Learning the inverse kinetics of an octopus-like manipulator in three-dimensional space

Latent Representation-Based Learning Controller for Pneumatic and Hydraulic Dual Actuation of Pressure-Driven Soft Actuators

The pneumatic and hydraulic dual actuation of pressure-driven soft actuators (PSAs) is promising because of their potential to develop novel practical soft robots and expand the range of soft robot applications. However, the physical characteristics of air and water are largely different, which makes it challenging to quickly adapt to a selected actuation method and achieve method-independent accurate control performance. Herein, we propose a novel LAtent Representation-based Feedforward Neural Network (LAR-FNN) for dual actuation. The LAR-FNN consists of an autoencoder (AE) and a feedforward neural network (FNN). The AE generates a latent representation of a PSA from a 30-s stairstep response. Subsequently, the FNN provides an individual inverse model of the target PSA and calculates feedforward control input by using the latent representation. The experimental results with PSAs demonstrate that the LAR-FNN can meet the requirements of dual actuation control (i.e., accurate control performance regardless of the actuation method with a short adaptation time) with a single neural network. The results suggest that a LAR-FNN can contribute to soft dual-actuation robot development and the field of soft robotics.

A Twisted and Coiled Polymer Artificial Muscles Driven Soft Crawling Robot Based on Enhanced Antagonistic Configuration

Twisted and coiled polymer (TCP) actuators are becoming increasingly prevalent in soft robotic fields due to their powerful and hysteresis-free stroke, large specific work density, and ease of fabrication. This paper presents a soft crawling robot with spike-inspired robot feet which can deform and crawl like an inchworm. The robot mainly consists of two leaf springs, connection part, robot feet, and two TCP actuators. A system level model of a soft crawling robot is presented for flexible and effective locomotion. Such a model can offer high-efficiency design and flexible locomotion of the crawling robot. Results show that the soft crawling robot can move at a speed of 0.275 mm/s when TCP is powered at 24 V.

Individual deformability compensation of soft hydraulic actuators through iterative learning-based neural network

Robotic devices with soft actuators have been developed to realize the effective rehabilitation of patients with motor paralysis by enabling soft and safe interaction. However, the control of such robots is challenging, especially owing to the difference in the individual deformability occurring in manual fabrication of soft actuators. Furthermore, soft actuators used in wearable rehabilitation devices involve a large response delay which hinders the application of such devices for at-home rehabilitation. In this paper, a feed-forward control method for soft actuators with a large response delay, comprising a simple feed-forward neural network (FNN) and an iterative learning controller (ILC), is proposed. The proposed method facilitates the effective learning and acquisition of the inverse model (i.e. the model that can generate control input to the soft actuator from a target trajectory) of soft actuators. First, the ILC controls a soft actuator and iteratively learns the actuator deformability. Subsequently, the FNN is trained to obtain the inverse model of the soft actuator. The control results of the ILC are used as training datasets for supervised learning of the FNN to ensure that it can efficiently acquire the inverse model of the soft actuator, including the deformability and the response delay. Experiments with fiber-reinforced soft bending hydraulic actuators are conducted to evaluate the proposed method. The results show that the ILC can learn and compensate for the actuator deformability. Moreover, the iterative learning-based FNN serves to achieve a precise tracking performance on various generalized trajectories. These facts suggest that the proposed method can contribute to the development of robotic rehabilitation devices with soft actuators and the field of soft robotics.

A Survey for Machine Learning-Based Control of Continuum Robots

Soft continuum robots have been accepted as a promising category of biomedical robots, accredited to the robots’ inherent compliance that makes them safely interact with their surroundings. In its application of minimally invasive surgery, such a continuum concept shares the same view of robotization for conventional endoscopy/laparoscopy. Different from rigid-link robots with accurate analytical kinematics/dynamics, soft robots encounter modeling uncertainties due to intrinsic and extrinsic factors, which would deteriorate the model-based control performances. However, the trade-off between flexibility and controllability of soft manipulators may not be readily optimized but would be demanded for specific kinds of modeling approaches. To this end, data-driven modeling strategies making use of machine learning algorithms would be an encouraging way out for the control of soft continuum robots. In this article, we attempt to overview the current state of kinematic/dynamic model-free control schemes for continuum manipulators, particularly by learning-based means, and discuss their similarities and differences. Perspectives and trends in the development of new control methods are also investigated through the review of existing limitations and challenges.

Review of machine learning methods in soft robotics

Soft robots have been extensively researched due to their flexible, deformable, and adaptive characteristics. However, compared to rigid robots, soft robots have issues in modeling, calibration, and control in that the innate characteristics of the soft materials can cause complex behaviors due to non-linearity and hysteresis. To overcome these limitations, recent studies have applied various approaches based on machine learning. This paper presents existing machine learning techniques in the soft robotic fields and categorizes the implementation of machine learning approaches in different soft robotic applications, which include soft sensors, soft actuators, and applications such as soft wearable robots. An analysis of the trends of different machine learning approaches with respect to different types of soft robot applications is presented; in addition to the current limitations in the research field, followed by a summary of the existing machine learning methods for soft robots.

Cable Tension Analysis Oriented the Enhanced Stiffness of a 3-DOF Joint Module of a Modular Cable-Driven Human-Like Robotic Arm

Inspired by the structure of human arms, a modular cable-driven human-like robotic arm (CHRA) is developed for safe human–robot interaction. Due to the unilateral driving properties of the cables, the CHRA is redundantly actuated and its stiffness can be adjusted by regulating the cable tensions. Since the trajectory of the 3-DOF joint module (3DJM) of the CHRA is a curve on Lie group SO(3), an enhanced stiffness model of the 3DJM is established by the covariant derivative of the load to the displacement on SO(3). In this paper, we focus on analyzing the how cable tension distribution problem oriented the enhanced stiffness of the 3DJM of the CHRA for stiffness adjustment. Due to the complexity of the enhanced stiffness model, it is difficult to solve the cable tensions from the desired stiffness analytically. The problem of stiffness-oriented cable tension distribution (SCTD) is formulated as a nonlinear optimization model. The optimization model is simplified using the symmetry of the enhanced stiffness model, the rank of the Jacobian matrix and the equilibrium equation of the 3DJM. Since the objective function is too complicated to compute the gradient, a method based on the genetic algorithm is proposed for solving this optimization problem, which only utilizes the objective function values. A comprehensive simulation is carried out to validate the effectiveness of the proposed method.

Controlling of Pneumatic Muscle Actuator Systems by Parallel Structure of Neural Network and Proportional Controllers (PNNP)

This article proposed a novel controller structure to track the non-linear behavior of the pneumatic muscle actuator (PMA), such as the elongation for the extensor actuator and bending for the bending PMA. The proposed controller consists of a neural network (NN) controller laid in parallel with the proportional controller (P). The parallel neural network proportional (PNNP) controllers provide a high level of precision and fast-tracking control system. The PNNP has been applied to control the length of the single extensor PMA and the bending angle of the single self-bending contraction actuator (SBCA) at different load values. For further validation, the PNNP has been applied to control a human-robot shared control system. The results show the efficiency of the proposed controller structure.

Learning the Kinematics of a Manipulator Based on VQTAM

The kinematics of a robotic manipulator is critical to the real-time performance and robustness of the robot control system. This paper proposes a surrogate model of inverse kinematics for the serial six-degree of freedom (6-DOF) robotic manipulator, based on its kinematics symmetry. Herein, the inverse kinematics model is derived via the training of the Vector-Quantified Temporal Associative Memory (VQTAM) network, which originates from Self-Organized Mapping (SOM). During the processes of training, testing, and estimating of this neural network, a priority K-means tree search algorithm is utilized, thus improving the training efficacy. Furthermore, Local Linear Regression (LLR), Local Weighted Linear Regression (LWR), and Local Linear Embedding (LLE) algorithms are, respectively, combined with VQTAM to obtain three improvement algorithms, all of which aim to further optimize the prediction accuracy of the networks for subsequent comparison and selection. To speed up the solving of the least squared equation, which is common among the three algorithms, Singular Value Decomposition (SVD) is introduced. Finally, data from forward kinematics, in the form of the exponential product of a motion screw, are obtained, and are used for the construction and validation of the VQTAM neural network. Our results show that the prediction effect of the LLE algorithm is better than others, and that the LLE algorithm is a potential surrogate model to estimate the output of inverse kinematics.

Octopus Arm-Inspired Tapered Soft Actuators with Suckers for Improved Grasping

Octopuses can employ their tapered arms to catch prey of all shapes and sizes due to their dexterity, flexibility, and gripping power. Intrigued by variability in arm taper angle between different octopus species, we explored the utility of designing soft actuators exhibiting a distinctive conical geometry, compared with more traditional cylindrical forms. We find that these octopus-inspired conical-shaped actuators exhibit a wide range of bending curvatures that can be tuned by simply altering their taper angle and they also demonstrate greater flexibility compared with their cylindrical counterparts. The taper angle and bending curvature are inversely related, whereas taper angle and applied bending force are directly related. To further expand the functionality of our soft actuators, we incorporated vacuum-actuated suckers into the actuators for the production of a fully integrated octopus arm-inspired gripper. Notably, our results reveal that because of their enhanced flexibility, these tapered actuators with suckers have better gripping power than their cylindrical-shaped counterparts and require significantly larger forces to be detached from both flat and curved surfaces. Finally, we show that by choosing appropriate taper angles, our tapered actuators with suckers can grip, move, and place a remarkably wide range of objects with flat, nonplanar, smooth, or rough surfaces, as well as retrieve objects through narrow openings. The results from this study not only provide new design insights into the creation of next-generation soft actuators for gripping a wide range of morphologically diverse objects but also contribute to our understanding of the functional significance of arm taper angle variability across octopus species.

Soft Robotics as an Enabling Technology for Agroforestry Practice and Research

The shortage of qualified human labor is a key challenge facing farmers, limiting profit margins and preventing the adoption of sustainable and diversified agroecosystems, such as agroforestry. New technologies in robotics could offer a solution to such limitations. Advances in soft arms and manipulators can enable agricultural robots that can have better reach and dexterity around plants than traditional robots equipped with hard industrial robotic arms. Soft robotic arms and manipulators can be far less expensive to manufacture and significantly lighter than their hard counterparts. Furthermore, they can be simpler to design and manufacture since they rely on fluidic pressurization as the primary mechanisms of operation. However, current soft robotic arms are difficult to design and control, slow to actuate, and have limited payloads. In this paper, we discuss the benefits and challenges of soft robotics technology and what it could mean for sustainable agriculture and agroforestry.

A generalizable equilibrium model for bending soft arms with longitudinal actuators

Current models of bending in soft arms are formulated in terms of experimentally determined, arm-specific parameters, which cannot evaluate fundamental differences in soft robot arm design. Existing models are successful at improving control of individual arms but do not give insight into how the structure of the arm affects the arm’s capabilities. For example, omnidirectional soft robot arms most frequently have three parallel actuators, but may have four or more, while common biological arms, including octopuses, have tens of distinct longitudinal muscle bundles. This article presents a quasi-static analytical model of soft arms bent with longitudinal actuators, based on equilibrium principles and assuming an unknown neutral axis location. The model is presented as a generalizable framework and specifically implemented for an arm with [Formula: see text] fluid-driven actuators, a subset of which are pressurized to induce a bend with a certain curvature and direction. The presented implementation is validated experimentally using planar (2D) and spatial (3D) bends. The planar model is used to initially estimate pressure for a closed-loop curvature control system and to bound the accessible configurations for a rapidly-exploring random trees (RRT) motion planner. A three-segment planar arm is demonstrated to navigate along a planned trajectory through a gap in a wall. Finally, the model is used to explore how the arm morphology affects maximum curvature and directional resolution. This research analytically connects soft arm structure and actuator behavior to unloaded arm performance, and the results may be used to methodically design soft robot arms.

A Soft Parallel Kinematic Mechanism

In this article, we describe a novel holonomic soft robotic structure based on a parallel kinematic mechanism. The design is based on the Stewart platform, which uses six sensors and actuators to achieve full six-degree-of-freedom motion. Our design is much less complex than a traditional platform, since it replaces the 12 spherical and universal joints found in a traditional Stewart platform with a single highly deformable elastomer body and flexible actuators. This reduces the total number of parts in the system and simplifies the assembly process. Actuation is achieved through coiled-shape memory alloy actuators. State observation and feedback is accomplished through the use of capacitive elastomer strain gauges. The main structural element is an elastomer joint that provides antagonistic force. We report the response of the actuators and sensors individually, then report the response of the complete assembly. We show that the completed robotic system is able to achieve full position control, and we discuss the limitations associated with using responsive material actuators. We believe that control demonstrated on a single body in this work could be extended to chains of such bodies to create complex soft robots.