The research programme has four major innovative aspects: 1) it will combine computational intelligence, dynamical systems theory and orbital mechanics to quantify the uncertainty of rare, high risk events and correlate spatially and temporally distant incidents, 2) it will develop the concept of Space Traffic Management to improve the resilience of the space environment, 3) it will develop the concept of criticality of asteroids in view of future exploration, exploitation and deflection, 4) it will develop the concept of exploration with small spacecraft.
The following are brief overviews of the state of the art and major innovative aspects of every Work Package.
COMPUTATIONAL INTELLIGENCE AND THE QUANTIFICATION OF UNCERTAINTY
The prediction of rare and catastrophic events, like collisions with space debris or impacts with asteroids, and anomalies that can lead to catastrophic events is key to improve the resilience of the space environment and planetary defence. This requires an element of uncertainty quantification and an element of operation planning and optimization. In both contexts, computational intelligence can offer machine learning, optimisation and statistical analysis techniques to study the global behavior of space objects, identify rare events and quantify the associate uncertainty. It can also be used to provide optimal planning and scheduling of collision and impact avoidance maneuvers, fundamental in Space Traffic Management and Planetary Defence. The problem of correlating spatially and temporally distant events translates into processing multiple sources of information to learn from past data about patterns and early warning signals that can predict rare events and plan recovery measures. Data fusion techniques able to deal with imprecise, incomplete and spurious data have been lately subject of an intense research and techniques coming from the field of fuzzy set theories, probabilistic methods and evidence theory have been applied also in this context. Moreover recent advances in deep belief networks for unsupervised anomaly detection, and time series analysis for anomaly prediction, have proved to be more effective in modeling of complex nonlinear systems.
The major innovations come from the combination of computational intelligence techniques with rigorous mathematical modelling for the treatment of uncertainty and the use of artificial intelligence techniques to support Space Traffic Management. This WP will develop key enabling computational tools in support to other WPs, and computational intelligence techniques to quantify the probability of rare, high risk events and correlate spatially and temporally distant events.
THE ART OF DEMISE
On average, over the past decade, a space object above 800 kg has been re-entering every week, e.g. ESA’s GOCE in 2013 and NASA’s UARS in 2011. Most of these objects do not totally demise during atmospheric re-entry. Fragments may survive and reach the ground where they pose a risk to people and things. Space agencies are currently enforcing constraints on the casualty risk for the re-entry event and different re-entry analysis tools have been developed by agencies, industries and research centres to assess this risk. At the same time, new approaches to the design of spacecraft, aimed at increasing the probability of complete demise during re-entry, offer an effective way to meet these constraints. Thus re-entry analyses approaches have to be embedded into the design process since the very beginning, to evaluate the demise probability and casualty risk of different configurations. Currently, most of the efforts are focused on structural design for demise techniques, the improvement of high fidelity and low fidelity models used to predict the aero-thermal performance and thermo-mechanical response of objects, the implementation of efficient uncertainty quantification techniques for a more correct characterisation of the risks, and better understanding and modelling of material properties. Plus, since demisability competes against the survivability of the object during its mission lifetime, design for demise can be seen as a multi-criteria design problem.
The major innovation is the development and implementation of uncertainty based multi-fidelity approaches for the effective and efficient prediction of re-entry demisability and in-orbit survivability performance, compared to current approaches, which are either not really effective nor efficient.
THE DYNAMICS OF CHAOS AND THE DISPOSAL OF SPACE DEBRIS
Recent studies on space debris show the occurrence of a plethora of dynamical phenomena: the overlapping of resonances and the onset of chaos, the existence of a web-like structure of luni-solar resonances, the chaotic variation of the orbital elements, the occurrence of bifurcations of equilibria, the existence of libration regions which lead to excursions in the eccentricity, the chaotic transport in the phase space. These intricate dynamics exists, for examples, in the neighbourhood of GNSS constellations. Despite the fact that the MEO region will be populated by four complete constellations, namely GPS, GLONASS, Galileo and BeiDou, there are no internationally agreed mitigation guidelines, as for LEO and GEO. Thus, it is mandatory to understand the dynamics of these objects. Past approaches were based mainly on deriving single or double-averaged (over short period terms) forms of the equations of motion. Such derivations, dating back to the 60’s, have been refined over the years. These approaches are able to identify critical solutions depending on the particle’s major semi-axis, eccentricity and inclination, as well as the main secular frequencies of motion, as induced by the Earth’s oblateness, as well as luni-solar secular perturbations. However, an important drawback of these analytical approaches is their failure to characterise the chaotic nature of the orbital dynamics close to particular resonances. The study of the chaotic dynamics, as well as its consequences for the long term diffusion of populations of space debris, becomes particularly important for highly inclined orbits, where, luni-solar resonances lead, in particular, to domains of extended resonance overlap. A complete investigation for every Earth orbital region is still incomplete. Furthermore, the use of perturbation theory will allow the study of the separation of nearby orbits in the fully non-linear regime.
The major innovation is in the analysis of the dynamics of space debris, exploiting the regular and chaotic character to devise disposal orbits and in the classification of debris according to suitably defined orbital elements.
THE LINKAGE PROBLEM
The current surveys, like Pan-STARRS, produce an always increasing number of asteroid observations. Several of them are not yet used in the current Orbit Determination (OD) procedures, but they are stored in a database: the isolated tracklet file (ITF), available at the Minor Planet Center website. To the present date the ITF file contains more than 12 million of observations. In view of the next generation asteroid surveys, like LSST or the European Fly-Eye telescope, if more efficient algorithms for OD are not developed, the new observation technologies will surpass the capability of processing the collected data, and this can make questionable the construction of new powerful telescopes. The identification and correlation problem is very relevant in the context of space debris. New algorithms have been recently introduced but they have not been tested yet on large data sets. Another problem is how to patch different dynamics. Patched dynamics is used as a first approximation of more complex models: the solutions of the simpler problem can be used as a starting guess to compute the solutions of the more difficult one, or sometimes they can be used to obtain the qualitative behaviour of the more complex system. Some examples can be found in the field of astrodynamics, and in celestial mechanics, to study close approaches of asteroids with the Earth or the passage through shadow for an Earth satellite.
The major innovation is in the application of new OD methods to large data sets, the creation of a complete OD pipeline which is able to deal with very large database of observations, the inclusion, in the existing OD methods, of perturbation effects, and an alternative definition of region of influence of the Earth.
THE CRITICALITY OF SMALL ASTEROIDS
Our observational knowledge of the asteroid populations is far from being complete. Therefore, asteroid population models are the main source of information about the number of objects in the size-range below a certain threshold. The main belt and NEAs do not represent independent populations, as they are closely connected by evolutionary processes and dynamical transport mechanisms associated to orbital resonances. Recent modeling of the steady state orbital distributions of NEAs extends up to H = 25 mag, i.e. down to a few tens of meters. Despite its great success, this model may still have some shortcomings. As pointed out in, the main belt escape rate does not match very well the observed flux of NEAs. Moreover, for a very small objects it may not be fully appropriate, as it assumes the same orbit distribution for large and small objects in the main belt that may not be necessarily true because for the very small sizes non-gravitational forces become comparatively more important. In order to have the orbital distributions of small NEAs, first the orbit and size distribution of these objects in the main belt is required. However, both distributions are currently poorly known, if known at all. Some available estimations are based only on indirect evidence, such as disruption rate of small main belt asteroids. Last but not least, the asteroid belt is subject to evolutionary processes. So far, no model has included all the important processes affecting asteroid evolution. Besides, new models still have to be successfully tested against known constraints.
The major innovation is in modelling of the population distribution of small size asteroids and their transport process from main belt to NEO including available constraints. The new model will be validated using historical data on the impact rate on Earth, Moon and Mars. From WP5 and WP7 we will derive a criticality index to make decision on future observations, methods of deflection, the possible exploitation and the need to explore each asteroid.
THE MANIPULATION OF NON-COOPERATIVE TARGETS AND ON ORBIT SERVICING
Robotic manipulators are envisaged to be used to capture and repair satellites or remove space debris. This ambitious goal presents unique challenges and requires high levels of autonomy. Two successful technology demonstrators involving unmanned manipulator-equipped satellite have been performed so far: ETS-VII in 1997 and Orbital Express in 2007. Currently, ESA is working on the e.Deorbit mission to demonstrate active debris removal by capturing and removing the satellite Envisat while Airbus DS is developing SpaceTug. The free-floating nature of the satellite-manipulator system must be taken into account for the control of such systems. In 1989 Umetani and Yoshida proposed a resolved rate and acceleration control based on the new Jacobian matrix in generalised form for satellite-manipulator systems. More recently it was presented a control-oriented dynamic modeling framework for constrained multibody systems, which was applied to free-floating satellite-manipulator systems. In addition to autonomous robotics, proximity navigation to a non-cooperative orbiting object is one of the key technologies required to realise on-orbit-servicing (OOS) and active debris removal. The navigation system has to identify the target, estimate its relative motion from far-range to close-range distance, and provide full translational and rotational motion, mass and inertia parameters estimation. For cooperative scenarios the NASDA ETS-VII and DARPA Orbital Express missions already demonstrated OOS. The PRISMA formation flying test-bed performed a wide range of proximity activities, reaching an inter-satellite distance below 1m, using relative GPS metrology and a monocular close-range camera system. The AVANTI experiment showed that the inspection-range region around a noncooperative target can be safely achieved, relying solely on line-of-sight observations from a monocular far-range camera sensor, even in harsh environments. The AFRL XSS (Experimental Spacecraft System) program, with the XSS-10 and 11 endeavours, performed on-orbit inspection of noncooperative rocket bodies with high levels of autonomy. Current and future projects mainly address noncooperative scenarios, to apply OOS to satellites not designed to be serviced or damaged (e.g., NASA Restore-L), as well as active removal of debris (e.g., e.Deorbit, RemoveDEBRIS, CleanSpace One). For such projects close-range relative navigation systems exploiting sensor data-fusion, in line with the H2020-PERASPERA programme, is a key enabling technology. Multi-sensors’ close-range relative navigation systems demand for the development of dedicated algorithms, with special focus on embedded solutions for Low Earth Orbit.
The major innovation is the development of autonomous navigation and control solutions for proximity operations and the manipulation of debris and satellites in view of future on orbit servicing and their experimental validation in a laboratory environment.
THE EXPLORATION AND EXPLOITATION OF COMETS AND ASTEROIDS
The exploration of minor bodies can contribute substantially to our knowledge of the solar system and to plan for the exploitation and manipulation of comets and asteroids. A potential low-cost exploration and prospection technology is offered by CubeSats. A variety of CubeSat platforms are available for applications around the Earth. Proposals exist to extend these capabilities to interplanetary space including asteroids. This extension requires the development of guidance, navigation and control (GNC), power and propulsion systems for deep space missions that account for the limited resources on board small spacecraft. A further key exploration technology is offered by landers. Landing on a rotating object in microgravity is a challenging task because of the lack of a priori precise information. The recent example of the ESAs cornerstone mission Rosetta deploying the lander Philae on a comet in 2014 demonstrates that key developments are required in GNC and system design in combination with prior prospection of the landing site to guarantee a successful landing. Once new information from prospection is available, the challenge is to use this information to improve our ability to manipulate the rotational and orbital dynamics of minor bodies.
The major innovation involves formulating novel methodologies to approach, orbiting, and landing on minor bodies by using new space platform characterized by limited resources in terms of mass and power. Furthermore, this WP will contribute, together with WP5, to create a workbook assigning a criticality index to each asteroid to assess the possibility to explore, exploit or deflect that minor body.
SPACE TRAFFIC MANAGEMENT AND THE RESILIENT SPACE ENVIRONMENT
The publication of the IAA Cosmic Study on STM 2006 defined Space traffic management (STM) as: “the set of technical and regulatory provisions for promoting safe access into outer space, operations in outer space and return from outer space to Earth free from physical or radio-frequency interference.” This lead to the subsequent creation of the Space Debris Mitigation Guidelines of UNCOPUOS of 2008 and the ISO 24113:2011 Space Systems – Space Debris Mitigation standards. These standards are necessary but not sufficient to reduce the risk of collisions. A viable STM program faces a great barrier caused by the ever-increasing number and variety of orbiting objects ranging in size from a few microns to several meters and the planned future large constellations. In addition, most debris objects cannot be tracked and motion cannot be accurately measured or simulated. The current two line element (TLE) sets and associated processes used for disseminating data about space debris are not adequate for precision conjunction analysis or accurate long term prediction. In order to ensure that the space environment is resilient to anomalies and catastrophic events and its exploitation is sustainable a comprehensive STM system must be implemented, integrating improved space situational awareness (SSA), long term orbit prediction, debris removal, re-entry risk assessment and new tools to support space operators.
The major innovation is in the development of the concept of Space Traffic Management and the related technologies required to detect and avoid collisions increase of resilience of the space environment and implement preventive actions including active and passive disposal with particular focus on future large constellations.