PRISMA

Predictive maintenance for individual robot cells

As part of the project, a demonstrator for a predictive maintenance algorithm for robot cells is to be realised, especially in special machine construction. The core component will be early warning systems and error detection at component level.

Thus, a range of physical models including statistical and machine learning methods (e.g. neural networks) will be adapted to a wide variety of robot cells. Data from different individual cells will be used to create these models at component level.

The project aims to minimise downtimes and maintenance times and maximise the life expectancy of such unique items.