PRepAIr

Ziel/Beschreibung

The PRepAIr project is researching data-driven automation of, among other things, robotic surface repair processes for vehicle paint damage (spot repairs) through the use of machine learning models. The aim of the project is to achieve a higher quality of repair decisions with lower start-up costs and to enable scalability through the possibility of adaptation to OEM requirements. However, there are specific challenges when using machine learning models, such as the challenge of label quality and predictability, the consideration of costs and secondary objectives, and the possibility of unwanted data feedback loops that can produce bias in the data. The project is based on the findings of a feasibility study conducted in fall 2022, which has already led to a joint publication at ISR 2023 and a nomination as a finalist for the euRobotics Technology Transfer Award 2023.