The BaSys überProd project, funded by the German Federal Ministry of Education and Research under the code 01IS20094T, ran from 1 January 2021 to 31 March 2023. It brought together Fraunhofer IESE, Hörmann Automotive St. Wendel GmbH, the Rheinland‑Pfälzische Technische Universität Kaiserslautern‑Landau (RPTU), Levicron GmbH and Sensitec GmbH. Hörmann led the first overall goal of digitalising an existing mechanical press through a Digital Twin, Sensitec headed the second goal of virtual design of passive magnetic sensors, and Levicron was responsible for the third goal of AI‑supported condition monitoring of a high‑frequency spindle. RPTU contributed to all three goals by developing predictive‑maintenance monitoring concepts, AI‑based sensor‑data analysis, computationally efficient sensor models and hidden‑information extraction from sensor signals. Levicron withdrew early in the project, so the third goal was not pursued further.
The technical outcomes of the first goal are centred on the installation of a comprehensive sensor suite on the mechanical press. The system continuously monitors operating parameters and flags any deviations beyond predefined tolerances. By feeding this data into a predictive‑maintenance framework, the project demonstrated that potential production failures caused by unplanned downtime could be avoided. The second goal yielded a simulation tool called “Virtual Magnetic Sensor.” This software dramatically improves the efficiency of magnetic‑sensor design in the manufacturer’s development department. Its sales mode allows the manufacturer to generate an initial sensor concept for a production plant with minimal effort and system knowledge. A key scientific contribution is a mathematical model that predicts the constant eigenmode of gear‑modulated magnetic fields. The model can estimate the offset of a GMR‑based gear sensor within seconds, providing a sufficiently accurate result for practical use. The third goal, which would have focused on extracting hidden information from spindle‑sensor signals using AI, was not realised due to the early exit of Levicron.
Collaboration was organised through a series of regular meetings. Monthly sessions titled “Sensorik AW 2‑4” and “AW 2” involved all partners, while biweekly meetings focused on each specific goal. A separate biweekly management meeting for the BaSys überProd system ensured cross‑company coordination. Early in the project, pandemic‑related restrictions limited in‑person workshops with suppliers such as Siemens, IFM and X‑Sensors, which slowed initial communication. Despite these challenges, the partners maintained a steady exchange of data and expertise, culminating in a final report published on 1 September 2023. The project’s outcomes are publicly available, reflecting its commitment to open dissemination of research findings.