The ModIA project has developed a methodology for creating hybrid digital twins designed to improve prescriptive maintenance in industrial production systems. By combining physical simulations with machine learning algorithms, the project established a structured framework to efficiently acquire, process, and analyze vibration data from machinery. Two industrial case studies—an Asyril vibrating hopper and a Starrag drive system—were used to demonstrate the approach, enabling predictions about machine conditions based solely on vibration measurements. This hybrid method facilitates real-time integration into the LYSR analytics platform and highlights potential improvements in predictive maintenance accuracy.

Collaboration with industry partners HID Global, LYSR, Asyril, and Starrag allowed the identification and resolution of critical measurement challenges, particularly in handling high-frequency data acquisition. The project outcomes not only include the successful implementation of operational digital twins for the studied machines but also contributed to enhancing the measurement technologies and analytical tools offered by the involved partners. These results position participating companies to further advance predictive maintenance applications, enhancing reliability and performance in their production environments.