CNR – INM, Italy
The talk focus on intelligent control systems for different robotics applications, particularly manipulation with a low-cost low-accuracy tendon-driven 3D-printed small arm, and cooperative grasping and transportation of a bulky object executed by a swarm of innovative marine robots. Concerning the arm manipulation, two different approaches have been tested and evaluated: a stochastic control algorithm based on Belief Space Planning (BSP) and a neural controller that learned the control policy from the previous BSP tests, through an imitation learning scheme. The problem of swarm cooperation in a marine context has been then investigated, within the Italian BEASTIE (roBotic undErwater Autonomous Social Team for cooperative manipulation and IntelligencE) project, and suitable distributed control approaches based on BSP and Reynolds’ Boids model are being developed.
Enrica Zereik earned her Ph.D. in Space Robotics from the University of Genova in 2010. She joined the Italian National Research Council (CNR) in 2012 and is currently affiliated with the Institute of Marine Engineering (INM). Her research spans various topics in robotics and control, with a focus on space robotics and marine robotics. She has worked on numerous European and National research projects and currently coordinates the Italian PRIN project BEASTIE. She was the PI for CNR of the project HumaBeliefs and is the CNR Head of the heron@cnr Joint Lab since 2020. She serves as an Associate Editor for IEEE Robotics and Automation Magazine and the International Conference on Robotics and Automation 2024. She has co-authored about 100 papers and was awarded the IEEE Transactions on Control Systems Technology Outstanding Paper Award in 2017. She is a member of the IFAC Technical Committee on Marine Systems (TC 7.2).
TU Delft, Netherlands
The acquisition and self-improvement of novel motor skills is among the most important problems in robotics. I will discuss various learning techniques we developed that enable robots to have complex interactions with their environment and humans. Complexity arises from dealing with high-dimensional input data, non-linear dynamics in general and contacts in particular, multiple reference frames, and variability in objects, environments, and tasks. A human teacher is always involved in the learning process, either directly (providing data) or indirectly (designing the optimization criterion), which raises the question: How to best make use of the interactions with the human teacher to render the learning process efficient and effective? I will discuss various methods we have developed in the fields of supervised learning, imitation learning, reinforcement learning, and interactive learning. All these concepts will be illustrated with benchmark tasks and real robot experiments ranging from fun (ball-in-a-cup) to more applied (sorting products).
Jens Kober is an associate professor at the TU Delft, Netherlands. He worked as a postdoctoral scholar jointly at the CoR-Lab, Bielefeld University, Germany and at the Honda Research Institute Europe, Germany. He graduated in 2012 with a PhD Degree in Engineering from TU Darmstadt and the MPI for Intelligent Systems. For his research he received the annually awarded Georges Giralt PhD Award for the best PhD thesis in robotics in Europe, the 2018 IEEE RAS Early Academic Career Award, the 2022 RSS Early Career Award, and has received an ERC Starting grant. His research interests include motor skill learning, (deep) reinforcement learning, imitation learning, interactive learning, and machine learning for control.
Zurich University of Applied Sciences, Switzerland
Inspired by the structure and dynamics of biological brains, neuromorphic hardware has been used as a tool in computational neuroscience. Today, this technology has matured to enable scalable applications and a new challenge of programming neuromorphic devices to solve real-world sensing, learning, and control tasks is driving modern research in this field. In this talk, I will give a historic overview of the field and its applications, with an emphasis on event-based vision and robotics.
Yulia Sandamirskaya is the Head of the Research Center “Cognitive computing in life sciences” at ZHAW Zurich University of Applied Sciences. She has a PhD in Physics/Neural Computational from the Ruhr-University Bochum and was leading the Applications research team at the Neuromorphic Computing Lab at Intel and the research group “Neuromorphic cognitive robots” at INI, UZH/ETH Zurich. She was the coordinator of EU CSA NEUROTECH, shaping the Neuromorphic computing technology community in Europe and co-chaired the European Network for the Advancement of Artificial Intelligent Systems, EUCog.
ERA Chair holder, UNIZG-FER
While it is clear to many that merging AI and Robotics would open new extremely promising opportunities for basic research, technological development and for coping with our days dramatic societal challenges, the still open research issues are still significant and will require significant work. We will discuss the open challenges, the possible solution path, and the different research strategies that researchers and organizations may pursue to cope with them. Which are the main roadblocks? Which competences and skill will the next generation of researchers need?"
Prof. Fabio Bonsignorio is the ERA Chair in AI for Robotics and head of the AIFORS Research Group at the University of Zagreb. He is also the CEO and Founder of Heron Robots. He has held professorships at the Biorobotics Institute of the Scuola Superiore Sant’Anna in Pisa and the University Carlos III of Madrid. He is a Founding Director of euRobotics aisbl and has been the coordinator of the ShanghAI Lectures since 2013. He is a leading expert in Reproducible Research and Benchmarking in Robotics and AI. He coordinated the EURON Special Interest Group on Good Experimental Methodology and Benchmarking in Robotics and was a board member of EURON III. He is a pioneer in the applications of Blockchain technologies in Robotics. He is a senior member and a Distinguished Lecturer of IEEE/RAS,
TUM, Chair for Cognitive Systems, Germany
For machines to learn, and especially robots to learn from humans, machines should go beyond learning just the simple action of humans. The Next Frontier in AI I ascribe to is “Purposive-AI”, which should enable machines to learn with purposes. Making use of this new AI will change the way we interact with machines for a Better Future. I will give examples of this in my talk to illustrate the potential power of early applications of Purpose-AI in robot learning.
Prof. Gordon Cheng has made pioneering contributions in Humanoid Robotics, Neuroengineering, Artificial Intelligence for the past 20 years. Since 2010, Gordon Cheng has been holding the Chair for Cognitive Systems, which he also founded, as a part of the School of Computation, Information and Technology at Technical University of Munich (TUM), Germany. Since 2016, he has been the Program Director of the Elite Master of Science program in Neuroengineering (MSNE) of the Elite Network of Bavaria, a unique study program in Germany. He is also the coordinator of the CoC for Neuro-Engineering - Center of Competence Neuro-Engineering with the School. Formerly, he was the Head of the Department of Humanoid Robotics and Computational Neuroscience, ATR Computational Neuroscience Laboratories, Kyoto, Japan. He has also been designated as a Project Leader/Research Expert for National Institute of Information and Communications Technology (NICT) of Japan. He is also involved in a number of major European Union projects. Over the past ten years Gordon Cheng has been the co-inventor of approximately 20 patents and is the author of approximately 350 technical publications, proceedings, editorials and book chapters.