Simultaneous localization and mapping technology is the basis for multi-robot systems to complete navigation, path planning, and autonomous exploration in complex, dynamic, and Global Positioning System (GPS)-denied environments. This paper reviews the current status and progress of multi-robot simultaneous localization and mapping (SLAM) technology based on LiDAR. First, this paper studies the basic principles of LiDAR SLAM. It analyzes the system model construction of LiDAR SLAM, including the mobile robot coordinate system model, kinematic model, sensor model, map presentation, LiDAR SLAM framework, and classic algorithms. Then, this paper discusses the basic framework of collaborative SLAM, analyzes the key issues such as data association, loop closure detection, and global graph optimization in collaborative SLAM, and conducts a detailed literature review on the solutions to key problems in sub-fields of multi-robot SLAM such as frontier detection, task allocation, map fusion, and compares the advantages and disadvantages of various algorithms. Finally, this paper outlines the challenges and future research directions of multi-robot LiDAR SLAM.

In recent years, the deterioration of infrastructure facilities, such as bridges, has caused several problems. Currently, human inspectors conduct periodic inspections to identify damaged areas. However, this process is expensive and time-consuming. Therefore, robotic inspection has received significant attention. This study focused on magnet-wheeled inspection robots operating along complex multilevel paths. The movement from the bottom to the top surface of the flanges was particularly difficult, similar to that of an overhanging steel plate. As the motor drives the robot wheel, gravity and anti-torque interfere with the robot’s movement along its path. However, static analysis shows that the impact can be reduced depending on the robot’s posture relative to that of the flange. Therefore, a magnetic-wheeled robot with a posture-changing pushing mechanism is proposed. This study confirms that the proposed robot can travel along its path using a pushing mechanism while carrying a 1.0 kg weight. Therefore, the robot’s ability to conduct inspections while carrying heavy equipment, such as inspection devices, was confirmed.

Capturing dynamic targets is particularly challenging for either rigid or soft grippers, as impact buffering should be completed in a short time to ensure the reliability of the robotic system. At collision onset, to deal with relatively low contact forces, adopting low stiffness and damping can effectively mitigate the rebound of the dynamic targets. As the contact area and forces increase, employing high stiffness and damping becomes necessary for absorbing high energy. This paper proposed a novel robotic gripper whose stiffness and damping follow a predefined profile “low stiffness and damping for low impact and high stiffness and damping for high impact.” The variable effects of impact buffering and energy dissipation in a collision process were modeled and analyzed. Then, a passive variable stiffness and damping regulator (P-VSDR) was developed where tendons and pulleys are used to generate a nonlinear motion from a linear spring-damper unit. The contact dynamics model of the robotic gripper equipped with P-VSDR was established. Simulated and experimental results show that this gripper enables reliable capture of dynamic targets with different velocities.

The effectiveness of robotic grippers is critical for the secure and damage-free manipulation of objects with diverse geometries and material properties. This paper presents the design, analysis, and experimental evaluation of a novel reconfigurable four-finger robotic gripper. The proposed design incorporates two stationary fingers fixed to a circular base and two movable fingers repositioned and reoriented via a face gear mechanism, enabling multiple finger configurations to enhance adaptability. A single geared motor drives the opening and closing motions of all four fingers, simplifying the actuation mechanism. The robotic gripper was fabricated using 3D printing technology, ensuring cost-effective and precise manufacturing. Experimental tests were conducted to evaluate the robotic gripper’s reconfigurability and grasping performance across a range of objects, demonstrating its effectiveness in various configurations. Additionally, a closed-loop force control system was implemented to assess the grasping performance of a soft reconfigurable variant. Grasping force measurements were performed on three distinct objects, yielding a grasping curve that confirmed successful adaptation and secure handling. While the results validate the robotic gripper’s performance, further refinement of the control algorithm is recommended to optimize its capabilities. Compared to conventional three-finger designs, the proposed robotic gripper offers superior reconfigurability and adaptability, making it suitable for a broader range of industrial and research applications. The innovative face gear mechanism and modular design expand the robotic gripper’s functionality, positioning it as a versatile tool for advanced robotic manipulation tasks.

This paper presents four new monolithic continuum robot designs that can be 3D printed in a single piece and with TPU or similar elastic filaments for either educational or experimental applications. Similar tendon-driven continuum robots are usually made of a flexible backbone (often in NiTi alloys) and rigid vertebrae, with tens of components in a robot segment resulting in time-consuming manual assembly and high costs. Conversely, the proposed designs achieve equivalent functionality while avoiding the manufacturing challenges. Additionally, by removing the need for coupled features for assembly and 3D-printing backbones and vertebrae as a single part, new geometries are possible and can be explored to tailor robot performance to specific requirements. To validate the proposed design, four sample prototypes have been manufactured and experimentally tested. The obtained results, when compared to the piecewise constant curvature model, demonstrate a 3.06% tip positioning error and limited reduction of the workspace area of 23.07%, which compares favorably to similar but more expensive and complex tendon-driven robots.

The control of shipborne stabilisation platforms is challenging due to the effects of platform dynamic characteristics and unpredictable wave disturbances in operational environments. This paper proposes an integrated control strategy that combines dynamic feedforward and fuzzy gain control. Based on the derived dynamic model of the shipborne stabilisation platform, a dynamic feedforward controller is designed to mitigate the effects of platform dynamics on motion accuracy. In the fuzzy gain control design, scaling modules are proposed to enhance the fuzzy controller’s adaptability to varying operating conditions and unpredictable wave disturbances. The motion of the stabilisation platform is simulated by taking the motion of the lower platform calculated based on the wave fluctuations in marine environments as the input. The prototype experiment is conducted by using a large-scale parallel mechanism to simulate the wave environments. Simulation and experimental results indicate that the proposed control strategy achieves real-time disturbance compensation without precise mathematical modelling or pre-training, and demonstrates good adaptability.

6D pose estimation can perceive an object’s position and orientation in 3D space, playing a critical role in robotic grasping. However, traditional sparse keypoint-based methods generally rely on a limited number of feature points, restricting their performance under occlusion and viewpoint variations. To address this issue, we propose a novel Neighborhood-aware Graph Aggregation Network (NGANet) for precise pose estimation, which combines fully convolutional networks and graph convolutional networks (GCNs) to establish dense correspondences between 2D–3D and 3D–3D spaces. The $K$-nearest neighbor algorithm is integrated to build neighborhood relationships within isolated point clouds, followed by GCNs to aggregate local geometric features. When combined with mesh data, both surface details and topological shapes can be modeled. A positional encoding attention mechanism is introduced to adaptively fuse these multimodal features into a unified, spatially coherent representation about pose-specific features. Extensive experiments indicate that our proposed NGANet achieves a higher estimation accuracy on LINEMOD and Occlusion-LINEMOD datasets. In addition, its effectiveness is also validated under real-world scenarios.

In this study, we introduce a real-time pose estimation for a class of mobile robots with rectangular body (e.g., the common automatic guided vehicles), by integrating odometry and RGB-D images. First, a lightweight object detection model is designed based on the visual information. Then, a pose estimation algorithm is proposed based on the depth value variations within the target region that exhibit specific patterns due to the robot’s three-dimensional geometry and the observation perspective (termed as “differentiated depth information”). To improve the robustness of object detection and pose estimation, a Kalman filter is further constructed by incorporating odometry data. Finally, a series of simulations and experiments are conducted to demonstrate the method’s effectiveness. Experiments show that the proposed algorithm can achieve a speed over 20 Frames Per Second (FPS) together with a good estimation accuracy on a mobile robot equipped with an Nvidia Jetson Nano Developer KIT.

Kinematically redundant parallel mechanisms (PMs) have attracted extensive attention from researchers due to their advantages in avoiding singular configurations and expanding the reachable workspace. However, kinematic redundancy introduces multiple inverse kinematics solutions, leading to uncertainty in the mechanism’s motion state. Therefore, this article proposes a method to optimize the inverse kinematics solutions based on motion/force transmission performance. By dividing the kinematically redundant PM into hierarchical levels and decomposing the redundancy, the transmission wrench screw systems of general redundant limbs and closed-loop redundant limbs are obtained. Then, input, output, and local transmission indices are calculated, respectively, to evaluate the motion/force transmission performance of such mechanisms. To address the problem of multiple inverse kinematics solutions, the local optimal transmission index is employed as a criterion to select the optimal motion/force transmission solution corresponding to a specific pose of the moving platform. By comparing performance atlas before and after optimization, it is demonstrated that the optimized inverse kinematics solutions enlarge the reachable workspace and significantly improve the motion/force transmission performance of the mechanism.

The paper presents an enhanced method for unknown parameter estimation and nonlinear controller adaptation that combines the concept of unfalsification with the genetic algorithm (GA). This approach is based on the measured data and employs a bank of nonlinear controllers designed to dynamically adjust to the system’s evolving conditions. The controllers in the bank can be switched to meet the system’s requirements. This method is applied to an autonomous underwater vehicle (AUV) with uncertain parameters. Using the unfalsification method, these uncertain parameters are estimated, and a suitable controller is selected from the bank to guide the AUV along a desired trajectory. Additionally, an artificial intelligence technique, such as the GA, is employed to update the controller bank, resulting in versatile and optimised candidates. The simulation results obtained in the MATLAB/Simulink environment show that in the environment considered in this paper, the Adaptive Unfalsification algorithm in conjunction with GA estimates the unknown parameter values better than the sole GA-optimised values. Also, the convergence of the actual trajectory of the AUV using the Adaptive Unfalsification algorithm in conjunction with GA is faster and better than the sole GA-optimised algorithm. Furthermore, a survey of experimental results from established literature is included to evaluate the practical implementation of the proposed design, which concludes that within a reasonable time, the Adaptive Unfalsification algorithm in conjunction with GA can be implemented in commercially available processors.

In complex work environments, improving efficiency and stability is an important issue in robot path planning. This article proposes a new path optimization algorithm based on pseudospectral methods. The algorithm includes an adaptive weighting factor in the objective function, which automatically adjusts the quality of the path while satisfying the performance indicators of the shortest time. It also considers kinematic, dynamic, boundary, and obstacle constraints, and applies the Separating Axis Theorem collision detection method to improve computational efficiency. To discretize the continuous path optimization problem into a nonlinear programming problem, the algorithm utilizes Chebyshev polynomials for the interpolation of state and control variables, along with the adoption of the Lagrange interpolation polynomial to approximate the curve. Finally, it solves the nonlinear programming problem numerically using CasADi, which supports automatic differentiation. The results of the simulation demonstrate that the path optimized by the adaptive-weight pseudospectral method can satisfy various constraints and optimization objectives simultaneously. Experimental verification confirms the efficiency and feasibility of the proposed algorithm.

Robot hands are essential components of robots; however, the hand of more complex spatial mechanisms with coupling chains is rarely proposed. This paper proposes a hybrid hand with three underactuated finger plane limbs connected by a flexible closed-loop chain. The degree of freedom (DOF) of the hybrid hand is equal to the number of motors before grasping the object. When the contact force appears between the fingertips and the object, the flexible linkages deform, allowing the hybrid hand to maintain adaptability during contact. As the three fingers make contact with the object, the hybrid hand forms a closed-loop chain with the object, ensuring that the overall DOF remains consistent with the number of motors. Firstly, the hybrid hand’s structural characteristics and DOF are analyzed. Secondly, the kinematics of the hybrid hand are derived, and the relationships among the spring deformation, the kinematics of the fingertip and the input of the hybrid hand are obtained according to the geometric constraints. Thirdly, based on the kinematic results and the principle of virtual work method, the coupling dynamics formula of the hybrid hand is established, and the relationship between the dynamic driving force, dynamic constrained force, spring force and the force acting on the object is solved. Finally, the simulation model of the hybrid hand is constructed in MATLAB to validate the theoretical solution, and the merits of the hybrid hand were confirmed by prototype experiments. This paper aims to support a theoretical foundation for the intelligent control of novel hybrid hands.

Automatic visual localization of electric vehicle (EV) charging ports presents significant challenges in uncertain environments, such as varying surface textures, reflections, lighting and observation distance. Existing methods require extensive real-world training data and well-focused images to achieve robust and accurate localization. However, both requirements are difficult to meet under variable and unpredictable conditions. This paper proposes a 2-stage vision-based localization approach. Firstly, the image synthesis technique is used to reduce the cost of real-world data collection. A task-oriented parameterization protocol (TOPP) is proposed to optimize the quality of the synthetic images. Secondly, an autofocus and servoing strategy is proposed. A hybrid detector is employed to enhance sharpness assessment performance, while a visual servoing method based on single exponential smoothing (SES) is developed to enhance stability and efficiency during the search process. Experiments were conducted to evaluate image synthesis efficiency, detection accuracy, and servoing performance. The proposed method achieved 99% detection accuracy on the real-world port images, and guided the robot to the optimal imaging position within 16 s, outperforming comparable approaches. These results highlight its potential for robust automated charging in real-world scenarios.

With the increasing manufacturing of electric vehicles, car battery recycling is crucial for environmental sustainability. The disassembly of car batteries includes critical health hazards for the operator, due to potential chemical reactions or physical injuries. These reasons make robots particularly interesting for automatic disassembly. This paper proposes a systematic approach to automation and human–robot cooperation in car battery disassembly tasks with a case study on screw removal. A novel parameter is proposed to evaluate whether a human operator or a robot is more appropriate for each specific task, considering both performance and associated risks. The proposed metrics are validated with an experimental example, in which the performance of a robot and a human on a screw-removal task is compared numerically using statistical methods. The advantages and disadvantages of both options are examined through the application and show how the new performance criterion effectively provides insights into the distribution of tasks between humans and robots.

The human hand’s exceptional dexterity and compliance, derived from its rigid-soft coupling structure and tendon-driven interphalangeal coordination, inspire robotic grippers capable of versatile grasping and force adaptation. Traditional rigid manipulators lack compliance for delicate tasks, while soft robots often suffer from instability and low load capacity. To bridge this gap, we propose a biomimetic multi-joint composite finger integrating a 3D-printed rigid phalanges (46–51 mm) with dual fabric-reinforced pneumatic bladders, mimicking human finger biomechanics. This hybrid design combines hinge-jointed rigidity and anisotropic fabric constraints, enabling two rotational degrees of freedom with higher radial stiffness, achieving 2.18× higher critical burst pressure (240 kPa) than non-reinforced bladders, while preserving axial compliance. Experimental validation demonstrates a 4.77 N maximum fingertip force at 200 kPa and rapid recovery (< 2s) post-impact. The composite finger exhibits human-like gestures (enveloping, pinching, flipping) and adapts to irregular/fragile objects (e.g., eggs, screws) through coordinated bladder actuation. Assembled into a modular gripper, it sustains 1 kg payloads and executes thin-object flipping via proximal-distal joint synergy. This rigid-soft coupling design bridges compliance and robustness, offering high environmental adaptability for applications in industrial automation, human–robot interaction, and delicate manipulation.

Underwater robots conducting inspections require autonomous obstacle avoidance capabilities to ensure safe operations. Training methods based on reinforcement learning (RL) can effectively develop autonomous obstacle avoidance strategies for underwater robots; however, training in real environments carries significant risks and can easily result in robot damage. This paper proposes a Sim-to-Real pipeline for RL-based training of autonomous obstacle avoidance in underwater robots, addressing the challenges associated with training and deploying RL methods for obstacle avoidance in this context. We establish a simulation model and environment for underwater robot training based on the mathematical model of the robot, comprehensively reducing the gap between simulation and reality in terms of system inputs, modeling, and outputs. Experimental results demonstrate that our high-fidelity simulation system effectively facilitates the training of autonomous obstacle avoidance algorithms, achieving a 94% success rate in obstacle avoidance and collision-free operation exceeding 5000 steps in virtual environments. Directly transferring the trained strategy to a real robot successfully performed obstacle avoidance experiments in a pool, validating the effectiveness of our method for autonomous strategy training and sim-to-real transfer in underwater robots.

A wrist-hand exoskeleton designed to assist individuals with wrist and hand limitations is presented in this paper. The novel design is developed based on specific selection criteria, addressing all the Degrees of Freedom (DOF). In the conceptual design phase, design concepts are created and assessed before being screened and scored to determine which concept is the most promising. Performance and possible restrictions are assessed using kinematic and dynamic analysis. Using polylactic acid material, the exoskeleton is prototyped to ensure structural integrity and fit. Manual control, master-slave control, and electroencephalography (EEG) dataset-based control are among the control strategies that have been investigated. Direct manipulation is possible with manual control, nevertheless, master-slave control uses sensors to map user motions. Brain signals for hand opening and closing are interpreted by EEG dataset-based control, which manages the hand open-close of the exoskeleton. This study introduces a novel wrist-hand exoskeleton that improves usefulness, modularity, and mobility. While the numerous control techniques give versatility based on user requirements, the 3D printing process assures personalization and flexibility in design.

A finite-time adaptive composite integral sliding mode control strategy based on a fast finite-time observer is proposed for trajectory tracking of the Stewart parallel robot, considering unmodeled uncertainties and external disturbances. First, a global finite-time converging sliding mode surface composed of intermediate variables and integral terms is established to eliminate steady-state tracking errors. Next, a fast finite-time extended state observer is designed to compensate for uncertainties and external disturbances, improving the robustness of the control system. Finally, based on this, a finite-time sliding mode control rate is designed. The gain value is adjusted through an adaptive reaching law to reduce sliding mode chattering, and global finite-time convergence of the system is theoretically proven using Lyapunov theory. Experimental verification shows that the proposed control strategy has stronger robustness to uncertainties and external disturbances, faster error convergence, less chattering, and higher stability accuracy.