The traditional pneumatic muscle robot joint has weak load capacity and low control precision. This paper proposes a bionic robotic ankle driven by multiple pneumatic muscle actuators. The structural design of the bionic robotic ankle and the drive mechanism that imitates human muscle recruitment are introduced. A dynamic model of the ankle and a static model of the pneumatic muscle actuator are established to analyze the driving characteristics. The multi-muscle recruiting strategy and load matching control method are optimized, and the output characteristics are simulated, including the robotic ankle driven by a single pneumatic muscle actuator, the robotic ankle driven by dual pneumatic muscle actuators, and the bionic ankle driven by multiple pneumatic muscle actuators. A prototype and testing platform are developed, and experimental research is carried out to validate the theoretical analysis and simulation. The results show that the bionic robotic ankle driven by multiple pneumatic muscle actuators can match varied loads, effectively reducing angle error and increasing output force.

Earthquake and other disasters nowadays still threat people’s lives and property due to their destructiveness and unpredictability. The past decades have seen the booming development of search and rescue robots due to their potential for increasing rescue capacity as well as reducing personnel safety risk at disaster sites. In this work, we propose a spider-inspired wheeled compliant leg to further improve the environmental adaptability of search mobile robots. Different from the traditional fully-actuated method with independent motor joint control, this leg employs an under-actuated compliant mechanism design with overall semi-tendon-driven control, which enables the passive and active terrain adaptation, system simplification and lightweight of the realized search robot. We have generalized the theoretical model and design methodology for this type of compliant leg, and implement it in a parametric program to improve the design efficiency. In addition, preliminary load capacity and leg-lifting experiments are carried out on a one-leg prototype to evaluate its mechanical performance. A four-legged robot platform is also fabricated for the locomotion tests. The preliminary experimental results have verified the feasibility of the proposed design methodology, and also show possibilities for improvements. In future work, structural optimization and stronger actuation elements should be introduced to further improve the mechanical performance of the fabricated wheeled leg mechanism and robot platform.

Human-robot collaboration fully leverages the strengths of both humans and robots, which is crucial for handling large, heavy objects at construction sites. To address the challenges of human-machine cooperation in handling large-scale, heavy objects — specifically building curtain walls — a human-robot collaboration system was designed based on the concept of “human-centered with machine support”. This system allows the handling of curtain walls according to different human intentions. First, a robot trajectory learning and generalization model based on dynamic motion primitives was developed. The operator’s motion intent was then characterized by their speed, force, and torque, with the force impulse introduced to define the operator’s intentions for acceleration and deceleration. Finally, a collaborative experiment was conducted on an experimental platform to validate the robot’s understanding of human handling intentions and to verify its ability to handle curtain wall. Collaboration between humans and robots ensured a smooth and labor-saving handling process.

We present a novel approach to predicting the pressure and flow rate of flexible electrohydrodynamic pumps using the Kolmogorov-Arnold Network. Inspired by the Kolmogorov-Arnold representation theorem, KAN replaces fixed activation functions with learnable spline-based activation functions, enabling it to approximate complex nonlinear functions more effectively than traditional models like Multi-Layer Perceptron and Random Forest. We evaluated KAN on a dataset of flexible EHD pump parameters and compared its performance against RF, and MLP models. KAN achieved superior predictive accuracy, with Mean Squared Errors of 12.186 and 0.012 for pressure and flow rate predictions, respectively. The symbolic formulas extracted from KAN provided insights into the nonlinear relationships between input parameters and pump performance. These findings demonstrate that KAN offers exceptional accuracy and interpretability, making it a promising alternative for predictive modeling in electrohydrodynamic pumping.

In the dynamic and unstructured environment of human-robot symbiosis, companion robots require natural human-robot interaction and autonomous intelligence through multimodal information fusion to achieve effective collaboration. Nevertheless, the control precision and coordination of the accompanying actions are not satisfactory in practical applications. This is primarily attributed to the difficulties in the motion coordination between the accompanying target and the mobile robot. This paper proposes a companion control strategy based on the Linear Quadratic Regulator (LQR) to enhance the coordination and precision of robot companion tasks. This method enables the robot to adapt to sudden changes in the companion target’s motion. Besides, the robot could smoothly avoid obstacles during the companion process. Firstly, a human-robot companion interaction model based on nonholonomic constraints is developed to determine the relative position and orientation between the robot and the companion target. Then, an LQR-based companion controller incorporating behavioral dynamics is introduced to simultaneously avoid obstacles and track the companion target’s direction and velocity. Finally, various simulations and real-world human-robot companion experiments are conducted to regulate the relative position, orientation, and velocity between the target object and the robot platform. Experimental results demonstrate the superiority of this approach over conventional control algorithms in terms of control distance and directional errors throughout system operation. The proposed LQR-based control strategy ensures coordinated and consistent motion with target persons in social companion scenarios.

This paper investigates the utilization of large language models (LLMs) for the comprehensive control of humanoid robot locomotion. Traditional reinforcement learning (RL) approaches for robot locomotion are resource-intensive and rely heavily on manually designed reward functions. To address these challenges, we propose a method that employs LLMs as the primary designer to handle key aspects of locomotion control, such as trajectory planning, inverse kinematics solving, and reward function design. By using user-provided prompts, LLMs generate and optimize code, reducing the need for manual intervention. Our approach was validated through simulations in Unity, demonstrating that LLMs can achieve human-level performance in humanoid robot control. The results indicate that LLMs can simplify and enhance the development of advanced locomotion control systems for humanoid robots.

In the field of autonomous robots, achieving complete precision is challenging, underscoring the need for human intervention, particularly in ensuring safety. Human Autonomy Teaming (HAT) is crucial for promoting safe and efficient human-robot collaboration in dynamic indoor environments. This paper introduces a framework designed to address these precision gaps, enhancing safety and robotic interactions within such settings. Central to our approach is a hybrid graph system that integrates the Generalized Voronoi Diagram (GVD) with spatio-temporal graphs, effectively combining human feedback, environmental factors, and key waypoints. An integral component of this system is the improved Node Selection Algorithm (iNSA), which utilizes the revised Grey Wolf Optimization (rGWO) for better adaptability and performance. Furthermore, an obstacle tracking model is employed to provide predictive data, enhancing the efficiency of the system. Human insights play a critical role, from supplying initial environmental data and determining key waypoints to intervening during unexpected challenges or dynamic environmental changes. Extensive simulation and comparison tests confirm the reliability and effectiveness of our proposed model, highlighting its unique advantages in the domain of HAT. This comprehensive approach ensures that the system remains robust and responsive to the complexities of real-world applications.

In the field of security, intelligent surveillance tasks often involve a large number of dense and small objects, with severe occlusion between them, making detection particularly challenging. To address this significant challenge, Dense and Small YOLO (DS-YOLO), a dense small object detection algorithm based on YOLOv8s, is proposed in this paper. Firstly, to enhance the dense small objects’ feature extraction capability of backbone network, the paper proposes a lightweight backbone. The improved C2fUIB is employed to create a lightweight model and expand the receptive field, enabling the capture of richer contextual information and reducing the impact of occlusion on detection accuracy. Secondly, to enhance the feature fusion capability of model, a multi-scale feature fusion network, Light-weight Full Scale PAFPN (LFS-PAFPN), combined with the DO-C2f module, is introduced. The new module successfully reduces the miss rate of dense small objects while ensuring the accuracy of detecting large objects. Finally, to minimize feature loss of dense objects during network transmission, a dynamic upsampling module, DySample, is implemented. DS-YOLO was trained and tested on the CrowdHuman and VisDrone2019 datasets, which contain a large number of densely populated pedestrians, vehicles and other objects. Experimental evaluations demonstrated that DS-YOLO has advantages in dense small object detection tasks. Compared with YOLOv8s, the Recall and mAP@0.5 are increased by 4.9% and 4.2% on CrowdHuman dataset, 4.6% and 5% on VisDrone2019, respectively. Simultaneously, DS-YOLO does not introduce a substantial amount of computing overhead, maintaining low hardware requirements.

The friction between the joints of the continuum manipulator with discrete joints brings great difficulties to kinematic modeling. The traditional driving wire arrangement limits the load capacity of the manipulator. A cable-stayed notch manipulator for transluminal endoscopic surgery is proposed, and a driving force coupling kinematic mode is established. The manipulator is fabricated from a superelastic Nitinol tube with bilaterally cut rectangular notches and is actuated by a stay cable. By applying the comprehensive elliptic integral solution (CEIS) for large deformation beams, the bending angle of each elastic beam is obtained, and the kinematics from the driving space to the joint space is formed. According to the bending angle of each elastic beam, the expression of the manipulator in Cartesian space can be obtained by geometric analysis. The kinematics from the joint space to the Cartesian space is established. The outer diameter of the manipulator is only 3.5 mm, and the inner diameter can reach 2 mm, allowing instruments to pass through. The maximum error of the manipulator movement is less than 5%. The load capacity of the manipulator has been verified through the stiffness experiments, and the maximum load of the manipulator can reach 400 g. The cable-stayed notch manipulator can be accurately modeled on the base of CEIS, and its motion accuracy can meet the needs of engineering applications. The compact size and excellent load capacity of the manipulator make it potential for application in transluminal endoscopic surgical robots.

In recent years, a variety of pneumatic soft actuators (PSAs) have been proposed due to the development of soft robots in biomimetic robots, medical devices, etc. At the same time, the modeling and control of PSAs remains an open question. In this paper, a spatial bending pneumatic soft actuator (SBPSA) modeling method based on the Prandtl-Ishlinskii (PI) model is proposed, and the inverse model is designed to compensate for hysteresis nonlinearity. Furthermore, an adaptive feedback controller combined with a hysteresis compensator is proposed for the precise control and tracking of SBPSAs. Finally, an experimental platform is built, and experimental results demonstrate the effectiveness of the proposed method for precise tracking.