Information processing apparatus, method for controlling information processing apparatus, and storage medium

Abstract

To provide an information processing device and the like capable of simulating robot control without exceeding a limit of a movable range of an operation in simulation of robot control of a multi-axis robot.SOLUTION: The information processing device includes a learning unit configured to generate a learning model learned by simulation in a virtual space, a storage unit configured to store robot configuration information, a reception unit configured to receive work space information for executing teaching and a work requirement set in advance for a workpiece, and a generation unit configured to generate a three dimensional model of a work space including the robot and the workpiece in the virtual space based on the robot configuration information and the work space information. And a simulation unit that outputs a result of simulating an operation of the robot satisfying the work requirements as control information of the movable range in the real space based on the learning model.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an information processing device, a control method for an information processing device, and a control program for an information processing device. [Background technology]

[0002] Traditionally, offline teaching of robots at production and manufacturing sites has been a time-consuming and labor-intensive process for operators. In response to this, in recent years, progress has been made in the development of information processing devices capable of simulating robot movements in 3D. By using such information processing devices, it is possible to reduce the time and effort required for offline teaching by using machine learning to simulate the robot's movements relative to the workpiece.

[0003] In a simulation of robot movement, soft limits are used to prevent the robot's movement from exceeding the limits of its range of motion. A soft limit is a control that stops the robot's control program if the robot moves beyond that limit, preventing it from being established.

[0004] As an example of a soft limit in the operation of a robot, Patent Document 1 discloses a technology in which, in a control program for controlling a machining center, when a command value for the operation of a control axis exceeds a soft limit value, which is the limiting value of the soft limit, the execution of the control program is stopped and the operation of the machining center is stopped.

[0005] Furthermore, when simulating the movements of a robot, a system capable of providing feedback to the robot simulator used in the robot movement simulation has been developed, and such systems have been provided by various robot manufacturers.

[0006] To control the soft limits within a simulation using a game development platform, it was proposed to use one of the assets of the game development platform, for example, Rotation Limit Hinge, which is a function of Final IK (inverse kinematics).

[0007] Final IK is an asset that provides the ability to freely move characters and other objects within Unity, and Rotation Limit Hinge provides a function to control the amount of rotation of the joints of characters and other objects, and is used to express the movements of characters and other objects more naturally. [Prior art documents] [Patent documents]

[0008] [Patent Document 1] Japanese Patent Application Publication No. 11-90770 Summary of the Invention [Problem to be solved by the invention]

[0009] In the robot control program in the above-described simulation, if the movable range of one of the axes of a multi-axis robot exceeds the soft limit value, which is a soft limit restriction, and further execution of the control program is stopped as in the above-described conventional technology, there is a risk that machine learning in the robot simulation using the control program will not progress.

[0010] The present invention has been made in consideration of the above-mentioned problems, and aims to provide an information processing device or the like that enables simulation of robot control of a multi-axis robot without exceeding the limits of the movable range of motion in the simulation of robot control. [Means for solving the problem]

[0011] The information processing device of the first aspect is characterized by comprising: a learning unit that generates a learning model learned by simulating in a virtual space using information about past work performed in real space by a multi-joint, multi-axis robot on a specified workpiece; a memory unit that stores robot configuration information including the shape of each arm that constitutes the robot, the connection relationships between the arms, and the movable range of each arm; a reception unit that receives workspace information including information about the shape of the workpiece to be taught and the workspace of the robot that processes the workpiece, as well as work requirements that have been set in advance for the workpiece; a generation unit that generates a three-dimensional model of the workspace including the robot and workpiece in the virtual space based on the robot configuration information and the workspace information; and a simulation unit that outputs the results of a simulation of operations performed by the robot in virtual space that satisfy the work requirements as control information for the movable range in real space based on the learning model.

[0012] In a second aspect, in the information processing device according to the first aspect, the learning may be reinforcement learning.

[0013] In a third aspect, in the information processing device according to the second aspect, if the cumulative rotation amount of each axis of the robot in the virtual space obtained from the accumulation of the network output for each reinforcement learning episode exceeds a range of a preset soft limit value, the cumulative rotation amount may be adjusted to be within the range of the soft limit value.

[0014] In a fourth aspect, in the information processing device according to the first aspect, the simulation unit may use a learning model to obtain control information that satisfies predetermined work requirements for the workpiece to be taught.

[0015] In a fifth aspect, in the information processing device according to the first aspect, the learning unit may generate a learning model by machine learning using at least the shape of a predetermined workpiece, work requirements previously set for the predetermined workpiece, and past work spaces as input data, and work performed by a robot on the predetermined workpiece as output data.

[0016] In a sixth aspect, in the information processing device according to the first aspect, the robot may be a painting robot that paints a workpiece.

[0017] As a seventh aspect, in the information processing device according to the first aspect, the information processing device may further include a setting unit that sets a connection relationship between the arms that configure the robot.

[0018] The control method of the information processing method according to the eighth aspect is characterized in that the information processing device executes the following steps: a learning step in which an information processing device generates a learning model that is learned by simulating in a virtual space using information about past work in real space by a multi-joint, multi-axis robot on a specified workpiece; a storage step in which robot configuration information including the shape of each arm that constitutes the robot, the connection relationships between the arms, and the movable range of each arm; a reception step in which workspace information including information about the shape of the workpiece to be taught and the workspace of the robot that processes the workpiece, as well as predetermined work requirements for the workpiece; a generation step in which an information processing device generates a three-dimensional model of the workspace including the robot and workpiece in the virtual space based on the robot configuration information and the workspace information; and a simulation step in which the results of simulating the robot's operations in the virtual space that satisfy the work requirements are output as control information for the movable range in real space based on the learning model.

[0019] The control program of the information processing device according to the ninth aspect is characterized in that the information processing device is provided with the following functions: a learning function that generates a learning model learned by simulating in a virtual space using information about past work performed in real space by a multi-joint, multi-axis robot on a specified workpiece; a memory function that stores robot configuration information including the shape of each arm that constitutes the robot, the connection relationships between the arms, and the movable range of each arm; a reception function that receives workspace information including information about the shape of the workpiece to be taught and the workspace of the robot that processes the workpiece, as well as work requirements that have been set in advance for the workpiece; a generation function that generates a three-dimensional model of the workspace including the robot and workpiece in a virtual space based on the robot configuration information and the workspace information; and a simulation function that outputs the results of a simulation in virtual space of the robot's operations that satisfy the work requirements as control information for the movable range in real space based on the learning model. [Effects of the Invention]

[0020] The information processing device etc. of the present invention is characterized by comprising: a learning unit that generates a learning model that is learned by simulating in a virtual space using information about past work performed in real space by a multi-joint, multi-axis robot on a specified workpiece; a memory unit that stores robot configuration information including the shape of each arm that constitutes the robot, the connection relationships between the arms, and the movable range of each arm; a reception unit that receives workspace information including information about the shape of the workpiece to be taught and the workspace of the robot that processes the workpiece, as well as work requirements that have been set in advance for the workpiece; a generation unit that generates a three-dimensional model of the workspace including the robot and workpiece in a virtual space based on the robot configuration information and the workspace information; and a simulation unit that outputs the results of a simulation of the robot's operations in the virtual space that satisfy the work requirements as control information for the movable range in real space based on the learning model, thereby providing an information processing device etc. that enables simulation of robot control without exceeding the limits of the movable range of operations in the simulation of robot control of a multi-axis robot. [Brief explanation of the drawings]

[0021] [Figure 1] FIG. 1 is a schematic configuration diagram of a simulation system according to one embodiment of the present invention and a block diagram of an information processing device. [Figure 2] FIG. 2 is a diagram illustrating an example of the configuration of a robot. [Figure 3] Figure 3 shows an example of a screen used when setting the connection relationships between the components of a robot. [Figure 4] FIG. 4 is a flowchart of a joint rotation amount control program of an information processing device according to an embodiment of the present invention. [Figure 5] FIG. 5 is a control flow of an information processing device according to an embodiment of the present invention. [Figure 6] FIG. 6 is a diagram illustrating an outline of a simulation system according to one embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0022] An information processing device according to one embodiment of the present invention will be described in detail below with reference to the drawings. Note that the drawings are merely examples, and the information processing device of the present invention is not limited to those shown. Furthermore, the drawings are schematic, and the size ratios of the components of the robot, the number of workpieces, the workspace, and objects present around the robot, and their positional relationships are not strictly accurate.

[0023] <Embodiment> <System configuration> Fig. 1 is a schematic diagram of a simulation system including an information processing device 100 according to one embodiment of the present invention. The simulation system 500 includes the information processing device 100, an articulated, multi-axis robot 10, and a robot control device 200. The robot 10 is a painting robot that drives each arm constituting the robot under the control of the robot control device 200 to paint a workpiece 12. Note that Fig. 1 shows a painting robot with a spray gun attached to its tip as an example, but the robot 10 is not limited to this and may be a welding robot, a transport robot, a pick-up robot, or the like.

[0024] The information processing device 100 reproduces the robot 10, the workpiece (workpiece) 12, and the object 14 contained in the robot 10's workspace in real space as three-dimensional models in a virtual space, and simulates the operation of the robot 10. Here, the object 14 is an object present around the robot 10, and may be another workpiece, another robot, control equipment, or the like. The virtual space in which the simulation is performed is displayed on a display unit 141. The display unit 141 is a monitor having a function of displaying various information output from the information processing device 100, and can be realized by, for example, an LCD (Liquid Crystal Display), an organic EL display, or the like. Note that the display unit 141 may also be, for example, a monitor of a mobile terminal or a tablet terminal. An operator can input instructions to the information processing device 100 using an input device such as a keyboard 131 or a mouse (not shown). Note that in the information processing device 100 according to this embodiment, the display unit 141 is not essential for performing machine learning (reinforcement learning), which will be described later.

[0025] The robot control device 200 drives the robot 10 based on robot control information output by a simulation in the information processing device 100. The control information may include, for example, values ​​such as the target angle and angular velocity of each axis (joint) constituting the robot 10, and the torque and drive power of the drive motor. Note that while FIG. 1 shows a configuration in which the robot control device 200 is connected to the information processing device 100, the present invention is not limited to this. In other words, the robot control device 200 may exist separately from the information processing device 100.

[0026] In the simulation system 500, the information processing device 100 reduces the load in simulating a plurality of movements of the robot 10 and provides an environment for efficient machine learning.

[0027] Next, the configuration of the robot 10 will be described with reference to Figures 2 and 3. Figure 2 is a diagram for explaining an example of the configuration of the robot, and Figure 3 is an example of a screen when setting the connection relationships of the components of the robot. The robot 10 is a multi-jointed, six-axis arm robot. The joints of the arm robot are movable parts, and each rotates around an axis. The robot 10 comprises a base 10a, a first arm 10b, a second arm 10c, a third arm 10d, and an end effector 10e. The base 10a is the base of the robot 10, which is placed directly on the installation location. The first arm 10b is connected to the base 10a via a first axis. The second arm 10c is connected to the first arm 10b via a second axis. The third arm 10d is connected to the second arm 10c via a third axis and a fourth axis. The end effector 10e is connected to the third arm 10d via a fifth axis and a sixth axis.

[0028] A rotation angle range is set for each joint. For example, the rotation angle range of the third arm 10d is set to, for example, -110 degrees to 150 degrees with the third axis 10f as the center of rotation, and the limit value 10g of the movable range is set to -110 degrees to 150 degrees (see FIG. 2). The soft limit value is set so that the third arm 10d of the robot model 20 in the virtual space does not exceed the limit value 10g of the movable range. The soft limit value is not a physical limit at which the motors installed at each joint (axis) of the robot can rotate, but a value at which the soft limit control set on software is executed.

[0029] Conventionally, the soft limit of the robot model 20 in the virtual space has been controlled using Rotation Limit Hinge, which is one of the functions of Final IK (inverse kinematics), which is one of the assets of the game development platform Unity (registered trademark). Final IK is an asset that provides the ability to freely move characters and other objects within Unity, and Rotation Limit Hinge provides a function to control the amount of rotation of the joints of characters and other objects, and is used to express the movements of characters and other objects more naturally.

[0030] When using the Rotation Limit Hinge to control the soft limit of the robot model 20's movement, if any of the six axes of the robot model 20 reaches the soft limit value, a phenomenon may occur in which control becomes ineffective thereafter. For example, when the rotation of the third axis (soft limit range -110 degrees to 150 degrees) reaches the lower limit of -110 degrees, the third axis continues to stop (the third axis becomes fixed) even when set to -110 degrees or higher in a subsequent action (see Figure 2). The above-mentioned conventional method of controlling the robot model 20 using soft limits causes the movement of the robot model 20 in virtual space to stop during reinforcement learning, preventing learning from progressing.

[0031] Furthermore, the amount of rotation of an axis controlled by Rotation Limit Hinge can be controlled from 0 degrees to values ​​of 360 degrees or more. In contrast, rotational movements of game development platforms (for example, Unity) are expressed in degrees from -180 degrees to 180 degrees. For this reason, when the amount of rotation handled by Rotation Limit Hinge is handed over to the game development platform, it is converted into the expression of the rotational movements of the game development platform, which has resulted in a problem of a discrepancy between the amount of rotation of the axis controlled by Rotation Limit Hinge and the amount of rotation of the axis expressed by the game development platform.

[0032] In addition, when resetting the rotation amount of the axis, even if the rotation was more than one revolution, the rotation amount of 0 degrees and 180 degrees would be expressed as the same value, which caused the reset to not work properly. To solve this problem, the information processing device 100 uses a joint rotation amount control program that controls the rotation amount of the joints of the robot model 20 so that it does not exceed the soft limit range, instead of the conventional control of the soft limit of the rotation amount of the axis using a Rotation Limit Hinge. This joint rotation amount control program will be described with reference to Fig. 4. Fig. 4 is a flowchart of the joint rotation amount control program of the information processing device 100 according to one aspect of the present invention.

[0033] If the cumulative rotation amount of each axis of the robot 10 in the virtual space obtained from the accumulation of the network output for each reinforcement learning episode exceeds a preset soft limit value range, the information processing device 100 adjusts the cumulative rotation amount so that it is within the soft limit value range.

[0034] (Step S50) The information processing device 100 acquires the rotation amount increment for each axis and the distance from the target to the tip of the gun. The information processing device 100 acquires the increment of rotation amount for each axis of the robot model 20 obtained from the network output for each episode (one unit of learning) of reinforcement learning (step T13) described below, and the distance between the target and the tip of the gun. The tip of the gun is also called the aim point. The rotation increment of each axis of the robot model 20 obtained from the network output is a value in which the action value of the action of the agent in reinforcement learning is reflected on each axis of the robot model 20.

[0035] (Step S51) The information processing device 100 calculates the cumulative rotation amount for each axis. The information processing device 100 calculates the torque increment from the rotation amount increment of each axis, adds the torque increment to the cumulative torque from one episode ago to calculate the cumulative torque of each axis at the current episode point in time, and converts this into the cumulative rotation amount of each axis at the current episode point in time.

[0036] (Step S52) The information processing device 100 checks whether the accumulated rotation amount has reached the soft limit value. The information processing device 100 checks whether the cumulative rotation amount of each axis at the time of the current episode has reached the soft limit value set for each axis.

[0037] (Step S53) If the accumulated rotation amount reaches the soft limit value, the information processing device 100 adjusts the rotation amount increment. If the cumulative rotation amount of any one of the axes at the time of the current episode has reached the soft limit, the information processing device 100 adjusts the rotation amount increment of each axis of the robot model 20 obtained from the network output of the current episode of reinforcement learning (step T13) so that the cumulative torque at the current time is within the range of the soft limit value.

[0038] (Step S54) The information processing device 100 expresses the angle on each axis as -360 degrees to 360 degrees, and replaces it with the cumulative rotation amount for updating. The information processing device 100 expresses the angle of the current episode on each axis in the angle expression of the game development platform, from -360 degrees to 360 degrees, and updates it to the accumulated rotation amount at the time of the current episode.

[0039] The joint rotation amount control program of the information processing device 100 adjusts the angle of each axis so that it is within the soft limit range even if the angle of each axis exceeds the soft limit range. Therefore, the information processing device 100 can prevent the simulation of the robot model 20 from stopping due to the soft limit, and can eliminate factors that hinder the progress of machine learning.

[0040] <Hardware configuration> The information processing device 100 includes a control unit 110, a communication I / F (interface) 120, an input I / F 130, an output I / F 140, and a storage unit 170 as hardware components.

[0041] The control unit 110 is typically a processor, and includes a central processing unit (CPU), a micro processing unit (MPU), a graphics processing unit (GPU), etc., and may be realized by a logic circuit (hardware) formed in an integrated circuit (an integrated circuit (IC) chip, a large scale integration (LSI)), etc., or a dedicated circuit. The control unit 110 may execute the functions and methods described in each embodiment by reading a program stored in the storage unit 170 and executing code or instructions included in the read program.

[0042] The storage unit 170 stores various programs and various data required for the operation of the information processing device 100. The storage unit 170 may include, for example, a hard disk drive (HDD), a solid state drive (SSD), a flash memory, etc. The storage unit 170 may also include memory (random access memory (RAM), read only memory (ROM), etc.) that provides a working area for the control unit 110. The information processing device 100 according to this embodiment functions as a reception unit 112, a generation unit 113, a simulation unit 114, a setting unit 115, and a learning unit 116 (described later) when the control unit 110 executes a program loaded into the memory of the storage unit 170. For example, the storage unit 170 may store motion analysis software for simulating the movement of the robot 10 in a virtual space.

[0043] The storage unit 170 stores robot configuration information including the shape of each arm constituting the robot 10, the connection relationships between the arms, and the range of motion of each arm. The robot configuration information may include the length, width, and length of the arms constituting the robot, the distance between joints (effective length), the connection relationships between the arms, and identification information of the connecting joints. For example, the robot configuration information may be extracted from CAD (Computer Aided Design) data of the robot 10. Furthermore, the robot configuration information may include information indicating the range of motion of each arm, such as soft limit values, the rotation direction of each joint, the range of possible rotation angles, the maximum angular velocity, and the maximum angular acceleration. Note that this data may be received and stored by the receiving unit 112 (described later) via the input device (keyboard) 131 input by an operator. Although FIG. 1 illustrates the storage unit 170 as being integrated with the information processing device 100, the data stored in the storage unit 170 may be stored in a database server separate and independent from the information processing device 100.

[0044] The communication I / F 120 is implemented as hardware such as a network adapter, communication software, or a combination of these, and transmits and receives various data to and from an external device such as the robot control device 200, for example.

[0045] The input I / F 130 is a connection interface with an input device that inputs various operations to the information processing device 100. The input device includes, for example, a touch panel, hardware keys such as a keyboard 131, and a pointing device such as a mouse.

[0046] The output I / F 140 is a connection interface with an output device that outputs processing results such as simulation results by the information processing device 100. The output device includes, for example, a display unit (display) 141, a speaker, and the like.

[0047] <Functional configuration> The control unit 110 includes a receiving unit 112, a generating unit 113, a simulation unit 114, a setting unit 115, and a learning unit .

[0048] The receiving unit 112 receives workspace information, including information about the shape of the workpiece 12 to be taught and information about the workspace 300 of the robot 10 that processes the workpiece 12, as well as task requirements previously set for the workpiece 12. The workspace information and task requirements may be input by an operator via the keyboard 131. The workpiece 12 is an object that the robot 10 works on, and may be, for example, a part of a vehicle or other workpiece. The receiving unit 112 may receive information about the shape of the workpiece 12, such as the length, width, and height of the workpiece 12. Furthermore, the information about the workspace 300 may include information about an object 14 that exists in the workspace 300 of the robot 10 in real space. The object 14 is an object that exists in a position that may interfere with the robot 10's work, such as another workpiece, peripheral equipment, or another adjacent robot. The receiving unit 112 may receive information about the length, width, height, placement position, etc. of the object 14. The receiving unit 112 may acquire the above-mentioned workspace information from a design drawing (CAD data, etc.) of the workpiece 12 or the object 14, or from a moving image of the workpiece 12 or the object 14.

[0049] The work requirement is information indicating the work to be performed on the workpiece 12, and may be, for example, the locations to be painted on the workpiece 12 and the painting route (painting line) connecting the locations to be painted.

[0050] The generation unit 113 generates a 3D model of the workspace 300 including the robot 10 and workpiece 12 in the virtual space 400 based on the robot configuration information and workspace information. Hereinafter, the 3D models of the robot 10, workpiece 12, and object 14 will also be referred to as the robot model, workpiece model, and object model, respectively. Fig. 1 shows the display unit 141 on which the virtual space 400 including the robot model 20, workpiece model 22, and object model 24 is displayed.

[0051] Here, the generation unit 113 generates a 3D model by voxelizing (discretizing) the object model 24. Voxelization is a representation format in which the 3D model is divided into cubic regions of a predetermined size and information is reduced. Note that existing technologies may be used for voxelizing the object model 24 and generating the 3D model.

[0052] The simulation unit 114 may use a three-dimensional model of the robot 10 in the virtual space 400 to simulate the actions of the robot 10 that satisfy the task requirements, and output the results as control information for the range of motion in the real space based on the learning model.

[0053] The setting unit 115 sets the connection relationships between the arms that make up the robot 10. Fig. 3 shows an example of a setting screen for setting the connection relationships between the arms. The setting screen 40 is displayed on the display unit 141, and the operator can define the connection relationships between the arms on the setting screen 40.

[0054] The learning unit uses information about past work performed by the robot 10 on a predetermined workpiece in the real space to generate a learning model learned by simulating in the virtual space 400. The learning unit will be described later.

[0055] <Control flow> The above-mentioned processing by the information processing device 100 will be described with reference to the flowchart of FIG. (Step S110) The learning unit 116 uses information about past work performed in real space by a multi-joint, multi-axis robot on a predetermined workpiece to generate a learning model learned by simulating it in the virtual space 400. More specifically, the learning unit 116 may perform learning by referring to a program based on offline teaching that was used for the work in real space. In this case, the learning unit 116 may generate a learning model by machine learning using at least the shape of a specified workpiece in past work, the work requirements previously set for the specified workpiece, and the past work space as input data, and the work performed by the robot 10 on the specified workpiece as output data.

[0056] (Step S111) The storage unit 170 stores robot configuration information including the shape of each arm constituting the robot 10, the connection relationship between the arms, and the range of motion of each arm.

[0057] (Step S112) The receiving unit 112 receives workspace information including information about the shape of the workpiece 12 to be taught and the workspace 300 of the robot 10 that processes the workpiece 12, as well as work requirements that have been set in advance for the workpiece 12.

[0058] (Step S113) The generation unit 113 generates a three-dimensional model of the workspace including the robot and the workpiece in the virtual space 400 based on the robot configuration information and workspace information.

[0059] (Step S114) The simulation unit 114 outputs the results of simulating the operation of the robot 10 that satisfies the task requirements in the virtual space 400 as control information for the robot 10 in the real space.

[0060] Learning by the information processing device 100 according to one aspect of the present invention will be described with reference to FIG. 6. The storage unit (database) 170 stores information about work previously performed by the robot 10. The database 170 may include, for example, the painting posture, entry / exit posture, intermediate posture, arm type and its soft limit value, vehicle type as the workpiece, the work environment including objects present in the work space, and the painting route of the robot 10. Note that a separate database 170 may exist for each robot. Specifically, the learning unit 116 uses the information contained in the database 170 to learn a model for generating an optimal painting route. The display unit 141 displays the three-dimensional environment (virtual space) constructed by the generation unit 113 of the information processing device 100 and a simulation performed by the simulation unit 114, in which the arm of the robot model 20 is operated (steps T11 and T12). The learning unit 116 performs learning to generate an optimal painting route using a reinforcement learning algorithm, such as Q-learning, ε-greedy algorithm, or policy gradient method (step T13). The generated learning model may be stored in the database 171, and control data for controlling the robot 10 may be output (step T14).

[0061] While the present invention has been described based on the drawings and examples, it should be noted that those skilled in the art would readily be able to make various modifications and alterations based on the present disclosure. Therefore, it should be noted that these modifications and alterations are within the scope of the present invention. For example, the functions included in each component, step, etc. may be rearranged so as not to cause logical inconsistencies, and multiple components, steps, etc. may be combined or separated into one. Furthermore, the configurations described in the above embodiments may be appropriately combined. For example, each component described as being included in the information processing device 100 may be realized in a distributed manner across multiple information processing devices.

[0062] Although a painting robot has been described above as an example of a robot, the present invention is not limited to painting robots and may be applied to other robots as long as they operate to perform some task. For example, the robot may be a welding robot or a transport robot that grasps a workpiece and moves it to a predetermined location.

[0063] Furthermore, in the above description, the method by which the information processing device automatically executes processing is described as the processors functioning as each functional unit constituting the information processing device 100 executing a robot simulation program (processing program) or the like to execute predetermined processing, but this may also be realized by a logic circuit (hardware) or dedicated circuit formed in an integrated circuit (IC (Integrated Circuit) chip, LSI (Large Scale Integration)) or the like in the device.

[0064] The processing program may be recorded on a processor-readable recording medium, and the recording medium may be a "non-transitory tangible medium" such as a tape, disk, card, semiconductor memory, or programmable logic circuit. The processing program may also be supplied to the processor via any transmission medium capable of transmitting the processing program (such as a communications network or broadcast waves). The present invention may also be realized in the form of a data signal embedded in a carrier wave, in which the processing program is embodied by electronic transmission.

[0065] The processing program can be implemented using, for example, a scripting language such as ActionScript or JavaScript (registered trademark), an object-oriented programming language such as Python (registered trademark), Objective-C or Java (registered trademark), or a markup language such as HTML5.

[0066] Although the present invention has been described based on the drawings and examples, it should be noted that various modifications and alterations would be readily apparent to those skilled in the art based on the present disclosure. Therefore, it should be noted that these modifications and alterations are within the scope of the present invention. For example, the present invention is applicable to various multi-axis robots, such as six-axis arm robots including a traveling axis and seven-axis arm robots. [Explanation of symbols]

[0067] 10. Robot 10a Pedestal 10b First arm 10c Second Arm 10d Third Arm 10e End Effector 10f 3rd axis 10g limit of movement range 12 Workpiece 14 Object 20 Robot Models 22 Workpiece model 24 Object Models 40 Settings screen 100 Information processing device 110 control section 112 Reception 113 Generation part 114 Simulation Department 115 Setting section 116 Learning Department 120 Communication I / F (Interface) 130 Input I / F 131 Keyboard (input section) 140 Output I / F 141 Display section 170 Storage section 200 Robot control device 300 workspace 400 Virtual Space 500 Simulation System

Claims

1. a learning unit that generates a learning model by simulating in a virtual space using information about past work performed in real space by a multi-joint, multi-axis robot on a predetermined workpiece; a storage unit that stores robot configuration information including the shape of each arm that configures the robot, the connection relationship between the arms, and the range of motion of each arm; a receiving unit that receives workspace information including information about the shape of a workpiece to be taught and a workspace of the robot that processes the workpiece, as well as preset work requirements for the workpiece; a generation unit that generates a three-dimensional model of the workspace including the robot and the workpiece in the virtual space based on the robot configuration information and the workspace information; a simulation unit that outputs a result of simulating an operation of the robot that satisfies the task requirements in the virtual space based on the learning model as control information for the movable range in the real space; An information processing device comprising:

2. 2. The information processing apparatus according to claim 1, wherein the learning is reinforcement learning.

3. 3. The information processing device according to claim 2, wherein, when the cumulative rotation amount of each axis of the robot in the virtual space obtained from the accumulation of network outputs for each episode of the reinforcement learning exceeds a range of a preset soft limit value, the cumulative rotation amount is adjusted so that it is within the range of the soft limit value.

4. the simulation unit uses the learning model to obtain control information that satisfies the work requirements that are preset for the workpiece on which the teaching is performed. The information processing device according to claim 1 .

5. the learning unit generates the learning model by machine learning using at least the shape of the predetermined workpiece, the work requirements preset for the predetermined workpiece, and the past work space as input data, and the work performed by the robot on the predetermined workpiece as output data. The information processing device according to claim 1 .

6. The robot is a painting robot that paints the workpiece. The information processing device according to claim 1 .

7. a setting unit that sets a connection relationship between the arms that configure the robot, The information processing device according to claim 1 .

8. The information processing device a learning step of generating a learning model by simulating in a virtual space using information about past work performed in a real space by a multi-joint, multi-axis robot on a predetermined workpiece; a storage step of storing robot configuration information including the shape of each arm constituting the robot, the connection relationship between the arms, and the movable range of each arm; a receiving step of receiving workspace information including information about the shape of a workpiece to be taught and a workspace of the robot that processes the workpiece, as well as preset work requirements for the workpiece; a generation step of generating, in the virtual space, a three-dimensional model of the workspace including the robot and the workpiece based on the robot configuration information and the workspace information; a simulation step of simulating, in the virtual space, an operation of the robot that satisfies the task requirements, and outputting the result as control information for the movable range in the real space based on the learning model; 2. A method for controlling an information processing device, comprising:

9. In the information processing device, A learning function that generates a learning model by simulating in a virtual space using information on past work performed in the real space by a multi-joint, multi-axis robot on a specified workpiece; a storage function for storing robot configuration information including the shape of each arm constituting the robot, the connection relationship between the arms, and the range of motion of each arm; a receiving function for receiving workspace information including information on the shape of a workpiece to be taught and the workspace of the robot for processing the workpiece, as well as preset work requirements for the workpiece; a generation function for generating a three-dimensional model of the workspace including the robot and the workpiece in the virtual space based on the robot configuration information and the workspace information; a simulation function that outputs, in the virtual space, a simulation result of an operation of the robot that satisfies the task requirements, as control information for the movable range in the real space based on the learning model; and A control program for an information processing device, which realizes the above.

Images