Christian Duriez
Defrost research group, Inria Université de Lille, France
Bio: Christian Duriez received the engineering degree from the Institut Catholique d’Arts et Métiers of Lille, France, and the Ph.D. degree in robotics from University of Evry, France, His thesis work was realized at CEA/ Robotics and Interactive Systems Technologies followed by a Postdoctoral Position with the CIMIT SimGroup in Boston. He arrived at INRIA in 2006 to work on interactive simulation of deformable objects, haptic rendering and medical simulation. He is currently the head of DEFROST team, created in January 2015. He is also Chief Executive Officer in the Spin-off Compliance Robotics which develops soft demonstrators. His research interests are soft robot simulation, FEM modeling, inverse deformable model, contact and friction.
Title of the presentation: Modeling, Sensing, and Control of Soft Robot: towards new haptic devices
Abstract: Soft robotics opens new perspectives in the design of machines capable of safe, adaptive, and rich interactions with their environment. Unlike traditional rigid robots, soft robots exploit material compliance to create motions and physical interactions that are inherently safe for humans. In this presentation, we explore how advances in modeling, sensing, and control of soft robots open ways toward the development of a new generation of haptic devices. We will show how our modeling framework has enable the creation of soft tangible objects capable of complex, controllable motions, whose behavior can be further enhanced through integrated force sensing, in particular in the context of medical reharsal. Finally, we will present the design and control of a 5-degrees-of-freedom (5-DOF) haptic interface, illustrating how soft robotics can bring a new design approach in this domain.
Adrien Koessler
Control research group, LCFC, France
Bio: Adrien Koessler received the M.S. degree in 2015, from the Institut Français de Mécanique Avancée, Clermont-Ferrand, France, and the Ph.D. degree in engineering science from Université Clermont Auvergne, Clermont-Ferrand, in 2018. He is currently an associate professor in Robotics at Université de Lorraine/ENIM and LCFC. His research interests are robot modeling and model-based control applied to manipulation tasks. In particular, he is investigating collaborative robots and manipulation of deformable objects.
Title of the presentation: From multi-robot to human-robot deformable object manipulation.
Abstract: Although manipulation is vastly automated in industrial manufacturing and assembly processes, manual labour is heavily relied on when flexible objects are manipulated. Such objects can be cables, sheets of material, cloth, or even food products. This gap drove the rise of a new research area in the late 2010s, called Deformable Object Manipulation (DOM). Firstly, I will present my work on the deformation models for the manipulated objects and their integration into model-based deformation control, up to the milestone of multi-robot object shape control. However, the limitations in dexterity and situational awareness of robotic systems are amplified when dealing with flexible objects. Hence, the usual motivations for HRI are enhanced in a DOM context. This leads us to consider collaborative tasks where humans and robots manipulate flexible bodies. I will present the approach of the ANR HERR-MAN project on that topic, as well as its very first results.
Sven Lilge
Autonomous Space Robotics Lab, Canada
Bio: Sven Lilge is currently a Postdoctoral Fellow at the Autonomous Space Robotics Laboratory at the University of Toronto Robotics Institute, Canada. He obtained his Ph.D. in Computer Science working at the Continuum Robotics Laboratory at the University of Toronto under the supervision of Jessica Burgner-Kahrs. His research is primarily concerned with continuum robots focusing on their design, modeling, control and kinematic evaluation and particularly parallel structures.
Title of the presentation: Sensing and State Estimation for Continuum Robots: A Probabilistic Perspective
Abstract: One major challenge in continuum and soft robots is to efficiently capture and describe their non-linear shapes. While traditional modeling approaches for continuum robots often achieve good agreement with physical prototypes, they can exhibit significant residual errors. These errors typically stem from unmodeled effects, uncertainties in material properties or unknown interactions with the environment – issues that are particularly pronounced in human-robot interaction scenarios where contact occurs naturally. One promising approach to address this challenge is the application of probabilistic state estimation methods that fuse additional sensor information with a feasible prior model to estimate the shape of continuum robots, along with associated uncertainties. In this talk, we will discuss the application of Gaussian process regression to the state estimation of continuum and soft robots on SE(3) as an example of such methods.
Lukas Lindenroth
King’s College London, United Kingdom
Bio: Lukas Lindenroth is a Lecturer in Surgical Robotics in the Department of Surgical and Interventional Engineering within the School of Biomedical Engineering and Imaging Sciences. His research focuses on using novel robotics technologies such as soft robotics to develop medical devices for improved patient outcomes and clinician experience. Prior to this role he was postdoctoral research fellow at the Wellcome / EPSRC Centre for Interventional and Surgical Sciences at University College London. He obtained his PhD in Robotics from KCL where he developed soft robotic solutions for diagnostic ultrasound.
Title of the presentation: Enabling more efficient patient interaction in medical interventions through soft robot design, intrinsic sensing, and control.
Abstract: Soft robotics is an exponentially growing field with great promise for medical interventions. It involves the development of flexible and adaptable robotic systems that can effortlessly interact with patients and users alike. By leveraging soft materials and advanced control algorithms, soft robots offer the potential to revolutionize medical interventions, addressing existing challenges and improving patient outcomes. In this talk I will discuss the use of soft robot in medical applications, present challenges in existing designs and demonstrate through multiple use cases how with the help of parallel topologies, some of these challenges can be alleviated. I will, moreover, discuss the potential of intrinsic force sensing and control in fluid driven soft robots.
Mattia Franchi De’ Cavalieri
Soft Mechatronics for BioRobotics Lab , Scuola Superiore Sant’Anna, Italy
Bio: Mattia Franchi de’ Cavalieri is currently a Post-Doc researcher at The BioRobotics Institute, Scuola Superiore Sant’Anna in the Soft Mechatronics for BioRobotics Laboratory. He has a PhD in Neuroscience from University of Florence, collaborating with the IRCCS Stella Maris and the BioRobotics Institute. He received the BSc degree in Mechanical Engineering from the Sapienza University of Rome and the MSc degree in Biomedical Engineering (cum laude) from the Università Campus Bio-Medico, Italy.
Title of the presentation: Innovative 3D printed Soft Robots with integrated sensing abilities
Abstract: Soft robotics is becoming increasingly important worldwide, but how these innovative robots can sense their bodies and interact with the external environment remains an open question and a challenge, due to their continuously deformable structure and materials. Additionally, an important challenge lies in the automatization and production standardization of these robots to achieve higher performances and repeatability. The SOFTNESS project aims at developing a new generation of 3D printed modular soft robot with proprioception and exteroception abilities, starting from the recent advances in material science and signal modelling. We implemented fabric soft sensors and embedded the sensorial system directly in the fabrication process to create an easy-to-build and repeatable soft manipulator that can sense external touch and its spatial presence, allowing to reach higher performances in the control, the reproducibility, and interaction with the environment of soft robots in respect to the state of the art.
Felix Vanneste
ENSADLab, Ecole Nationale Supérieure des Arts Décoratifs, Paris
Bio: Félix Vanneste is currently a postdoctoral researcher in the Reflective Interaction group at EnsadLab in PSL Paris. He completed his PhD at Inria Lille in 2022, where he focused on the numerical simulation of mesostructured materials for soft robotic systems. His research centers on the use of finite element modeling (FEM) to design and optimize mechanical properties, particularly anisotropic behaviors, in order to create and control new degrees of freedom in soft robots. He combines these simulation approaches with additive manufacturing and metamaterials to physically realize and prototype novel robotic systems.
Title of the presentation: Leverage additive manufacturing capabilities to program matter’s compliance
Abstract: Soft robotics initially leveraged additive manufacturing (AM) primarily for producing molds used in silicone casting. Now with modern AM techniques and progress in material science, we can directly print a flexible material (like TPU for example) and have precise control over its distribution within a geometry. This enables the fine-tuning of compliance and the creation of advanced mechanical behaviors, such as anisotropy. This specific matter layout (periodic or not) creates what is called flexible mechanical metamaterial. In this presentation, I will introduce a specific metamaterial architecture and demonstrate how it can be used to create and control new behavior, how to actuate it as well as model it. These advances open new avenues for building soft robots with optimized, task-specific structures, enhancing performance while maintaining the inherent safety and compliance that make soft robots ideal for human interaction.
