Granted Patent B2

US 12,280,507 · App. 17/784,916 · Granted Apr 22, 2025

Inventors: Andreas Spenninger (Karlsfeld, DE); Jose Ramon Medina Hernandez (Munich, DE)

Assignee: Franka Emika GmbH

B25J9/1664B25J9/161B25J13/085B25J13/088G05B2219/40599G05B2219/40607

Abstract

Claims (68)

1. A method of generating a control program for a second robot manipulator based on empirical data from executing a predetermined application by a first robot manipulator, the method comprising:

executing the predetermined application by the first robot manipulator;

during execution of the predetermined application: determining time series of trajectory data by joint angle sensors of the first robot manipulator and/or time series of wrench data by a sensor unit of the first robot manipulator for detecting forces and/or torques, and storing the determined time series in a storage unit, the trajectory data comprising kinematic data relating to a reference point of the first robot manipulator or relating to the joint angles of the first robot manipulator, and the wrench data comprising forces and/or torques acting between the first robot manipulator and an object in the surroundings;

determining robot commands from the stored time series, and storing the determined robot commands in the storage unit, the robot commands being principal elements of a control program for a respective robot manipulator without reference to design conditions of the first robot manipulator; and

generating the control program for the second robot manipulator based on the stored robot commands and based on design conditions of the second robot manipulator.

2. The method of claim 1 , wherein the robot commands comprise at least one category of different categories comprising:

a predetermined path curve of a reference point of the respective robot manipulator from a predetermined start point to a predetermined end point of the predetermined path curve;

a velocity of the reference point on the predetermined path curve;

an acceleration of the reference point on the predetermined path curve;

a force and/or a torque that the reference point of the respective robot manipulator exerts on an object in the surroundings of the respective robot manipulator; and

target torques for rotational actuators of the respective robot manipulator.

3. The method of claim 2 , wherein the method comprises:

determining at least two successive robot commands from the different categories; and

determining a blending transition between the at least two successive robot commands from the different categories.

4. The method of claim 3 , wherein the method further comprises performing the blending transition by a continuous, and over time of the blending transition, a time-dependent predetermined function course.

5. The method of claim 1 , wherein the method further comprises performing nonlinear optimization in determination of the robot commands from the stored time series.

6. The method of claim 1 , wherein in determination of the robot commands from the stored time series, the method further comprises:

applying a predetermined artificial neural network, an input variable of the artificial neural network being the stored time series and an output variable of the artificial neural network being a respectively selected one of a plurality of structurally predetermined robot commands; and

adapting parameters of the respectively selected one of the predetermined robot commands based on the stored time series.

7. The method of claim 1 , wherein determination of the time series of trajectory data is additionally performed by a camera unit.

8. The method of claim 7 , wherein the camera unit is an external camera unit.

9. The method of claim 1 , wherein the design conditions of the first robot manipulator and/or the design conditions of the second robot manipulator comprise at least one of the following:

distances between joints of the respective robot manipulator;

number of joints of the respective robot manipulator;

maximum torque that is capable of being applied by rotational actuators of the respective robot manipulator;

type and configuration of an end effector of the respective robot manipulator;

virtual stiffness of a closed-loop control of the respective robot manipulator;

material stiffness of links and/or joints of the respective robot manipulator;

a geometrically maximum possible working space of the respective robot manipulator;

time constants and/or bandwidths of actuators of the respective robot manipulator;

safety level, and/or current safety configuration, and/or residual risk of the respective robot manipulator;

physical existence and/or configuration of communication interfaces of the respective robot manipulator;

number of robot arms of the respective robot manipulator; and

mass and/or inertia of components, in particular links, of the respective robot manipulator.

10. A robot system to generate a control program for a second robot manipulator of the robot system based on empirical data from executing a predetermined application by a first robot manipulator of the robot system, the robot system comprising:

a first control unit configured to:

control the first robot manipulator to execute the predetermined application, and further designed to determine time series of trajectory data by joint angle sensors of the first robot manipulator and/or time series of wrench data by a sensor unit of the first robot manipulator during execution of the predetermined application and to store the determined time series in a storage unit, the trajectory data comprising kinematic data relating to a reference point of the first robot manipulator or relating to the joint angles of the first robot manipulator, the wrench data comprising forces and/or torques acting between the first robot manipulator and an object in the surroundings; and

determine robot commands from the stored time series and to store the determined robot commands in the storage unit, the robot commands being principal elements of a control program for a respective robot manipulator without reference to design conditions of the first robot manipulator; and

a second control unit configured to generate the control program for the second robot manipulator based on stored robot commands and based on design conditions of the second robot manipulator.

11. The robot system of claim 10 , wherein the robot commands comprise at least one category of different categories comprising:

a predetermined path curve of a reference point of the respective robot manipulator from a predetermined start point to a predetermined end point of the predetermined path curve;

a velocity of the reference point on the predetermined path curve;

an acceleration of the reference point on the predetermined path curve;

a force and/or a torque that the reference point of the respective robot manipulator exerts on an object in the surroundings of the respective robot manipulator; and

target torques for rotational actuators of the respective robot manipulator.

12. The robot system of claim 11 , wherein the first control unit is further configured to:

determine at least two successive robot commands from the different categories; and

determine a blending transition between the at least two successive robot commands from the different categories.

13. The robot system of claim 12 , wherein the first control unit is further configured to perform the blending transition by a continuous, and over time of the blending transition, a time-dependent predetermined function course.

14. The robot system of claim 10 , wherein the first control unit is further configured to perform nonlinear optimization in determination of the robot commands from the stored time series.

15. The robot system of claim 10 , wherein in determination of the robot commands from the stored time series, the first control unit is further configured to:

apply a predetermined artificial neural network, an input variable of the artificial neural network being the stored time series and an output variable of the artificial neural network being a respectively selected one of a plurality of structurally predetermined robot commands; and

adapt parameters of the respectively selected one of the predetermined robot commands based on the stored time series.

16. The robot system of claim 10 , wherein determination of the time series of trajectory data is additionally performed by a camera unit.

17. The robot system of claim 16 , wherein the camera unit is an external camera unit.

18. The robot system of claim 10 , wherein the design conditions of the first robot manipulator and/or the design conditions of the second robot manipulator comprise at least one of the following:

distances between joints of the respective robot manipulator;

number of joints of the respective robot manipulator;

maximum torque that is capable of being applied by rotational actuators of the respective robot manipulator;

type and configuration of an end effector of the respective robot manipulator;

virtual stiffness of a closed-loop control of the respective robot manipulator;

material stiffness of links and/or joints of the respective robot manipulator;

a geometrically maximum possible working space of the respective robot manipulator;

time constants and/or bandwidths of actuators of the respective robot manipulator;

safety level, and/or current safety configuration, and/or residual risk of the respective robot manipulator;

physical existence and/or configuration of communication interfaces of the respective robot manipulator;

number of robot arms of the respective robot manipulator; and

mass and/or inertia of components, in particular links, of the respective robot manipulator.

Assignments (3)

Priority Claims (1)

Continuity (1)

References Cited (31)