»We couldn’t have done it by ourselves.«
Thomas Mader, Head of Automation and Controls at GEA Group, talks about the leading-edge cluster “It’s OWL”, which is revolutionizing food production processes using intelligent technology. Since 1893, German technology supplier GEA has been building food processing machines for the food industry, and now serves several industries including pharmaceutical, chemical and marine. Three years ago the company’s engineers joined forces with its cluster partner Fraunhofer IEM to develop a system based on machine learning.
Mr. Mader, tell us about the technology you developed?
For the past three years, GEA has been working on what we call an intelligent separator. Centrifugal separators separate solids and liquids by centrifugal force. Together with Fraunhofer IEM, which is also a member of the cluster, we have developed a software system based on machine learning that detects anomalies in the machine’s operation and compensates for these automatically.
The machines are usually monitored by an expert who controls and operates the system and fixes errors as necessary. Traditionally, the engineer is required to monitor the machines regularly, but this demands the full attention of the engineer who could be used elsewhere. This costs money and can result in production delays.
How does the technology work?
The system collects data about the condition of the separator through sensors. If the system detects certain abnormal patterns, it intervenes without the assistance of an engineer. The AI-based system also ensures process security in future. The intelligent separator is currently a first-of-its-kind prototype. We believe there is huge market potential.
How did you collaborate with cluster partner Fraunhofer IEM?
We shared both the coordination of the project and the implementation of the actual technology. We are experts in mechanical engineering but it was Fraunhofer that provided the data-science expertise. They taught us how to analyze datasets and together we created a great model for how to leverage these technologies in our future portfolio.
In the long term, we really have to focus on being more data-driven and hire more data scientists ourselves. And this is where the cluster with its many partners comes into play: we all share our experience and knowledge to tackle the challenges ahead.
Photo: GEA Group