Christian Bertsch
Bosch Research Project Leader
FMI Project Leader
Modelica Association Board Member
Background: Christian Bertsch studied Mathematics and Physics at Heidelberg University
Sessions
The Functional Mock-up Interface (FMI) standard plays a central role in enabling interoperability between simulation tools, particularly in cyber-physical and embedded system development. This tutorial introduces the core concepts of FMI and provides hands-on experience in exporting, simulating, and co-simulating Functional Mock-up Units (FMUs). Designed for both beginners and intermediate users, the tutorial is divided into three main sessions. The first session presents the motivation and history of FMI, key technical concepts, and an overview of FMI 3.0, followed by live demonstrations of FMU export using Dymola and Simulink. The second session dives into working with FMUs using Python and the FMPy library, including validation, simulation, and integration with Jupyter notebooks. Finally, we will set up co-simulations involving multiple FMUs, highlighting both graphical and command-line workflows, and demonstrating the impact of co-simulation parameters on accuracy and stability.
The tutorial concludes with a joint session summarizing the role of FMI in digital twins and highlighting real-world applications. Throughout the tutorial, attendees will have opportunities to ask questions and follow along with guided examples and shared materials.
The Functional Mock-up Interface (FMI) standard plays a central role in enabling interoperability between simulation tools, particularly in cyber-physical and embedded system development. This tutorial introduces the core concepts of FMI and provides hands-on experience in exporting, simulating, and co-simulating Functional Mock-up Units (FMUs). Designed for both beginners and intermediate users, the tutorial is divided into three main sessions. The first session presents the motivation and history of FMI, key technical concepts, and an overview of FMI 3.0, followed by live demonstrations of FMU export using Dymola and Simulink. The second session dives into working with FMUs using Python and the FMPy library, including validation, simulation, and integration with Jupyter notebooks. Finally, we will set up co-simulations involving multiple FMUs, highlighting both graphical and command-line workflows, and demonstrating the impact of co-simulation parameters on accuracy and stability.
The tutorial concludes with a joint session summarizing the role of FMI in digital twins and highlighting real-world applications. Throughout the tutorial, attendees will have opportunities to ask questions and follow along with guided examples and shared materials.
This session is chaired by:
Over the last fifteen years, the Functional Mock-up Interface (FMI) has evolved by addressing real industrial needs for simulation interoperability. What began as a solution mainly for the automotive industry has now firmly established itself as the de-facto industrial standard for tool-independent model exchange and co-simulation across numerous domains. This keynote will trace the journey of FMI from its initial concepts in version 1.0 to the powerful ecosystem of today. The release of FMI 3.0 marks a pivotal milestone in this journey, enabling major improvements that redefine the boundaries of system simulation. This new version introduces advanced co-simulation capabilities, facilitates the development and deployment of virtual ECUs, and seamlessly integrates FMI models into complex optimization and machine learning workflows. These enhancements empower engineers to tackle new challenges in digitalization and automation with greater efficiency and fidelity. Looking to the road ahead, the FMI standard continues to expand through the development of Layered Standards. These domain-specific extensions, built on top of the core FMI framework, enable further improvements and unlock entirely new use cases for virtual engineering. This talk will provide a comprehensive look at FMI’s past, present, and future, celebrating its role as a cornerstone of modern model-based development.
This session is chaired by:
