Search

Realizing high-performance soft robots is challenging because many existing soft or compliant actuators exhibit limitations like fabrication complexity, high power requirement, slow actuation, and low force generation. Due to their high force output and power efficiency, compactness, and simplicity in fabrication, twisted string actuators (TSAs) have exhibited strong potential in mechatronic and robotic applications. However, they have had limited uses in soft robotics. Consequently, modeling and control of TSA-driven soft robots have not been sufficiently studied. This paper presents the first study on the modeling and control of a TSA-driven soft robot manipulator. A physics-based model was developed to predict the manipulator's kinematic motion. An inverse model was derived to realize open-loop control. Models which describe the behavior of TSAs were utilized in a novel way to develop the proposed kinematic and inverse mod- els of the soft robot. The proposed modeling and control approaches were experimentally verified to be effective. For example, the modeling and control errors of the bending angle were 1.60°(3.11%) and 2.11°(3.68%), respectively.

Realizing high-performance soft robotic grippers is challenging because of the inherent limitations of the soft actuators and artificial muscles that drive them, including low force output, small actuation range, and poor compactness. Despite advances in this area, realizing compact soft grippers with high dexterity and force output is still challenging. This paper explores twisted string actuators (TSAs) to drive a soft robotic gripper. TSAs have been used in numerous robotic applications, but their inclusion in soft robots has been limited. The proposed design of the gripper was inspired by the human hand. Tunable stiffness was implemented in the fingers with antagonistic TSAs. The fingers' bending angles, actuation speed, blocked force output, and stiffness tuning were experimentally characterized. The gripper achieved a score of 6 on the Kapandji test and recreated 31 of the 33 grasps of the Feix GRASP taxonomy. It exhibited a maximum grasping force of 72 N, which was almost 13 times its own weight. A comparison study revealed that the proposed gripper exhibited equivalent or superior performance compared to other similar soft grippers., 19 pages, 15 figures

Twisted string actuators (TSAs) have exhibited great promise in robotic applications by generating high translational force with low input torque. Despite great success, it remains a challenge to reliably estimate the strain of TSAs using compact solutions while maintaining actuator compliance. The inclusion of position sensors not only increases system complexity but also decreases system compliance, a property often crucial in soft robots. We recently constructed a compliant TSA with self-sensing capability by adopting conductive and stretchable super-coiled polymer (SCP) strings; however, only quasi-static measurements of the strain-resistance correlation were obtained. This study proposes a strategy to experimentally characterize and model the transient self-sensing property in compliant TSAs. The correlation between resistance and strain is characterized under different motor twisting sequences and step durations, and exhibited transient decay, hysteresis, and creep. A self-sensing model that consists of a log-based nonlinear term, a rate-dependent Prandtl-Ishlinskii hysteresis term, and a creep term is proposed for the compliant TSAs. Experimental results confirm the high effectiveness of the proposed self-sensing approach, with the average model validation error less than 0.036 cm.

The nature-inspired behavior of collective motion is found to be an optimal solution in swarming systems for predator avoidance and survival. In this work, we propose a two-level control architecture for multi-robot systems (MRS), which leverages the advantages of flocking control and function approximated reinforcement learning for predator avoidance task. Reinforcement learning in multi-agent systems has gained a tremendous amount of interest in recent few years. Computationally intensive architectures such as deep reinforcement learning and actor-critic approaches have been extensively developed and have proved to be extremely efficient. The proposed approach, comprising of cooperative function approximated Q-learning, is applied such that it ensures formation maintenance in MRS while predator avoidance. A consensus filter is incorporated into the control architecture, to sense predators in close vicinity in a distributed and cooperative fashion to ensure consensus on states among the robots in the system. The proposed approach is proved to be convergent and results in superior performance in unexplored states and reduced number of variables. Simulation results confirm the effectiveness of the proposed approach over existing methods. We expect that the proposed approach can be conveniently applied to many other areas with little modifications, such as fire-fighting robots, surveillance and patrolling robots.

Supercoiled polymer (SCP) actuator is a recently developed thermally-driven artificial muscle that has shown large strain, high power density, and strong promise in robotics and intelligent systems. Termed as lonely stroke, the first cycle of SCP actuators is inconsistent with subsequent cycles that are repeatable and exhibit hysteresis. The lonely stroke not only affects SCP actuators' performances, but also presents coupling with hysteresis. It is thus crucial to capture the lonely stroke and hysteresis of SCP actuators to fully unleash their potential. However, the existing modeling and control strategies of SCP actuators fail to consider the lonely stroke nonlinearity. In this study, a modeling strategy is proposed to capture the quasi-static voltage-strain lonely stroke and hysteresis in SCP actuators. This is realized by expanding the input range of the Preisach operator, a widely used hysteresis model, to physically infeasible region to account for the lonely stroke. An iterative algorithm is proposed to compensate for the lonely stroke and hysteresis by approximately inverting the proposed model. For comparison purposes, a conventional Preisach operator and a polynomial model are considered. The modeling and control performance of the proposed approach is evaluated in experiments, and the superiority of the proposed scheme is demonstrated.

Supercoiled polymer (SCP) actuators belong to a recently discovered class of artificial muscles that show strong promise in various robotic applications. Extensive studies have been conducted on various aspects of SCP actuators, including the characterization of their hysteresis nonlinearity and dynamical behaviors. However, the existing strategies cannot effectively capture the first cycle of SCP actuators, which, under the application of certain sequences of inputs, gives rise to a “lonely stroke”. For example, under the application of continuous voltage oscillation sequences, the first output cycle is different from the successive repeatable cycles. This lonely stroke behavior in SCP actuators needs to be better analyzed since no comprehensive experimental investigations have been conducted. Furthermore, the lonely stroke affects SCP actuator’s performances in repeatable cycles. In this study, we conduct experimental investigations on SCP actuator’s lonely stroke under different loading and temperature conditions. The experimental procedures and measurements are presented and discussed. The results show that the coupling between lonely stroke and hysteresis is complex and history-dependent. For example, the maximum output strain discrepancy due to the lonely stroke was found to be 3.5% for an SCP actuator with a maximum strain of 15%.

Artificial muscles are inherently compliant actuators that exhibit high power density and impressive stroke and force outputs. They have found numerous applications in advanced mechatronic and robotic systems. Termed as lonely stroke, the first quasi-static input–output cycle of artificial muscles is inconsistent with subsequent cycles that are repeatable and exhibit hysteresis. The lonely stroke not only affects the behavior of artificial muscles, but also presents coupling with the subsequent repeatable hysteresis cycles. While lonely stroke has been reported in a number of artificial muscles, existing studies lack systematic experimental characterization procedure. Furthermore, no models or compensation schemes have been proposed. In this work, we propose the first study to systematically characterize, model, and compensate for the hysteresis with lonely stroke property in artificial muscles. The experimental procedure to characterize the hysteresis with lonely stroke behavior is presented. A modeling approach is developed through the expansion of the input range of the Preisach operator, a widely adopted hysteresis model, to physically infeasible region. The effects of such expansion on the model accuracy and computational cost are studied. An iterative algorithm is proposed to compensate for the hysteresis with lonely stroke by approximately inverting the proposed expanded Preisach operator. The effectiveness of the proposed scheme is validated by comprehensive simulation and experimental results. While this study primarily considers super-coiled polymer (SCP) actuators, the proposed model is also validated for several popular artificial muscles.

Supercoiled polymers (SCP) actuator, as a recently discovered artificial muscle, has attracted a lot of attention as a compliant and compact actuation mechanism. SCP actuators can be fabricated from nylon polymer threads, and generates up to 20% strain under thermal activation. A common challenge, however, is to accurately and efficiently estimate the performance of SCP actuators considering their significant hysteresis among loading, strain, and power input. Previous studies adopted either linear models that failed to capture the hysteresis or phenomenological models that required tedious procedures for identification and implementation. In this paper, a physics-inspired model is presented to efficiently capture and estimate SCP actuators’ strain – loading hysteresis by analyzing the properties of nylon threads from which they are fabricated. The strains of SCP actuators are found to be linear to that of the nylon threads under the same loading conditions. An efficient approach is proposed to characterize and estimate the strain – loading hysteresis of SCP actuators fabricated with different numbers of nylon threads. A helical spring model is adopted to obtain the stiffness of SCP actuators with different configurations. Experimental validation involving two-ply, four-ply, and six-ply nylon threads and SCP actuators are provided to confirm the effectiveness of the proposed model.