Abstract

Quadrupeds jumping on rocky terrains, humanoid robots performing summersaults, robot manipulators folding laundry: it is an exciting time in the field of robotics.

Still, we do not have a robot filling the dishwasher at home or a humanoid robot building a scaffold on a construction site. Where do we really stand? What have been the enablers of the undeniable progress in robotics research in the last five year? And what is still lacking to build trustworthy AI robotics systems for the industry and our homes?

The talk will provide a perspective on pixel-to-action robot control, the role of physics simulation, the use of machine learning for providing unprecedented levels of fast but potentially faulty perception and motion planning, and the surge of new computational and mechatronics hardware that is meant to be tolerant to collisions and exploit physical contact with the environment, instead of fearing and avoiding it.

Alessandro Saccon received a PhD in control system theory from the University of Padova, Italy, in 2006. He also holds a degree with honors in computer engineering from the same University. Following his PhD, he held a research and development position at the University of Padova in collaboration with racing motorcycle company Ducati Corse. From 2009 to early 2013, he held a post-doctoral research position at the Instituto Superior Técnico, Lisbon, Portugal working on geometric numerical methods for solving trajectory optimization problems.

Alessandro has held visiting positions at the University of Colorado, Boulder (2003 and 2005), at the California Institute of Technology (2006), and at the Australian National University (2011). He is the author and co-author of more than 50 peer-reviewed scientific publications across his various research areas.

In this talk, I will give an overview on the development of autonomous driving technologies, and their evolution from classical modular stacks to end-to-end stacks. I will highlight the safety assurance challenges stemming from their monolithic structures and discuss how they are being addressed by their embedment within safety halos.

Finally, I will step back to reason on the broader deployment of autonomous vehicles within mobility systems. I will propose a path forward that is safe—not only in the technical sense, but also in a societal one: A path that incorporate halos from the social sciences that help us avoid the engineer trap and ensure deployments that transcend technology-for-technology paradigms and are instead rooted in the wellbeing of humans and the planet.

Mauro Salazar is an Assistant Professor in the Control Systems Technology section at Eindhoven University of Technology (TU/e) in the Netherlands, where he leads the MOVEMENT Research Group. 

He is also affiliated with Eindhoven AI Systems Institute (EAISI). He earned his PhD in Mechanical Engineering from ETH Zürich in collaboration with the Ferrari Formula 1 team in 2019, before moving to Stanford University for a postdoctoral position on future mobility systems until 2020.
His research focuses on optimization models and methods for cyber-socio-technical systems and control, with applications on sustainable energy and mobility systems that foster justice and wellbeing. He was awarded the ETH Medal for his MSc and PhD theses, and his papers earned several awards, including the Best Student Paper award at the 2018 IEEE Intelligent Transportation Systems Conference and at the 2022 European Control Conference, as well as the Best Paper Award at the 2024 IEEE Vehicle Power and Propulsion Conference.
He was nominated for TU/e Young Researcher Award in 2022 and 2025, and, in 2024, for membership in The Young Academy of KNAW. In 2025, he received the Best Master Teacher Award from the Department of Mechanical Engineering.

In our paper, we propose a model-agnostic post-hoc explanation procedure devoted to computing feature attribution. The proposed method, termed Sparseness-Optimized Feature Importance (SOFI), entails solving an optimization problem related to the sparseness of feature importance explanations. The intuition behind this property is that the model's performance is severely affected after marginalizing the most important features while remaining largely unaffected after marginalizing the least important ones. Existing post-hoc feature attribution methods do not optimize this property directly but rather implement proxies to obtain this behavior. Numerical simulations using both structured (tabular) and unstructured (image) classification datasets show the superiority of our proposal compared with state-of-the-art feature attribution explanation methods.

Isel Grau received her Ph.D. in Computer Science from Vrije Universiteit Brussel (VUB), Belgium, where her research focused on machine learning interpretability and semi-supervised classification. During her postdoctoral work at the Artificial Intelligence Laboratory of VUB, she collaborated on interdisciplinary projects with organizations such as the Interuniversity Institute of Bioinformatics Brussels, Universitair Ziekenhuis Brussel, and Collibra BV.

Isel has co-authored over 75 peer-reviewed publications and has been actively involved in organizing thematic workshops. She frequently serves as a reviewer for leading AI conferences and journals, including ECAI, BNAIC, LION, and IEEE Transactions on Fuzzy Systems. She has also been a visiting researcher at institutions such as Warsaw University of Technology and the University of Lisbon.

Dominique Fürst is a Boundary Spanner in Artificial Intelligence and moderator of the EAISI Research track. She is an alumna of the EAISI Academy professional education program in Data Science & AI, where she and her team received the Best Project Award for their work with Enexis, applying reinforcement learning to enable smart charging strategies for electric vehicles.

In her role as Boundary Spanner in AI, Dominique connects students, researchers, and external partners around AI-driven challenges or real-world problems that can be addressed using AI. In close collaboration with diverse stakeholders, she translates these into meaningful challenges within Challenge-Based Learning (CBL) education that align with scientific research. She facilitates multidisciplinary collaboration and ensures that societal and academic goals are brought together to create tangible impact by empowering the next generation of AI-enabled engineers and advancing responsible innovation.

In addition to her AI-focused work, Dominique coordinates the Boundary Spanner Program at TU/e innovation Space, supporting a growing network of TU/e staff who span boundaries across disciplines, institutes, and sectors. TU/e innovation Space is the center of expertise for CBL and student entrepreneurship at Eindhoven University of Technology. It serves as a learning hub for educational innovation and an open community where students, researchers, industry, and societal organizations exchange knowledge and co-create responsible solutions to real-world challenges.

Our paper examines the critical role of trust in artificial intelligence systems that manage multi-source energy integration in sustainable urban environments. Building on our ongoing machine learning analyses of photovoltaic (PV) and building-integrated PV systems, we investigate how AI can optimize interconnected energy networks. Our design-driven research adopts an interdisciplinary approach to model the complex relationships between diverse generation sources (solar, wind, nuclear, thermal), energy storage systems, and consumption sectors (residential, commercial, industrial) at appropriate urban scales. We demonstrate how these energy sources connect through existing and new infrastructure networks for both electricity and heat distribution. The complexity of these systems necessitates advanced AI-driven analytics for optimal operation—defined through multiple objectives including flexibility, sustainability, and cost-effectiveness. This research addresses the central question: "How can we design trustworthy AI systems that balance automation with human oversight in complex urban energy networks?" We propose a comprehensive framework for establishing trust in AI-powered energy integration hubs, focusing on transparency in decision-making, validation through real-world testing, and human-AI collaborative approaches. Through case studies utilizing multi-source energy data, we demonstrate how properly designed trust mechanisms enhance energy efficiency, grid resilience, and stakeholder acceptance while addressing concerns regarding data privacy, algorithmic bias, and system reliability. We invite collaboration from stakeholders willing to share operational energy system data to further validate and refine our models. The findings suggest that trustworthy AI implementation is essential for maximizing the potential of integrated energy systems in sustainable urban development, particularly as cities face increasing challenges from climate change and growing energy demands.

Guang Hu obtained his Ph.D. with distinction from Tsinghua University in 2019. Prior to joining TU/e, he held positions as a postdoctoral researcher at the Karlsruhe Institute of Technology in Germany and as a researcher at the Paul Scherrer Institut (PSI) in Switzerland, building a strong international research profile.

Guang has published in several respected journals in the fields of computational modeling, machine learning applications, and sustainable/nuclear energy. His research contributions span theoretical advancements in modeling techniques and practical applications for energy systems. He actively participates in international research collaborations and serves as a reviewer for journals in his field, contributing to the academic community's knowledge development and quality assurance.

Valentina Breschi received her B.Sc. in Electronic and Telecommunication Engineering and her M.Sc. in Electrical and Automation Engineering from the University of Florence (Italy) in 2011 and 2014, respectively. She received her Ph.D. in Control Systems from IMT School for Advanced Studies Lucca (Italy) in 2018, being a visiting scholar at the University of Michigan (USA). From 2018 to 2023, she was with Politecnico di Milano (Italy), first as a post-doctoral researcher and then as a junior assistant professor from 2020 to 2023. In 2023, she joined the Control Systems Group at TU/e as an Assistant Professor.

Prediction-based decision-making systems are becoming increasingly prevalent in various domains. Previous studies have demonstrated that such systems are vulnerable to runaway feedback loops which exacerbate existing biases. The automated decisions have dynamic feedback effects on the system itself.

In this talk we will show how existence of feedback loops in the machine learning-based decision-making pipeline can perpetuate and reinforce machine learning biases and propose strategies to counteract their undesired effects. Understanding and mitigating these feedback mechanisms is a key step toward developing AI systems we can trust to make fair and reliable decisions.

Giulia De Pasquale is an Assistant Professor in the Control Systems group at TU Eindhoven since 2025. Before that she was a PostDoc and Lecturer at ETH Zurich.

She took the PhD in Control and Systems Engineering from the University of Padova, Italy,  in 2023.  During PhD, she spent an abroad period as a Visiting Research Scholar at the University of California, Santa Barbara.
She took both the Master Degree in Control Engineering and the Bachelor Degree in Information Engineering from the University of Padova in 2019 and 2017 respectively.  
During her Master Degree she took part in the Erasmus program, with a research experience at the Luleå Tekniska Universitet, Sweden, in collaboration with RI.SE.SICS Luleå and ABB Västerås working on system identification for the airflow in datacenters. She also spent six months at ETH Zurich, Switzerland,  working on her Master Thesis.

What would it take for us to invest our trust in AI wisely? This question is frequently taken to be about the trustworthiness of AI systems. Depending on one’s definition of trustworthiness, an AI system is worthy of our trust if it is reliable and/or meets certain ethical constraints, such as the absence of pernicious biases.

While we should design systems that are trustworthy in this way, in this talk I will shift our focus away from the qualities of AI systems to the qualities of the people developing, deploying, and interacting with these systems.

Trusting wisely, I will argue, requires that we develop the skills, capacities, and character traits – in short, virtues – that enable epistemic attunement to the capabilities and limitations of AI systems. Construed as a techno-epistemic virtue, trust is a cultivated, context-sensitive capacity to discern when, how, and to what extent we can rely on AI outputs.

Barend is an assistant professor of ethics at Tilburg University, where he does research on AI ethics and human-centered technology. He is especially interested in the role technology plays in (re-)producing racial, gender, and other injustices.

Barend de Rooij joined the Department of Philosophy in January of 2022 after completing a joint PhD in Philosophy at the University of Sheffield (UoS) and the University of Groningen (UG). In his doctoral research, which was supervised by Boudewijn de Bruin (UG) and Miranda Fricker (NYU), he studied the notion of group character, or the idea that collectives can instantiate virtuous or vicious character states qua group. 

Before moving to Sheffield, he completed an MA in Philosophy and European Studies at KU Leuven. He also holds an MLitt in Philosophy from the University of St. Andrews. During his  BA at Leiden University College, he spent some time studying philosophy at Rutgers University, where he discovered my passion for analytic philosophy.