Search Results (2,503)

The quality of underwater imagery is inherently degraded by light absorption and scattering, a challenge that severely limits its application in critical domains such as marine robotics and archeology. While existing enhancement methods, including recent hybrid models, attempt to address this, they often struggle to restore fine-grained details without introducing visual artifacts. To overcome this limitation, this work introduces a novel hybrid U-Net-Transformer (UTR) architecture that synergizes local feature extraction with global context modeling. The core innovation is a Recurrent Multi-Scale Feature Modulation (R-MSFM) mechanism, which, unlike prior recurrent refinement techniques, employs a gated modulation strategy across multiple feature scales within the decoder to iteratively refine textural and structural details with high fidelity. This approach effectively preserves spatial information during upsampling. Extensive experiments demonstrate the superiority of the proposed method. On the EUVP dataset, UTR achieves a PSNR of 28.347 dB, a significant gain of +3.947 dB over the state-of-the-art UWFormer. Moreover, it attains a top-ranking UIQM score of 3.059 on the UIEB dataset, underscoring its robustness. The results confirm that UTR provides a computationally efficient and highly effective solution for underwater image enhancement. Full article

Insulators on power lines require regular maintenance by operators in high-altitude hazardous environments, and the emergence of aerial manipulators provides an efficient and safe support for this scenario. In this study, a lightweight telescopic aerial manipulator system is developed, which can realize the installation and retrieval of insulator inspection robots on power lines. The aerial manipulator has three degrees of freedom, including two telescopic scissor mechanisms and one pitch rotation mechanism. Multiple types of cameras and sensors are specifically configured in the structure, and the total mass of the structure is 2.2 kg. Next, the kinematic model, dynamic model, and instantaneous contact force model of the designed aerial manipulator are derived. Then, the hybrid position/force control strategy of the aerial manipulator and the visual detection and estimation algorithm are designed to complete the operation or complete the task. Finally, the lifting external load test, grasp and installation operation test, as well as outdoor flight operation test are carried out. The test results not only quantitatively evaluate the effectiveness of the structural design and control design of the system but also verify that the aerial manipulator can complete the accurate automatic grasp and installation operation of the 3.6 kg target device in outdoor flight. Full article

Digital transformation powered by Building Information Modeling (BIM) and Artificial Intelligence (AI) is reshaping renovation practices by addressing persistent challenges such as fragmented records, scheduling disruptions, regulatory delays, and inefficiencies in stakeholder coordination. This study explores the integration of these technologies through a case study of a Catholic church renovation (2022–2023) in Hong Kong, supplemented by insights from 10 comparable projects. The research proposes a practical framework for incorporating digital tools into renovation workflows that focuses on diagnosing challenges, defining objectives, selecting appropriate BIM/AI tools, designing an integrated system, and combining implementation, monitoring, and scaling into a cohesive iterative process. Key technologies include centralized BIM repositories, machine learning-based predictive analytics, Internet of Things (IoT) sensors, and robotic process automation (RPA). The findings show that these tools significantly improve data organization, proactive planning, regulatory compliance, stakeholder collaboration, and overall project efficiency. While qualitative in nature, this study offers globally relevant insights and actionable strategies for advancing digital transformation in renovation practices, with a focus on scalability, continuous improvement, and alignment with regulatory frameworks. Full article

Navigating autonomous robots in confined channels is inherently challenging due to limited space, dynamic obstacles, and energy constraints. Existing sensor fusion strategies often consume excessive power because all sensors remain active regardless of environmental conditions. This paper presents an energy-aware adaptive sensor fusion framework for channel robots that deploys RGB cameras, laser range finders, and IMU sensors according to environmental complexity. Sensor data are fused using an adaptive Extended Kalman Filter (EKF), which selectively integrates multi-sensor information to maintain high navigation accuracy while minimizing energy consumption. An energy management module dynamically adjusts sensor activation and computational load, enabling significant reductions in power consumption while preserving navigation reliability. The proposed system is implemented on a low-power microcontroller and evaluated through simulations and prototype testing in constrained channel environments. Results show a 35% reduction in energy consumption with minimal impact on navigation performance, demonstrating the framework’s effectiveness for long-duration autonomous operations in pipelines, sewers, and industrial ducts. Full article

This study focuses on human awareness, a critical component in human–robot interaction, particularly within agricultural environments where interactions are enriched by complex contextual information. The main objective is identifying human activities occurring during collaborative harvesting tasks involving humans and robots. To achieve this, we propose a novel and lightweight deep learning model, named 1D-ResNeXt, designed explicitly for recognizing activities in agriculture-related human–robot collaboration. The model is built as an end-to-end architecture incorporating feature fusion and a multi-kernel convolutional block strategy. It utilizes residual connections and a split–transform–merge mechanism to mitigate performance degradation and reduce model complexity by limiting the number of trainable parameters. Sensor data were collected from twenty individuals with five wearable devices placed on different body parts. Each sensor was embedded with tri-axial accelerometers, gyroscopes, and magnetometers. Under real field conditions, the participants performed several sub-tasks commonly associated with agricultural labor, such as lifting and carrying loads. Before classification, the raw sensor signals were pre-processed to eliminate noise. The cleaned time-series data were then input into the proposed deep learning network for sequential pattern recognition. Experimental results showed that the chest-mounted sensor achieved the highest F1-score of 99.86%, outperforming other sensor placements and combinations. An analysis of temporal window sizes (0.5, 1.0, 1.5, and 2.0 s) demonstrated that the 0.5 s window provided the best recognition performance, indicating that key activity features in agriculture can be captured over short intervals. Moreover, a comprehensive evaluation of sensor modalities revealed that multimodal fusion of accelerometer, gyroscope, and magnetometer data yielded the best accuracy at 99.92%. The combination of accelerometer and gyroscope data offered an optimal compromise, achieving 99.49% accuracy while maintaining lower system complexity. These findings highlight the importance of strategic sensor placement and data fusion in enhancing activity recognition performance while reducing the need for extensive data and computational resources. This work contributes to developing intelligent, efficient, and adaptive collaborative systems, offering promising applications in agriculture and beyond, with improved safety, cost-efficiency, and real-time operational capability. Full article

Background: Nephrectomy in patients with polycystic kidney disease (PKD) is typically arduous due to the considerable size of the kidneys. The laparoscopic method has arisen as a minimally invasive substitute for open surgery. Nonetheless, conventional laparoscopic methods may be inadequate for tackling the distinct anatomical complexities of a large polycystic kidney. This study presents a unique method, “the hilum first technique”, specifically designed for nephrectomy in patients with PKD, emphasizing its safety and efficacy in addressing this intricate condition. Methods: A retrospective analysis of patients with PKD who underwent minimally invasive nephrectomy using “the hilum first technique” at our hospital between 2020 and 2025. Data on operative time, blood loss, conversion rates, hospital stay, and outcomes were analyzed to evaluate this technique’s safety and efficacy. Results: Minimally invasive nephrectomy using the “hilum first technique” was successfully performed in 16 cases; the mean age of patients was 56.3 years. Two of which were robot-assisted, in which one of them, a bilateral nephrectomy was done, with no conversions to open surgery, even for huge kidneys. The mean operative time was 159.6 min, with an estimated blood loss of 50 mL. Postoperatively, the median duration of hospital stay was 4 days (range: 3–10 days), and 75% of patients experienced no complications. Three patients (18.7%) were readmitted within 30 days due to surgical site infection, subcutaneous hematoma, and pneumonia. Seven patients (43.7%) underwent kidney transplantation within a median duration of 132 days post-nephrectomy. Conclusions: This retrospective study, although limited by a small sample size, demonstrated significant promise as a novel strategy for tackling the challenges of huge polycystic kidneys. The findings suggest its feasibility and safety, although further validation is required. Full article

With the growing application of soft robot technology in complex, dynamic environments, the limitations of traditional rigid robots have become increasingly prominent, urgently demanding novel soft actuation technologies. Dielectric elastomer actuators (DEAs) have gradually emerged as a research focus in soft robotics due to their high energy density, rapid response, low noise, and excellent compliance. This paper systematically reviews the research progress of DEA-based soft robots over the past decade. Using classification and comparative analysis, DEAs are categorized into four basic types according to their initial shape—planar, saddle-shaped, cylindrical, and conical—with detailed elaboration on their working principles, structural features, and typical applications. Furthermore, from two major application scenarios (underwater and terrestrial), this paper analyzes the adaptability of various DEAs in robot design and corresponding optimization strategies and summarizes their performance and research challenges in bionic propulsion, multi-modal motion, and environmental adaptability. Finally, it provides the prospective future research directions of DEAs in material development, structural design, intelligent control, and system integration, providing theoretical support and technical references for their wide application in fields such as medical treatment, detection, and human–robot interaction. Full article

This study introduces an autonomous framework for grasping moving objects on a conveyor belt, enabling unsupervised detection, grasping, and categorization. The work focuses on two common object shapes—cylindrical cans and rectangular cartons—transported at a constant speed of 3–7 cm/s on the conveyor, emulating typical scenarios. The proposed framework combines a vision-based neural network for object detection, a target localization algorithm, and a deep reinforcement learning model for robotic control. Specifically, a YOLO-based neural network was employed to detect the 2D position of target objects. These positions are then converted to 3D coordinates, followed by pose estimation and error correction. A Proximal Policy Optimization (PPO) algorithm was then used to provide continuous control decisions for the robotic arm. A tailored reinforcement learning environment was developed using the Gymnasium interface. Training and validation were conducted on a 7-degree-of-freedom (7-DOF) robotic arm model in the PyBullet physics simulation engine. By leveraging transfer learning and curriculum learning strategies, the robotic agent effectively learned to grasp multiple categories of moving objects. Simulation experiments and randomized trials show that the proposed method enables the 7-DOF robotic arm to consistently grasp conveyor belt objects, achieving an approximately 80% success rate at conveyor speeds of 0.03–0.07 m/s. These results demonstrate the potential of the framework for deployment in automated handling applications. Full article

This work presents an open-source simulation framework designed to extend the capabilities of the VRX environment for developing and validating control strategies for surface robotic vehicles. The platform features a custom monohull, kayak-type USV with four thrusters in differential configuration, represented with a complete graphical mockup consistent with its physical design and modeled with realistic dynamics and sensor integration. A thrust mapping function was calibrated using manufacturer data, and the vehicle’s behavior was characterized using a simplified Fossen model with parameters identified via Least Squares estimation. Multiple motion controllers, including velocity, position, trajectory tracking, and path guidance, were implemented and evaluated in a variety of wave and wind scenarios designed to test the full vehicle dynamics and closed-loop behavior. In addition to extending the VRX simulator, this work introduces a new USV model, a calibrated thrust response, and a set of model-based controllers validated in high-fidelity marine conditions. The resulting framework constitutes a reproducible and extensible resource for the marine robotics community, with direct applications in robotic education, perception, and advanced control systems. Full article

This paper presents tilt-and-pivot manipulation, a robotic technique for picking large, thin objects resting on hard supporting surfaces. The method employs in-hand dexterous manipulation by reorienting the gripper around the object’s contact point, allowing a finger to enter the gap between the object and the surface, without requiring relative sliding at the contact. This finally facilitates reliable pinch grasps on the object’s faces. We investigate the kinematic principles and planning strategies underlying tilt-and-pivot, discuss effector design considerations, and highlight the practical advantages of the strategy, which is applicable to a variety of low-profile objects. Experimental results, incorporating vision and force–torque sensing, demonstrate its effectiveness in bin-picking scenarios and its applicability to more complex object-handling tasks. Full article

Rigid–flexible coupled robots hold significant potential for operating in unstructured environments, but a systematic analysis of their actuation strategies across diverse physical fields is notably lacking in the literature. This review addresses this gap by introducing a novel taxonomy based on field-controlled evolutionary pathways—mechanical → electromagnetic → chemical → biohybrid—and critically examining over 100 seminal studies through a six-dimensional framework encompassing design, dynamics, and performance. We demonstrate that hybrid field integration (e.g., pneumatic-chemical synergy) improves grasping robustness by 40% in cluttered environments compared to single-field systems. Notably, biohybrid actuators, which integrate living cells, exhibit over 90% motion similarity to biological models, while phase-transition materials allow for adaptive stiffness tuning (0.1–5 N·mm⁻¹) in medical applications. Radar chart analysis further reveals fundamental trade-offs between energy efficiency, response speed, and scalability across the various fields. These insights provide a clear roadmap for the development of next-generation robots with embodied intelligence, emphasizing multi-field synergies and bio-inspired adaptability. Full article

The advancement of adaptable industrial robots in intelligent manufacturing is hindered by the inefficiency of traditional motion planning methods in high-dimensional spaces. Therefore, a Dempster–Shafer evidence theory-based hybrid motion planner is proposed, in which a probabilistic sampling network (PSNet) and an enhanced artificial potential field (EAPF) cooperate with each other to improve the planning performance. The PSNet architecture comprises two modules: a motion planning module (MPM) and a fusion sampling module (FSM). The MPM utilizes sensor data alongside the robot’s current and target configurations to recursively generate diverse multimodal distributions of the next configuration. Based on the distribution information, the FSM was used as a decision-maker to ultimately generate globally connectable paths. Moreover, the FSM is equipped to correct collision path points caused by network inaccuracies through Gaussian resampling. Simultaneously, an augmented artificial potential field with a dynamic rotational field is deployed to repair local paths when worst-case collision scenarios occur. This collaborative strategy harmoniously unites the complementary strengths of both components, thereby enhancing the overall resilience and adaptability of the motion planning system. Experiments were conducted in various environments. The results demonstrate that the proposed method can quickly find directly connectable paths in diverse environments while reliably avoiding sudden obstacles. Full article

The adoption of robots in the construction phase can improve safety by replacing hazardous tasks and enhance productivity by automating repetitive work. Despite these advantages, adoption remains slow, constrained by economic, industrial, institutional, socio-cultural, and technological barriers. Wider acceptance is particularly urgent in construction, where fragmented processes, low profit margins, and safety risks make innovation both necessary and challenging. This study identified 22 critical barriers through a systematic literature review and categorized them into five dimensions. Beyond identification, the study prioritized these barriers using ISM and MICMAC analysis, clarifying which factors are fundamental drivers and which are outcome-related. The results showed that economic drivers occupy the base of the hierarchy and exert the greatest systemic influence, socio-cultural barriers emerge as highly dependent outcomes, and software usability acts as a linkage factor connecting technological immaturity with social acceptance. These findings reveal that barriers are interdependent rather than isolated and underscore the need for a structured prioritization framework. By applying ISM and MICMAC, this study presents a stepwise roadmap that differentiates fundamental drivers from outcome-related constraints, offering academic insights and practical guidance for policymakers to design strategies such as investment incentives, standardization, legal frameworks, and R&D expansion to accelerate adoption. Full article

To achieve a complete circular economy for used electric vehicle batteries, it is essential to implement a disassembly step. Given the significant diversity of battery geometries and designs, a high degree of flexibility is required for automated disassembly processes. The incorporation of human–robot interaction provides a valuable degree of flexibility in the process workflow. However, human behavior is characterized by unpredictable timing and variable task durations, which add considerable complexity to process planning. Therefore, it is crucial to develop a robust strategy for coordinating human and robotic tasks to manage the scheduling of production activities efficiently. This study proposes a global optimization approach to the scheduling of production activities, which employs a genetic algorithm with the objective of minimizing the total production time while simultaneously reducing the idle time of both the human operator and robot. The proposed approach is concerned with optimizing the sequencing of disassembly tasks, considering both temporal and exclusion constraints, to guarantee that tasks deemed hazardous are not executed in the presence of a human. This approach is based on a two-level adaptation framework developed in RoboDK (Robot Development Kit, v5.4.3.22231, 2022, RoboDK Inc., Montréal, QC Canada). At the first level, offline optimization is performed using a genetic algorithm to determine the optimal task sequencing strategy. This stage anticipates human behavior by proposing disassembly sequences aligned with expected human availability. At the second level, an online reactive adjustment refines the plan in real time, adapting it to actual human interventions and compensating for deviations from initial forecasts. The effectiveness of this global optimization strategy is evaluated against a non-global approach, in which the problem is partitioned into independent subproblems solved separately and then integrated. The results demonstrate the efficacy of the proposed approach in comparison with a non-global approach, particularly in scenarios where humans arrive earlier than anticipated. Full article

The construction industry faces persistent productivity shortfalls and rising carbon dioxide emissions, which drives a shift toward the use of low-carbon materials and higher degrees of automation. Timber, a renewable and carbon-sequestering material, becomes especially compelling when combined with robotic fabrication. Although rapid advances have been implemented in the last decade, research and practice remain fragmented, and systematic evaluations of technological readiness are scarce. This gap is addressed in this review through critical literature synthesis of robotic timber construction, combining bibliometric analysis with a comparative evaluation of twelve representative case studies from 2020 to 2025. Computational and robotic tools are mapped across the design to fabrication pipeline, and emerging advancements are identified such as digital twins, real-time adaptive workflows, and machine learning driven fabrication, alongside discrete and circular strategies. Barriers to scale up are also assessed, including mid-level technology readiness, regulatory and safety obligations for human–robot interaction, evidence on cost and productivity, and workforce training needs. By clarifying the current level of robotization and specifying both research gaps and industrial prerequisites, this study provides a structured foundation for the next phase of development. It helps scholars by consolidating methods and metrics for rigorous evaluation, and it helps practitioners by highlighting pathways to scalable, certifiable, and circular deployment that align cost, safety, and training requirements. Full article