@inproceedings{GalkaMeissner, author = {Galka, Stefan and Meißner, Sebastian}, title = {Einsatz von Digitalen Zwillingen f{\"u}r die operative Steuerung von innerbetrieblichen Transportsystemen}, series = {Simulation in Produktion und Logistik 2025}, booktitle = {Simulation in Produktion und Logistik 2025}, editor = {Schmidt, Thorsten and Rank, Sebastian and K{\"u}hn, Mathias}, publisher = {Technische Universit{\"a}t Dresden, Professur f{\"u}r Technische Logistik}, address = {Dresden}, isbn = {978-3-86780-806-4}, doi = {10.25368/2025.280}, pages = {47-1 -- 47-11}, abstract = {This paper presents the architecture and functionality of a simulation-based Digital Twin designed to support decision-making in the control of internal transport systems. The developed Digital Twin leverages a wide range of data from operational information systems to predict future system loads. Specific methods for forecasting transport orders have been developed. An integrated simulation framework enables analysis of potential control decisions, such as resource allocation or the timing of order releases. The simulation results are used for multi-criteria evaluation of the available options for preventive action, thus, providing users with concrete recommendations to support their operational decision-making processes.}, language = {de} } @inproceedings{GalkaWeigert, author = {Galka, Stefan and Weigert, Alexander}, title = {Petri Net Structures for Automated Model Generation through Process Descriptions}, series = {IFAC Papers OnLine}, volume = {59}, booktitle = {IFAC Papers OnLine}, number = {10}, publisher = {Elsevier}, doi = {10.1016/j.ifacol.2025.09.093}, pages = {542 -- 547}, abstract = {Creating simulation models and running experiments requires expertise and significant effort. Modelling complex production systems by using Petri Nets, which have a simple formalism, leads to considerable modelling effort. To reduce the workload required, fundamental processes within a production system must be analyzed and represented as generic Petri Net structures to enable their automatic generation. These structures can serve as modular building blocks to construct arbitrarily complex systems. The broader objective is to enable the automatic generation of such, based on process descriptions within a simulation framework. Built on these findings, the necessary information for developing a specialized formal process description is derived, enabling the automated generation of Petri Net simulation models for production systems.}, language = {en} } @inproceedings{Galka, author = {Galka, Stefan}, title = {Integration and usage of simulation-based digital twins in internal logistics systems}, series = {IFAC Papers OnLine}, volume = {59}, booktitle = {IFAC Papers OnLine}, number = {10}, publisher = {Elsevier}, doi = {10.1016/j.ifacol.2025.09.152}, pages = {897 -- 902}, abstract = {This paper presents the workflows of Digital Twins in practice, with reference to three case studies. The objective is to present the processing of data and information flow. The most essential requirements for the design of a Digital Twin are synthesized from the case studies and a layer model is presented, which was developed based on the three case studies. The model consists of four functional layers and an overarching data- and workflow-management-backbone. A front-end layer is proposed for communication with the end user.}, language = {en} } @article{GinsterWeigertHeymetal., author = {Ginster, Vinzenz and Weigert, Alexander and Heym, Maximilian Klaus and Beier, Christoph J{\"u}rgen Anton and Schiebahn, Alexander and Galka, Stefan}, title = {Factory simulation of adhesive bonding processes considering pot-life-induced waste}, series = {Journal of Advanced Joining Processes}, volume = {13}, journal = {Journal of Advanced Joining Processes}, publisher = {Elsevier}, issn = {2666-3309}, doi = {10.1016/j.jajp.2026.100375}, pages = {7}, abstract = {The simulation of production systems in the planning phase is an established tool for evaluating system performance under dynamic conditions (Da Silva et al., 2018). When planning a production system that includes time-critical processes such as adhesive bonding, exceeding time constraints (e.g., excessive lead times relative to the adhesive pot-life) can lead to product rejection and scrap. These risks should be analyzed by means of simulation, as it enables the consideration of stochastic effects such as random machine breakdowns, which significantly influence product lead times. This study focuses on the production of adhesively bonded electrolyzer-cells. Due to the cell design, several handling and stacking operations must be carried out between adhesive application and final joining, so that immediate joining is not feasible and pot-life-induced scrap becomes a critical risk factor. We therefore conducted an exploratory investigation that combines experimental pot-life characterization with discrete-event simulation of a planned electrolyzer-cell production system. A two-component epoxy adhesive was characterized using differential scanning calorimetry with kinetic analysis, rotational rheometry, and 90° peel tests on application-representative joints. Based on these tests, an application-specific pot-life was derived and implemented as a time-dependent quality criterion in the discrete-event simulation model. A full-factorial simulation study with varied buffer size and mean time to repair at a station availability of 98 \% was carried out. The results demonstrated a pronounced trade-off between throughput and pot-life-induced waste and identified a buffer capacity of one and short repair times as the most favorable configuration. Extending pot-life due to different processing temperatures substantially reduced scrap, highlighting the benefit of integrating curing kinetics and dynamic simulation in early-stage production system design for time-critical adhesive processes.}, language = {en} }