Demonstrating Data-to-Knowledge Pipelines for Connecting Production Sites in the World Wide Lab: Data and Software

Gorißen, Leon Michel; Behery, Mohamed Anwar Abdellatif; Schneider, Jan-Niklas; Ziefle, Martina; Brauner, Philipp; Kötter, Ender David; Lakemeyer, Gerhard; Häfner, Constantin Leon; Hinke, Christian Rüdiger; Kaster, Thomas; Lennartz, Moritz; Brecher, Christian; Gries, Thomas; Kastner, Daniel

Demonstrating Data-to-Knowledge Pipelines for Connecting Production Sites in the World Wide Lab: Data and Software

Gorißen, Leon Michel^RWTH* ; Schneider, Jan-Niklas^RWTH* ; Behery, Mohamed Anwar Abdellatif^RWTH* ; Brauner, Philipp^RWTH* ; Lennartz, Moritz^RWTH* ; Kötter, Ender David^RWTH* ; Kaster, Thomas^RWTH* ; Kastner, Daniel^RWTH* ; Hinke, Christian Rüdiger^RWTH* ; Gries, Thomas^RWTH* ; Lakemeyer, Gerhard^RWTH* ; Ziefle, Martina^RWTH* ; Brecher, Christian^RWTH* ; Häfner, Constantin Leon^RWTH*

2025

Online
URL: https://s.fhg.de/gorissen-2025a

Projekte

Inhaltliche Beschreibung (Schlagwörter)
Industry 4.0 (frei) ; Transfer Learning (frei) ; Foundation Models (frei) ; Digital Shadow (frei) ; Digital Twin (frei)

Kurzfassung
This dataset accompanies the publication Demonstrating Data-to-Knowledge Pipelines for Connecting Production Sites in the World Wide Lab. It includes robot trajectory data, trained models, and source code used to develop and validate a foundation model for inverse dynamics control. The dataset captures motion from multiple Franka Emika Panda operating in distinct manufacturing use cases, enabling cross-domain learning and model adaptation. In the project all data is stored in a research data repository with semantic annotations, ensuring findability and interoperability. A release of the raw data is made available here. The foundation model can be fine-tuned for specific applications, supporting efficient robotic control and advancing data-driven manufacturing strategies. As of December 2024 the project is not concluded and future updates might be provided. Links to the datasets and code are provided below with DOIs for accessibility and reuse.

Funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germanys Excellence Strategy – EXC-2023 Internet of Production – 390621612.

Externer link:

Volltext