This article presents a novel vascular interventional guidewire and catheter manipulator based on purely compliant mechanisms, adopting a technical approach different from that of existing surgical robots. Inspired by human manual manipulation of guidewires and catheters, the proposed manipulator adopts a biomimetic operation paradigm, enabling axial and rotational actuation while simultaneously measuring the interaction force and torque in real time. By employing purely compliant mechanisms, the design is highly simplified, with integrally manufactured core components and no rigid kinematic pairs. A symmetric architecture is adopted to enable a modular design, facilitating assembly and maintenance. Experimental results validate the performance of the proposed biomimetic compliant manipulator, and operational safety is demonstrated through a force-limiting control strategy based on the integrated force-sensing capability. The proposed design concept provides guidance for the development of vascular interventional surgical robots and is potentially applicable to mechanical systems requiring multiaxis manipulation and sensing.

In-situ manufacturing with mobile equipment offers an effective solution for producing large-scale, complex components. In mobile manufacturing, measurement is essential for achieving coordinate system (CS) alignment between the equipment and workpiece. However, obtaining precise and efficient measurements is challenging. To address this, a visual-tactile fusion CS alignment method is proposed, utilizing fast visual recognition to guide the tactile probe for accurate measurement. To resolve datums’ misrecognition caused by lighting conditions and material variations, a structured-light aided visual recognition method with color-space enhancement is introduced, enabling high-efficiency and robust datum recognition. To counteract deviations due to the probe’s anisotropic behavior in tactile measurement, an analytical mapping model between the probe’s anisotropic trigger error and the parallel robot’s parasitic motion is established, and a precise trigger error compensation table for CS alignment is generated. Experiments conducted on a mobile hybrid machining robot demonstrate that the proposed method achieves an alignment accuracy of 0.026 mm and 0.013°, with measurement times under 10 min. The accuracy fully meets the machining requirements for large-scale structures at 0.2 mm, and efficiency is improved threefold compared to traditional laser tracker measurement methods.

Force and tactile sensing play increasingly important roles in robotics hands to complete complex tasks safely and effectively, complementing the last piece of the puzzle of human-level dexterity. Despite the advances in flexible sensors and skins, tactile sensing at the fingertip of robotic hands remains an open question. Exploiting embedded six-axis force/torque (F/T) sensors for external stimuli estimation, namely, intrinsic tactile sensing, seems to be an effective solution. Here, we present OriCube, an origami-inspired, fingertip-size (14.5 × 14.5 × 12.5 mm<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup>), lightweight (4 g), fully integrated, inductive six-axis F/T sensor. It utilizes an origami coil array to achieve self-decoupled six-axis F/T measurements by detecting the displacements and rotations of a metal shell via eddy–current coupling. A modular silicone spring-based elastomer design allows fine-tuning of the stiffness and performance of each axis independently. The calibrated OriCube exhibits a high resolution of 0.01% F.S. (23 N), low crosstalk (2%), high stability (<0.2% drift in 6 h), and insensitivity to strong magnetic field changes. It is capable of detecting light touch by a feather as well as high impact force when hammered. We demonstrate real-time tactile sensing (contact position and force vector) at the curved surface of a robotic fingertip via one embedded OriCube sensor.

Rolling element bearings are commonly used in rotating machines. However, reliable and real-time bearing condition monitoring and fault diagnosis is a challenging problem in industry, due to limitations associated with reliable real-time data acquisition (DAQ) and signal processing in noisy environments where bearings are often housed. In this article, a smart sensor-based bearing fault detection technology is proposed. The proposed method introduces a DAQ design for high resolution vibration signal measurement. A wavelet swarm decomposition (WSD) technique is also developed for stand-alone bearing fault detection. The proposed WSD technique adopts a new wavelet filter to adaptively determine center frequencies in the time-frequency domain. Moreover, an ensemble function is introduced to select decomposition modes and improve processing efficiency. The effectiveness of the proposed smart sensor bearing fault detection method is validated experimentally to detect different types of bearing faults. Test results show that the proposed method powered by the developed WSD technique performs better (6% higher) than other related techniques in bearing fault detection and diagnosis.

Hybrid continuum robots with concentric tubes and cable-driven mechanisms offer compactness and dexterity. However, they face limited adaptability and modeling challenges. This article proposes, for the first time, a multimode operation strategy for the composite continuum robot, enabling switching among high stiffness, high flexibility, and high precision modes. By adjusting the relative position between the concentric tube and the notched continuum, the robot achieves flexible transitions among these modes. A static model based on Euler–Bernoulli beam theory is established, where the bending moments from the concentric tube acting on the notched continuum segment are analytically integrated to reflect coupling effects. Algorithms are designed to achieve the mapping between the driving force and the shape, as well as the mapping relationship between the driving cable length and the robot shape. Based on the Kepler oval and the established static model, the inverse kinematics model of the robot is constructed. Simulations and experimental results show a modeling error of only 0.042 mm (0.084%), outperforming conventional models. The robot demonstrates convenient mode switching, showing performance characteristics, such as high stiffness, high flexibility, and high precision. This work contributes to improved modeling and control strategies for continuum robots.

Assisting divers with robotic exoskeletons would increase their working efficiency and reduce the risk of accidents. Nonetheless, the synchronization of assistance with underwater movements has not been investigated and remains a complex issue. Due to the agility and individuality of these movements in aquatic environments, conventional methods utilized on land may not work properly. In this article, we proposed an ultrarobust adaptive oscillator (AO) (URAO) method for predicting the kicking phase for the flutter kick and breaststroke kick, two main underwater motion patterns in diving activity. The URAO works without the need for data collection or training and can adapt to different users, motion patterns, and rapid changes in kicking frequency, offering greater precision than methods previously utilized for land-based exoskeletons. Eight participants were tested while performing fluctuating-frequency flutter and breaststroke kicks in indoor and outdoor swimming pools. The results showed that URAO outperformed both conventional AO and state-of-the-art estimators in predicting underwater kicking phases, reducing <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">RMS_EWS</i> prediction error by up to 49.5% for flutter kicks and up to 44.8% for breaststroke kicks.

This article proposes a backstepping integral sliding mode control (SMC) scheme augmented with an adaptive bandwidth-tuned NESO for valve positioners with time-varying dynamics and disturbances. First, cascaded tracking differentiators filters based on the Fhan function generate smooth reference trajectories and PD-spectrum required by the observer with low phase lag. Next, an improved time-varying multiplier, driven by the observer’s tracking error and desired velocity, schedules the observer gains and bandwidth online, balancing high-precision state estimation and noise suppression during composite motion. The resulting adaptive bandwidth-tuned NESO supports the backstepping integral SMC, in which an integral sliding surface and a smoothed <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$fa{{l}^{*}}$</tex-math></inline-formula> nonlinearity is embedded to shape cascaded error dynamics, suppress overshoot, and improve steady-state accuracy. Experimental comparisons have demonstrated the efficacy of the proposed nonlinear control strategy, which has been shown to enhance control speed and accuracy while effectively resisting perturbations and reducing the overshoot phenomenon.