Journal of Field Robotics最新文献

This study investigates the landing dynamics of a telescopic-legged robotic rover on granular surfaces of small celestial bodies, addressing the challenges posed by its high-degree-of-freedom structure. Using a custom module within the PolyDEM discrete element framework, the interaction between complex rigid bodies and nonspherical particles was accurately simulated and validated through ground-based experiments. The results reveal that retracted legs exhibit superior damping performance on coarse gravel layers by increasing localized contact area, enhancing friction, and promoting internal particle rearrangement. Conversely, extended legs perform better on fine-grained regolith by increasing penetration depth and enabling multi-point interactions, which amplify energy absorption. Additionally, this study examines the effects of impact velocity and angle on rover stability, providing actionable insights for terrain-specific deployment strategies. These findings advance our understanding of high-DOF robotic systems in extraterrestrial environments and offer practical guidance for future mission planning.

The development of maritime trade and operations has led to a gradual increase in maritime accidents. Research on the specificity of maritime search and rescue (SAR) missions is essential to improve mission efficiency; however, traditional SAR programs usually use predefined paths to conduct searches, which is difficult to meet the timeliness and uncertainty requirements. To solve the challenge, we propose a Boundary Adaptive Neural Network Coverage Path Planning Scheme based on Target Drift Prediction (BANCP-TDP). The framework includes three modules: drift trajectory prediction, optimal region determination, and coverage search. First, the Limited Red-billed Blue Magpie Optimizer Back Propagation drift model is used to predict the drift trajectory of the wrecked target. Subsequently, we use the Multiphysics Monte Carlo Gravitational Search prediction model to determine the distribution of targets at different moments and the optimal SAR region for guiding an autonomous underwater vehicle to carry out SAR missions. Then, we propose a BANCP for the SAR regions with complex boundaries, aiming to minimize the path length and maximize the coverage ratio. The comparative field experiments at Qingdao Jin Cao Gou reservoir and simulation results show that the proposed framework can effectively shorten path length while minimizing the repeated paths.

This paper presents an innovative strategy to improve the landing capability of the solar-powered unmanned aerial vehicles (UAVs) with a low-speed and high-aspect-ratio blended wing body configuration by an innovative combination between the swallow tails and distributed differential propellers. It is proved that adjusting the opening angle of the swallow tails improves the UAV's stability, while higher yawing control efficiency of the differential throttle enhances the control efficiency at small throttle during the landing approach phase. The innovative yawing control method combining the opening of the swallow tails and a moderate separation of local airflow is validated via numerical flight simulations and landing flight tests. Thus, the damping ratio of the phugoid mode significantly increases, the lateral–directional control ability of the UAVs is enhanced in a simple but efficient way, the response speed is shortened, and the trajectory tracking accuracy is improved, even at higher altitudes and strong wind disturbances.

Precise robotic manipulation driven by visual perception is essential for a robot arm to autonomously execute tasks in unstructured environments. Nevertheless, uncertainties including inaccurate calibration, imperfect kinematic modeling, and dynamic environmental disturbances pose significant challenges to existing manipulation strategies. Considering that a robot arm finally performs tasks through the end-effector, we propose a monocular end-effector pose estimation-based visual servoing (EEVS) method. Our method first uses a lightweight detection network to obtain a coarse pose and a mask of the end-effector, and then refines the pose according to an optimal segmentation model. The visual servoing control of the robot arm is realized based on the relative pose between the end-effector and the target. We theoretically demonstrate that this end-to-end control scheme can effectively mitigate the impact of the uncertainties. Experiments on data set demonstrate that our method achieves more accurate end-effector pose estimation than state-of-the-art approaches, especially in the scenes with significant occlusion. Moreover, real-world robotic experiments prove that our method provides higher manipulation accuracy compared with traditional methods even in the presence of substantial kinematic errors and environmental changes.

Swarm robots provide significant promise for autonomous exploration, disaster response, environmental monitoring, and military operations. Nevertheless, most existing platforms are limited to simulations or lack the full-stack integration necessary for practical deployment. Although our prior simulation-based research validated the theoretical viability of decentralized swarm behaviour, its actual execution continues to pose challenges. To bridge the gap, in the current research work, Woxbots, an economical and scalable swarm robotics platform designed for the experimental validation of pattern formation and collision-aware route planning using velocity obstacle techniques, has been introduced. The current system was equipped with strong multiagent communication, real-time obstacle avoidance, and velocity-dependent trajectory regulation. The modular design, constructed using Raspberry Pi and ESP32 microcontrollers, guarantees dependable odometry, easy interface, and adaptable robot coordination. A PID controller improves dynamic responsiveness and trajectory accuracy. Extensive experimental trials with various autonomous agents confirm the platform's reliability, flexibility, and real-time efficacy in dynamic, limited settings. Woxbots provide a cost-effective and practical solution, facilitating the development of scalable swarm robotic systems for academic research, industrial automation, and mission-critical operations.

Autonomous underwater vehicles (AUVs) have become indispensable tools for exploring marine environments. However, most AUVs lack the ability to perform operations directly on the seabed. To address this limitation, we have designed a specialized AUV for seabed operations: the autonomous underwater helicopter (AUH). The AUH-HY-1 is a seabed operational vehicle specifically designed for complex near-seabed exploration tasks. It adopts a disc-shaped configuration and incorporates an Acoustic-Inertial-Optical integrated positioning and navigation system. The vehicle is equipped with multiple detection sensors and operates on a software framework developed using MOOS-IvP, which provides robust mission management and control. The AUH-HY-1 operates in two modes: the traditional AUV mode and the AUH mode, where it collaborates with a Subsea Docking Station (SDS) for long-term resident operations. A docking control strategy with orientation constraints ensures precise and reliable docking of the AUH. The integration of sonar with a YOLO-based recognition algorithm enhances the vehicle's target detection performance. Field trials conducted in controlled pool environments, lakes, and open sea settings have demonstrated the AUH-HY-1's operational reliability and task execution capabilities, including waypoint-based path planning, accurate docking, and seabed detection. It can perform near-seabed detection tasks at a height of approximately 3 to 5 meters above the seabed, with a maximum speed of around 3 knots. The vehicle achieved a notable docking success rate of over 85% and a sonar image detector MAP of 80.2% with an average processing speed of 0.025 s per image. These results indicate that the AUH-HY-1 is well suited for supporting seabed exploration and advancing applications in marine technology and oceanographic research.

A vision-based, walk-behind fertilizer applicator for cotton crops was designed to address the limitations of conventional systems. Traditional applicators face multiple operational issues such as high discharge rates, uneven fertilizer distribution, input wastage, and increased weed infestation. Field evaluations were carried out to measure parameters including missing plant index, uniformity of application, precision, fertilizer savings, and overall field performance. The applicator integrates three main systems: a cotton detection unit, an electronic controller, and an automatic fertilizing mechanism. The cotton detection unit captures and processes images to identify cotton plants while excluding weeds; the control system governs the stepper motor of the fertilization unit based on cotton detection signals; and the fertilizing mechanism dispenses micro-granular fertilizer near cotton plants by activating a metering unit through the stepper motor. The detection unit showed high accuracy with bounding box loss at 2.14% and object loss at 1.56% over 100 training epochs, confirming effective real-time detection. Small variations in discharge were observed, with Urea ranging between 5.59 and 8.9% and DAP between 6.89% and 9.25%. The left-to-right distribution ratios for Urea and DAP ranged from 0.89 to 1.05 and 0.90 to 1.04, respectively, with average values of 0.95 (Urea) and 0.97 (DAP). The system demonstrated high precision, delivering fertilizer with an accuracy between 90 and 93 percent, and an overall average of 91%. Theoretical and actual field capacities were in the range of 0.07–0.14 and 0.045–0.078 ha/h, respectively, while field efficiency (FE) varied from 69.39% at 0.5 km/h to 58.21% at 1 km/h. This study effectively developed a precision applicator that ensures targeted fertilization, adaptive delivery, reduced labor and time, better weed control, and promotes the use of modern technology in farming.

In the deep ocean, where light cannot penetrate and GPS coverage is unavailable, sonar remains the principal modality for underwater perception, navigation, and threat detection. In these acoustically complex environments, traditional sonar signal processing methods face critical limitations characterized by multipath propagation, Doppler shifts, ambient noise, and adversarial stealth. Reverberant littoral zones, low-observable platforms, and time-varying interference reduce the effectiveness of classical beamformers, matched filters, and deterministic classifiers. This paper presents a systematic review of recent advances in underwater sonar systems and artificial intelligence (AI)-driven signal processing for naval and autonomous applications. We trace the evolution from model-based frameworks to data-driven architectures, highlighting the growing role of convolutional and recurrent neural networks, deep Kalman filters, transformer-based classifiers, and multi-sensor fusion methods. These approaches are assessed in the context of GPS-denied navigation, constrained bandwidth, and dynamic acoustic conditions. Particular emphasis is placed on AI-driven motion estimation, where modern models increasingly surpass traditional methods in mitigating inertial drift, enhancing trajectory prediction, and improving operational resilience. This review synthesizes current capabilities and identifies unresolved challenges in model explainability, real-time adaptability, adversarial resilience, and energy-aware computation. Beyond summarizing recent developments, the paper offers a forward-looking perspective on intelligent sonar systems that seamlessly integrate sensing, inference, and decision-making positioning them as pivotal enablers in the future architecture of autonomous maritime operations.

To address the challenges of marine emergency rescue missions, this paper investigates the fixed-time consensus control problem of the multiple autonomous underwater vehicle systems (MAUVS) subjected to the system uncertainties, the nonlinear external disturbances, and the limited computational resources. First, a novel fixed-time radial basis function disturbance observer is developed to deal with the lumped disturbances composed of the dynamic uncertainties and the external ocean disturbances. Subsequently, an adaptive fixed-time event-triggered distributed control scheme based on command filter backstepping and a compensation mechanism is proposed to achieve the precise control and ensure the formation converge within a fixed time. Furthermore, a novel fixed-time compensation signal is introduced to address the potential errors from the filter process. Additionally, an event-triggered mechanism is employed to reduce the number of controller updates and computational consumption while maintaining control performance. Theoretical analysis clearly demonstrates that the Zeno behavior can be excluded, and the MAUVS can complete the formation task within a fixed time. Finally, the effectiveness of the consensus controller is validated through simulations. Furthermore, the applicability of the scheme is verified through the pool experiment.