The nature of continuum soft robots calls for novel perception solutions, which can provide information on the robot's shape while not substantially modifying their bodies' softness. One way to achieve this goal is to develop innovative and completely deformable sensors. However, these solutions tend to be less reliable than classic sensors for rigid robots. As an alternative, we consider here the use of monocular cameras. By admitting a small rigid component in our design, we can leverage well-established solutions from mobile robotics. We propose a shape sensing strategy that combines a SLAM algorithm with nonlinear optimization based on the robot's kinematic model. We prove the method's effectiveness in simulation and with experiments of a single-segment continuous soft robot with a camera mounted to the tip. We achieve mean relative translational errors below 9% simulations and experiments alike, and as low as 0.5% on average for some simulation conditions.@en

Accurate and complete terrain maps enhance the awareness of autonomous robots and enable safe and optimal path planning. Rocks and topography often create occlusions and lead to missing elevation information in the Digital Elevation Map (DEM). Currently, these occluded areas are either fully avoided during motion planning or the missing values in the elevation map are filled-in using traditional interpolation, diffusion or patch-matching techniques. These methods cannot leverage the high-level terrain characteristics and the geometric constraints of line of sight we humans use intuitively to predict occluded areas. We introduce a self-supervised learning approach capable of training on real-world data without a need for ground-truth information to reconstruct the occluded areas in the DEMs. We accomplish this by adding artificial occlusion to the incomplete elevation maps constructed on a real robot by performing ray casting. We first evaluate a supervised learning approach on synthetic data for which we have the full ground-truth available and subsequently move to several real-world datasets. These real-world datasets were recorded during exploration of both structured and unstructured terrain with a legged robot, and additionally in a planetary scenario on Lunar analogue terrain. We state a significant improvement compared to the baseline methods both on synthetic terrain and for the real-world datasets. Our neural network is able to run in real-time on both CPU and GPU with suitable sampling rates for autonomous ground robots. We motivate the applicability of reconstructing occlusion in elevation maps with preliminary motion planning experiments.@en

Actuators’ dynamics have been so far mostly neglected when devising feedback controllers for continuum soft robots since the problem under the direct actuation hypothesis is already quite hard to solve. Directly considering actuation would have made the challenge too complex. However, these effects are, in practice, far from being negligible. The present work focuses on model-based control of piston-driven pneumatically-actuated soft robots. We propose a model of the relationship between the robot’s state, the acting fluidic pressure, and the piston dynamics, which is agnostic to the chosen model for the soft system dynamics. We show that backstepping is applicable even if the feedback coupling of the outer on the inner subsystem is not linear. Thus, we introduce a general model-based control strategy based on backstepping for soft robots actuated by fluidic drive. As an example, we derive a specialized version for a robot with piecewise constant curvature.@en

Sensing the shape of continuum soft robots without obstructing their movements and modifying their natural softness requires innovative solutions. This letter proposes to use magnetic sensors fully integrated into the robot to achieve proprioception. Magnetic sensors are compact, sensitive, and easy to integrate into a soft robot. We also propose a neural architecture to make sense of the highly nonlinear relationship between the perceived intensity of the magnetic field and the shape of the robot. By injecting a priori knowledge from the kinematic model, we obtain an effective yet data-efficient learning strategy. We first demonstrate in simulation the value of this kinematic prior by investigating the proprioception behavior when varying the sensor configuration, which does not require us to re-train the neural network. We validate our approach in experiments involving one soft segment containing a cylindrical magnet and three magnetoresistive sensors. During the experiments, we achieve mean relative errors of 4.5%.@en

Integrating Brain-Machine Interfaces into non-clinical applications like robot motion control remains difficult - despite remarkable advancements in clinical settings. Specifically, EEG-based motor imagery systems are still error-prone, posing safety risks when rigid robots operate near humans. This work presents an alternative pathway towards safe and effective operation by combining wearable EEG with physically embodied safety in soft robots. We introduce and test a pipeline that allows a user to move a soft robot's end effector in real time via brain waves that are measured by as few as three EEG channels. A robust motor imagery algorithm interprets the user's intentions to move the position of a virtual attractor to which the end effector is attracted, thanks to a new Cartesian impedance controller. We specifically focus here on planar soft robot-based architected metamaterials, which require the development of a novel control architecture to deal with the peculiar nonlinearities - e.g., non-affinity in control. We preliminarily but quantitatively evaluate the approach on the task of setpoint regulation. We observe that the user reaches the proximity of the setpoint in 66% of steps and that for successful steps, the average response time is 21.5s. We also demonstrate the execution of simple real-world tasks involving interaction with the environment, which would be extremely hard to perform if it were not for the robot's softness.@en

Parallel robots based on Handed Shearing Auxetics (HSAs) can implement complex motions using standard electric motors while maintaining the complete softness of the structure, thanks to specifically designed architected metamaterials. However, their control is especially challenging due to varying and coupled stiffness, shearing, non-affine terms in the actuation model, and underactuation. In this paper, we present a model-based control strategy for planar HSA robots enabling regulation in task space. We formulate equations of motion, show that they admit a collocated form, and design a P-satI-D feedback controller with compensation for elastic and gravitational forces. We experimentally identify and verify the proposed control strategy in closed loop.@en

Electrically-actuated continuum soft robots based on Handed Shearing Auxetics (HSAs) promise rapid actuation capabilities while preserving structural compliance. However, the foundational models of these novel actuators required for precise control strategies are missing. This paper proposes two key components extending discrete Cosserat rod model (DCM) to allow for modeling HSAs. First, we propose a mechanism for incorporating the auxetic trajectory into DCM dynamical simulations. We also propose an implementation of this extension as a plugin for the Elastica simulator. Second, we introduce a Selective Piecewise Constant Strain (SPCS) kinematic parameterization that can describe an HSA segment's shape with fewer configuration variables. We verify both theoretical contributions experimentally. The simulator is used to replicate experimental data of the mechanical characterization of HSA rods. For the second component, we attach motion capture markers at various points to a parallel HSA robot and find that the shape of the HSAs can be kinematically represented with an average accuracy of 0.3 mm for positions and 0.07 rad for orientations.@en

Model-based strategies are a promising solution to the grand challenge of equipping continuum soft robots with motor intelligence. However, finite-dimensional models of these systems are inherently inaccurate, thus posing pressing robustness concerns. Moreover, the actuation space of soft robots is usually limited. This article aims at solving both these challenges by proposing a robust model-based strategy for the shape control of soft robots with system uncertainty and input saturation. The proposed architecture is composed of two key components. First, we propose an observer that estimates deviations between the theoretical model and the soft robot, ensuring that the estimation error converges to zero within finite time. Second, we introduce a sliding mode controller to regulate the soft robot shape while fulfilling saturation constraints. This controller uses the observer's output to compensate for the deviations between the real system and the established model. We prove the convergence of the closed-loop with theoretical analysis and the method's effectiveness with simulations and experiments.@en

Model-based strategies are a promising solution to the grand challenge of equipping continuum soft robots with motor intelligence. However, finite-dimensional models of these systems are inherently inaccurate, thus posing pressing robustness concerns. Moreover, the actuation space of soft robots is usually limited. This article aims at solving both these challenges by proposing a robust model-based strategy for the shape control of soft robots with system uncertainty and input saturation. The proposed architecture is composed of two key components. First, we propose an observer that estimates deviations between the theoretical model and the soft robot, ensuring that the estimation error converges to zero within finite time. Second, we introduce a sliding mode controller to regulate the soft robot shape while fulfilling saturation constraints. This controller uses the observer's output to compensate for the deviations between the real system and the established model. We prove the convergence of the closed-loop with theoretical analysis and the method's effectiveness with simulations and experiments.@en