Journal of Field Robotics最新文献

Unmanned surface vehicles (USVs) applied in urban waterways may suffer from inaccurate localization due to the Global Navigation Satellite System (GNSS) attenuation, and be susceptible to collision threats from vessels of human-induced violations and piloting errors. This paper proposes an enhanced navigation framework capable of stable continuous localization, dynamic obstacle perception, and collision-free motion planning. A tightly coupled LiDAR-Visual-Inertial Odometry via Smoothing and Mapping (LVI–SAM) is selected as the fundamental framework of localization and mapping subsystem. An incrementally mapping data structure is incorporated to improve the computation efficiency and accuracy of the LiDAR odometry optimization process. To mitigate the long-term accumulating odometry drift, valid GNSS measurements are introduced to provide absolute reference in the factor graph optimization framework, which can achieve optimum state estimation by maximum a posteriori given all the noisy measurements from multiple sensors. Furthermore, a dynamic occupancy grid map framework, based on sequential Monte Carlo and probability hypothesis density method, is developed to enhance situational awareness of USVs for risk anticipation of dynamic obstacles and facilitate predictive avoidance. Extensive real-world experiments have been carried out to demonstrate that the proposed autonomous navigation system is capable of robust and accurate localization over long-term urban waterway navigation, and dynamic obstacle avoidance through a safer predictive strategy.

Robotic technology in precision crop farming has the potential to minimize inputs, such as labor, fertilizer, or plant protection products, maximizing the net yield while reducing the environmental impact. To maximally exploit the benefits of precision crop farming, it has to be applied continuously over multiple years, which requires (robotic) technology for a wide range of agricultural operations. Researchers need access to (noncommercial) robot platforms with complete mechanical and software controllability to investigate new applications that could unlock the true potential of precision farming. This study presents the agricultural robot taskmap operation framework (ARTOF), which provides common functionality for robots with different vehicle configurations to execute task maps in crop farming applications based on global navigation satellite system positioning. The two-layered software stack has a mechatronic layer and an operational layer. The mechatronic layer performs motion control and includes machine safety to meet the required performance level in correspondence with European regulations. The operational layer performs autonomous implement and navigation control. Add-ons interact with the operational layer using the ARTOF Redis interface and increase flexibility. Hardware-in-the-loop testing enables static end-to-end testing and minimizes the developing time and operational faults when developing new functionality. To demonstrate the framework's flexibility, it was integrated into four in-house developed and modified agricultural robots with four-wheel drive, four-wheel steering (4WD4WS), skid steering, and Ackerman steering vehicle configurations. These robots performed 11 applications under real practice conditions in arable farming and horticulture for—in total—more than 11 km of field application. The power consumption, navigation accuracy, and software usability were evaluated. An average navigation accuracy of 1.0 cm was achieved during hoeing with a 4WD4WS robot using the newly developed navigation controller. This new open-source software framework enables the rapid validation of agricultural robotic research to broaden the number of precision crop farming applications and fully exploit their potential.

This study proposes a novel framework for large-scale scene reconstruction based on 3D Gaussian splatting (3DGS) and aims to address the rendering deficiency and scalability challenges faced by existing embodied AI tasks. For tackling the scalability issue, we split the large scene into multiple cells, and the candidate point-cloud and camera views of each cell are correlated through a visibility-based camera selection and a progressive point-cloud extension. To reinforce the rendering quality, three highlighted improvements are made in comparison with vanilla 3DGS, which are a strategy of the ray-Gaussian intersection and the novel Gaussians density control for learning efficiency, an appearance decoupling module based on ConvKAN network to solve uneven lighting conditions in large-scale scenes, and a refined final loss with the color loss, the depth distortion loss, and the normal consistency loss. Finally, the seamless stitching procedure is executed to merge the individual Gaussian radiance fields for novel view synthesis across different cells. Evaluation of Mill19, Urban3D, and MatrixCity datasets shows that our method consistently generates more high-fidelity rendering results than state-of-the-art methods of large-scale scene reconstruction. We further validate the generalizability of the proposed approach by applying it to self-collected video clips recorded by a commercial drone.

The ability to perceive environments supports an important foundation for our self-developed robotic rat to improve kinematic performance and application potential. However, the existing visual perception of quadruped robots suffers from poor perception accuracy in real-world dynamic environments. To mitigate the problem of erroneous data association, which is the main cause of low accuracy, the work presents an approach that combines leg odometry (LO) and inertial measurement unit (IMU) measurements with visual simultaneous localization and mapping to provide robust localization capabilities for small-scale quadruped robots in challenging scenarios by estimating the depth map and removing moving objects in dynamic environments. The method contains a depth estimation network with higher accuracy by combining the attention mechanism in the Transformer with the RAFT-Stereo depth estimation algorithm. Besides, the method combines target identification and segmentation with 3D projection of feature points to remove moving objects in dynamic environments. In addition, LO and IMU data are fused in the modified framework of ORB–SLAM3 to achieve highly accurate localization. The proposed approach is robust against erroneous data association due to moving objects and wobbles of quadruped robots. Evaluation results on multiple stages demonstrate that the system performs competitively in dynamic environments, outperforming existing visual perception methods in both public benchmarks and our custom small-scale robotic rat.

This study presents an innovative motion planning approach for underwater robotic arms, grounded in the multistrategy improved particle swarm optimization (PSO) (strategy adaptive particle swarm optimization [SAPSO]) algorithm. The SAPSO algorithm amalgamates the sine–cosine algorithm with the sparrow search algorithm, thereby enhancing the convergence efficiency and the capability to escape local optima inherent in PSO. Through the implementation of a 3–5–3 polynomial trajectory planning method, the proposed approach ensures a seamless transition from the initial to the target position while maintaining the continuity and fluidity of movement. Both simulation and underwater experimental analyses have validated the precision and efficacy of the SAPSO algorithm in collision detection, joint parameter optimization, and target capture operations. The outcomes underscore that the SAPSO algorithm considerably amplifies the speed and stability of trajectory planning and exhibits innovation and efficiency in the domain of underwater robotic arm motion planning.

This paper presents a nonlinear robust fast finite-time controller for three-dimensional (3D) path-following of underactuated Autonomous Underwater Vehicles considering parametric uncertainties, external disturbances, and the actuator's dynamics, fault, and saturation. A path-following error system is built using the virtual guidance method. The proposed cascaded closed-loop control scheme is composed of two layers: (1) first, a kinematic layer including an improved 3D approach-angle–based guidance law employs the Lyapunov theory and a fast finite-time backstepping control to transform the 3D path-following position errors into the command velocities; (2) then, a kinetic layer is designed to compute the actual control inputs using an integral fast terminal sliding mode control and the Lyapunov theory. A nonlinear fast finite-time disturbance observer enhances stability by estimating the lumped uncertainties in the sliding surface dynamics. A force/moment control loop equipped with a novel antiwindup system and a recursive least squares estimator compensates for the actuator's undesirable effects, including the internal dynamics, fault, and saturation. It is proved that the path-following errors uniformly globally converge to zero within a finite time. Comparative simulations illustrate that the proposed control provides better dynamic response and robustness compared with the asymptotic and conventional finite-time controllers.

Autonomous underwater vehicles (AUVs) face substantial challenges in obstacle avoidance due to the complex, dynamic nature of underwater environments and inherent sensing limitations. This study introduces a novel optimization framework that addresses these challenges by synergistically integrating advanced sampling strategies with reinforcement learning (RL) and model predictive path integral (MPPI) algorithms. The proposed framework strategically leverages the complementary strengths of both approaches: MPPI's proficiency in short-term trajectory prediction combined with RL's exploratory capabilities and end-to-end training paradigm. This integration enables AUVs to rapidly adapt to environmental perturbations, make efficient real-time obstacle avoidance decisions, continuously adjust to increasingly complex underwater scenarios, and achieve long-term safe navigation objectives. To evaluate the efficacy of this RL-MPPI hybrid approach, comprehensive numerical simulations were conducted across diverse underwater environmental conditions, encompassing both static and dynamic obstacles. The simulation results demonstrate enhanced adaptability and responsiveness in complex underwater environments, improved predictive accuracy and stability in obstacle avoidance maneuvers, and effective navigation through static and dynamic underwater scenarios while maintaining robust predictive characteristics. Quantitatively, the proposed method reduces the average cost value by 9.3% and average execution time by 2.9% compared with traditional MPPI in water-free environments. Furthermore, in the presence of unknown water flow, it achieves a 7.2% reduction in average cost value and a 1.6% decrease in average execution time. This study contributes to the advancement of underwater robotics by offering a robust, adaptive, and computationally efficient approach to collision prevention for AUVs. The proposed framework demonstrates considerable promise for enhancing AUV capabilities in safe and efficient navigation through increasingly challenging underwater environments.

In this paper, an adaptive tracking controller for the propeller-driven wall-climbing robot is developed, which is subject to velocity-related input saturation and velocity constraint. First, the model of the propeller-driven wall-climbing robot is established, where actuator dynamics and input saturation are considered with velocity constraints. The strategy of active gravity balance is put forward, which simplifies the modeling but leads to the problem of velocity-related input saturation. Second, the Gauss integration function is used to approximate the velocity-related input saturation. The velocity constraint would be handled by employing the barrier Lyapunov-based transformation rather than the barrier Lyapunov function (BLF) method. Thirdly, the tracking controller is developed based on the dynamic surface control method, where the adaptive robust controller and neural networks are combined to deal with unmodeled dynamics and external disturbances. According to the Lyapunov stability theory, it is proved that the propeller-driven robot system will be stable under the developed controller, while signals in the closed-loop system are ultimately uniformly bounded. Finally, simulation results show the effectiveness of the proposed tracking control scheme.