Conference Opening

Connecting Dyad and Modelica with FMI and AI
Dyad is a new language for acausal modeling and simulation which incorporates many features such as automatic differentiation, dynamic optimization, LLM-based generative AI integration, scientific machine learning (SciML), and embedded code generation. In this talk we will focus on the aspects which connect the Dyad system to the Modelica ecosystem. We will showcase generative AI tooling for high-accuracy translation between Modelica and Dyad, FMI import capabilities as an alternative to bring models in from the Modelica world, and the creation of lean FMUs from Dyad for export into Modelica workflows.

The vendor session of LTX offers a condensed update for recent versions of the following tools:
- Dymola, our workhorse for Modelica development & FMI simulations
- TIL (Thermal systems, heat pumps) and PSL (Process Systems Library for Power-to-X processes)
- DLR libraries (Cables, Visualization, ...)
- ReSim: Our Python test tooling which has been employed for MSL 4.1.0 release tests

System Modeler, together with Wolfram Language, redefines what's possible—combining intuitive, drag-and-drop modeling with the world's most powerful computational engine.

You will discover the latest news and perspectives regarding Dymola and underlying technologies, as well as the portfolio of libraries. Also included: latest standards support and examples of workflows involving native and web clients of the Dassault Systèmes offer.

In this presentation, we will start with presenting the significant enhancements in the classic XRG Modelica Libraries for thermal energy system simulation, relevant use cases and how to deploy results to industrial product development. These libraries are a foundation to next-level surrogate and machine-learning modelling which is enabled by XRG’s newest Modelica library SMArtInt+. The multi-tool library closes the gap between data-driven and physical Modelica modelling with its powerful tool-chain and user assistance which is crucial for real-life product integration.

Think you’ve seen Modelon Impact? Think again. Over the past few years, we’ve taken a major leap forward reimagining the platform to deliver a faster, smarter, and more flexible system modeling experience for Modelica experts.
In this session, we’ll unveil the next-generation features that make Modelon Impact a tool built not just for today, but for the future of simulation. Get an exclusive look at our AI-powered Modelica code editor—a game-changing capability that helps users write and edit custom Modelica models in a fraction of the time.
We’ll also highlight our newest collaboration features designed to help teams work together seamlessly on complex simulation projects, and we’ll share the latest third-party library integrations that open doors to new and advanced applications.
Whether you’re returning to Modelon Impact or seeing what’s new for the first time in a while, this session is your opportunity to experience how we’re redefining what’s possible with Modelica. Join us and see how we’re making flexibility, extensibility, and future-readiness a reality.

ODE: AI-First Systems Engineering Platform for Modelica
ODE introduces the world's first AI-native systems engineering platform that brings Modelica modeling and simulation directly into the browser, eliminating traditional installation and deployment barriers. Our cloud-based platform combines a native Modelica compiler and simulation engine with advanced AI capabilities that understand physics, engineering constraints, and domain-specific modeling patterns to accelerate model development and analysis workflows. Built on proven enterprise-grade infrastructure, ODE seamlessly integrates with our scientific computing platform Paper, enabling engineers to transition from Modelica simulation results to interactive documentation, collaborative analysis, and publication-ready reports within a unified environment. The platform leverages web technology to deliver near-native performance for complex multi-physics simulations while maintaining the collaborative advantages of modern cloud computing. Engineers can develop, simulate, analyze, and share Modelica models through natural language interfaces, automated optimization workflows, and one-click deployment of interactive applications without requiring local software installations. ODE represents a paradigm shift in systems engineering workflows, making advanced Modelica capabilities accessible to broader engineering teams while dramatically reducing the complexity of model sharing, collaboration, and knowledge transfer across organizations.
Keywords: Modelica, AI-assisted modeling, cloud simulation, browser-based engineering, collaborative systems engineering, WebAssembly simulation

Confident product decisions require trust in the underlying simulations. This talk presents how SSP Traceability—based on the System Structure and Parameterization (SSP) standard— creates a digital thread of trust—linking simulation models, processes, requirements, and KPIs. We demonstrate how orchideo | easySSP implements this approach by enabling modular, step-by-step traceable simulation workflows, that document and connect all relevant artifacts, providing the basis for trustworthy, simulation-driven product development.

Cyber-Physical Systems (CPS) simulation provides a unified paradigm for equipment verification. While Modelica excels in multi-physics modeling (e.g., mechanical, electrical, and thermal systems), it lacks native support for signals, communications, control, imaging, data processing, and AI.
To overcome these limitations, MWORKS delivers:
Enhanced Modelica Support – After a decade of R&D and five years of refactoring, MWORKS offers improved Modelica compilation and solver capabilities.
Extended Block-Diagram Modeling – Building on Modelica’s unified framework, MWORKS enables embedded code generation and bidirectional model-code traceability.
Seamless Scientific Computing Integration – By leveraging Julia, MWORKS incorporates toolboxes for signals, control, imaging, data, and AI, achieving tight integration between system modeling and scientific computing for full-spectrum CPS simulation.
MWORKS has been successfully deployed in aerospace, aviation, nuclear, shipbuilding, and automotive industries.

Claytex specialises in modelling and simulation of complex systems and providing consulting services, training and software solutions. We focus on simulation for systems engineering and understanding the interactions of systems with their environment.
Everything we do at Claytex is about modelling and simulation. Systems Engineering is a very broad topic and here at Claytex we focus on the modelling and simulation of the systems to understand their behaviour and dynamics. Our work covers multiple industries and domains such as vehicle dynamics and pwoertrain, energy management of buildings, flight dynamics of drones, thermal management and extends into the virtual testing of ADAS and autonomous vehicles with physics based sensor models. We help our customers understand their designs to make better products, faster.
We develop simulation solutions using the open standards of Modelica, FMI, OpenSCENARIO and OpenDRIVE; distribute systems engineering solutions from Dassault Systemes, rFpro and specialist Modelica library developers. Our portfolio of tools includes Dymola, Reqtify, rFpro as well as our own software products including the VeSyMA suite of solutions. We also build bespoke simulation tools for specialist applications.

MathWorks FMI support overview: FMI import, FMI export, Cross-platform support, FMI 3 support, Etc.

An overview of the latest OpenModelica functionality will be given, including the new backend, OMSimulator for FMI-based simulation and composition, Julia interoperability, CRML support and industrial applications of OpenModelica.

Conference opening by local chair Ulf Christian Müller

AI is increasingly being used in the fast and accurate simulation of physical systems. In this keynote, we will discuss how knowledge of the underlying physics can be explicitly incorporated into AI systems for physical simulations. Key examples will highlight the potential gains as well as the involved challenges.

News by Dirk Zimmer

This session is chaired by

This session is chaired by

This session is chaired by

This session is chaired by

Add your biography.

This paper will present a new unified algorithm for unit checking and inference, and showing the benefits for various libraries.

The Modelica Language supports declaring units for variables using the SI-standard. This allows dimensional checking to detect possible errors in equations. The units for variables make it easier to interpret, input and plot their values. When we infer the unit of a variable we get the same benefits also for variables without a declared unit. We will use unit inference and checking for the combination, even if the check is primarily a dimensional check.

Both dimensional checking and unit inference are already implemented in several Modelica tools, but not consistently. The original motivation for this paper was to understand the different approaches, and demystify the unit handling with the goal of making it more available. Based on that understanding this paper will also present a new unified algorithm combining the different strengths, and showing the results for various libraries.

This paper presents Liaison (Liaison, 2025) a new open- source tool designed to address challenges related to portability, security, and intellectual property when sharing Functional Muckup Units. Built on the Functional Muckup Interface 3.0 standard, Liaison has a server-client architecture that leverages Zenoh for communication (Corsaro et al., 2023. Zenoh is a novel pub/sub/query protocol that supports a variety of network topologies such as peer-to-peer, routed, mesh and brokered communication with minimal configuration, setting Liaison apart from similar tools. An example of a case that led the authors to develop this tool (Software-in-the-Loop testing of sail control systems for wind assisted propulsion of ships (Laursen et al. 2023)) is presented.

In the past decades, a large number of power electronic converter-based energy resources have been connected to grids around the world. The increasing number of such devices is pushing for both electromechanical and electromagnetic dynamic simulations to be part of routine stability studies for transmission planning studies. Hybrid wave-phasor simulations have been proposed in the literature to address this challenge. This paper describes Open-Instance Wave-Phasor Interface (OpenIWPI), a Modelica library used to couple phasor-based and electromagnetic power system models, allowing simulation and linearization of such hybrid models to be performed altogether. Examples are presented to describe how the library can be useful in power system studies.

Large gaps persist between the environmental impact of state-of-the-art solid waste treatment technologies and the net-zero decarbonization goals outlined in the Paris Agreement. This results in a critical opportunity to improve waste management systems by advancing circular economy objectives for material recovery whilst avoiding burden shifting.
Waste management systems are multi-domain, multi-energy and multi-product (“Waste-to-X”): they include logistics, storage units, chemical conversion processes of waste materials into valuable resources such as energy, fuels, chemicals, or recycled materials, and all types of related processing for these commodities. The integration of waste treatment technologies via superstructure optimization has been extensively studied to improve resource valorization, but the drawback of integrated mathematical models is the lack of flexibility for refined process design. The evaluation of site-specific Waste-to-X systems critically requires the inclusion of dynamic models, with the capability to include any type of complexity such as the modeling of heterogeneous flows to track all impurities and compounds across a system, as well as the implementation of control strategies for the operations of the units. The analysis of such systems should handle input data uncertainties and provide transparent performance indicators computation to increase stakeholders’ confidence in the results.
In this contribution, we present specific tool requirements to assess waste management systems, and introduce a corresponding design of a simulation and optimization framework. The tool (“platform”) relies on OpenModelica for waste treatment process modeling and simulation, and on Python interfaces for data management, process and system co-optimization, and post-processing workflows to provide key performance indicators. We show how the platform structure allows to account simultaneously for the three pillars of sustainability (economic, environmental and social) for decision-making in an industrial R&D context. We provide insights on modeling conventions and emphasize their importance when establishing the Life-Cycle Inventory, for instance when defining heterogeneous streams (which can be environmental emissions, exchanged commodities or waste streams). We present how the Modelica-Python interface is built, showing how input data and modeling documentation are handled systematically to increase transparent evaluation of Waste-to-X systems.
We illustrate platform features (“workflows”) such as pinch analysis for process heat integration, Life-Cycle Impact Assessment (LCIA), and cost breakdown analysis, on municipal solid waste treatment systems including sorting technologies, storage tanks, waste incineration, and post-combustion carbon capture.

Hybrid modeling – the combination of first-principle models and machine learning – offers the potential to increase model accuracy while reducing modeling effort. Although approaches for creating hybrid models from system simulation models exist, the unique characteristics of Modelica-based, object-oriented models – such as modularity and reusability – can, as of today, not be utilized. In this contribution, we explore approaches for bridging this gap to enable the use of hybrid models with Modelica. Key challenges of architecture definition, training environment and reintegration of the trained machine learning parts into a Modelica model are addressed. To illustrate our approach, we present a case study involving a SCARA robot. This example demonstrates a partially integrated workflow for hybrid modeling, intended to serve as a foundation and motivation for further research.

The Functional Mock-up Interface (FMI) standard is a flagship in the co-simulation and model exchange domain. However, the integration of graph-based computational models—particularly neural networks—into Functional Mock-up Units (FMUs) has remained a technical challenge due to interoperability and platform-specific limitations. To address this, we propose ONNX2FMU, a command-line Python tool that facilitates the deployment of Open Neural Network Exchange (ONNX) models into FMUs. According to FMI's good practices, ONNX2FMU generates C source code to wrap ONNX models in Functional Mockup Units, supports FMI versions 2.0 and 3.0, and provides multi-platform compilation capabilities. The tool simplifies the mapping process between model description and ONNX model inputs and outputs via JSON files, ensuring accessibility and flexibility. This paper presents the tool architecture and methodology and showcases its applicability through illustrative examples, including a reduced-order model powered by a recurrent neural network.

We propose a novel approach for training Physics-enhanced Neural ODEs (PeN-ODEs) by expressing the training process as a dynamic optimization problem. The full model, including neural components, is discretized using a high-order implicit Runge-Kutta method with flipped Legendre-Gauss-Radau points, resulting in a large-scale nonlinear program (NLP) efficiently solved by state-of-the-art NLP solvers such as Ipopt. This formulation enables simultaneous optimization of network parameters and state trajectories, addressing key limitations of ODE solver-based training in terms of stability, runtime, and accuracy. Extending on a recent direct collocation-based method for Neural ODEs, we generalize to PeN-ODEs, incorporate physical constraints, and present a custom, parallelized, open-source implementation. Benchmarks on a Quarter Vehicle Model and a Van-der-Pol oscillator demonstrate superior accuracy, speed, generalization with smaller networks compared to other training techniques. We also outline a planned integration into OpenModelica to enable accessible training of Neural DAEs.

Power systems undergoing large-scale renewable deployment require accurate dynamic models of inverter-based technologies such as solar photovoltaic plants, wind turbine generators, and battery energy storage systems to perform the require studies that would allow to assess and maintain power grid stability. The second generation of the Western Electricity Coordinating Council (WECC) generic renewable energy system (RES) models provides a general framework for modeling inverter-based resources in power system dynamics and stability studies. In this paper, we expand the existing Modelica-compliant open-source OpenIPSL.Electrical.Renewables package by implementing the Renewable Energy Generator/Converter B (REGC_B) model along with the Renewable Energy Electrical Controller (REEC\D) model, according to the second-generation WECC RES framework. Using Modelica's object-oriented features, the implementation emphasizes modularity and reusability in building scalable power system models using the OpenIPSL library. This work highlights the potential of using Modelica and OpenIPSL to support a standardized, scalable development of inverter-based RES models within the WECC framework, and extents the only fully Modelica-compliant open-source package that implements these models.

There are many situations in which a Modelica model needs to handle quantities which are not expressed in the often preferred unscaled SI units. Applying correct unit conversions is extremely important in such situations, and the risk of human error needs to be mitigated using unit-aware technology. Considering the power of unit checking mechanisms in several Modelica tools today, one can be surprised that unit conversion in Modelica still needs to be performed using error-prone user-written formulas and functions. It is demonstrated how automatic and implicit unit conversion can be introduced in Modelica, and that this can be done safely. The benefits of this approach are illustrated in a variety of examples and applications.

Operator training simulators are one of the cornerstones of Model-Based Systems Engineering and a key technology to enable cost-efficient, fast, safe and customizable training. Such simulators often require real-time, large-scale multiphysics models tightly integrated with Decentralized Control Systems (DCS), which poses challenges in both process modeling and communication setup. This paper identifies two major challenges: ensuring simulation efficiency by balancing model fidelity and computational cost, and achieving modeling efficiency by streamlining communication between the process model and the DCS. We highlight the complexity of managing thousands of signals and metadata, and demonstrate how communication architectures aligned with DCS templates can reduce engineering workload and support project scalability.

Stray flux tubes around cylindrical poles are commonly modelled starting from the results for planar flux tubes using the circumference of the cylinder as depth. While this is a tried and tested approach, we here discuss analytical expressions using the actual axisymmetric geometry of a fraction of a hollow torus and compare their results to those of the accepted approach.

We present two technologies for speeding up co-simulations under the FMI standards. By smoothing the input signals inside each FMU, the internal integrator may avoid re-initialization. This can significantly reduce the number of model and Jacobian evaluations. To further help the integrator we also propose a predictor compensation technique tailored to the input smoother. The main benefit of our technologies is the ease-of-use, requiring no model manipulations, nor any special co-simulation master algorithms. The technologies are implemented in Dymola~2025x and validated with both an academic mechanical model as well as thermo-fluid examples where we can observe performance gains with factor up to 100, and often around 5-10. One of these thermo-fluid examples is used in the \emph{OpenSCALING} research project to generate training data for constructing surrogate models, for which the input smoothing is especially important to speed up the dataset creation.

Modelica models exhibit excellent cross-platform compatibility (they can be compiled and simulated on any platform supporting Modelica). However, experiments have revealed that simulation results of the same Modelica model may vary across different platforms (under identical simulation algorithm configurations). The root cause of such discrepancies lies in modeling uncertainties introduced by improper modeling practices, such as insufficient initial constraints or ambiguous state variables selection. Different Modelica platforms employ unique translation strategies when handling model uncertainties. Therefore, consistent model translation must be ensured to achieve aligned simulation results. Model Disambiguation can be done in two ways: by improvements in language, such as propose relevant annotations; by improvements in vendor tools. This paper presents a model disambiguation technology in MWORKS.Sysplorer that enables modelers to automatically correct model text based on translation information, eliminating uncertainties and ensuring model portability across Modelica platforms.

This paper presents the current status of the open-source Modelica library SMArtInt (Simple Modelica Artificial Intelligence Interface), which offers users a straightforward and efficient approach of integrating artificial intelligence via neural networks directly into Modelica models. We provide a detailed overview of the library’s features, area of application, and development process. The primary focus is on the diverse use cases of SMArtInt. The new version 0.5.1 brings an exciting feature with the support of the ONNX format. ONNX (Open Neural Network Exchange) is an open, cross-platform format for the representation and exchange of deep learning models. This means that in addition to TensorFlow models (TFLite), many other models, such as PyTorch can now also run in SMArtInt via the ONNX format.

The first part of this work highlights the development of a custom system cost of hydrogen calculation that was applied as a model optimization target while accounting for total system costs, as well as hydrogen production. This custom estimation was incorporated into Modelon Impact via customized Modelica language & economy components to optimize the capacity, balance of plant, and operation of hybrid energy systems. Results of this optimization are compared to total cost of ownership (TCO) optimization for a particular case study including photovoltaics, battery energy storage, electrolysis, and hydrogen storage components, for various sensitivities. The second part of this work describes the development of a custom Proton Exchange Membrane (PEM) electrolyzer as well as a custom Alkaline electrolyzer in order to better inform hardware selection for hybrid energy system applications. These components, developed in Modelon Impact, are able to capture transient behavior of PEM and Alkaline electrolyzer technologies, such as minimum electricity limits and ramp rates. With this added functionality, a detailed comparative analysis is currently underway, investigating the operational profiles, as well as hydrogen production, of a PEM electrolyzer system versus an Alkaline system for various hybrid energy system configurations. As a preliminary investigation, 5-minute level PV data was utilized within a PV + electrolysis configuration, in order to capture the generation variability necessary to result in observed differences between a PEM and Alkaline electrolyzer. Preliminary differences in the hydrogen production and operational responses for each configuration are highlighted as part of the industrial user presentation.

LookAhead is a lightweight algorithm that improves event handling in co-simulation by predicting events and adjusting the communication step size beforehand. It operates without requiring subsystem event handling capabilities. This paper compares LookAhead with other event handling methods, namely Rollback and Early Return, from the perspective of performance and applicability. Results from the presented example show that LookAhead performs on par with iterative co-simulation methods and is particularly well-suited for handling shared state events.

The paper presents the use of data reconciliation to better characterize the operating state of experimental installations in an industrial context. The paper focuses on the development of a data reconciliation approach using the OpenModelica prototype to study first the detection of the defects in an experimental hydraulic test loop, but also the characterization of good measurement data in an HVAC testing facility. Data reconciliation is shown to be effective for the most pronounced defects intentionally introduced in the system. Regarding the characterization of measurements, data reconciliation identified two initially unnoticed invalid experiments.

The commercial building engineering design world requires analysis tools that can be learned quickly while still delivering accurate and actionable results. This typically means applying rigid, application specific tools with limited system analysis flexibility. In contrast, Modelica offers a multi-domain and flexible modeling environment for building mechanical systems, but its learning curve can be daunting for new users, especially in the commercial building space where engineers juggle multiple project deadlines with short design windows. This may leave engineers wondering where to start when embarking on their first modelica modeling project. This case study reviews the process and lessons learned from a mechanical engineer’s first modelica-based plant energy model. The journey starts with selection of a modelica platform and continues through the project deliverable that evaluates the energy use of a chilled water plant, covering modelica system development complete with system controls and equipment performance matching the constructed project.

A generic model of a master controller concept is presented to demonstrate the advantages of aggregating multiple flexible electrical power units. The master controller consists of various submodels with different features. A Power Balancing Controller ensures that the operation does not cause any imbalance with the traded power. A waterfall method that ensures that any ancillary service activations are handled using the appropriate units to obtain cheapest operation. Simulations of the concept include examples of step responses on aFRR activations, as well as actual aFRR data from the Danish TSO. The results suggest that if the system consists of slow units, the aggregation of units can increase the capacity offered to faster types of ancillary services. Additionally, by allowing certain units to shut down automatically more capacity can be offered as the units are no longer bound by minimum loads.

This paper introduces a transformer-based generative network for rapid parameter estimation of Modelica building models using simulation data from a Functional Mock-up Unit (FMU). Utilizing the \texttt{MixedAirCO2} model from the Modelica Buildings library, we simulate a single-zone mixed-air volume with detailed thermal and \cotwo dynamics. By varying eight physical parameters and randomizing occupancy profiles across 100 simulated systems, we generate a comprehensive dataset. The transformer encoder, informed by room temperature and \cotwo concentration outputs, predicts the underlying physical parameters with high accuracy and without re-tuning (hence, ``zero-shot''). This approach eliminates the need for iterative optimization or can be used to warm-start such optimization-based approaches, enabling real-time control, monitoring, and fault detection in FMU-based workflows.

This session is chaired by

This session is chaired by

This session is chaired by

This session is chaired by

This session is chaired by

Designing complex systems like aircraft is increasingly challenging due to the growing number of systems and stakeholders involved. These systems comprise numerous interconnected subsystems developed by different teams, requiring designs that meet high-level requirements while managing constraints to ensure safety, performance, and compliance. This paper presents findings from various projects focused originally on enhancing simulation capabilities for an aircraft systems developer and integrator. The project's scope then expanded to improving collaborative design practices for greater efficiency and agility. The goal is also to better align systems engineering and simulation activities, ensuring continuity and improving overall effectiveness. The developed processes and tools presented in this paper address real-world complexity and are designed for industrial use, including robust configuration management and data protection. Built on open standards (SysML, Modelica, FMI etc.), they try to integrate smoothly into industrial platforms. The current prototype reflects years of adaptation to constraints and evolving practices at Dassault Aviation and among its design partners.

Free open-source Modelica library called Chemical 2.0 (https://github.com/MarekMatejak/Chemical) provides expressions between chemical substances and processes. These robust and unified definitions allow users to choose whether define processes or substances in their dynamic (electro-)chemical models. Propagation of substance definition and chemical solution through connected components simplify configuration. Chemical pathways can start even with unknown substances. Chemical kinetics was rewritten. The possibilities and performance of chemical pathways modeling are increased using a new type of connectors based on inertial electro-chemical potential. Chemical processes can be directly connected without need to add unsignificant states. Parameterization of chemical reactions is also streamlined, e.g. using forward rate and dissociation coefficient.

The development of Cyber-Physical Systems, particularly in the field of autonomous mobile robotics, often relies on proprietary environments, which limit flexibility and interoperability. This paper proposes a modular and open-source methodology that enables the modeling and simulation of such systems using non-proprietary tools. The methodology integrates Functional Mock-up Interface standard for model exchange, Open Neural Network Exchange standard for Convolutional Neuronal Networks algorithms integration, ROS2 as robotic middleware, and Gazebo as the simulation environment. To validate the approach, we applied it in the development of a mobile robot that navigates autonomously by following traffic signals. This implementation demonstrates that these open technologies can be effectively combined, overcoming common integration barriers among proprietary tools. The proposed workflow provides a practical alternative to proprietary solutions and demonstrates the feasibility of integrating open standards for the development of autonomous robotic systems.

This publication presents models for dynamic, 3-dimensional simulation of the thermal runaway, thermal propagation and the electro-thermal behaviour of battery cells. An optional and replaceable electrical equivalent circuit diagram describes the electrical behaviour of the battery cell. Coupled with the 3-dimensional, thermal equivalent circuit diagram, consisting of a resistance-capacitance network, the thermo-electric behaviour is represented. A special feature of the battery cell model is the optional, replaceable and physical model of the thermal runaway based on Arrhenius equations and the associated mass loss. This means that either the thermo-electric or the thermal runaway behaviour or both can be simulated in combination if required. The anisotropic thermal conductivity of the battery cell and the modelling of relevant heat conduction paths, such as the path across the housing of the battery cell, is made possible by 3D dis-cretization of the battery cell.

Uncertainty Quantification (UQ) studies allow us to determine whether a model is fit for a particular purpose, as well as the operational domain in which it can be used. Standardising the UQ analysis setup and result summary enables the iterative composition of UQ information, which is a crucial step in evaluating model credibility. In this paper, we present an initial attempt to specify UQ information as a cross-layer standard for Modelica-, FMI-, and SSP-based workflows subject to two essential restrictions: (a) uncertainties can only be described in terms of parameters, and (b) analysis is limited to forward uncertainty propagation and sensitivity analysis of nonlinear models. More analysis features are planned for the future. The approach is illustrated using both a simple example and an industrial use case.

This work introduces the first validated, user-friendly, and accurate open-source photovoltaic-thermal (PVT) collector model in Modelica, tailored for system-level simulation and optimization. Current state-of-the-art PVT collector Modelica models are largely limited to oversimplified, steady-state representations that fail to capture the dynamic thermal behavior inherent to real PVT systems. A comprehensive Modelica model is developed based on the ISO 9806 standard (test method for the quasi-dynamic thermal performance of solar thermal collectors), coupled with an electrical system model through an internal heat transfer coefficient. The model calibration relies exclusively on manufacturer datasheet parameters, thereby eliminating the need for parameter estimation from measurement data. The model is validated using experimental results from an unglazed PVT collector, demonstrating strong agreement for various (weather) conditions. The findings highlight that, while steady-state models may suffice for conventional solar thermal collectors (STCs), accurate PVT modeling necessitates a dynamic approach, particularly for the thermal aspects. The electrical output of the PVT collector is less sensitive to transient effects. In addition to the model formulation and validation, this work presents a user-friendly automated calibration method based on manufacturer data, and critically addresses the limitations and potential trade-offs of using exclusively datasheet-derived parameters, thereby providing a transparent tool for PVT system simulation, design, and optimization within the open-source IDEAS library.

The Functional Mock-up Interface (FMI) is the standard for exchanging industrial simulation models in a variety of different applications. Although sensitivity analysis for continuously differentiable systems is directly supported by the standard, for systems with state discontinuities, it is only possible to determine correct sensitivities to a limited extent. In this position paper, we investigate how sensitivity analysis for discontinuous Functional Mock-up Units (FMUs), i.e. including state and time events, works in theory and which additional steps are required to obtain correct results in practice. We further investigate that these steps are unnecessarily computationally intensive from a mathematical point of view, but cannot be implemented in a more efficient way under the current restrictions of the standard. We therefore make a concrete proposal for the new layered standard sensitivity analysis (LS-SA) that remedies the current deficits of FMI in the sensitivity analysis of discontinuous systems. In this way, LS-SA opens FMI towards a variety of next-level applications — including (scientific) machine learning and optimal control — by providing fully differentiable FMUs under high computational performance.

This work describes a novel hybrid reinforcement learning enhanced trajectory planning algorithm for an active debris removal scenario for combined control of a satellite with a 7-axis robotic arm. A reinforcement learning algorithm is combined with a correction algorithm and classical trajectory planning to handle the collision free approach of a chaser satellite to a target, and placing the gripper at the robots near the grasping point for use with a combined controller, which commands the satellite and its robotic arm simultaneously. The algorithm is verified using a complex simulation scenario study implemented in Modelica/FMI.

This presentation provides an overview of Saab Aeronautics’ perspective and involvement in the ITEA4 project Open Standards for SCALable Virtual Engi- neerING and Operation (OpenSCALING). The project explores a range of re- search topics, from hybrid modeling techniques—integrating physics-based and data-driven approaches—to the development of traceable, model-based engi- neering processes. The focus of this presentation is on the processes, methods, and tools required to establish credible models and simulations. This includes both Modeling & Simulation (M&S) operations and the numerical aspects of comprehensive credibility assessments. A central tenet of the project is that stakeholders should communicate digitally using stand-alone simulation arti- facts that are iteratively refined and applicable throughout all phases of the system life cycle. In essence, models and simulators should encapsulate infor- mation about their own credibility. Key features of relevant Modelica Associ- ation standards will be highlighted, including specific elements of the recently released SSP 2.0 and the emerging format for exchanging numerical credibility data developed within the project. Finally, the use of Modelica standards at Saab Aeronautics will be illustrated—ranging from early subsystem concept de- sign to full-scale aircraft simulators used during operational and maintenance phases.

Mechanistic modeling of drug behavior and response is essential for rational drug development and personalized therapy, yet constructing, maintaining, reusing and customizing complex pharmacokinetic–pharmacodynamic and physiologically based pharmacokinetic models can be error-prone when implemented solely via equations or code. We introduce Pharmacolibrary, a free Modelica library offering standardized acausal components for pharmacokinetics, pharmacodynamics, toxicokinetics/toxicodynamics and pharmacogenomics from compartmental and physiologically based templates to effect models and genotype–phenotype records—to simplify model reuse, customization, and interoperability. Its utility is showcased with gentamicin, midazolam, and fentanyl case studies, including pharmacogenomics-driven clearance adjustments and pharmacodynamics simulations.

Creating correct Modelica models and packages from templates and data files describing a multi-physics system is useful in numerous situations. For example, in laying out power plants, designing airplane air conditioning or chemical reactions. This publication shows two possibilities of doing this with an example from robotics design and simulation. The URDFModelica library contains templates for links, joints and whole robots that can be mobile or stationary. The library also has a Python script that takes a valid URDF file as input. With minimal manual processing of this input file, a complete robot simulation package is created automatically. Alternatively, the input file, translated to a Modelica record, can be used to set the parameters and connections of a generic robot model. The URDFModelica library has already been successfully used for quick generation of first Modelica simulations of existing robots or robots yet to be fully developed. The general structure and approach can be adapted to other application domains without much effort. It is planned to release URDFModelica as open source Modelica library.

In response to staff shortages in hospitals, healthcare providers aim at increasingly automating their systems. Defining automated system paths – with collisions avoidance – is a critical step towards automation. In this paper, three different approaches for path planning are investigated: a static, a dynamic and a hybrid approach. The hybrid approach, that sequentially combines part of the static and dynamic approaches, results in improved accuracy against the static approach and improved performance against the dynamic approach.

This paper introduces the FMI 3.0 Layered Standard for Network Communication (FMI-LS-BUS), an extension of the Functional Mock-up Interface 3.0 (FMI 3.0) standard designed to address interoperability challenges in simulating distributed, networked systems, particularly in automotive applications. By leveraging FMI 3.0 features such as clocks, clocked variables, and hierarchical terminals, the standard defines two complementary abstraction layers: Physical Signal Abstraction (High-Cut): Representing physical signal values as clocked variables. Network Abstraction (Low-Cut): Emulates hardware-level bus protocols (e.g., CAN, Ethernet) using FMI 3.0’s clocked binary variables. Aligning with the V-model development process, we demonstrate how these layers address distinct challenges in different design phases: High-Cut supports require- ments engineering and functional testing by simplifying signal exchange during Virtual Electronic Control Unit (vECU) integration. Low-Cut enables later phases of the design validation by replicating network timing and protocol specific properties, such as error handling. The standard’s applicability currently focuses on automotive use cases (e.g., CAN, CAN FD, CAN XL, Ethernet, FlexRay, LIN) but can be extended to industrial au- tomation and IoT, facilitated by its domain-agnostic structure.

The objective of this communication is to present to the FMI community the importance for Airbus to be close to the FMI Standard. As it can be used in a wide range of our Engineering Activities (as exposed later in this short paper), it is crucial for us to share to the all community how we are using today and to expose how we want to use it in the near future to improve our way of working.

Aquifer Thermal Energy Storages (ATES) as long-term storages have a strong potential to address the seasonal discrepancy of supply and demand of thermal energy. The operation of high-temperature ATES (HT-ATES) and their integration into district heating systems are the subject of current research projects. Buoyancy plays an important role in determining how HT-ATES performs. The target of this paper is to present a system model in Modelica that takes these buoyancy effects into account. The validation with experimental data and numerical simulations shows that the system model represents the buoyancy effects well. A sensitivity analysis underlines the importance of optimizing the grid structure and shows that a high resolution in the aquifer is necessary, especially in the vertical direction. Finally, a 10-year simulation shows the deviation of the heat recovery factor, i.e. the ratio of the amount of heat extracted to the amount of heat injected, between a model with and without buoyancy effects.

This session is chaired by

This session is chaired by

This session is chaired by

This session is chaired by

This session is chaired by

In this paper, a new modeling approach combining native Simcenter Amesim submodels and Modelica submodels is presented. FMI terminals are used to enable the physical connection of fluids between Simcenter Amesim and Modelica. The innovation lies in the fluid properties that are computed in the causal and acausal worlds using the same technology. The architecture of a new Modelica AmeTpfMedia library developed to ensure continuity of fluid properties is presented, along with its accuracy and performance. Using this new approach, a demonstrator of a closed-loop heterogeneous cooling system for battery thermal management is built, opening the door to a new way of thinking complex multi-physics systems.

The automotive industry faces two critical challenges in product development: (1) Consistency and traceability across all involved domains / disciplines and throughout the whole development process to manage the complex interrelationships among architecture, components, features, and functions. (2) The pressure to increase development speed and efficiency to stay competitive. There are solutions addressing both challenges: Model-based systems engineering (MBSE) addresses consistency and traceability, while virtualization and simulation address development speed and efficiency. In state-of-the-art industrial practice, however, these two solutions are still applied in isolation from one another, leading to inconsistencies, missing traceability and high effort and resource investment (due to redundant or manual modeling work, finding and correcting inconsistencies etc.). To address these industry challenges, the prostep ivip SmartSE project—an established multi-year cross-industry initiative—has been working on integrating MBSE with simulation to enhance traceability, consistency and efficiency in product development processes. We will present our industry experiences and the obtained results from the SmartSE project, where our companies collaborate on solving the real-world challenges that drive the urgent need for a standards-based integration of MBSE and simulation. Related standardization efforts include the "Simulation Credibility Standard and Recommendation" [1] from SmartSE, the SysML specification for creating MBSE system architecture models [2], and the SSP Traceability Standard [3]. The unique value of this presentation lies in bringing together industry perspectives from OEMs, suppliers and tool vendors to share their experiences, challenges and targets with the integration of MBSE and simulation. Rather than only presenting theoretical solutions, we will also show working proof-of-concepts created within the SmartSE project. The proposed solutions are based on standards (SysML, XMI, SSP, SSP-Traceability, FMI) and thus applicable across industries, irrespective of the specific tool set used in an organization. Our presentation will reveal how the integration of MBSE and simulation enables consistency and traceability between quantitative (i.e., system architecture model) and qualitative (i.e., simulation models) models, whilst increasing development speed and efficiency due to auto-generation of artifacts. We provide industry insights into why the integration of MBSE and simulation is an essential means for traceability, consistency and development speed, and how it can be put into practice.

References [1] White Paper: SmartSE - Guard Rails for "Simulation Credibility Standards and Recommendation", prostep ivip Association. (2024) [2] OMG Systems Modeling Language (OMG SysML), Object Management Group (2019) SYSML SPEC. Available at https://www.omg.org/spec/SysML/1.6/About-SysML/ [3] Prototype Standard: SSP Traceability Standard (An SSP layered Standard). (2024, in development). Available at: https://github.com/modelica/ssp-ls-traceability

Today's process for designing, specifying, installing and testing building HVAC control is not digitalized, leading to expensive manual workflows and missed operational performance. To digitalize the design-build-operate process for building HVAC control, the authors developed a process, associated tools and initiated the voluntary ASHRAE Standard 231P. This paper describes this process and tools, which are both based on Modelica. Standardization through a proposed voluntary ASHRAE Standard and the existing Modelica Language Specification provides a robust technology foundation for industry investment. It is based on declarative specification of the control logic, and allows reuse of existing technologies for open-loop and closed-loop control testing through coupling with an HVAC system or a whole building energy model. It supports control testing using MIL, SIL and HIL, and export of digital twins for operational support. The process and ASHRAE Standard 231P have been designed to accommodate existing Building Automation Systems product lines, while also enabling direct code generation such as by using FMI or eFMI. Control deployment can be digital or manual and conformance to the digital specification can be tested formally and programmatically at each step of the control delivery.

A digital twin of the overall aircraft Environmental Control System is being developed as part of TheMa4HERA, a large European Research Initiative. It shall support verification and validation by virtually demonstrating the behavior of the complete system in various conditions. To this end, also a prototypical control scheme needs to be developed so that a dynamic simulation through complete flight missions is enabled. The prototypical control scheme is tuned using a simplified version of the Digital Twin which focuses on robustness and fast computation time, while making it robust enough to be stable when used with the high-fidelity Digital Twin. An already stable and working controller for simulating the detailed Digital Twin, provides a significant gain of time, allowing for immediate preliminary results on both steady-state and transient system behaviors. This paper describes the base model of the Digital Twin and the methodology used to design the prototypical control architecture.

TAeZoSysPro is a Modelica library developed for thermal and fluid dynamics simulations in industrial applications. The library provides components and models for heat transfer, fluid flow, and mass-energy release calculations. It includes modules for simple thermal modeling, fluid dynamics phenomena, specialized media, and partial differential equations. This paper presents the motivations behind developing this specific library and demonstrates its capabilities through industrial use cases, particularly for applications involving ventilation systems, equipment thermal behavior, and accident scenarios.

This papers reports on activities in the European project MOTIONAL that aims at the development of a digital twin environment which facilities the modularity, interoperability and composability of complex digital twin assemblies of railway systems. The approach that refers to the Functional Mock-up Interface is justified by a discussion of the comparable activities in industry and in the automotive field compared to particularities in the railway system. The work was initiated by the selection and analysis of nine use cases. An introductory digital twin example illustrates the current implementation status and related aspects, while an outlook presents the integration into the Federated Rail Data Space as the business case and as a vision of the activity.

This paper presents the CDL-PLC translator, a tool to convert control sequences for building heating, ventilation, and air conditioning (HVAC) systems expressed in a subset of Modelica called Control Description Language (CDL) which is undergoing standardization via the voluntary ASHRAE Standard 231P into the IEC 61131-10 XML format. Such a translation is a step in a model-based design workflow and contributes to digitalizing the control delivery process in the building sector. The translation into the IEC 61131-10 XML is the last step in the implementation of control logic developed in CDL on programmable logic controllers (PLCs) standardized in IEC 61131. The paper presents the details of the translator, an example for the validation of a CDL block vs. the corresponding IEC block, and a practical application example for the translation of a control sequence for a hybrid heat pump plant from a case study neighborhood in Belgium. Practical applications, use cases, and future developments are discussed in the context of building industry processes.

The auto-cascade refrigeration cycle offers higher reliability and lower manufacturing costs compared to the cascade refrigeration cycle, making it particularly attractive for applications involving significant temperature differences between the heat source and sink. However, this advantage comes with increased complexity in its operation. This article aims to develop a media package in Modelica to simulate auto-cascade refrigeration cycles, with the goal of enhancing understanding of their operation and control. The method is demonstrated using an R23/R134a mixture, a commonly used refrigerant combination in such applications. The media package is compatible with the Modelica Standard Library and is constructed using curve fits based on the REFPROP dataset. Unlike typical refrigeration systems, the media package for auto-cascade components includes the independent mixture composition (Xi) as an additional variable in function calls. A combination of polynomial and Chebyshev polynomial curve-fits for the refrigerant properties has been shown to provide an optimal balance between accuracy and computational efficiency. The article presents example simulations performed to demonstrate use of the media package with component models from the Modelica.Fluid package. A lumped model for the phase separator is developed and simulated with the R23/R134a media to demonstrate ideal phase separation.

New York State (NYS) faces significant challenges in meeting the Climate Act’s bold goals of 70% renewable energy generation by 2030 and total decarbonization of the electric grid by 2040. Extensive simulations are required to assess the impact of numerous inverter-based resources (IBRs) deployed to the large-scale NYS power grid, aiming to evaluate their dynamic behavior and mitigate any negative interactions with their control schemes. However, the modeling efforts required are huge and the computational burden of large-scale simulations is extensive, and often limited by the capabilities of domain- specific tools. This work addresses these limitations by developing a Functional Mockup Unit (FMU) of Grid- Forming (GFM) Inverters for IBR control and integrating them with an electromechanical phasor-domain power system solver. The proposed FMU facilitates the simulation and parametric studies needed to analyze large- scale IBR usage with significantly improved manual modeling and computational efforts. The paper details the process of developing and FMU model for GFM IBRs, including all relevant control loops implemented in the Modelica language and FMU integrated in OPAL-RT’s ePHASORSIM software. Our FMU models are used to successfully deploy and study the impacts of up to 6, 200+ MVA from IBRs on the 5000-bus NYS transmission system.

Extended abstract for a User Presentation at the SSP User Meeting. The potential of the SSP standard to describe system structures to drive an end-to-end credible simulation process from the definition of an abstract analysis architecture to the evaluation of the overall system behavior in a co-simulation setup, is evaluated in this application to a heat pump system. From practitioners perspective the benefits and short-comings are compared against current best practices using proprietary solutions.

Motivated initially by the specific needs of Small Modular Reactors (SMR) transient analysis, including Molten Salt Reactors (MSR) draining or leak scenarios, or Sodium Natural Circulation Reactor (SNCR) scenarios, this paper presents the development of a generic two-substance liquid-gas medium and an approach used for a liquid Sodium medium. Such media have a low isothermal compressibility, which is a key issue for the numerical robustness of the simulation. These media are designed for compatibility with standard libraries like Modelica.Fluid. We detail their formulation, capabilities, and limitations. A significant portion of the paper is then dedicated to investigating a crucial aspect arising in such models: the representation of the liquid phase’s compressibility. We compare different approaches, analyzing their impact on the resulting equation system structure and numerical robustness using the OpenModelica debugger. This analysis demonstrates the trade-offs involved and provides insights into selecting appropriate liquid models for dynamic simulations involving non-miscible flows.

The paper presents the collaborative development of a digital twin for railway braking and traction applications as part of the ERJU MOTIONAL project. The project’s goal is to enhance the efficiency, user-friendliness, and maintainability of European railways through the application of digital twin concepts across nine distinct use cases. Contributing components are shared in a common library as Functional Mock-up Units (FMUs), enabling the creation of digital twins for various applications. Specifically, this study focuses on the braking and traction use case for virtual validation purposes. The paper details the vehicle dynamics model, control units, and their interfaces, presenting simulation results under diverse adhesion conditions. Additionally, it highlights the potential benefits of digital twins in optimizing railway operations and monitoring system integrity, ultimately enhancing the reliability and efficiency of railway transportation.

Steam power systems, as one of the critical power systems in industrial applications, require rigorous design and verification processes. Model-Based Systems Engineering (MBSE) provides a structured approach to decomposing system architecture from top to bottom and enabling multi-disciplinary collaborative design, ensuring precise requirement management and efficient design processes. To accelerate the iterative design and verification of steam power systems, this paper employs the SysML language to conduct requirement, functional, structural, and parametric analyses, thereby completing the system architecture design. The seamless transformation from system design architecture to system simulation architecture is achieved based on the SysML-to-Modelica tool. Additionally, Modelica simulation technology is utilized to construct simulation models and perform dynamic scenario analyses of the system. Innovatively, this paper proposes a closed-loop technical approach for steam power system design, simulation, and verification, which effectively optimizes system design and improves the efficiency of both design and verification processes.

Virtualization in development of ever more complex products is becoming increasingly important. Therefore, the use of modeling & simulation activities as part of product development and release is also increasing. Traceability is for these activities a keystone for quality tracking and reuse for efficiency. This presentation shows how this can be realized by applying the SSP Traceability layered standard (i.e., which information is involved related to the modeling and simulation activity, where does it come from and where shall it be propagated) in conjunction with the MIC Core standard for metadata. At the beginning, the Credible Simulation Process Framework developed as part of the prostep SmartSE project is presented, which enables integration into company processes. We will show this using the example of the challenges of developing a simulation model Due to different reasons, the development of a simulation model typically consists of the design and the implementation in a simulation tool (e.g. Dymola), the model being then immediately used for simulation purposes, skipping essential activities like the requirement specification (e.g. relevant application area, relevant physical effects, …), the documentation for the end-user and/or the model verification respectively the model validation. Often, the user documentation is written afterwards from the model implementation, leading, e.g. to potential transcription issues, inconsistent content respectively versions between implementation and documentation, and high review efforts to ensure the high quality of the developed model. We will then present the solution approach based on the SSP 2.0 and SSP Traceability Layered Standard and show it as a demonstration. Another important point for traceability and reuse is the use of standardized metadata to find and evaluate information. We will present a solution based on the MIC-Core Metadata Standard. We will close with an overview, which of these solution elements are implemented and publicly available.

Innovation in residential energy systems drives research efforts towards novel building and district control strategies. Development and testing of such strategies requires advanced, interactive building and district simulators. This paper presents yards, an interactive simulator that combines the modelling capabilities of Modelica with a language-agnostic and cross-platform interface for controller development. yards offers modelling flexibility beyond that of existing tools, as any custom, user-provided building or district model can be implemented easily. A case study demonstrates how yards can be used to simulate a tiny cluster of buildings.

Panel discussion on the value of open standards

Lucerene Trainstation
3 minute foot walk to landing stage 5 or 6

Boat Cruise

The complexity of modern integrated energy systems demands the systematic use of systems engineering methods and tools to address key challenges across product lifecycle. This keynote will explore three related and critical areas: 1) the demand for diverse model fidelities and analysis, 2) maintaining consistency across design layers, and 3) the importance of seamless tooling and integration. Real-world, HVAC-specific examples will highlight how these challenges are being tackled in practice and the opportunities they present. For over two decades, Carrier has relied on Modelica as a cornerstone technology for model-based product design. The journey began with the development of control systems for transportation refrigeration, where transient simulations allowed for rapid control function development and verification. Today, while controls development remains a vital application, Carrier has expanded Modelica’s use across the entire product lifecycle—from conceptual design, through testing and verification, to business sales tools, and into operation with monitoring and diagnostics. Design and operation of highly integrated energy systems such as data centers and district heating systems also calls for new methods and tools. Differentiated applications has driven diversified analysis with Modelica models. Modelica-based steady-state simulation has emerged as a back-bone in product design, sales tools, as well as in field diagnostics. Steady-state and transient optimization is also a necessary element across the product life cycle. The need for diversified analysis and computation is further amplified by application of systems engineering methods where requirements, design space exploration, and validation and verification are key elements.

This session is chaired by

This session is chaired by

This session is chaired by

This session is chaired by

This session is chaired by Johan Andréasson.

Johan is head of strategy and innovation at Modelon and has a long history with Modelica and FMI and has contributed with various papers on these topics since the first Modelica conference in Lund, year 2000.

Modeling pumps in 1D flows in Modelica is not new. In this paper a new approach is presented that can be used for general centrifugal pumps as well as centrifugal pumps based on measurement data for the DLR Thermofluid Stream Library. The presented model is based on dimensionless numbers for good scaling. This approach allows to predict the pump behavior also for pumps where no measurement data exist based on data from similar pumps. The models are designed with focus on numerically robustness. They avoid non-linear equation systems and work for zero mass flow rate and/or zero angular velocity. Furthermore care has been taken that extrapolation of the pump data outside the data range is robust without interpolation artifacts. A pump surge example demonstrates that the TFS can handle instationary behavior like limit cycles by default without any further modifications needed.

The convergence failure of iterative Newton solvers during the initialization of Modelica models is a serious show-stopper, particularly for inexperienced users. This paper presents the implementation in the OpenModelica tool of methods presented by two of the authors in a previous paper, to help diagnosing and resolving these convergence failure by providing ranked lists of potentially critical start attributes that might need to be fixed in order to successfully achieve convergence. The method also provides library developers with useful information about critical nonlinear equations, that could be replaced by equivalent, less nonlinear ones, or approximated by homotopy for more robust initialization.

Physics-based simulations (PBS) are increasingly valuable for real-time applications in embedded systems, yet integrating them on resource-constrained devices remains challenging. This paper presents ufmu, a lightweight framework that enables execution of FMI 2.0-compliant Functional Mock-up Units (FMUs) within the MicroPython environment, targeting platforms such as the ESP32. The proposed approach translates FMU model descriptions into C structures, integrates them into MicroPython firmware, and exposes a minimal Python API for simulation control, enabling model-based computations on-device without cloud dependencies. We evaluate the framework using a standard FMU model, comparing performance across ESP32, Unix, and plain C environments in terms of memory usage, execution time, and firmware size. Despite the ESP32's hardware limitations, the results demonstrate that meaningful simulations can be achieved efficiently, with minimal memory overhead. All code, documentation, and experiment instructions are freely available under an MIT license, supporting reproducibility and adoption in education, prototyping, and embedded research. This work also lays the foundation for future integration with eFMI and the FMI 3.0 standard.

Motivated by real use cases from Bosch for optimization-based functions and controllers in the context of energy optimal operation of vehicles and buildings, we investigate the usage of FMI for optimization purposes via the open-source tool CasADi. We implemented a framework in Python to automatically compare optimization results and computation times of optimal control problems in CasADi. We are able to compare the results generated by different implementations: a) including the system dynamics as FMUs and b) a native implementation where the system dynamics is realized by Python code for CasADi. We present results from two use cases: the trajectory following of a single track vehicle model and the optimal control of a building’s chiller system. Detailed analysis of the split of the execution time of one optimization run gives valuable insight which kind of FMI function calls or derivatives are competitive and which one have bottlenecks compared to the native solution in CasADi without FMUs.

Significant opportunities exist to preserve and reuse simulation and analytical models and data. Based on the capabilities developed by the Modelica Association (MA), LOTAR International (Lotar), and other contributing tool developers, multiple engineering and manufacturing industries can compile extensive archives of reusable and interoperable performance, behavior and integrated product information. Through the development and implementation of a preservation plan, the use of compliant off-the shelf software applications, and packaging using compatible data standards, a repository of analytical interactions, simulations and functional prototypes can be archived, maintained, and resourced by future users. This paper identifies progress made since the publication of (Coïc 2021, Coïc 2023) with respect to maturity of the LOTAR draft standards – ASD-STAN 9300 series, (ASD-STAN, 2025) and accompanying prototype implementations. There has also been significant progress on the side of Modelica Association relevant standards with the standardization of layered standards in FMI and SSP (Modelica Association 2024-11., Modelica Association 2024-12), and new versions of both standards. The very recently released layered standard SSP Traceability (Modelica Association 2025-04) provides mechanisms to document and preserve relevant metadata for model archiving – Long Term Archiving and Retrieval (LOTAR). The FMI layered standard FMI-LS-Ref standardizes parameter sets and reference results, which are an important subset of the required LOTAR International archiving data.

DNV is a global assurance and risk management company, and our research division has been involved in co-simulation efforts in the Norwegian maritime industry for more than 10 years, since the inception of the Open Simulation Platform initiative. In this time, we have worked with several challenges inherent to co-simulation, and its application to model-based assurance and secure industry collaboration. For the past three years, we have had a central role in the SEACo (Safer, Easier, and more Accurate Co-simulations) research project, from which we will present some of the results that we find relevant to the FMI community.

The coupling problem in co-simulations encompasses not only the challenge of correctly connecting the different simulation models in a larger system but also dealing with the coupling errors that are introduced during the data exchange [2] [3]. In recent work, the concept of power bonds, introduced in bond graph modelling of physical systems, has been applied to co-simulations, resulting in several experimental methods for error quantification and control. In particular, the energy-conservation-based co-simulation method (ECCO) [1] is attractive for co-simulation as it relies only on the coupling signals as power bonds and assumes no additional constraints on the models, co-simulation solver, or FMI version in use. The impact of this approach is highest for couplings with large numerical contributions, where the simplistic discretization found in many co-simulation solvers leads to correspondingly large numerical errors. For stiff couplings, where we here refer to stiffness in the numerical sense, this is of particular interest as an error reducing step size control algorithm may be used to identify a suitable step size for integrating the system equations based on minimizing the observed coupling error for each timestep.

As a part of case-study in the SEACo project, the ECCO algorithm is applied to the connection between a deck crane and hull. This system is known to be poorly suited for co-simulation because of the large inertias inherent to this coupling. We look at the effect of applying ECCO to such a system, starting with a simplified configuration of two connected masses that can be actuated in one degree of freedom, before considering the same problem under multiple degrees of freedom.

In SEACo, we also consider assuring complex operations based on simulations. This requires building trust in the chosen simulation methods, the system models, as well as the system integration as a co-simulation. The value of assurance in such a context can only be realized with sufficiently efficient and practical methods, which are currently lacking. We will present from the assurance activities performed in the project, centered on a case study involving a marine lifting operation between two floating structures. We will also connect this work to our recommended practice for the assurance of simulation models, and connected methodology for model qualification, verification, and validation.

References [1] Sadjina, S., Kyllingstad, L. T., Skjong, S. and Pedersen, E. “Energy Conservation and Power Bonds in Co-Simulations: Non-Iterative Adaptive Step Size Control and Error Estimation”. Engineering with Computers 33(3), 607–620 (2017).

[2] Sadjina, S., Pedersen, E. “Energy conservation and coupling error reduction in non-iterative co-simulations”. Engineering with Computers 36, 1579–1587 (2020).

[3] Sadjina, S., Kyllingstad, L. T., Pedersen, E. and Skjong, S. “Energy Conservation and Co-simulation: Background and Challenges”. Proceedings of the 2024 International Conference on Bond Graph Modeling and Simulation (ICBGM’24). Simulation Series Vol 56(2). San Diego, CA, USA. July 1–3, 155–168 (2024).

In the realm of model verification, ensuring traceability and repeatability is paramount for achieving Modeling & Simulation (M&S) credibility and development efficiently. This paper explores challenges and solutions for expressing complex model Verification & Validation (V&V) workflows that ensures readability while still enables automation, parallelization and advanced customization of verification activities. This paper contributes a practical solution that is evaluated through a detailed implementation within an in- house developed multi-purpose automation and Verification & Validation (V&V) framework.

As systems become more complex, the demand for thorough testing and virtual prototyping grows. To simulate whole systems, multiple tools are usually needed to cover different parts. These parts include the hardware of a system and the environment the system interacts with. The Functional Mock-up Interface (FMI) standard for co-simulation can be used to connect those tools.

The control part of modern systems is usually a computing unit, such as a System-on-a-Chip (SoC) or Microcontroller Unit (MCU), which executes software from a connected memory and interacts with peripherals. To develop the software without requiring access to the physical hardware, full-system simulators, so-called Virtual Platforms (VPs), are commonly used. The industry standard framework for VP development is SystemC TLM. SystemC provides interfaces and concepts that enable modular design and model exchange. However, SystemC lacks native FMI support which limits the integration into broader co-simulation environments.

This paper presents a novel framework to control and interact with SystemC-based VPs using the FMI. We present a case study showing how a simulated temperature sensor in a SystemC simulation can obtain temperature values from an external tool via FMI. This approach allows the unmodified target software to run on the VP and receive realistic environmental input data such as temperature, velocity, or acceleration values from other tools. Thereby, extensive software testing and verification is enabled. By having tests ready and the software pre-tested using a VP once the physical hardware is available, certifications like ISO 26262 can be done earlier.

Cyclic frosting and defrosting operations constitute a common characteristic of air-source heat pumps in cold climates during winter. Simulation models that can capture simultaneous heat and mass transfer phenomena associated with frost/defrost behaviors and their impact on the overall heat pump system performance are of critical importance to improved controls of heat delivery and frost mitigation. This paper presents a novel frost formulation using an enthalpy method to systematically capture all phase-change behaviors including frost formation and melting, retained water refreezing and melting, and water drainage during cyclic frosting and defrosting operations. A Fuzzy modeling approach is proposed to smoothly switch source terms when evaluating the dynamics of frost and water mediums for numerical robustness. The proposed frost/defrost model is incorporated into a flat-tube outdoor heat exchanger model of an automotive heat pump system model to investigate system responses under cyclic operations of frosting and reverse-cycle defrosting.

Tasks involving Modelica models often do not simply investigate the dynamic behavior of a system, but rather want to characterize also possible optimal control strategies according to suitable criteria. Unfortunately, since Modelica does not support out-of-the-box optimization features, users are often forced to use other tools to code again the system model for optimization studies. For this reason, the authors present Modelica2Pyomo, an open-source tool to translate Modelica models into Pyomo optimization programs, leveraging on their flat Base Modelica representation. This work illustrates the main features of Modelica2Pyomo, including automatic variables and constraints normalization, expressions manipulation and initialization via Modelica simulation results. To demonstrate the capabilities of this framework, two examples are showcased, including an industrial relevant open-loop optimal control problem of a solid-oxide fuel cell.

Numerical simulation of a thermofluid vapor compression cycle (VCC) model in Modelica, for example, can exhibit a variation in the total fluid (refrigerant) mass. This paper provides a dynamic analysis of a commonly used VCC model, identifies and analyzes the root cause of this variation, and proposes a number of remedies. The cause lies within the dynamic equations that result from application of the principle of mass conservation. In many common formulations, these equations express the conservation of mass as one or more differential equations that equate the time derivative of mass to zero. The resulting set of n ordinary differential equations (and a number of auxiliary algebraic equations) include the time derivative of a mass constraint function, but not the actual mass constraint function itself. As a result, this modeling formulation has the following properties: (1) equilibrium solutions of the system are neither isolated, nor exponentially stable; (2) a linearization about any equilibrium solution has at least one eigenvalue equal to zero, making an equilibrium solution stable, but not exponentially stable; (3) for a VCC model formulated using two fluid states per control volume, a one-dimensional equilibrium manifold exists containing all of the equilibrium solutions, and is parameterized by the total fluid mass; (4) an (n-1) dimensional, stable, invariant manifold exists transverse to the equilibrium manifold, defined by the mass constraint function, and on which analytic solutions to the model evolve and the total fluid mass remains constant; and (5) numerical solutions may drift off of this manifold, resulting in an observed drift of fluid mass. These properties have consequences for simulation, control design, numerical model reduction, and state estimation. A number of methods to stabilize the mass constraint are proposed and a number of examples that illustrate the behavior, analysis and remedies are provided.

Direct collocation-based dynamic optimization plays an important role in the optimization of equation-based models. With this approach, continuous problems are transcribed into sparse nonlinear programs (NLPs) that can be solved efficiently. The open-source Modelica environment OpenModelica provides an implementation using Radau IIA collocation, but has major limitations, such as the lack of parameter optimization, no adaptive mesh refinement, and no support for higher-order integration schemes. This paper presents (1) a comprehensive reimplementation that addresses these limitations and (2) a novel $h$-method mesh refinement algorithm. Implemented in the custom Python / C++ optimization framework GDOPT, the approach demonstrates significant performance improvements, solving typical problems 2 to 3 times faster than OpenModelica under equivalent conditions. Using the proposed mesh refinement algorithm, the framework correctly identifies non-smooth regions and increases resolution accordingly, requiring only a small increase in computation time. The implementation lays the foundation for a future integration into the OpenModelica toolchain.

Hybrid simulation models combine physics equations with trainable components to improve simulation results and performance. Physics-enhanced neural ordinary differential equations (PeN-ODE) are a promising type of hybrid models that combine artificial neural networks (NN) with the differential equations of a dynamic system. Dynamical simulation models are often part of embedded control algorithms of cyber-physical systems (CPS); compliance with the safety and real-time requirements of such embedded environments is, however, challenging.

In this work, we propose a workflow to incorporate trained NNs in Modelica models to form hybrid simulation models that are PeN-ODEs. We thereby focus on the transformation steps from equation-based trained PeN-ODEs in Modelica towards causal solutions suited for the embedded domain -- up to and including MISRA C:2023 compliance checks and final software-in-the-loop (SiL) tests of generated production code in the modeling environment -- for which we leverage eFMI standard compliant tools (Dymola and Software Production Engineering). It is of particular interest, how the trained NNs of the hybrid model are implemented. We present two approaches: (1) generation of C code using existing Open Neural Network Exchange (ONNX) tooling and (2) pure Modelica code with the tensor-flow represented as multi-dimensional equations. Both approaches are discussed, highlighting why (2) is, in the long run, a better option given the eFMI technology space.

This paper presents a concept for integrating geometric tolerance analysis into system simulations using Functional Mock-up Units (FMUs). While various methods and tools for tolerance analysis exist in the mechanical domain, there is currently no standardized or widely established approach for their integration into multi-domain system simulation. This work proposes a structured FMU interface based on the Functional Mock-up Interface (FMI) standard, enabling a modular and reusable representation of tolerance behaviour. The concept is demonstrated in a case study, in which an FMU for static tolerance calculation was implemented and successfully verified against a commercial analysis tool. Furthermore, the use of the FMU with FMUGym illustrates the potential of the FMI ecosystem to flexibly combine tolerance models with other simulation environments and analysis methods. The results demonstrate that FMUs can provide a suitable and tool-independent interface for integrating geometrical tolerance effects into system-level simulations and model-based engineering processes.

The increasing complexity of automotive design requires more frequent and integrated use of simulation for cross-domain development and validation. To support this need Renault has formalized the concept of simulation platform that integrates five key elements: digital vehicle, driver, simulation environment, simulation means, and assessments. The FMI standard facilitates interoperability among these components.
A case study on thermal management highlights challenges with varying co-simulation step sizes, impacting performance and accuracy. Introducing FMU Containers, which embed multiple FMUs and manage their connections, significantly improves simulation efficiency.

This session is chaired by

This session is chaired by

This session is chaired by

This session is chaired by

This session is chaired by

This study presents a method for predicting vehicle suspension component loads at the early design stage. A hybrid road simulation combines road load data acquired from a reference vehicle with the Time Waveform Replication (TWR) technique to generate virtual equivalent road profiles. The TWR was implemented in Python, and a multibody dynamics vehicle model developed using Modelon's Vehicle Dynamics Library was used to simulate chassis response. Integration and iterative simulation between the TWR system and the vehicle model were conducted via Functional Mock-up Units using the Python FMI library, FMPy. These virtual inputs were applied to a virtual test rig. In this study, road load data from a reference vehicle were used to derive the input signals, which were then applied to simulate the suspension loads of a target vehicle. Simulation results were validated against measurement data to confirm the effectiveness of the proposed method.

This paper investigates the hydraulic short circuit (HSC) operation of pumped hydro energy storage systems, specifically how it can be modelled and its impact on energy capacity, efficiency and operating range. Hydraulic short circuit operation enables power modulation during charging for fixed-speed pumped hydro systems, thereby allowing them to operate more flexibly. In this work, a pumped hydro storage system with 140.5 MW nominal charging power and 10 h nominal charging time is analysed. The storage is simulated via a dynamic, system-level model composed of detailed physical component models to accurately depict the HSC operating behaviour. HSC enables part-load operation from 39 MW to 107 MW but with significantly lower overall system efficiencies. The HSC operating limits are governed by the turbine's minimum flow and maximum power. The custom-built OpenModelica library employed for this analysis can be downloaded from https://github.com/ibeyers/pumped-storage-analysis.

The global expansion of data center construction is fueled by the rising demand for AI applications. These energy-intensive facilities face increasing pressure to operate efficiently due to the European Efficiency Directive (ENER, 2024), for instance. Minimizing their environmental impact is therefore critical. To address this challenge, the System Simulation team in Danfoss Climate Solutions and TLK Energy are focusing in this joint presentation on innovative heat recovery approaches in modern, liquid-cooled data centers as a key strategy for energy-efficient operation. Using a holistic model of the overall system from liquid cooled server racks via the cooling loop to chillers, dry coolers and heat recovery including heat consumers, we can determine the system efficiency covering all interactions of the sub-systems for fully transient conditions. The backbone of the system models are the validated models of the relevant components from Danfoss’ portfolio. By setting the simulation results in relation to the needed CAPEX we are going to draw a picture of an optimized system setup for the presented use case.

Marine operations are often developed with the use of numeric simulation, in particular lifting operations and transfer of cargo between different units at sea. The effect of the environmental conditions is often the limiting factor and must be included together with models of different components and sub-systems. This paper describes an approach to synchronize spatial and temporal environment information such as universal constants, current, wind, and wave for use in co-simulations of marine operations. Co-simulation models in marine operations will inherently use physical constants, wind and current velocities to calculate forces. Wind and current velocities can have spatial and temporal variations that require the models to synchronize the values. In the event of simulations in waves, the position of the ocean surface, wave particle velocities must be coherent between individual co-simulation models. Further complicating matters is the case of propulsors that produce forces by creating local currents. This paper suggests a structured description of an environment for co-simulation of marine operations. This is exemplified by the implementation of a co-simulation of an offshore lifting operation where vessel, crane, propulsors and lifting load are all integrated with a common environment.

As the maritime industry evolves, there is a focus on simulation-driven design, testing, and validation using novel technology solutions. Simulation models designed to represent the behaviour and features of real systems are increasingly available for testing during the early phase of the full development, but in many cases, their testing suffers from the availability of test oracles. Metamorphic testing has become increasingly used in different application domains as an approach to test systems when an explicit test oracle is unavailable. In order to increase its adoption by domain experts, we combine metamorphic testing with Behaviour Driven development for the verification and validation of simulation models. The tool-based approach facilitates automated test generation based on domain-specific custom metamorphic transformations to generate meaningful test inputs for metamorphic input relations. The method also uses features and scenarios extracted from system requirements and domain expertise to define metamorphic output relations. By automating test generation based on system behaviours as features, scenarios, metamorphic transformations, and output candidates for metamorphic relations in a Gherkin-like format, the tool enables practitioners to verify models based on domain-specific constraints and metamorphic relation checks. Our preliminary evaluation shows that the tool can detect MR violations in the simulation models under test and that automated test generation provides improved coverage.

Testing real hardware and simulation models in combination in a software- or hardware-in-the-loop set-up is challenging. One of the key factors is the high demand for accuracy in the simulation model. If classical modeling based on physical principles is not sufficient to reach the desired level of accuracy, hybrid modeling, the combination of physical simulation models and machine learning can be applied. In this publication, we train a hybrid model for a controlled electric motor within the electro-hydraulic braking system of a car under the conditions and restrictions of a real engineering application in the field. We apply state-of-the-art modeling patterns for this, and further extend them with application specific methodological optimizations. Finally, we investigate and show the quantitative and qualitative advantages of the proposed approach for this specific application, resulting in a gain in accuracy by multiple factors.

Optimization and stress testing are key aspects of the design and verification process for large, high-risk systems. Optimization is about improving the capabilities and performance of a system; stress testing is about uncovering its weaknesses and faults. Both require a quantitative representation of the system's behavior, and for complex, multi-physical systems, co-simulation can be a very powerful method to create such a representation. However, co-simulation frequently involves the use of black-box subsystem models, which poses challenges to traditional optimization and stress testing methods. Here, we review the state of the art in co-simulation-based optimization and stress testing, focusing especially on \emph{adaptive stress testing} in the latter case, and discuss open research questions and promising research directions. In particular, we make the case that a co-simulation is not an entirely black box even when some or all of its subsystems are; it may be possible to exploit the visible system structure, coupling variable values, and partial subsystem information. We use examples from the maritime industry to motivate and illustrate the discussion, centering on the highly contemporary design case of an autonomous ferry.

Solar thermal technology is a promising solution for decarbonizing heat production in industrial applications and district heating networks. When combined with heat storage and advanced control strategies, it can cover a significant share of heat demands. However, designing and optimizing such systems is complex due to their dynamic behavior and the interplay of multiple physical phenomena. To better understand and design these systems, modeling tools are essential. Modelica is particularly well-suited for this purpose. At Newheat, large scale solar thermal field models have been developed in the Modelica language using the Dymola environment. These models represent the thermal and hydraulic behavior of a solar thermal field at two different levels of complexity. Each is designed for different project phases-fast simulations for early-stage feasibility studies and slower but more detailed simulations for the engineering phase. To assess the accuracy of both models, comparisons with measured data on an operational solar plant were performed. Results indicate that both models achieve high thermal accuracy, with errors of less than 4% in annual heat production. On the hydraulic side, the detailed model provides more precise results than the simplified one. The main drawback of this model being slow simulations in case of very complex solar field layouts. Moving forward, these models will support various applications and enable scalable modeling of complex solar thermal systems, adapting to different project phases and requirements.

This work presents a comprehensive simulation and optimal control framework for the energy management of a biogas generation plant. Using the Modelon platform and the Optimica library, a dynamic model of the plant was developed, capturing the behavior of digesters, gas storage, renewable energy and cogeneration units. The optimization targets the real-time scheduling of energy production and consumption, aiming to minimize operational costs and maximize revenue from electricity trading.

A model predictive control (MPC) approach was implemented, utilizing the IPOPT solver to compute optimal control trajectories over a defined prediction horizon, updated every hour. The control variables include the partialization of cogenerators, the decision to buy or sell electricity from the grid, and the management of energy buffers, all while ensuring compliance with process constraints and system dynamics.

To enhance forecast accuracy and system responsiveness, real-time weather data and electricity price forecasts are integrated into the control algorithm. This allows the plant to proactively adapt to variations in ambient temperature (affecting biogas production) and market price volatility.

The implementation resulted in a 6% improvement in overall energy efficiency and increase in profitability compared to baseline rule-based control strategies. These results validate the use of predictive and optimization-based approaches for advanced energy management in renewable biogas systems.

Self-restriction to a certain modeling style can enable the modeling of large-scale systems and the robust modeling of complex system architectures. This paper discusses how such a self-restriction can be achieved within the Modelica language and provides a corresponding example.

Cyber-physical systems are self-adaptive, changing their behavior at run time to adapt to their context. Hence, their simulations must also handle variability at run time. The lack of support for variability in industrial equation-based modeling languages, such as Modelica, causes problems when simulating self-adaptive systems, e.g., they only limitedly support structural variability, and state transitions are based on if-then-else conditions that can cause conflicts, especially for complex control mechanisms. We present a modeling technique for equation-based models containing variability by implementing concise and dedicated language constructs to express state space and transitions via contextual modeling. Contextual modeling abstracts the modeled world and allows the definition of constraints that reduce the risk of reaching conflicting states. We demonstrate the feasibility of our approach on a case study, presenting the advantages of our modeling technique regarding the definition of state control and the reduction of risk for reaching conflicting states.

This study proposes an optimal temperature feedback control aimed at minimizing the electricity consumption of condenser water loops, saving a significant amount of the total energy compared to the existing control. The developed control can be applied across multiple plants irrespective of their diverse configurations, such as sizes and performances, because the functions used in the energy-efficient control are grounded on the fundamental physics that take place in general condenser water loops. Specifically, we analytically derived the physical equations for the optimal flow rate and temperature of condenser water, identifying key influencing factors thereof. The key factors were then verified using a high-fidelity system simulator using Dymola. Lastly, optimal functions used in the proposed control were suggested using regression analysis, which is consistent with the derived equations. The developed control including optimal functions is expected to be implemented in twelve semiconductor plants in 2025.

System simulation is inherently an applied technology, driven by specific application demands. The Modelica language enables equation-based, system-level, multi-domain, and visual modeling, facilitating researchers conducting system simulation studies of nuclear power systems. This paper introduces the custom-built AESE library on the OpenModelica platform. Examples in AESE are provided to illustrate the development and simulation of system models for conventional pressurized water reactors (PWRs) and high-temperature gas-cooled reactors (HTGRs). The paper also describes tasks completed under different application scenarios, including hardware-in-the-loop simulation, multi-objective optimization, system identification, and rapid optimization, using model-driven and data-driven approaches with the AESE library. The further application of simulation models and data has significant practical and engineering value. This paper serves as a valuable reference for the intelligent application of energy system models in the context of advanced engineering challenges.

This paper introduces the ShipSIM, a novel free (standard conforming) Modelica library for modeling and simulation of ship maneuverability. This first Modelica library in the field of ship propulsion and maneuverability provides the components and flexibility to model the ship propulsion, hydrostatics and hydrodynamics to develop ship maneuvering simulations fully compatible with the Modelica Standard Library components. In this paper is presented the library’s key features and structure and introduce the underlying physical and mathematical foundations and modeling approaches. In addition, current implementation status, applicability limits and future development is discussed. The library development is coordinated by the authors and it is used in several MSc thesis. ShipSIM library is available on http://modelica-spain.org:3000/ Basilio/ShipSIM.

The ability to systematically compare and evaluate diverse control strategies is essential for the development of effective control algorithms in autonomous driving. This contribution presents the VDCWorkbench Modelica Library, a unified platform designed to support the development, testing, validation and verification of vehicle dynamics controllers and energy management strategies. The presented library is an extension of the IEEE VTS Motor Vehicle Challenge 2023 models and offers multi-physical component modeling, including a hybrid energy storage system (fuel cell & hydrogen tank and battery with aging model), as well as vehicle dynamics control for autonomous driving research projects. Two path-following approaches are featured: an open-loop lateral controller with a static inversion of a single-track model, and a closed-loop state-dependent geometric path-following controller with static control allocation. The library may also serve as the foundation for development of vehicle control methods, such as two-degree-of-freedom control approaches concepts. One example is given for this combination of a feedforward controller with residual reinforcement learning, where a learned agent improves the performance of the open loop controller. The entire library will be released as open source on GitHub in September 2025.

In a context of massive penetration of Renewable Energy Sources (RES) in the grid, it is of prime importance for Transmission System Operator (TSO) to correctly represent these components in their day-to-day dynamic simulations used to assess system stability, both in operational and planning situations. Well-parametrized generic Power Electronics Interfaced Resources (PEIR) models have emerged as a viable option for TSOs for large-scale stability situations, and are largely available in commercial power system software. However, despite previous efforts in the open-source community, especially in the Modelica one, there is a lack of a complete, up-to-date, and industrial-grade library. This paper presents the current status of the Dynaωo Modelica library for standard PEIR models that tries to fill this gap. The library’s architecture and content, the implementation choices, and already existing industrial usages of the models are presented in detail, alongside validation cases and results.

Context-aware systems are crucial in modern cyber-physical systems (CPS), enabling dynamic adaptation to changing conditions. Many such systems involve complex variability. Modelica, a leading equation-based language for modeling and simulating physical systems, struggles to manage this complexity. As variability management becomes more complex, traditional Modelica constructs such as conditional statements, state machines, and state graphs become increasingly inadequate and difficult to maintain. This work introduces a context-oriented modeling paradigm for Modelica by integrating Petri Net-based context control with FMI-based Variable Structure Systems (VSS). Specifically, we present Context Petri Nets (CoPN) as a formal mechanism for representing and managing contextual dependencies. By combining CoPN with VSS, our approach enables advanced variability management, supporting the modeling and simulation of complex, context-aware CPS.

This paper describes the use of a Modelica system model to support the development of a heat management control system for a fossil-free sailing yacht. Due to tight project timelines, the control system was developed and tested virtually, avoiding delays associated with waiting for the physical system to become available. The system model covers key functions such as heat recovery and heat dumping, enabling automated testing of various operational scenarios. This approach not only accelerates development but also reveals early insights into interactions between the control logic and system dynamics. The model is designed to be seamlessly replaced by the real system once it is built. Future comparisons between simulated and real-world performance will guide refinements to improve model accuracy and support model-based tuning of the control system.

This paper proposes a new paradigm of the Embodied Cyber-Physical System (Embodied CPS, ECPS) to address the issues of the disconnection between physical laws and intelligent decision-making and the insufficient interaction with dynamic environments in the modeling and simulation of traditional CPS. ECPS achieves unified modeling of physical laws and autonomous decision-making through the "perception-decision-action" closed loop.To verify ECPS, an embodied space framework based on Modelica/MWorks is designed. Through three major technological innovations: constructing an embodied domain modeling specification and embedding the Navier-Stokes equations into the training of the policy network; expanding the syntax and semantics of Modelica, encapsulating physical constraint reinforcement learning components, and establishing a gradient interaction protocol between the Physics-Informed Neural Network (PINN) and Modelica equations; building a digital twin-hardware-in-the-loop co-simulation platform based on the FMI/SSP protocol to establish a collaborative verification link between high-precision physical simulation and real-time decision-making.Taking the Unmanned Surface Vehicle (USV) as the carrier, the full-process method from dynamic modeling, reinforcement learning strategy training to virtual-real environment co-simulation is demonstrated. Experiments verify the effectiveness of this framework in achieving the closed-loop coupling of physical simulation and intelligent decision-making under complex sea conditions, providing a methodological foundation for interpretable modeling and verifiable simulation in the development of embodied intelligence.

Announcment of Awards, Changes of Board Members and new Conferences

This session is chaired by

This session is chaired by

This session is chaired by

This session is chaired by

This session is chaired by

In a 2023 Modelica Conference paper, we proposed a novel method for the modular structural analysis of DAE systems, in which the structural analysis is not performed on flattened models, but rather at the class level. A new notion of structural interface was proposed, in which classes are enriched with context information. That paper developed our approach based on a few illustrative examples.

In this paper, we provide the details of our algorithm. Its performance depends on the system architecture: the analysis of models having a small number of classes (possi- bly instantiated many times), with a low treewidth system architecture, scales up very efficiently with this approach. We then present additional benchmarks, among which a urban heating network, a representative real-life example on which a near-logarithmic scaling up is shown.

Ammonia is a promising zero carbon and sustainable hydrogen carrier that can be used as a fuel in solid oxide fuel cells (SOFC) by offering advantages related to the ease of storage and the possibility of being used directly without an external reformer. In this study, a Modelica-based dynamic model of an 'Ammonia to Power' (A2P) system was developed by integrating ammonia decomposition kinetics, electrochemical reactions, all the system-level components and the main control loops. A novel Balance of Plant (BoP) configuration is proposed, featuring a five-way heat exchanger that recovers waste heat primarily using the fuel stream as the thermal energy vector instead of air. The model evaluates transient responses to operational perturbations, the behavior of the different control loops, and recirculation percentage rates to optimize system performance. Efficiency is calculated as the ratio of the power output from the SOFC to the power derived from the fresh ammonia line.

With the help of MBSE (Model-Based Systems Engineering), development and validation of ADAS (Advanced Driver Assistance Systems) can be planned and implemented in a structured and consistent manner across collaboration partners and departments. The consistent use of standards supports this approach: Based on a requirement formulated using the ReqIF (Requirements Interchange Format) standard, the ADAS architecture can be specified using the SysML (Systems Modeling Language) standard. Corresponding standards are also available for the planning and implementation of tests at component level in this context, particularly for simulation-based test environments (SiL, Software in the Loop): The SSP standard (System Structure and Parameterization) defines systems of simulation models and their parameterization. The integration and coupling of corresponding models are defined by the FMI (Functional Mock-up Interface) standard. However, the question remains open as to how specifications at the requirements level can be transferred into concrete system architectures for the execution of tests at the component level. This presentation addresses this question and describes a suitable procedure using open-source, royalty-free specifications, and software artifacts: Based on the requirements of UNECE Regulation No. 157 - Automated Lane Keeping Systems (ALKS), a system architecture is defined that results in the implementation of an automated test using freely available simulation frameworks and models.

This paper highlights some still-open issues in the Modelica representation of industrial controllers, particularly when simulation efficiency must balance with fidelity to their real-world implementation as digital components of cyber-physical systems. The treatise focuses on PID controllers due to their predominant role, though the ideas readily extend to virtually any other control structure. The aim is to stimulate discussion around this important yet often overlooked topic, while also suggesting directions for future development.

Overall vehicle performance optimization is the main target in race car and road hypercar development. Considering the complexity of current vehicles, a holistic approach to analyse the interaction of vehicle dynamics, powertrain cooling system dynamics, brake cooling and human drivers in the same simulation can be vital to maximizing the overall performance (Bouvy et al, 2012). This article is a continuation of a previous article written by Dallara and Claytex (“Race Car Cooling System Model for Real Time use in a Driving Simulator”, Stellato et al, 2023). The previous collaboration describes the implementation of a 1D cooling system model integrated with a vehicle multibody model to be used in the Dallara dynamic driving simulator with human drivers. This collaboration has continued into a new phase, where Dallara has developed a model to optimize the brake cooling of its vehicles, with Claytex’s VeSyMA suite use for the auxiliary vehicle systems and code compilation. The model has been validated through comparison with real data of an existing vehicle, showing an acceptable accuracy to size a race car braking system and for a refined assessment of the global vehicle performance on the driving simulator.

One of the most popular ways of measuring signals for various components in vehicle on durability tracks will be Road Load Data Acquisition (RLDA) which gains load data through accelerated customer usages tests on either proving ground or field, using an instrumented vehicle. However, under the current circumstances, not only does a vehicle development cycle get shorter but also new mobility concepts keep being developed, a methodology is needed to cope with overall durability assessment in advance of a physical prototype being made for testing. In this study, this virtual testing methodology will be called Virtual RLDA. n this study, one of the most well-known electric vehicles in market is from Hyundai, with 2 in-wheel motors in rear which was chosen as the target vehicle. First, 6 major sub-systems such as vehicle controller, suspension, hydraulic brake, motor, battery and thermal management are modelled and validated separately, then they are integrated into one large vehicle model as a part of establishing the Virtual RLDA Platform which will be used for overall vehicle level durability development in early stage in Hyundai Kia Motor Company (HKMC).

Interest in ammonia as an energy carrier is growing due to its superior storage and transport properties compared to hydrogen. The objective of this work is to construct a useful tool for predicting the behavior of a solid oxide fuel cell (SOFC) stack fed directly with ammonia. This configuration is particularly interesting because the internal cracking of ammonia eliminates the need for an external cracker, thus reducing the overall cost of the system. The ammonia decomposition reaction was implemented in the anode channel of the stack and calibrated against literature results. The model was then validated in the ohmic region only by calculating the area specific resistance (ASR) and comparing the results with experimental data collected at the Bruno Kessler Foundation (FBK) laboratory. This SOFC model can therefore be used as a starting point for the analysis of a scale-up application.

Offshore power generation and transmission re-quires long subsea cables. When using AC, the in-ductance of the cables results in reactive power that contributes to the load of the cables and increases the voltage at busbars. At the same time, offshore power systems are becoming increasingly complex. They are evolving from dedicated collector grids and transmission cables for wind farms through shared use of transmission cables by multiple wind farms to hybrid offshore grids that connect wind farms to multiple countries. This paper discusses operational challenges of off-shore grids and how they are solved with a model-based master controller that solves optimal power flow problems in real-time. The Kriegers Flak Combined Grid Solution serves as example. It transports wind power from the four offshore wind farms (Baltic 1 and 2, Kriegers Flak A and B) with a total capacity of 950 MW to shore. Additionally, it promotes energy trade between Germany and Denmark. At the heart of the connec-tion of these two energy grids through the Baltic Sea there is a Master Controller for Interconnected Operations (MIO). Based on a Modelica model and using the PowerSystems library, it solves several state estimation and optimal power flow problems in real-time. MIO is implemented as geo-redundant system with ABB AbilityTM OPTIMAX® and System 800xA.

The Modelica language (Modelica.org) makes it easy to build large, complex models by allowing the instantiation of reusable component models. Modelica tools typically expand arrays of variables, equations and components and perform symbolic transformations on the scalar elements. This reduces the efficiency of the translation process and makes it difficult to change array dimensions after translation.

Scalarization can be avoided by imposing certain restrictions on the way models are written. This paper describes such restrictions, and the algorithms needed to be applied during the translation. As a result, arrays are resizable after translation and also during simulation. Several examples demonstrate the approach with the Web App Modiator. As a side effect, it is also shown how to provide meaningful diagnostics for erroneous models.

Simulation-based validation and verification (V&V) of the overall system safety and functionality is crucially important to reduce development times while maintaining the product quality. Dealing with such complex problems usually requires several teams collaborate during the development process. Standardized interfaces and data formats like FMI, SSP are key enablers for such collaborative development environments. The authors developed a Transmission Control Unit (TCU) proof-of-concept to evaluate how SSP and FMI standards can facilitate such collaboration development tasks. For example we analyzed: (1) sharing models with an agreed architecture, while applying IP protection, (2) performing parameter variations for testing and calibration of a component; (3) converting a given simulation architecture and interface definition to an SSP architecture in an automated and incrementally updatetable manner. Very recently, we started to extend this concept with virtual ECU implementations, utilizing the features of FMI 3.0 and FMI layered standards such as FMI-LS-XCP and FMI-LS-BUS.

Equation-based modeling that utilizes reusable components to represent real-world systems can result in excessively large models. This, in turn, significantly increases compilation time and code size, even when employing state-of-the-art scalarization and causalization techniques. This paper presents an algorithm that leverages repeating patterns and uniform causalization to enable array-size-independent constant time processing. Allowing structural parameters that govern array sizes to remain resizable during and after the causalization process enables the formulation of an integer-valued nonlinear optimization problem. This approach identifies the minimal model configuration that preserves the required structural integrity, which can subsequently be resized as needed for simulation. The proposed method has been implemented in OpenModelica and builds upon preliminary work aimed at preserving array structures during causalization, while still resolving the underlying problem in a scalarized manner.

FMI and SSP are interface standards for exchanging simulation models between heterogeneous tools.SmartSE project of the prostep ivip Association is studying the Credible Simulation Process Framework for reliable system development through simulation-based decision making which utilizes FMI and SSP. JAMBE (Japan Automotive Model-Based Engineering center) has developed interface guidelines for simulation model exchange between multiple companies while utilizing SmartSE recommendations. It also publishes reference models for vehicle system simulation of HEVs and EVs.These new standards and design methodologies continue to be demonstrated by JAMBE working activities and are increasingly recognized by Japanese automotive industry as providing a reliable and effective way for cross-team/cross-company co-development. This presentation also introduces the model exchange and co-simulation environment built on cloud server and show example use cases of simulation-based decision making using SmartSE and JAMBE standards, as well as FMI and SSP standards.

Electrolysis systems are the key technology to produce hydrogen from renewable energy sources. The modeling and simulation of proton exchange membrane (PEM) electrolyzers and the connected power grid are a challenge due to highly fluctuating loads in the available electrical power. Our study focuses on the modeling of the electrolyzer and its balance of plant with varying loads resulting from different energy sources. The balance of plant contains a water supply system, the electrical power supply and processes for treating the produced hydrogen, in particular drying. The electrolyzer model is validated using steady-state and dynamic data from the literature. Thereby, it is shown that the developed model accurately represents the measurement results from the literature. Furthermore, we implement an empirical degradation model to assess the effects of different use-cases on the efficiency of the electrolyzer.

Solid-state batteries are promising for electric mobility as they potentially offer higher energy densities and therefore driving range, long-life and safety. They come with changed thermal properties and thermal management requirements. In this paper a comparison of a solid-state battery system with a state-of-the-art liquid battery system is presented. This study compares a solid-state battery with a state-of-the-art liquid NMC battery for an electric coach on typical real-world long-distance routes, including fast-charging. The thermal management is performed by a reversible R744 heat pump. By designing both battery systems for the same energy capacity, the solid-state battery releases more heat during fast-charging. At the same battery system size, the solid-state battery significantly outperforms the liquid battery system in driving range and does not need more cooling.

Controller development of aerial vehicles is a long-standing task during aircraft design and has a large impact on the resulting systems performance. In today's integrated design loops, ideally all components of the aircraft must be considered in simulation and testing in order to develop complex system architectures and meet performance requirements. One of the tools that are suited for modeling and simulation of a multidisciplinary aircraft and systems assembly is DLR's FlightDynamics Library, which takes advantage of the capabilities of the Modelica modeling language. In this work, a new augmentation to the library is discussed, which implements a range of common aircraft control concepts which can be used for design of new controllers, closed-loop simulation and experimental testing via code-export. The library is set up in a modular way, so that flight guidance and flight control systems can be developed for multiple aerial platforms, including manned/unmanned fixed- and rotary-wing aircraft. For this paper, a simulation example is provided by means of an autoland controller that shall be designed for a high-fidelity 6-DoF fixed-wing aircraft model in Modelica. The combination of aircraft and controller models is subjected to a three step synthesis process, which yields a controller that is robust against external and internal disturbances.

In response to growing interest in high-temperature electrolysis, CEA-Liten has developed multi-stack Solid Oxide Cell (SOC) testing equipment comprising four electrolyser stacks. Each stack includes multiple cells, and together they form a module housed within a shared thermal enclosure and connected to a balance of plant. This setup supports the development of innovative control strategies, essential due to the stacks’ sensitivity to operating conditions. To mitigate risks and costs associated with testing new strategies directly on the physical system, a Model in the Loop approach was implemented. The model replicates the real module’s characteristics and operational capabilities, allowing safe and efficient design and validation of control strategies. Various transition scenarios between operating conditions—tailored to diverse production needs and constraints—were developed and validated using the model before real-world implementation. This paper presents the model-based control methodology and compares experimental results with simulations.

The Modelica language (www.modelica.org) has become a de facto standard for systems modeling and many tools exist. This paper describes certain modern enhancements and a static web app implementation called Modiator (Modelica Instant Simulator). It allows an improved immediate first-time user experience since the web app is available in seconds and simulations can be done directly in the browser. State of the art numerical solvers from the Sundials suite have been compiled into WebAssembly. The Modelica model is translated into Javascript code using techniques such as sorting, tearing, index reduction, state selection, etc. A subset of Modelica is supported with some extensions, for example, support for self-modifying models. This paper also presents the Fluid1D and Model3D libraries.

This study presents a reduced-order model (ROM) of a large-scale transcritical CO2 heat pump designed by MAN Energy Solutions, currently operating in Esbjerg and under construction in Aalborg, Denmark. The original high-fidelity model, implemented in Modelon Impact and presented at Modelon Innovate, captures the complex thermodynamic and control dynamics of the heat pump but comes with high computational cost, making it impractical for real-time applications. To address this issue, we trained a neural network ROM using simulation data from the Modelica-based model, enabling significant reduction in simulation time without compromising accuracy. We demonstrate how the ROM is integrated back into the Modelica framework to facilitate fast simulations, paving the way for real-time control scenarios and digital twin implementations. Performance evaluations show that the ROM retains key dynamic behavior and delivers simulation speed-ups by more than two orders of magnitude. The model captures start-up and shut-down scenarios, assessing system reliability and transient behavior, and evaluates the heat pump’s flexibility for grid balancing, responding to variable load demands and grid needs. This approach not only accelerates system-level studies but also supports online applications such as fault detection, predictive control, and system optimization. Our work highlights the potential of combining physics-based modeling with data-based machine learning techniques to bridge the gap between simulation fidelity and computational efficiency, crucial for the development of next-generation digital twins in the energy domain.

Mitigating periodic oscillations (e.g. in rotating systems) is a common control engineering problem. Fast Fourier Transform (FFT)-based methods are well-suited for respective analysis. While FFT algorithms inherently assume signal periodicity, rotating systems often exhibit true periodic behavior (e.g., shaft rotation frequencies). Using angle-sampled data rather than time-sampled data allows direct analysis of oscillations relative to rotational cycles, which is particularly useful for tracking unbalance or periodic external excitation in rotating assemblies. Modelica provides several built-in resources to address these challenges. First of all, inverse models have the potential to derive an ideal control signal in time domain. For periodic disturbances, this ideal control is likely to be approximated well by a periodic, i.e. Fourier-transformable signal. Modelica is an appropriate model environment to store and retrieve tabled FFT data depending on operating conditions such as rotational speed. In a real-time application, synthesizing control signals from precomputed Fourier tables offers a practical alternative to executing potentially complex inverse models online, reducing computational effort and system complexity. The paper demonstrates this approach using the example of mitigating oscillations induced by an internal combustion engine in a hybrid automotive power train.