Method and device for operating a machine

Patent metadata

Field	Value
Publication number	US-12304087-B2
Application number	US-202117355961-A
Country	US
Kind code	B2
Filing date	Jun 23, 2021
Priority date	Jun 30, 2020
Publication date	May 20, 2025
Grant date	May 20, 2025

Abstract

A device for and method of operating a machine. The method includes providing a sequence of skills of the machine for executing a task, selecting a sequence of states from a plurality of sequences of states, depending on a likelihood, wherein the likelihood is determined depending on a transition probability from a final state of a first sub-sequence of states of the sequence of states for a first skill in the sequence of skills to an initial state of a second sub-sequence of states of the sequence of states for a second skill in the sequence of skills.

First claim

What is claimed is:

1. A method of operating a machine, the method comprising the following steps:
providing a sequence of skills of the machine for executing a task;
selecting a sequence of states from a plurality of sequences of states, depending on a likelihood, wherein the likelihood is determined depending on a transition probability from a final state of a first sub-sequence of states of the sequence of states for a first skill in the sequence of skills to an initial state of a second sub-sequence of states of the sequence of states for a second skill in the sequence of skills; and
executing the task based on at least the first skill and the second skill, wherein the task includes physically manipulating an object by the machine, wherein the final state of the first sub-sequence is determined depending on a final component of an instance of the first skill, wherein the initial state of the second sub-sequence is determined depending on an initial component of an instance of the second skill, and wherein the transition probability is determined depending on a divergence between a first Gaussian mixture Model for the final component and a second Gaussian mixture Model for the initial component.

2. The method according to claim 1 , further comprising:
providing a first model for the first skill, wherein a parameter of the first model defines a trajectory in a state of the first sub-sequence of states, or
providing a first model for the second skill, wherein a parameter of the first model defines a trajectory in a state of the second sub-sequence of states;
wherein the method further comprises determining for the state of the sequence of states, the parameter, and determining a control signal for operating the machine according to the trajectory depending on the parameter.

3. The method according to claim 1 , further comprising:
providing a second model including the plurality of sequences of states;
determining the likelihood for at least one sequence of the plurality of sequences depending on the transition probabilities between the sub-sequences of states for the skills in the at least one sequence; and
selecting the sequence of states from the second model having a higher likelihood compared to at least one other sequence of states of the plurality of sequences of states.

4. The method according to claim 3 , wherein the providing of the second model includes determining the transition probability from a final state of the first sub-sequence according to the first skill to a plurality of initial states of different second sub-sequences according to different instances of the second skill.

5. The method according to claim 4 , wherein an instance of the first skill and/or an instance of the second skill is defined by a task-parameterized hidden semi Markov model.

6. The method according to claim 1 , further comprising:
mapping a first observation for the machine to a first state, wherein the sequence of states starts at the first state, and/or mapping a second observation to a second state;
wherein the sequence of states ends at the second state.

7. The method according to claim 6 , wherein:
the first observation characterizes the machine or an environment of the machine before executing a first skill in the sequence of skills, and/or
the second observation characterizes the machine or the environment of the machine after executing a last skill in the sequence of skills.

8. A device for operating a machine, the device configured to:
provide a sequence of skills of the machine for executing a task;
select a sequence of states from a plurality of sequences of states, depending on a likelihood, wherein the likelihood is determined depending on a transition probability from a final state of a first sub-sequence of states of the sequence of states for a first skill in the sequence of skills to an initial state of a second sub-sequence of states of the sequence of states for a second skill in the sequence of skills; and
execute the task based on at least on the first skill and the second skill, wherein the task includes physically manipulating an object by the machine, wherein the final state of the first sub-sequence is determined depending on a final component of an instance of the first skill, wherein the initial state of the second sub-sequence is determined depending on an initial component of an instance of the second skill, and wherein the transition probability is determined depending on a divergence between a first Gaussian mixture Model for the final component and a second Gaussian mixture Model for the initial component.

9. A non-transitory computer-readable medium on which is stored a computer program including computer readable instructions for operating a machine, the instructions, when executed by a computer, causing the computer to perform the following steps:
providing a sequence of skills of the machine for executing a task;
selecting a sequence of states from a plurality of sequences of states, depending on a likelihood, wherein the likelihood is determined depending on a transition probability from a final state of a first sub-sequence of states of the sequence of states for a first skill in the sequence of skills to an initial state of a second sub-sequence of states of the sequence of states for a second skill in the sequence of skills; and
executing the task based on at least on the first skill and the second skill, wherein the task includes physically manipulating an object by the machine, wherein the final state of the first sub-sequence is determined depending on a final component of an instance of the first skill, wherein the initial state of the second sub-sequence is determined depending on an initial component of an instance of the second skill, and wherein the transition probability is determined depending on a divergence between a first Gaussian mixture Model for the final component and a second Gaussian mixture Model for the initial component.

Assignees

• Bosch Gmbh Robert

Inventors

• Kupcsik Andras Gabor
• Rozo Leonel
• Guo Meng

Classifications

• G06F9/4498

  Finite state machines · CPC title

• G05B13/042 Primary

  in which a parameter or coefficient is automatically adjusted to optimise the performance · CPC title

• B25J9/1661

  characterised by task planning, object-oriented languages · CPC title

• B25J9/1664 Primary

  characterised by motion, path, trajectory planning · CPC title

• G05B2219/39205

  Markov model · CPC title