MAGICIAN, artificial intelligence applied to aesthetic quality

The European project MAGICIAN, an ambitious initiative that corrects aesthetic imperfections in industrial processes of vehicles with artificial intelligence, has just started. It will run until September 2027 within the framework of the Horizon Europe programme, with a budget of 9.1 million euros.

Zabala Innovation’s Brussels office is part of this project, coordinated by the University of Trento and with partners from Italy, Sweden, Belgium, Greece, Germany, the Netherlands and Turkey. In particular, our consultancy is in charge of the management of the cascade funding and the organisation and implementation of all activities related to the success of Magician.

“We have extensive experience in cascade funding with EU projects such as DigiFed, Blockchers, NESOI or EDI, so we will lead the management of two calls in the third and fourth year of the project to deepen the functionalities implemented in the main pilots of the project and extend the applications of the intelligences integrated in the systems towards new use cases”, explains Marina Coloma, consultant at Zabala Innovation and participant in the project.

Increasing quality with MAGICIAN

The increasing attention of consumers to the high aesthetic quality of products means that quality standards in production processes are rising, with the resulting physical strain on workers. MAGICIAN’s challenge is to automate some processes by using robots to detect and touch up production defects before finalising the aesthetics of the vehicles.

MAGICIAN will produce two robotic solutions, one for defect analysis (SR) and one for retouching (CR), which can work together or separately. In both cases they will use artificial intelligence modules trained with machine learning algorithms, always supervised by the workers.

“This project can provide more information on whether and how human-robot collaborative systems can provide better working environments in automotive manufacturing for people of all genders and ages. This knowledge is crucial and can result in better human-robot collaborative assistance systems being developed and introduced to support workers of all genders and ages in manufacturing,” concludes Marina Coloma.