The increasing complexity of industrial production systems requires efficient methods for model development in simulation-based planning processes. Conventional approaches to building discrete-event simulation models are often associated with significant effort for data preparation and modeling, which in practice frequently represents a barrier to the effective use of simulation. This paper demonstrates how Process Mining can be used to address these challenges and to support the development of data-driven simulation models in production and logistics. For this purpose, a generic data-driven procedure model is outlined, describing the structured transformation of event logs into simulation-ready models. In an initial prototypical use case from a production context, it is examined to what extent and with what quality data can be extracted from event logs using Process Mining to make them usable for simulation-based planning. Finally, the developed approach is methodically evaluated regarding its applicability and added value in the context of simulation-supported production planning. The results illustrate that Process Mining can reduce modeling effort and improve data quality, provided that an appropriate data foundation is available. Furthermore, the approach opens new potentials for objective and iteratively adaptable process design in dynamic production environments.