Name of participant: Lukas Rauh
Project’s name: QUAIL: Qualifiable AI Lifecycle-Management for Production Systems
Project description:
Artificial intelligence holds enormous potential for industrial manufacturing, yet more than half of all AI projects fail to make the transition from prototype to full-scale operation. Heterogeneous IT/OT system landscapes, a lack of automation, regulatory uncertainties and a shortage of expertise are holding back small and medium-sized enterprises in particular, but also large corporations. QUAIL addresses precisely this gap.
QUAIL is developing a comprehensive blueprint for the operationalisation of AI applications in cyber-physical production systems (CPPS). Rather than focusing on a new algorithm, the project aims to develop an integrated software system that connects existing tools and structures, automates processes and documents the entire AI lifecycle, from data collection and model development to ongoing operations.
QUAIL combines three previously isolated components into a single, modular solution:
- MLOps for production: Proven DevOps principles are applied to the specific requirements of industrial manufacturing, including real-time capability, safety and availability
- A Semantic data model: A standards-based database (e.g. based on the Asset Administration Shell) describes all AI artefacts in an interoperable and traceable manner, serving as the backbone for transparency and reusability
- Integrated compliance mechanisms: Regulatory requirements from the EU AI Act, EU Data Act or ISO standards are supported directly within the process through automated documentation and quality assurance
The system is complemented by an LLM-based assistance feature that simplifies the complex tool landscape, enabling users without AI expertise to confidently engage with the system and participate in decision-making processes.
The project involves developing a technology demonstrator comprising a backend, an integration layer for common MLOps tools and a role-based web interface. Iterative validation is carried out in close collaboration with industry partner TRUMPF, under realistic production conditions.
Benefits of the project outcomes
- For SMEs: A low-threshold, compliant introduction to AI usage via the software system
- For large enterprises: A consistent framework for scaling and efficiently managing AI applications across departments
- For Germany as a business location: Strengthening technological sovereignty through open standards rather than proprietary dependencies
Software Campus Partner: Fraunhofer-Verbund IUK-Technologie und TRUMPF
Implementation period: 01.01.2026 – 31.12.2027



































![[KOM,BI]Co-citation-based machine learning to determine promising research projects](https://softwarecampus.de/wp-content/uploads/2022/02/KOMBI-768x768.png)




































































