The project will be implemented in cooperation with Futtec and the Signal Lab research team, led by Radek Martink from the Faculty of Electrical Engineering and Informatics (FEECS). "The AI-MATTERS project under the Digital Europe Programme focuses on experimental testing and validation of artificial intelligence methods for industrial applications. In this phase, we focus on testing AI-assisted surface defect detection and analysis for mobile devices. Our goal is to increase the accuracy of defect detection over existing methods and enable their effective use in industrial and logistics applications," says Radek Martinek.
The testing involves several key phases. First, exploratory data analysis is performed to assess the quality of the delivered dataset and identify any anomalies. "We then focus on preprocessing the image data, optimizing input parameters for AI models, for example by standardizing formats and reducing noise. The third phase is testing the neural network architecture, where we verify the performance of the models and their adaptation to mobile detection systems," explains Dominik Vilímek from FEECS.

The research team also includes Boris Pustějovský and Ondřej Svoboda, who will test various neural network models and optimize their outputs.
"Our deliverables will include comparative analysis of models, optimization of algorithms and creation of test reports according to AI TEF Manufacturing standards. The project brings major innovations in the field of AI for industrial manufacturing and logistics, where accurate surface defect detection improves the efficiency and safety of operations," adds Robert Šamárek from FEI.
The AI-MATTERS project represents an important step towards the wider implementation of artificial intelligence in industrial practice and opens up new possibilities in the field of automation and optimisation of manufacturing processes.