Multi-arm harvesting robots offer a promising solution to the labor shortage in fruit harvesting, due to their ability to improve harvesting efficiency. However, multi-arm harvesters necessitate additional visual sensors to acquire distribution information of fruits within larger working spaces. Greater demands are consequently imposed on graphics computation, leading to increased costs in computing hardware of robot system. To balance the graphics computing cost and reduce energy consumption, distributed graphics computation frameworks for multi-arm robot vision system are proposed in this study. First, a host-edge framework is proposed to assign the tasks of image inference and depth alignment to host computer and edge computing modules through a decentralized mode of local connection. Moreover, to increase the endurance time of robot in application, the edge computing modules are reduced and the fifth generation mobile communication is integrated into robot graphics computing system to transfer on-board image processing to a remote computing server with MQTT protocol. To verify the effectiveness of the proposed framework, comprehensive experiments were performed, demonstrating that, compared with traditional computing framework, the proposed local distributed framework reduced 35.6% average time consumption, and over 20 FPS average processing speed can be achieve. The remote distributed framework has reduced the computational power consumption of the on-board system by approximately 23.1% while ensuring the performance is not lower than the local distributed framework. Finally, by discussing the two frameworks in terms of stability and cost, we present the commercial viability for the application of multi-arm harvesting robot.

In shallow water environments, traditional technologies face challenges in retrieving unexploded ordnance (UXO). We present an innovative and safe alternative robotic platform to replace traditional divers in this dangerous task. A waterproof eighteen-DOF legged robot MR. Crab has been developed specifically for the purpose of locating and recovering UXO. Utilizing walking legs for both locomotion and manipulation, the robot demonstrates the capability to pick up inert UXO weighing at 2.78kg (half robot’s mass) with one pair of walking legs and 10.5kg (184% of robot’s mass) when utilizing all three pairs of legs. To enhance our understanding of the gripping capacity, static analysis was conducted to predict the maximum weight of graspable UXO with respect to UXO radii. For easier control, a user-friendly GUI had been built for remote teleoperation. Additionally, an automatic gripping algorithm was developed to efficiently place the robot in the optimal gripping position and then grab the inert UXO.

The precise localization of mobile robots in unstructured environments is of utmost importance for many industrial and field applications, especially when the mobile robot is part of a more complex kinematic chain, such as a mobile manipulator. Being able to precisely localize affects the outcome of tasks that rely on an open-loop kinematic computation, such as work-station docking procedures. To achieve a repeatable and precise localization and positioning, mobile robots generally rely on onboard sensors, most commonly 2D laser scanners, whose readings are subjected to noise and numerous disturbing factors (e.g., materials reluctance). Problems arise when precise localization is needed in dynamic and unstructured environments where generally applicable methods won’t perform adequately or might be time-consuming to set up. In this work, we propose a cloud-edge computing architecture to deploy a recurrent neural network (RNN) based registration system, which uses a pair of consecutive LiDAR readings to estimate a fixed transformation. The capability of RNNs to process contiguous inputs will help neglect errors embedded in punctual laser scanner readings and output a more precise registration estimation. In such a way, the RNN can estimate a displacement error based on multiple consecutive readings and act as a sensor to be employed in a closed-loop control scheme. To tackle the dynamic and unstructured environments, the model is firstly tuned on synthetic LiDAR data to embed rigid transformations into the deep learning model, for then rapidly fine-tuned on local scenarios. After model architecture and optimization of hyperparameters, the devised model is tested in different scenarios, comparing the precise positioning capability of the AMR(autonomous mobile robot) with that of a classical registration algorithm. The results suggest that an RNN model can greatly improve the registration precision of laser scanner signals and, consequently, the precise positioning efficiency of AMRs.

This paper studies the coverage path planning and path following problems for an underwater biomimetic robotic fish to finish the deep-sea net cage water quality monitoring. Firstly, with a focus on minimizing total path length, repetition rate, and turning occurrences to enhance path coverage efficiency and reduce energy consumption, we propose a novel coverage path planning strategy. This strategy incorporates DQN, a reward function, and a rewrite strategy inspired by RRT*. Secondly, a high-performance path-following method, which takes into account robot performance, is designed to cope with adverse conditions in net cages. Finally, the simulation and field experiments demonstrate significant improvements over existing methods and the effectiveness in practical applications, showcasing its applicability in aquaculture management. The proposed algorithm offers valuable insights into optimizing coverage path planning for underwater robots in practical scenarios.

This article presents the first ever fully autonomous UAV (Unmanned Aerial Vehicle) mission to perform gas measurements after a real blast in an underground mine. The demonstration mission was deployed around 40 minutes after the blast took place, and as such realistic gas levels were measured. We also present multiple field robotics experiments in different mines detailing the development process. The presented novel autonomy stack, denoted as the Routine Inspection Autonomy (RIA) framework, combines a risk-aware 3D path planning D + ∗ , with 3D LiDAR-based global relocalization on a known map, and it is integrated on a custom hardware and a sensing stack with an onboard gas sensing device. In the presented framework, the autonomous UAV can be deployed in incredibly harsh conditions (dust, significant deformations of the map) shortly after blasting to perform inspections of lingering gases that present a significant safety risk to workers. We also present a change detection framework that can extract and visualize the areas that were changed in the blasting procedure, a critical parameter for planning the extraction of materials, and for updating existing mine maps. As will be demonstrated, the RIA stack can enable robust autonomy in harsh conditions, and provides reliable and safe navigation behavior for autonomous Routine Inspection missions.

Aiming at the problem of weak yawing controllability of solar-powered Unmanned Aerial Vehicles (UAVs) with low-speed and high aspect-ratio Blended Wing Body (BWB) configuration, based on the large-scale aerodynamic characteristics of the biomimetic swallow tail, the influence of different opening methods and angles of the swallow tail on the aerodynamic characteristics and stability and controllability of an UAV was derived. The mechanism of the higher yawing control efficiency of the differential throttle compared to the conventional rudder is analyzed in order to significantly decrease the control efficiency at small throttle. We innovatively propose a yawing control method combining the opening of the swallow tail to a moderately separation state of local airflow and we validate it for the case of the landing control of a solar-powered UAV. The flight dynamics mode characteristics show that this control method has little impact on the lateral-directional stability of the UAV and that the control efficiency can be reserved to the same level of cruise state. Therefore, the control law of cruise state can be applied to landing state with swallow tail open directly, which enhances the lateral-directional control ability of the UAVs in a simple but efficiently way. Flight simulation and flight test results show that the proposed bionic control method of the swallow tail combined with differential throttle can effectively enhance the landing control ability of the UAVs, improve the response speed, and reduce the trajectory error.

The use of automation in mining can be found at all stages of the mining process, covering exploration, excavation, loading, transportation, mineral processing, and search and rescue missions in emergency situations. Due to the harsh conditions during an underground mine disaster, robots can be of great assistance to rescue teams by entering areas that are unsafe for human rescuers, locating trapped workers and collecting valuable data. The design and implementation of mine rescue robots is characterized by great complexity since it encompasses the consideration of a great range of components and requirements. However, comprehensive guidelines for design and deployment of robots in harsh underground environments have not been established. As a result, the developers of mine rescue robots (MRRs) are to exercise their own judgment on the development and evaluation of the platform functionalities. This review attempts an extensive discussion of design and functionality requirements based on the common practices and lessons learned from the existing MRRs. Common considerations, as well as current and future trends are identified with the aim of facilitating future research and development on mine search and rescue robots.

Agricultural robots offer a viable solution to the critical challenges of productivity and sustainability of modern agriculture. The widespread deployment of agricultural robotic fleets, however, is still hindered by the overall system’s complexity, requiring the integration of several non-trivial components for the operation of each robot but also the orchestration of robots working with each other and human workers. This paper proposes a topological map as the unifying representation and computational model to facilitate the smooth deployment of robotic fleets in agriculture. This topological abstraction of the system state results in an efficient representation of large-scale environments, but also offers the scalable and efficient operation of the entire fleet and allows for ex-situ modelling and analysis of operations. The practical use of the proposed framework is demonstrated in a horticultural use case with a fleet of robots supporting the work of human fruit pickers. The critical components of the system are analysed and evaluated in deployment in both realistic digital twin and real-life soft fruit farms of different scales, demonstrating the scalability and effectiveness of the proposed framework. The presented framework is general and should be easy to adopt in other multi-robot/multi-human scenarios such as warehouse logistics, cleaning and maintenance of public spaces.

Modern mobile robots require precise and robust localization and navigation systems to achieve mission tasks correctly. In particular, in the underwater environment, where Global Navigation Satellite Systems (GNSSs) cannot be exploited, the development of localization and navigation strategies becomes more challenging. Maximum A Posteriori (MAP) strategies have been analyzed and tested to increase navigation accuracy and take into account the entire history of the system state. In particular, a sensor fusion algorithm relying on a MAP technique for Simultaneous Localization and Mapping (SLAM) has been developed to fuse information coming from a monocular camera and a Doppler Velocity Log (DVL) and to consider the landmark points in the navigation framework. The proposed approach can guarantee to simultaneously locate the vehicle, thanks to the onboard sensors, and map the surrounding environment with the information extracted from the images acquired by a bottom-looking optical camera. Optical sensors can provide constraints between the vehicle poses and the landmarks belonging to the observed scene. The DVL measurements have been employed to solve the unknown scale factor and to guarantee the correct vehicle localization even in absence of visual features. After validating the solution through realistic simulations, an experimental campaign at sea was conducted in Stromboli Island (Messina), Italy. In conclusion, an algorithm, which works with the Poisson surface reconstruction method to obtain a smooth seabed surface, for mesh creation has been developed.

Published version:  Journal of Field Robotics https://doi.org/10.1002/rob.22329

This paper presents the development and implementation of a multi-UAV system focused on coverage path planning on multiple separated areas capable of re-planning the collective mission in case of unexpected events. For this purpose, we present a distributed-centralized architecture that uses heuristic and computationally efficient methods to perform the planning/re-planning and decision-making tasks during the control of the mission execution. We performed a computational evaluation of the algorithms, comparing them with other proposals, together with experiments in simulated and real flights. The results show that the system can distribute tasks equitably among the aircraft in an efficient way, even in the middle of the flight, when facing unexpected events; and show a higher computational efficiency when compared to multiple proposals in the state of the art.

Point cloud registration is a vital task in 3D perception, with several different applications in robotics. Recent advancements have introduced neural-based techniques that promise enhanced accuracy and robustness. In this paper, we thoroughly evaluate well-known neural-based point cloud registration methods using the Point Clouds Registration Benchmark, which was developed to cover a large variety of use cases. Our evaluation focuses on the performance of these techniques when applied to real-complex data, which presents a more challenging and realistic scenario than the simpler experiments typically conducted by the original authors. With few exceptions, the results reveal a significant performance gap, with most neural-based methods performing poorly on real-complex data. However, amidst the underwhelming results, 3DSmoothNet stands out as an exception, demonstrating excellent performance across the majority of benchmark sequences. Remarkably, it outperforms a state-of-the-art feature extractor, namely FPFH, showcasing its effectiveness in capturing informative and discriminative features from point clouds. Given these results, we assert that while neural-based techniques could have advantages over conventional methods, further research is necessary to fully unlock their practical utility. We advocate for more extensive studies employing unbiased and realistic testing procedures, akin to the rigorous evaluation framework employed in this work. This paper represents a crucial step towards establishing a more robust literature in the field of point cloud registration. By exposing the limitations and advantages of existing neural-based methods, we aim to inspire future research and drive the development of more effective techniques for real-world applications.

Unmanned aerial vehicles (UAVs) offer many advantages over ground vehicles, including quadruped robots, based on high maneuverability when performing exploration in complex and unknown environments. However, due to their limited computational capability, UAVs require light-weight but accurate state estimation algorithms for reliable exploration. In this paper, we propose an segmented map based exploration system based on LiDAR-based state estimation for UAVs. The proposed system includes capabilities such as exploration, obstacle avoidance, and object detection with localization using 3D dense maps generated by tightly coupled LiDAR Inertial Odometry (LIO). Our proposed system is a hybrid system that can switch between guided and exploration modes, making it practical for search and rescue missions in disaster scenarios. The proposed LIO algorithm adapts to its surroundings, allowing for fast and accurate state estimation in complex environments. The proposed exploration algorithm is designed to cover specific regions in the 3D dense map generated by proposed LIO, with the UAV determining if map points are included within the coverage area. We tested the proposed system in both simulation and real-world environments and validated that proposed system outperforms state-of-the-art algorithms in various aspects such as localization accuracy and exploration efficiency in complex environments.

Multi-sensor fusion-based localization technology has achieved high accuracy in autonomous systems. How to improve the robustness is the main challenge at present. The most commonly used LiDAR and camera are weather-sensitive, while the FMCW Radar has strong adaptability but suffers from noise and ghost effects. In this paper, we propose a heterogeneous localization method called Radar on LiDAR Map (RoLM), which aims to enhance localization accuracy without relying on loop closures by mitigating the accumulated error in Radar odometry in real time. Our approach involves embedding the data from both Radar and LiDAR sensors into a density map. We calculate the spatial vector similarity with an offset to determine the corresponding place index within the candidate map and estimate the rotation and translation. To refine the alignment, we utilize the Iterative Closest Point (ICP) algorithm to achieve optimal matching on the LiDAR submap. We conducted extensive experiments on the Mulran Radar Dataset, Oxford Radar RobotCar Dataset, and our dataset to demonstrate the feasibility and effectiveness of our proposed approach.

This article presents a study on the world's first unmanned machine designed for autonomous forestry operations. In response to the challenges associated with traditional forestry operations, we developed a platform equipped with essential hardware components necessary for performing autonomous forwarding tasks. Through the use of computer vision, autonomous navigation, and manipulator control algorithms, the machine is able to pick up logs from the ground and manoeuvre through a range of forest terrains without the need for human intervention. Our initial results demonstrate the potential for safe and efficient autonomous extraction of logs in the cut-to-length harvesting process. We achieved a high level of accuracy in our computer vision system, and our autonomous navigation system proved to be highly efficient. This research represents a significant milestone in the field of autonomous outdoor robotics, with far-reaching implications for the future of forestry operations. By reducing the need for human labour, autonomous machines have the potential to increase productivity and reduce labour costs, while also minimizing the environmental impact of timber harvesting. The success of our study highlights the potential for further development and optimization of autonomous machines in the forestry industry.

This paper presents the design and development of a funnel-shaped Sparus Docking Station (SDS) intended for the non-holonomic torpedo-shaped Sparus II Autonomous Underwater Vehicles (AUV). The SDS is equipped with sensors and batteries, allowing for a stand-alone long-term deployment of the AUV. An inverted Ultra Short BaseLine (USBL) system is used to locate the Docking Station (DS) as well as to provide long-term drift-less AUV navigation. The SDS is able to observe the ocean currents using a Doppler Velocity Log (DVL), being motorized to allow its self-alignment with the current. Moreover, a docking algorithm accounting for the current is used to guide the robot during the docking maneuver. The paper reports experimental results of the docking maneuver in sea trials.

A new form of waypoint navigation controller for a skid-steer vehicle is presented, which consisted of a multiple input-single output nonlinear fuzzy angular velocity controller. The mem- bership functions of the fuzzy controller employed a trape- zoidal structure with a completely symmetric rule-base. No- tably, Hierarchical Rule-Base Reduction (HRBR) was incorpo- rated into the controller to select only the rules most influen- tial on state errors. This was done by selecting inputs/outputs and generating a hierarchy of inputs using a Fuzzy Relations Control Strategy (FRCS). Similar to some traditional fuzzy con- trollers, the system provided coverage for the global operat- ing environment. However, a rule for every possible combi- nation of variables and states was no longer necessary. Con- sequently, HRBR fuzzy controllers effectively increase both the number of inputs and their associated fidelity without the rule-base dramatically increasing. To contextualize the performance of the controller, a background on vehicle dy- namic modeling methodologies and an in-depth explanation of the related simulation model are provided. An examina- tion of the proposed controller is then completed employing test courses. The test courses examine the effects of steer- ing disturbance, phase lag, and overshoot as expressed in Root Mean Square Error (RMSE), Max Error (ME), and Course Completion Time (CCT). Finally, simulation and experimental results for the controller’s performance were compared with a state-of-the-art waypoint navigation vehicle controller, ge- ometric pure pursuit. The fuzzy was found to outperform the pure pursuit experimentally by 52.1 percent in RMSE, 26.8 percent in ME, and 1.07 percent in CCT, on average, validat- ing the viability of the controller.

A document by Maria Höffmann. Click on the document to view its contents.