sensor 22 is shortened for fluctuations in the conveyance direction, the inclination angle θ of the scale 20 is increased for fluctuations in the horizontal direction, and fluctuations in the height direction are assumed. The accuracy of the displacement sensor 22 may be increased.

  Next, referring to FIG. 9 to FIG. 16, the position of the workpiece 100 that is repeatedly executed in the fluctuation amount calculation unit 24 from when the displacement sensor 22 detects the one end 20 a of the scale 20 until the other end 20 b is detected. An arithmetic process for detecting the amount of fluctuation will be described. Note that the detection of the one end 20a of the scale 20 is possible depending on whether or not the signal output from the displacement sensor 22 has changed to the first signal level L1 after maintaining the state of the second signal level L2 for a predetermined time. . Further, the detection of the other end 20b of the scale 20 is performed by, for example, determining whether or not a predetermined time has elapsed since the signal output from the displacement sensor 22 has changed from the first signal level L1 to the second signal level L2. Is possible.

FIG. 9 shows the content of the calculation process for calculating the moving speed of the workpiece 100 in the conveyance direction.
In step 1001 (abbreviated as “S1001” in the figure. The same applies hereinafter), as shown in FIG. 10, the second signal level corresponding to the distance to the upper surface 102a of the pallet 102 is obtained for the signal waveform output from the displacement sensor 22. The rising edge (upward thick arrow) at the time of changing to the first signal level L1 corresponding to the distance from L2 to the upper surface 20d of the scale 20 is detected, and the time tu at this time is determined until at least step 1003 is executed. The data is stored in a storage device such as a RAM (Random Access Memory) provided in the fluctuation amount calculation unit 24.

  In step 1002, a falling edge (downward thick arrow) when changing from the first signal level L1 to the second signal level L2 is detected, and the time td at this time is stored in the RAM until at least step 1003 is executed. To remember.

In step 1003, a time Δt _ud = td−tu from the time tu stored in step 1001 to the time td stored in step 1002 is calculated.
In step 1004, the moving speed of the scale 20, that is, the moving speed S of the workpiece 100 is calculated. Specifically, the width B of the main body of the scale 20 between the adjacent through slits 20c is divided by the time Δt _ud . In particular, when the width B is equal to the predetermined gap A of the through slit 20c, the moving speed S of the workpiece 100 is S = A / Δt _ud . In short, the moving speed S is calculated based on the pulse interval of the pulse signal waveform output from the displacement sensor 22. The calculated movement speed S of the workpiece 100 is stored in a RAM or the like until at least the next movement speed S is calculated.

  The calculated movement speed S is an average speed in a minute time, and by multiplying the movement speed S by the minute time, the movement distance D in the transport direction of the workpiece 100 for each minute time can be obtained. The calculation corresponds to the calculation of the amount of change in the transport direction of the workpiece 100 position.

When the rising edge time stored when step 1001 is executed again to make the movement of the robot 12 follow the movement speed of the workpiece 100 more accurately, the time between steps 1001 and 1002 is assumed. , The time Δt _du = tu′−td may be calculated, and a step of calculating the moving speed by dividing the time Δt _du by the predetermined gap A of the through slit 20c may be added.

In addition, in this way, the processing for calculating the time Δt _du from the time tu at the rising edge to the time td at the falling edge is performed as the main calculation for calculating the time Δt _ud from the time td at the falling edge to the time tu at the rising edge. It may be executed in parallel with the processing.

FIG. 11 shows the content of the calculation process for calculating the amount of variation in the horizontal direction of the workpiece 100 position.
In step 2001, as shown in FIG. 12, the peak value V1 of the signal waveform output from the displacement sensor 22 is measured and stored in the aforementioned RAM or the like.

  In step 2002, it is determined whether or not a rising edge (upward thick arrow) is detected. If the rising edge is detected, the process proceeds to step 2003 (Yes). On the other hand, if the rising edge is not detected, the laser beam diffuses and reflects on the upper surface 102a of the pallet 102 through the through slit 20c. In order to detect the distance to the upper surface 102a of the pallet 102 just before the diffused reflection from the upper surface 20d of the scale 20 is detected, the process returns to step 2001 (No).

In step 2003, the peak value V2 is measured and stored in a RAM or the like.
In step 2004, the first difference Vd1 (= V2−V1) is calculated from the latest peak value V1 stored in the RAM or the like and the peak value V2 stored in step 2003.

In step 2005, the peak value V3 is measured and stored in a RAM or the like.
In step 2006, it is determined whether a falling edge (downward thick arrow) is detected. When the falling edge is detected, the process proceeds to step 2007 (Yes). On the other hand, when the falling edge is not detected, the laser light is diffusely reflected from the upper surface 20d of the scale 20; In order to detect the distance to the upper surface 20d of the scale 20 just before passing through the through slit 20c and diffusely reflecting on the upper surface 102a of the pallet 102, the process returns to step 2005 (No).

In step 2007, the peak value V4 is measured and stored in a RAM or the like.
In step 2008, the second difference Vd2 (= V3−V4) is calculated from the latest peak value V3 stored in the RAM or the like and the peak value V4 stored in step 2007.

  In step 2009, a first lateral variation ΔY in the lateral direction of the pallet 102 immediately below the displacement sensor 22 is calculated. The first lateral variation ΔY is a subtraction value (Vd2−Vd1) obtained by subtracting the first difference Vd1 from the second difference Vd2, and an inclination angle θ at which the scale 20 is inclined with respect to the upper surface 102a of the pallet 102. The tangent angle is calculated by dividing by tan θ. That is, the first lateral fluctuation amount ΔY in the lateral direction of the pallet 102 immediately below the displacement sensor 22 is expressed by the equation: ΔY = (Vd2−Vd1) / tan θ. The calculated first lateral variation ΔY is stored in the RAM or the like until at least the next first lateral variation ΔY is calculated.

  In short, the amount of variation in the lateral direction of the workpiece 100 position is regarded as the first lateral variation ΔY in the lateral direction of the pallet 102 immediately below the displacement sensor 22, and the inclination angle θ of the scale 20 and the pallet This is calculated based on the amount of change in the distance from the upper surface 102a of the 102 to the upper surface 20d of the scale 20.

FIG. 13 shows the content of the calculation process for calculating the amount of variation in the height direction of the workpiece 100 position.
In step 3001, as shown in FIG. 14, it is determined whether or not a falling edge (downward thick arrow) is detected in the signal waveform output from the displacement sensor 22. If it is determined that a falling edge has been detected, the process proceeds to step 3002 (Yes). On the other hand, if it is determined that a falling edge has not been detected, step 3001 is executed again (No).

In step 3002, the peak value V5 is measured and stored in a RAM or the like.
In this way, the peak value of the second signal level L2 is measured after the falling edge is detected, and the peak value is not measured at the first signal level L1, as described above, the pallet 102 fluctuates in the horizontal direction. In this case, since the first signal level L1 changes, even if the peak value of the first signal level L1 is measured, the fluctuation in the height direction of the pallet 102 and the fluctuation in the horizontal direction of the pallet 102 can be distinguished. Because it is difficult.

In step 3003, the peak value V6 is measured and stored in a RAM or the like.
The time from the measurement in step 3002 to the measurement in step 3003 may be managed by a timer as long as the peak value V6 can be measured before the next rise of the signal. It may depend on the data sampling period.

  In step 3004, a first height direction variation ΔZ in the height direction of the pallet 102 immediately below the displacement sensor 22 is calculated. The first height direction variation ΔZ is calculated as a subtraction value (V6−V5) obtained by subtracting the peak value V5 from the peak value V6. In short, the amount of variation in the height direction of the workpiece 100 position is regarded as the first height direction variation amount ΔZ in the height direction of the pallet 102 immediately below the displacement sensor 22, and the displacement sensor 22 to the pallet 102. Is calculated based on the amount of change in the distance to the upper surface 102a. The calculated first height direction variation ΔZ is stored in the RAM or the like until at least the next first height direction variation ΔZ is calculated.

  Next, FIG. 15 shows an instruction signal output to the robot 12 (servo motor) until the displacement sensor 22 detects one end 20a of the scale 20 until the other end 20b is detected in the robot controller 14. An example of the correction process which correct | amends based on the variation | change_quantity of a position is shown.

  In step 4001, the position of the displacement sensor 22 in the X direction when the displacement sensor 22 detects one end 20a of the scale 20 is set to 0, the distance Y0 from the displacement sensor 22 to the upper surface 20d of the scale 20, and the displacement sensor 22 to the pallet. The distance Z0 to the upper surface 102a of 102 is detected, and an initial deviation amount indicating how much deviation has occurred with respect to the reference position of the scale 20 is calculated.

  Here, as shown in FIG. 16, the reference position of the scale 20 is a distance Xs in the X direction from one end 20a of the scale 20 to the displacement sensor 22 when the robot teaching described above is performed (FIG. 16A). Reference), a distance Ys from the displacement sensor 22 to the upper surface 20d of the scale 20 (see FIG. 16B), and a distance Zs from the displacement sensor 22 to the upper surface 102a of the pallet 102 (see FIG. 16B). The reference position (Xs, Ys, Zs) is stored in advance in the robot controller 14.

  Accordingly, assuming that the initial deviation amount in the X direction is ΔX0, the initial deviation amount in the Y direction is ΔY0, and the initial deviation amount in the Z direction is ΔZ0, the initial deviation amounts are ΔX0 = 0−Xs, ΔY0 = Y0−Ys, And ΔZ0 = Z0−Zs.

In step 4002, the teaching data that defines the operation of the robot 12 for each point in the teaching data described above is corrected by the initial shift amount calculated in step 4001, and static correction teaching data P _i (i : Integer of 0 to n).

  In step 4003, as shown in FIG. 17, the latest fluctuation amount stored in the fluctuation amount calculation unit 24, that is, the movement speed S that indirectly represents the fluctuation amount in the transport direction, and the horizontal direction at every predetermined time Δt. Fluctuation amount ΔY and height fluctuation amount ΔZ are read.

In step 4004, by correcting the interpolated data IPD1 between two static correction which is corrected teaching data _{P 0} and _{P 1} in step 4002, obtaining the dynamic correction interpolation data CIPD1.

Here, the interpolation data IPD1 between corrected two static correction teaching data _{P 0} and _{P 1} in step 4002, the robot hand 18, the second static from the position of the first static correction teaching data _{P 0} correction to the position of the teaching data _{P 1,} for example, when moving linearly, a parameter representing the position of the six parameters representing the position and orientation of the robot hand 18 (X, Y, Z, Rx, Ry, Rz) (X, Y, Z) is a parameter function that is a linear function with respect to time t between the positions specified by the two static correction teaching data P ₀ and P ₁ , X ₁ (t), Y ₁ ( t) and Z ₁ (t).

The parameter function, X ₁ (t), Y ₁ (t), and Z ₁ (t) are corrected parameter function, X ₁ c (t), Y ₁ c (t), and dynamic correction interpolation data cIPD1, respectively. It is corrected to Z ₁ c (t). For example, X ₁ c (t) = X ₁ (t) + S × Δt, Y ₁ c (t) = Y ₁ (t) + ΔY, and Z ₁ c (t) = Z ₁ (t) + ΔZ. It is corrected.

Then, at each time t _m , X ₁ c (t _m ), Y ₁ c (t _m ), and Z ₁ c (t _m ) are calculated, and the calculated X ₁ c (t _m ), Y ₁ c Based on (t _m ) and Z ₁ c (t _m ), the rotation angles ANGLE1 (J1, J2, J3, J4, J5, J6) of the six servo motors of the robot 12 are calculated by a known method. However, m is m = 0, (1 / n) × (T /, where T (= n × Δt) is the movement time between the positions specified by the two static correction teaching data P ₀ and P _1. Δt), (2 / n) × (T / Δt),..., (K / n) × (T / Δt),..., T / Δt (n is a natural number, k is an integer of 0 or more). The rotation angle ANGLE1 (J1, J2, J3, J4, J5, J6) is output to the robot 12 as an instruction signal.

The above-described correction of the interpolation data is sequentially performed on the interpolation data between the two static correction teaching data P _i-1 and P _i (i = 2,..., N). Twelve six servomotor rotation angles ANGLEi (J1, J2, J3, J4, J5, J6) are calculated.

  According to the workpiece position recognition device 16 and the robot 12 system 10 including the workpiece position recognition device 16, the variation amount of the workpiece 100 position in the conveyance direction, the lateral direction, and the height direction is calculated from the distance signal detected by one displacement sensor 22. The instruction signal for the robot 12 is corrected based on the calculated fluctuation amount. Therefore, the followability of the robot 12 with respect to the moving workpiece 100 can be improved without increasing the number of sensors for detecting the change in the workpiece 100 position. As a result, when the robot 12 performs assembly work as in this embodiment, assembly errors when assembling parts to the workpiece 100 are reduced, and the assembly work line including the robot system 10 and the transfer device 104 is reduced. It is possible to suppress the stop.

Next, a second embodiment for carrying out the present invention will be described. In addition, about the same structure as 1st Embodiment, the description is abbreviate | omitted or simplified by attaching | subjecting the same code | symbol.
FIG. 18 illustrates a state in which the pallet 102 that is transported by the transport device 104 is rotated at an angle α from the transport direction to the lateral direction on the transport device 104.

  As shown in FIG. 18A, the workpiece position recognition device 16 is located above the trajectory of the scale 20 and, more specifically, above the trajectory of the through slit 20c, than the length of the scale 20 in the transport direction. It further has an additional displacement sensor 32 (second sensor) installed at the same height while being separated from the displacement sensor 22 by a short first predetermined interval F1. The additional displacement sensor 32 detects the distance to the upper surface 20d of the scale 20 and the upper surface 102a of the pallet 102, and outputs a signal corresponding to the distance to the fluctuation amount calculation unit 24.

  As shown in FIG. 18B, the fluctuation amount calculation unit 24 sets the distance HL from the upper surface 102a of the pallet 102 to the upper surface 20d of the scale 20 as step 2001 to step 2004 or step 2005 to step 2008 in FIG. Similarly, calculation is performed based on a signal output from the displacement sensor 22. Then, based on the calculated distance HL and the inclination angle θ of the scale 20, the fluctuation amount calculation unit 24 calculates the Y from the intersection line between the upper surface of the inclined scale 20 and the upper surface of the pallet 102 to the displacement sensor 22. The direction distance YL is calculated. Specifically, the distance YL can be calculated from the equation YL = HL / tan θ.

  Further, as shown in FIG. 18 (c), the fluctuation amount calculation unit 24 determines the distance HR from the upper surface 102a of the pallet 102 to the upper surface 20d of the scale 20 from step 2001 to step 2004 or step 2005 to step 2005 in FIG. Similar to 2008, calculation is performed based on the signal output from the additional displacement sensor 32. Based on the calculated distance HR and the inclination angle θ of the scale 20, the fluctuation amount calculation unit 24 extends from the intersection line between the upper surface of the inclined scale 20 and the upper surface of the pallet 102 to the additional displacement sensor 32. The Y direction distance YR is calculated. Specifically, the distance YR can be calculated from the equation YR = HR / tan θ.

The fluctuation amount calculation unit 24 calculates the angle α in FIG. 18A based on the calculated distance YL and distance YR and the first predetermined interval F1 between the displacement sensor 22 and the displacement sensor 32. Specifically, the angle α can be calculated from the equation: α = tan ⁻¹ ((YL−YR) / F1). As a result, the rotation angle α of the pallet 102 around the Z-axis and hence the rotation angle α of the workpiece 100 can be detected. Therefore, in step 4003 of FIG. 15, the robot controller 14 sets the rotation angle α from the variation calculation unit 24 for a predetermined time. In step 4004, the interpolation data IPD1 is corrected to obtain dynamic correction interpolation data cIPD1. For example, the parameter function Rz (t) related to the posture of the robot hand 18 is corrected to the correction parameter function Rzc (t), so that Rzc (t) = Rz (t) + α.

  According to the workpiece position recognition device 16 of the second embodiment and the robot 12 system 10 including the workpiece position recognition device 16, by using the two sensors of the displacement sensor 22 and the displacement sensor 32, the conveyance direction and the lateral direction of the workpiece 100 position. Since not only the fluctuation in the height direction but also the rotation angle α of the workpiece 100 around the Z axis can be detected, the followability of the robot 12 to the moving workpiece 100 can be further improved with a small number of sensors.

Next, a third embodiment for carrying out the present invention will be described. In addition, about the same structure as 1st Embodiment and 2nd Embodiment, the description is abbreviate | omitted or simplified by attaching | subjecting the same code | symbol.
As shown in FIG. 19, in addition to the displacement sensor 22 and the displacement sensor 32 in the second embodiment, the workpiece position recognition device 16 is located above the pallet 102 at a second predetermined interval F2 in the Y direction with the displacement sensor 22. It further has a displacement sensor 34 (third sensor) that is spaced and installed at the same height. As shown in FIG. 20, the displacement sensor 34 detects the distance HHT to the upper surface 102 a of the pallet 102 and outputs a signal corresponding to the distance HHT to the fluctuation amount calculation unit 24. Further, as in the first and second embodiments, the displacement sensor 22 and the displacement sensor 32 detect the distance HHL and the distance HHR to the upper surface 102a of the pallet 102, respectively, Thus, signals corresponding to the distance HHL and the distance HHR are output.

Based on the signal output from the displacement sensor 22, the signal output from the displacement sensor 32, and the signal output from the displacement sensor 34, the fluctuation amount calculation unit 24 determines the inclination of the upper surface 102 a of the pallet 102 with respect to the horizontal plane as X The angle β for the direction and the angle γ for the Y direction are calculated. Specifically, it can be calculated from the following equations: β = tan ⁻¹ (HHR−HHL) / F1 and γ = tan ⁻¹ (HHT−HHL) / F2. Thereby, the posture variation amount of the pallet 102 with respect to the horizontal plane, and hence the posture variation amount of the workpiece 100 is calculated.

  In step 4003 of FIG. 15, the robot controller 14 reads the angle β and the angle γ from the variation calculation unit 24 every predetermined time Δt. In step 4004, the robot controller 14 corrects the interpolation data IPD1 to obtain the dynamic correction interpolation data cIPD1. Ask. For example, the parameter functions Ry (t) and Rz related to the posture of the robot hand 18 are corrected to the correction parameter function Rzc (t), and Rxc (t) = Rx (t) + γ, Ryc (t) = Ry (t) + β And

  According to the workpiece position recognition device 16 of the third embodiment and the robot system 10 including the workpiece position recognition device 16, the three positions of the displacement sensor 22, the displacement sensor 32, and the displacement sensor 34 are used to convey the position of the workpiece 100. It is possible to detect not only the variation in the direction, the lateral direction and the height direction and the rotation angle α of the workpiece 100 around the Z axis, but also the posture variation amount of the workpiece 100. Accordingly, in order to detect the fluctuation amount of the workpiece 100, conventionally, five to six sensors are required, but the followability of the robot 12 to the moving workpiece 100 can be further improved with a small number of sensors. it can.

Regarding the above embodiment, the following additional notes are disclosed.
(Supplementary Note 1) A first surface that is installed at a second predetermined position on a work conveyance pallet on which a work is placed at a first predetermined position, is conveyed together with the work, and is inclined up and down in a lateral direction with respect to the conveyance direction; A second surface having a lower vertical slope than the first surface is formed so that the second surface alternately appears in the transport direction when viewed from above, and the first surface is disposed above the locus along which the scale moves. A first sensor that detects a distance to a surface or the second surface and outputs a signal in accordance with the distance; and an output signal output from the first sensor to determine a variation amount of the workpiece position. A fluctuation amount calculation unit to calculate,
A workpiece position recognizing device.

(Supplementary note 2) The scale is a plate-like body extending in the transport direction, and is arranged so as to incline up and down in the lateral direction, and a through slit having a predetermined gap extending in the lateral direction is formed on the scale. Repeatedly formed in the direction at the predetermined interval,
The work position recognition apparatus according to claim 1, wherein the first surface is an upper surface of the scale, and the second surface is an upper surface of the work conveyance pallet.

(Additional remark 3) The said variation | change_quantity calculating unit is the said output signal of a said 1st sensor about the inclination angle of the said scale inclined up and down in the said horizontal direction, and the distance from the upper surface of the said workpiece conveyance pallet to the upper surface of the said scale. The workpiece position recognizing device according to appendix 2, wherein a variation amount in the lateral direction of the workpiece position is calculated on the basis of the amount of change obtained from the above.

(Additional remark 4) The said fluctuation amount calculating unit is the height of the said workpiece | work position based on the variation | change_quantity calculated | required from the said output signal of the said 1st sensor about the distance from the said 1st sensor to the upper surface of the said workpiece conveyance pallet. The workpiece position recognizing device according to appendix 3, wherein a variation amount in the vertical direction is calculated.

(Additional remark 5) The said variation | change_quantity calculation unit calculates the variation | change_quantity in the conveyance direction of the said workpiece | work position based on the pulse interval in the said output signal of a said 1st sensor. Position recognition device.

(Appendix 6) Above the trajectory where the scale moves, the scale is disposed at a predetermined interval shorter than the length of the scale in the transport direction and spaced apart from the first sensor, and the upper surface of the scale and the workpiece transport The sensor further includes a second sensor that detects a distance to the upper surface of the pallet and outputs a signal according to the distance, and the variation amount calculation unit includes the signal output from the first sensor, the second sensor, and the second sensor. 6. The workpiece according to claim 5, wherein a rotation angle of the workpiece from the transport direction to the lateral direction is calculated based on a signal output from the sensor, an inclination angle of the scale, and the predetermined interval. Position recognition device.

(Additional remark 7) The said fluctuation | variation amount calculating unit is calculated from the signal output from the said 1st sensor, The 1st distance to the upper surface of the said workpiece conveyance pallet, and the signal output from the said 2nd sensor The supplementary note 6 is characterized in that the rotation angle of the workpiece from the conveyance direction to the lateral direction is calculated based on a difference between the second distance to the upper surface of the workpiece conveyance pallet calculated from The workpiece position recognition device described.

(Additional remark 8) It is installed in the upper part of the said workpiece conveyance pallet, spaced apart from the said 1st sensor and the said 2nd sensor, The distance to the upper surface of the said workpiece conveyance pallet is detected, According to the said distance A third sensor that outputs a signal; and the fluctuation amount calculation unit outputs a signal output from the first sensor, a signal output from the second sensor, and an output from the third sensor. The workpiece position recognizing device according to appendix 6 or appendix 7, wherein the posture variation amount of the workpiece with respect to the horizontal is calculated based on the received signal.

(Supplementary Note 9) The fluctuation amount calculation unit calculates a first distance to the upper surface of the workpiece transfer pallet calculated from a signal output from the first sensor, and a signal output from the second sensor. Based on the difference between the second distance to the upper surface of the workpiece conveyance pallet calculated from the first, the first upper surface of the workpiece conveyance pallet in the direction from the first sensor toward the second sensor. An inclination is calculated, and a difference between the first distance or the second distance and a third distance to the upper surface of the workpiece conveyance pallet calculated from a signal output from the third sensor is calculated. Based on this, the second inclination of the upper surface of the work transport pallet in the direction from the third sensor toward the first sensor or the second sensor is calculated to change the posture of the work relative to the horizontal. Work position recognition device according to note 8, characterized by calculating the amount.

(Supplementary Note 10) A first surface that is installed at a second predetermined position on a work conveyance pallet on which a work is placed at a first predetermined position, is conveyed together with the work, and is inclined up and down in a transverse direction with respect to the conveyance direction; Up to the first surface or the second surface above the movement trajectory of the scale formed so that the second surface having a lower vertical slope than the first surface appears alternately in the transport direction when viewed from above. A sensor that detects the distance of and outputs a signal according to the distance is installed,
A workpiece position recognition method, wherein a fluctuation amount calculation unit connected to the sensor calculates a fluctuation amount of the workpiece based on a signal output from the sensor.

(Additional remark 11) The 1st surface which is installed in the 2nd predetermined position on the work conveyance pallet which mounts a work in the 1st predetermined position, is conveyed with the said work, and inclines up and down in the transverse direction to the conveyance direction, and the above-mentioned A scale that is formed so that the second surface having a lower vertical slope than the first surface appears alternately in the transport direction when viewed from above;
A sensor that is installed above the movement locus of the scale, detects a distance to the first surface or the second surface, and outputs a signal according to the distance;
Based on the signal output from the sensor, a fluctuation amount calculation unit that calculates the fluctuation amount of the workpiece position;
A robot for taking out the workpiece on the workpiece conveying pallet;
A robot controller that outputs an instruction signal to perform a predetermined operation on the robot based on data stored in advance;
Have
The robot system, wherein the robot controller corrects the instruction signal based on a variation amount of the work position calculated by the variation amount calculation unit.

DESCRIPTION OF SYMBOLS 10 Robot system 12 Robot 14 Robot controller 16 Work position recognition apparatus 20 Scale 20c Through slit 20d Upper surface 22 Displacement sensor 24 Fluctuation amount calculation unit 26 First plate 28 Second plate 30 Scale 32 Displacement sensor 34 Displacement sensor 100 Work 102 Work conveyance pallet 102a Upper surface 104 Conveying device

Claims (9)

From a first surface and a first surface that are installed at a second predetermined position on a work conveyance pallet on which a work is placed at a first predetermined position, are conveyed together with the workpiece, and are inclined up and down in a lateral direction with respect to the conveyance direction. And a scale formed such that the second surface with a small vertical inclination appears alternately in the transport direction when viewed from above,
A first sensor that is installed above a trajectory of the scale and detects a distance to the first surface or the second surface, and outputs a signal according to the distance;
A fluctuation amount calculation unit for calculating a fluctuation amount of the workpiece position based on the output signal output from the first sensor;
A workpiece position recognizing device. The scale is a plate-like body extending in the transport direction, and is arranged so as to incline up and down in the lateral direction. The scale has a through slit having a predetermined gap extending in the lateral direction in the transport direction . Are repeatedly formed at predetermined intervals ,
The workpiece position recognition apparatus according to claim 1, wherein the first surface is an upper surface of the scale, and the second surface is an upper surface of the workpiece conveyance pallet.   The fluctuation amount calculation unit is a change obtained from the output signal of the first sensor with respect to an inclination angle of the scale that is inclined up and down in the lateral direction and a distance from an upper surface of the work transfer pallet to the upper surface of the scale. The workpiece position recognition apparatus according to claim 2, wherein a fluctuation amount in the lateral direction of the workpiece position is calculated based on the amount.   The variation amount calculation unit varies the height of the workpiece position in the height direction based on a variation amount obtained from the output signal of the first sensor with respect to the distance from the first sensor to the upper surface of the workpiece conveyance pallet. The workpiece position recognition apparatus according to claim 3, wherein an amount is calculated.   The workpiece position recognition apparatus according to claim 4, wherein the fluctuation amount calculation unit calculates a fluctuation amount in the transport direction of the workpiece position based on a pulse interval in the output signal of the first sensor. . The upper surface of the scale and the workpiece transport pallet are disposed above the trajectory of the scale and spaced apart from the first sensor at a second predetermined interval shorter than the length of the scale in the transport direction. A second sensor that detects a distance to the upper surface of the sensor and outputs a signal according to the distance;
The fluctuation amount calculation unit is configured to move from the transport direction based on the signal output from the first sensor, the signal output from the second sensor, the inclination angle of the scale, and the second predetermined interval. 6. The workpiece position recognition apparatus according to claim 5, wherein a rotation angle of the workpiece in the lateral direction is calculated. The distance between the first sensor and the second sensor is set above the work transfer pallet, and the distance to the upper surface of the work transfer pallet is detected, and a signal is output according to the distance. A third sensor;
The fluctuation amount calculation unit is configured to determine the posture of the workpiece with respect to the horizontal based on a signal output from the first sensor, a signal output from the second sensor, and a signal output from the third sensor. The work position recognizing device according to claim 6, wherein a fluctuation amount is calculated. From a first surface and a first surface that are installed at a second predetermined position on a work conveyance pallet on which a work is placed at a first predetermined position, are conveyed together with the workpiece, and are inclined up and down in a lateral direction with respect to the conveyance direction. The distance to the first surface or the second surface is detected above the movement trajectory of the scale formed so that the second surface having a small vertical inclination appears alternately in the transport direction when viewed from above. Then, a sensor that outputs a signal according to the distance is installed,
A workpiece position recognition method, wherein a fluctuation amount calculation unit connected to the sensor calculates a fluctuation amount of the position of the workpiece based on a signal output from the sensor. From a first surface and a first surface that are installed at a second predetermined position on a work conveyance pallet on which a work is placed at a first predetermined position, are conveyed together with the workpiece, and are inclined up and down in a lateral direction with respect to the conveyance direction. And a scale formed such that the second surface with a small vertical inclination appears alternately in the transport direction when viewed from above,
A sensor that is installed above the movement locus of the scale, detects a distance to the first surface or the second surface, and outputs a signal according to the distance;
Based on the signal output from the sensor, a fluctuation amount calculation unit that calculates the fluctuation amount of the workpiece position ;
A robot for taking out the workpiece on the workpiece conveying pallet;
A robot controller that outputs an instruction signal to perform a predetermined operation on the robot based on data stored in advance;
Have
The robot system, wherein the robot controller corrects the instruction signal based on a fluctuation amount of the work position calculated by the fluctuation amount calculation unit.