Purpose and goal
For additive manufacturing to fulfill its potential in safety-critical production, quality assurance must be calibrated, integrated and traceable in real time for sensitive production environments. TRUSTAM applies federated learning to additive manufacturing quality assurance – a framework where AI models are improved without raw data ever leaving the place where it was generated. Only model updates are exchanged, enabling shared intelligence while maintaining complete data confidentiality.
Expected effects and result
TRUSTAM will deliver a comprehensive demonstrator covering local model adaptation, federated model enhancement, new applied knowledge of AI and control in safety-critical industrial conditions. This will be the basis for new products and services, enabling the safe implementation of AI-based quality assurance in sensitive industrial sectors. During the project, finely calibrated AI models are developed, fine-tuned for specific machines, application areas and production conditions.
Planned approach and implementation
TRUSTAM is delivered through a collaboration between Interspectral AB (data fusion, analysis, and visualization software), AMEXCI AB (AM services), Saab AB (industrial end-user), and Scaleout Systems AB (federated learning). The work is organized into six modular work packages, with regular online coordination complemented by physical meetings, workshops, and evaluation events throughout the project period. The outcome is a demonstrator presented to both the Swedish and global AM markets.
The project description has been provided by the project members themselves and the text has not been looked at by our editors.
