Method and system for handling a load arrangement with a robot gripper

Patent metadata

Field	Value
Publication number	US-12583102-B2
Application number	US-202118042051-A
Country	US
Kind code	B2
Filing date	Aug 12, 2021
Priority date	Aug 19, 2020
Publication date	Mar 24, 2026
Grant date	Mar 24, 2026

Abstract

A method for handling a load arrangement with a robot includes:
activating a lifting state of a gripper of the robot for load lifting; determining a parameter of a time profile of a load arrangement-dependent force variable using at least one sensor of the robot during a movement of the lifted load arrangement; classifying a load arrangement lifted by the gripper using a machine-learned model on the basis of the determined parameter, in particular during a movement of the lifted load arrangement and/or over the pick-up area in which the load arrangement has been situated for lifting, in particular a pick-up area of a pick-up station and/or over or in a pick-up container;

and at least one of the steps of:
carrying out a first process with the robot if the load arrangement has been classified into a first class; and/or carrying out a second process with the robot if the load arrangement has been classified into a second class.

First claim

What is claimed is:

1 . A method for handling a load arrangement with a robot, the method comprising:
activating a lifting state of a gripper of the robot for load lifting; determining a parameter of a time profile of a load-arrangement-dependent force variable using at least one sensor of the robot during a movement of the lifted load arrangement; classifying the load arrangement lifted by the gripper using a machine-learned model on the basis of the determined parameter; and at least one of:
carrying out a first process with the robot in response to the load arrangement being classified into a first class, or
carrying out a second process with the robot in response to the load arrangement being classified into a second class;
wherein the robot is controlled to carry out the first process or the second process based on a distribution of the load relative to the gripper.

2 . The method of claim 1 , wherein the load arrangement is at least one of:
classified during a movement of the lifted load arrangement; classified during a movement of the lifted load arrangement over a pick-up area in which the load arrangement has been situated for lifting; classified during a movement of the lifted load arrangement over a pick-up area of a pick-up station; or classified during a movement of the lifted load arrangement over or in a pick-up container.

3 . The method of claim 1 , further comprising at least one of:
carrying out a third process with the robot in response to the load arrangement being classified into a third class; or carrying out a fourth process with the robot in response to the load arrangement being classified into a fourth class.

4 . The method of claim 1 , wherein the classes differ from one another in at least one of the number, distribution, or type of loads lifted by the gripper.

5 . The method of claim 3 , wherein the classes differ from one another in at least one of the number, distribution, or type of loads lifted by the gripper.

6 . The method of claim 1 , wherein at least one of:
the gripper has at least two lifting devices; the sensor is arranged on the gripper, between the gripper and a robot arm to which it is fastened, or on the robot arm; or the force variable depends on at least one of:
at least one load-arrangement-dependent force component, or
at least one load-arrangement-dependent torque component.

7 . The method of claim 6 , wherein at least one of:
the at least two lifting devices of the gripper are suction devices; or the at least one load-arrangement-dependent force component is a vertical load-arrangement-dependent force component.

8 . The method of claim 3 , wherein at least one of:
the first process comprises a transfer movement of the lifted load arrangement using the robot; the second process comprises at least one of:
a complete or partial deactivation of the lifting state of the gripper,
a different picking operation from the first process or,
a repetition of a lifting attempt; or
the third process comprises at least one of:
a complete or partial deactivation of the lifting state of the gripper,
a different picking operation from the first and second processes, or
a repetition of a lifting attempt.

9 . The method of claim 8 , wherein at least one of:
the transfer movement of the first process is at least one of a movement out of the pick-up area, or a movement to a picking station; the deactivation of the gripper in at least one of the second or third processes comprises at least one of deactivating one or more of the lifting devices of the gripper, or deactivating the gripper over the pick-up area; or the different picking operation of at least one of the second or third processes is a sequential picking operation.

10 . The method of claim 8 , wherein at least one of:
at least one of the second or third processes comprises a guided deposition movement of the lifted load arrangement using the robot; or the partial deactivation of the lifting state of the gripper comprises a deactivation of at least one of the lifting devices, while at least one further of the lifting devices which is identified as load-bearing remains activated.

11 . The method of claim 10 , wherein the guided deposition movement of the lifted load arrangement is at least one of a movement in the pick-up area, or a movement before deactivation of the lifting state.

12 . Method of claim 1 , wherein at least one of:
the machine-learned model has at least one artificial neural network; or the machine-learned model is machine-learned by lifting at least one learning load arrangement with the gripper of the robot.

13 . The method of claim 12 , wherein at least one of:
the class of the at least one learning load arrangement is input manually or automatically, or is specified by only partial activation of the gripper; or the model is mechanically learned by lifting multiple different learning load arrangements with the gripper of the robot.

14 . The method of claim 13 , wherein specifying the class of the at least one load arrangement by partial activation of the gripper comprises specifying by activation of only one of the lifting devices of the gripper.

15 . A system for handling a load arrangement with a robot, the system comprising:
means for activating a lifting state of a gripper of the robot for load lifting; means for determining a parameter of a time profile of a load-arrangement-dependent force variable using at least one sensor of the robot during a movement of the lifted load arrangement; means for classifying the load arrangement lifted by the gripper using a machine-learned model on the basis of the determined parameter; and means for at least one of:
carrying out a first process with the robot in response to the load arrangement being classified into a first class, or
carrying out a second process with the robot in response to the load arrangement being classified into a second class;
wherein the robot is controlled to carry out the first process or the second process based on a distribution of the load relative to the gripper.

16 . The system of claim 15 , wherein the means for classifying is configured for at least one of:
classifying the load arrangement during a movement of the lifted load arrangement; classifying the load arrangement during a movement of the lifted load arrangement over a pick-up area in which the load arrangement has been situated for lifting; classifying the load arrangement during a movement of the lifted load arrangement over a pick-up area of a pick-up station; or classifying the load arrangement during a movement of the lifted load arrangement over or in a pick-up container.

17 . A computer program product for handling a load arrangement with a robot and comprising a program code stored on a non-transitory, computer-readable medium, the program code, when executed by a computer, causing the computer to:
activate a lifting state of a gripper of the robot for load lifting; determine a parameter of a time profile of a load-arrangement-dependent force variable using at least one sensor of the robot during a movement of the lifted load arrangement; classify the load arrangement lifted by the gripper using a machine-learned model on the basis of the determined parameter; and at least one of:
carry out a first process with the robot in response to the load arrangement being classified into a first class, or
carry out a second process with the robot in response to the load arrangement being classified into a second class;

Assignees

• Kuka Deutschland Gmbh

Inventors

• Kaspar Manuel

Classifications

• B25J15/0616

  with vacuum · CPC title

• B25J9/1653

  parameters identification, estimation, stiffness, accuracy, error analysis · CPC title

• B25J9/1638

  compensation for arm bending/inertia, pay load weight/inertia · CPC title

• B25J9/1612

  characterised by the hand, wrist, grip control · CPC title

• B25J9/0096

  co-operating with a working support, e.g. work-table · CPC title