Search Results (1,150)

This paper proposes AutoRL, a high-fidelity Robot Operating System 2 (ROS2)–Gazebo simulation pipeline that addresses a critical reproducibility gap in learning-based flight control: existing reinforcement learning (RL) frameworks for unmanned aerial vehicles (UAVs) lack deterministic, step-level coupling between control actions and physics updates. AutoRL enforces a strict one-to-one correspondence between agent actions and physics updates via blocking ROS2 service calls, preserving the Markov property required for stable policy learning and enabling verifiable reproducibility independent of the learning algorithm. A composite reward function jointly optimizes attitude tracking accuracy, oscillation suppression, actuator smoothness, and disturbance robustness. Its modular, service-oriented architecture provides a reusable framework for offline-trained RL research. A proximal policy optimization (PPO) controller trained within AutoRL validates the framework, demonstrating consistent convergence and stable performance across multiple independently seeded runs. Determinism was experimentally verified across two regimes: with Gaussian IMU noise disabled, repeated rollouts produced bit-identical trajectories, while with noise enabled, the measured distribution of trajectory divergence agreed with a reference distribution drawn from the declared sensor model, together confirming that the blocking service call architecture eliminates all non-stochastic sources of nondeterminism between the agent and the physics engine. The trained model is exported in a lightweight form compatible with embedded flight control firmware and remains adaptable across airframe configurations by automatically recomputing the control allocation matrix from configuration files. The total policy network contains only $\mathrm{10,628}$ trainable parameters, and inference was measured on a Cortex-M7 microcontroller at $598.7$ µs per step, $15.0\%$ of the 4 ms control period. Full article

Affective computing has become increasingly relevant in child–robot interaction (CRI), particularly in social robotics, emotion recognition, engagement assessment, and autism-related interventions. This systematic review with a critical and integrative synthesis analyzes 105 included studies to examine how affect is sensed, represented, processed, expressed, and evaluated in CRI. The literature search was conducted in IEEE Xplore, Web of Science, Scopus, and PubMed, following a systematic screening process guided by the review objectives. A descriptive and structured narrative synthesis was conducted considering publication characteristics, robot platform and morphology, target population, sensing modalities and observed affect-relevant features, affective constructs and representation models, computational and control mechanisms, robot affective expression, evaluation strategies, and remaining research gaps. The findings show a strong emphasis on ASD-related contexts, visually observable and behavioral features, facial emotion recognition, body movement analysis, and engagement assessment. The review also identifies important limitations, including reliance on camera-based affect recognition, comparatively limited use of physiological and other complementary sensing modalities, unclear alignment between robot roles and interaction strategies, insufficient reporting of robot emotional expressiveness and control mechanisms, and limited attention to explainability, data governance, and long-term ethical implications. Based on these findings, an integrative taxonomy of affective computing in CRI is proposed, comprising six interconnected dimensions: interaction context; sensing modalities and observed features; affective constructs and representation models; computational and control mechanisms; robot affective expression; and evaluation and adaptation strategies. Rather than treating these dimensions as entirely novel categories, the taxonomy consolidates and extends previously fragmented classifications into a child-centered representation of the affective interaction process. Overall, this review argues that affective CRI should move beyond automatic emotion recognition toward multimodal, embodied, developmentally appropriate, explainable, and ethically grounded robot interaction. Full article

The identification and spatial positioning of welds are key links in welding process matching and automatic welding path planning. Therefore, achieving automatic recognition and spatial positioning of weld point cloud features under online scanning imaging conditions is of great significance for the promotion and application of teaching free automatic welding technology. This article is based on a point cloud neural network model and conducts in-depth research on the recognition and spatial positioning algorithm of weld point cloud features. Specifically, the PointNet++ network, which is a point cloud neural network model, is first used to perform feature recognition on the three-dimensional point cloud of the welded parts obtained by laser line scanning. PointNet++ can distinguish different types of welded joints based on point cloud features and further perform preliminary rough positioning of the spatial position of the weld seam. After retaining the coarse positioning point cloud containing weld seam features, different algorithms are used to accurately locate the spatial position of the weld seam based on different weld seam features. The experimental results show that based on the PointNet++model for rough positioning, the weld length error does not exceed 0.3 mm, and the recognition accuracy and efficiency are much higher than traditional algorithms. The research results of this article can provide important references for the further intelligent development of automatic welding robots. Full article

Coding and educational robotics are increasingly promoted in schools as catalysts for creative thinking, yet recent reviews suggest that this link is not automatic and depends critically on pedagogical design. Empirical evidence from Italian compulsory education remains scarce. We report preliminary findings from the Coding&Learning project, a quasi-experimental study involving 291 enrolled students (138 primary, 153 lower-secondary) and 23 trainers across four Italian schools. Verbal divergent thinking was assessed pre- and post-intervention with the Divergent Association Task (DAT); complete pre–post data were available for 217 students. All four groups improved between pre- and post-test, and within-group gains were statistically significant in both experimental groups (primary: +3.27 points, p = 0.031; lower-secondary: +2.61 points, p = 0.007) and in the lower-secondary control group. When experimental and control classes were compared directly—through gain-score t-tests and ANCOVA controlling for pre-test—the between-condition differences did not reach significance, and effect sizes were small (primary g = 0.18; lower-secondary g ≈ 0), as is not unexpected for a brief, statistically underpowered preliminary study. At the primary level, a larger share of experimental students improved (66.7% vs. 52.3%). Trainers reported coherent perceived gains in dispositions adjacent to verbal divergent thinking. Findings are specific to verbal divergent thinking and should not be generalised to creativity as a broader construct. Full article

Robotic bin-picking of disordered, randomly stacked workpieces remains challenging because reliable grasping depends on an accurate estimate of object pose, yet many established solutions require high-precision 3D sensing, detailed object models, or large annotated datasets that raise the cost and effort of deployment on a new production line. This work presents a complete binocular vision framework that estimates workpiece pose by image matching and executes vision-guided grasping on a 6-DOF manipulator. A pose-annotated multi-view template library is constructed automatically through robot-driven image acquisition and compressed by a coarse-to-fine clustering scheme, and object pose is estimated by discriminative template matching with rigid refinement. To characterize the geometric reliability of the matched poses, an offline cross-modal analysis relates the 2D templates to a 3D reference model of the object and measures their agreement through region and contour reprojection metrics. Grasp configurations are then generated under orientation and collision constraints and corrected online by closed-loop visual feedback. Experiments on two representative workpieces show template-matching accuracy of 89–90% against classical and learned similarity measures, and grasp success between 81 and 87% across single-object and mixed scenes, outperforming the GraspNet baseline under the tested conditions. The framework offers an accurate and deployment-friendly route to robotic bin-picking. Full article

Long-horizon robotic manipulation requires a policy to bridge task-level semantic reasoning with metric three-dimensional interaction geometry. Existing vision–language–action policies usually acquire geometry implicitly from visual tokens or introduce deterministic intermediate variables only in the image plane, which rely on expensive human annotations and training cost. This article presents a spatial Gaussian-guided hierarchical framework that uses ordered 3D Gaussian interaction regions as an explicit planning interface between vision–language reasoning and action generation. The proposed framework enables efficient adaptation of a pretrained vision–language model for robotic manipulation tasks. First, an automatic geometric enhancement pipeline converts raw robot demonstration videos into near-, mid-, and late-stage Gaussian supervision through foreground extraction, metric depth estimation, stable camera aggregation, end-effector localization, 3D lifting, and temporal grouping, without requiring manual 3D interaction annotation. The generated Gaussian representations provide structured spatial guidance, where their covariance characterizes interaction-region extent and variability rather than fully calibrated physical uncertainty. Second, a shared vision–language backbone predicts structured subtasks and Gaussian interaction regions, while a conditional diffusion executor generates future action chunks under these semantic and geometric conditions. A trajectory-to-Gaussian likelihood objective explicitly encourages consistency between generated motions and the predicted spatial interaction plan. Experiments on a mixed real-robot dataset derived from LHManip and RH20T show that our method improves trajectory tracking success from 55.7% to 70.8% over a same-backbone direct VLA baseline. Closed-loop simulation evaluation on LIBERO with 80% backbone parameter frozen achieves 85.3% average task success, demonstrating the effectiveness of explicit 3D interaction representations for spatial reasoning and long-horizon manipulation. Full article

Smart and sustainable construction increasingly depends on automation, yet robotic steel assembly still lacks task-specific perception data for bespoke connection systems. The Intermeshed Steel Connection (ISC) is a novel steel connection system that can reduce bolting effort and support faster, more reusable assembly, but dependable perception for ISC-aware robotic assembly remains underdeveloped. No public image corpus exists for ISC components, and collecting real site imagery is constrained by access, safety, privacy, and the limited deployment of ISC in practice. This paper introduces ISC-Perception, a hybrid vision dataset for near-field robotic assembly with novel Intermeshed Steel Connections. The dataset combines photorealistic CAD renders from SolidWorks Visualize, automatically annotated synthetic scenes generated in Unity, and a limited curated set of real ISC and human images. For a normalised 10,000-image Unity-based pipeline example, the proposed pipeline reduces estimated human effort to 30.5 h compared with 166.7 h for manual labelling, while the full training and validation set contains 15,928 images. Detectors trained on the hybrid dataset outperform synthetic-only and photorealistic-only alternatives, achieving mAP@0.50 of 0.756 on the complete test set. A near-size-matched comparison indicates that the improved performance is associated with the hybrid composition rather than dataset size alone under the evaluated training conditions. In a 1200-frame multi-view benchtop robotic assembly experiment, the detector achieves mAP@0.50/mAP@[0.50:0.95] of 0.943/0.823. These results show that ISC-Perception provides a practical route to data generation for emerging construction robotics applications where real imagery is scarce and supports the development of perception modules for robotic steel assembly in smart construction. Full article

Climate change is increasing the frequency of heat stress events in dairy farming, adversely affecting milk production, milk composition, cow behavior, and animal health. However, many previous studies relied on distant weather-station data and low-frequency milk recording systems, limiting the assessment of short-term and delayed heat stress responses. Therefore, the aim of this study was to combine dense longitudinal data from automatic milking systems (AMS) with continuously recorded on-farm meteorological measurements to investigate the immediate and lagged effects of heat stress on Holstein dairy cows. The study included 386,587 AMS visit records from 790 cows on three commercial dairy farms in Germany, corresponding to up to 127,310 cow-day records collected between August 2022 and August 2025. Temperature–humidity index (THI) values were calculated based on dense on-farm temperature and relative humidity records and were evaluated for multiple lag periods prior to AMS recordings. Linear mixed models were applied to infer the effects of THI on production, physiological, behavioral, and milking process traits. Increasing THI was associated with reduced daily milk yield, altered milk fat and protein percentages, decreased AMS visit frequency, prolonged milking intervals, and increased milk temperature. For contemporaneous THI, an increase from THI 50 to THI 70 corresponded to model-estimated declines of −0.86 kg in daily milk yield, −0.20% in milk fat content and −0.06% in milk protein content, −0.12 daily AMS visits, and +1.14 °C in milk temperature. The strongest associations were generally observed for prompt and short-term lagged THI windows. In contrast, longer lag periods were associated with weaker and less distinct trait responses. Rather than merely confirming the established decline in milk yield under heat stress, the integrated and temporally resolved analysis revealed trait-specific response patterns across production, behavioral, physiological, health-related, and milking-process traits. In particular, milk temperature and voluntary AMS attendance showed pronounced associations with contemporaneous and short-term THI, demonstrating the value of combining AMS-derived phenotypes with high-resolution on-farm climate data for heat stress monitoring. Full article

In industrial robot trajectory planning, trajectory segmentation has crucial importance in distinguishing between different geometric primitives, such as straight lines and arcs. Separating these elements facilitates allocating optimized motion instructions, customized to each segment category. This distinction ensures both improved execution smoothness and better operational performance. For this purpose, the Bidirectional Long Short-Term Memory (Bi-LSTM) machine learning algorithm has been implemented to segment trajectories into linear and arc-shaped parts, for which dedicated robotic commands can be used. First, several Bi-LSTM models with different architectures were trained using a synthetic dataset containing different shapes with labelled segments. Then, a theoretical study was performed to evaluate the accuracy of recognizing different shape segments using a test dataset. Finally, the generated trajectories, which implemented the best machine learning model, were transferred into the RoboDK software to launch the robot. Two different methods were used to generate trajectories for the UR3 industrial robot. The original trajectory was generated using linear interpolation only, while the second was generated using the machine learning algorithm. The experimental results show significant differences in terms of the smoothness and velocity profiles between these two trajectory generation methods. By enabling automatic classification of trajectory segments into line and arc primitives using the Bi-LSTM-based approach, the execution time is reduced by up to 43.5% and the vibration amplitude by up to 27.4% at higher speeds around 250 mm/s. However, this came at the cost of reduced accuracy at high speed, with reproduction error reaching 1.4–1.9 mm versus 0.46–0.7 mm for linear interpolation. Full article

Manual handling and packing of thin metal plates remains a labour-intensive operation in the automotive manufacturing sector, frequently requiring multiple operators and resulting in limited productivity, reduced process repeatability, and ergonomic constraints. This paper presents a preliminary virtual and experimental proof-of-concept for the automation of such a packing task using a Delta robot, motivated by a representative industrial scenario involving automotive heat exchanger plates. The proposed study adopts an intentionally simplified problem formulation to support early-stage feasibility assessment and system design. In the considered scenario, nominal pickup and placement coordinates are predefined in the robot control program. The plates are manually aligned with marked pickup areas, while their actual physical poses are not measured or automatically corrected, allowing the study to focus on the robotic packing stage rather than on perception-based localization. A virtual robotic packing cell is developed using MATLAB-based simulation tools, with emphasis on functional sequence verification, workspace reachability, and end-effector integration. A set of preliminary feasibility indicators is considered, including reachability of the predefined positions, execution of the manipulation sequence, cycle time, end-effector behaviour, and correspondence between the planned and experimentally achieved layouts. The virtual model is used to support planning and preliminary analysis, while a simplified experimental demonstrator is implemented to verify the physical execution of the inclined-tray packing task using representative plate geometries. The experimental results provide a controlled functional feasibility baseline and identify part-presentation accuracy, sensing, calibration, and cycle-time reduction as the main requirements for further development. Full article

Objectives: To evaluate the clinical accuracy and patient-reported outcomes (PROMs) of robotic-assisted full-arch implant rehabilitation performed with an autonomous robotic implant system. Materials and Methods: Six consecutive edentulous patients requiring implant-supported full-arch rehabilitation were treated using the Yakebot task-autonomous robotic implant system following a fully digital workflow. A total of 36 Straumann Bone Level Tapered implants were placed using a flapless approach. Implant placement accuracy was assessed by comparing planned and postoperative CBCT datasets. Coronal, apical, depth, and angular deviations were automatically calculated using dedicated robotic planning software. Patient-reported outcomes were prospectively evaluated through a structured questionnaire assessing preoperative perceptions, postoperative morbidity, and overall treatment satisfaction. Results: All implants were successfully placed according to the robotic-assisted workflow without intraoperative complications or conversion to conventional surgery. Mean three-dimensional coronal and apical deviations were 0.45 ± 0.18 mm and 0.47 ± 0.19 mm, respectively, while mean total angular deviation was 1.30 ± 0.63°. Patients reported low preoperative anxiety (0.5 ± 0.8/10), limited postoperative pain (2.2 ± 3.5/10), swelling (1.3 ± 2.2/10), and interference with daily activities (1.3 ± 2.2/10). Overall surgical experience was highly rated (9.7 ± 0.8/10). Final satisfaction scores were exceptionally high, with all patients indicating they would undergo robotic-assisted surgery again and recommend the procedure to others. Conclusions: Robotic-assisted full-arch implant rehabilitation demonstrated a high level of clinical accuracy, with submillimetric coronal and apical deviations and low angular discrepancies. In addition, treatment was associated with low postoperative morbidity, excellent patient acceptance, and very high satisfaction. These findings support the potential of autonomous robotic implant surgery as a predictable and patient-centered approach for full-arch rehabilitation, although larger controlled studies are required to confirm these results. Full article

Ultrasound examination for developmental dysplasia of the hip (DDH) in infants is highly dependent on operator experience, leading to inconsistent imaging quality and poor reproducibility between sonographers. This study proposes an autonomous robotic ultrasound system to improve the standardization and automation of hip ultrasound examinations. The system consists of a robotic arm, a six-axis force/torque sensor, an RGB-D camera and an ultrasound probe, integrating multiple functions including contact force control, visual localization, deep-learning-based segmentation and ultrasound image screening. To ensure stability and safety during scanning, an admittance-based hybrid force/position control strategy is adopted to achieve constant contact force control. For Graf standard plane acquisition, a stage-wise search strategy is designed, in which the search space is progressively narrowed through femoral head searching and multi-angle scanning. The optimal Graf standard plane is then automatically selected by combining image segmentation with a scoring mechanism. A customized hip phantom was used for validation. Experimental results show that the Dice coefficient for femoral head segmentation reaches 0.872, while the average Dice coefficient for multi-structure segmentation reaches 0.866. In 30 autonomous scanning trials, the success rate of Graf standard plane acquisition is 90.0%. Meanwhile, the system can maintain the contact force stably within the target range during scanning, validating the effectiveness of the force control strategy. These results indicate that the proposed robotic system, image recognition algorithm and visual servo control strategy exhibit favorable safety and feasibility, providing an innovative solution for automated infant hip ultrasound examination of DDH. Full article

Robot-assisted surgery has significantly expanded the possibilities of minimally invasive procedures by providing enhanced dexterity and visualization. However, the lack of direct haptic feedback still limits the surgeon’s ability to accurately assess instrument–tissue interactions, motivating the need for automatic intraoperative assistance systems. During vesicourethral anastomosis (VUA) in robot-assisted radical prostatectomy (RARP), accurate engagement of the bladder and urethral mucosa is essential to ensure proper tissue approximation and watertight closure. Nevertheless, automatic identification of fine-grained needle–tissue interactions during this phase remains largely unexplored. In this work, we propose a proof-of-concept framework for vision-based needle–tissue interaction analysis in RARP endoscopic videos, combining semantic segmentation, geometric proximity analysis, and motion coherence estimation to identify biomechanically plausible interaction events. Two independent transformer-based models were fine-tuned for semantic segmentation of the mucosal tissue and the surgical needle using a patient-level split of six real-world RARP procedures, comprising four procedures for training, one for validation, and one for independent testing. The models achieved Dice scores of 0.837 and 0.774, respectively. The segmentation outputs were subsequently used to drive a motion-aware interaction analysis pipeline, combining geometric proximity estimation between the needle endpoint and the mucosal tissue with optical-flow motion coherence analysis. The proposed interaction framework was evaluated on an independent test set, achieving a specificity of 0.933 and a recall of 0.667. An ablation study further demonstrated the complementary contribution of geometric proximity and motion coherence cues for needle–tissue interaction detection. Although limited by the retrospective nature and size of the dataset, this study introduces a low-latency, end-to-end framework for interaction-aware surgical scene understanding during RARP. The proposed approach represents an initial step toward the development of context-aware intraoperative guidance systems for robotic urologic surgery. Full article

With the growing global fleet of motor vehicles and rising demand for unmanned services, enhancing the efficiency and intelligence of refueling operations at gas stations has become a critical industry priority. This review focuses on refueling robots as its core research subject, providing a systematic review of its developmental history and system architecture. Building upon this foundation, this review conducts an in-depth analysis and synthesis of three key enabling technologies: (1) the end effector—integrating multi-degree-of-freedom actuators and sensor modules to precisely control fuel tank lid actuation and fuel nozzle insertion/removal; (2) refueling interface identification—enabling vehicle-type classification, refueling interface location extraction, and recognition of refueling interface features; and (3) refueling interface localization—determining the 6 DoF pose of the refueling interface relative to the robot. Through this technical analysis, it is shown that refueling robots have attained an initial level of intelligence; however, significant challenges remain in achieving high precision and robust performance, ensuring safety and reliability, and establishing standardization and broad interoperability. Future research efforts should therefore prioritize improving environmental adaptability—particularly in complex, unstructured settings—advancing autonomous decision-making capabilities, and enhancing product universality, thereby accelerating the commercial deployment of refueling robots. Full article

This paper presents a quantitative evaluation of an automatic translation system between industrial robot programming languages based on a sequence-to-sequence (Seq2Seq) neural architecture using Long Short-Term Memory (LSTM) networks. The study addresses the interoperability problem between proprietary robot programming languages by proposing a data-driven approach capable of learning correspondences between structured code instructions. A parallel dataset of 28,000 aligned instruction pairs was constructed and preprocessed through tokenization and normalization to enable structured sequence learning. The model was trained under four configurations (50,100, 150 and 200 epochs) to analyze the impact of training duration on performance and generalization capability. The system was evaluated using multiple quantitative metrics, including accuracy, loss, BLEU, and Exact Match (EM), allowing assessment of both structural similarity and exact sequence correctness. Experimental results demonstrate that the 200-epoch configuration improves the performance across all metrics, achieving an accuracy of 0.9943, a BLEU score of 0.682, and an Exact Match of 0.970 on the test set. These results indicate that the model is capable of generating both structurally consistent and syntactically correct translations. The analysis shows that while BLEU captures structural similarity, EM provides a stricter evaluation of exact sequence correctness, which is critical in structured code translation tasks where minor variations may affect execution. The proposed approach demonstrates the feasibility of applying neural machine translation techniques to industrial robot programming, contributing to improved interoperability and reduced manual effort in multi-platform robotic environments. Full article