Abstract: Today's world faces significant sustainability challenges, from climate change and resource depletion to the need for cleaner energy and efficient production processes. Technology plays a critical role in addressing those challenges. Siemens has always led transformations with pioneering innovations and cutting-edge solutions which also positions us at the forefront of sustainable change. Siemens' approach, grounded in our concept of "technology with purpose," drives sustainable transformation across various industries. We focus on three key areas: digitization, automation, and electrification. Digitization is a powerful lever in achieving sustainability goals. By combining real and digital worlds, businesses gain the transparency needed to optimize operations, reduce waste, and make informed decisions that align with their sustainability targets. The proper configuration of products has a significant impact on their environmental footprint. Through green configuration, individualized products not only meet the functional requirements but also minimize their environmental impact. This holistic approach ensures that sustainability is embedded in every stage of the product lifecycle, from design and manufacturing to operations and beyond.

Short Bio: Sophie Rogenhofer is Senior Research Scientist and Project Manager at Siemens Foundational Technologies in Vienna, Austria, specializing in Data Analytics, Artificial Intelligence, and Configuration Technologies. She has more than 10 years of experience in leading and conducting research topics in an international industrial company. As project manager in research and pre-development, her responsibility is to harden new technologies, assess their suitability for industrial use, and support business units in integrating them into their products and services. Sophie Rogenhofer is leading project teams in the areas of Symbolic AI and Configuration Technologies to support business units in the design and implementation of product configurators by integrating specific technical expertise and data into powerful AI methods. Currently, her research focus is on AI for Sustainability, specifically on evaluating and hardening Green Configuration technologies for industrial use. This involves integrating life cycle assessment (LCA) calculations into product configurators, enabling the consideration of environmental impacts for highly complex products at the push of a button. Sophie Rogenhofer holds a degree in Mechanical Engineering and Economics from the Technical University of Vienna and is certified Project Manager (IPMA) and Scrum Master.

Abstract: Combinatorial (testing) methods have attracted attention as a means of providing high assurance at reduced cost and are considered a proven method for system testing. In this talk, we take a leap towards the application of such methods beyond system testing, and in particular examine their adaptation for use-cases typically encountered as part of the disaster management cycle with the aim to advance disaster preparedness and enhance the resilience of (cyber-)physical systems. Any efforts in the ex-ante phase of a disaster requires detailed understanding of potential disaster scenarios, which makes them an integral part of activities around advancing disaster preparedness, prevention, and resilience. For this reason, we consider different discrete mathematical models and generation techniques for descriptive disaster scenario generation.

Short Bio: Dimitris E. Simos is Key Researcher for the Applied Discrete Mathematics for Information Security research area with SBA Research located in Vienna and leads its Mathematics for Testing, Reliability and Information Security (MATRIS) research group. He is also the Head of Strategic Research at SBA Research responsible for shaping and implementing the strategic R&D agenda of the research center. He is also an Associate Professor (non-tenured track, habilitation in Applied Computer Science) with Graz University of Technology and holds a Guest Researcher appointment with the US National Institute of Standards and Technology (NIST), Applied Computational Mathematics Division (ACMD). During his career Dimitris has (co)-authored over 150 papers in Discrete Mathematics and their applications to Computational and Computer Science and has been awarded the rank of Fellow of the Institute of Combinatorics and its Applications (FTICA) and the Applications of Computer Algebra Early Researcher Award (ACA-ERA 2024). His research interests include Combinatorial Designs and their applications to Software Testing, Symbolic Computation, Optimization, Disaster Management and all aspects of Information Security.

Abstract: As the space industry moves towards reusability and cost efficiency, liquid rocket engines face new operational challenges. Reusable rockets require rapid and precise thrust control across a broad throttling range, while repeated use leads to wear and tear, altering engine dynamics over time. The increasing adoption of additive manufacturing adds further complexity by introducing variability in component geometry. To address these challenges, closed-loop control systems offer potential solutions by adapting to changes in system dynamics caused by both engine reuse and manufacturing variability. However, designing conventional control systems is difficult due to the need to manage multiple interdependent variables, combined with the limited computational capacity of space-grade processors. As an alternative, neural networks trained via deep reinforcement learning (DRL) present a promising approach. This presentation explores the use of DRL for controlling reusable liquid rocket engines, focusing on the LOX/LNG expander-bleed LUMEN engine. Key control challenges such as wide-range throttling, constraint handling, and efficiency optimization are addressed. The performance of DRL-based controllers is experimentally tested, achieving an average control error of 1.4%. A comparison with model predictive control is provided, followed by an outlook on future developments in fault- tolerant control systems, ensuring safety and reliability in next-generation rocket engines.

Short Bio: Dr. Günther Waxenegger-Wilfing is a Professor of Intelligent Space and Energy Systems at the University of Würzburg and the Head of the Applied AI Group at the DLR Institute of Space Propulsion. His research focuses on leveraging AI for sequential decision-making and advanced control systems, with applications in space propulsion, test bench operations, and intelligent energy management. Furthermore, he collaborates closely with the University of Würzburg’s Institute of Computer Science to advance autonomous decision-making in satellite systems. Before transitioning to AI and space research, Dr. Waxenegger-Wilfing earned his PhD in Theoretical Physics and gained experience as a Quantitative Analyst in the finance sector, where he applied data- driven models to financial systems.