Tian Xu

This research proposes a multi-modal pipe-climbing robot design and a position-based force-tracking adaptive impedance controller for pipe surface inspections and maintenance in unstructured industrial environments. The driving foot combines underactuated linkages with four Mecanum wheels for secure gripping and omnidirectional mobility. A two-level compliance strategy addresses the challenge of maintaining a stable gripping force on pipes of different diameters. Passive compliance provides preliminary mechanical adaptation. Inspired by the regulation of muscle spindle sensitivity by γ motor neurons (γ-MN), a three-parameter adaptive impedance controller was designed at the control level. By continuously estimating the contact environment parameters online and dynamically updating the impedance reference trajectory, the controller achieved zero steady-state gripping force error under the stated steady-state conditions. Lyapunov analysis established uniform stability of the ideal estimator for constant parameters and convergence of the force prediction error under the stated assumptions, and global uniform asymptotic stability (GUAS) of the parameter error under persistent excitation. Simulations verified the force-tracking performance of the controller under the specified contact parameter variations and measurement noise conditions. Experiments separately evaluated the locomotion performance of the prototype driving foot and its fixed-base force-tracking performance on pipes of three diameters. At target forces of 120 N and 200 N, the across-trial averages of the relative deviations of the steady-state means for the different pipe diameters ranged from 0.125% to 0.336%, and the across-trial standard deviations of the steady-state means ranged from 0.19 to 0.96 N, confirming small mean force deviations and good repeatability for the tested pipe diameters.

Penetrant Testing (PT) is a critical non-destructive inspection method in industry. While robotic automation improves PT repeatability and safety, it demands accurate and scalable spatial measurement. Traditional high-precision instrumentation lacks scalability and vision-based methods struggle to balance computational efficiency and the preservation of spatially critical information. This paper develops a monocular vision target localization method for automated PT systems, focusing on deterministic spatial-cue preservation during feature compression and motion-consistent depth refinement. The proposed approach integrates a feature-based weighting module into a depth estimation network, selectively preserving essential spatial cues while suppressing irrelevant details to enhance robustness under constrained representation lengths. A motion-aware depth optimization step then enforces inter-frame consistency by leveraging known manipulator motions as a strong geometric prior. Experiments on standard indoor benchmarks show consistent accuracy gains. More importantly, evaluations on a novel simulated PT dataset show a significant performance improvement (∼27.2% RMSE reduction) over representative baselines. A full workflow demonstration further confirms its operational efficacy and measurement reliability in real-world PT scenarios.

Monocular SLAM has gained widespread popularity due to its numerous advantages. However, this approach is susceptible to appearance changes and often experiences problems such as sensitivity of dynamic objects and scale drift. Using semantics as an advanced visual feature is highly effective since it remains unaffected by appearance and viewpoint changes. This makes it easier to establish a more robust data association. In this paper, we represent the semantic information of features as the semantic probability that they belong to each category, thus reducing the adverse effects of semantic segmentation noise and increasing the ability of semantic representation. Furthermore, we propose a method that combines the semantic probability with geometric features to enhance the accuracy and robustness of the SLAM system. Specifically, the odometry module chooses geometric-semantic feature point descriptors, which help achieve higher accuracy in feature matching and eliminate the influence of dynamic features. It also provides an initial estimation of the real scale. The back-end optimization module utilizes semantic probability to establish medium-term constraints, effectively mitigating the cumulative drift of attitude and scale. The loop detection module constructs global and local descriptors using semantic information and incorporates the DBOW2 algorithm based on geometric features, enabling accurate and robust loop detection. The experimental results demonstrate that the proposed algorithm achieves high-performance pose estimation on the test sequences compared to other advanced monocular SLAM methods. Compared with DBOW2 based on appearance, the AUC (area under the curve) of our loop detection method is increased by 28.1% while remaining lightweight.

This article presents a multisensory measurement synchronization approach that combines a laser tracker and multiple photoelectric switches, which easily embed laser tracker into robotic machining systems to achieve high-accuracy error compensation and trajectory tracking. Previous composite error compensation systems rely on limited interfaces provided by equipment producers and require careful temporal logic arrangement, making the whole system complicated and task-specific. The proposed method utilizes portable photoelectric switches that work as an intermediary between the robot system and the laser tracker system to match the corresponding data from both sides. The switches could be mounted anywhere along the robot task trajectories, and a critic trigger adjustment is designed to effectively determine the trigger position. Various types of trajectory are executed and measured to validate the effectiveness and robustness of the proposed method, and subsequent error compensation experiments demonstrate a significant improvement in robot accuracy.

This paper introduces a novel climbing robot for tube-sheet inspection (CRTI) that uses inner wall grippers (IWGs) to grasp tubes, enabling it to hang and crawl beneath the tube-sheet plane. The robot is designed primarily for inspecting steam generators in nuclear power plants. The CRTI based on a planar parallel mechanism is proposed, with its workspace optimized through the solution space method to address the limitations of existing systems in adaptability, load capacity, and efficiency. A modular IWG based on a wedge mechanism is introduced, featuring a load capacity exceeding 100 kg without damaging tubes, with its configuration optimized through a geometric method to enable insertion into smaller apertures. Furthermore, a forward kinematic solution method for parallel robots based on coordinate transformation is proposed, providing a unique analytical solution for the forward kinematics of all parallel configurations of this CRTI. The performance evaluation, which is grounded in singularity analysis and typical experiments, reveals that the proposed CRTI offers high efficiency, substantial load capacity, and exceptional adaptability, indicating its significant potential for future applications.

Dynamic identification is an important issue in robotics. Typically, using outlier data weighting and nonlinear friction models can effectively enhance the accuracy of dynamic identification. However, existing data weighting methods can cause data discontinuity and even regressor singularity. The modeling accuracy of the traditional Stribeck friction model can be further improved. Moreover, specifying arbitrary initial values for the nonlinear friction coefficients can decrease the overall identification efficiency of the convex optimization algorithm. To address these issues, this article proposes an identification framework based on data weighting protection and nonlinear friction separation. To deal with the outliers, a novel data weighting method is developed, which replaces outliers with the weighted average values of nearby nonoutlier data points, thus avoiding data discontinuity and regressor singularity. To improve the accuracy of the traditional Stribeck friction model, the arctangent function is used to replace the sign function, and the exponential velocity function is used to replace the linear one to describe the viscous friction phenomenon. In addition, a symmetric excitation trajectory is designed to separate the friction torques and then estimate the initial values of the nonlinear friction coefficients by data fitting, which can improve the computation efficiency of the proposed algorithm. Experimental results validate the correctness and superiority of the proposed algorithm. For instance, compared to other two methods, the proposed approach improves joint torque prediction accuracy by more than 12% and enhances identification efficiency by more than 92%.

Accurate payload dynamics benefit to robotic applications, such as model-based control. However, existing methods rely on joint torque signals for identifying payload dynamics. For these robots which do not equip with joint torque sensors, the torques can be determined merely via the multiplication of actuator currents and joint torque constants. Unfortunately, the uncertainty and identification errors of joint torque constants can lead to deviations in joint torque estimation, which in turn increase the cumulative identification errors for payload dynamics. To tackle this problem, a new approach about identification for payload dynamics employing joint currents rather than joint torques is developed in this article. Unlike traditional methods that use estimated joint torques, the proposed method only uses measured joint currents. It avoids the identification errors of joint torque constants and the estimation errors of joint torques, thereby reducing the cumulative identification errors in payload dynamics. To the best of authors’ knowledge, it is an initial effort in identifying payload’s current dynamics. This article has four main contributions. The first one is to develop a novel model for payload identification according to current dynamics. The second one is to propose an identification method for estimating payload’s current dynamics based on iterative weighting. The third one is to integrate a zero-velocity continuous nonlinear friction model and apply it to identification of current dynamics. Ultimately, experimental evaluations are performed to compare the proposed approach with four other methods, and the results demonstrate the advantages of the proposed approach.