Project


GratKI
AI-supported deburring of components using a modular deburring machine and reinforcement learning

In metal processing, deburring is often manual and error-prone. Conventional automation fails due to a lack of precision and flexibility for small series. In the "GratKI" project, an AI-supported 5-axis CNC deburring machine with intelligent process control is being developed. BIBA is responsible for the optical quality inspection and uses reinforcement learning to dynamically adjust process parameters. The result is an adaptive, fully automated solution that reduces scrap and guarantees maximum cost-effectiveness and precision for industrial deburring.

Contact persons:
M. Lütjen eMail schickenWebseite betreten (Project manager)

W. Zeitler eMail schickenWebseite betreten

Funded by:
BMWE / IGF

Duration:
16.04.2026 - 15.04.2028

See project's publications
List all projects

nach oben
Printed: 25.05.2026
© 2011 by Universität Bremen, Germany
Source: http://www.ips.biba.uni-bremen.de/projekt.html?&L=1.