Prof. Dr.-Ing. Michael Freitag
Project
|
La2-Mo2
Development of an LLM pipeline for processing and transforming natural language commands into context-dependent robot commands
The assembly industry faces increasing product variety and customization, which demands production flexibility, complicating Cobot automation, especially for SMEs. Despite Cobot market growth, frequent reconfiguration remains a challenge. Natural language programming could simplify this, but current models lack contextual understanding and precision.
The La2-Mo2 project aims to develop a system that uses LLMs for programming Cobots through natural language. By interpreting spoken instructions and converting them into precise robot commands, the system will make Cobot programming more accessible, reducing complexity and increasing flexibility, particularly benefiting SMEs in assembly processes.
Contact persons:
D. Niermann  
Funded by:
BMWK / IGF
Duration:
01.10.2024 - 30.09.2026
See project's publications
List all projects
|
