Invention Title:

ROBOT SYSTEMS, METHODS, CONTROL MODULES, AND COMPUTER PROGRAM PRODUCTS THAT LEVERAGE LARGE LANGUAGE MODELS

Publication number:

US20250100131

Publication date:
Section:

Performing operations; transporting

Class:

B25J9/1602

Inventors:

Applicant:

Smart overview of the Invention

The patent application describes innovative robot control systems and methods that utilize large language models (LLMs) to enhance robotic autonomy. These systems allow robots to understand and execute tasks specified in natural language, enabling them to operate with minimal human intervention. The integration of LLMs facilitates the conversion of natural language queries into actionable task plans, providing a seamless interface for robot operation.

Functionality

Key functionalities include capturing environmental data through sensors and generating natural language descriptions of the environment. The system then uses these descriptions to query an LLM for work objectives, which are expressed in natural language. This process enables the robot to autonomously identify and execute tasks by following detailed task plans generated by the LLM.

Operational Method

The operational method involves several steps: capturing sensor data, generating environment descriptions, querying the LLM for work objectives, and receiving task plans. The robot can execute these plans autonomously or seek operator confirmation before proceeding. This flexibility allows for varying levels of human oversight based on operational needs.

Control Module

The control module comprises processor-executable instructions stored on a non-transitory medium, enabling the robot to perform tasks autonomously. These instructions facilitate interaction with the LLM, allowing the robot to receive and execute multiple work objectives efficiently. Operators can also intervene to select specific task plans for execution, offering a balance between automation and manual control.

Applications and Benefits

By leveraging LLMs, this technology addresses the need for automation across diverse tasks traditionally performed by humans. It offers significant potential in various domains, from household chores to complex industrial operations, enhancing productivity and allowing humans to focus on more strategic activities. The system's adaptability is a key advantage, providing tailored solutions for different operational contexts.