[Final] Results 2022 CORE Call

Many of the recent scientific works in signal processing have been increasingly influenced by the most complex and least explored signal processing system – the human nervous system. Cognition refers to the process through which humans and animals sense and interact with their environment and cognitive radar signal processing aims to parallels to neurobiological cognition to learn from the sensed environment and act accordingly e.g., better focus the sensing on particular areas of interest while nulling others. While the radar gets data on the entire surrounding, it’s the cognitive signal processing that brings about the decision on how to act after inferring from the data.

Similarly, metacognition is a well-studied concept in both neurobiology and educational psychology, and can be termed as a higher order thinking which can be summarized as learning about learning or knowing about knowing. It is pertinent to remark the importance of metacognition as separate from plain cognition. Since the cognitive cycle is a closed loop system, without any provision of altering any of the steps once the cycle has kicked in a operational system. This leads to an inherent inflexibility of the system to adapt to drastic change in the channel conditions, change of engineering modules, or the operating objective or all of these. Hence, the radar must include multiple strategies with their own cognitive cycles. The selection of the appropriate strategy is handled by metacognition.

Consider the case of vehicles equipped with Automatic Emergency Braking (AEB) systems. It has been reported that many situations like bridges, railroad tracks and parking garages can trick the vehicles into breaking by making them believe that they are about to crash. Even steam coming out of underground tunnels and ducts in large cities can trigger the system, apparently. A standard operational cognitive radar cannot fully “learn” this situation. In a metacognitive radar, on the other hand, this flaw can be monitored and a learning strategy for this situation devised by selecting the optimization objective and solution. Furthermore, the learned information can be transferred in a connected networked so that other cars do not make the same mistake.

The project formalises such a learning for the different tasks of radar (detection, estimation, classification and tracking) with the transmit waveforms and receiver algorithms as the means of implementing the learned rules. The project aims to create a system that acts in optimal manner based on various possible strategies that can be learned from the environment.

Software Testing forms one of the key quality assurance methods and is a vital ingredient of the modern software development practices. Deciding if a program execution corresponds to a correct/incorrect behaviour is currently a manual task. To this end, MeMoRIA will automatically generate a set of (metamorphic) properties that distinguishes good from bad software behaviours. Hence, MeMoRIA aims at investigating and solving this real-world industry relevant problem, and contributing with the adoption of automated testing.

Currently, there are a lot of investments and research to build a full-scale quantum computer. Advances in this realm have two drastically different impacts: 1) Positive: The way a quantum computer doing computations is fundamentally different from a classical computer and this will help to solve some problems that are intractable for a classical computer and to open up new technological applications and opportunities. 2) Negative: A full-scale quantum computer can be used to attack classical cryptographic constructions. This compromises the security and privacy of our digital communication.

A proof (argument) system is a cryptographic protocol that will be affected negatively by a quantum hacker. To illustrate a proof (argument) system, let say one wants to prove that she/he possesses a password to log in to an online bank account without revealing her/his password. This task can be done by a proof (argument) system. Considering many real-world applications of succinct proofs, e.g. electronic voting systems and cryptocurrencies, it is essential to investigate the effect of a quantum adversary to the proof (argument) systems. In case of a security break, it is necessary to develop proof (argument) systems secure against a quantum hacker. In this project, we conduct this investigation and will develop quantum secure proof (argument) systems. Beside the scientific contribution of our project, we expect a post-quantum secure cash system as an output of the project. In addition, our result can be used in any application that uses a proof (argument) systems as a building block to guarantee the (post-) quantum security.

Proliferation of radar sensing in novel applications imposes stringent requirements on performance while exploiting only a limited number of resources and with a limited amount of power. Further, the scenarios being considered are dynamic requiring the systems to be flexible and be adaptable with negligible or no human intervention. For example, indoor sensing require detailed information about the interiors, separating the different closely objects, irrespective of the lighting conditions while having constraints on placement, number of sensors and potential interference to mobile communications. Further, the scenarios are dynamic, with the number of objects and people varying in number, orientation etc. The radar systems need to distinguish between different closely spaced objects, recognize the number, posture, orientation without the semantic information made available by a camera. While radar sensing offers a cost-effective and small form-factor solution, current technology needs to be improves to offer the required performance and flexibility. Delving deeper, the radar system comprises of the Radio Frequency (analogue) components and digital processing. Traditionally, the digital processing offers flexibility exploiting the available information, while the radio frequency parts are kept fixed. To enable the radar systems to meet the earlier said demands, a synergetic adaptation of the digital processing and the radio frequency parts are mandated. In this context, in addition to the advances in digital radar signal processing, reconfigurability of the Radio Frequency (analogue) components offers additional degrees of freedom to overcome the shortcomings of radar.

Thus far, the radar systems had Radio Frequency components separate from the digital processing. Integrated chipsets incorporating the key analogue processing and antennas together with digital processing are being increasingly considered for the newer bands to be used for sensing and communications. The use of novel materials for such chipsets, known as metamaterials, enable the reconfigurability of analogue components at low-cost and low power. The properties of these metamaterials can be controlled rendering their intelligent operation in a manner required by the application. This project considers the use of such intelligent surfaces for radar sensing, addresses the challenges of their incorporation (handling flexibility needs to be aced as in other fields) and demonstrate the gains for indoor imaging and monitoring. The project also aims to validate the gains through in-lab activities lending the project outcomes closer to reality.

In the last 30 years, the world has become densely connected. Most ICT systems address a complicated network of users, vastly distributed over geographical locations and cultural contexts. What is more, ICT services are usually done by people, with people, and for people. The intensive human involvement makes them hard to describe and analyse with the standard tools of formal verification.

Voting and elections are good examples of services that are difficult to specify, hard to verify, and extremely important to the society. The democratic societies are currently facing a number of serious threats. Electoral fraud, manipulation of voters, fake news and disinformation are used to swing the outcome of elections and severely undermine the voters’ trust. If democracy is to be effective, it is essential to assess and mitigate those threats. In the course of our previous projects, we have used ideas from game theory, multi-agent systems, and theory of socio-technical systems towards the goal, with considerable success. It is now time to move on to richer formalizations that allow for reasoning about more realistic models of voters and attackers, and more mathematical analysis of their possible behaviour. To this end, we will extend our techniques to quantitative analysis of voting, and combine them with the recent advances in probabilistic verification.

We will also show that the methodology can be applied well beyond voting procedures, for example to better synthesize strategies for small autonomous rockets navigating in an uncertain environment.

As more activities are performed online, many of them related to government services, there are expectations that elections should follow the trend. In recent years, governments in conjunction with the industry started running pilots to asses the feasibility and security of electronic voting, some prominent examples being in Norway, Estonia and Switzerland. In fact, internet voting is already used in the real world for elections with relatively lower stakes than political elections, e.g. for memberships in international organisations and leadership positions in universities. However, secure electronic voting is a more difficult problem than, say, secure online banking or secure messaging. In addition to external attackers and network threats, electronic voting is subject to threats from parties and infrastructure running the election: e.g. voting platforms and servers, parties responsible for voter registration, parties collecting or tallying the ballots, etc. The adversary may also coerce voters, or be a powerful state actor investing significant resources towards interfering in another country’s election process.

Considering this complex threat landscape, we need systematic methods to ensure future electronic voting systems are indeed secure. We need to reason in advance about all possible attacker actions and rigorously conclude whether a desired property is satisfied or not. Automated tool support is an essential element in this task. The main goal of our project is to develop methods that allow realistic, rigorous and automated formal verification of security for electronic voting protocols, with a focus on vote privacy. We will address several challenges related to this objective: we will develop precise specifications for the properties that we want to achieve, for the protocols implementing them, and for the possible actions of the adversary. Furthermore, we will also develop techniques that improve automated verification such that we can have high assurance automated guarantees of security for electronic voting systems.

Logs are special text files produced by computing systems: you can find them on your mobile phone, on your laptop, and on large enterprise information systems. They are used by software developers to record important events of applications and of the computer, such as the arrival of a network message, the change of a setting, or an unexpected situation (like an error). Logs are also used by special type of programs – called log-driven analyses – to assure the reliability of the computing system, for example to detect an anomaly in the system.
The effectiveness and the efficiency of log-driven analyses depend on the quality of the logs, which can be summarized as the level of information about the system execution contained in a log. This quality attribute depends on the way developers have written the code producing the information recorded in the logs, which we call the “logging code”. Unfortunately, developers can make mistakes when programming (for example, because they have to quickly release a new version of an application) or can neglect to update the logging code when they change other parts of an application. As a result, the logging code is not aligned anymore with the core functionality, leading to low-quality logging code that produces low-quality logs.

To address the above issue, LOGODOR will leverage the concept of “smells” (as in “code smells”). In the same way “code smells” represent any characteristic in the source code of a program that possibly indicates a deeper problem, we will define the novel notion of “log smells””, i.e., any characteristic in the logs of a system that possibly indicate a deeper problem related to logging. The notion of log smells introduced by LOGODOR will cover both code-level log smells, i.e., smells specific to the logging code, and message-level log smells, i.e., smells related to the log messages and their textual content.

Based on the notion of log smells, LOGODOR aims to develop automated and scalable approaches for detecting and removing a wide range of log smells, by focusing on four specific activities: (1) development of a comprehensive catalogue of log smells, (2) automated detection and removal of code-level log smells by enhancing the logging code, (3) automated detection and removal (or mitigation) of message-level log smells by cleaning log messages when accessing the logging code is not feasible or possible, and (4) evaluation of the developed techniques and tools in terms of their effectiveness, efficiency, and benefits.

Being compliant with regulations in general, and the GDPR in particular, is required from almost all companies. Nevertheless, the legal language itself is usually complex and requires the intervention of legal experts. This problem has prevented AI from being used efficiently to help solving such compliance issues in applications intended for non-legal experts. The objective of the research is to simplify the access to the legal knowledge via an artificial intelligence system, able to explain abstract legal concepts with easy to understand examples. Let us consider the problem of digital marketing in the context of GDPR. The regulation makes references to the concept of legitimate interest, yet a person running a small company may struggle understanding what does it mean, to have a legitimate interest. The proposed solution, guides the user in their journey towards compliance, but at the same time, uses state-of-the-art artificial intelligence in order to explain the legal concepts via relevant examples.

Computer systems in banks, insurance companies, but also in autonomous vehicles or satellites remain worthwhile targets for cyberattacks and must be protected to withstand, tolerate and safely operate through such attacks. Unfortunately, due the growing complexity of systems, adversaries have gained an advantange, which requires us, defending these systems, to anticipate that some attacks might be successful. Fortunately, resilience techniques, such as triplicating the actual computer system and protocols for seeking a majority of matching responses, exists that allow systems to continue to work correctly, even if cyberattacks have been partially successful. But for these techniques to be applicable, they have to match the system’s structure, in particular how the individual components of such a system interact.

The resilience techniques we have developed so far are limited in these interaction patterns and developing new techniques for more elaborate patterns remains a difficult and error prone task. In particular, tools that support us in developing such protocols by allowing us to check whether what we have constructed is correct, work
only on the final protocols and they require a rare expertise to be used.

In the FM-CReST project, researchers from CISPA, Germany, and from the SnT of University of Luxembourg have joined forces to research a new class of highly automated, easy-to-use tools to assist developers in constructing provably correct resilience protocols. We do so by co-designing protocols for systems with complicated interaction patterns, while observing this protocol construction and developing the tools needed to simplify such developments in the future.

Emergent quantum technologies such as quantum computation have a unique potential for industrial and service transformation. Their commercialization has already had a major impact on higher education, research, and academia. A particularly advanced form is that of adiabatic quantum computation that relies on quantum annealing. It provides a way of solving complex optimization problems. Its applications are manifold and include portfolio optimization, forecasting election polls, prediction of financial crashes, the evolution of financial derivatives, optimization of investment, image acquisition, traffic flow, and nurse scheduling, among others. For example, several major banks have started exploring quantum optimization in their business model. Devices performing quantum annealing are available in laboratories across the world and have been commercialized by companies such as D-Wave. Cloud quantum computing facilities online such as those created by IBM can also be applied to quantum annealing. Deepening the understanding of quantum annealing as a physical nonequilibrium process is fundamental to further advance this technology. AQCQNET aims at achieving this goal by combining tools of network theory, quantum many-body physics, and nonequilibrium scaling theory. The project Central to quantum annealing is the mapping of the optimization problem to that of finding the lowest energy configuration of a physical system with many discrete degrees of freedom known as spins. The interactions between spins can be described as a graph and characterized using the tools of network theory. AQCQNET aims at relating the performance of quantum annealing to solve optimization problems with the underlying network topology of the interacting spins in which the problem is encoded. By doing so, it aims at introducing novel algorithms for quantum annealing using local protocols exploiting the network topology. By benchmarking their performance against classical methods, AQCQNET aims at identifying scenarios in which the quantum properties provide an advantage. Overall, the project focuses on fundamental phenomena defining the functioning and performance of devices for quantum computing and quantum technologies.

Most materials in nature exhibit phase transitions between different phases of matter. Water can be found in ice, liquid, or vapor form. A piece of iron can acquire or loss its magnetic properties as a function of temperature. Understanding the dynamics across phase transitions is crucial for many applications ranging from the exploration of novel materials to quantum computation. Even when a material is driven slowly across a phase transition, excitations appear in it. Indeed, the crossing of so-called continuous phase transitions results in topological defects, a kind of excitation that is particularly robust and stable. As a result, topological defects govern many of the physical properties of a given phase of matter. For example, the observation of vortices can inform about the superfluid or superconducting character of a given sample. Topological defects in solids can seed cracks leading to material failure. In addition, topological defects shed new light on the fundamental laws of physics such as the occurrence of the Higgs mechanism, the study of which was recognized by the Nobel Prize.

A theory known as the Kibble-Zurek mechanism (KZM) predicts that the density of topological defects created in the course of a phase transition follows a universal law, increasing with the rate at which the transition is crossed. The KZM formula governing this increase can be calculated using solely the equilibrium properties of a system, providing a precious bridge between in and out of equilibrium phenomena. Moreover, such behavior is shared by many materials alike, grouped in so-called universality classes, making the results universal and of broad applicability. Since its conception by Kibble and Zurek, the KZM has been the subject of intense study sustained over decades of research. It was first explored in a cosmological setting and has by now been applied to colloids, liquid crystals, superfluids, superconductors, multiferroics, and many other materials.

However, findings to date have been focused on the characterization of the mean density and quantities depending upon it. The Project BeyondKZM aims at providing a paradigm shift, by opening a new research focus on the statistical properties of topological defects generated across phase transitions. The working hypothesis, motivated by preliminary findings, is that not only the density of topological defects is universal, but the fluctuations governing the distribution of the number of topological defects and their location are also universal. A crucial lesson from twentieth-century physics is that many phenomena are not dictated by average properties but are governed instead by fluctuations. The Project “BeyondKZM” aims at exploring and describing such new physics by combining two disparate fields of research, namely, the theory of phase transition dynamics (KZM) and a branch of mathematics dealing with random processes. In doing so, Physics unveiled by the project BeyondKZM will be directly applicable to the plethora of systems in which topological defects can be imaged. We thus expect this project to motivate a new generation of experiments. Overall, this project will advance the frontiers of nonequilibrium phenomena and the prediction of material properties governed by topological defects.”

One of the most ambitious goals in structural materials is to achieve exceptional mechanical properties, with an ideal combination of light weight, high strength, and high toughness.For many biomaterials, strategies to achieve the advantageous unification between structure and mechanical properties have been through millions of years of evolution, as in the case of nacre, one good example of a material that combines high strength, high stiffness, and high toughness. Inspired by nature strategies to increase toughness, BioInter proposes to engineer the interface of composites by creating alternate sections around the fibre and along the fibre length of high and low adhesion with the matrix using micro-plasma surface deposition and activation, promoting intermittent sections of high and low shear strength during the process of debonding. The strong regions ensure that the filament strength is taken, while the weak areas weaken the running crack, creating a complex pattern for crack evolution.

Optical spectroscopy (light absorption and light emission) is an important method to gain information about the microscopic structure of materials. The link between experimental data and a full understanding of the atomic structure of a material is usually made via theoretical calculations. The development of computational methods for the description of the complex optical properties of materials is thus an important ingredient for the characterization and the design of novel materials with tailored optical properties. These calculations are by now rather standard for materials with direct band gap where a photon (light particle) can be directly absorbed or emitted. However, for the large class of materials with indirect band gap, where photon absorption or emission is accompanied by excitation of lattice vibrations (phonons), a reliable and efficient calculation method is still missing. We will address in particular the class of layered materials and single-layer 2D materials, where strong excitonic effects (states where the excited electron in the conduction band is bound to the hole left behind in the valence band) are often occurring. The combination of strong excitonic effects and indirect band gaps requires the development of a new computational approach where absorption and emission of photons is coupled to the excitation of lattice vibrations (phonons). In this project, we will develop the methodology and implement it in an efficient computer code. We will study some seemingly simple prototype materials (such as diamond and hexagonal boron nitride) where the luminescence spectra contain complex features that are still not understood. The final goal is to calculate phonon-assisted optical spectra routinely of a large number of materials in an automatized (“high-throughput”) way, creating a reliable data base of optical properties of materials.

The law of physics pose certain limitations to the efficiency of solar cells. The efficiency of a solar cell is the share of solar energy falling on the cell that is transformed into electricity. For example the laws of thermodynamics limit the efficiency of a single solar cell to about 33%. Just for completeness: combining different solar cells in tandem cells or multi-junction solar cells allows to overcome this limit. The best solar cell so far has an efficiency of more than 47%.

These limitations are well understood for ideal solar cells. However in real solar cells many more limitation appear. Some of them can be overcome by optimisation. Other are more fundamental and due to the very nature of the structures and materials used in the solar cell.

Accepted models exist to describe these fundamental limitations of real solar cells. However, they have only partially been checked experimentally. In this project we aim to check these models fully experimentally and to understand better which limitations are fundamental and how the different aspects of solar cell design interact with each other.

This endeavor will allow us to define new design rules for solar cells and to improve them.

Non-equilibrium physics in general owes a lot to the fascinating phenomenon of fluid turbulence. Both the natural and artificial worlds are full of examples of how turbulence plays a significant role in our lives, ranging from small and large-scale climate effects, mixing of dyes in solvents, to aerodynamics of cars and jet engines, to name a few. Fluid turbulence has traditionally sat at the interdisciplinary junction of fluid dynamics and statistical mechanics, where statistical properties of fluid flow are studied under the applications of external drive.

Fluid Turbulence has many intriguing properties: for example, the symmetry breaking at small length scales (close to boundaries), only to be restored once again at large length scales, away from boundaries; or the self-similarity of flows across length scales. These and other interesting phenomena associated with fluid turbulence remained the core of non-equilibrium physics until a couple of decades ago, when studies on ”active matter’ exploded onto the research scene. Usually active matter refers to ‘self-sustained’ flows, where the systems spontaneously take-in and dissipate energy to and from their surroundings for their mechanical motion. Examples include naturally occurring bacteria or sperm swarms, and even artificial counterparts, such as Janus particles. State of the art non-equilibrium physics aims to answer one question : Can we understand fundamental principles of life through a statistical physics approach? It is in this context that the discovery of turbulent-like flows in such living and active systems some years back, led to the creation of the field of ‘Active Turbulence’, which deals with characterizing features of turbulence for active systems. Our CORE-junior project fits exactly to this theme.

Our aim is two-fold: We are proposing a novel theoretical approach for characterizing universal properties of Active Turbulence , and thereby exploiting such models to quantify the associated energy dissipation in order to design efficient active engines. The study of engines lies at the heart of equilibrium thermodynamics. It is imperative now that ideas such as power fluctuations and efficiency are considered for stochastic active processes, to lead to a fundamental understanding of non-equilibrium physics in general. We propose to use analytical and numerical techniques which have proved to be very successful in studies of turbulence of passive systems, and expect our new results to be verifiable in experiments on bio-physical systems.

The VERITAS project is about using lasers, which are high power light beams, in a specific configuration for the manipulation of motion of charged particles (ions). Due to this light-matter interaction, it is possible to create different phenomena: These particles can for example be accelerated by the lasers so that they achieve higher velocities. On the other hand, deceleration of the particles by the use of the lasers can also be achieved so that the particles are slowed down to a standstill. A third type of interaction that can envisioned is that a group of ions having an initially large range of velocities, after interacting with the specially structured laser light, can be brought down to a smaller range of velocities, leading towards all ions having only one single velocity in the end.

All of these aspects will provide scientists novel ways to precisely handle charged particles and will lead to numerous improvements for existing scientific and industrial instruments. For example, it will be possible to perform microscopy imaging of smaller material features with even higher resolution, revealing more details about the structure of these materials. Furthermore, studies about the chemical composition of materials will be improved in order to understand even better of what these materials are made of at a smaller scale.

In general, these achieved instrumental optimisations will lead to a globally better understanding and characterization of materials involved in various relevant technologies of today. For example, the functionality of solar panels, batteries, fuel cells or other electronic devices can in this way be even better understood and improved. This will lead at long term to a reduced power consumption of electronic devices, improve the production efficiency of renewable energy devices or lead to more sustainable solutions during the production phase of these devices. Hence, the VERITAS project can at the end have a considerable impact on energy and environmental related topics, having therefore an immediate impact on the society of tomorrow.

Non-diffusive fluctuations are ubiquitous in nature and provide a crucial link to unravelling new dynamical phenomena. They have deep consequences in various aspects of physics, biology and engineering. For example, non-diffusive fluctuations explain how the thermal conductivity at nanoscale systems is enhanced or the birds optimise resources when non-diffusively foraging for food. The project aims at understanding analytically tractable microscopic classical and quantum models which shows non-diffusive behaviour. WP1 and WP2 relate to the study of non-diffusive heat fluctuation in classical many-body systems in the interdisciplinary aspects of the stochastic process and Hamiltonian systems which are externally driven. In WP3 we will extend the analysis to the quantum system being randomly driven which characterises the dynamic fluctuations of energy in the heating dynamics of the system. In summary, the project will improve our understanding of non-diffusive dynamical fluctuations arising in classical and quantum systems. The project will add to the knowledge in specific cases for the question of the microscopic origin of non-diffusive fluctuations and a relevant macroscopic framework using hydrodynamics or fractional operators to describe it.

Electronics are everywhere, from our cars to our kitchens – and unfortunately, electronic waste (e-waste) is as well. 53.6 million tons was produced worldwide in 2019, up 21% from 2014, and 74 million tons is predicted to be produced by 2030. This has substantial environmental and health consequences, especially since so much of what is made is not recycled or recovered. A typical printed circuit board is mostly made up of glass-fiber reinforced epoxy resin – a three dimensional network of large molecules linked to one another irreversibly via strong chemical bonds. Such materials are extremely robust and perform well, but they are often composed of compounds that are non-renewable, harmful and/or persistent, a problem exacerbated by the fact that they cannot be melted or dissolved making recycling impossible. Instead, they are typically incinerated or landfilled, which can lead to the release of harmful compounds and cause harm to human health and the environment.

With this as the backdrop, we continue to witness increasing levels of functionality and performance when it comes to electronics, with one of the most recent developments being flexible systems. To emphasize how fast things move, the first foldable smartphone was demonstrated in 2006. The first commercially available model hit the market in 2018, and by 2021, 7.1 million units were shipped (up 264% from 2020). While this constitutes only a small fraction of the total smartphone market, this brief history highlights the growth of flexible electronics. If things are not done differently, such growth will not only lead to amazing new devices, but also to significant amounts of additional e-waste.

This project aims to address these issues by demonstrating the possibility to integrate two highly complementary technologies: flexible transient electronics and vitrimer networks. The materials that make up flexible devices (conductors, semiconductors and insulators) are carefully formulated and selected such that, in contrast with their conventional analogs, they can be efficiently dissolved in minimally harmful media (e.g. water, aqueous solutions, low-hazard organic solvents). Vitrimers are three dimensional networks of large molecules linked to one another by chemical bonds, but with a twist: these chemical bonds are strong and stable when the materials are in use, but when they are heated above their use temperatures, these same bonds become dynamic, mobile and reactive – allowing us to reuse, repair, recycle and recover these materials in ways that were formerly impossible with such networks. Given strong experience in both areas, our teams at the University of Bozen-Bolzano (UniBz) (South Tyrol, Italy) and the Luxembourg Institute of Science and Technology (LIST) (Luxembourg) aim to combine these technologies to produce next generation flexible electronics that are designed not only with performance in mind, but with the end-of-life accounted for as well. In this way, we avoid the generation of new e-waste in favor of a future where such systems are designed with greater sustainability for ourselves and our environment.

Skyrmions are localized magnetic vortices in ferromagnetic materials. Their properties have been the subject of intense investigation over the past decade, mainly thanks to their potential use in next-generation magnetic storage devices.

Typical skyrmion sizes are between nanometers and micrometers. However, at the smallest length scales, quantum effects should be taken into account. Indeed, recent years have witnessed a growing interest in the quantum-mechanical properties of skyrmions.

In this project, we will therefore investigate skyrmions in the quantum regime. We will study suitable quantum-mechanical model systems using a combination of numerical and analytical approaches, which will allow us to obtain a comprehensive theory of skyrmions ranging from the quantum regime to the classical regime. Moreover, classical continuum calculations will serve as a gauge to the quantum calculations. We will use this to calculate experimentally observable quantities, such as the spin structure factor, which can be determined in neutron-scattering experiments. Hence, our project will propose materials for the existence of quantum skyrmions, study their unique quantum-mechanical properties, and propose ways to detect them experimentally.

DNA and proteins are the molecules that rule life. Reading their sequential composition has allowed us to understand their role in life to tackle health issues and make new products based on natural components that have become commodities of modern life.

Specifically, DNA sequencing has allowed the irruption of new medicines like the development of RNA vaccines or the new personalised therapies based in genetics.

Meanwhile, proteomics is lagging behind because the methods to “read” the composition of proteins are still cumbersome, expensive and require too many molecules compared to DNA sequencing. However, proteins are the real architects of life, thus, many prospect applications in precision medicines and protein engineering for green industries demand improving the performance of protein analyses while working with fewer molecules, increasing the throughput and at a lower costs.

In this project we approach this need with a new design of a lab on a chip technology that uses a novel electrochemical methods to disassemble the proteins in an ordered array of sensitive sensors. The constituent amino acids become immobilised into very reactive sensors that detect their fingerprints employing very few molecules to reconstruct the protein sequence.

This technology works with minimum amounts of samples and uses electronic components that can be mass produced at low costs to democratise the access to protein sequencing.

The Global Navigation Satellite System and Earth Observation market had revenues of around €200 billion in 2021 and is expected to reach €500 billion by 2031. Many satellites from different vendors are carrying out their work in space. Additionally, all the data they acquire (images, recordings, analysis, raw data, etc.) may be offered by different providers. The huge amount of collected data requires appropriate storage, classification, and further processing for final use. Data lakes are the place where any data can be stored. The relative initial maturity of this market shows many aspects that can be dramatically improved to achieve much better end benefits – and for all parties to the transaction.

In the current context, many customers might be interested in combining data of two or more satellite data providers, e.g., to obtain more frequent data or to merge data of several types. To this end, the customer needs to request data separately from each provider. Such a scenario is sub-optimal since it requires the customer to know the different providers. A data marketplace is one promising solution for this problem. It is an online platform where users can sell or buy data from different sources.

The SpacE data bRokEriNg optImizaTion sYstem (SERENITY) project, aims to propose a novel type of marketplace where space data providers and consumers interact through a data broker that relies on a data lake storage. To this end, the SERENITY project will develop novel artificial intelligence and data storage approaches that will permit to tackle objectives like maximising providers’ profits, optimising data latency and minimising total purchase cost for customers.

The beyond 5G (B5G) networks are designed to serve both human-oriented services, e.g., 4K video streaming and augmented reality, and cyber and machine-type applications, e.g., smart cities, automotive and industrial IoTs. These different and sometime conflicting QoS requirements from vertical services, e.g., ultra-high data rate and stringent latency constraints, pose unprecedented challenges in B5G network design and optimization under limited bandwidth and power resources. To tackle these challenges in addition to enhancing ubiquitous coverage, B5G systems are built upon multi-layer ground-air-space network architecture to efficiently harness the spatial densification along with advanced PHY signal transmission techniques. The multi-layer architecture, in one hand, provides the agility and flexibility in meeting diverse QoSs from heterogeneous services, on the other hand, extremely increases the system dynamics and complexity. Conventional model-based methods seem unfortunately insufficient to cope with such stringent requirements and sheer complexity of B5G, which is mainly due to i) lack of precise modelling for highly complex and heterogeneous systems and ii) limited time to obtain the system-wise state information as well as the real-time optimum solution.
In this project, we investigate the data-driven paradigm for B5G design and optimization by proposing novel risk-aware and distributed DRL-based resource management to overcome the limitations of the conventional RL in the B5G context. One key issue is how to determine and present the risk factor in the learning model to simultaneously accelerate the learning performance and minimize the chance of violating the QoS. Although there exist risk-aware RL in the literature, they evaluate the risk by the universal metric as the variance of the reward function, which ignores the inherent network topology and the eventual goal of the tasks.

Luxembourg has been successful in creating in a very short time, a space ecosystem with over 60 startups, addressing diverse short-term to long-term business cases. This has positioned Luxembourg as a global hotspot for established and emerging actors from the international space economy such as iSpace, Redwire, Bradford Space, etc.

However, an architecture (high-level design) of Luxembourg’s space program is currently missing, which describes which space missions are conducted in which order and which space systems, technologies, and actors exist, will exist, or should be developed. A space program architecture is a high-level description of a space program. It would support steering the strategy of the Luxembourg space ecosystem but also guiding companies where to position themselves within that ecosystem. While a high-level space roadmap exists for Luxembourg, it is rather schematic and does not provide much detail on how the underlying actors, technologies, space missions, and their sequencing are linked to address key economic and societal factors.

A space program in the following is understood as a combination of top-level strategic goals with lower-level combinations of space missions, hardware, and technology. A space program goes beyond an individual space mission (e.g. an Earth observation satellite mission) and usually combines multiple missions in a sequence along with the used systems and technologies. It links strategic objectives (political, economic, societal, environmental, etc.), which are usually high-level (often national) with means to achieve them (systems, technologies, etc.).
Representing a single mission is complex but has been treated in the existing literature abundantly. However, the sequencing of missions in time, with respect to the underlying systems and availability of technologies in an ecosystem is much more complex but has great strategic value in supporting the decision-making process of key stakeholders such as the Luxembourg Space Agency and various ministries (Ministry of Economy and Ministry of Defense).

Designing a space program is a highly collaborative activity, bringing together diverse stakeholders and a key benefit lies in developing a shared understanding during the process. Contrary to traditional space program architectures which are paper-based, recent advances in this field and in technology roadmapping indicate that a model-based approach to defining space program architectures seems to be promising. Instead of a textual document, a model of the space program architecture allows to integrate and link diverse data and to run simulations to analyze the impact of one or multiple space program architectures. Such a model-based approach enables to design and evaluate space program architectures during workshops with stakeholders, analyzing the impact of changes (political changes, technological changes, etc.) and opens the door to a more thorough approach to strategic decision-making, also in domains beyond space, e.g. energy, mobility, healthcare.

How can a space program architecture be collaboratively designed, which integrates strategic goals, technologies, space missions, actors, etc. in order to facilitate decision-making?
SAMP will address the above research question by proposing a number of computer models and approaches that will allow decision-makers (e.g. engineers and managers from industry, government officials, representatives of local communities) to collaboratively design a space program architecture.

These expected outputs will support decision-makers and stakeholders to explore different space program options and the potential impact of changes to the program and with respect to external, e.g. political events. The method presented in SAMP will provide a basis on which future methods for other domains such as healthcare, energy, and mobility could be based, supporting strategic decision-making.

Robotics is considered the Holy Grail that has promised to change the world as we currently know it, becoming a central part of our lives in a fully robotized society. Service robots assist humans at work in non-industrial settings or in the home, needing to operate in complex dynamic unstructured environments. Robots need to continuously acquire a complete situational awareness, by perceiving the environment within time and space, comprehending its meaning, and projecting it in the future, to enable intelligent decision-making and autonomous tasks execution.

The project DEUS (which means God in Latin), focuses on the development of a novel robotic situational awareness as it is an essential missing CORE capability on autonomous and intelligent robots which is blocking their further development. We will provide the robots with an understanding of the situation with multiple levels of abstraction such as geometric (e.g. shape of the objects), semantic (e.g. type of the objects), or relational (e.g. relationships between the objects). This understanding will not be limited to the current situation (i.e. what is the situation that the robot perceives at the present moment), but it will create a long-term situational understanding (i.e. what is going on around the robot) by integrating the perceived measurements from multiple sensor sources with past observations and background knowledge.

We will develop a robotic situational awareness by combining adapted versions of novel machine learning based techniques with extended versions of traditional optimization based algorithms, to get the best out of each of them. Our research will focus on two different aspects. On the first hand, we will research how to acquire a deep long-term situational understanding from the sensor measurements. On the other hand, we will focus on the acquisition of semantic-relational situational understanding by reasoning on the knowledge that the robot previously acquired.

Our research will be experimentally validated using our ready-to-use robotic platforms and we will prove its technological relevance by demonstrating a proof-of-concept in two real-world use-cases we have selected, namely a UAS for surveillance of critical infrastructure, and a walking robot for inspection of construction sites.

Our expected open access publications, our datasets and one PhD thesis will generate high scientific impact. We also expect that the DEUS project will bring high long-term socio-economical impact, by providing the foundations of an essential component of the robotic artificial intelligence that will enable a limitless number of potential applications for service robots.

While Unmanned Aerial Vehicles (UAVs) were initially applied as flying sensors, recent years have also seen a rapidly increasing interest in Aerial Manipulators (AM) by attaching a robotic arm and gripper to a UAV. AMs dramatically increase the spectrum of UAV applications, but especially the use of rigid manipulators is difficult: they are safety-critical especially when interacting with humans, difficult to control and have problems when gripping of irregular shaped or unknown objects is required. In order to cope with these challenges, there is a very recent and novel interest in using soft grippers and soft robotic arms on UAVs, forming Soft Areal Manipulators (SAM). However, feedback control of soft grippers and arms in combination with UAV flight control is also a very challenging task and requires the integration of sensors into these soft-bodied systems. In general soft robotics, there is a main focus on mechanical sensing using strain sensors, where many different sensing principles such as resistance, capacitance, piezoelectricity, optical fibers, or magnetics are proposed. However, the main problems of these sensors are their wiring and integration on or in the soft structure.

An alternative are optical sensors based on vision, since they provide easier fabrication and less wiring combined with relatively high spatial precision. Vision-based tactile sensors usually use a deformable body as the sensing medium and apply a camera to capture the deformation of the medium. Marker patterns may be printed on or in the sensing medium, and the motion of the markers are tracked from the embedded camera. However, all these sensors only detect deformations of surfaces with complex computer vision but do not directly measure strain or forces. Recent research in soft matter physics on Cholesteric Liquid Crystal (CLC) Elastomer (CLCE) mechanochromism offers a revolutionary novel approach to vision-based strain sensing. CLCs spontaneously form a supramolecular helix, leading to Bragg reflection (structural colour) of circularly polarized light. When a CLC is polymerized into a CLCE, a very interesting material arises, combining the striking optical properties of a cholesteric phase with the viscoelasticity of a rubber. Because of the coupling between the cholesteric helix and the elastomeric network, a mechanical deformation changes the reflection colour, as well as the polarization. This mechanochromic response of a CLCE can enable tensile and/or pressure sensors capable of visualizing complex mechanical strain in many objects. The response is continuous and local and can be detected with cameras, hence CLCEs allow the design of novel vision-based strain and force sensors, which might also be very beneficial in soft robotics and soft aerial manipulation.

Therefore, this interdisciplinary project COSAMOS focuses on the development of a novel vision-based strain and force sensing concept using CLCEs, the design of first soft aerial grippers and robotic arms with integrated CLCE sensors and the development of a related suitable control concept for SAMs. The solutions will be integrated to experimental SAM prototypes and finally tested in two demonstrations based on realistic use cases.

SMOD-SHA project aims to develop a smart system, bringing together, in a SM system, the multi-fault diagnosis, the self-healing and fault-tolerant approaches. The proposed system is challenging, especially, when integrating the multi-fault aspect of the SM system and its complex and heterogeneous structure that can be described by multi-models. SMOD-SHA project has a significant positive impact on sustainability, i.e. improves response time and enables for minimization of potential failures in the manufacturing process thus saving cost and time while maintaining operational efficiency, making accurate predictions when the machines need to be fixed and reducing energy consumption. As a result, we will have a high-quality manufacturing process optimization resource and energy consumption. In another word, robust and resilient manufacturing process.

The theory of financial contracting offers powerful tools to implement efficient decision making in private and public organizations. Optimal contracts provide arrangements that are compatible with the incentive constraints induced by private interests that hinder organizational efficiency. However, the applicability of the theory is still limited as quantifying the predictions of contracting and incentive models using available data poses technical and conceptual challenges.

This project takes a step towards closing the gap between theory and practice by enlarging the empirical content and applicability on the theory. The project takes advantage of modern computational and econometric tools to offer applications to four broad areas in financial economics. In corporate finance, I assess the quantitative importance of managerial beliefs and optimally-designed compensation contracts in shaping corporate policies. In auction theory, I design optimal procurement auctions for a financially-constrained buyer with implications for large-scale applications such as governmental purchases of pharmaceuticals (e.g. Covid19 vaccines). In public policy, I assess whether interventions are effective to address credit market failures and stimulate recovery during the Covid19 crisis. In financial markets, I show that the study of financial contracts between firms and external lenders provide helpful tools to track price fluctuations of financial securities.

In the past 10 years, blockchain technology (which underpins decentralized digital currencies and smart contracts) has been one of the most fertile research domains in computer science. Due to their decentralized nature and the need for a consensus mechanism, blockchains have created a number of novel and unique research challenges. The FNR CORE project FinCrypt tackled some of these challenges and came up with promising solutions to real-world problems in the context of blockchain applications and smart contracts. However, along with the rapid growth of digital currencies and the ever-increasing value of assets linked to blockchains, various new attack vectors have emerged and the limitations of current blockchain technologies regarding scalability and privacy become more and more apparent. Consequently, the need for research in the blockchain and smart contract domain is bigger than ever before. The proposed research builds on the success of the earlier FNR FinCrypt project but shifts the research focus towards scaling solutions and advanced cryptography for blockchains. The mission of the CryptoFin project is to seek innovative solutions for some of the most pressing open research problems in the blockchain domain, especially in the context of Layer-two (L2) protocols for off-chain transactions (e.g. Ethereum) and the design of advanced cryptographic techniques like Verifiable Delay Functions (VDFs), Proofs-of-Work (PoW) with special features and new MPC/SNARK-friendly primitives. Due to its timely nature and practical relevance, the proposed research has the potential to advance the state-of-the-art and create significant real-world impact.

Artificial Intelligence is playing an increasingly important role in our lives: from recommending specific products and websites to us, to predicting how we will vote in elections to driving our vehicles. It is also being used in ethically and socially important domains like healthcare, education and criminal justice. AI has the potential to greatly increase our knowledge by helping us make new scientific discoveries, prove new theorems and spot patterns hidden in data. But it also poses a potential threat to our knowledge and reasoning by ‘nudging’ us towards some kinds of information and away from others, creating ‘internet bubbles’, by reinforcing biases that are present in ‘Big Data’, by helping to spread and target political propaganda, by creating ‘deep-fake’ images and videos as well as increasingly sophisticated and human-like texts and conversations. The fundamental aim of this project is to investigate how we can rationally respond to the outputs of artificial intelligence systems and what is required to understand and explain AI systems. This is a topic that requires an inter-disciplinary approach, drawing on both computer science to investigate the details of recent AI advances but also on philosophy to investigate the nature of rationality, understanding and explanation. The issues here are especially important since many of the most powerful recent advances in AI have been achieved by training ‘Deep Neural Networks’ on vast amounts of data using machine learning techniques. This creates the unusual situation where even the designers and creators of these AI systems admit that they do not fully understand its internal processes or how the system will process new data. It is vital then that we investigate how we might produce explanations of the behaviour of these systems that humans can actually use and understand. It is also vitally important to investigate when and how it can be rational for human consumers to trust the outputs of systems trained via machine learning despite the fact that we lack full knowledge of their internal functioning or the data that was used to train them. One of our main hopes for this project is that we will be able to develop new ways of measuring how explain-able or how trust-worthy an AI system is that could eventually be implemented by computers.

In this project, we deal with different aspects of the design of a complex system that requires a global view of a given problem. Let’s take for instance the case of mobility: usually we solve local traffic congestion (public or private) with ad-hoc solutions. However, the solution impacts on the other transports means, over transportation network, and on the passengers, is rarely taken into account before problems occurs elsewhere. The pain-point to address is then to build a holistic representation for the overall mobility, on the country and its borders; thus requiring to collect all the viewpoints from all the stakeholders and then associating them consistently. The challenge addressed here is to be able to associate all the different viewpoints into a coherent whole, where each party understand each other and the related impacts: a model federation. This model federation can be used to answer all the questions we can ask about a system, in the context of mobility: it should provide answers to the passengers, up to the policy maker (efficiency of a regulation) passing by the drivers, urban planners and network managers.