The Speakers

Kirstin Petersen is an Associate Professor and Aref and Manon Lahham Faculty Fellow in the ECE Dept. at Cornell University. Her lab is focused on design and coordination of robot collectives able to achieve complex behaviors beyond the reach of an individual, and corresponding studies on how social insects do so in nature. Petersen did her postdoc at the Max Planck Institute for Intelligent Systems and her PhD at Harvard University with the Wyss Institute. Her graduate work was featured in and on the cover of Science, she was elected among Robohub’s 2018 top 25 women to know in robotics, and received the Packard Fellowship in Science and Engineering in 2019, the NSF CAREER award in 2021, and the Cornell CoE Research Excellence Award in 2022.

Environmentally-Mediated Coordination in Swarms

Natural swarms exhibit sophisticated colony-level behaviors with remarkable scalability and error tolerance. Their evolutionary success stems from more than just intelligent individuals, it hinges on their morphology, their physical interactions, and the way they shape and leverage their environment. I will discuss some examples from our lab related to how bees, termites, and robot swarms can leverage their shared environment to achieve distributed coordination.

Andy Philippides and Rachael Stentiford

Andy is a Professor of Biorobotics at the University of Sussex where he co-directs the Centre for Computational Neuroscience and Robotics, an interdisciplinary research group at the interface of Artificial Intelligence and Neuroscience, and be.AI: the Leverhulme Doctoral Centre for Biomimetic Embodied AI. His research aims to both better understand intelligence and develop novel AI and biorobotic algorithms, by considering intelligence as an active process in which adaptive behaviour emerges from the interaction of body, brain and environment. This is exemplified by his work on robotic visual navigation and exploration inspired by the remarkable visual navigation and learning abilities of ants and bees.

Rachael is a Research Fellow at the University of Sussex where she works on spiking neural network models of the Insect head direction system. Prior to Sussex she worked as a Postdoc at the Bristol Robotics Laboratory working on Mammalian models of head direction. With a background in both experimental and computational neuroscience methods, she is interested in interdisciplinary approaches to exploring visual navigation across species.

Insect visual navigation in natural environments: lessons from small-brained cognition

Visual navigation is an important task for autonomous robots and animals. However, the natural world is a complex and dynamic visual environment. Despite having small brains (~10^6 neurons) and low-resolution eyes (~10^3 pixels), insects learn to navigate in highly dynamic environments extremely rapidly and robustly. What lessons should engineers learn from these animals? Such remarkable learning is possible because cognition originates not from the brain in isolation, but emerges from animals interacting with the environment through specialised visual systems and active vision strategies.

In this talk we highlight the neural, sensory and behavioural organisation of two navigation tasks: maintaining an estimate of head direction and route following. In both cases we have built spiking neural networks (SNNs) inspired by the connectivity of insect brains, and show how these networks can be embodied on small robots. These SNN models can track head directions or navigate routes through dynamic visual environments without requiring complex processing to recognise specific objects or to specify when or what to learn. Thus the algorithms are efficient and all computation can be performed on-board the small robots, highlighting the potential for low-energy dedicated neuromorphic circuitry.

Martin Pearson

Dr. Martin Pearson is a Senior Lecturer at the University of the West of England, Bristol and theme leader for Neurorobotic and Biomimetic research at the Bristol Robotics Laboratory. He holds a BEng. In Electronic and Electrical Engineering from the University of Manchester and MSc. and PhD. from University of the West of England, Bristol. His research interest lies at the intersection of robotics, animal behaviour, neuroscience and AI and has published this work in over 100 peer reviewed articles.

Navigating the world through touch: whiskered robotic models of rodent CNS

The Rodents and many other mammals use their tactile facial whiskers to navigate the world, enabling them to thrive in confined and visually occluded habitats. The potential for mobile robots to exploit a similar capability has motivated a long running biomimetic study of whisker based touch. This has encompassed models of the electro-mechanical structure and morphology of the sensory apparatus itself, the ethology of numerous species of whiskered animals, and the underlying neural mechanisms of tactile and spatial cognition. What has been revealed is that the whisker sensory system is instrumental in a large number of cognitive processes and is represented in most regions of the brain. It has also been shown that modelling the neuro-ethology of rat whisker exploration can inform intrinsically safe robot tactile interaction and robust object/place recognition. Highlights from this broader study will be presented along with a roadmap for future directions of investigation.

Cecilia Laschi

Cecilia Laschi is Provost’s Chair Professor of robotics at the National University of Singapore, leading the Soft Robotics Lab. She is Co-Director of CARTIN – Centre for Advanced Robotics Technology and Innovation. She is on leave from Scuola Superiore Sant'Anna, Italy, The BioRobotics Institute (Dept. of Excellence in Robotics & AI). She graduated in Computer Science at the University of Pisa and received a Ph.D. in Robotics from the University of Genoa. She received an Honorary Doctorate from the University of Southern Denmark, Odense, in 2023. She was JSPS visiting researcher at the Humanoid Robotics Institute of Waseda University, Tokyo, Japan.

Cecilia Laschi is best-known for her research in soft robotics, an area that she pioneered and contributed to develop at international level. She investigates fundamental challenges for building robots with soft materials, with a bioinspired approach which started with a study of the octopus as a model for robotics. She explores marine applications of soft robots and their use in the biomedical field, with a focus on eldercare. She has worked in humanoid and neuro-robotics, applying brain models in humanoid robots.

She is Editor-in-Chief of Bioinspiration & Biomimetics and Specialty Chief Editor of Soft Robotics in Frontiers in Robotics & AI. She is Editorial Board member of Science Robotics, IEEE Robotics & Automation Letters, International Journal of Robotics Research, and the Intelligent Robotics and Autonomous Agents (IRAA) Series of MIT Press. She serves as evaluators for the EC (incl. ERC programme), HFSP and national research agencies.

She is IEEE Fellow and member of other scientific societies, like AAAS (American Association for the Advancement of Science), I-RIM (Italian Institute of Intelligent Machines) and GNB (Italian National Group of Bioengineering). She is member of the IEEE Robotics & Automation Society (RAS), where she was elected twice as AdCom member and co-founded the Technical Committee (TC) on Soft Robotics. She founded and chaired the 1st IEEE-RAS International Conference on Soft Robotics (RoboSoft) in 2018, serving now in its Steering Committee. She was Co-Chair of Gordon Research Conference on Robotics 2024 and Program Chair of the IEEE/RSJ International Conference on Intelligent Systems – IROS 2024.

She co-founded the spin-off company RoboTech, in edutainment robotics.

Robotics and the embodied side of intelligence

The quest for providing robots with intelligence has long pursued the study of human intelligence. Implementing brain models into humanoid robots has shown that intelligent behavior can be achieved, especially for ancient, stably-evolved mechanisms of sensory-motor coordination. They are common to most animal species, with proper consideration of diverse morphologies and environments, and provide animals with the adaptive behavior necessary for survival and reproduction. It is today clear that such adaptive behavior is also shaped by the body morphology itself, and by its interaction with the environment. In this respect, on the robotics side, many control tasks can be simplified if such embodied intelligence is taken into account and proper robot design accounts for it, instead of leaving the whole control burden to the robot brain. Evidence of embodied intelligence are devised across species, beyond the animal kingdom. Designing robots with embodied intelligence requires including external interactions and providing the robot body with proper compliance to receive them, as made possible by modern paradigms of soft robotics. The challenge ahead is to mend the gap between robot body and brain, and to smoothly integrate brain-inspired intelligence with embodied intelligence, in our future robots.

Kenji Doya

Professor

Neural Computation Unit

Okinawa Institute of Science and Technology Graduate University

Kenji Doya is a Professor of Neural Computation Unit, Okinawa Institute of Science and Technology (OIST) Graduate University. He studies reinforcement learning and probabilistic inference and how they are realized in the brain. He took his PhD in 1991 at the University of Tokyo, worked as a postdoc at U. C. San Diego and the Salk Institute, and joined Advanced Telecommunications Research International (ATR) in 1994. In 2004, he was appointed as a Principal Investigator of the OIST Initial Research Project, and as OIST established itself as a Graduate University in 2011, he became a Professor and served as the Vice Provost for Research till 2014. He served as a Co-Editor in Chief of Neural Networks from 2008 to 2021 and currently serves as the President of the Japanese Neural Network Society (JNNS). He received the INNS Donald O. Hebb Award in 2018 and the age-group 2nd place at Ironman Malaysia in 2022.

Evolution, Reinforcement Learning, and Mental Simulation

Living creatures can take different forms of adaptation, from evolution of body, behaviors and learning capabilities, reinforcement learning of novel behaviors during lifetime, and model-based inference for real-time decision and planning. This talk will present our efforts to understand how these are realized in animals and to implement similar capabilities in robots.

Barbara Mazzolai

Barbara Mazzolai is the Associate Director for Robotics and the Director of the Bioinspired Soft Robotics Laboratory at the Istituto Italiano di Tecnologia (IIT), Genoa. From February 2011 to March 2021, she served as the Director of the IIT Center for Micro-BioRobotics (CMBR). She earned her Bachelor's degree in Biology (with Honours) from the University of Pisa, Italy, and obtained her Ph.D. in Microsystems Engineering from the University of Rome Tor Vergata. Between July 2012 and 2017, she held the position of Deputy Director for the Supervision and Organization of the IIT Centers Network. In 2017, she was a Visiting Faculty at the Aerial Robotics Lab, Department of Aeronautics, at Imperial College London. Since 2024, she has been a contract professor for a course in soft robotics in the Department of Mechanics at the Polytechnic of Milan.

Barbara Mazzolai is a member of the Scientific Advisory Board (SAB) of the Max Planck Institute for Intelligent Systems (Tübingen and Stuttgart, Germany), of the SAB of the Max Planck Queensland Centre (MPQC) for the Materials Science of Extracellular Matrices, as well as of the Advisory Committee of the Cluster on Living Adaptive and Energy-autonomous Materials Systems - livMatS (Freiburg, Germany).

Her research work revolves around bioinspired soft robotics, where she combines principles from both biology and engineering to advance technological innovation and scientific knowledge. She has been the Coordinator of several EU-funded projects in this field, including PLANTOID, GrowBot, and I-SEED. In May 2021, she began her European Research Council (ERC) Consolidator Grant titled "I-Wood," focusing on Forest Intelligence: robotic networks inspired by the Wood Wide Web. Throughout her career, she has received numerous awards for her contributions, including the Marisa Bellisario Award and the Medal of the Italian Senate. She is an author and co-author of more than 260 papers published in international journals, books, and conference proceedings.

Can Plants Serve as a Model for Developing Intelligent Robots?

TBD

Yasmine Meroz

School of Plant Science and Food Security,

Center for Physics and Chemistry of Living Systems,

Tel Aviv University, Tel Aviv, Israel

Yasmine investigates the physics underlying computational and behavioral processes in plants, such as decision-making, memory phenomena, and collective behavior. Yasmine has a B.Sc. in physics and mathematics, a M.Sc. in physics, and a Ph.D. in physical chemistry from Tel Aviv University. She then continued to a postdoc in physics at that Weizmann Institute, and a second postdoc in applied mathematics at Harvard, where she made a shift and studied memory and decision-making in cellular chemotaxis. This deeply influenced her view on intelligent behavior as a computational process, and became central to her work as an independent researcher.

Plant Tropisms as a Window on Emergent Memory and Computation in Distributed Systems

Plants survive in a harsh and fluctuating environment, optimizing their search for fluctuating nutrients, and predicting danger. They achieve this through complex response processes, such as decision-making, based on memory, or the capability to accumulate and compare past stimuli. For example, a plant shoot accumulates sensory information from various fluctuating light sources, decides which direction yields consistently most light for photosynthesis, and grows in that direction. While they have no brain, computations occur in a decentralized manner at the tissue level. Here we propose a reverse-engineering approach to investigating the underlying rules for the accumulation and integration of sensory inputs. Our theoretical model, based on response theory, predicts that plants respond to the sum of stimuli at short timescales, and to the difference in stimuli at longer timescales. We confirm this experimentally, and suggest that this process may be essential for navigational problem-solving capabilities of plants.

Paco Calvo

P.I. Minimal Intelligence Laboratory (MINT Lab)

University of Murcia (Spain)

Paco Calvo’s research interests range broadly within the cognitive sciences, with special emphasis on plant intelligence, ecological psychology, embodied cognitive science, robotics and AI. He uses time-lapse photography to explore perception-action and learning in plants. His scientific articles have appeared in Annals of Botany, Biochemical and Biophysical Research Communications, Frontiers in Neurorobotics, Frontiers in Robotics and AI, Journal of the Royal Society, Plant, Cell & Environment, Plant Signaling & Behavior, Scientific Reports, and Trends in Plant Science, among other journals. He is co-author with Natalie Lawrence of the popular book Planta Sapiens (W.W. Norton, 2023; Japanese translation with Kadokawa, 2023).

Soft, Green, and Ecological: Merging Plant-Inspired and Ecological Robotics

Bioinspired robotics can emulate the growth-based movement of plants, for example, through the use of pneumatic actuators that extend the tip of a soft, everting growbot. This growth is part of a broader system for movement guidance and control. My talk will explore how principles of Gibsonian ecological psychology can be applied to the active, directional control of a plant-inspired climbing growbot, which aims to reach a specific target, mimicking natural plant behavior. We will discuss how the ecological concepts of "information" and "control" elucidate the navigation of both plants and their robotic counterparts through their environments. I propose investigating the timing control of growbots that utilize energy flows, which could provide all the necessary information for guiding their phototropic and target-oriented behaviors. Importantly, the control laws do not necessitate optic energy flows. Reality's specification might not be confined to any single ambient energy array but could occur across sensory modalities in a higher-order format. To demonstrate perception through the growbot's global ambient energy array, I will present research on the selective foraging behavior of Cuscuta parasites, which integrate various cues like light, volatile substances, and acoustic signals. A working hypothesis suggests that a Cuscuta-inspired growbot could modulate its growth patterns using the global energy array. By combining plant-inspired and ecological robotics, we aim to develop an unconventional toolkit for the robotics community, potentially opening new avenues for research and application.

Letizia Zullo

Researcher

IRCCS Ospedale San Martino, Genova, Italy

Letizia Zullo is a neuroscientist investigating the neuroethology of motor control. After receiving her PhD from the University Federico II of Naples, she moved to the Hebrew University of Jerusalem and then to the Italian Institute of Technology studying the neural bases of Octopus motor control and its translation into bio-inspired robotics. She is currently Researcher at the IRCCS Ospedale Policlinico San Martino of Genova where she studies aquatic animals’ neurophysiology toward human health, surgery and prosthetics.

Control Architecture of Movements in the Octopus

The very large and complex nervous system of Octopus vulgaris supports the wide repertoire of flexible behaviors of this animal species. From a motor control perspective, it is challenging to understand how these capabilities are achieved in a soft-bodied animal where the number of controlled variables is not restricted, as in skeletal animals, by a fixed number of joints. Soft-bodied animals also manifest, during movements, considerable variations in their body shape and stiffness/softness ratio. All these properties are especially interesting in the field of soft robotics, but their full translation is hindered by the complexity of controlling motion in soft deformable elements, intrinsically carrying a high computational load.

The octopus’ solution to this problem arise from a unique brain, body and environment relationships where sensory information is not represented centrally by using a brain-to-body coordinates system typical of vertebrates. Indeed, the octopus shows the emergence of an embodied organization of flexible intelligent behaviors where motor actions results from both internal command and body-environment relationship. The octopus embodied principles can be a source of inspiration to designing autonomous soft robots in which complex behaviors emerge from the body dynamic physical and sensory interactions with the environment.