While the mission of Clusters of Excellence is basic and fundamental research, a transfer of results to industrial partner is also expected. While SC SimTech maintains links to companies interested in simulation in general via its Industrial Consortium, EXC SimTech supports several measures to foster transfer to industrial partners.
SimTech Transfer Projects
SimTech Projects
METEOR Projects 2023
Projects with SimTech researchers
The team of Prof Morrison develops simulation methods for the in silico optimisation of drug transport from the circulatory system into the brain. These methods are applied to improve the design of therapeutic antibodies for the treatment of neurodegenerative disorders, which are developed, produced and functionally tested in collaboration with academic, clinical and private sector partners.
Contact: Markus Morrison
Active learning aims at providing a training set of machine-learning applications that covers as many relevant scenarios as possible. In that project, researchers from SimTech and NEC Laboratories Europe joined forces to develop a method to bias molecular dynamics simulations towards regions in which an existing model lacks training data. It shows unprecedented efficiency for the simulation of peptides. NEC’s business areas include drug development and material discovery and the research conducted with SimTech is used to improve the simulations of atomistic systems using machine learning. Active learning plays an important role in this context as it enables a more intelligent and resource-efficient way to generate training data.
Contact: Johannes Kästner, Mathias Niepert
preCICE is an open-source coupling library for partitioned multi-physics simulations used in more than 100 research groups in academia and industry. Through TTI GmbH (Technology Transfer Initiative of the University of Stuttgart), we offer priority support and bespoke training on preCICE.
Contact: Benjamin Uekermann
The results of the BMBF-funded project "Geomechanical Integrity of Host and Barrier Rocks – Experiment, Modeling, and Analysis of Discontinuities (GeomInt)" make significant contributions to the analysis of the formation and development of discontinuities in clay, salt, and crystalline rocks. In GeomInt, we (a group of different institutions from BGR Hannover, UFZ Leipzig, CAU Kiel, IFG Leipzig TUBAF Freiberg, MIB Stuttgart ) consider different types of discontinuities in damaged rocks, such as those occurring in the excavation-disturbed zone of crystalline and sedimentary rocks, discontinuities that can form at phase/grain boundaries, as well as discrete fracture and fracture networks e.g. from faults. The hydraulic pathways created or expanded by these discontinuities pose the risk of migration of fluid phases from deep geological layers to near-surface ones and play therefore a significant role in geomechanical safety analyses.
Contact: Holger Steeb
In EU KDT Smarty, the University of Stuttgart researches the applicability and usage of knowledge technologies in the distributed setting of a joint Cloud-IOT infrastructure.
Contact: Steffen Staab
The Analytic Computing research group investigates new methods for the representation and querying of uncertain knowledge and its applicability to the distributed management of knowledge required in a highly individualised factory for circular product design, production, and automated re-manufacturing.
Contact: Steffen Staab
The Analytic Computing research group collaborates with the team of EXC2075/1 participating researcher Prof. Tim Ricken, the Humboldt-University Berlin and Jena University Hospital to integrate simulation models and graph neural networks to individualize diagnosis and treatment decisions for patients with liver cancer.
Contact: Steffen Staab, Tim Ricken
The group of Bernard Haasdonk was part of a BMBF project, in which Universities Muenster, Darmstadt and Fraunhofer ITWM Kaiserslautern and Umicore GmbH have collaborated from April 2020 - December 2023. The topic was "ML-MORE: Maschinelles Lernen und Modellordnungsreduktion zur Vorhersage der Effizienz katalytischer Filter".
Together with the group of Prof. Dr. Martin Votsmeier (TU Darmstadt and Umicore AG & Co. KG, Hanau, Germany) we have a joint publication:
Döppel, F., Wenzel, T., Herkert, R., Haasdonk, B., Votsmeier, M.: Goal-Oriented Two-Layered Kernel Models as Automated Surrogates for Surface Kinetics in Reactor Simulations. Chemie Ingenieur Technik, 2024.
Contact: Bernard Haasdonk
About 68 percent of German employees currently commute to work by car without being able to use this time productively. In the future, autonomous vehicles could enable passengers to engage in various activities during the ride, requiring new software tools. The project aims to further develop the muscle-activated multi-body human model EMMA to assess safety and ergonomics in semi- or fully autonomous vehicles, with EMMA4Drive using optimization algorithms to efficiently simulate new postures and movements.
Contact: Jörg Fehr
The use of digital twins can be extended throughout the lifetime of a product, if it represents the systems dynamics accurately. Thus, a more realistic modelling approach with adjustable model depth is the approach of the research project "HybridCtrl-Digital Twin for model-based drive control of order picking robots".
Contact: Jörg Fehr
In the "Safe Motorcycle" project, a novel safety concept featuring seat belts, airbags, and protectors is being investigated, which secures riders to the vehicle during accidents, turning the motorcycle into a protective shell. The goal is to enhance the passive safety of motorcycles and increase their suitability as commuter and shared mobility vehicles. Simulative approaches and biomechanical analyses are used to optimize the concept, as experimental investigations are costly and time-consuming.
Contact: Jörg Fehr
Human-centered mobility focuses on the behavior of vehicle occupants to improve driving safety. In this research project, we are therefore investigating the creation of digital twins to assess injury risks when occupants adopt different postures in the vehicle. To this end, we use camera and lidar sensors of a smartphone to reconstruct the kinematics of occupant posture and surrogate models to predict dynamic behavior.
Innovation Campus Future Mobility, BUP20 Bottom-UP Project by Jonas Kneifl, real-time capable mobile digital twins – for the safe and convenient operation of new mobility concepts (TEDZ)
Contact: Jörg Fehr, Jonas Kneifl
The automotive industry is undergoing a transformation, with vehicle safety remaining of utmost importance. A comprehensive safety strategy that integrates all available sensor data and optimizes both active and passive safety systems is crucial to protect occupants effectively. The knowledge and models available in the automotive industry are to be used to start interdisciplinary modelling, simulation and optimization.
Contact: Jörg Fehr
The research project "Simulative system and algorithm design in automation technology" explores the use of CNC controllers to manage industrial robots, aiming to synchronize robot operations with machine tools and unify programming and coordinate systems. The novelty lies in leveraging digital twins to enhance trajectory planning by integrating dynamic models of robots, enabling better performance and ease of use for laypersons. This approach was validated through both virtual commissioning simulations and real-world tests.
Contact: Jörg Fehr
As part of the nanodiagBW initiative, we find algorithms for efficient detection of protein translocation events in biological nanopores. Using state-of-the-art Machine Learning Methods we investigate characteristics in these signals that allow us to uniquely identify proteins translocating through the pore. We further apply extensive Molecular Dynamic simulation methods to model the translocation process of the peptide within the nanopore at the atomistic level in order to better study the interactions that occur between the peptide and the pore and understand the underlying molecular mechanism of nanopore sensing.
Contact: Christian Holm
The EuroHPC Center of Excellence MultiXScale develops and provides infrastructure, software, and training for multiscale simulations, emphasizing scalable and high-performance coupling of simulation schemes across various length and time scales. The Stuttgart team acts as work package lead and prepares the ESPREsSo package for particle and lattice-based models at the coarse-grained scale, targeting pre-exascale HPC systems, with a focus on energy materials applications.
Contact: Christian Holm
In the project “fair-flexi - A fully reproducible CFD, we create a fully reproducible CFD framewotj for simulation and training. The key feature is that the whole simulation chain (not just the code itself) follows the FAIR principles. This makes a lot of implicit user knowledge (how to set certain parameters, how to build a good grid, how to do nice postprocessing) explicit, accessible and reproducible. This will help both the scientific community as well as newcomers and learners.
Contact: Andrea Beck, Bernd Flemisch
The pan-European Center of Excellence for Exascale CFD is funded by the EU and the national funding agencies to identify and support the next generation of computational fluid dynamics codes that are capable of bridging the gap to the exascale era. FLEXI has been identified as the leading code for compressible aerodynamics. In the project "EuroHPC JU", we port the ecosystem to GPU based architectures (tier-0 EuroHPC systems), demonstrate its exascale readiness and apply it to the currently largest simulation of shock / boundary layer interactions.
Contact: Andrea Beck
The goal of the overal project was to augment GNC systems by using Artificial Intelligence techniques in order to improve their performance, flexibility, autonomy and capability to handle failures and performance degradation. The team of the University of Stuttgart and RWTH Aachen developed novel robust optimization-based guidance techniques which integrate model uncertainties and generate distinct robust behavior at the guidance level. The results highlight significant performance improvements over non-robust baseline solutions and with high design flexibility, while significant steps have been taken towards implementability for real-time operation.
Contact: Carsten Scherer
In order to better understand and optimize the durability of laser melt injection generated coatings, a thorough understanding of the eigenstresses is needed. Experiments are very limited in resolving such stresses such that we need to understand through simulations the eigenstress state in tungsten carbide particles on a copper substrate undergoing rapid thermo-mechanical processes. In order to achieve this goal, a combination of multiple micromechanical model order reduction methods and a data-driven approach are combined.
Contact: Felix Fritzen
Dirk Pflüger is leading the Graduate School "Intelligent Methods for Test and Reliability" (GS-IMTR), Germany's only fully industry-financed Graduate School. It is funded by Advantest, one of the world leaders in Automated Test Equipment. It bundles the wide competencies of different institutes at the University’s Faculty for Computer Science, Electrical Engineering and Information Technology, and it is also open for other Faculties. Around ten PhD candidates and one Junior Professor will work together towards new solutions. The Graduate School’s scope includes topics such as design for test and diagnosis; post-silicon validation; test generation and optimization; robust device tuning; system-level test; lifetime test and reliability management; and test automation. A modern understanding of these topics demands novel artificial intelligence methods and has tight connections to data science, data analytics, data understanding, visualization, security, and privacy.
Contact: Dirk Pflüger
The project deals with the modelling and experimental analysis of condensation in porous media and it is therefore of relevance for the wet engine application. The goals are however very different. In Syntrac we deal with the exhaust gas treatment, in the Wet Engine the injection of water is used to improve the engine performances. The re-condensation of water vapor is then part of the wet engine cycle. However, strong synergies are envisaged between TP-CO2 and RU 2.
Contact: Grazia Lamanna, Kathrin Schulte
In Firefox a structure-integrated skin heat exchanger is to be developed. This consists of a fibre composite sandwich structure with a folded core. This work is relevant for the design of modern aircrafts with electrified propulsion systems, e.g. fuel cells. The latter generate a large heat flow of unusable heat due to the low temperature difference, which leads to a greatly increased cooling requirement and require innovative cooling concepts.
Contact: Rico Poser
DETERMINISTIC6G addresses central challenges of future deterministic end-to-end communication enabled by 6G. In particular, partners from USTUTT investigate robust end-to-end scheduling approaches for dependable real-time communication over wireless networks with stochastic packet delay.
Contact: Frank Dürr