SMARTFACT brings together 17 Slovenian research and industrial partners to develop and validate technologies for more efficient, flexible and sustainable manufacturing. The programme is coordinated by the Jožef Stefan Institute. Within this wider consortium, ViCoS at the University of Ljubljana, Faculty of Computer and Information Science (UL FRI), contributes targeted research in computer vision and machine learning for industrial inspection.

Programme overview

SMARTFACT integrates artificial intelligence, robotics, machine vision, laser processing and plasma technologies. Its activities progress from industrial research at TRL 3–4 to prototype development and demonstrations at TRL 5–6. The programme is organised around four interconnected thematic pillars:

• AI-supported process control,
• robotics of the future,
• laser processing for electronics, and
• plasma technologies.

Expected programme-level results include self-learning systems for monitoring manufacturing processes, energy-efficient production scheduling, automated visual quality inspection, adaptable robotic systems, automatically optimised laser micro- and nanoprocessing, and thermochromic sensors for recording the temperature history of food products.

UL FRI (ViCoS) contribution

The ViCoS team focuses on AI-based visual inspection for automated quality control, particularly where defects are rare, annotations are limited, or inspection conditions change. Planned work includes:

• Adaptable surface-defect detection (in collaboration with Gorenje and TECOS): developing and evaluating fully supervised, weakly supervised, few-shot and one-class unsupervised methods, including approaches that combine defect-free data with only a few defective examples.
• Configuration of machine-vision solutions (in collaboration with SICK): developing novel models for defining configurations of solutions to vision problems within existing software frameworks for machine-vision solutions.
• Inspection for micro- and nanomanufacturing (in collaboration with RETTRO): adapting deep-learning-based defect detection to microscopic structures, using the desired nanostructure geometry to guide inspection, and supporting visual feedback for laser-process optimisation and integration into RETTRO’s CAD/CAM software.

The work is planned for validation with industrial partners in representative use cases, including metallic appliance surfaces, configuration of machine-vision solutions, and optical inspection in micro- and nanomanufacturing.

Funding and duration

SMARTFACT runs from July 2026 to June 2029. It was selected under the public call JR RR PROGRAMI (TRL 3-6). The total programme budget is €14,691,133.33, comprising €8,832,720.83 in ERDF co-financing and €5,858,412.50 from private sources.

Further information

See the official SMARTFACT programme website for consortium-wide information and updates.