ABSTRACT Vision‐based three‐dimensional (3D) reconstruction plays an important role in robotic grasping and manipulation by providing accurate 3D information for target localization and motion planning. These tasks are usually performed in confined environments, where the vision‐based measurement systems are required to provide higher measurement accuracy while occupying a smaller volume. However, traditional stereo vision systems struggle to simultaneously achieve high precision and small volume, which limits their deployment in confined environments. In this paper, we developed a rotatable‐lens monocular vision system (RLMVS) consisting of a monocular camera, a wedge‐shaped lens, and an automatic rotation mechanism for the wedge‐shaped lens. A flexible learning‐based calibration method was proposed for the RLMVS. Furthermore, we proposed a rotatable lens‐based 3D reconstruction (RLR) method for RLMVS, which is the first time using the refraction of the light induced by a rotating wedge‐shaped lens to enable 3D reconstruction. Experimental results demonstrated that, compared with the traditional stereo vision system, the RLMVS reduced the RM measurement error by 61.2%, and reduced the overall volume of the 3D sensing modules by 46.5%. These advantages make the RLMVS particularly suitable for robotic manipulation in confined environments. We integrated the RLMVS with a robotic arm and performed grasping of small objects in a confined environment, demonstrating that the RLMVS provided a compact, low‐cost, yet accurate 3D sensing modality for robotic manipulation in confined environments.

This paper develops a predefined-time convergent and noise-tolerant fractional-order zeroing neural network (PTC-NT-FOZNN) model, innovatively engineered to tackle time-variant quadratic programming (TVQP) challenges. The PTC-NT-FOZNN, stemming from a novel iteration within the variable-gain ZNN spectrum, known as FOZNNs, features diminishing gains over time and marries noise resistance with predefined-time convergence, making it ideal for energy-efficient robotic motion planning tasks. The PTC-NT-FOZNN enhances traditional ZNN models by incorporating a newly developed activation function that promotes optimal convergence irrespective of the model's order. When evaluated against six established ZNNs, the PTC-NT-FOZNN, with parameters $0 < \alpha \leq 1$, demonstrates enhanced positional precision and resilience to additive noises, making it exceptionally suitable for TVQP tasks. Thorough practical assessments, including simulations and experiments using a Flexiv Rizon robotic arm, confirm the PTC-NT-FOZNN's capabilities in achieving precise tracking and high computational efficiency, thereby proving its effectiveness for robust kinematic control applications.

Camera model calibration establishes an accurate mapping between the 3D physical world and 2D images, enhancing the precision and reliability of vision measurements. Conventional calibration methods typically require multiple images to estimate initial values for nonlinear optimization, otherwise resulting in convergence to local optimal solutions. In addition, it also suffers from parameter coupling issues, limiting measurement accuracy and efficiency. To address these challenges, this study introduces a novel calibration method based on a physics-informed neural network (PINN). The proposed method reformulates the camera model as a set of constraint equations integrated into the neural network loss function. Additionally, a multilevel neural network architecture is designed to model the hierarchical transformations across coordinate systems in the camera’s imaging process. This design enables the neural network to capture and preserve physical relationships, ensuring accurate propagation of spatial information through the imaging pipeline. The method reduces dependency on initial values in nonlinear optimization and enhances calibration accuracy, even under physics-informed few-shot learning scenarios, while maintaining physical interpretability. A custom vision experiment system was developed to evaluate the method. Experimental results demonstrate that the proposed approach improves calibration accuracy by 29.82% compared to traditional methods and by 30.89% compared to the BPNN-based approach. These results confirm the effectiveness and precision of the calibration strategy.

While the manufacturing industry demands high-precision and efficient measurement of high-reflective workpieces, the low dynamic range of images depicting such workpieces leads to the loss of fringe information. Thus, achieving accurate measurement becomes challenging. Existing three-dimensional (3D) measurement methods have inherent limitations, such as being time-consuming, exhibiting poor precision, and poor dynamics. This paper proposes a double-exposure branch fusion network (DBF-Net) for high-dynamic-range (HDR) fringe projection profilometry. First, based on the pinhole imaging principle, the 3D measurement of fringe projections and the camera-projector inverse calibration model was explored to establish the object-image relationship accurately. Then, images of two exposure levels were randomly selected as the input for DBF-Net, and a sinusoidality-preserving algorithm was embedded into the network to obtain HDR fringe images. A weighted fusion loss function was also designed, which allows for dynamically adjusting the weighting of each term to improve the generalization of the method. By combining the aforementioned components with a phase-demodulation method, high-precision 3D reconstruction for high-reflectivity workpieces was achieved. The fringe projection profilometry system was established, and 3D accuracy verification experiments were conducted. And comparative experiments focusing on three aspects were conducted: image quality, phase demodulation accuracy, and 3D reconstruction of high-reflectivity workpieces. Based on the results, the proposed method achieves a measurement accuracy of 0.06 mm, effectively resolving the point-cloud loss issue encountered when measuring high-reflectivity surfaces.

Non-destructive testing of carbon steel components is critical for ensuring structural integrity and operational safety. Magneto-optical imaging (MOI) offers high detection accuracy and superior visualization. However, complex system configurations and high parameter sensitivity, which lead to inconsistent detection results, hinder its widespread adoption. Therefore, this study proposes a parameter design method for the MOI detection system driven by an end-to-end magneto-optical dual-field coupling model (E2E-MOM). First, an integrated MOI probe with a detection system layout based on a normal-incidence, high-fidelity optical path was proposed. This design addresses the problems of irrational optical path layouts, which reduce imaging fidelity and low efficiency in detection caused by weak system integration. Subsequently, an E2E-MOM was developed to guide the establishment of a finite-element-based simulation model that visualizes the entire detection process. Thereafter, evaluation criteria for magneto-optical images were established. The system's detection performance was assessed by evaluating various excitation and optical path parameters within the simulation model. This led to an optimized parameter set designed to ensure high-sensitivity detection capability. Finally, the experimental system was constructed and validated through testing. Results demonstrated a significant improvement in system performance following parameter design: detection efficiency reached 41.47 m2/h, sensitivity achieved 60 mu m, and detection results exhibited high fidelity. In conclusion, the design methodology proposed in this paper can significantly enhance the system's detection capabilities and shorten the equipment development cycle.

The manipulation of objects of varying sizes is crucial for grippers used in logistics and smart home applications. However, conventional grippers lack cross‐scale grasping capabilities. This study presents a novel single‐pump‐actuated transformable dual‐mode gripper (SATDG) that combines vacuum suction and granular jamming to effectively grasp objects ranging from 10 to 300 mm in size, covering a vast 30‐fold range, with forces from 0.5 to 21 N. Using a bistable mechanism and origami structure, the SATDG can transform between two modes and activate them with one pressure source. Subsequently, a finite‐element analysis of the characteristics of the bistable frame and the origami structure is presented. The performances of both grasping modes of the SATDG are validated. Demonstrations have shown that the SATDG performs outstandingly in logistics and smart home scenarios.

Teleoperation offers a promising approach to robotic data collection and human-robot interaction. However, existing teleoperation methods for data collection are still limited by efficiency constraints in time and space, and the pipeline for simulation-based data collection remains unclear. The problem is how to enhance task performance while minimizing reliance on real-world data. To address this challenge, we propose a teleoperation pipeline for collecting robotic manipulation data in simulation and training a few-shot sim-to-real visual-motor policy. Force feedback devices are integrated into the teleoperation system to provide precise end-effector gripping force feedback. Experiments across various manipulation tasks demonstrate that force feedback significantly improves both success rates and execution efficiency, particularly in simulation. Furthermore, experiments with different levels of visual rendering quality reveal that enhanced visual realism in simulation substantially boosts task performance while reducing the need for real-world data.

To investigate the influences of radial tears of the medial meniscus and subsequent surgical treatment (repair or meniscectomy) on the knee joint mechanics during normal walking. An original multiscale finite element musculoskeletal model was developed to model and study an intact knee model (no RT), four meniscal RT knee models (30

Stochastic resonance (SR) is widely used to weak signal detection and early fault diagnosis, but it often processes a complex signal into a sine shaped one, resulting in the loss of multi-harmonic signature embedded in the raw signal. To avoid this, this paper investigates underdamped SR induced by a symmetric triple-well potential with a uniform depth. Then, a SR array method is proposed to enhance weak multi-harmonic fault characteristics for early fault diagnosis of machinery. Theoretical results using spectral amplification factor indicate that the designed SR is superior to that induced by a symmetric triple-well potential with a different depth, increasing the potential-well width would maximize the spectral amplification factor of the tristable SR but keep the optimal noise intensity unchanged basically, and increasing the potential-well depth would decrease the spectral amplification factor of the tristable SR but enlarge the optimal noise intensity obviously. Experimental results demonstrate that the proposed method is able to enhance weak multi-harmonic fault characteristics for early fault diagnosis of roller bearings and gearboxes. Comparing with accugram and minimum entropy deconvolution combined with spectral kurtosis methods, the proposed method could detect weak fault characteristics and their harmonics successfully but other two methods fail to extract fault signature.

Robotic manipulators, traditionally designed with classical joint-link articulated structures, excel in industrial applications but face challenges in human-centered and general-purpose tasks requiring greater dexterity and adaptability. To address these challenges, we propose the prismatic-bending transformable (PBT) joint—a novel, scissors-inspired mechanism with directional maintenance capability that provides bending, rotation, and elongation/contraction within a single module. This design enables transformable kinematic chains that are modular, reconfigurable, and scalable for diverse tasks. We detail the mechanical design, optimization, kinematic and dynamic modeling, and experimental validation of the PBT joint, demonstrating its integration into foldable, modular robotic manipulators. The PBT joint functions as a single stock keeping unit, enabling manipulators to be constructed entirely from standardized PBT joints. It also serves as a modular extension for existing systems, such as wrist modules, streamlining design, deployment, transportation, and maintenance. Three joint sizes have been developed and tested, showcasing enhanced dexterity, reachability, and adaptability, particularly in confined and cluttered spaces. This work presents a promising approach to robotic manipulator development, providing a compact and versatile solution for operation in dynamic and constrained environments.

The thin-walled workpiece with non-equal thickness and closed section, such as a blade, constantly suffers from vibration and deformation in the Machining process. Supporting the workpiece during Machining to improve its local stiffness is an effective solution to this problem. Therefore, an asymmetrical support machining strategy by collaboration of a 6-DOF robot and a machine tool is proposed in this paper. In order to implement this collaborative machining strategy, several unconventional issues need to be solved: (1) The end-supporter path of the robot should be harmonized with the machining path of the machine tool strictly; (2) In the process of collaborative machining, the motion of the workpiece and the multi-DOF structure of the robot will lead to dynamically varying interference between the links and the workpiece, so it is essential to schedule the interference-free joint path of the robot inside the time-varying feasible regions. Aiming at these two problems, this paper proposes a robot path scheduling method with variable interference avoiding for collaborative machining. First, a method based on neutral-surface mapping is presented to schedule the end-supporter path of the robot. Then, the variable feasible space without interference is transformed to the path feasible domain of the interference-avoiding joints using the conformal geometric algebraic theory. Finally, a digraph model is established to obtain the optimal variable-interference-free joint path of the robot. The verification test shows that the scheduled end-supporter and joint paths can effectively harmonize the machining path during collaborative machining.

A k-winners-take-all (k-WTA) model is capable of selecting k winners from n participants via competitive operations for dynamic task allocation in a multirobot system (MRS). In this article, an event-triggered k-WTA (ET-kWTA) model is proposed to improve communication efficiency and alleviate the computational burden of competitive coordination in an MRS. By detecting the rate of change on k-WTA outputs of each robot, an ET condition based on Lyapunov stability is derived. Through the ET condition, different communication and computation frequencies are selected so that the computational performance and communication frequency are equilibrated. Thus, communication interactions and the computational burden of the MRS are subtly reduced. Theoretical analyses are provided to guarantee the convergence and robustness of the proposed ET-kWTA model. Furthermore, comparative simulations and experiments are conducted to illustrate the effectiveness and superiority of the proposed ET-kWTA model.

Soft robotics represents an emerging field that offers inherent compliance and adaptability. Over the past five years, research on actuation for soft robotic hands has grown explosively. Classic tendon-motor and pneumatic systems now incorporate high-torque micro-servos, textile bellows rated for millions of cycles, and millisecond-level smart valves. Meanwhile, new actuation methods such as electro-hydraulic HASEL pouches, photothermal liquid-crystal elastomers, and bio-hybrid muscle strips bring additional advantages to the robotic hands, including high energy density, cable-free light activation, self-repair capabilities, and quiet operation. This review first examines the fundamental designs of soft dexterous hands and soft metacarpophalangeal joints, looking at their working principles and key implementations for each actuation type. Then, this review explores their performance metrics, advantages, limitations, and suitable application scenarios. Finally, this review comprehensively compares critical parameters like actuation force, speed, and control complexity, highlighting the complementary strengths of these approaches and identifying areas for potential integration and future development.

This paper proposes discrete-time learning algorithms that utilize a data-driven technology to address the uncertain issues of optimization and structure. The main challenge lies in acquiring accurate optimization indices and Jacobian matrix, which can be addressed through iterative estimations enabled by these algorithms. On this basis, we propose a new model-adaptive kinematic control (MAKC) scheme for redundant manipulators without prior structure knowledge, incorporating the estimated optimization index and Jacobian matrix. To solve this scheme, a discretized data-driven neural dynamics (D3ND) controller is proposed based on the 94LVI algorithm, Kalman filter, and discrete-time learning algorithms. Theoretical analysis is provided to demonstrate its convergence. Subsequently, simulations and experiments are carried out on redundant manipulators using manipulability and joint drift as performance criteria. The results substantiate the robustness, practicability, and superiority of the proposed controller when encountering uncertain issues.

A novel binary channel fuzzy self-adjusted neural network (BCF-SANN) is proposed and researched for solving time-changing quadratic programming (QP) problems in this article. Unlike the fixed parameters of the typical zeroing neural network, the main parameters of the proposed BCF-SANN are time-changing, and its errors are adaptively quickly convergent. The biggest advantage of the novel neural network is that it combines a fuzzy self-adjusted controller, which takes the errors and derivatives of errors as fuzzy inputs and neural networks, further improving the convergence and robustness of the neural networks. To design the novel neural network, a time-changing QP problem is first established; then, using Lagrange’s law, the time-changing QP problem is transformed into a time-changing matrix equation; and finally, based on the time-changing parameter neural dynamics method, a novel BCF-SANN is proposed. The detailed design process is given in this article, and the convergence and robustness of the proposed BCF-SANN are proved by theoretical analysis. Through comparative experiments, it is demonstrated that the proposed BCF-SANN has a faster convergence rate and stronger robustness than the traditional zeroing neural network and 1-D fuzzy recurrent neural network (RNN).

For objects with complex topological and geometrical features, stochastic topological grasping can be executed without the necessity for feedback or precise planning. However, this grasping method has two significant limitations. First, the technique's effectiveness is reduced when interacting with topologically and geometrically simple objects like spheres, cubes, and cylinders, due to the inherent variability in grasping patterns. Additionally, the method's low stiffness restricts its ability to securely handling heavier objects. To address these challenges, this paper proposes an entanglement soft robotic gripper with variable stiffness and two transformed grasping modes (entanglement and clamping modes). The gripper contains three filaments, which can enhance the stiffness through the mechanism of layer jamming. Furthermore, the entanglement mode and the clamping mode, can be transformed by adjusting the working length of the filaments. The grasping performance comparison with and without variable stiffness was carried out, and the results indicated that the implementation of variable stiffness led to a 149 % increase in payload weight. Through experimental validation, we successfully employed the gripper in variable stiffness and transformed modes to grasp items with various shapes and weights. Demonstration of grasping heavier objects and transforming between two grasping modes were also conducted to showcase the adaptability and versatility of the gripper.

Navigating complex environments-ranging from confined spaces to open areas blocked by obstacles or out of reach of traditional robotic arms-remains a major challenge for conventional robotic systems. These challenges include limited dexterity, adaptability, and accessibility. To address these issues, this article proposed a dexterity and exploration enhanced quadruped robot (DEEQR), which was designed with a quadruped base and a multifunctional retractable variable stiffness manipulator (MRVSM). The MRVSM features a compact active-passive composite continuum mechanism with a 12.8 mm diameter, a high-stiffness telescopic mechanism, and a 2.5 mm diameter working channel that supports interchangeable tools such as a 2-DOF grasper or other miniaturized tools. This design enables precise manipulation and exploration in narrow and obstructed environments. Experiments demonstrated that DEEQR can traverse uneven terrain, navigate confined gaps, and perform complex tasks such as a bomb disposal operation inside a backpack with a 5 cm diameter entrance. The robot also successfully explored and delivered water in blocked spaces beyond the direct reach of the quadruped base. These results highlight DEEQR's strong potential for anti-terrorism applications and search and rescue missions in challenging environments.

In nature, prehensile tails serve as versatile and essential appendages for animals, facilitating both grasping and enhanced mobility. Although existing robotic tails effectively contribute to mobility across a range of behaviors, they lack versatile object-grasping capabilities. Inspired by these biological capabilities, a soft robotic prehensile tail is presented that uniquely integrates object manipulation with dynamic mobility enhancement for quadrupedal robots. This robotic tail offers a threefold stiffness variation and achieves large-angle bending ($\sim 636^{\circ }$) at the tail tip, thereby enabling secure and adaptable grasping. By adjusting its stiffness, the tail can conform to various shapes in a soft state and lift objects of different weights in a stiff state, demonstrating versatile grasping. The stiffened tail reliably supports the robot's body load (e.g., when hanging on a rod) and facilitates rapid, precise dynamic adjustments. Moreover, a novel synergy is revealed whereby grasped objects increase the tail's inertial effects, thereby enhancing the robot's dynamic capabilities during rapid maneuvers—a unique feature that transforms manipulation tasks into mobility advantages.

Accurate rigid-body dynamics is crucial for serial industrial robot applications, such as force control and physical human-robot interaction. Despite decades of research, the precise identification of dynamic parameters-particularly low-magnitude inertia parameters-remains a challenge for serial industrial robots. Researchers usually focus on developing various parameter estimation methods, while optimizing exciting trajectories in similar ways, typically minimizing the condition number of the information matrix. However, such optimization usually fails to ensure sufficient excitation for each parameter, due to nonconvex coupling effects. To address this limitation, we propose a fully decoupled rigid-body dynamics identification (FDRDI) method in this article. This approach innovatively eliminates coupling effects by using novel symmetrical exciting trajectories based on reciprocating S-curve. This innovation enables the independent identification of dynamic parameters associated with joint friction, as well as the gravity and inertia of links and payloads. Comparative experiments show that FDRDI achieves superior identification accuracy, evidenced by reduced joint torque prediction errors and payload parameter estimation errors.