Reduced-order models are central to motion planning and control of quadruped robots, yet existing templates are often hand-crafted for a specific locomotion modality. This motivates the need for automatic methods that extract task-specific, interpretable low-dimensional dynamics directly from data. We propose a methodology that combines a linear autoencoder with symbolic regression to derive such models. The linear autoencoder provides a consistent latent embedding for configurations, velocities, accelerations, and inputs, enabling the sparse identification of nonlinear dynamics (SINDy) to operate in a compact, physics-aligned space. A multi-phase, hybrid-aware training scheme ensures coherent latent coordinates across contact transitions. We focus our validation on quadruped jumping—a representative, challenging, yet contained scenario in which a principled template model is especially valuable. The resulting symbolic dynamics outperform the state-of-the-art handcrafted actuated spring-loaded inverted pendulum (aSLIP) baseline in simulation and hardware across multiple robots and jumping modalities.

Segmenting gas bubbles in multiphase flows is a critical yet unsolved challenge in numerous industrial settings, from metallurgical processing to maritime drag reduction. Traditional approaches — and most recent learning-based methods — assume near-spherical shapes, limiting their effectiveness in regimes where bubbles undergo deformation, coalescence, or breakup. This complexity is particularly evident in air lubrication systems, where coalesced bubbles form amorphous and topologically diverse patches. In this work, we revisit the problem through the lens of modern vision foundation models. We cast the task as a transfer learning problem and demonstrate, for the first time, that a fine-tuned Segment Anything Model (SAM v2.1) can accurately segment highly non-convex, irregular bubble structures using as few as 100 annotated images.

Recent advances in machine learning have begun to embed oscillatory network principles within neural architectures, aiming to enhance computational efficiency and robustness in time-series regression. Building on these developments, we take a step toward applying such principles to the learning of physical dynamics. We introduce Neural Linear Oscillator Networks (nLON): a vision-based framework that extracts compact latent representations of complex motion directly from image sequences. A convolutional autoencoder encodes position and velocity into a low-dimensional manifold, whose temporal evolution is governed by coupled linear mechanical oscillators driven by a linear combination of the inputs. This strong structural prior not only promotes sample efficiency and interpretability but also guarantees that the learned model remains a mechanical, Wiener-type system. From this formulation, we derive closed-loop controllers that ensure stable regulation. We focus on soft robots-systems whose nonlinear, continuous, and high-dimensional dynamics make them both a challenging and ideal testbed for our approach. Using tentacle robots in high-fidelity simulations and real-world experiments, we validate that our framework delivers accurate long-horizon predictions and consistently surpasses state-of-the-art baselines, achieving superior structural fidelity and final-step accuracy. Finally, we leverage the learned dynamics for model-based control, demonstrating in simulation that the resulting scheme achieves robust and reliable tracking.

Soft robots' ability to safely navigate complex environments motivates the development of algorithms for accurate environmental interaction assessment, enabling greater autonomy. Specifically, strain-based shape and force estimation of continuum robots with embedded soft sensors poses an open challenge mainly owing to continuous softness, anisotropic deformation, and non-linear properties. Mathematical description of deformable soft bodies and accurate estimation of external forces are crucial for achieving controllable and intelligent behaviors of these robots. In this paper, a kinetostatic strain-based modeling for rod-driven soft robots (RDSR) with embedded stretch sensors is proposed, which incorporates local strains, actuation variables, and external interactions. The strain model enables full shape estimation of the robot and prediction of strain variations in soft bodies. Building on this, we develop a force estimator based on predicted and measured sensor and actuator lengths to evaluate 3D external forces, accounting for both orthogonal and tangential components relative to the backbone. Moreover, we introduce a methodology using a novel ellipsoid representation to handle tangential forces that may become insensitive in certain singular configurations. This estimator allows us to either disregard such forces when they do not influence deformation or estimate them when they become observable. Our simulations and experiments demonstrate how this approach can be used to analyze the robot's configuration and successfully estimate external forces. Finally, it is demonstrated that when the continuum arm follows trajectories with higher strain sensitivity, tangential force estimation is significantly improved.

Learning from Interactive Demonstrations has revolutionized the way nonexpert humans teach robots. It is enough to kinesthetically move the robot around to teach pick-and-place, dressing, or cleaning policies. However, the main challenge is correctly generalizing to novel situations, e.g., different surfaces to clean or different arm postures to dress. This article proposes a novel task parameterization and generalization to transport the original robot policy, i.e., position, velocity, orientation, and stiffness. Unlike the state of the art, only a set of keypoints is tracked during the demonstration and the execution, e.g., a point cloud of the surface to clean. We then propose to fit a nonlinear transformation that would deform the space and then the original policy using the paired source and target point sets. The use of function approximators like Gaussian Processes allows us to generalize, or transport, the policy from every space location while estimating the uncertainty of the resulting policy due to the limited task keypoints and the reduced number of demonstrations. We compare the algorithm’s performance with state-of-the-art task parameterization alternatives and analyze the effect of different function approximators. We also validated the algorithm on robot manipulation tasks, i.e., different posture arm dressing, different location product reshelving, and different shape surface cleaning.

Using the language of dynamical systems, Imitation learning (IL) provides an intuitive and effective way of teaching stable task-space motions to robots with goal convergence. Yet, IL techniques are affected by serious limitations when it comes to ensuring safety and fulfillment of physical constraints. With this work, we solve this challenge via TamedPUMA, an IL algorithm augmented with a recent development in motion generation called geometric fabrics. As both the IL policy and geometric fabrics describe motions as artificial second-order dynamical systems, we propose two variations where IL provides a navigation policy for geometric fabrics. The result is a stable imitation learning strategy within which we can seamlessly blend geometrical constraints like collision avoidance and joint limits. Beyond providing a theoretical analysis, we demonstrate TamedPUMA with simulated and real-world tasks, including a 7-DoF manipulator.

Motor imagery (MI), an important category in electroencephalogram (EEG) research, often intersects with scenarios demanding low energy consumption, such as portable medical devices and isolated environment operations. Traditional deep learning (DL) algorithms, despite their effectiveness, are characterized by significant computational demands accompanied by high energy usage. As an alternative, spiking neural networks (SNNs), inspired by the biological functions of the brain, emerge as a promising energy-efficient solution. However, SNNs typically exhibit lower accuracy than their counterpart convolutional neural networks (CNNs). Although attention mechanisms successfully increase network accuracy by focusing on relevant features, their integration in the SNN framework remains an open question. In this work, we combine the SNN and the attention mechanisms for the EEG classification, aiming to improve precision and reduce energy consumption. To this end, we first propose a noniterative leaky integrate-and-fire (NiLIF) neuron model, overcoming the gradient issues in traditional SNNs that use iterative LIF neurons for long time steps. Then, we introduce the sequence-based attention mechanisms to refine the feature map. We evaluated the proposed noniterative SNN with attention (NiSNN-A) model on two MI EEG datasets, OpenBMI and BCIC IV 2a. Experimental results demonstrate that: 1) our model outperforms other SNN models by achieving higher accuracy and 2) our model increases energy efficiency compared with the counterpart CNN models (i.e., by 2.13 times) while maintaining comparable accuracy.

Human fingers exhibit remarkable dexterity and adaptability through a combination of structures with varying stiffness levels, ranging from soft tissues (low stiffness) to tendons and cartilage (medium stiffness) to bones (high stiffness). This paper focuses on the development of a robotic finger that emulates these multi-stiffness characteristics. Specifically, we propose utilizing a lattice configuration, parameterized by voxel size and unit cell geometry, to achieve fine-tuned stiffness properties with high precision. A key advantage of this approach is its compatibility with single-process 3D printing, which eliminates the need for manual assembly of components with varying stiffness. Using this method, we present a novel, human-like robotic finger and a soft gripper. The gripper is integrated with a rigid manipulator and demonstrated in pick-and-place tasks, showcasing its effectiveness.

Robotics is shifting from rigid, articulated systems to more sophisticated and heterogeneous mechanical structures. Soft robots, for example, have continuously deformable elements capable of large deformations. The flourishing of control techniques developed for this class of systems is fueling the need of efficient procedures for evaluating their inverse dynamics (ID), which is challenging due to the complex and mixed nature of these systems. As of today, no single ID algorithm can describe the behavior of generic (combinations of) models of soft robots. We address this challenge for generic series-like interconnections of possibly soft structures that may require heterogeneous modeling techniques. Our proposed algorithm requires as input a purely geometric description (forward-kinematics-like) of the mapping from configuration space to deformation space. With this information only, the complete equations of motion can be given an exact recursive structure which is essentially independent from (or “agnostic” to) the underlying reduced-order kinematics. We achieve this by exploiting Kane’s method to manipulate the equations of motion, showing then their recursive structure. The resulting ID algorithms have optimal computational complexity within the proposed setting, that is, linear in the number of distinct modules. Further, a variation of the algorithm is introduced that can evaluate the generalized mass matrix without increasing computation costs. We showcase the method applicability to robot models involving a mixture of rigid and soft elements, described via possibly heterogeneous reduced order models (ROMs), such as Volumetric FEM, Cosserat strain-based, and volume-preserving deformation primitives. None of these systems can be handled using existing ID techniques.

Learning from demonstration is a promising approach for teaching robots new skills. However, a central challenge in the execution of acquired skills is the ability to recognize faults and prevent failures. This is essential because demonstrations typically cover only a limited set of scenarios and often only the successful ones. During task execution, unforeseen situations may arise, such as changes in the robot's environment or interaction with human operators. To recognize such situations, this paper focuses on teaching the robot situational awareness by using a camera input and labeling frames as safe or risky. We train a Gaussian Process (GP) regression model fed by a low-dimensional latent space representation of the input images. The model outputs a continuous risk score ranging from zero to one, quantifying the degree of risk at each timestep. This allows for pausing task execution in unsafe situations and directly adding new training data, labeled by the human user. Our experiments on a robotic manipulator show that the proposed method can reliably detect both known and novel faults using only a single example for each new fault. In contrast, a standard multi-layer perceptron (MLP) performs well only on faults it has encountered during training. Our method enables the next generation of cobots to be rapidly deployed with easy-to-set-up, vision-based risk assessment, proactively safeguarding humans and detecting misaligned parts or missing objects before failures occur.

Soft robots promise inherent safety via their material compliance for seamless interactions with humans or delicate environments. Yet, their development is challenging because it requires integrating materials, geometry, actuation, and autonomy into complex mechatronic systems. Despite progress, the field struggles to balance task-specific performance with broader factors like durability and manufacturability—a difficulty that we find is compounded by traditional sequential design processes with their lack of feedback loops. In this perspective, we review emerging co-design approaches that simultaneously optimize the body and brain, enabling the discovery of unconventional designs highly tailored to the given tasks. We then identify three key shortcomings that limit the broader adoption of such co-design methods within the soft robotics domain. First, many rely on simulation-based evaluations focusing on a single metric, while real-world designs must satisfy diverse criteria. Second, current methods emphasize computational modeling without ensuring feasible realization, risking sim-to-real performance gaps. Third, high computational demands limit the exploration of the complete design space. Finally, we propose a holistic co-design framework that addresses these challenges by incorporating a broader range of design values, integrating real-world prototyping to refine evaluations, and boosting efficiency through surrogate metrics and model-based control strategies. This holistic framework, by simultaneously optimizing functionality, durability, and manufacturability, has the potential to enhance reliability and foster broader acceptance of soft robotics, transforming human-robot interactions.

While consciousness has been historically a heavily debated topic, awareness had less success in raising the interest of scholars. However, more and more researchers are getting interested in answering questions concerning what awareness is and how it can be artificially generated. The landscape is rapidly evolving, with multiple voices and interpretations of the concept being conceived and techniques being developed. The goal of this paper is to summarize and discuss the ones among these voices connected with projects funded by the EIC Pathfinder Challenge “Awareness Inside” callwithin Horizon Europe, designed specifically for fostering research on natural and synthetic awareness. In this perspective, we dedicate special attention to challenges and promises of applying synthetic awareness in robotics, as the development of mature techniques in this new field is expected to have a special impact on generating more capable and trustworthy embodied systems.

Planning methods often struggle with computational intractability when solving task-level problems in large-scale environments. This work explores how the commonsense knowledge encoded in Large Language Models (LLMs) can be leveraged to enhance planning techniques for such complex scenarios. Specifically, we propose an approach that uses LLMs to efficiently prune irrelevant components from the planning problem's state space, thereby substantially reducing its complexity. We demonstrate the efficacy of our system through extensive experiments in a household simulation environment as well as real-world validation on a 7-DoF manipulator (video: https://youtu.be/6ro2UOtOQS4).

Model-based manipulation of deformable objects has traditionally dealt with objects while neglecting their dynamics, thus mostly focusing on very lightweight objects at steady state. At the same time, soft robotic research has made considerable strides toward general modeling and control, despite soft robots and deformable objects being very similar from a mechanical standpoint. In this work, we leverage these recent results to develop a control-oriented, fully dynamic framework of slender deformable objects grasped at one end by a robotic manipulator. We introduce a dynamic model of this system using functional strain parameterizations and describe the manipulation challenge as a regulation control problem. This enables us to define a fully model-based control architecture, for which we can prove analytically closed-loop stability and provide sufficient conditions for steady state convergence to the desired state. The nature of this work is intended to be markedly experimental. We provide an extensive experimental validation of the proposed ideas, tasking a robot arm with controlling the distal end of six different cables, in a given planar position and orientation in space.

Motivated by the need for efficiency and robustness in repetitive robotic tasks such as locomotion, this study introduces the concept of Natural Motion Manifolds (NMMs) and presents a control method to stabilize and excite motions based on these structures. By considering the intersection of a Poincaré section with a surface comprising a continuum of autonomous evolutions, the proposed controller extends the linearized Poincaré map control from a single orbit to a family of orbits. This allows us to derive simple controllers to excite intrinsic nonlinear resonances and exploit the natural dynamics when varying the energy target (or the running velocity). We validated the method through simulations and experiments on a serial elastic quadruped. Relying on natural dynamics and minimal motor commands, we could implement a bounding gait at desired velocities without needing dynamic compensations. The experiments provide a thorough validation of the feasibility and the benefits of controlled, predictable, and purposeful oscillatory behavior via explicit excitation of a quadruped’s natural dynamics.

Generating precise motions with continuum soft robots calls for ways of closing the loop through nonconventional sensors - like cameras. This paper considers, for the first time, model-based visual servoing control of tendon-driven continuum soft robots. We take into account both regulation and trajectory tracking. We especially focus on a system inspired by the human neck, which employs an eye-in-hand configuration. The considered control architecture maps visual feedback to motor commands using the overall system's Jacobian, aiming for accurate end-effector (i.e., robot's head) positioning. This is made possible by blending classic results in visual serving with reduced-order models for continuum soft robots. In this work, we place significant focus on evaluating the controller's effectiveness on the physical prototype, for which we develop a tailored testing setup, which is a novel contribution of this work. Central to this setup is a tendon-driven soft robotic neck. We extensively characterize the control algorithm's performance for several control gains and different operational scenarios. We show that incorporating feedforward velocity estimation into the controller consistently improves performance in trajectory tracking tasks.

Deep reinforcement learning (DRL) has emerged as a promising solution to mastering explosive and versatile quadrupedal jumping skills. However, current DRL-based frameworks usually rely on pre-existing reference trajectories obtained by capturing animal motions or transferring experience from existing controllers. This work aims to prove that learning dynamic jumping is possible without relying on imitating a reference trajectory by leveraging a curriculum design. Starting from a vertical in-place jump, we generalize the learned policy to forward and diagonal jumps and, finally, we learn to jump across obstacles. Conditioned on the desired landing location, orientation, and obstacle dimensions, the proposed approach yields a wide range of omnidirectional jumping motions in real-world experiments. In particular, we achieve a 90 cm forward jump, exceeding all previous records for similar robots. Additionally, the robot can reliably execute continuous jumping on soft grassy grounds, which is especially remarkable as such conditions were not included in the training stage.

Soft robotics integrates engineering, materials science, and biology to tackle challenges that conventional robotics cannot solve. Alongside the advancements in soft robot technology, there is also a need for a standardized hardware platform that can enable benchmarking of various control methods developed for soft-bodied robots. This paper contributes to the state-of-the-art by designing a testbed that features a tendon-driven soft-bodied robot with integrated closed-loop force control.

Inspired by the dynamic coupling of moto-neurons and physical elasticity in animals, this work explores the possibility of generating locomotion gaits by utilizing physical oscillations in a soft snake by means of a low-level spiking neural mechanism. To achieve this goal, we introduce the Double Threshold Spiking neuron model with adjustable thresholds to generate varied output patterns. This neuron model can excite the natural dynamics of soft robotic snakes, and it enables distinct movements, such as turning or moving forward, by simply altering the neural thresholds. Finally, we demonstrate that our approach, termed SpikingSoft, naturally pairs and integrates with reinforcement learning. The high-level agent only needs to adjust the two thresholds to generate complex movement patterns, thus strongly simplifying the learning of reactive locomotion. Simulation results demonstrate that the proposed architecture significantly enhances the performance of the soft snake robot, enabling it to achieve target objectives with a 21.6% increase in success rate, a 29% reduction in time to reach the target, and smoother movements compared to the vanilla reinforcement learning controllers or Central Pattern Generator controller acting in torque space.