Search

Spacecraft formation flying has attracted notable interest in recent years, thanks to the properties of flexibility to different mission needs and enhanced robustness and capabilities compared to monolithic space systems. A key element for the full exploitation of this technology consists in the automation of formation management and control, allowing higher independence from ground stations as well as prompter reactions and higher autonomy in task-assignment and decision making. In this context, this work proposes a distributed architecture to autonomously coordinate and control formations containing any number of members applying low-thrust propulsion and employing finite state machines to make decisions and assign tasks to the elements of the formation based on their current state. The nominal relative trajectories (called slots) are represented within the architecture through a set of geometric parameters characterizing their stability, size, and shape. The architecture can autonomously maintain the spacecraft on a set of nominal slots, as well as reconfigure the formation to a newly assigned geometry and deal with contingency situations, performing collision avoidance manoeuvres and keeping track of the slots of the formation that remain empty to autonomously define a course of action allowing the evading spacecraft to re-join the formation safely. At the same time, a strategy for the identification of the geometrical conditions to be violated to efficiently perform formation maintenance is proposed. The architecture is tested by means of numerical simulations within a GMAT-MATLAB integrated environment considering a quasi-linear safe ellipse formation aimed at distributed synthetic aperture radar applications as reference test case. Simulation results demonstrate that the proposed architecture allows effective autonomous coordination within tight formations, with cm-level ΔV expenses, while being fully scalable with the number of satellites.

Missions requiring autonomous, close-proximity operations of spacecraft, such as On-Orbit Servicing, On-Orbit Assembly and Active Debris Removal, have become a thriving topic in the aerospace research community over the last decades, not only from an economic, operative, and scientific perspective, but also as a mean of ensuring the sustainability of the space environment. These operations involve a variety of technological challenges, most of which are related to the need of autonomous and safe Guidance, Navigation and Control systems. Since the future of these mission scenarios is strictly tied to spacecraft standardisation and modularity, relative navigation employing monocular cameras on servicing platforms to approach targets equipped with artificial markers for pose estimation purposes has drawn great attention. Following this trend, this paper presents an original vision-based pose estimation architecture for relative navigation with respect to passively cooperative targets equipped with ArUco markers. The proposed architecture foresees two operative modes, namely Acquisition and Tracking. The first features ArUco's detection through hue-saturation-value image representation, their identification by reading their built-in code and the computation of the pose without a-priori knowledge. The second, instead, takes advantage of prior pose estimates to speed up the entire processing pipeline. Performance is assessed through an extensive numerical simulation campaign, considering as test scenario the final approach phase of a rendezvous manoeuvre to reach a satellite belonging to a large constellation in Low Earth Orbit, and using the Planet and Asteroid Natural scene Generation Utility tool for realistic synthetic image generation. The dedicated tests on the Acquisition mode show that successful marker detection and pose initialization is achieved from up to 99.76% of the possible relative position and attitude states of the chaser with respect to the target at beginning of the final approach trajectory. As the chaser gets closer to the target, results highlight significant robustness of both operative modes against illumination conditions and uncertainties in the knowledge of camera intrinsic parameters. Overall, the architecture shows pose estimation accuracies up to millimetric and sub-degree levels.

This paper presents an experimental activity aimed at assessing performance of techniques for inertia and attitude parameters estimation of an uncooperative but known space target. The adopted experimental set-up includes a scaled-down 3D printed satellite mock-up, a spherical air bearing and a low-cost solid-state LIDAR. The experimental facility also comprises a motion capture system to obtain a benchmark of the pose (position and attitude) parameters and an ad-hoc designed passive balancing system to keep the centre of mass as close as possible to the centre of rotation. The LIDAR-based 3D point clouds, collected while the target rotates on the spherical air-bearing to reproduce the rotational dynamics of an almost-free rigid body, are processed to obtain pose and angular velocity estimate. Those data are used to determine the inertia properties of the target by solving a linear system based on the integral form of Euler equations. Performance and robustness of the algorithm for attitude and inertia properties estimation are assessed considering both the inertia benchmark provided by a high-fidelity CAD model of the target and the pose solution obtained from the motion tracking system and its extrinsic calibration with respect to the LIDAR.

This paper presents an innovative multi-sensor relative navigation module conceived to enable close-proximity operations for on-orbit servicing of small satellites. The module is composed of a Near-Infrared Laser Range Finder, acting both as range measurer and illuminator, and a monocular camera able to detect a set of reflecting fiducial markers installed on the target for pose estimation. Sensor fusion is performed in several points of the proposed processing pipeline, first following a cross-cueing logic in which the range aids the pose determination process, and later through the integration of both measurements into a Multiplicative Extended Kalman Filter for relative state estimation. The entire relative navigation architecture is first tested by numerically reproducing a close-range rendezvous trajectory of a chaser spacecraft moving toward the target, and using the open-source software Blender for synthetic image generation. Simulations are aimed at determining nominal relative state estimation accuracy (also considering variable sensors' acquisition rates) as well as performance under non-nominal conditions (e.g., due to a temporary loss of measurements). Focusing on the image processing and pose estimation pipeline, an in-depth analysis of the computational efficiency is then carried out by running the proposed algorithms on an embedded processing unit (i.e., the Raspberry Pi 4 Model B) using both synthetic data and real images collected within a dedicated experimental testbed. Numerical and experimental results demonstrate capability to attain pose and relative state estimates in line with accuracy requirements to ensure autonomous and safe close-proximity operations, e.g., up to sub-cm and sub-degree error levels in relative position and attitude respectively. In addition, the algorithm's computational runtime has shown to be compatible with real-time implementation at an acquisition frequency of 2 Hz.

The future of UAS operations requires adequate and efficient Sense and Avoid strategies to ensure their safe and secure integration in both controlled and uncontrolled airspace. To make onboard implementation of these strategies feasible, research must focus on the development of sensing and tracking solutions, taking size, weight and power (SWaP) constraints as well as the challenging scenarios characterizing low altitude operations, into account. In this framework, visual cameras and low SWaP radars are among the most popular sensing choices. Hence, exploiting such sensing solutions, this paper proposes both single and multi-sensor detection and tracking approaches and compares their performance. Specifically, the developed techniques are tested on data retrieved during ground-to-air tests which involve a small UAV flying near collision geometries, starting from a range of about 550 m. Different tracking strategies are considered including standalone visual, standalone radar, and fused radar-visual. Concerning intruder detection, several visual-based techniques are investigated based on machine learning, morphological filtering and template matching. Radar detections are filtered and centroided with ad hoc algorithms. While in clear air conditions comparable declaration ranges, larger than 500 m, are provided by all the tested approaches, results show the advantages of using a fused strategy to attain sub-degree angular and angular rate tracking accuracy coupled with the highly accurate range and range rate, around 2 m and 1 m/s, respectively, typical of radars. Conflict detection performance is proven to also benefit from the use of a fused strategy in terms of smaller errors in the estimation of the distance at closest point of approach.

Improvements in low altitude, non-cooperative sense and avoid are of major interest for collision hazard mitigation within the UTM/UAM/U-Space framework. In this regard, the sensing architecture must be carefully designed so that its detection and tracking performance is suitable for timely and reliable conflict assessment, while respecting size, weight, power and costs constraints, which are particularly strict for small aerial vehicles. Within this framework, an experimental assessment of non-cooperative sensing solutions based on a lightweight radar and a visual camera, respectively, is presented in this paper. Visual detections are obtained by using a Deep Learning-based neural network, while raw detections produced by the radar are first filtered based on Doppler information to remove ground clutter, and then clustered by means of a centroiding approach. The resulting detection sets are used to generate tentative and firm tracks using customized Kalman filtering techniques. Following a research plan that foresees data gathering with incremental complexity, ground-to-air tests have been carried out using a small UAV as flying intruder, and Carrier-Phase Differential GNSS to get a reference solution and assess visual-based and radar-based detection and tracking performance. Results achieved by standalone radar and visual sensing solutions clearly highlight the potential of sensor fusion strategies to take advantage of their complementary characteristics.

Future Active Debris Removal missions will require an autonomous spacecraft (chaser) to safely monitor at close distance, and then approach and dispose an inactive artificial space object (target). Since these targets are uncooperative, meaning that they cannot provide any hint or help to the chaser navigation system, such operations require the target-chaser relative state and the target inertia parameters to be accurately estimated relying only on measurements from active or passive Electro-Optical sensors. In this framework, this paper proposes an original multi-step architecture for the estimation of the relative motion and inertia parameters of an uncooperative target during a close-range monitoring trajectory. In the first phase, LIDAR-based pose measurements and a smoothing approach are used to retrieve accurate, linearly independent estimates of the target angular velocity. These estimates are then used to compute the target's moments of inertia ratios solving a linear system based on the conservation equation for the angular momentum. Once the inertia parameters are accurately estimated, the LIDAR-based pose measurements are used to feed an Unscented Kalman Filter to determine the full relative state according to a loosely coupled configuration. The architecture foresees autonomous failure detection strategies to avoid divergence in the relative state estimation error caused by unavoidable, unfavorable target observation conditions occurring during the monitoring trajectory. Performance assessment is carried out through numerical simulations realistically reproducing close-range relative motion dynamics and LIDAR sensor operation, and considering targets characterized by highly variable size, shape, and orbital dynamics as test cases.

This paper deals with the problem of electrical asset mapping with LiDAR-equipped UAVs. Compared with standard solutions relying on ground-based augmentation systems and high value payloads integrated on single assets, two main innovations are proposed and discussed. First, the possibility to fulfill georeferencing accuracy and precision requirements without ground-based GNSS stations/networks is explored, exploiting multi-frequency multi-constellation receivers and the added value of the European GNSS Galileo. Precise Point Positioning processing is used to mimic the High Accuracy Service, which will be made available by Galileo in the near future providing decimeter-level absolute accuracy. GNSS estimates are fused with inertial measurements to the aim of positioning and attitude reconstruction. Second, the application potential of multi-drone systems is analyzed. A cooperative navigation strategy is adopted which exploits drone-to-drone visual tracking and differential GNSS processing to provide high accuracy attitude information. Formation geometry of the cooperative platforms is investigated with the aim of minimizing the attitude error. Navigation and georeferencing performance are tested on synthetic and experimental data using error metrics relevant to powerline reconstruction.

This paper describes the design, characterization and preliminary testing of a relative navigation module based on electro-optical sensors and tailored for close-proximity operations with respect to passively cooperative targets of the CubeSat class. The proposed unit relies on a laser range finder with a relatively large beam divergence and a visual camera. The laser range finder operates in the near-infrared band and is exploited to both provide direct range measurements and actively illuminate the scene. The visual sensor and its optics are coherently chosen and allow obtaining full pose (i.e., relative position and attitude) measurements by processing the acquired target’s images. The pose estimation procedure is based on the detection and identification of a set of fiducial markers installed on the target surface and highly reflective in the near-infrared band. Experimental tests are carried out at components level to both characterise the laser range finder and preliminarily assess the performance of the proposed pose determination approach. The results show that the proposed approach is robust to a large number of outliers, produced by highly reflective insulation layers typically covering satellite surfaces. Also, in all the considered test cases pose estimates with pixel-level reprojection errors, corresponding to sub-millimetre accuracy, are obtained.

This paper addresses the possibility to exploit Enhanced Vision Systems (EVS) for landing operations within the Urban Air Mobility (UAM) framework. First, an analysis of sensing requirements is carried out in terms of sensors’ Field of View (FOV), operational range, resolution, and frame rate. These requirements are estimated considering different factors such as airspace structure and possible landing trajectories, navigation performance, and wind conditions. A preliminary assessment of the level of maturity of the available technologies is obtained as a result of the comparison of the EVS state of the art and the estimated visual requirements for the UAM scenarios. Finally, the performance of sensing systems fulfilling the previously defined requirements is assessed within a numerical simulation framework able to realistically reproduce UAM landing scenarios including trajectory/attitude disturbances. Specifically, visual-based techniques for pose estimation with respect to a vertiport are developed and then integrated within a multi-sensor fusion architecture combining other sources such as inertial sensors.

This paper discusses the integration of air-to-air visual measurements within a cooperative multi-vehicle architecture conceived to improve navigation performance of small Unmanned Aerial Vehicles in GNSS-challenging environments. The key concept is to exploit cooperation with other aircraft flying under better GNSS coverage conditions by exchanging navigation data and using a monocular camera system for relative sensing purposes. Specifically, accurate line of sight, obtained using Deep Learning-based detectors and local image analysis, are complemented with distance information achieved through an innovative shape-based passive ranging approach accounting for both the target attitude and its position in the field of view. Camera-based measurements are combined with the estimates generated by the other onboard sensors, within a customized Extended Kalman Filter in a closed loop fashion, since navigation estimates are used in feedback as hints for visual processing. An experimental flight test campaign is carried out using two quadcopters. A comparison between filter performance achievable as a function of the specific set of available information sources, i.e., bearing-only vs. line of sight and ranging, is carried out. Results show that the trade-off between correct, false, and missed detections, as well as the passive ranging accuracy allow the filter ensuring metric-level positioning error within GNSS-challenging areas. The added value of using both bearing and range measurements strongly depends on the formation geometry and the GNSS coverage conditions and can be predicted thanks to the “generalized dilution of precision”.

The continuous growth of space debris motivates the development and the improvement of tools that support the monitoring of a more and more congested space environment. Satellite breakup models play a key role to predict and analyze orbital debris evolution, and the NASA Standard Breakup Model represents a widely used reference, with current activities relevant to its evolution and improvements especially towards fragmentation of small mass spacecraft. From an operational perspective, an important point for fragmentation modelling concerns the tuning of the breakup model to achieve consistency with orbital data of observed fragments. In this framework, this paper proposes an iterative approach to estimate the model inputs, and in particular, the parents’ masses involved in a collision event. The iterative logic exploits the knowledge of Two Line Elements (TLE) of the fragments at some time after the event to adjust the input parameters of the breakup model with the objective of obtaining the same number of real fragments within a certain tolerance. Atmospheric re-entry is accounted for. As a result, the breakup model outputs a set of fragments whose statistical distribution, in terms of number and size, is consistent with the catalogued ones. The iterative approach is demonstrated for two different scenarios (i.e., catastrophic collision and non-catastrophic collision) using numerical simulations. Then, it is also applied to a real collision event.

This paper presents a strategy to adaptively select the field of view and the resolution of a scanning LIDAR in the frame of close-proximity operations with non-cooperative space targets. Specifically, the idea is to relate the selection of these operational parameters to the estimated target-chaser relative distance as well as to the size occupied by the target in the sensor field of view, with the goal to improve the performance of the relative navigation system. The proposed method is supported by a chaser attitude control strategy which allows keeping the sensor boresight axis always pointed towards the center of mass of the target. The performance achieved by the adoption of this strategy is evaluated by simulating two types of close-range trajectories in a numerical simulation environment which reproduces both the operation of a scanning LIDAR and the relative motion between two spacecraft.

The AMPERE project aims at engineering and starting to commercialize a dedicated solution, to be used for electrical power network information gathering. AMPERE will support decision making actors (e.g., institutions and public/ private companies in charge to manage electrical network) to collect all needed info to plan electrical network maintenance and upgrade. In particular, the need for such a solution comes in emerging countries (worldwide) where, despite global electrification rates are significantly progressing, the access to electricity is still far from being achieved in a reliable way, and, there is lack of knowledge about already deployed infrastructure. The AMPERE data collection solution relies on drones equipped with GNSS/INS units, LIDAR, optical and thermal cameras. In such a context, Galileo is a key enabler - especially, considering its free-of-charge High Accuracy Service (HAS) and its highly precise E5 AltBOC code measurements- as a core component to map electric utilities, optimize decision making process about the network development and therefore increase time and cost efficiency, offering more convenient way to manage energy distribution. The project is also aimed at investigating the potential of cooperative multi-UAV solutions.

In many Space Surveillance and Tracking (SST) operations, from the cataloguing of trackable space objects to orbit determination and correlation, ground-based measurements play a central role. In this frame, the monitoring and, if necessary, the update of the sensors' calibration parameters is of utmost importance to correctly evaluate the positions of tracked objects. In this work, a method for the metrological characterization of ground-based sensors for SST applications is implemented, and the performance of the developed sensor calibration tool is assessed. Starting from the acquisition of information about the sensor and data from Tracking Data Messages and Orbit Ephemeris Messages, and assuming a Gaussian distribution for all the observed parameters, the tool computes the residuals and outputs the main statistics for all the parameters of interest. The distributions of the residuals are analyzed to evaluate the statistical significance of the results and the robustness of the hypothesis of Gaussian distribution.

This paper discusses an approach conceived to improve navigation performance of small Unmanned Aerial Vehicles (UAVs) in GNSS-challenging environments by exploiting cooperation with other aircraft flying in better GNSS coverage conditions. Cooperation is realized by exchanging navigation data (i.e., GNSS observables when available) and exploiting a monocular camera system for relative vision-based tracking. Cooperative measurements are used within an Extended Kalman Filter, developing a solution potentially ready for real-time applications. The visual algorithm exploits both Deep Learning-based detectors and standard machine vision techniques to provide not only accurate line-of-sight but also distance estimates, and it is designed to deal with targets placed both above and below the horizon. The two algorithmic blocks are integrated in a closed loop fashion since navigation estimates are used in feedback to support visual processing. An experimental flight test campaign is carried out using two quadcopters to assess attainable navigation performance in terms of attitude and positioning. Results compare filter performance when using line-of-sight only with the case of using line-of-sight and ranging measurements altogether. They demonstrate that reliability and integrity of visual algorithms are good enough for the navigation filter needs, and that metric positioning error is achieved within GNSS-challenging areas by using the proposed cooperative strategy. The added value of range estimation strongly depends on the formation geometry and the GNSS coverage conditions.

This paper lies in the framework of mission scenarios, such as Active Debris Removal and On-Orbit Servicing, which require an active spacecraft (chaser) to orbit in close-proximity with respect to a space target. Specifically, these activities involve relative orbital maneuvers, such as monitoring, rendezvous and docking, in which the target-chaser distance ranges from a few tens of meters (depending on the target size) up to contact (in the case of docking). A critical challenge related to the realization of these maneuvers is the need to minimize the risk of collision, considering that the target is a non-cooperative object which may be characterized by uncontrolled rotational dynamics. This goal can be achieved by designing relative trajectories which satisfy specific constraints in terms of safety and stability, on one side, as well as by exploiting relative navigation technologies and algorithms which provide highly accurate estimates of the target-chaser relative motion parameters thus allowing to relax the control requirements. Both these aspects are addressed by this paper with focus on Geostationary Earth Orbits since they represent a particularly crowded orbital region in which the possibility to remove large debris and to extend the operative life of spacecraft, such as telecommunication ones, may have a significant scientific and economic benefit. Hence, an original method is presented to design safety ellipses for target monitoring around GEO targets, which, simultaneously, can provide optimal relative observation geometry for relative navigation (pose determination) using Electro-Optical sensors. The design approach is formulated in mean orbit parameters and it is based on a relative motion model relevant to two-satellite formations which includes the non-Keplerian perturbations due to secular Earth oblateness, as well as the possibility of considering targets moving along a small-eccentricity orbit. An example of trajectory design is shown considering a GEO target as test case. Given this trajectory, pose determination performance is also evaluated within a numerical simulation environment capable of realistically reproducing target-chaser relative dynamics, the operation of a scanning LIDAR selected on board the chaser as relative navigation sensor, and pose estimation algorithms based on the processing of 3D point clouds.

Autonomous relative navigation is a critical functionality which needs to be developed to enable safe maneuvers of a servicing spacecraft (chaser) in close-proximity with respect to an uncooperative space target, in the frame of future On-Orbit Servicing or Active Debris Removal missions. Due to the uncooperative nature of the target, in these scenarios, relative navigation is carried out exploiting active or passive Electro-Optical sensors mounted on board the chaser. The focus here is placed on active systems, e.g., LIDARs. In this paper, an original loosely-coupled relative navigation architecture which integrates pose determination algorithms designed to process raw LIDAR data (i.e., 3D point clouds) within a Kalman filtering scheme is presented. Pose determination algorithms play a twofold role being used to initialize the filter state and covariance as well as in the update phase of the Kalman filter. The proposed filtering scheme is an Unscented Kalman Filter designed to use, as measurements for the update phase, relative position, attitude and angular velocity estimates. Performance assessment is carried out within a simulation environment realistically reproducing the operation of a scanning LIDAR and the relative motion between two spacecraft during a target monitoring maneuver. The numerical simulation campaign demonstrates robustness of the proposed approach even when dealing with challenging conditions (e.g., low range measurement accuracy, low update rate and high point-cloud sparseness) determined by the LIDAR noise level and operational parameters.

The interest of the space community toward missions like On-Orbit Servicing of functional satellite to extend their operative life, or Active Debris Removal to reduce the risk of collision among artificial objects in the most crowded orbital belts, is significantly increasing for both economical and safety aspects. These activities present significant technical challenges and, thus, can be enabled only by increasing the level of autonomy and robustness of space systems in terms of guidance, navigation and control functionalities. Clearly this goal requires the design and development of ad-hoc technologies and algorithms. In this framework, this paper presents an original architecture for relative navigation based on a single passive camera able to fully reconstruct the relative state between a chaser spacecraft and a non-cooperative, known target. The proposed architecture is loosely coupled, meaning that pose determination and full relative state estimation are entrusted to separate, but rigidly interconnected processing blocks. Innovative aspects are relevant to both the pose determination algorithms and the filtering scheme. Preliminary performance assessment is carried out by means of numerical simulations considering multiple realistic target/chaser relative dynamics and target geometries. Results allow demonstrating robustness against measurement error sources caused possibly by image processing as well as fast rotational dynamics.

This paper describes a calibration technique aimed at combined estimation of onboard and external magnetic disturbances for small Unmanned Aerial Systems (UAS). In particular, the objective is to estimate the onboard horizontal bias components and the external magnetic declination, thus improving heading estimation accuracy. This result is important to support flight autonomy, even in environments characterized by significant magnetic disturbances. Moreover, in general, more accurate attitude estimates provide benefits for georeferencing and mapping applications. The approach exploits cooperation with one or more “deputy” UAVs and combines drone-to-drone carrier phase differential GNSS and visual measurements to attain magnetic-independent attitude information. Specifically, visual and GNSS information is acquired at different heading angles, and bias estimation is modelled as a non-linear least squares problem solved by means of the Levenberg–Marquardt method. An analytical error budget is derived to predict the achievable accuracy. The method is then demonstrated in flight using two customized quadrotors. A pointing analysis based on ground and airborne control points demonstrates that the calibrated heading estimate allows obtaining an angular error below 1°, thus resulting in a substantial improvement against the use of either the non-calibrated magnetic heading or the multi-sensor-based solution of the DJI onboard navigation filter, which determine angular errors of the order of several degrees.

Very low altitude non-cooperative sense and avoid is considered as an essential component of multi-layered mitigation strategies for collision hazard within the UTM/U-Space framework. However, it currently represents a challenging problem with open issues concerning performance trade-offs and technological options. In fact, low altitude operations emphasize sensing challenges such as ground clutter removal for radar, and below-the-horizon detection for optical sensors. These issues have a significant impact in scenarios involving small UAS, which may be characterized by low detectability though potentially generating relatively fast encounter geometries. As a contribution to this framework, this paper proposes some improved sensing strategies for these scenarios. The first idea is to optimize architectural and algorithmic trade-offs by accounting for the possible closure rates corresponding to near collision scenarios. This approach can be used to develop adaptive sensing strategies, as well as to select different sensors to cover different areas of the aircraft field of regard. The concept is demonstrated in numerical analyses focused on the computation of probability of missed and false conflict detection. Then, with regards to purely visual architectures, the possibility to extract and use shape-based ranging information to improve conflict detection performance is investigated through flight tests with small UAS. Finally, the paper describes first experimental activities conducted with a compact collision avoidance radar.

There is currently no metric to discern the relative importance of an asteroid close approach with the Earth from the rest. This typically leads to communication media releasing information on close approach events under undefined criteria. This further means that, on many occasions, they inform about events that lack relevance or are publicised to levels which are not justified in terms of objective considerations. The evaluation of that relevance is further complicated by the fact that there are many factors that play a role in this subject, as asteroid size, minimum close approach distance, close approach velocity, actual NEO population, etc. In this paper, we establish an objective criterion to evaluate the expected frequency for the close approach of an NEO, based on the current estimates of the population of such objects, the object absolute magnitude and the parameters of the close approach. In addition, we propose a scalar close approach index that can be used to discriminate the relevance of an NEO close approach with the Earth, and we apply it to the current list of close approaches reported in ESA’s NEO Coordination Centre web portal.

This paper presents an original approach for hazard detection and landing site selection, exploiting LIDAR measurements, to be used in the framework of autonomous planetary exploration scenarios. The proposed algorithm is designed aiming to reduce the computational effort typically required to process highly dense LIDAR point cloud data for the generation of a hazard map. In this respect, the proposed logic consists in selecting the safest landing region by diving the hazard detection process in two phases. First, a coarse hazard map is generated to discard most of the dangerous regions of terrain from the selection process. Second, a fine hazard map is generated over the safe areas to identify the best landing spot. The level of hazard is measured taking the local surface slope, height, and roughness into account. Preliminary performance assessment of the proposed technique is carried out using point clouds acquired imaging a mockup of the Lunar terrain with the RVS3000-3D ® scanning LIDAR (developed by Jena Optronik GmbH). In addition, the algorithm is tested processing a high-density terrestrial DEM, available through the Open Topography website, to evaluate its performance against a realistic landing terrain scenario.

The paper focuses on guidance and relative navigation aspects in the framework of an Active Debris Removal mission. Specifically, an original multi-step strategy, inspired by the one designed for the OSAM-1 mission, is proposed to perform rendezvous toward an uncooperative target reaching a relative distance suitable for monitoring using Electro-Optical sensors. At each step, passive safety is guaranteed by setting pre-defined regions of interest that cannot be breached by the chaser until the mission control on ground provides an Authority-to-Proceed command. As regards the navigation aspects, particular attention is paid to the problem of obtaining highly accurate relative state estimates in close-proximity conditions using a LIDAR. Finally, an ad-hoc reconfiguration strategy is proposed to keep the chaser within a limited along-track distance from the target. The reconfiguration manoeuvre is commanded on the basis of the LIDAR-based relative state estimates. Performance assessment is carried out by means of numerical simulations.

The research project named "CREATEFORUAS" aims at enhancing "trustable" flight autonomy for small Unmanned Aircraft Systems (UAS), by tackling different enabling technologies. The paper briefly describes the main research areas of the project which are relevant to vision-based environment understanding, sense and avoid, robust control, and multi-UAV cooperation. Then, it also provides a snapshot of current developments presenting simulation and experimental results.

The concepts of On-Orbit Servicing and Active Debris Removal have been introduced to increase the operating life of active satellites and as a debris remediation measure, respectively. For these operations, a servicing spacecraft (chaser) must orbit safely and autonomously in close proximity to a target that is typically known, but non-cooperative. So, advanced guidance, navigation and control functions are required to carry out these missions. This work focuses on the problem of estimating the target-chaser relative state and the target inertia properties exploiting measurements from a LIDAR during the different phases of a close-range rendezvous maneuver. Specifically, the relative trajectory is designed to provide both safety and optimal observation conditions. To this aim, it is composed of multiple safety ellipses for monitoring purposes, in which the target-chaser relative distance is of the order of tens of meters, followed by a R-bar trajectory for the final approach. In this scenario, the LIDAR-based relative navigation solution is a multi-step processing architecture, which exploits pose determination techniques, unscented Kalman filtering and a linear least squares solver. The performance of this approach is evaluated by means of numerical simulations.

Being able to perform autonomous relative navigation with respect to an uncooperative and possibly unknown space target is a critical task in the framework of on-orbit servicing and active debris removal applications. Electro-Optical sensors are required because of the non-cooperative nature of the target. Specifically, the solution proposed in this work is a LIDAR-based relative navigation architecture, which combines, in a loosely-coupled configuration, a pose determination algorithm, designed to process raw LIDAR data (i.e. point clouds), with an unscented Kalman filter to estimate relative motion parameters as well as inertia properties of the target. The developed algorithm is tested in a numerical simulation environment realistically reproducing the operation of a scanning LIDAR and the relative motion between two spacecraft.

This paper presents an innovative approach to design relative trajectories suitable for close-proximity operations in orbit, by assigning high-level constraints regarding their stability, shape and orientation. Specifically, this work is relevant to space mission scenarios, e.g. formation flying, on-orbit servicing, and active debris removal, which involve either the presence of two spacecraft carrying out coordinated maneuvers, or a servicing/recovery spacecraft (chaser) performing monitoring, rendezvous and docking with respect to another space object (target). In the above-mentioned scenarios, an important aspect is the capability of reducing collision risks and of providing robust and accurate relative navigation solutions. To this aim, the proposed approach exploits a relative motion model relevant to two-satellite formations, and developed in mean orbit parameters, which takes the perturbation effect due to secular Earth oblateness, as well as the motion of the target along a small-eccentricity orbit, into account. This model is used to design trajectories which ensure safe relative motion, to minimize collision risks and relax control requirements, providing at the same time favorable conditions, in terms of target-chaser relative observation geometry for pose determination and relative navigation with passive or active electro-optical sensors on board the chaser. Specifically, three design strategies are proposed in the context of a space target monitoring scenario, considering as design cases both operational spacecraft and debris, characterized by highly variable shape, size and absolute rotational dynamics. The effectiveness of the proposed design approach in providing favorable observation conditions for target-chaser relative pose estimation is demonstrated within a simulation environment which reproduces the designed target-chaser relative trajectory, the operation of an active LIDAR installed on board the chaser, and pose estimation algorithms.

This paper proposes an original approach for visual-based obstacle detection and tracking, and conflict detection, suitable to endow small Unmanned Aircraft Systems with non-cooperative Sense and Avoid capabilities. Specifically, it is designed to detect and track an uncooperative flying object (intruder), and to establish whether or not it represents a collision threat. It is based on three main algorithmic steps, each aided by the use of ownship navigation data. First, visual detection is carried out using two Deep Learning based neural networks (operating above and below the horizon, respectively) followed by local image analysis to improve the accuracy in the detected intruder position on the image plane. Second, tentative track generation and firm tracking are executed exploiting local association, multi-temporal frame differencing, and Kalman filtering. Finally, conflict detection is applied to each firm track based on the estimate of line of sight and line of sight rate in stabilized coordinates. An experimental dataset, which reproduces realistic low-altitude encounter geometries with two customized quadcopters, is collected to assess proposed approach performance. The two vehicles, equipped with high-resolution color cameras, are flown at slightly different altitudes so that the intruder is located above and below the horizon in the two respective image subsets. The proposed approach allows the target to be declared at relatively long range and to be tracked with a line of sight rate accuracy of the order of tenths of degrees per second, which is effective for conflict detection.

This paper presents an innovative method to estimate the systematic error, i.e., bias, affecting the measurements of magnetometers on board of small UAVs, thus significantly improving magnetic heading accuracy. Excluding the presence of other external disturbances, this bias is largely produced by the magnetic field generated on board due to the presence of electric equipment, such as the payload and the engines, especially in very compact configurations (besides eventual uncompensated sensor offsets). Consequently, ground calibration techniques have limits. The proposed approach is conceived to rely on a simple and fast flight procedure during which all the devices operate in nominal conditions. It exploits the capability of the UAV whose magnetometers need to be calibrated (chief) to detect and track a cooperative vehicle (deputy) using a visual camera. Also, the two UAVs must fly under nominal GNSS coverage thus enabling relative positioning. By integrating deputy line-of-sight and chief/deputy differential GNSS information, the magnetic biases’ determination problem can be expressed as a system of non-linear equations which is solved using a customized implementation of the well-known Levenberg-Marquardt algorithm. The calibration can be carried out either off-line, using the data collected in flight, or directly on board, i.e., in real time. In the latter case, the two UAVs must be able to exchange navigation data through a reliable communication link. Demonstration of feasibility and performance assessment of this method are done using data collected by means of a flight test campaign.

This paper deals with the problem of 3D surface reconstruction processing point clouds produced by a LIDAR. In the framework of an algorithmic architecture for pose determination of an uncooperative spacecraft, this functionality is required to update the existing model of the target geometry (reference model), if a non-negligible inconsistency with respect to its real geometry is detected. Hence, two original approaches to generate triangular mesh models from point clouds are developed, which are based on the marching-cube and Delaunay triangulation concepts, respectively. The output of these techniques can then be exploited within an original algorithmic strategy conceived to carry out the update function on the reference model, aiming at improving pose estimation accuracy. The two 3D surface reconstruction approaches are tested over point clouds acquired using a new prototype of the RVS 3000-3D® LIDAR produced by Jena-Optronik GmbH and a satellite mock-up as target. Their compatibility for real-time implementation on FPGA hardware is also preliminary evaluated. Finally, performance assessment of the proposed strategy for on-line update of the reference model is carried out within an ad-hoc experimental setup, using a Kinect V2 sensor to acquire point clouds of a scaled satellite mock-up.

The safe integration of Unmanned Aircraft Systems in the civil airspace requires the development of Sense and Avoid (also named Detect and Avoid) systems, which represent a technology needed to detect a threat (e.g., an intruder vehicle) and, consequently, command a maneuver suitable to avoid collision within a small amount of time. In this framework, this paper presents a quantitative assessment of conflict detection performance achievable by different sensing architectures, which can be designed and installed on board of small Unmanned Aerial Vehicles. The attention is focused on non-cooperative systems, which can rely on active sensors (such as radars and LIDARs), passive sensors (such as visual cameras) or multi-sensor-based solutions. The aim is to quantitatively identify the potentialities and limitations of each technological solution. The analysis is carried out by means of a numerical simulation environment able to reproduce horizontal collision scenarios. Specifically, conflict detection performance is expressed in terms of two probabilities, namely the probability of missed conflict detection and false conflict detection, which are evaluated as a function of three parameters which allow uniquely defining the collision scenario, i.e., the time to and distance at the closest point of approach and the azimuth of the intruder with respect to the velocity vector of the ownship. Since, the probability determination process includes the effect of tracking uncertainties (i.e., probability of firm tracking and false track generation), this analysis also allows highlighting the diverse but interconnected effects of sensing accuracy and detection range.

This paper presents first results of an experimental flight-test campaign aimed to gather data for performance assessment of non-cooperative Sense and Avoid architectures for small Unmanned Aircraft Systems (UAS). The attention is here focused on vision-based approaches. An innovative sensing technique is proposed which exploits a Deep Learning (DL) network as the main processing block of the detector algorithm, and a multi-temporal strategy for track generation and confirmation. Both the detection and tracking phases foresee ad-hoc solutions to deal with the presence of intruders either above or below the horizon. Two customized small quadcopters, equipped with high-resolution color cameras, are used to reproduce in flight low-altitude, near-collision scenarios characterized by different speed and height above ground, thus being able to act simultaneously as ownship and intruder. Results demonstrate the capability of the DL-based detector to provide maximum declaration range around 300 m and 100 m, above and below the horizon, respectively. The tracker can robustly produce firm track of the intruder while rejecting many false positives, particularly occurring in below-the-horizon scenarios.

This paper presents the results of research activities carried out to characterize and test the operation of a latest generation commercial, high-resolution Time-of-Flight (TOF) camera. The aim is to preliminary evaluate the achievable performance as well as potential limitations related to the use of this instrument for autonomous navigation purposes. Two fields of investigation have been identified: autonomous navigation of Unmanned Aerial Vehicles flying in GPS-denied environments; autonomous relative navigation between non-cooperative space objects. With reference to these applications, first, a metrological characterization has been operated within a laboratory setup. Second, experimental tests have been carried out by processing point clouds acquired by the TOF sensor with state-of-the-art algorithms (for depth-based odometry and non-cooperative pose determination, respectively).

This paper presents processing architectures and algorithms for airborne visual tracking in cooperative scenarios, such as the ones relevant to vision-aided swarming of Unmanned Aerial Vehicles. While in the considered applications it is possible to exploit a priori information relevant to the physical configuration of the target UAVs, and coarse knowledge of the relative position among aircraft, significant requirements exist in terms of the tradeoff between false positives and missed detections. The focus is set on medium/large distances among UAVs, and the presented solution combines adaptive template matching and morphological filtering to achieve reliable performance for variable background and illumination conditions. In order to minimize requirements related to computational load and vision system performance (e.g., minimum frame rate), navigation and cooperative information is used within the tracking algorithm. The adopted approach is tested on experimental data acquired in multi-UAV flight tests. First results demonstrate the potential of fusing information derived from cameras, onboard sensors, and target navigation system.