TrustME - Trustworthy AI for Manufacturing & Engineering Processes

The TrustME research project aims to make a significant contribution to the safe and certifiable use of artificial intelligence in aviation production. The focus is on the development of a central AI platform that consistently guarantees quality assurance, data provenance and traceability of AI workflows - in compliance with the EU AI Act and the certification requirements of the European Aviation Safety Agency (EASA).

The project partners consider the entire life cycle of AI applications - from data acquisition to model development and validation through to certification. This end-to-end approach avoids isolated solutions and creates the basis for industrially deployable, trustworthy AI systems.

Central AI platform with integrated quality assurance

As part of TrustME, a reusable platform for managing training data and AI workflows is being developed. These integrated mechanisms for monitoring data quality and data provenance as well as for versioning models and data sets. The platform is being expanded to include modular components for automated validation and verification (V&V), which enable AI models to be checked against legal and regulatory requirements. The system is supplemented by automatic report generation, which documents the entire development process in a traceable manner.

Strength prediction for thermoplastic CFRP welded joints

DLR is contributing its expertise in the field of thermoplastic CFRP welded joints and is building up a complete data chain - from material properties to process data and ultrasonic testing to destructive testing. On this basis, an AI-supported prediction model for weld strength is being developed, which should significantly reduce the physical testing effort. Two measurement campaigns will also make it possible to analyse the effects of retraining on existing models and thus map the complete life cycle of AI applications.

Sustainability through AI

TrustME makes a direct contribution to climate targets and sustainable production processes in aviation. Precise AI-based prediction models reduce material consumption, physical inspections are reduced and energy consumption in production is lowered through optimised processes. The platform also enables the definition and verification of application-specific sustainability criteria, such as expected CO₂ savings in AI-based component optimisation.

Cooperation and promotion

The partners involved are:

• Airbus Operations GmbH
• Airbus Aerostructures GmbH
• Fraunhofer Gesellschaft e.V.
• German Aerospace Centre (DLR) e.V.
• Helmut-Schmidt-University Hamburg
• Technical University Hamburg
• University Augsburg