Pneumatic muscles with similar characteristics to biological muscles have been widely used in robots, and thus are promising drivers for frog inspired robots. However, the application and nonlinearity of the pneumatic system limit the advance. On the basis of the swimming mechanism of the frog, a frog-inspired robot based on pneumatic muscles is developed. To realize the independent tasks by the robot, a pneumatic system with internal chambers, micro air pump, and valves is implemented. The micro pump is used to maintain the pressure difference between the source and exhaust chambers. The pneumatic muscles are controlled by high-speed switch valves which can reduce the robot cost, volume, and mass. A dynamic model of the pneumatic system is established for the simulation to estimate the system, including the chamber, muscle, and pneumatic circuit models. The robot design is verified by the robot swimming experiments and the dynamic model is verified through the experiments and simulations of the pneumatic system. The simulation results are compared to analyze the functions of the source pressure, internal volume of the muscle, and circuit flow rate which is proved the main factor that limits the response of muscle pressure. The proposed research provides the application of the pneumatic muscles in the frog inspired robot and the pneumatic model to study muscle controller.

Inspired by visual-tactile cross-modal bidirectional mapping of the human brain, this paper introduces a novel approach to bidirectional mapping between visual and tactile data, an area not fully explored in the predominantly unidirectional existing studies. First, we adopt separate Variational AutoEncoder (VAE) models for visual and tactile data. Furthermore, we introduce a conditional flow model built on the VAE latent feature space, enabling cross-modal bidirectional mapping between visual and tactile data using one model. The experimental results show that our method achieves excellent performance in terms of the similarity between the generated data and the original data (Structural Similarity Index (SSIM) of visual data: 0.58, SSIM of tactile data: 0.80), the classification accuracy on generated data (visual data: 91.60%, tactile data: 88.05%), and the zero-shot classification accuracy between generated data and language (visual data: 44.49%, tactile data: 45.03%). To the best of our knowledge, the method proposed in this paper is the first one to utilize a single model to achieve bidirectional mapping between visual and tactile data. Our model and code will be made public after the acceptance of the paper.

The Stewart platform is a typical parallel manipulator. It is used as a space docking mechanism whose requirement for movement scope is determinate. This article addresses the dimension synthesis of the space docking mechanism. First, this article compares the common indexes used to evaluate the performance of parallel manipulator. These indexes can mainly be divided into two types. The evaluation indexes based on the Jacobian matrix, including the singular value index, the manipulability index, and condition index, are derived from the Jacobian matrix of the manipulator, and the transmissibility indexes, such as pressure angle, transmission angle, and motion/force transmission index, lay emphasis on the power transmission ability of manipulator. This article proposes the global transmission index under determinate constraint of workspace on the basis of the previous transmissibility indexes to evaluate the performance of parallel manipulator whose workspace is determinate and movement is the combination of translation and rotation. And the optimization design of the manipulator is carried out by taking the global transmission index under determinate constraint of workspace into account. The detailed process of determining the optimal configuration for the Stewart manipulator used as space docking mechanism is presented. The optimal result is gotten, which has a large value of global transmission index under determinate constraint of workspace in the required workspace, and a simulation is carried out to verify the validity of the optimal result.

In manufacturing, traditional task pre-programming methods limit the efficiency of human–robot skill transfer. This paper proposes a novel task-learning strategy, enabling robots to learn skills from human demonstrations flexibly and generalize skills under new task situations. Specifically, we establish a markerless vision capture system to acquire continuous human hand movements and develop a threshold-based heuristic segmentation algorithm to segment the complete movements into different movement primitives (MPs) which encode human hand movements with task-oriented models. For movement primitive learning, we adopt a Gaussian mixture model and Gaussian mixture regression (GMM-GMR) to extract the optimal trajectory encapsulating sufficient human features and utilize dynamical movement primitives (DMPs) to learn for trajectory generalization. In addition, we propose an improved visuo-spatial skill learning (VSL) algorithm to learn goal configurations concerning spatial relationships between task-relevant objects. Only one multioperation demonstration is required for learning, and robots can generalize goal configurations under new task situations following the task execution order from demonstration. A series of peg-in-hole experiments demonstrate that the proposed task-learning strategy can obtain exact pick-and-place points and generate smooth human-like trajectories, verifying the effectiveness of the proposed strategy.

The robotic hand plays an extremely significant role in completing various robotic tasks, especially for a robot to complete different grasping tasks. In this paper, we propose a novel multi-fingered bionic hand with variable stiffness for robotic grasp. First, based on the analysis of the human hand, we design the modular finger with variable stiffness mechanism and the flexible thumb with dual-link independent mechanism. Second, the multi-fingered bionic hand composed of three modular fingers and one flexible thumb is presented, which possesses 14 degrees-of-freedom (DoFs). Then, the parameters of the designed hand are analyzed to obtain the stiffness characteristics and working space. Furthermore, the grasping control method based on position control and force control is proposed for robotic grasp. Finally, a series of experiments are designed to verify the effectiveness of the designed hand and the proposed method. The experimental results show that (1) the designed variable stiffness mechanism can be used to adjust the stiffness of hand fingers, (2) the proposed method can effectively control the grasping force under different stiffness conditions, and (3) the designed hand can successfully grasp different types of objects by various grasping strategies.