The development of a Smart Factory Grid results in specific challenges. In particular, the heterogeneity of information sources poses a central problem. The data is often high-dimensional, formatted differently and varies in terms of availability and quality. This makes it considerably more difficult to combine and use this data for status diagnosis and forecasting.

Another obstacle is the lack of a comprehensive methodology for information fusion. While initial solutions already exist for individual aspects, such as the processing of heterogeneous data formats or transfer learning, the systematic integration of these approaches remains an open research question. In addition, the limited amount of available data restricts the use of data-driven models.

In addition, the flexible adaptation of operating strategies requires the optimization of maintenance planning, which can no longer be carried out using conventional methods due to the complexity of production.

The research approach aims to develop a holistic methodology for condition diagnosis and prognosis in a Smart Factory Grid. A central step is the classification of the relevant information sources, such as sensor and operating data, simulation results or physical models. Based on this, problem-related methods for information fusion are developed. The aim is to consistently merge data from different formats and sources and to combine both physical and data-driven approaches.

The research approach aims to address health management in a Smart Factory Grid as a complex optimization problem. Here, the integration of a manufacturing simulation, which has been lacking in the literature to date, is being investigated. Predictive to prescriptive maintenance is intended to enable the automated generation of maintenance proposals in order to make precise and efficient maintenance decisions.