Scientific contact person: Ezio Bartocci
Webpage of your present research group: https://www.eziobartocci.com/team.php, öffnet eine externe URL in einem neuen Fenster

Research Field: This Vienna Research Group will explore integrating advanced AI techniques, particularly large language models (LLMs), to transform cyber-physical systems' design and development processes. The group's research will focus on leveraging LLMs to formalize complex system requirements, enabling precise and automated interpretation of high-level specifications. This approach bridges the gap between human-centric design intents and machine-interpretable representations.

Additionally, the group will investigate how AI-driven methods can accelerate the design cycle of CPS by automating traditionally labour-intensive tasks, such as model generation, system synthesis, and validation processes. By embedding LLMs and other AI tools into the CPS design pipeline, the group seeks to improve CPS development's efficiency, reliability, and scalability for critical applications like autonomous vehicles, robotics, smart grids, and healthcare systems.

Embedding: The group will align with the Trustworthy Cyber-Physical Systems group (within the CPS Research Unit), complementing its formal methods, safety, and dependability expertise. Collaborative efforts will integrate AI-driven innovations with established verification frameworks to create robust and efficient CPS design methodologies.

Scientific contact person: Agata Ciabattoni
Webpage of your present research group: https://www.logic.at/staff/agata/team.html, öffnet eine externe URL in einem neuen Fenster

Laws whether social, ethical, or legal—are foundational to human society and are increasingly pivotal in the design and deployment of Artificial Intelligence (AI) systems. The interdisciplinary field of AI and Law explores this relationship in two complementary directions. First, AI tools, such as Large Language Models (LLMs), can significantly enhance the formalization and analysis of complex legal systems expressed in natural language—such as AI regulations, or street codes—enabling these laws to be checked, processed, and automated. Second, embedding compliance with legal, ethical, and social norms into AI systems is essential to ensure their safe, fair, and responsible use in society.

Scientific contact person: Peter Szmolyan
Webpage of your present research group: https://www.tuwien.at/mg/asc/dds

The envisioned research group seeks to bridge two foundational areas at the scientific forefront: Artificial Intelligence (AI) and the computational study of Differential Equations and Dynamical Systems (DE/DS). DE/DS is governed by complex workflows that are necessary to analyse differential equations. Indi- vidual steps in these workflows range from elementary (dimensionality reduction, non-dimensionalization, linearization at a fixed point, etc.) to advanced (bifurcation analysis, asymptotic expansions, blow-up, etc.) Parts of these workflows are of numeric nature, while other parts are of analytic and symbolic nature. Soft- ware tools at varying levels of specialization and complexity, such as Maple or pde2path, XPPAUT support mathematicians and applied scientists in navigating these workflows faster, as they cover individual parts of these workflows. However, currently, no unified set of tools exists, and using these tools successfully requires expert mathematical knowledge. The aim of this proposal is to leverage the capabilities of AI to lay the foundations for integrating the many existing tools for solving and analysing differential equations in order to increase support for mathematicians and applied scientists. A synergistic LLM-assisted approach that enhances the robustness and efficiency of the workflow is envisioned. The final goal is to be able to perform qualitative and quantitative analyses of differential equations at a significantly higher level of automation than currently possible. This approach will also have an educational dimension, as, once trained, the de- veloped models can be used to make parts of these complex workflows accessible to a wider audience of mathematicians, applied scientists and students – which helps drive the entry barriers towards research in DE/DS down. In establishing this research group, the department, as well as the wider faculty, would be at the forefront of the AI wave that is currently transforming many scientific disciplines but has not yet fully reached mathematics. This project is embedded into the research group Differential Equations and Dynamical Systems of Prof. Peter Szmolyan at the Institute of Analysis and Scientific Computing (ASC) at the Vienna University of Technology. Peter Szmolyan is an expert in dynamical systems and qualitative theory of differential equations. We envision fruitful collaborations with other members of the ASC in the areas of partial differential equations, for example, by exploring the integration of tools such as Netgen/NGSolve into our AI approach.

Scientific contact person: Eduard Gröller
Webpage of your present research group: https://www.cg.tuwien.ac.at/, öffnet eine externe URL in einem neuen Fenster

Our group performs extensive fundamental and applied research in computer graphics. Our areas of expertise are modelling and rendering for computer graphics, visualization, visual computing, virtual environments, and color. The new group aims to pioneer advancements in generative AI, enabling the synthesis, understanding, and manipulation of complex visual data across 2D and 3D domains. By leveraging Vienna's dynamic research ecosystem, this work will target high-impact societal applications while fostering collaborations with TU Wien and beyond. Objectives will be: High-Resolution Multi-Modal Generation, Image, Scene and Video Decomposition and Relighting, 3D and Video Synthesis, Cross-Domain Adaptation. The proposed research directly addresses key challenges in industries such as urban planning, medical imaging, and cultural heritage preservation. By enhancing generative AI’s interpretability and domain adaptability, the project will contribute tools that empower professionals in these areas, while ensuring ethical and equitable AI deployment.

The Research Unit of Computer Graphics (193-02) [1] performs extensive fundamental and applied research in computer graphics. Our areas of expertise are modelling and rendering for computer graphics, visualization, visual computing, virtual environments, and color. Besides our research projects, we specialize in consulting and technology transfer as well as computer graphics related education on both undergraduate and graduate levels. Visualization has been a major part of the research activities at our research unit over the last 30 years. The visualization group [2] is a sub-group of our research unit and performs basic and applied research projects in all areas of visualization (scientific visualization, information visualization, visual analytics) and visual computing.

[1] https://www.cg.tuwien.ac.at/, öffnet eine externe URL in einem neuen Fenster
[2] https://www.cg.tuwien.ac.at/group/Visualization-Group, öffnet eine externe URL in einem neuen Fenster

Scientific contact person: Andreas Otto

The laser is a versatile manufacturing tool used for various processes such as welding, cutting, additive manufacturing, ablation or drilling, across time- and length scales ranging from femtoseconds to hours and nanometers to meters. The underlying mechanisms of laser-material interaction and thus resulting process characteristics and properties of the final manufactured part are governed by a highly nonlinear, coupled multiphysical problem. Artificial Intelligence has a great potential to help gaining process understanding, predicting process outcomes, optimizing existing processes or to solve the inverse problem of finding the necessary process parameters to obtain desired processing results. Within the research group’s activities, AI methods shall be established and used at the intersection of experimental work and model-based simulation. On the modeling side, advances in physics-informed machine learning and hybrid data- and physics-driven simulation approaches will be utilized to drastically reduce computational efforts while enhancing predictive capabilities. The nonlinear and coupled nature of the underlying physical problem makes optimization tasks especially challenging, both in simulations and experimentally. Machine learning methods will be used to overcome these barriers, with the aim of eventually replacing trial-and-error experiments or large Design-of-Experiment campaigns with solutions for inversely formulated questions.

The research group will be situated at the intersection of two existing research groups which focus on the experimental investigation of laser-based processing and physics-based simulation of the same.

Scientific contact person: Sabine Andergassen
Webpage of your present research group: https://informatics.tuwien.ac.at/orgs/e194-06, öffnet eine externe URL in einem neuen Fenster

Our research group leverages artificial intelligence and machine learning to uncover new physical phenomena in complex quantum systems and transform them into real-world technologies. With applications in quantum-enhanced sensing, imaging, communication and networks, we aim to design technologies and systems that outperform classical counterparts, with blueprints ready for implementation in existing laboratories. To realize this vision, we will integrate knowledge from scientific literature, develop highly efficient auto-differentiable simulators, and employ advanced machine-learning exploration techniques. These methods will open new directions in quantum science and contribute to the development of next-generation quantum hardware. By bridging quantum physics and computer science, our interdisciplinary approach drives the creation of transformative technologies that push the limits of what is currently achievable.

Our research group will be embedded within the Research Unit Machine Learning at TU Wien Informatics.

Scientific contact person: Paweł W. Woźniak

Dark patterns are manipulative design practices that influence users to make decisions that may not align with their best interests. Commonly found in social media, e-commerce, and gaming platforms, these deceptive interfaces exploit cognitive biases, limit user choices, and undermine informed decision-making. Despite regulations such as the EU’s Digital Services Act and California’s Consumer Privacy Act, users face challenges in identifying and navigating these subtle and sophisticated practices. As a result, dark patterns have emerged as a critical research focus in HCI, with an urgent need to mitigate their negative impacts.

Artificial intelligence and machine learning present transformative opportunities to detect and address dark patterns. Advanced AI models in computer vision and natural language processing can analyse visual and textual elements to identify manipulative features in real time. By proactively flagging such practices and providing immediate feedback, AI systems can shift the burden of detection from users to technology, enabling safer and more trustworthy digital interactions.

The proposed research group would focus on integrating HCI approaches with advanced AI methods to better understand dark patterns and develop tools for their real-time detection and mitigation. Situated within the Research Unit Human-Computer Interaction at TU Wien, the group would benefit from the unit's outstanding expertise in evaluation methods, social interactions, and accessibility. This embedding offers a robust foundation for conducting interdisciplinary research that combines human-centred design principles with computational techniques, establishing Vienna as a leader in ethical and responsible design research.

Scientific contact person: Johannes Böhm
Webpage of your present research group: https://www.tuwien.at/mg/geo/hg

The research unit Higher Geodesy at the Department of Geodesy and Geoinformation deals with dynamic processes in the Earth system as manifested in changes of its form, rotation, gravity field, and atmosphere. We use space geodetic observations to and from satellites, most notably the Global Navigation Satellite Systems (GNSS).

The Vienna Research Group for Young Investigators will develop cutting-edge AI/ML concepts, approaches, and methods to advance and improve geodetic observations, models, and products. For example, we do have observations from tens of thousands of geodetic GNSS stations and potentially billions of smartphones with GNSS receivers and other sensors. This huge amount of data, together with machine learning techniques, will significantly enhance our understanding of the system Earth and the atmosphere in particular. Moreover, meteorological data along with geodetic observations can improve geodetic and geophysical models, and machine learning techniques applied with gravity field observations bear the potential of extending our knowledge in global water storage. We solicit applications by scientists with experience in geodetic Earth system modeling with AI and experience/plans on the advancement of AI/ML concepts, approaches and methods for the discipline.

Scientific contact person: Allan Hanbury
Webpage of your present research group: https://informatics.tuwien.ac.at/orgs/e194-04, öffnet eine externe URL in einem neuen Fenster

Short description of your research field and where you would like to embed the group: The current scientific publication process is essentially a digital version of the traditional hardcopy publication process, with the main unit of publication being online PDFs. The aim of this VRG is to develop methods and technology to revolutionise the scientific publication process through the use of AI and Data Science. These developed methods should support rapid access to scientific information, such as automated State-of-the-Art reports or Systematic Reviews. They should also support multiple types of scientific output, such as code, data, and lab protocols. Working together with multiple scientific disciplines is key to developing approaches that are widely accepted by the scientific community.

Scientific contact person: Silvia Miksch
Webpage of your present research group: https://www.cvast.tuwien.ac.at/, öffnet eine externe URL in einem neuen Fenster

The Visual Analytics  Group (CVAST) is a research unit of the Institute of Visual Computing and Human-Centered Technology at TU Wien Informatics. Visual Analytics as ”the science of analytical reasoning facilitated by interactive visual interfaces” aims to enable the exploration and the understanding of large and complex data sets by intertwining interactive visualization, data analysis, human-computer interaction, and cognitive and perceptual science.
CVAST  is concerned with computer-tools, methods, and concepts that support humans in coping with complex information spaces. We strive to make complex information structures more comprehensible, facilitate new insights, and enable process of information and knowledge discovery. At this, human abilities as well as users' needs and tasks are central issues to assist in situations where complex decisions need to be made. Data and information are a broad field – we focus particularly on the temporal, spatial and spatial-temporal dimensions.

The new WWTF VRG group will complement CVAST by means of mixed-initiative Artificial Intelligence (AI)  and Visualization strategies to push our problem-solving and decision-making capabilities beyond the traditional AI and Machine Learning (ML) approaches (in the vision of Human-Centered AI, explainable AI, and interactive ML). Such an approach strategically harnesses the collective strengths of human expertise and cutting-edge AI/ML and Visualization techniques to forge a path toward more sophisticated and dynamic problem-solving methodologies. Possible application domains are Health Care/Medicine and Cultural Heritage.  

The WWTF VRG will be embedding in the research unit "Visual Analytics" # 193-07

Scientific contact person: Thomas Eiter

Webpage of your present research group:  Knowledge-Based Systems Group, öffnet eine externe URL in einem neuen Fenster

This research group shall provide foundations and practical tools for systems that deal with knowledge dynamically and adaptably. To that end, the researchers in this group will investigate how machine learning, logic, and theory change can be jointly employed to make knowledge representation systems more adaptable and contextualizable.

In the context of the VRG25 call, the group will be dedicated to developing novel ways of dealing with scientific theories and knowledge from a knowledge-representation perspective. The vision is that scientific theories and scientific knowledge, like experimental designs, research data, and scientific infrastructure, become first-class objects in scientific knowledge representation systems, making them easier to understand, exchange, and discuss. Furthermore, because knowledge representation systems are capable of reasoning and adapting, those capabilities will carry over to scientific knowledge, ultimately supporting the advancement of theories and planning future research. With a partner from science, such as cognitive sciences, life science, or physics, it shall be explored and evaluated how the foundational results will be put into usable tools in scientific practice.

Scientific contact person: Wolfgang Wagner
Webpage of your present research group: https://www.tuwien.at/en/mg/geo/rs

The use of Artificial Intelligence (AI) to model the Earth’s system has recently made big headlines, with some research teams and companies having demonstrating that AI-based forecast can be more accurate and faster than physics-based predictions at a fraction of the compute power. This has also spurred the expectation that it will be possible to use AI for combining Earth observation data collected by satellites with Earth system models to improve the spatial resolution of digital replica of the Earth to 1km and better. This is deemed an essential step for being able to model the combined effect of natural and anthropogenic processes, which is e.g. necessary for planning climate change adaptation and mitigation measures. However, the use of large AI models in Earth observation and Earth system modelling can be problematic for many reasons, including a tendency for overfitting, high correlation between input data sets, and the often strong sensitivity of AI models to unknown processes that are neither described by the Earth system models nor are captured by the satellite observations. These problems are compounded by the fact that AI model outputs are notoriously difficult to interpret, and do not offer easy routes for making them explainable. Therefore, we propose to establish a new Vienna Research Group that carries out basic research for making AI models used to model Earth observation data explainable. This is crucial in applications where transparency, interpretability, and accountability are essential, such as forecasting climatic extremes or quantifying the effectiveness of climate change adaptation and mitigation measures.

Scientific contact person: Stefanie Elgeti

Efficient experimental design under uncertainty is a fundamental challenge across scientific and engineering disciplines. Uncertainty arises from incomplete knowledge of complex systems, variability in experimental conditions, and constraints on available resources. Addressing these challenges requires foundational research into sequential decision-making processes to optimize experiments that balance performance, efficiency, and robustness. The research group on foundations of experimental design will focus on advancing the theoretical foundations and applicability of design of experiments. The group will develop novel methods that account for uncertainty in decision-making, incorporating multi-fidelity modeling and adapting to time-varying conditions. These approaches aim to integrate heterogeneous data sources effectively, improving resource allocation and experimental outcomes.

The research aims to bridge theoretical advancements in optimization and machine learning with realworld experimental needs. Mechanical engineering applications, such as structural testing, material characterization, and system optimization, will serve as key motivating domains. By addressing core challenges in uncertainty quantification, model selection, and data-efficient learning, the proposed research aims to improve the principles and practices of sequential experimental design. Ultimately, the group’s work will contribute to making experimental sciences more precise, efficient, and adaptable, with broad implications for both fundamental research and practical applications.

Scientific contact person: Fazel Ansari
Webpage of your present research group: https://www.tuwien.at/mwbw/im/pim

The Foundation Models in Manufacturing (F2M) research group aims to explore the thematic areas of NLP (Natural Language Processing) and KGs (Knowledge Graphs) in engineering context. The goal is to tackle tangible challenges associated with unstructured and multimodal data, in particular text, in industrial settings within the domain of smart manufacturing. Recent foundation language models, as a dominant paradigm for NLP, in synergy with KGs, represent a major step forward towards enhancing cognitive capabilities in cyber physical production systems (CPPS). These complementary technologies provide significant advantages in developing innovative, data-driven solutions to optimize industrial operations, improve policy and strategy selection, and enhance informed decision-making processes across industrial sectors, particularly production, maintenance and logistics.

The F2M group will be embedded within the Research Unit of Production and Maintenance Management (PIM, E330-06) at the Faculty of Mechanical and Industrial Engineering, TU Wien. This collaborative environment establishes NLP as a key contributor to technical language processing (TLP), tailoring it for technical engineering settings. We focus on enhancing the reliability, interpretability, and scalability as well as industrial deployment of foundation language models for engineering use cases (up to TRL 5) by incorporating high-quality explicit knowledge, formalizing implicit and experiential knowledge, and leveraging the symbolic reasoning provided by KGs. This research is relevant to sectors such as semiconductors, aviation, railways, and automotive, aligning with current ICT trends to address real-world manufacturing challenges. The research group also provides the opportunity to transfer scientific findings to the educational portfolio, under the motto of research-driven teaching, and introduce new courses on TLP and Foundation Models for bachelor and master students of the mechanical and digital engineering as well as cross-faculty courses to the business informatics and management.

Scientific contact person: Michael Wimmer
Webpage of your present research group: https://www.cg.tuwien.ac.at/, öffnet eine externe URL in einem neuen Fenster

The research unit Computer Graphics at the Institute of Visual Computing & Human-Centered Technology performs extensive fundamental and applied research in computer graphics. Our areas of expertise are modelling and rendering for computer graphics, visualization, visual computing, virtual environments, and color, with a focus on 3D modeling techniques.

The Vienna Research Group for Young Investigators will study 3D data analysis and generation, with a primary focus on developing cutting-edge methods in generative models and representation learning. We aim to create robust frameworks that effectively capture the geometric, structural, and texture-related intricacies of 3D data. By bridging the gap between theoretical innovation and practical application, our work supports the development of intelligent systems capable of accurate 3D modeling, simulation, and visualization. Our research not only addresses fundamental challenges in high-dimensional data representation but also explores interdisciplinary applications, such as immersive environments, virtual and augmented reality, and scientific computing. By leveraging advancements in machine learning, we strive to push the boundaries of how 3D data is analyzed, synthesized, and utilized in real-world scenarios.

Scientific contact person: Florian Michahelles
Webpage of your present research group: media.tuwien.ac.at, öffnet eine externe URL in einem neuen Fenster

Artifact-based Computing and User Research Unit
Focuses on integrating physical and digital prototyping with human-centered design to develop seamless collaborations between humans and AI. The unit emphasizes mutual enhancement, fostering co-learning and maintaining user autonomy to ensure AI systems amplify human skills rather than replacing them.

Motivation
With automation, we might lose skills, e.g., being only the editor but not the writer of text with LLM. When we put the user in full control, there is the possibility for co-learning and amplifying skill. Thus, people should continue to preserve autonomy and control over algorithmic systems and their outputs and suggestions, by maintaining their self-determined agency when using AI systems. While it is easy to hand over control to a seemingly superior and unbiased AI system, ignoring the limits of AI systems is fraught with risks.

New Embedded Research Group
Explores the impact of warning cues in AI systems, integrating ICT and psychology perspectives. It aims to enhance user autonomy by studying how cues trigger constructive skepticism and empower users. Using qualitative and quantitative methods, the group will advance understanding of how to design AI systems that promote informed, self-determined interaction.

Scientific contact person: Ezio Bartocci
Webpage of your present research group: https://www.eziobartocci.com/team.php, öffnet eine externe URL in einem neuen Fenster

Research Field: The proposed Vienna Research Group will focus on integrating advanced machine learning methodologies into developing and analyzing autonomous cyber-physical systems (CPS). These systems form the backbone of essential domains, including self-driving vehicles, autonomous robotics, smart agriculture, Industry 4.0, and intelligent infrastructure. In particular, the research will emphasize machine-learning techniques, leveraging vast data generated by CPS/IoT systems to enhance predictive accuracy, model inference, and optimal control. We will give to scaling these methods for complex, large-scale applications while improving the interpretability and reasoning over high-level data features. The goal is to establish a foundation for cognitive CPS systems capable of lifelong learning.

Embedding: The group will be embedded within the Trustworthy Cyber-Physical Systems group (within the CPS Research Unit), where the research will build on established expertise in formal verification, safety assurance, and secure system design. The group will collaborate closely with interdisciplinary teams across TU Wien and contribute to advancing dependable, scalable, and intelligent CPS.

Scientific contact person: Thomas Gärtner

Our main research interests are efficient and effective machine learning and data mining algorithms. By efficiency we mean on the one hand the classical computational complexity of decision, enumeration, etc.problems but on the other hand also a satisfactory response time that allows for effectiveness. By effectiveness we mean how well an algorithm helps to solve a real world problem.

The aim of the proposed VRG is to advance the state-of-the-art of machine learning in terms of theory/algorithms and to demonstrate the practical effectiveness of the novel learning algorithms in scientific areas such as Chemistry, Material Science, Physics, Biology, or any other area of the Sciences. In line with the WWTF VRG25 call, interested candidates will  already have demonstrated their potential for making theoretical/algorithmic advances in machine learning and for successfully applying machine learning algorithms in a scientific research field.

Scientific contact person: Stefan Woltran
Webpage of your present research group: https://dbai.tuwien.ac.at/, öffnet eine externe URL in einem neuen Fenster

In this research group we shall study fine-grained complexity of counting problems and how to exploit these for counting-based techniques in order to efficiently solve reasoning problems in artificial intelligence. While machine learning has reached almost every aspect of society, current techniques fall short of correctness guarantees. In contrast, modern #SAT model counters can provably count solutions for instances beyond thousands of variables. The main focus of the QuantAI group shall, therefore, be on neurosymbolic approaches that aim to combine quantitative techniques with machine learning in order to leverage the best of both worlds. The overall goal is to bridge formal verification and machine learning by pioneering counting-based quality guarantees for AI. Given the breadth of this research, we foresee novel applications in data analysis, especially in safety-critical domains where reliability and confidence guarantees are crucial, but also in other areas of natural science like physics or computational biology.

The QuantAI group will be embedded into the databases and artificial intelligence group (Reserch Unit 192-02 / DBAI). Synergies are expected, as DBAI hosts experts in data management, counting, (fine-grained) complexity theory, as well as artificial intelligence.

Scientific contact person: Ivona Brandic

Machine learning has revolutionized society in recent years, allowing humanity to address problems previously considered intractable. The junction of data with large computational capabilities made this breakthrough possible. However, there are still problems at which machine learning cannot succeed for fundamental reasons, for instance those related to quantum physics. On the other hand, quantum computing promises enhanced computational capability as compared to traditional methods. The development of this technology has potential to enhance existing computational science beyond its current capabilities, through the attainment of quantum advantages. However, only few and specific examples are known where quantum advantages can be demonstrated. In this VRG, we will develop the machinery to amalgamate quantum computing and machine learning with the goal to extend the applicability of quantum machine learning to broad problems. To this end, variational approaches, which have demonstrated a renowned success record in classical computing, will be combined with the enhanced computational capabilities of quantum computing, including in the fault-tolerant regime. This research will escape from established general-purpose variational algorithms from the community of quantum computing. In contrast, we will focus on identifying cases where quantum machine learning can arguably provide advantages over traditional methods. Investigation on this topic will allow for an enhancement in the applicability of quantum computing.

Scientific contact person: Ioannis Giannopoulos

Webpage of your present research group: https://geoinfo.geo.tuwien.ac.at, öffnet eine externe URL in einem neuen Fenster

The amount of human geo-data is growing exponentially because of recent developments in areas of IoT and urban computing. At the same time, AI is taking off as a tool in spatio-temporal analysis providing new insights but lacking in transparency and ease of access. To address both the questions on the data and the analysis thereof, Digital Twins could offer a prime venue to enable accessible, reproducible, and explainable Geo-AI. However, Digital Twins are also a buzz word that requires more rigorous definition to underpin Geo-AI research. Nowadays, Digital Twins are specific artifacts created to represent one particular system instead of a generic tool that helps researchers to manage complexity in their work. The new group would develop a consistent theoretical approach for AI-driven Digital Twins and their application as a framework for accessible, reproducible, and explainable Geo-AI on concrete examples of human activity on a geographic scale.