Tiago Pereira do Nascimento, Ph.D.

In this work, we propose an improved artificially weighted spanning tree coverage (IAW- STC) algorithm for distributed coverage path planning of multiple flying robots. The proposed approach is suitable for environment exploration in cluttered regions, where unexpected obstacles can appear. In addition, we present an online re-planner smoothing algorithm with unexpected detected obstacles. To validate our approach, we performed simulations and real robot experiments. The results showed that our proposed approach produces sub-regions with less redundancy than its previous version.

This letter proposes an approach for estimating the loss of effectiveness of UAV actuators while also considering the presence of aerodynamic effects, most specifically those regarding blade-flapping. The fault detection scheme is based on a Reduced Order Extended Kalman Filter Tolerant to Aerodynamic Effects (ROEKF-TAE) capable of detecting loss-of-effectiveness (LOE) faults in more than one actuator at the same time. The experimental results were performed using a Parrot Mambo micro drone and compared to those of a traditional EKF, as well as to those of a state-of-the-art adaptive Kalman Filter available in the literature. The filter proposed in this work could better distinguish between the effects of the actual fault and those of the blade-flapping disturbance when compared to the other filters. Also, it was able to identify with more precision which actuators were truly defective and which were not.

In conjunction with the growth of automated warehouses, a logistical problem also increases. The automated inventory counting problem emerges from the difficulty of managing the products of these large distribution centers. Usually, these centers have long corridors and high shelves with many different products. To solve this problem, this work proposes an approach of a highly-scalable low-cost plug-and-play multirobot system for inventory management. Our approach is composed of a set that includes a micro-drone, an embedded camera module, and a ground mobile robot. In our tests, two situations are analyzed: first with one heterogeneous multirobot system set and a second situation with two heterogeneous multi-robot system sets. The results demonstrate the advantage of having the interconnected multi-robot system to reduce the time of the inventory management task.

Landing an unmanned aerial vehicle unmanned aerial vehicle (UAV) on top of an unmanned surface vehicle (USV) in harsh open waters is a challenging problem, owing to forces that can damage the UAV due to a severe roll and/or pitch angle of the USV during touchdown. To tackle this, we propose a novel model predictive control (MPC) approach enabling a UAV to land autonomously on a USV in these harsh conditions. The MPC employs a novel objective function and an online decomposition of the oscillatory motion of the vessel to predict, attempt, and accomplish the landing during near-zero tilt of the landing platform. The nonlinear prediction of the motion of the vessel is performed using visual data from an onboard camera. Therefore, the system does not require any communication with the USV or a control station. The proposed method was analyzed in numerous robotics simulations in harsh and extreme conditions and further validated in various real-world scenarios.

This paper presents a modular autonomous Unmanned Aerial Vehicle (UAV) platform called the Multi-robot System (MRS) Drone that can be used in a large range of indoor and outdoor applications. The MRS Drone features unique modularity changes in actuators, frames, and sensory configuration. As the name suggests, the platform is specially tailored for deployment within a MRS group. The MRS Drone contributes to the state-of-the-art of UAV platforms by allowing smooth real-world deployment of multiple aerial robots, as well as by outperforming other platforms with its modularity. For real-world multi-robot deployment in various applications, the platform is easy to both assemble and modify. Moreover, it is accompanied by a realistic simulator to enable safe pre-flight testing and a smooth transition to complex real-world experiments. In this manuscript, we present mechanical and electrical designs, software architecture, and technical specifications to build a fully autonomous multi UAV system. Finally, we demonstrate the full capabilities and the unique modularity of the MRS Drone in various real-world applications that required a diverse range of platform configurations.

The intuitive control of robot swarms becomes crucial when humans are working in close proximity with the swarm in unknown environments. In such operations, it is necessary to maintain the autonomy of the swarm while giving the human operator enough means to influence the decision-making process of the robots. This paper presents a human-swarm interaction approach using full-body action recognition to control an autonomous flock of unmanned aerial vehicles. We estimate the full-body pose of the human operator and use a k-nearest neighbor algorithm to classify the action made by the humans. Finally, the swarm uses the identified action to decide its goal direction. We demonstrate the practicality of our approach with a multi-stage experimental setup to evaluate the prediction accuracy and robustness of the system.

This paper presents a family of autonomous Unmanned Aerial Vehicles (UAVs) platforms designed for a diverse range of indoor and outdoor applications. The proposed UAV design is highly modular in terms of used actuators, sensor configurations, and even UAV frames. This allows to achieve, with minimal effort, a proper experimental setup for single, as well as, multi-robot scenarios. Presented platforms are intended to facilitate the transition from simulations, and simplified laboratory experiments, into the deployment of aerial robots into uncertain and hard-to-model real-world conditions. We present mechanical designs, electric configurations, and dynamic models of the UAVs, followed by numerous recommendations and technical details required for building such a fully autonomous UAV system for experimental verification of scientific achievements. To show strength and high variability of the proposed system, we present results of tens of completely different real-robot experiments in various environments using distinct actuator and sensory configurations.

This work presents an approach for dealing with suspended-cable load transportation using unmanned aerial vehicles (UAVs), specifically when the cargo overcomes the lifting capacity. Herein, this approach is referred to as the Side-Pull Maneuver (SPM). This maneuver is an alternative and viable strategy for cases where there is no impediment or restriction to dragging the load along a surface, such as with pastures or marine environments. The proposal is based on a joint observation of the thrust and altitude of the UAV. To make this possible, the high-level rigid-body dynamics model is described and represented as an underactuated system. Its altitude-rate control input is then analyzed during flight. A flight state supervisor decides whether the cargo should be carried by lifting or by side-pulling, or whether it should be labeled as nontransportable. Comparative real experiments validate the proposal according to which maneuver (lifting or dragging) is performed for transport.

This paper presents a multi-layer software architecture to perform cooperative missions with a fleet of quad-rotors providing support in electrical power line inspection operations. The proposed software framework guarantees the compliance with safety requirements between drones and human workers while ensuring that the mission is carried out successfully. Besides, cognitive capabilities are integrated in the multi-vehicle system in order to reply to unforeseen events and external disturbances. The feasibility and effectiveness of the proposed architecture are demonstrated by means of realistic simulations.

The problem of safe interception of multiple intruder UAVs by a team of cooperating autonomous aerial vehicles is addressed in this paper. The presented work is motivated by the Mohamed Bin Zayed International Robotics Challenge (MBZIRC) 2020 where this task was simplified to an interaction with a set of static and dynamic objects (balloons and a UAV), and by a real autonomous aerial interception system of Eagle.One that our team has been working on. We propose a general control, perception, and coordination system for the fast and reliable interception of targets in a 3D environment relying only on onboard sensors and processing. The proposed methods and the entire complex multi-robot sys- tem were successfully verified in demanding desert conditions, with the main focus on reliability and fast deployment. In the MBZIRC competition, the proposed approach exhibited the greatest reliability and fastest solution. It was crucial to our team in winning the entire competition and achieving the second place in the intruder UAV interception scenario.

An approach that enables autonomous Unmanned Aerial Vehicles (UAV) with onboard sensor-based force control to interact with the indoor walls of historical buildings is proposed in this paper. The motivation for enabling UAVs to be pressed against walls is twofold: 1) it enables providing strong-side lighting on places where a light source needs to be remotely pressed against the wall for documentation by another drone with a camera and 2) it is a technique for enabling remote placement of infrastructure in difficult-to-access indoor locations, e.g., smart sensors for continuous monitoring of temperature and humidity. We propose therefore an admittance force-based control system that enables a UAV to interact with a wall in a stabilized manner at a pre-defined location. The UAV is coupled with a mechanism that can measure the interacting force, allowing the proposed controller to be in constant contact with the wall based on a measured force, and to regulate the force to the amount required by a given application. The proposed approach has been verified through numerous simulations in Gazebo and experiments with real robots in GNSS-denied environments relying solely on onboard sensors.

Very small size or micro, aerial vehicles are being recently studied due to the large influence of environmental disturbances. The multirotor aerial vehicle (MAV) usually requires control approaches that can guarantee a safe operation. However, limitations with respect to the embedded system (i.e. energy, processing power, memory, etc.) are usually present. In this work, we propose the use of Nonlinear model predictive control (NMPC), which can safely respect input constraints. In contrast, the application of NMPC in embedded systems of Micro-MAV is typically challenging. To solve this issue, we propose a modification on the NMPC called Embedded Fast NMPC that can ensure the implementation of the position controller safely and stably. Micro Multirotor Aerial Vehicles (Micro-MAVs) use low processing power boards. These boards usually rely solely on on-board sensors to perform localization and target detection, which in turn makes this platform suitable for experiments in GNSS-denied environments. We validate our approach with real robot experiments using a Micro-MAV.

In Mobile Robotics, visual tracking is an extremely important sub-problem. Some solutions found to reduce the problems arising from partial and total occlusion are the use of multiple robots. In this work, we propose a three-dimensional space target tracking based on a constrained multi-robot visual data fusion on the occurrence of partial and total occlusion. To validate our approach we first implemented a non-cooperative visual tracking where only the data from a single robot is used. Then, a cooperative visual tracking was tested, where the data from a team of robots is fused using a particle filter. To evaluate both approaches, a visual tracking environment with partial and total occlusions was created where the tracking was performed by a team of robots. The result of the experiment shows that the non-cooperative approach presented a lower computational cost than the cooperative approach but the inferred trajectory was impaired by the occlusions, a fact that did not occur in the cooperative approach due to the data fusion.

The use of robotic systems, especially multi-rotor aerial vehicles, in the documentation of historical buildings and cultural heritage monuments has become common in recent years. However, the teleoperated robotic systems have significant limitations encouraging the ongoing development of autonomous unmanned aerial vehicles (UAVs). The autonomous robotic platforms provide a more accurate and safe measurement in distant and difficult to access areas than their teleoperated counterpart. Through the use of autonomous aerial robotic systems, access to such places by humans and building of external infrastructures like scaffolding for documentation purposes is no longer necessary. In this work, we aim to present a novel autonomous unmanned aerial vehicle designed for the documentation of hardly attainable areas of historical buildings. The prototype of this robot was tested in several historical monuments comprising scanned objects located in dark and hardly accessible areas in the upper parts of tall naves. This manuscript presents the results from two specific places: the Church of St. Anne and St. Jacob the Great in Stará Voda, and St. Maurice Church in Olomouc, both in the Czech Republic. Finally, we also compare the three-dimensional map obtained with the measurements made by the 3D laser scanner carried onboard UAV against the ones performed by a 3D terrestrial laser scanner.

—A decentralized algorithm for flocking of Unmanned Aerial Vehicles (UAV) in environments with high obstacle density is proposed in this work. The method combines a local planning loop with bio-inspired swarming rules for navigating a compact UAV flock in a real workspace without relying on external infrastructures, such as motion capture system and GNSS. The group stability and coherence are achieved by employing a purposely designed onboard UVDAR system for mutual localization of teammates in local proximity of each UAV. The required robustness and scalability of the multi-UAV system are therefore achieved without any need for communication among the swarm particle. Such minimal sensory and communication requirements have allowed the system to become a backup technique for centralized multirobot systems in case of communication and GNSS dropout. The proposed approach has been verified in numerous simulations and real experiments inside a forest that represents one of the most challenging environments for deployment of compact groups of aerial vehicles.

In this work, we address the problem of achieving cohesive and aligned flocking (collective motion) with a swarm of unmanned aerial vehicles (UAVs). We propose a method that requires only onboard sensing of the relative range and bearing of neighboring UAVs, and therefore requires only proximal control for achieving formation. Our method efficiently achieves flocking in the absence of any explicit orientation information exchange (alignment control), and achieves flocking in a random direction without externally provided directional information. To implement proximal control, the Lennard-Jones potential function is used to maintain cohesiveness and avoid collisions. Our approach may be used independently from any external positioning system such as GNSS or Motion Capture, and can therefore be used in GNSS-denied environments. The performance of the approach was tested in real-world conditions by experiments with UAVs that rely only on a relative visual localization system called UVDAR, proposed by our group. To evaluate the degree of alignment and cohesiveness, we used the order metric and the steady-state value.

Highly dynamic systems such as Micro Multirotor Aerial Vehicles (Micro-MAVs) require control approaches that enable safe operation where extreme limitations in embedded systems, such as energy, processing capability and memory, are present. Nonlinear model predictive control (NMPC) approaches can respect operational constraints in a safe manner. However, they are typically challenging to implement using embedded computers on-board of Micro-MAVs. Implementations of classic NMPC approaches rely on high-performance computers. In this work, we propose a fast nonlinear model predictive control approach that ensures the stabilization and control of Micro Multirotor Aerial Vehicles (Micro-MAVs). This aerial robotic system uses a low processing power board that relies solely on on-board sensors to localize itself, which makes it suitable for experiments in GPS-denied environments. The proposed approach has been verified in numerical simulations using processing capabilities that are available on Micro-MAVs.

This paper presents a novel approach for control and motion planning of formations of multiple unmanned micro aerial vehicles(multi-rotor helicopters, in the literature also often called unmanned aerial vehicles—UAVs or unmanned aerial system—UAS) in cluttered GPS-denied on straitened environments. The proposed method enables us to autonomously design complexmaneuvers of a compact Micro Aerial Vehicles (MAV) team in a virtual-leader-follower scheme. The results of the motionplanning approach and the required stability of the formation are achieved by migrating the virtual leader along with the hullsurrounding the formation. This enables us to suddenly change the formation motion in all directions, independently from thecurrent orientation of the formation, and therefore to fully exploit the maneuverability of small multi-rotor helicopters. Theproposed method was verified and its performance has been statistically evaluated in numerous simulations and experimentswith a fleet of MAVs.

Cooperation of humans workers and a team of UAV co-workers for inspection and maintenance of electrical power is the main motivation of research presented in this paper. Collaborative human-UAV works at height are beneficial from several reasons including providing images from the ideal point of view, monitoring of the safety of individual workers, and even aerial delivering of required tools. These tasks also involve cognitive capabilities in the monitoring of the workers and the detection of unsafe behaviors, transportation of tools or parts needed by the workers and collective manipulation with the workers. In general, interaction of humans and teams of UAVs becomes an important task as aerial robots are widely spread in various applications that require the presence of people in their workspace. To achieve such interaction, group control of multiple UAVs must take states of workers (e.g. position relative to aerial co-workers and prediction of worker's future behavior), maintaining an adaptable formation and maximizing the observation of the worker. Thus, we propose in this work, a distributed vision-based nonlinear formation control (DVNFC) approach that results in an adaptable formation where the controller minimizes the error in observation always maintaining the visualization of the human by the whole formation. We performed several numerical simulations using ROS/Gazebo with real-time visual feedback to validate our approach.

Motion control theory applied to multi-rotor aerial vehicles (MAVs) has gained attention with the recent increase in the processing power of computers, which are now able to perform the calculations needed for this technique, and with lower cost of sensors and actuators. Control algorithms of this kind are applied to the position and the attitude of MAVs. In this paper, we present a review of recent developments in position control and attitude control of multi-rotor aerial robots systems. We also point out the growth of related research, starting with the boom in multi-rotor unmanned aerial robotics that began after 2010, and we discuss reported field applications and future challenges of the control problem described here. The objective of this survey is to provide a unified and accessible presentation, placing the classical model of a multi-rotor aerial vehicle and the proposed control approaches into a proper context, and to form a starting point for researchers who are initiating their endeavors in linear/nonlinear position, altitude or attitude control applied to MAVs. Finally, the contribution of this work is an attempt to present a comprehensive review of recent breakthroughs in the field, providing links to the most interesting and most successful works from the state-of-the-art.