Science Minister Mohrs and high-ranking acatech delegation visit IIP-Ecosphere at Hannover Messe

The IIP-Ecosphere project can look back on its successful participation at Hannover Messe 2023. Lower Saxony’s Minister of Science Falko Mohrs and a high-ranking acatech delegation were among the many visitors of the IIP-Ecosphere stand to find out more about the platform demonstrator for AI-based quality assurance.

If production companies rely on optical quality assurance through artificial intelligence and federated learning across machines and locations, new productivity advantages will result. Lower Saxony’s Science Minister Falko Mohrs could see this at the IIP-Ecosphere stand at the Hannover Messe. The AI project, jointly coordinated by the L3S Research Center and the Institute for Manufacturing Engineering and Machine Tools (IFW) at Leibniz Universität Hannover, was an exhibitor at the Lower Saxony stand of the Ministry of Science. The AI experts from IIP-Ecosphere explained the various options of the project’s IIoT platform to the minister using the demonstator for AI-based quality assurance. In conversation with Dr Holger Eichelberger (University of Hildesheim), Dr Claudia Niederée (L3S), Per Schreiber (IFW) and Dr Johannes Winter (L3S), the minister was very interested in the potential of federated learning, which allows data holders to train a common model without physically sharing their data.

The IIP-Ecosphere team also welcomed other high-profile guests to the stand, including Prof. Wolfgang Wahlster, founding director of the German Research Centre for Artificial Intelligence (DFKI), Prof. Thomas Weber, President of acatech, and Prof. Henning Kagermann, Chairman of the Board of Trustees of acatech and, with Wolfgang Wahlster, one of the pioneers of Industrie 4.0. They, too, were given an insight into the options of the IIP-Ecosphere platform which is distinguished from existing Industrie 4.0 platforms by the consistent use of the Industrie 4.0 management shell – from the device via application services to the application. In addition, the low-code-based configuration allows flexible adaptation of the platform to company-specific requirements. The automated generation of applications from the configuration ensures efficient development, robust software, and easy customisation.

Data security through federated learning

The platform demonstrator for quality assurance consists of two robotic arms equipped with cameras. The AI-based visual system detects workpiece defects, such as scratches or shape deviations, at an early stage. New this year: a second robot. Both robots are able to learn from each other without having to exchange data. So-called federated learning means more data security for AI users. Users can also use AI when data is scarce, for example, when errors rarely occur.

The handling of data has so far been a stumbling block for the use of AI. Federated learning enables machine learning across locations and company boundaries, for example, between machine builders and machine users, without having to exchange data. This makes the use of AI more interesting for companies.