Search Results (254)

A safe physical human–robot interaction (pHRI) in rehabilitation requires reliable perception and low-latency decision making under heterogeneous and unreliable sensor inputs. This paper presents a multimodal sensor-fusion-based safety framework that integrates physical state estimation, semantic information fusion, and an edge-deployed large language model (LLM) for real-time pHRI safety control. A dynamics-based virtual sensing method is introduced to estimate internal joint torques from external force–torque measurements, achieving a normalized mean absolute error of $18.5\%$ in real-world experiments. An asynchronous semantic state pool with a time-to-live mechanism is designed to fuse visual, force, posture, and human semantic cues while maintaining robustness to sensor delays and dropouts. Based on structured multimodal tokens, an instruction-tuned edge LLM outputs discrete safety decisions that are further mapped to continuous compliant control parameters. The framework is trained using a hybrid dataset consisting of limited real-world samples and LLM-augmented synthetic data, and evaluated on unseen real and mixed-condition scenarios. Experimental results show reliable detection of safety-critical events with a low emergency misdetection rate, while maintaining an end-to-end decision latency of approximately 223 ms on edge hardware. Real-world experiments on a rehabilitation robot demonstrate effective responses to impacts, user instability, and visual occlusions, indicating the practical applicability of the proposed approach for real-time pHRI safety monitoring. Full article

Background: Embodied artificial intelligence (EAI), integrating advanced AI algorithms with robotic platforms capable of sensing, planning, and acting, has emerged as a transformative approach in healthcare delivery. This systematic review synthesizes evidence on robotic perception, decision-making, and clinical impact of EAI systems in healthcare settings. Methods: Following PRISMA 2020 guidelines, we searched PubMed/MEDLINE, Scopus, Web of Science, IEEE Xplore, and ACM Digital Library for studies published between January 2020 and August 2025. Seventeen studies met eligibility criteria, spanning four domains: surgical assistance, rehabilitation, hospital logistics, and telepresence. The protocol was prospectively registered in PROSPERO under ID: CRD420261285936. Results: Perception architectures predominantly employed multimodal sensor fusion, combining vision with force/torque, depth, and physiological signals. Decision-making approaches included imitation learning, reinforcement learning, and hybrid symbolic-neural control. Key findings indicate that surgical robots demonstrated consistency advantages in specific experimental tasks, rehabilitation robotics produced statistically significant improvements (SMD = 0.29) across 396 randomized controlled trials, and both logistics and telepresence systems achieved very high operational success levels. Nonetheless, important barriers remain, including limited external validation, small sample sizes, and insufficient cost-effectiveness data. Conclusions: Future research should prioritize standardized benchmarks, prospective multicenter trials, and patient-centered outcome measures to facilitate clinical translation of EAI technologies. Full article

In this paper, we present a physical interaction scheme for omnidirectional multirotor aerial vehicles (MRAVs) equipped with fixedly tilted non-coplanar propellers, based on an admittance control architecture. An external wrench observer is employed to estimate the interaction wrench at the end-effector, hence eliminating the need for an additional force/torque sensor. We show that using the nominal allocation matrix in this class of admittance controllers can lead to a contact loss during complex interaction scenarios due to unmodeled and state-dependent aerodynamics effects. To address this issue, we propose a method for identifying the wrench map across different regions of the vehicle’s orientation in $SO\left(3\right)$ using free-flight experimental data. This is achieved by formulating a Quadratic Programming (QP) optimization whose solution provides the best approximation of the wrench map for a given orientation of the MRAV. The effectiveness of this approach is experimentally demonstrated, including static point contacts at various orientations, sliding contact, and peg-in-hole tasks. Full article

Upper limb rehabilitation exoskeletons form a spatial closed kinematic chain with the human arm, where inevitable joint-center and axis misalignment can generate hyperstatic interaction forces and torques. Passive degrees of freedom (DOF) are widely introduced to improve kinematic compatibility, yet different compatible configurations may exhibit distinct wearable performance. This study experimentally compares three compatible four-degree-of-freedom exoskeleton configurations derived from the synthesis of Li et al. using a single reconfigurable rehabilitation robot. The platform is assembled into each configuration through modular passive units and instrumented with two six-axis force–torque sensors at the upper-arm and forearm interfaces. Interaction forces and torques are measured in passive training mode during eating and combing trajectories. For each configuration, tests are performed with passive joints released and with passive joints locked to quantify the effect of passive motion accommodation. Directional and resultant metrics are computed using mean and peak values over movement cycles. Results show that releasing passive joints consistently reduces interaction loading, and Category 2 achieves the lowest forces and torques with the strongest peak suppression, indicating the best practical compatibility. Full article

Capacitive sensors are widely adopted in compact robotic systems due to their simple structure, ease of fabrication, and scalability for miniaturized designs. However, sensor miniaturization inevitably leads to reduced sensitivity and increased sensitivity imbalance, particularly in torque measurements, due to limited electrode area and spatial constraints. To address these limitations, this paper presents a compact six-axis force/torque (F/T) sensor based on a redundant capacitive sensing architecture. The proposed sensing architecture employs a symmetric arrangement of multiple capacitive electrodes, providing redundant capacitance measurements that enhance sensitivity while reducing coupling errors under multi-axis loading conditions. By exploiting redundant capacitive responses rather than relying on complex mechanical separation, the proposed design effectively improves measurement robustness. Based on this architecture, a compact six-axis F/T sensor with a diameter of 20 mm and a height of 12 mm is developed. Experimental validation demonstrates that the proposed sensor achieves linearity (>98.2%) with reduced cross-axis interference, confirming improved sensitivity and reliable multi-axis F/T measurement. This work provides a practical and scalable solution for integrating high-performance six-axis F/T sensing into space-constrained robotic systems. Full article

This study investigated the effects of 6 weeks of plyometric training (PT) performed on soft (unstable) and hard (stable) surfaces compared with conventional training on the balance, explosive power, and muscle strength of adolescent female basketball players. The participants were randomly assigned to three groups: soft-surface PT (n = 14), hard-surface PT (n = 14), and conventional training (n = 14). Performance outcomes included 30 m sprint time, vertical jump height, plantar flexion and dorsiflexion maximal voluntary isometric contraction (MVIC) torque, Y-balance dynamic balance, and center of pressure-based static balance. Ground reaction forces, MVIC torques, and balance parameters were measured using high-precision force sensors to ensure accurate quantification of biomechanical performance. Statistical analyses were performed using two-way repeated-measures ANOVA with post hoc comparisons to evaluate group × time interaction effects across all outcome variables. Results demonstrated that soft- and hard-surface PT significantly improved sprint performance, vertical jump height, and plantar flexion MVIC torque compared with conventional training, while dorsiflexion MVIC increased similarly across all the groups. Notably, soft-surface training elicited greater enhancements in vertical jump height, dynamic balance (posteromedial and posterolateral directions), and static balance under single- and double-leg eyes-closed conditions. The findings suggest that PT on an unstable surface provides unique advantages in optimizing neuromuscular control and postural stability beyond those achieved with stable-surface or conventional training. Thus, soft-surface PT may serve as an effective adjunct to traditional conditioning programs, enhancing sport-specific explosive power and balance. These results provide practical guidance for designing evidence-based and individualized training interventions to improve performance and reduce injury risk among adolescent female basketball athletes. Full article

Objectives: Shoulder joint strength and muscle activation during overhead reaching are critical for ergonomic task design, rehabilitation, and exoskeleton support. The objective of this study was to characterize maximum shoulder torque and flexor muscle activation profiles across functional elevation angles in healthy adult males. Methods: A total of 14 healthy male participants performed maximum voluntary isometric contractions at eight arm elevation angles (90–160°, sagittal plane, and standing). Shoulder torque was measured using a calibrated force sensor and normalized to each participant’s overall maximum. Electromyography (EMG) was recorded from the anterior deltoid, medial deltoid, biceps brachii, and clavicular pectoralis major; EMG for the medial deltoid, biceps brachii, and pectoralis major was normalized to muscle-specific isometric MVCs, whereas the anterior deltoid was normalized to the peak value at 90° during the main task. All EMG signals were smoothed using a 0.5 s RMS-based moving average window. Linear regression was used to analyze the torque–angle relationship, and linear mixed-effects models were used to test EMG differences across angles. Summary statistics included mean ± SD, coefficient of variation, R², p-values (significance threshold: p < 0.05), Cohen’s d, and 95% confidence intervals where appropriate. Results: Maximum torque declined with elevation angle (y = −0.6317x + 157.21; R² = 0.99), from 77.2 Nm at 90° to 43.2 Nm at 160°, with normalized values from 99.6% to 55.3%. Medial deltoid activation increased significantly with elevation (p < 0.001, from 87.5 ± 19.9% at 90° to 109.4 ± 25.6% at 150°), while pectoralis major declined sharply (p < 0.001, from 68.9 ± 24.2% at 90° to 19.8 ± 5.6% at 160°). Anterior deltoid and biceps brachii activations were high and showed no systematic change with angle (p = 0.37 and 0.81, respectively), remaining within approximately 95–102% and 70–85% of their reference levels across 90–160°. Normalization reduced inter-participant variability, clarifying muscle-specific trends. Conclusions: This study provides preliminary biomechanical reference values for shoulder torque and muscle activation across elevation angles in healthy males under isometric standing conditions, confirming an inverse torque–angle relationship and distinct muscle activation strategies at higher positions. These findings may inform ergonomic assessment and exoskeleton design, while recognizing that generalization to dynamic tasks and other populations requires caution. Full article

The underwater hexapod robot has advantages such as lower energy consumption and reduced environmental interference compared to ROVs and AUVs. The foot-end contact detection with the seabed is the key technology for adapting to complex terrains. This paper focuses on the ‘Dragon Crab’ underwater hexapod robot developed by Shanghai Jiao Tong University and proposes an indirect detection method that does not require foot-end contact sensors. By establishing the kinematic and dynamic models of the robot’s legs, combined with multi-order polynomial trajectory planning to reduce non-contact force interference, the foot-contact determination condition is defined. Through simulation experiments and force analysis of the legs, the contact detection parameters are estimated. Then, single-leg contact tests are conducted to obtain joint motor torque variation curves and foot-end height variation curves through the kinematic model, verifying the proposed contact detection conditions and parameters. Finally, the method is applied to underwater obstacle-crossing experiments of the underwater hexapod robot using triangular and wave gait patterns. Experimental results show that the method can accurately identify the foot-end contact state and has high applicability in complex underwater terrains. Full article

This paper introduces an adaptive impedance control strategy for robotic manipulators, developed through the extremum seeking technique. A model-based disturbance observer (DOB) is employed to estimate contact forces, removing the dependency on torque sensors. An impedance vector is constructed to correct the errors arising from motor uncertainties and unknown couplings, without considering the threshold value of the control parameters. Joint tracking errors and fluctuations in contact force are incorporated into the cost function. For various tasks, suitable control parameters are adaptively optimized in real time using an extremum seeking approach, which continuously evaluates the cost function. A rigorous analysis is conducted on the stability of the proposed controller. Compared to conventional approaches, the proposed adaptive impedance control offers a more streamlined design for adjusting the manipulator’s contact impedance. Experimental results confirm that the extremum seeking strategy successfully tuned the controller parameters online according to variations in the cost function. Full article

Monitoring spindle bearing load is essential for ensuring machining accuracy, reliability, and predictive maintenance in machine tools. This paper presents an approach that combines drive-based cutting force estimation with a finite element method (FEM) spindle model. The drive-based method reconstructs process forces from the motor torque signal of the feed axes by modeling and compensating motion-related torque components, including static friction, acceleration, gravitation, standstill, and periodic disturbances. The inverse mechanical and control transfer behavior is also considered. Input signals include the actual motor torque, axis position, and position setpoint, recorded by the control system’s internal measurement function at the interpolator clock rate. Cutting forces are then calculated in MATLAB/Simulink and used as inputs for the FEM spindle model. Rolling elements are replaced by bushing joints with stiffness derived from datasheets and adjusted through experiments. Force estimation was validated on a DMC 850 V machining center using a standardized test workpiece, with results compared against a dynamometer. The spindle model was validated separately on a MCV 754 Quick machine under static loading. The combined approach produced consistent results and identified the front bearing as the most critically loaded. The method enables practical spindle bearing load estimation without external sensors, lowering system complexity and cost. Full article

To address the automation of table grape harvesting, a clamping and cutting integrated, four-point flexible end-effector is designed, based on the biological and mechanical characteristics of grapes. The clamping device is validated in regard to force closure requirements using a force spiral. On this basis, a finite element model of the grape pedicel–blade system is established, and dynamic simulations of pedicel cutting are conducted using ANSYS 2021/LS-DYNA. The simulation results indicate that when the pedicel diameter is 10 mm, the maximum shear stress is 1.515 MPa. A kinematic simulation of the clamping device is performed using ADAMS, producing a contact force curve between the end effector’s finger joints and the grape during the clamping process. The simulation results show that the peak contact force of 11 N is lower than the critical rupture force of the grape (24.79 N), satisfying the requirements for flexible, low-damage harvesting. Furthermore, to address the vulnerability of grapes, a contact-force control system is designed, employing a position–speed–torque three-loop control strategy. Pressure sensors integrated into the four clamping fingers provide real-time feedback to adjust the contact force, ensuring precise clamping control. Finally, a physical prototype of the end effector and controller is developed, and harvesting trials are conducted in a vineyard. The harvesting success rate reaches 96.7%, with an average harvesting time of 13.7 s per trial. The grape cluster damage and berry drop rates are 3.2% and 2.8%, respectively, meeting the expected design requirements. Full article

Background: The back kick is a key scoring technique in taekwondo, often exhibiting bilateral asymmetry in lower limb function. Understanding these differences is crucial for optimizing training and minimizing injury risk. Methods: This study recruited twelve elite taekwondo athletes to perform back kicks using both their dominant and non-dominant legs under standardized conditions. Kinematic, kinetic, and surface electromyographic data were synchronously collected using a 3D motion capture system, force plate, and sEMG sensors. Paired t-tests and effect sizes assessed bilateral differences. Results: During the leg-lifting phase (P1), attacking leg peak hip power was significantly greater on the non-dominant side (p < 0.01); knee flexion angle was greater on the dominant side (p < 0.01), yet peak knee power was higher on the non-dominant side (p < 0.01). Support leg knee flexion angle was greater on the dominant side (p < 0.01), while knee flexion torque was higher on the non-dominant side (p < 0.05); ankle extension moment (p < 0.05) and plantar flexion power (p < 0.01) favored the dominant side. In the kicking phase (P2), dominant knee power was significantly higher (p < 0.01). The biceps femoris on the non-dominant side showed significantly higher iEMG and RMS values (p < 0.05), and dominant striking speed was faster (p < 0.05). Conclusions: These findings confirm marked functional asymmetry, suggesting training should emphasize non-dominant leg development to improve performance and reduce injury risk. Full article

Drilling is fundamental to the assembly of aerospace structures, where millions of fastening holes must meet stringent structural and geometric requirements. Despite significant technological advances, hole quality remains sensitive to nonlinear and stochastic interactions between mechanics, thermal effects, tribology, and structural configuration. This review consolidates recent advances in intelligent drilling, focusing on how sensors and artificial intelligence (AI) are integrated to enable process understanding, prediction, and control. In-process monitoring modalities (e.g., cutting forces/torque, vibration, acoustic emission, motor current/active power, infrared thermography, and vision) are examined with respect to signal characteristics, feature design, and modelling choices for real-time anomaly detection, tool condition monitoring, and phase/interface recognition. Predictive quality modelling of burr, delamination, roughness, and roundness is discussed across statistical learning, kernel methods, and neural and hybrid models. Offline parameter optimisation via surrogate-assisted and evolutionary algorithms is considered alongside adaptive control strategies. Practical aspects of robotic drilling and multifunctional end-effectors are highlighted as enablers of embedded sensing and feedback. Finally, cross-cutting challenges (e.g., limited, heterogeneous datasets and model generalisability across materials, tools, and geometries) are outlined, together with research directions including curated multi-sensor benchmarks, multi-source transfer learning, physics-informed machine learning, and explainable AI to support trustworthy deployment in aerospace manufacturing. Full article

The six-degree-of-freedom (6-DoF) interaction forces/torque of the tool-end play an important role in the robotic tool manipulation using a gripper, which are usually indirectly measured by a robot wrist force/torque sensor. However, the real-time decoupling of the tool’s inertial force remains a challenge when different tools and grasping postures are involved. This paper presents a universal tool-end interaction forces estimation approach, which is capable of handling diverse grippers and tools. Firstly, to address uncertainties from varying tools and grasping postures, an online-identifiable tool dynamics model was built based on the Newton–Euler approach for the integrated gripper–tool system. Sensor zero-drift caused by factors such as the tool weight and prolonged operation is incorporated into the dynamic model and identified online in real time, enabling a coarse estimation of the interaction forces. Secondly, a spiking neural network (SNN) is specially employed to compensate for uncertainties caused by the wrist sensor creep effect, since its temporal processing and event-driven characteristics match the time-varying creep effects introduced by tool changes. The proposed method is experimentally validated on a robotic arm with a gripper, and the results show that the root mean square errors of the estimated tool-end interaction forces are below 0.5 N with $x$, $y$, and $z$ axes and 0.03 Nm with ${\tau }_x$, ${\tau }_y$, and ${\tau }_z$ axes, which has a comparable precision with the in situ measurement of the interaction forces at the tool-end. The proposed method is further applied to robotic scraper manipulation with impedance control, achieving the interaction forces feedback during compliant operation precisely and rapidly. Full article

Magnetically controlled spinal growing rods, used for treating early-onset scoliosis (EOS), face a critical clinical limitation: insufficient distraction force. Compounding this issue is the inherent inability to directly monitor the mechanical output of such implants in vivo, which challenges their safety and efficacy. To overcome these limitations, optimizing the rotor’s maximum torque is essential. Furthermore, defining the “continuous rotation domain” establishes a vital safety boundary for stable operation, preventing loss of synchronization and loss of control, thus safeguarding the efficacy of future clinical control strategies. In this study, a transient finite element magnetic field simulation model of a circumferentially distributed permanent magnet–rotor system was established using ANSYS Maxwell (2024). The effects of the clamp angle between the driving magnets and the rotor, the number of pole pairs, the rotor’s outer diameter, and the rotational speed of the driving magnets on the rotor’s maximum torque were systematically analyzed, and the optimized continuous rotation domain of the rotor was determined. The results indicated that when the clamp angle was set at 120°, the number of pole pairs was one, the rotor outer diameter was 8 mm, the rotor achieved its maximum torque and exhibited the largest continuous rotation domain, while the rotational speed of the driving magnets had no effect on maximum torque. Following optimization, the maximum torque of the simulation increased by 201% compared with the pre-optimization condition, and the continuous rotation domain was significantly enlarged. To validate the simulation, a rotor torque measurement setup incorporating a torque sensor was constructed. Experimental results showed that the maximum torque improved from 30 N·mm before optimization to 90 N·mm after optimization, while the driving magnets maintained stable rotation throughout the process. Furthermore, a spinal growing rod test platform equipped with a pressure sensor was developed to evaluate actuator performance and measure the maximum distraction force. The optimized growing rod achieved a peak distraction force of 413 N, nearly double that of the commercial MAGEC system, which reached only 208 N. The simulation and experimental methodologies established in this study not only optimizes the device’s performance but also provides a viable pathway for in vivo performance prediction and monitoring, addressing a critical need in smart implantable technology. Full article