Justin Carpentier is a researcher at Inria and École Normale Supérieure, heading the Willow research team since 2023. He graduated from École Normale Supérieure Paris-Saclay in 2014 and received a Ph.D. in Robotics in 2017 from the University of Toulouse. He did his Ph.D. in the Gepetto team at LAAS-CNRS in Toulouse, working on the computational foundations of legged locomotion. In 2024, he received an ERC Starting Grant focusing on laying the algorithmic and computational foundations of Artificial Motion Intelligence. His research interests lie at the interface of optimization, machine learning, computer vision, simulation, and control for robotics, with applications ranging from agile locomotion to dexterous manipulation. He is also the leading developer and manager of widely used open-source robotics software, among them Pinocchio, ProxSuite, HPP-FCL, and Aligator.
website: https://jcarpent.github.io/
Title: Recent Progress on Differentiable Simulation and Optimization for Robotics
Over the past decades, optimization has emerged as a key enabler for many robotics applications, particularly humanoids and quadrupeds. While initial solutions largely relied on models, in today’s data-driven landscape, it is becoming increasingly common to seek data-driven extensions to classical approaches. Differentiable control architectures propose a principled way to achieve this evolution. In this talk, I will highlight recent contributions from our group towards this objective, notably our efforts to (i) lay the groundwork for the next generation of differentiable simulators for rigid and soft robotics and (ii) develop differentiable optimization solvers that are both fast and dependable. Hopefully, these contributions will catalyze the design of the next generation of data-driven planning and control methods for robotics.
Ludovic Righetti is an Associate Professor in the Electrical and Computer Engineering Department and in the Mechanical and Aerospace Engineering Department at New York University and holds an international chair in robotics at the Artificial and Natural Intelligence Toulouse Institute (ANITI). He currently is a vice-president of the IEEE Robotics and Automation Society. He received an Engineering Diploma in Computer Science and a Doctorate in Science from the Ecole Polytechnique Fédérale de Lausanne. He was previously a postdoctoral fellow at the University of Southern California and a group leader at the Max-Planck Institute for Intelligent Systems. His work has received several awards including the Georges Giralt PhD Award, IEEE RAS Early Career Award and Heinz Maier-Leibnitz Prize and NYU’s Jacobs Excellence in Education Innovation Award. His research focuses on the planning, control and learning of movements for autonomous robots, with a special emphasis on locomotion and manipulation. He is also interested in the broader societal impacts of robotics and AI and regularly works with international organizations on the topic, especially on issues related to peace and security.
Title: Policy optimization of contact-rich behaviors: lessons from legged locomotion
Over the past few years, it has become normal to see legged robots casually walk in conferences and videos of their impressive tricks have become viral. What are the technological breakthroughs that made this possible? This presentation will show how recent advances in policy optimization have made the design of versatile and robust locomotion controllers within everyone’s reach, leading to a rapid democratization of legged robots. It will then discuss how these ideas could be transferred to soft robots and outstanding challenges.
Mehmet Dogar is a Professor in Robotics at the School of Computer Science, University of Leeds. His research focuses on robotic object manipulation. He investigates taking a physics-based approach to developing planners, controllers, and perception systems for the manipulation of objects, particularly simultaneous manipulation of multiple-objects, as well as deformable/soft objects. Prof. Dogar is an Associate Editor for IEEE Transactions on Robotics and also the International Journal of Robotics Research. He was a Visiting Professor at ETH Zurich in 2023 and a post-doctoral researcher at CSAIL, MIT between 2013-2015. He received his PhD from the Robotics Institute at the Carnegie Mellon University in 2013.
Title: Model-reduction of deformable object dynamics for robotic manipulation
I will talk about robotic planners, controllers, and perception systems that use physics-based predictions about the motion of contacted objects. In my group at the University of Leeds, we are interested in developing such systems for cluttered scenes that include rigid and deformable objects. Predicting the dynamics of such high-dimensional objects (i.e., soft/deformable objects, as well as a collection of multiple in-contact rigid objects) is computationally extremely expensive. I will particularly talk about model-reduction approaches we investigated to address this problem.
Christian Duriez (Senior Member, IEEE) received his Ph.D. in robotics from the University of Evry and CEA in 2004. After a postdoc at CIMIT SimGroup in Boston, he joined Inria in 2006, focusing on interactive simulation of deformable objects and haptics, especially for surgical applications. He became Directeur de Recherche in 2014 and was a visiting researcher at Stanford University in 2018. He is currently CEO of Compliance Robotics, a spin-off from the DEFROST team he founded at Inria Lille. He contributed to the development of the open-source SOFA framework and co-founded InSimo. His research interests include soft robotics, real-time FEM, contact simulation, and haptic feedback.
Title: Motion Planning for Soft Robots with Contacts: A Compliance-Based Approach
In soft robotics, motion planning loses much of its relevance if it cannot handle contact with the environment. This talk presents an approach based on mechanical compliance, which provides a unified framework for both inverse kinematics and contact modeling. Compliance is computed using FEM, with reduced models and a Lagrangian formulation that naturally incorporates actuation, sensing, and contact. This allows for efficient, contact-aware motion planning by inverse kinematics. We will illustrate this approach with applications, including the navigation of a soft catheter driven by a robot inside blood vessels.
Dr. Jessica Burgner-Kahrs is Professor with the Departments of Mathematical & Computational Sciences, Computer Science, and Mechanical & Industrial Engineering, the founding Director of the Continuum Robotics Laboratory, and Associate Director of the Robotics Institute at the University of Toronto, Canada. She received her Diplom and Ph.D. in computer science from Karlsruhe Institute of Technology (KIT), Germany in 2006 and 2010 respectively. Before joining the University of Toronto, she was Associate Professor with Leibniz University Hannover, Germany and a postdoctoral fellow with Vanderbilt University, USA. Her research focus lies on continuum robotics and in particular on their design, modelling, planning and control, as well as human-robot interaction. Her fundamental robotics research is driven by applications in minimally-invasive surgery and maintenance, repair, and operations. Her research was recognized with the Heinz Maier-Leibnitz Prize, the Engineering Science Prize, the Lower Saxony Science Award in the category Young Researcher, and she was entitled Young Researcher of the Year 2015 in Germany. She was elected as one of the Top 40 under 40 in the category Science and Society in 2015, 2016, and 2017 by the business magazine Capital and elected one of 100 Young Global Leaders from the World Economic Forum in 2019. Jessica is a Senior member of the IEEE, a Distinguished Lecturer of IEEE Robotics & Automation Society, and serves as a senior editor for IEEE Robotics & Automation Letters.
Title: Leveraging Environmental Interaction
This talk advocates leveraging environmental contacts as opportunities to exploit additional passive degrees of freedom inherent in underactuated soft continuum robots. Highlighting results from our recent work, I present a motion planning algorithm that determines effective motion sequences for single-segment tendon-driven continuum robots to navigate complex, cluttered environments through deliberate contact interactions. I further discuss computational strategies, including heuristic-guided search and simplified curvature modelling. Finally, I outline current limitations and identify critical research directions needed to fully exploit contact-aided motion planning.
Dr. Anup Teejo Mathew received his Ph.D. degree in Mechanical Engineering from the National University of Singapore, Singapore, in 2019. He is currently a Postdoctoral Fellow at Khalifa University, Abu Dhabi, UAE. He has authored or co-authored several papers on topics including soft robotics modeling, underwater soft robotics, and electroactive polymers. His current research focuses on geometric mechanics-based modeling and differentiable simulation of soft and hybrid soft-rigid robots. He is the developer of SoRoSim, a differentiable MATLAB toolbox for soft-rigid robotic systems. His research interests include theoretical modeling of soft robotics and the design and development of underwater soft robotic systems.
Title: Motion Planning in Hybrid Soft-Rigid Robots Using Differentiable Strain-Based Dynamic Model
This talk will focus on the Geometric Variable Strain (GVS) model, a unified dynamic modeling framework for hybrid soft-rigid robots, where the soft components are modeled as slender, deformable rods. The GVS model extends the classical screw-based formulation of rigid body dynamics to capture the continuous deformation of soft structures. Recent advancements in this framework, particularly the derivation of analytical derivatives, have enabled differentiable simulation and gradient-based optimization. These developments are implemented in SoRoSim, a MATLAB-based, differentiable simulation toolbox for hybrid soft-rigid robots. In this talk, I will discuss how these modeling tools and differentiable formulations can be leveraged in various trajectory optimization algorithms, including direct collocation (NLP), Differential Dynamic Programming (DDP), and Model Predictive Control (MPC), to efficiently generate optimal motions for hybrid soft-rigid robotic systems.
Josie Hughes is an Assistant Professor at EPFL where she established the CREATE in 2021. She undertook her undergraduate, masters and PhD studies at the University of Cambridge, joining the Bio-inspired Robotics Lab (BIRL). Following this, she worked as a postdoctoral associate at the CSAIL, MIT in the Distributed Robotics Lab. Her research focuses on developing novel design paradigms for designing robot structures that exploit their physicality and interactions with the environment. This includes the development of robotic hands, soft manipulators and locomoting robots, and automation systems for applications focused on sustainability and science.
Title: Exploiting the body for control of soft robots
By exploiting the design, sensor placement and fusion algorithms, this talk will explore how body properties and dynamics can assist in motion generation, as we move from open motion to constraint, contact based ones.
Cosimo Della Santina (Senior Member, IEEE) received the Ph.D. degree (cum laude) in robotics from the University of Pisa, Pisa, Italy, in 2019. From 2017 to 2019, he was a visiting Ph.D. student and a Postdoc with Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology. He was a Senior Postdoc and a Guest Lecturer with the Department of Informatics, Technical University of Munich, in 2020 and 2021, respectively. He is currently an Associate Professor with TU Delft, Delft, The Netherlands, and a Research Scientist with German Aerospace Institute (DLR), Munich, Germany. His research interest includes providing motor intelligence to physical systems, focusing on elastic and soft robots. Dr. Santina was the recipient of several awards, including the euRobotics Georges Giralt Ph.D. Award in 2020 and the IEEE RAS Early Academic Career Award in 2023. He was a recipient of a NWO VENI. He is involved as PI in a number of European and Dutch Projects, he is the coDirector of Delft AI Lab SELF.
Title: Towards closed form {learned model}-based controllers for soft robots
Dr. Thomas George Thuruthel is an Assistant Professor in the Department of Computer Science at University College London. He received his B.Tech. degree in Mechanical Engineering from the Indian Institute of Technology Hyderabad, India, in 2013, followed by a Master's degree in BioRobotics from Waseda University, Tokyo, Japan, in 2015, and a Ph.D. degree with Honors from Scuola Superiore Sant’Anna, Pisa, Italy. He served as a postdoctoral researcher at the University of Cambridge from 2019 to 2022. His research interests include modeling and control, dexterous manipulation, and soft sensing. Currently, he is focusing on the control of soft robots, design optimization of soft-bodied systems, development of novel soft sensors, and dexterous manipulation using visuo-tactile information.
Title: Modelling soft bodied systems under contact
The modeling of soft-bodied systems has progressed significantly in recent years, with both model-free and model-based techniques providing accurate and reliable solutions. However, most of these methods are limited to no-contact scenarios and cannot accommodate contact physics. Even methods capable of simulating contact mechanics face challenges in integrating contact sensors into their computational models. This talk briefly introduces the challenge of modeling soft interactions and explores potential model-free techniques for handling interaction dynamics.
Moritz Bächer is the Associate Lab Director of Disney’s Zurich-based robotics team, where he leads a strategic program focusing on the development of novel model- and learning-based tools for the design and control of believable robotic characters. His core expertise is the optimal design and control of both soft and rigid systems, using a combination of differentiable simulation and reinforcement learning. Prior to joining Disney, Moritz received his Ph.D. from the Harvard School of Engineering and Applied Sciences and his master’s degree from ETH Zurich
Title: The Role of Soft Robotics in Disney’s Robotic Character Platform
At Disney, we are building artist-centric tools that provide creative control of robotic characters consisting of both rigid and flexible components. In this talk, I will provide insight into our model- and learning-based tools that are part of our robotic character platform, enabling us to rapidly design expressive new characters. I will discuss the role of differentiable simulation in the optimal characterization, control and design of rigid-flexible hybrid systems and will outline the role that learning-based tools will play in this context using a freely walking robotic character as an example.
Robert Katzschmann is an Assistant Professor of Robotics at ETH Zurich, where he leads the Soft Robotics Lab, focusing on soft, musculoskeletal, and bio-hybrid robots that interact safely with humans and the environment. Drawing inspiration from biological systems, his work employs soft, compliant materials to create lifelike, adaptive robots. Before joining ETH Zurich, Robert served as Chief Technology Officer at Dexai Robotics and as a Senior Applied Scientist at Amazon Robotics. He received his Ph.D. in Mechanical Engineering from MIT in 2018, and his research has been featured in The New York Times, BBC, and more. A TED Fellow since 2022, Robert’s contributions are widely recognized across the robotics community, including his editorship for journals and conferences such as IJRR, ICRA, IROS, RoboSoft, npj Robotics, and RSS. His lab collaborates extensively with the Center for Robotics (RobotX), the ETH AI Center, and the ETH Max Planck Institute Center for Learning Systems.
Title: Planning, Modeling and Control of Life-like Musculoskeletal Robots
Living robots represent a new frontier in the planning, modeling and control of robotic systems. These robots incorporate biological living cells and synthetic materials into their design and require drastically different ways when trying to model and command their maturation and behavior. These bio-hybrid robots are dynamic and intelligent, potentially harnessing living matter’s capabilities, such as growth, regeneration, morphing, biodegradation, and environmental adaptation. Such attributes position bio-hybrid devices as a transformative force in robotics, promising enhanced dexterity, adaptive behaviors, sustainable production, robust performance, and environmental stewardship. Nature’s musculoskeletal design serves as inspiration for both artificial and living robots. In this talk, I will explore some of our recent advances in electrohydraulic musculoskeletal robots, which employ electrohydraulic actuators to produce lifelike muscle contractions and adaptive motions. I will also discuss our vision-controlled inkjet printing for robotics, xolographic biofabrication for biohybrid swimmers, and learning-based control of musculoskeletal systems. Together, these projects showcase how musculoskeletal, bio-hybrid, and computational techniques are opening new frontiers in robotics interaction and manipulation.
Mattia Gazzola is the Charles Conrad Kritzer Associate Professor in the Mechanical Science and Engineering Department at the University of Illinois Urbana-Champaign. He joined UIUC in Fall 2016 after a postdoc at Harvard and a PhD at ETH Zurich. His work lies at the interface between mechanics, biology, robotics, and computing. His studies were awarded with the ETH Medal, Early and Advanced Swiss National Science Foundation Fellowships, NSF CAREER, and featured on the cover of several scientific journals including Science, Nature, PNAS, PRL. He is the Lead PI and co-director of the center-scale NSF Expedition "Mind in Vitro–Computing with Living Neurons".
Title: Modeling and control of biological and biohybrid soft arms
Through an approach based on assemblies of Cosserat rods, we illustrate the modeling of soft arms characterized by intricate muscle architectures and their control through a blend of topology, passive mechanics, and learning algorithms. We demonstrate how costly control computations can be effectively offloaded to the physics of the body as it interacts with its environment.