UROP Project

Optimization of Robot-based Machining using Machine Learning

Excellence Cluster: Internet of Production

Industrial robots (IRs) are an integral part of production systems for handling and welding applications. Their advantages include flexibility, relatively low investment costs and a large ratio between workspace and installation area. Therefore, IRs also offer potential for more demanding processes such as machining of large components. These processes are characterized by high process forces and small tolerances. IRs, however, show deficits in terms of rigidity and absolute accuracy compared to traditional machine tools, which affect workpiece quality.

Task

In this project, several methods for optimizing robot-based machining using machine learning are researched. Possible topics (depending on the interests / skills of the students) are: data pipelines, data analysis, continual learning, time series prediction, structured neural networks, explainable artificial intelligence and digital twins.

Requirements

Solid programming skills and knowledge about robot dynamics. Highly motivated and enthusiastic about robotics.