Journal of Intelligent & Robotic Systems最新文献

Hull inspection is an important task to ensure sustainability of ships. To overcome the challenges of hull structure inspection in an underwater environment in an efficient way, an autonomous system for hull inspection has to be developed. In this paper, a new approach to underwater ship hull inspection is proposed. It aims at developing the basis for an end-to-end autonomous solution. The real-time aspect is an important part of this work, as it allows the operators and inspectors to receive feedback about the inspection as it happens. A reference mission plan is generated and adapted online based on the inspection findings. This is done through the processing of a multibeam forward looking sonar to estimate the pose of the hull relative to the drone. An inspection map is incrementally built in a novel way, incorporating uncertainty estimates to better represent the inspection state, quality, and observation confidence. The proposed methods are experimentally tested in real-time on real ships and demonstrate the applicability to quickly understand what has been done during the inspection.

Insect-inspired sensor fusion algorithms have presented a promising avenue in the development of robust and efficient systems, owing to the insects' ability to process numerous streams of noisy sensory data. The ring attractor neural network architecture has been identified as a noteworthy model for the optimal integration of diverse insect sensors. Expanding on this, our research presents an innovative bio-inspired ring attractor neural network architecture designed to augment the performance of microsatellite attitude determination systems through the fusion of data from multiple gyroscopic sensors.Extensive simulations using a nonlinear model of the microsatellite, while incorporating specific navigational disturbances, have been conducted to ascertain the viability and effectiveness of this approach. The results obtained have been superior to those of alternative methodologies, thus highlighting the potential of our proposed bio-inspired fusion technique. The findings indicate that this approach could significantly improve the accuracy and robustness of microsatellite systems across a wide range of applications.

Vision-based deep learning perception fulfills a paramount role in robotics, facilitating solutions to many challenging scenarios, such as acrobatic maneuvers of autonomous unmanned aerial vehicles (UAVs) and robot-assisted high-precision surgery. Control-oriented end-to-end perception approaches, which directly output control variables for the robot, commonly take advantage of the robot’s state estimation as an auxiliary input. When intermediate outputs are estimated and fed to a lower-level controller, i.e., mediated approaches, the robot’s state is commonly used as an input only for egocentric tasks, which estimate physical properties of the robot itself. In this work, we propose to apply a similar approach for the first time – to the best of our knowledge – to non-egocentric mediated tasks, where the estimated outputs refer to an external subject. We prove how our general methodology improves the regression performance of deep convolutional neural networks (CNNs) on a broad class of non-egocentric 3D pose estimation problems, with minimal computational cost. By analyzing three highly-different use cases, spanning from grasping with a robotic arm to following a human subject with a pocket-sized UAV, our results consistently improve the R(^{2}) regression metric, up to +0.51, compared to their stateless baselines. Finally, we validate the in-field performance of a closed-loop autonomous cm-scale UAV on the human pose estimation task. Our results show a significant reduction, i.e., 24% on average, on the mean absolute error of our stateful CNN, compared to a State-of-the-Art stateless counterpart.

Robotics has been booming in recent years. Especially with the development of artificial intelligence, more and more researchers have devoted themselves to the field of robotics, but there are still many shortcomings in the multi-task operation of robots. Reinforcement learning has achieved good performance in manipulator manipulation, especially in grasping, but grasping is only the first step for the robot to perform actions, and it often ignores the stacking, assembly, placement, and other tasks to be carried out later. Such long-horizon tasks still face the problems of expensive time, dead-end exploration, and process reversal. Hierarchical reinforcement learning has some advantages in solving the above problems, but not all tasks can be learned hierarchically. This paper mainly solves the complex manipulation task of continuous multi-action of the manipulator by improving the method of hierarchical reinforcement learning, aiming to solve the task of long sequences such as stacking and alignment by proposing a framework. Our framework completes simulation experiments on various tasks and improves the success rate from 78.3% to 94.8% when cleaning cluttered toys. In the stacking toy experiment, the training speed is nearly three times faster than the baseline method. And our method can be generalized to other long-horizon tasks. Experiments show that the more complex the task, the greater the advantage of our framework.

Visual odometry (VO) is an important problem studied in robotics and computer vision in which the relative camera motion is computed through visual information. In this work, we propose to reduce the error accumulation of a dual stereo VO system (4 cameras) computing 6 degrees of freedom poses by fusing two independent stereo odometry with a nonlinear optimization. Our approach computes two stereo odometries employing the LIBVISO2 algorithm and later merge them by using image correspondences between the stereo pairs and minimizing the reprojection error with graph-based bundle adjustment. Experiments carried out on the KITTI odometry datasets show that our method computes more accurate estimates (measured as the Relative Positioning Error) in comparison to the traditional stereo odometry (stereo bundle adjustment). In addition, the proposed method has a similar or better odometry accuracy compared to ORB-SLAM2 and UCOSLAM algorithms.

Navigation and planning for unmanned aerial vehicles (UAVs) based on visual-inertial sensors has been a popular research area in recent years. However, most visual sensors are prone to high error rates when exposed to disturbances such as excessive brightness and blur, which can lead to catastrophic performance drops in perception and motion planning systems. This study proposes a novel framework to address the coupled perception-planning problem in high-risk environments. This achieved by developing algorithms that can automatically adjust the agility of the UAV maneuvers based on the predicted error rate of the pose estimation system. The fundamental idea behind our work is to demonstrate that highly agile maneuvers become infeasible to execute when visual measurements are noisy. Thus, agility should be traded-off with safety to enable efficient risk management. Our study focuses on navigating a quadcopter through a sequence of gates on an unknown map, and we rely on existing deep learning methods for visual gate-pose estimation. In addition, we develop an architecture for estimating the pose error under high disturbance visual inputs. We use the estimated pose errors to train a reinforcement learning agent to tune the parameters of the motion planning algorithm to safely navigate the environment while minimizing the track completion time. Simulation results demonstrate that our proposed approach yields significantly fewer crashes and higher track completion rates compared to approaches that do not utilize reinforcement learning.

Condition monitoring of power transmission lines is an essential aspect of improving transmission efficiency and ensuring an uninterrupted power supply. Wherein, efficient inspection methods play a critical role for carrying out regular inspections with less effort & cost, minimum labour engagement and ease of execution in any geographical & environmental conditions. Earlier various methods such as manual inspection, roll-on wire robotic inspection and helicopter-based inspection are preferably utilized. In the present days, Unmanned Aerial System (UAS) based inspection techniques are gradually increasing its suitability in terms of working speed, flexibility to program for difficult circumstances, accuracy in data collection and cost minimization. This paper reports a state-of-the-art study on the inspection of power transmission line systems and various methods utilized therein, along with their merits and demerits, which are explained and compared. Furthermore, a review was also carried out for the existing visual inspection systems utilized for power line inspection. In addition to that, blockchain utilities for power transmission line inspection are discussed, which illustrates next-generation data management possibilities, automating an effective inspection and providing solutions for the current challenges. Overall, the review demonstrates a concept for synergic integration of deep learning, navigation control concepts and the utilization of advanced sensors so that UAVs with advanced computation techniques can be analyzed with different aspects of implementation.

This paper proposes the PANORAMA approach, which is designed to dynamically and autonomously manage the allocation of a robot’s hardware and software resources during fully autonomous mission. This behavioral autonomy approach guarantees the satisfaction of the mission performance constraints. This article clarifies the concept of performance for autonomous robotic missions and details the different phases of the PANORAMA approach. Finally, it focuses on an experimental implementation on a patrolling mission example.

Aerial 3D printing is a pioneering technology yet in its conceptual stage that combines frontiers of 3D printing and Unmanned aerial vehicles (UAVs) aiming to construct large-scale structures in remote and hard-to-reach locations autonomously. The envisioned technology will enable a paradigm shift in the construction and manufacturing industries by utilizing UAVs as precision flying construction workers. However, the limited payload-carrying capacity of the UAVs, along with the intricate dexterity required for manipulation and planning, imposes a formidable barrier to overcome. Aiming to surpass these issues, a novel aerial decomposition-based and scheduling 3D printing framework is presented in this article, which considers a near-optimal decomposition of the original 3D shape of the model into smaller, more manageable sub-parts called chunks. This is achieved by searching for planar cuts based on a heuristic function incorporating necessary constraints associated with the interconnectivity between subparts, while avoiding any possibility of collision between the UAV’s extruder and generated chunks. Additionally, an autonomous task allocation framework is presented, which determines a priority-based sequence to assign each printable chunk to a UAV for manufacturing. The efficacy of the proposed framework is demonstrated using the physics-based Gazebo simulation engine, where various primitive CAD-based aerial 3D constructions are established, accounting for the nonlinear UAVs dynamics, associated motion planning and reactive navigation through Model predictive control.

High-reliability landing systems for unmanned aerial vehicles (UAVs) have gained extensive attention for their applicability in complex wild environments. Accurate locating, flexible tracking, and reliable recovery are the main challenges in drone landing. In this paper, a novel UAV autonomous landing system and its control framework are proposed and implemented. It’s comprised of an environmental perception system, an unmanned ground vehicle (UGV), and a Stewart platform to locate, track, and recover the drone autonomously. Firstly, a recognition algorithm based on multi-sensor fusion is developed to locate the target in real time with the help of a one-dimensional turntable. Secondly, a dual-stage tracking strategy composed of a UGV and a landing platform is proposed for dynamically tracking the landing drone. In a wide range, the UGV is in charge of fast-tracking through the artificial potential field (APF) path planning and the model predictive control (MPC) tracking algorithms. While the trapezoidal speed planning is employed in platform controller to compensate for the tracking error of the UGV, realizing the precise tracking to the drone in a small range. Furthermore, a recovery algorithm including an attitude compensation controller and an impedance controller is designed for the Stewart platform, ensuring horizontal and compliant landing of the drone. Finally, extensive simulations and experiments are dedicated to verifying the feasibility and reliability of the developed system and framework, indicating that it is a superior case of UAV autonomous landing in wild environments such as grasslands, slopes, and snow.