This paper proposes an event-triggered nonlinear visual predictive control strategy for image-based visual servoing of robots. It involves developing a nonlinear model of the visual servoing system and designing a predictive control strategy that addresses safety, real-time performance, robustness, and smooth motion control. Field-of-view constraints ensure image feature visibility, physical constraints respect joint limits, and smooth motion constraints protect hardware from excessive stress. The event-triggered mechanism activates control laws only when necessary, reducing the computational burden of continuous control adjustments and enhancing responsiveness and efficiency. This strategy supports robustness, mitigates issues arising from local minima, and maintains system stability, providing a practical solution for real-time visual servoing tasks. Furthermore, we compare the performance of the proposed strategy against conventional and modified model predictive control strategies in various visual servoing tasks through simulations. Finally, the experiment results demonstrate the effectiveness of the strategy.

This paper addresses the robust trajectory tracking control challenge for multiple mobile robots in complex environments, an increasingly critical issue as the number of robots grows and the demand for high tracking accuracy and efficiency increases. Existing methods are unable to strike a balance between safety and tracking precision in multi-robot trajectory tracking, with the requirement that robots should be as close as possible to their designated positions at all times during tracking. To bridge these gaps, we introduce Multi Mobile Robot Trajectory Model Predictive Control (MMRT-MPC) and the Trajectory Action Dependence Graph (TADG) framework. MMRT-MPC incorporates multiple indicators into the cost function to improve trajectory tracking accuracy and efficiency. Meanwhile, TADG ensures safety during trajectory tracking and is compatible with MMRT-MPC as well as other control algorithms. Simulations in Gazebo show that the TADG method ensures the safety of trajectory tracking control. Compared with applying TADG to Prioritized Trajectory Optimization (PTO) and Bellman Dynamic Programming with Model Predictive Control (BDP-MPC), MMRT-MPC+TADG reduces average delay by 17.7% and 11.6% respectively under different numbers of robots, and by 20.8% and 14.3% in the case of 30 robots with random delays added. Furthermore, the validity of our proposed method is confirmed through real-world experimental results.

Overall, a big challenge in building a sailboat USV relies on the development of an autonomous system for guidance, navigation, and control (GNC) because both sail and rudder angle must be cooperatively adjusted to correct the navigation direction — traditional propelled boats can be more easily controlled with a straightforward control task to set the rudder angle. Moreover, sailing upwind requires special maneuvers to reach a given target in that unfeasible direction. Reinforcement learning emerges as a promising technique for building autonomous GNCs for sailing robots, but training the neural network with a real sailboat is impractical due to long periods of training and safety reasons. Even traditional control-based approaches are mainly tested in simulated environments due to the difficulties in building and operating a real sailboat. The issue that arises is the fidelity of these simulated environments. In this context, we propose Yara, an oceanic virtual environment with a reliable physics simulation for developing, training, and evaluating autonomous agents to operate digital twins of sailing robots in reinforcement learning and other paradigms. An autonomous sailing robot digital twin is available within the virtual environment, with the foil dynamics constructed based on a real sailing robot. We coupled these foil dynamics in Gazebo’s physics engine to compute the lift and drag forces acting on the sail, rudder, and keel. The simulated world feeds sensors such as cameras, wind sensors, and GPS. The Robot Operating System communicates these sensors’ data through topics, facilitating users’ implementation and testing of new GNC solutions. Yara provides a reliable solution for foil dynamic simulated physics that achieves a simulation speedup of 300 times on an i7 laptop with 8 GB of RAM, powered by a Nvidia RTX 3060 and running Ubuntu 20.04. With this speedup, it is possible to complete a million time steps of deep reinforcement learning training in approximately eight hours. Evaluation scenarios were presented to highlight specific features of the simulator, like the maneuverability of the sailing robot digital twin and applications to train, evaluate, and compare reinforcement learning agents and other control solutions.

Modular robots are versatile systems whose compositions can be adapted and optimized for a wide range of applications and environments. However, the practicality of their reconfigurations is often limited by the necessity of manually deriving models and adapting control software for each new configuration. This manual process can be time-consuming and complex, making it less feasible to fully utilize the potential of modular robotics. Existing approaches for automatic derivation of models and effortless deployment of model-based controllers consider robotic manipulators only. In contrast, we consider modular wheeled mobile robots and present an approach for automatically recognizing their kinematics and deploying their navigation capabilities given mobility and perception modules. Our approach has been tested through both simulations and experiments. The simulations validate the approach across different platforms, each with varying compositions of modules. The experimental results further show the effectiveness of our approach, by deploying the autonomous navigation of a platform composed of four steering wheels, and by focusing on car-like, differential, and omni-like kinematic configurations.

Multi-Robot Systems (MRS) have shown great potential in object manipulation and transportation tasks by enabling the coordinated handling of complex and large objects. This paper presents a novel hybrid caging-based approach for cooperative object transport using only two omnidirectional mobile robots equipped with soft grippers inspired by the Fin-Ray effect. The proposed system integrates a Leader-Follower feedback controller that regulates distance and angle between robots to maintain the object closure condition throughout the transport. The soft grippers enhance adaptability to object contours and allow stable manipulation with fewer robots. Extensive real-world experiments were conducted with objects of various shapes, demonstrating the robustness and effectiveness of the proposed strategy. Results show accurate trajectory tracking and object retention under both static and dynamic orientation scenarios, highlighting the potential of soft robotic integration in cooperative manipulation tasks.

Bayesian Meta-Learning’s role in the autonomous navigation of mobile robots remains unexplored; this systematic review highlights the effectiveness of Bayesian Meta-Learning in enhancing probabilistic mapping and navigation in mobile autonomous robots. The paper initially compares Bayesian Meta-Learning with existing techniques that have been the cornerstone of probabilistic mapping and navigation, such as Kalman filters and particle filters. It then meticulously examines the critical metrics, including path planning, obstacle avoidance, active localization, navigation in dynamic environments, and mapping accuracy, highlighting the substantial impact of Bayesian Meta-Learning in enhancing these critical aspects of autonomous navigation in mobile robots. The approach enhances adaptability and learning in mobile robots, highlighting its potential to transform autonomous navigation. In addition to emphasizing the positive outcomes of the research, the review also acknowledges a substantial research gap and aims to provide novel insights for future exploration. Future research directions include lifelong learning, uncertainty-aware exploration, integration of prior knowledge, and improvements in human-robot interaction to enhance existing paradigms and advance robust autonomous robotic systems. Therefore, this study accentuates the affirmative findings regarding Bayesian Meta-Learning and recognizes and contributes to the broader research landscape within the field.

The article analyses the current aspects of the aviation sanctions imposed on Russia by the European Union, their objectives and effects in the context of international aviation cooperation. It also discusses Norwegian legal regulations on the use of drones, resulting from EU regulations implemented by Norway as a member of the European Economic Area. In addition, the analysis covers, among others, safety regulations and operational requirements for drones used by tourists. In addition, the author presents examples of incidents violating safety and privacy standards, pointing out the challenges posed by drones in urban areas and in airspace. The article also analyses an important ruling of the Court of Justice of the European Union, which shaped the interpretation of regulations on sanctions and the use of drones in member states, emphasizing the role of the CJEU in unifying the application of EU law. The research conducted indicates the need to adapt international regulations to the growing role of drones and the challenges related to their responsible use, as well as to further develop sanctions law as a tool of the European Union's foreign policy.

Dual mobile manipulators (DMMs) possess the advantages of high redundancy and extensive working space, enabling robot systems to undertake more complex tasks, such as the collaborative transportation of objects with stereoscopic shapes. Some existing works have addressed the collaborative transportation tasks under the tightly-coupled transportation model. Nevertheless, there have been few studies on the transportation issues in the loosely-coupled transportation model, where the problems such as the difficulty of maintaining formation configuration and high requirements for internal force stability exist. To this end, this paper presents a new event-based planning and control approach. First, the motion forms during the transportation process are classified for the collaborative transportation scenarios of objects with stereoscopic shapes, and an event-based motion planning strategy is provided. Then, by combining admittance control and nonlinear feedback control, a force-position hybrid control method is designed to control the force and position errors during the transportation process. Experimental results indicate that compared to the time-based method, the proposed event-based method can better guarantee the collaborative performance of the DMMs system during the transportation process. To the best of our knowledge, this may be the first attempt to solve the collaborative transportation problem of DMMs with loosely-coupled model by the event-based method.

The increasing demand for unmanned aerial vehicles (UAVs) in the aerospace industry highlights the need for precise simulation environments, especially in remote regions. This study develops an open-source framework for a customized UAV simulation environment using ROS-Gazebo and Ardupilot. The method includes a realistic reconstruction of the environment based on free satellite information, the ROS-Gazebo scheme for modeling and testing the UAV platform, and Ardupilot for mission design and deployment. The methodology aims to reduce costs in surveillance missions and conduct more efficient operations in high-risk and challenging areas. To evaluate the accuracy of UAV positioning for simulating real missions, an experimental validation of the digital twin was carried out using real flight data records from two self-built UAVs equipped with the open-source Ardupilot autopilot. One of the aircraft, a fixed-wing UAV, was tested near the Antisana volcano (4500 meters above sea level), where wind gusts reached speeds of 9 to 12 m/s. These tests revealed maximum errors of 9% in the Z-axis trajectory (altitude). The other aircraft, a quadcopter, was evaluated in the Parque Carolina in Quito (2800 meters above sea level), with wind gusts between 5 and 7 m/s, showing an error of 8% in the Z-axis trajectory. The presented results demonstrate the suitability of the proposed method for emulating complex missions using digital twin models. In this regard, the main contribution of this work lies in its potential for the precise prediction of flight missions in the Z-axis, a crucial variable for avoiding collisions in mountainous regions.