headers/packages headers/math headers/algorithms headers/names headers/revision headers/spacing In the dynamic field of unmanned aerial vehicles (UAVs), integrating biohybrid systems marks a significant shift, offering a transformative approach to aerial robotics. This paper examines the fusion of biological and synthetic components in UAVs, highlighting how biohybrid robots can enhance performance, resilience, and adaptability. It argues that this integration signifies a move from traditional mechanical systems to more versatile and sustainable solutions. We explore the theoretical foundations, practical applications, and ethical considerations of biohybrid UAVs, incorporating case studies and performance evaluations to present a well-rounded view of their current and future potential. The paper provides actionable insights for stakeholders in academia, industry, and policymaking, advocating for responsible innovation in this emerging area. By addressing the challenges and opportunities of biohybrid systems in UAVs, this research offers a balanced perspective on their potential impacts and ethical implications. ”Integrating Biohybrid Systems in UAVs: A Paradigm Shift for Aerial Robotics” stands as a comprehensive guide for advancing this field, highlighting a path toward a more efficient and ethically sound future in aerial robotics.

Bridge inspection constitutes a critical yet labor-intensive task in civil infrastructure maintenance, often requiring access to confined, structurally complex environments. Conventional manual inspection suffers from low efficiency and high operational risks, and the robotic solutions encounter limitations in GNSS-denied and low illumination environments with texture-deficient surfaces. This study proposes QRIVAS (quadruped robot based intelligent visual acquisition system), an autonomous framework for structural component image acquisition without relying on prior maps to reduce the workload for manual close-proximity inspection. QRIVAS integrates 3D LiDAR SLAM with real-time semantic segmentation, enabling reliable navigation and precise structural component identification. In this paper, we focus on the exploration and inspection of bridge column—a representative and critical structural component of bridge systems. Experimental validation across simulated concrete railway viaducts and physical laboratory-scale bridge models (1:3 scale) shows that QRIVAS achieved 100% navigation success rate in simulation environments and 96.7% average task navigation success rate across six bridge columns in laboratory-scale bridge specimen. Compared to the existing research, QRIVAS shows consistent performance improvements across varying tolerance conditions (25 cm and 50 cm radius), maintaining robust operation under both flat concrete floor and rough artificial grass terrain conditions. This work demonstrates the potential of AI-driven robotic systems to transform traditional infrastructure maintenance practices.

To advance the automation of timber harvesting, this paper addresses the challenges of high nonlinearity and uncertainty inherent in a harvester’s hydraulic manipulator by developing and validating a complete, high-performance trajectory control scheme. A kinematic model is established using the Denavit-Hartenberg (D-H) method, and an S-curve planning method generates smooth reference trajectories. A model-free Fuzzy PID adaptive controller is designed to achieve precise tracking. To validate its performance, a high-fidelity co-simulation platform was established. Simulation results demonstrate that, compared to a conventional PID controller’s 5% overshoot and 2.5s settling time, the proposed fuzzy PID controller eliminated overshoot and reduced settling time by 40% to 1.5s. However, in a more complex five-axis coordinated tracking task, a rigorous forward kinematics analysis revealed that while the controller ensures stable path following, the maximum end-effector tracking error reached approximately 1400 mm. This result highlights the significant challenge of kinematic error amplification in large-scale manipulators. It underscores that achieving high-precision positioning requires advanced control strategies beyond robust feedback alone.

Humanoid robots are expected to operate in unstructured environments where foot-soil interaction strongly affects mobility. This work develops a simulation framework that couples multibody dynamics with deformable terrain models to study locomotion under such conditions. We compare two approaches: the Discrete Element Method (DEM), which provides particle-level accuracy at high computational cost, and the Soil Contact Model (SCM), an empirical formulation calibrated through DEM-based virtual bevameter tests. Through a progression of experiments–from single-foot validation to full-robot walking, climbing, running, and jumping–we show that SCM predicts foot-soil forces and joint loads with accuracy sufficient for motion planning, while achieving near real-time efficiency. DEM, while more computationally demanding, remains essential for analyzing extreme maneuvers involving rapid soil deformation. Together, these results highlight a pathway for scalable, high-fidelity simulation of humanoid locomotion in granular environments, with direct implications for planetary exploration, disaster response, and other field robotics applications.

To address the inspection and exploration needs of complex hybrid scenarios integrating man-made structures, unstructured outdoor terrain, surface water, and underwater environments, this study develops a robot named Hexplorer—an amphibious biomimetic hexapod robot inspired by the fin movement of epaulette sharks, designed for mixed complex environment exploration. This study also proposes a corresponding gait control algorithm for its terrestrial and aquatic locomotion. Hexplorer has six amphibious legs (2 degrees of freedom each, driven by 12 motors total) with a simple, reliable mechanical structure that enables both land and water movement via a single mechanism. We evaluated its locomotion capabilities on structured ground, unstructured ground, surface water, and underwater; in repeated experiments, Hexplorer showed timely, accurate directional navigation and attitude maintenance. Its moving speed ranged from 0.025m/s to 0.125m/s, the mean absolute difference between measured and expected yaw angles was 0.11–0.61, and the variances of pitch/roll angles were 0.0931–2.1598 and 0.0126–1.5562, respectively. Results confirm Hexplorer can traverse various aquatic/terrestrial environments via its unique biomimetic mechanism and supporting control algorithm, while regulating navigation direction and maintaining body stability.

Autonomous landing for Unmanned Aerial Vehicles (UAVs) requires both precision and resilience against environmental uncertainties, capabilities that current approaches struggle to deliver. This paper presents a novel learning-based solution that combines an advanced multimodal transformer-based detector with a reinforcement learning formulation to achieve reliable autonomous landing behavior across varying scenario uncertainties. Beyond the integration of multimodality for robust target detection, this research incorporates a comprehensive analysis of the impact of state representation on decision-making performance. The proposed methodology is validated through extensive simulation studies and real-world field experiments conducted on physical UAV platforms under natural wind disturbances, demonstrating reliable transfer from simulated training environments to controlled outdoor conditions. Field experiments across varying initial conditions and wind stress confirm the system’s robustness, achieving landing precision of 0.10 ± 0.08 meters in outdoor trials, demonstrating centimeter-level accuracy that surpasses the meter-level precision of global positioning systems.

Robotic haircutting represents an emerging application of service robotics where safety is paramount. Unlike conventional collaborative or domestic tasks, haircutting requires prolonged operation in close proximity to the human head, one of the most vulnerable body regions, while relying on sharp or heated tools. These characteristics make safety concerns central not only to functional reliability but also to user acceptance. This paper reviews the safety landscape of haircutting robots by examining risks across the entire workflow and synthesizing lessons from related domains such as collaborative robotics, healthcare robotics, and autonomous systems. A three-layer framework---mechanical, sensor, and algorithmic---is used to organize existing approaches and identify how safety can be embedded from hardware design to control logic. Highlights opportunities and unresolved challenges in ensuring safe human--robot interaction in haircutting scenarios, providing guidance for future research and deployment.

Traditional simultaneous localization and mapping (SLAM) systems exhibit strong performance in static scenarios, while they continue to face significant difficulties in dynamic scenarios characterized by moving objects and sparse features, which adversely affect the robustness and accuracy of the systems. Current semantic dynamic visual SLAM approaches commonly incorporate either semantic segmentation or object detection techniques. However, the former tends to be computationally demanding, while the latter is often susceptible to errors and omissions in detection. This paper proposes a novel visual-inertial SLAM system designed for Dynamic environments integrating YOLOv5s-Asym and Scene flow, referred to as DAS-SLAM. Within the DAS-SLAM framework, a novel object detection method incorporating asymmetric module has been developed, resulting in enhanced accuracy in object detection. Furthermore, DAS-SLAM integrates the proposed YOLOv5s-Asym network with geometric constraints derived from scene flow to effectively remove dynamic feature points, thereby substantially improving the accuracy of the system. Extensive indoor and outdoor experiments indicate that the proposed approach exhibits enhanced accuracy and robustness in comparison to leading contemporary methods.

Compared to traditional predefined time, this paper proposes a novel predefined time stability theorem with separately configurable power terms. Based on this, a predefined time sliding mode control is designed. To address the adverse effects of output constraints on closed-loop system performance, a predefined time nonsingular terminal sliding mode control method based on a predefined constraint state barrier Lyapunov function is proposed for a two-joint uncertain robotic arm. This method employs an RBFNN to estimate the dynamic model and designs a predefined time sliding mode controller. By utilizing a barrier Lyapunov function, it ensures error convergence to a neighborhood of the origin within the predefined time without violating the constraint boundaries. Simulation results for the two-joint manipulator demonstrate that this method compensates for uncertainties while ensuring output constraints are always satisfied, achieving high-precision and robust trajectory tracking control. Convergence to a small neighborhood occurs within the user-specified time. Combined with RBFNN uncertainty compensation, this approach realizes high-precision, robust trajectory tracking control.

Reliable navigation of autonomous underwater vehicles (AUVs) depends on high-precision state estimation, while complex underwater environments render measurement data susceptible to interference from non-Gaussian noise, significantly degrading the performance of traditional filtering algorithms. This paper investigates nonlinear state estimation under heavy-tailed measurement noise using an improved generalized maximum correntropy criterion, especially in scenarios where the measurement noise characteristics are unknown. A generalized Student’s t maximum correntropy criterion is proposed to mitigate the impact of non-Gaussian noise disturbances in Doppler Velocity Log (DVL) measurements. Additionally, variational Bayesian inference is utilized to tackle the challenge posed by unknown prior characteristics. Subsequently, the mean error analysis and mean square error analysis of the proposed method are evaluated. Finally, the effectiveness of the algorithm is validated through numerical simulations and sea trial data.

Wildfires and deforestation pose significant threats to forest ecosystems, highlighting the need for large-scale initiatives to restore destroyed forest areas. Traditional reforestation methods are labor-intensive and require extensive infrastructure to cultivate a variety of native forest seedlings in nurseries. To enable large-scale direct seeding of native species, this paper presents the design, development, and field evaluation of a UAV-based seed-dropping system. The approach integrates a novel mechanical gravity-drop seeding mechanism with a multirotor UAV programmed to autonomously deploy individual seeds at target locations. The system uses biochar-coated seed balls, which facilitate the handling of multiple seed varieties and enhance overall germination success. Field evaluations at an active reforestation site in Northern Thailand have demonstrated the feasibility of the proposed solution in hilly remote areas representative of reforestation projects. Operating at a height of 4 m, the system achieves approximately 90% deployment success, with an accuracy within 1 meter of targeted locations. Unlike conventional aerial seeding systems, the proposed precision seed-dropping approach significantly reduces seed wastage, making the solution scalable to large-scale ecological restoration and biodiversity recovery initiatives.

Visibility underwater is challenging, and degrades as the distance between the subject and camera increases. That is why underwater computer vision tasks in the forward-looking direction are difficult. We have collected underwater forward-looking stereo-vision and visual-inertial image sets with two underwater imaging platforms, a stereo camera rig and an ROV in the Mediterranean and Red Sea. To our knowledge there are no other public datasets in the underwater environment with this camera-sensor orientation that have published ground-truth depth maps as well as pose. These datasets are critical for the development of several underwater applications, including autonomous obstacle avoidance, visual odometry, 3D tracking, Simultaneous Localization and Mapping (SLAM) and depth estimation through deep learning. The stereo datasets include synchronized stereo images in dynamic underwater environments with objects of known size. The visual-inertial datasets contain monocular images and IMU measurements, aligned with millisecond resolution timestamps and objects of known size which were placed in the scene. Both sensor configurations allow for scale estimation, with the calibrated baseline in the stereo setup and the IMU in the visual-inertial setup. Ground truth depth maps were created offline for both dataset types using a commercial photogrammetry software (Agisoft Metashape). The ground truth is validated with multiple known measurements placed throughout the imaged environment. There are four stereo and 12 visual-inertial datasets in total, each containing thousands of images, with a range of different underwater visibility and ambient light conditions, natural and man-made structures and dynamic camera motions. The forward-looking orientation of the camera plus the corresponding ground truth makes these datasets unique and ideal for testing underwater obstacle-avoidance algorithms and for navigation close to the seafloor in dynamic environments. We show results from an experiment with a monocular depth estimation algorithm to demonstrate the applicability of the datasets. With our datasets, we hope to encourage the advancement of autonomous functionality for underwater vehicles in dynamic and/or shallow water environments.

The human pose estimation technology based on robotic vision sensing systems has emerged as a critical perception approach for collision avoidance and safety assurance in human-robot collaboration (HRC). In dynamic occlusion scenarios, human pose estimation primarily relies on rapidly evolving deep learning models, yet it still faces challenges such as degraded recognition accuracy under prolonged occlusion and the high time cost of data collection and annotation. To address this issue, we proposed a novel skeletal pose and minimum-distance fusion (SPMF) approach which used a dual-RGB-D camera system to achieve robust human pose estimation under occlusion conditions. Firstly, the 3D coordinates of the joints were pre-estimated using the OpenPose-based skeletal model, in which the human body was represented by capsules at the joint positions. Following, the minimum human-robot distance was calculated via the Gilbert-Johnson-Keerthi (GJK) algorithm by combining the capsule model of the robot. In addition, a framework for human postural discrimination was established by combining the joint prediction confidence and minimum distance parameters, which can judge the reliability of pre-estimated pose data for occluded points. Finally, an algorithm for correcting misjudged joints was proposed which reconstructed the depth values for fully occluded joints. The experimental results demonstrated that the proposed human pose estimation framework achieved real-time collision detection under robot occlusion while delivering accurate pose recognition for industrial safety applications.

Autonomous Underwater Vehicles (AUVs) have emerged as indispensable tools for a variety of subsea tasks, from habitat monitoring and seabed mapping to infrastructure inspection and mine countermeasures. A fundamental challenge in this field is Coverage Path Planning (CPP), the problem of ensuring complete and efficient area coverage. Within this research activity, we propose a Deep Reinforcement Learning (DRL)-based framework for CPP in underwater environments using a Forward-Looking Sonar (FLS). We validate the proposed methodology through simulation experiments comparing it with the classical lawnmower path and a state-of-the-art sampling-based algorithm. Results demonstrate that our DRL-based solution outperforms these baseline approaches in terms of coverage time per unit area and path length. Additionally, we present on-field deployment outcomes on FeelHippo AUV, showcasing the feasibility and practicality of our framework in real-world underwater missions.

This work presents a non-geometrical navigation approach based on a purely topological understanding of underground environments. By conceptualizing subterranean scenarios as a set of tunnels that intersect with each other, and taking a navigation approach based on topological instructions, we simplify the navigation problem to the sequential execution of human-understandable instructions. This approach is built on top of a lightweight Convolutional Neural Network (CNN) that processes the readings of a 3D LiDAR sensor and produces an estimation of the angular positions of the surrounding tunnels with respect to the robot. As a result of this approach, our method can navigate these underground environments by only being provided with the necessary topological instructions, without the need for a map, or for building one during navigation. Additionally, it can also rely on a lightweight graph representation of the environment. This graph can be either defined by the user, generated during navigation or explicitly built in an exploration task. To showcase these capabilities, this article provides an experimental evaluation of the method both in simulation and in a real environment.

jabbrv-ltwa-all.ldf jabbrv-ltwa-en.ldf With the aging population and labor shortages, the proportion of labor costs in tomato harvesting is increasing, making the development of tomato harvesting robots imperative. This study developed an integrated tomato bunch harvesting robotic system for cherry tomatoes. A combined cutting and gripping end-effector powered by a single actuator, achieving a cutting success rate of 93.33% and a gripping capacity of 1600 g. A parameterized camera arrangement was employed to match the robotic arm’s field of view, thereby avoiding mutual interference. A tomato bunch and stalk recognition model was constructed based on the YOLOv4 algorithm to enable precise localization of picking points. The proposed tomato bunch–stalk matching method achieved a recall rate of 99.22%, while the fuzzy discrimination method for stalk growth posture attained an accuracy of 97%. Field experiments demonstrated that the system achieved an average harvesting time of 12.23 seconds per tomato bunch and an overall picking success rate of 70.77% in unstructured environments, improving automation and operational efficiency compared to existing solutions. This research offers a solution integrating hardware optimization and perception algorithms for greenhouse harvesting robots, demonstrating potential for commercial application.

Conventional satellite-based navigation systems do not provide required navigation performance and accuracy in indoor environments. This entails the development of indoor navigation systems that can either work independently or in conjunction with the satellite-based as well as on-board navigation systems. This paper discusses some of the existing indoor navigation systems before presenting a novel Acoustic Navigation and Guidance System (ANGuS). Echolocation exhibited by mammals like bats and dolphins as well as some visually impaired humans is discussed. Significant research efforts have focused on the dilution of precision for outdoor radio navigation systems, both terrestrial and satellite-based. However, indoor navigation systems pose considerably different challenges, which would potentially require a different approach from the techniques developed for outdoor navigation systems. One such approach would be how the issue of poor geometry is tackled for multistatic indoor navigation systems. Due to higher elevation angles, even a slight change in beacon arrangement indoors can significantly affect the geometry and consequently the position dilution of precision as well as the positioning error. An alternate positioning error estimation technique is proposed for high PDOP multistatic navigation systems, thus improving error budgeting and system design of indoor navigation systems. Experimental results of ANGuS, including static positioning tests, flight tests on an unmanned aerial platform, and ground tests on a tracked vehicle are discussed. A simulation case study looking at integration of on-board inertial measurement unit (IMU) with ANGuS is also discussed.

The exploration of Subsurface Access Points (SAPs), such as lava tubes on the Moon and Mars, has gained significant interest due to their potential as stable environments shielded from surface radiation and temperature extremes. These sites are considered high-value targets for detecting water, signs of ancient life, and assessing their suitability as habitats for human missions. However, SAP exploration presents significant challenges, including navigating unknown and hazardous terrains, operating in low-light conditions, and managing limited communication capabilities. Recent advances in high-resolution imaging, Synthetic Aperture Radar (SAR), and other sensing technologies have enabled better identification and characterization of SAPs, providing critical data for potential exploration missions. This review presents a structured critical analysis of the challenges in planetary cave exploration and evaluates the state of the art robotic platforms which offer a cost-effective and safe alternative to human exploration in hazardous environments, in addition to sensor technologies that aid the understanding of SAPs such as seismic studies, geological characterization, and biosignature detection. This article emphasizes the advantages of multi-robot teams in generating comprehensive datasets and improving mission resilience. By combining the unique capabilities of heterogeneous robotic systems, these teams represent a crucial step toward enabling the exploration of SAPs and advancing our understanding of planetary subsurface environments.