Soft Manipulator Research Articles

Deep Learning Methods in Soft Robotics: Architectures and Applications

The area of soft robotics has been subject to intense research efforts in the past two decades and constitutes a paradigm for advanced machine design in future robotic applications. However, standard methods for industrial robotics may be difficult to apply when analyzing soft robots. Deep learning, which has undergone rapid and transformative advancements in recent years, offers a set of powerful tools for analyzing and designing complex soft machines capable of operating in unstructured environments and interacting with humans and objects in a delicate manner. This review summarizes the most important state‐of‐the‐art deep learning architectures classified under supervised, unsupervised, semisupervised, and reinforcement learning scenarios and discusses their main features and benefits for different soft robotic applications, including soft robot manipulators, soft grippers, soft sensors, and e‐skins, as well as bioinspired soft robots. Specific properties of recent deep learning architectures and the usefulness of their features in addressing various types of issues found in soft robotics are analyzed. The existing challenges and future prospects are identified and discussed in view of the enhanced integration of both areas, which improves the performance of next‐generation soft machines operating in real‐world conditions.

Robust and Adaptive Control of a Soft Continuum Manipulator for Minimally Invasive Surgery

This article investigates the model-based control in configurations space of a soft continuum manipulator for minimally invasive surgery. The main control challenges for these types of systems are the presence of model uncertainties and nonlinearities. To this end, a sliding-mode controller, a Lyapunov redesign controller, and an adaptive controller have been designed and compared by means of simulations and experiments on a prototype. The results indicate that the adaptive controller yields better accuracy but a slower transient. Conversely, the sliding-mode controller and Lyapunov redesign yield a faster response but can result in chattering or steady-state errors.

Modeling and Compensation of Stiffness-Dependent Hysteresis Coupling Behavior for Parallel Pneumatic Artificial Muscle-Driven Soft Manipulator

Modeling and Compensation of Stiffness-Dependent Hysteresis Coupling Behavior for Parallel Pneumatic Artificial Muscle-Driven Soft Manipulator

Learning Controllers for Continuum Soft Manipulators: Impact of Modeling and Looming Challenges

Learning Controllers for Continuum Soft Manipulators: Impact of Modeling and Looming Challenges

Gecko-inspired adhesion enhanced by electroadhesive forces

Abstract Soft robotic manipulators have been increasingly adopted over the last decade due to their passive conformation to the shapes of objects, which can reduce control complexity. The performance of these grippers can be improved using flexible adhesive skins that increase tactile gripping forces, which is particularly important when grasping delicate objects and flexible substrates that are otherwise difficult to manipulate. In this work, we investigate how passive gecko-inspired fibrillar adhesion can be augmented by actively controlled electroadhesion (EA). The passive gecko-inspired outer skin (GS) enables adhesion with no power consumption while EA is controlled with an applied voltage. A numerical finite-element model is used to investigate how EA is affected by a commercially available gecko-inspired skin. The results show that the dielectric properties of the gecko-inspired skin reduce the magnitude of field intensity on the adhesive contact surface by only 2.1% at 3 kV. Compared with GS alone, EA with gecko-inspired skin (EAGS) increases the shear force by 66.8% and the normal force by 53.7% with an applied voltage of 4 kV. It is shown that the gecko skin’s adhesion force is enhanced by increased engagement of the fibrillar microstructure to object surfaces due to EA. This is experimentally demonstrated using frustrated total internal reflection imaging. This work shows that electroadhesive-enhanced gecko-inspired skin generates a greater adhesive force than the sum of forces from the separate gecko-inspired skin and EA. In this way, electrically controllable and passive adhesion mechanisms can be combined to improve the handling of flexible and delicate objects with smooth or rough surfaces.

Online Learning Based Shape Control for a Soft Manipulator Based on Spatial Features Feedback

Online Learning Based Shape Control for a Soft Manipulator Based on Spatial Features Feedback

Static Model of the Underwater Soft Bending Actuator Based on the Elliptic Integral Function

Underwater soft manipulators are increasingly used for grasping underwater organisms and cultural relics due to their compliance and ability to protect delicate objects. Unlike rigid manipulators, these soft manipulators feature underwater soft bending actuators that deform under external forces, making their shape and output force challenging to predict. Consequently, classical beam theory is no longer applicable. To address these issues, this article models the underwater soft bending actuator as a cantilever beam and establishes its shape and output force using elliptic integral functions. Additionally, this article proposes a method for determining the unknown parameters of the driver shape and output force model using optimization techniques and introduces a solution process based on the particle swarm optimization framework. Given the role of tensile and bending stiffness in the actuator’s performance, this paper employs a parameter identification method based on length and bending angle information. Tests demonstrate that the proposed method effectively estimates the deformation and output force of the underwater soft bending actuators, confirming its accuracy.

A Cable‐Actuated Soft Manipulator for Dexterous Grasping Based on Deep Reinforcement Learning

A Cable‐Actuated Soft Manipulator for Dexterous Grasping Based on Deep Reinforcement Learning

Vision-based reinforcement learning control of soft robot manipulators

PurposeThis study aims to tackle control challenges in soft robots by proposing a visually-guided reinforcement learning approach. Precise tip trajectory tracking is achieved for a soft arm manipulator.Design/methodology/approachA closed-loop control strategy uses deep learning-powered perception and model-free reinforcement learning. Visual feedback detects the arm’s tip while efficient policy search is conducted via interactive sample collection.FindingsPhysical experiments demonstrate a soft arm successfully transporting objects by learning coordinated actuation policies guided by visual observations, without analytical models.Research limitations/implicationsConstraints potentially include simulator gaps and dynamical variations. Future work will focus on enhancing adaptation capabilities.Practical implicationsBy eliminating assumptions on precise analytical models or instrumentation requirements, the proposed data-driven framework offers a practical solution for real-world control challenges in soft systems.Originality/valueThis research provides an effective methodology integrating robust machine perception and learning for intelligent autonomous control of soft robots with complex morphologies.

Biomimetic tapered soft manipulator with precision and load-bearing capacity

Biomimetic tapered soft manipulator with precision and load-bearing capacity

Multifunctional Magnetic Muscles for Soft Robotics

Despite recent advancements, artificial muscles have not yet been able to strike the right balance between exceptional mechanical properties and dexterous actuation abilities that are found in biological systems. Here, we present an artificial magnetic muscle that exhibits multiple remarkable mechanical properties and demonstrates comprehensive actuating performance, surpassing those of biological muscles. This artificial muscle utilizes a composite configuration, integrating a phase-change polymer and ferromagnetic particles, enabling active control over mechanical properties and complex actuating motions through remote laser heating and magnetic field manipulation. Consequently, the magnetic composite muscle can dynamically adjust its stiffness as needed, achieving a switching ratio exceeding 2.7 × 10³. This remarkable adaptability facilitates substantial load-bearing capacity, with specific load capacities of up to 1000 and 3690 for tensile and compressive stresses, respectively. Moreover, it demonstrates reversible extension, contraction, bending, and twisting, with stretchability exceeding 800%. We leverage these distinctive attributes to showcase the versatility of this composite muscle as a soft continuum robotic manipulator. It adeptly executes various programmable responses and performs complex tasks while minimizing mechanical vibrations. Furthermore, we demonstrate that this composite muscle excels across multiple mechanical and actuation aspects compared to existing actuators.

Spring-reinforced pneumatic actuator and soft robotic applications

Abstract Compared to rigid-structure robots, soft robots possess higher degrees of freedom and stronger environmental adaptability, which has aroused increasing attention in the robotic field. Among them, soft pneumatic robots have excellent performances in various practical applications. However, the nonlinearity and instability of pressure response of soft actuators caused by lateral expansion come to a great challenge. To address this problem, we proposed to embed a spring constraint layer around each single air chamber. Following the design concept, we obtained single-cavity and multi-DoF pneumatic actuators and evaluated their elongation and bending characteristics. Experimental results demonstrated that our proposed actuators have more linear pressure response as well as higher consistency. Eventually, through robotic applications, including soft robotic hand and gripper our proposed actuators could facilitate flexible manipulation and elaborate performance.

Data‐Driven Modeling and High‐Precision Tracking Control of a Soft Continuum Manipulator: Enabling Robotic Sorting of Multiwire Cables

As a new class of robots, soft continuum manipulators have attracted attention due to their flexibility and compliance. However, these characteristics create challenges for precise modeling and control. This study proposes a hybrid offline and online data‐driven scheme to achieve high‐precision tracking control of a soft continuum manipulator. First, a novel multiscale deep neural network learns the manipulator model offline. Specifically, the feature fusion module extracts highly discriminative features and captures long‐term dependencies from the temporal trajectory data. The self‐attention module strengthens the ability to represent fusion features and enhances the model prediction accuracy. Then, the learnt model is updated using multisensor data online, and the proposed controller further compensates for the updated model and enhances the tracking accuracy in the movement stage. Finally, the experimental results demonstrate a significant improvement in motion accuracy under different trajectory‐tracking scenarios (i.e., deviations of <1 mm in position and <0.8° in orientation). The example of the multiwire cable sorting proves the feasibility of the proposed scheme in high‐precision industrial applications.

Global Control of Soft Manipulator by Unifying Cosserat and Neural Network

Global Control of Soft Manipulator by Unifying Cosserat and Neural Network

Kinematic Modeling and Control for an Elephant-Trunk Soft Manipulator Considering Hysteresis

Kinematic Modeling and Control for an Elephant-Trunk Soft Manipulator Considering Hysteresis

Real-time control of a soft manipulator based on reduced order extended position-based dynamics

Soft robots are high-dimensional nonlinear systems coupled with both geometric and material nonlinearity. Control of such a system is complex and time-consuming. In this study, a real-time trajectory tracking control framework is established based on the reduced order extended position-based dynamics. In contrast to the common nonlinear model order reduction methods that require to collect a large number of data to create the motion subspace, this article's motion subspace is constructed based on the model configuration and material properties. The linear modes of the model and the related modal derivatives provide the reduced order matrix, which streamlines and increases the efficiency of model construction. Then, coupled with the instantaneous optimal control, a real-time reduced order model-based control framework of soft robots can be constructed. Experiments on trajectory tracking of a soft manipulator are conducted to verify the accuracy and efficiency of the proposed controller. The average error of all experiments is within 1 cm; and the single-step calculation time of the controller is about 0.057 s, which is less than the sampling period 0.1 s.

High-speed aerial grasping using a soft drone with onboard perception

Contrary to the stunning feats observed in birds of prey, aerial manipulation and grasping with flying robots still lack versatility and agility. Conventional approaches using rigid manipulators require precise positioning and are subject to large reaction forces at grasp, which limit performance at high speeds. The few reported examples of high-speed aerial grasping rely on motion capture systems, or fail to generalize across environments and grasp targets. We describe the first example of a soft aerial manipulator equipped with a fully onboard perception pipeline, capable of robustly localizing and grasping visually and morphologically varied objects. The proposed system features a novel passively closed tendon-actuated soft gripper that enables fast closure at grasp, while compensating for position errors, complying to the target-object morphology, and dampening reaction forces. The system includes an onboard perception pipeline that combines a neural-network-based semantic keypoint detector, a state-of-the-art robust 3D object pose estimator, and a fixed-lag smoother to estimate the pose of known objects. The resulting pose estimate is passed to a minimum-snap trajectory planner, tracked by an adaptive controller that fully compensates for the added mass of the grasped object. Finally, a finite-element-based controller determines optimal gripper configurations for grasping. Experiments on three different targets confirm that our approach enables dynamic, high-speed, and versatile grasping, all of which are necessary capabilities for tasks such as rapid package delivery or emergency relief. We demonstrate fully onboard vision-based grasps of a variety of objects, in both indoor and outdoor environments, and up to speeds of 2.0 m/s—the fastest vision-based grasp reported in the literature. Finally, we take a major step in expanding the utility of our platform beyond stationary targets, by demonstrating motion-capture-based grasps of targets moving up to 0.3 m/s, with relative speeds up to 1.5 m/s.

Quality and Testing of Red Pepper Soft Picking Manipulator Based on RD-DEM Coupling

Quality and Testing of Red Pepper Soft Picking Manipulator Based on RD-DEM Coupling

Planning and Motion Control for Underwater Bimanual Soft Manipulator in Underwater Grasping Task

Planning and Motion Control for Underwater Bimanual Soft Manipulator in Underwater Grasping Task

Analytical Inverse Kinematics for 2-Segment Extensible Continuum Manipulators

Abstract Fast inverse kinematics (IK) algorithm is significant for real-time precise motion control of continuum and soft manipulator. In this paper, we studied an explicit expression of two-segment continuum manipulator based on the constant curvature assumptions and discussed the existence of the IK solutions. This model used pseudo-rigid-body method, unlike 6-axis rigid industrial robot arm, we found that the two-segment extensible continuum manipulators have limited rotation angle around the direction vector of its tip, i.e., this type manipulator has limited dexterity. Then we pre-constrained five degrees of freedom (DOFs) constrained and discussed the definition of the left DOF. The IK solving process is simplified to realize small calculation cost for most applications. This model will provide a robust and real-time way to control the two-segment extensible continuum manipulators.