Search

Autonomous Micro Aerial Vehicles (MAVs), with a form factor of 10cm in diameter, are an emerging technology thanks to the broad applicability enabled by their onboard intelligence. However, these platforms are strongly limited in the onboard power envelope for processing, i.e., less than a few hundred mW, which confines the onboard processors to the class of simple microcontroller units (MCUs). These MCUs lack advanced security features opening the way to a wide range of cyber security vulnerabilities, from the communication between agents of the same fleet to the onboard execution of malicious code. This work presents an open source System on Chip (SoC) design that integrates a 64 bit Linux capable host processor accelerated by an 8 core 32 bit parallel programmable accelerator. The heterogeneous system architecture is coupled with a security enclave based on an open source OpenTitan root of trust. To demonstrate our design, we propose a use case where OpenTitan detects a security breach on the SoC aboard the MAV and drives its exclusive GPIOs to start a LED blinking routine. This procedure embodies an unconventional visual communication between two palm sized MAVs: the receiver MAV classifies the LED state of the sender (on or off) with an onboard convolutional neural network running on the parallel accelerator. Then, it reconstructs a high-level message in 1.3s, 2.3 times faster than current commercial solutions.

Geometric pattern formation is an important emergent behavior in many applications involving large-scale multi-agent systems, such as sensor networks deployment and collective transportation. Attraction/repulsion virtual forces are the most common control approach to achieve such behavior in a distributed and scalable manner. Nevertheless, for most existing solutions only numerical and/or experimental evidence of their convergence is available. Here, we revisit the problem of achieving pattern formation giving sufficient conditions to prove analytically that under the influence of appropriate virtual forces, a large-scale multi-agent swarming system locally converges towards a stable and robust triangular lattice configuration. Specifically, the proof is carried out using LaSalle's invariance principle and geometry-based arguments. Our theoretical results are complemented by exhaustive numerical simulations confirming their effectiveness and estimating the region of asymptotic stability of the triangular configuration.

We investigate the properties of some recently developed variable-order differential operators involving order transition functions of exponential type. Since the characterisation of such operators is performed in the Laplace domain it is necessary to resort to accurate numerical methods to derive the corresponding behaviours in the time domain. In this regard, we develop a computational procedure to solve variable-order fractional differential equations of this novel class. Furthermore, we provide some numerical experiments to show the effectiveness of the proposed techniques., Comment: 23 pages, 13 figures

We consider black holes generically sourced by quantum matter described by regular wavefunctions. This allows for integrable effective energy densities and the removal of Cauchy horizons in spherically symmetric configurations. Moreover, we identify the ultrarigid rotation of the Kerr spacetime as causing the existence of an inner horizon in rotating systems, and describe general properties for quantum matter cores at the centre of rotating black holes with integrable singularities and no Cauchy horizon., Comment: LaTeX, 12 pages, 1 figure, final version to appear in JHEP

Precise relative localization is a crucial functional block for swarm robotics. This work presents a novel autonomous end-to-end system that addresses the monocular relative localization, through deep neural networks (DNNs), of two peer nano-drones, i.e., sub-40g of weight and sub-100mW processing power. To cope with the ultra-constrained nano-drone platform, we propose a vertically-integrated framework, from the dataset collection to the final in-field deployment, including dataset augmentation, quantization, and system optimizations. Experimental results show that our DNN can precisely localize a 10cm-size target nano-drone by employing only low-resolution monochrome images, up to ~2m distance. On a disjoint testing dataset our model yields a mean R2 score of 0.42 and a root mean square error of 18cm, which results in a mean in-field prediction error of 15cm and in a closed-loop control error of 17cm, over a ~60s-flight test. Ultimately, the proposed system improves the State-of-the-Art by showing long-endurance tracking performance (up to 2min continuous tracking), generalization capabilities being deployed in a never-seen-before environment, and requiring a minimal power consumption of 95mW for an onboard real-time inference-rate of 48Hz., Comment: \c{opyright} 2023 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works

Miniaturized autonomous unmanned aerial vehicles (UAVs) are an emerging and trending topic. With their form factor as big as the palm of one hand, they can reach spots otherwise inaccessible to bigger robots and safely operate in human surroundings. The simple electronics aboard such robots (sub-100mW) make them particularly cheap and attractive but pose significant challenges in enabling onboard sophisticated intelligence. In this work, we leverage a novel neural architecture search (NAS) technique to automatically identify several Pareto-optimal convolutional neural networks (CNNs) for a visual pose estimation task. Our work demonstrates how real-life and field-tested robotics applications can concretely leverage NAS technologies to automatically and efficiently optimize CNNs for the specific hardware constraints of small UAVs. We deploy several NAS-optimized CNNs and run them in closed-loop aboard a 27-g Crazyflie nano-UAV equipped with a parallel ultra-low power System-on-Chip. Our results improve the State-of-the-Art by reducing the in-field control error of 32% while achieving a real-time onboard inference-rate of ~10Hz@10mW and ~50Hz@90mW., Comment: \c{opyright} 2023 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works

Offshore wind turbine towers are pre-assembled and temporarily held in close proximity to each other in group arrangements on port quaysides. Accurate estimates of the aerodynamic loads on the individual towers and on the overall group are essential for the safe and economic design of the quayside's supporting structures and foundations. Wind tunnel tests represent the main way to address the problem, but results may lead to overconservative designs due to inevitable mismatches in the Reynolds number. This crucial issue is dealt with here using an original engineering solution based on concentrated but discontinuous surface roughness, which allows, for the first time in the case of finite-height towers arranged in groups and subjected to an atmospheric boundary layer flow, the successful simulation of the target high Reynolds number regime. This case study assumes slender wind turbine towers with a height of 115 m, and for the sake of generality, the investigations focus principally on a cylindrical shape rather than the more complex real-world geometry. The constant diameter of the towers is determined based on the theoretical equivalence of the mean overturning moment. The rationality of this procedure is verified a posteriori using a set of measurements on the real-shape towers. The experiments show a regular behavior of the maximum mean base shear force and moment for towers arranged in double-row groups, while the results are more complicated for single-row groups, since biased flow sometimes occurs in symmetric or nearly-symmetric configurations. Dynamic loads are also inspected, and gust factors in good agreement with Eurocode 1's prescriptions are found. Several parametric studies are carried out, the most extensive of which is devoted to assessing the role of tower height. A complicated non-monotonic pattern of the load coefficients with the tower height is encountered., Comment: 29 pages, 31 figures

Stealth solutions of scalar-tensor gravity and less-known de Sitter spaces that generalize them are analyzed regarding their possible role as thermal equilibria at non-zero temperature in the new first-order thermodynamics of scalar-tensor gravity. No stable equilibria are found, further validating the special role of general relativity as an equilibrium state in the landscape of gravity theories, seen through the lens of first-order thermodynamics., Comment: 10 pages, no figures

Message Passing Neural Networks (MPNNs) are instances of Graph Neural Networks that leverage the graph to send messages over the edges. This inductive bias leads to a phenomenon known as over-squashing, where a node feature is insensitive to information contained at distant nodes. Despite recent methods introduced to mitigate this issue, an understanding of the causes for over-squashing and of possible solutions are lacking. In this theoretical work, we prove that: (i) Neural network width can mitigate over-squashing, but at the cost of making the whole network more sensitive; (ii) Conversely, depth cannot help mitigate over-squashing: increasing the number of layers leads to over-squashing being dominated by vanishing gradients; (iii) The graph topology plays the greatest role, since over-squashing occurs between nodes at high commute (access) time. Our analysis provides a unified framework to study different recent methods introduced to cope with over-squashing and serves as a justification for a class of methods that fall under graph rewiring., Comment: Accepted at ICML 2023; 21 pages

We present a study of lattice-QCD methods to determine the relevant hadronic form factors for radiative leptonic decays of pseudoscalar mesons. We provide numerical results for $D_s^+ \to \ell^+ \nu \gamma$. Our calculation is performed using a domain-wall action for all quark flavors and on a single RBC/UKQCD lattice gauge-field ensemble. The first part of the study is how to best control two sources of systematic error inherent in the calculation, specifically the unwanted excited states created by the meson interpolating field, and unwanted exponentials in the sum over intermediate states. Using a 3d sequential propagator allows for better control over unwanted exponentials from intermediate states, while using a 4d sequential propagator allows for better control over excited states. We perform individual analyses of the 3d and 4d methods as well as a combined analysis using both methods, and find that the 3d sequential propagator offers good control over both sources of systematic uncertainties for the smallest number of propagator solves. From there, we further improve the use of a 3d sequential propagator by employing an infinite-volume approximation method, which allows us to calculate the relevant form factors over the entire allowed range of photon energies. We then study improvements gained by performing the calculation using a different three-point function, using ratios of three-point functions, averaging over positive and negative photon momentum, and using an improved method for extracting the structure-dependent part of the axial form factor. The optimal combination of methods yields results for the $D_s^+ \to \ell^+ \nu \gamma$ structure-dependent vector and axial form factors in the entire kinematic range with statistical plus fitting uncertainties of order 5%, using 25 gauge configurations with 64 samples per configuration., Comment: 31 pages, 14 figures

Time-resolved dynamics in charge-density-wave materials have revealed interesting out-of-equilibrium electronic responses. However these are typically only performed in a single material possessing a CDW. As such, it is challenging to separate subtle effects originating from the CDW. Here, we report on the ultrafast dynamics of the GdSb$_{x}$Te$_{2-x-\delta}$ series of materials where E$_{F}$ can be tuned, resulting in a change from an undistorted tetraganal phase to a CDW with a wavevector that depends on $x$. Using mid-infrared, near-infrared, and visible excitation, we find the dynamics are sensitive to both E$_{F}$ and the presence of the CDW. Specifically, as the Sb content of the compounds increases, transient spectral features shift to higher probe energies. In addition, we observe an enhanced lifetime and change in the sign of the transient signal upon removing the CDW with high Sb concentrations. Finally, we reveal fluence- and temperature-dependent photo-induced responses of the differential reflectivity, which provide evidence of transient charge density wave suppression in related telluride materials. Taken together our results provide a blueprint for future ultrafast studies of CDW systems.

Indirect noise is a significant contributor to aircraft engine noise, which needs to be minimized in the design of aircraft engines. Indirect noise is caused by the acceleration of flow inhomogeneities through a nozzle. High-fidelity simulations showed that some flow inhomogeneities can be chemically reacting when they leave the combustor and enter the nozzle (Giusti et al., 2019). The state-of-art models, however, are limited to chemically non-reacting (frozen) flows. In this work, first, we propose a low-order model to predict indirect noise in nozzle flows with reacting inhomogeneities. Second, we identify the physical sources of sound, which generate indirect noise via two physical mechanisms: (i) chemical reaction generates compositional perturbations, thereby adding to compositional noise; and (ii) exothermic reaction generates entropy perturbations. Third, we numerically compute the nozzle transfer functions for different frequency ranges (Helmholtz numbers) and reaction rates (Damk\"{o}hler numbers) in subsonic flows with hydrogen and methane inhomogeneities. Fourth, we extend the model to supersonic flows. We find that hydrogen inhomogeneities have a larger impact to indirect noise than methane inhomogeneities. Both the Damk\"{o}hler number and the Helmholtz number markedly influence the phase and magnitude of the transmitted and reflected waves, which affect the sound generation and thermoacoustic stability. This work provides a physics-based low-order model, which can open new opportunities for predicting noise emissions and instabilities in aeronautical gas turbines with multi-physics flows.

Given a (smoothable) projective nodal K\"ahler Calabi-Yau threefold, we show, via a gluing construction, that all its - possibly non-K\"ahler - small resolutions admit Chern-Ricci flat balanced metrics, which among other things solve the dilatino equation appearing in the Hull-Strominger system., Comment: 12 pages. The issue in the proof of the previous version is now fixed. Comments are welcome!

We compute the standard Euclidean window of the hadronic vacuum polarization using multiple independent blinded analyses. We improve the continuum and infinite-volume extrapolations of the dominant quark-connected light-quark isospin-symmetric contribution and address additional sub-leading systematic effects from sea-charm quarks and residual chiral-symmetry breaking from first principles. We find $a_\mu^{\rm W} = 235.56(65)(50) \times 10^{-10}$, which is in $3.8\sigma$ tension with the recently published dispersive result of Colangelo et al., $a_\mu^{\rm W} = 229.4(1.4) \times 10^{-10}$, and in agreement with other recent lattice determinations. We also provide a result for the standard short-distance window. The results reported here are unchanged compared to our presentation at the Edinburgh workshop of the g-2 Theory Initiative in 2022., Comment: 24 pages, 15 figures

We investigate the specific attenuation factor for the Bessel models of viscoelasticity. We find that the quality factor for this class can be expressed in terms of Kelvin functions and that its asymptotic behaviours confirm the analytical results found in previous studies for the rheological properties of these models., Comment: 11 pages, 6 figures

A numerical framework for the 'real-time' estimation of the infection risk from airborne diseases (e.g., SARS-CoV-2) in indoor spaces such as hospitals, restaurants, cinemas or teaching rooms is proposed. The developed model is based on the use of computational fluid dynamics as a pre-processor to obtain the time-averaged ventilation pattern inside a room, and a post-processing tool for the computation of the dispersion of sprays and aerosols emitted by its occupants in 'real time'. The model can predict the dispersion and concentration of droplets carrying viable viral copies in the air, the contamination of surfaces, and the related spatially-resolved infection risk. It may therefore provide useful information for the management of indoor environments in terms of, e.g., maximum occupancy, air changes per hour and cleaning of surfaces. This work describes the fundamentals of the model and its main characteristics. The model was developed using open-source software and is conceived to be simple, user-friendly and highly automated to enable any potential user to perform estimations of the local infection risk., Comment: for associated video, see https://www.linkedin.com/posts/danielfredrich_we-have-developed-sprei-an-easy-to-use-computational-activity-6865360164510404608-ORyU?utm_source=share&utm_medium=member_desktop

Molecular dynamics (MD) simulations are performed to investigate the thermal and mass accommodation coefficients (TAC and MAC, respectively) for the combination of iron(-oxide) and air. The obtained values of TAC and MAC are then used in a point-particle Knudsen model to investigate the effect of chemisorption and the Knudsen transition regime on the combustion behavior of (fine) iron particles. The thermal accommodation for the interactions of $\mathrm{Fe}$ with $\mathrm{N_2}$ and $\mathrm{Fe_xO_y}$ with $\mathrm{O_2}$ is investigated for different surface temperatures, while the mass accommodation coefficient for iron(-oxide) with oxygen is investigated for different initial oxidation stages $Z_\mathrm{O}$, which represents the molar ratio of $\mathrm{O}/\left(\mathrm{O} + \mathrm{Fe}\right)$, and different surface temperatures. The MAC decreases fast from unity to 0.03 as $Z_\mathrm{O}$ increases from 0 to 0.5 and then diminishes as $Z_\mathrm{O}$ further increases to 0.57. By incorporating the MD-informed accommodation coefficients into the single iron particle combustion model,a new temperature evolution for single iron particles is observed compared to results obtained with previously developed continuum models. Specifically, results of the present simulations show that the oxidation process continues after the particle reaching the peak temperature, while previous models predicting that the maximum temperature was attained when the particle is oxidized to $Z_\mathrm{O} = 0.5$. Since the rate of oxidation slows down as the MAC decreases with an increasing oxidation stage, the rate of heat loss exceeds the rate of heat release upon reaching the maximum temperature, while the particle is not yet oxidized to $Z_\mathrm{O} = 0.5$. Finally, the effect of transition-regime heat and mass transfer on the combustion behavior of fine iron particles is investigated and discussed.

We present a strategy to study QCD non-perturbatively on the lattice at very high temperatures. This strategy exploits a non-perturbative, finite-volume, definition of the strong coupling constant to renormalize the theory. As a first application we compute the flavor non-singlet mesonic screening masses in a wide range of temperature, from $T\sim 1 $ GeV up to $\sim 160 $ GeV with three flavors in the chiral limit of QCD. Our results show very interesting features of the screening spectrum at very high temperatures. On one hand the mass splitting between the vector and the pseudoscalar screening masses is clearly visible up to the electroweak scale and cannot be explained by the known 1-loop perturbative result. On the other hand the restoration of chiral symmetry manifests itself through the degeneracy of the pseudoscalar and the scalar channels and of the vector and the axial-vector ones. This degeneracy pattern is the one expected by Ward identities associated to the presence of chiral symmetry., Comment: 9 pages, 4 figures, contribution to the 39th International Symposium on Lattice Field Theory (LATTICE2022), August 8-13, 2022, Bonn (Germany)

The aim of this work is to introduce MaRF, a novel framework able to synthesize the Martian environment using several collections of images from rover cameras. The idea is to generate a 3D scene of Mars' surface to address key challenges in planetary surface exploration such as: planetary geology, simulated navigation and shape analysis. Although there exist different methods to enable a 3D reconstruction of Mars' surface, they rely on classical computer graphics techniques that incur high amounts of computational resources during the reconstruction process, and have limitations with generalizing reconstructions to unseen scenes and adapting to new images coming from rover cameras. The proposed framework solves the aforementioned limitations by exploiting Neural Radiance Fields (NeRFs), a method that synthesize complex scenes by optimizing a continuous volumetric scene function using a sparse set of images. To speed up the learning process, we replaced the sparse set of rover images with their neural graphics primitives (NGPs), a set of vectors of fixed length that are learned to preserve the information of the original images in a significantly smaller size. In the experimental section, we demonstrate the environments created from actual Mars datasets captured by Curiosity rover, Perseverance rover and Ingenuity helicopter, all of which are available on the Planetary Data System (PDS)., Comment: ECCV 2022 (oral)

Few years ago we started the investigation of microscopic Optical Potentials (OP) in the framework of chiral effective field theories and published our results in a series of manuscripts. Starting from the very first work, where a microscopic OP was introduced following the multiple scattering procedure of Watson, and then followed by more recent works, where the agreement with experimental data and phenomenological approaches was successfully tested, we finally arrived at a description of elastic scattering processes off non-zero spin nuclei. Among our achievements, it is worth mentioning the partial inclusion of three-nucleon forces, and the extension of our OP to antiproton-nucleus elastic scattering. Despite the overall good agreement with empirical data obtained so far, we do believe that several improvements and upgrades of the present approach are still to be achieved. In this short essay we would like to address some of the most relevant achievements and discuss an interesting development that, in our opinion, is needed to further improve microscopic OPs in order to reach in a near future the same level of accuracy of the phenomenological ones., Comment: 8 pages, 4 figures, Conference proceedings of the 13th International Spring Seminar on Nuclear Physics, Sant'Angelo d'Ischia, May 15-20, 2022

A timely and long-term programme of kaon decay measurements at a new level of precision is presented, leveraging the capabilities of the CERN Super Proton Synchrotron (SPS). The proposed programme is firmly anchored on the experience built up studying kaon decays at the SPS over the past four decades, and includes rare processes, CP violation, dark sectors, symmetry tests and other tests of the Standard Model. The experimental programme is based on a staged approach involving experiments with charged and neutral kaon beams, as well as operation in beam-dump mode. The various phases will rely on a common infrastructure and set of detectors., Comment: Letter of Intent submitted to CERN SPSC. Address all correspondence to hike-eb@cern.ch

We present our progresses in the use of the non-perturbative renormalization framework based on considering QCD at finite temperature with shifted and twisted (for quarks only) boundary conditions in the compact direction. We report our final results in the application of this method for the non-perturbative renormalization of the flavor-singlet local vector current. We then discuss the more challenging case of the renormalization of the energy-momentum tensor, and show preliminary results on the relevant one-point functions for the computation of the renormalization constants of its non-singlet components., Comment: 9 pages, 2 figures, contribution to the 39th International Symposium on Lattice Field Theory, 8th-13th August, 2022, Bonn, Germany

We discuss a strategy to study non-perturbatively QCD up to very high temperatures by Monte Carlo simulations on the lattice. It allows not only the thermodynamic properties of the theory but also other interesting thermal features to be investigated. As a first concrete application, we compute the flavour non-singlet mesonic screening masses and we present the results of Monte Carlo simulations at 12 temperatures covering the range from T $\sim$ 1 GeV up to $\sim$ 160 GeV in the theory with three massless quarks. On the one side, chiral symmetry restoration manifests itself in our results through the degeneracy of the vector and the axial vector channels and of the scalar and the pseudoscalar ones, and, on the other side, we observe a clear splitting between the vector and the pseudoscalar screening masses up to the highest investigated temperature. A comparison with the high-temperature effective theory shows that the known one-loop order in the perturbative expansion does not provide a satisfactory description of the non-perturbative data up to the highest temperature considered., Comment: 6 pages, 3 figures; contribution to the proceedings of the International Conference on High Energy Physics (ICHEP 2022), July 6-13, 2022, Bologna (Italy)

In our everyday life, we intuitively use space to regulate our social interactions. When we want to talk to someone, we approach them; if someone joins the conversation, we adjust our bodies to make space for them. In contrast, devices are not as considerate: they interrupt us, require us to input commands, and compete for our attention. In this paper, we introduce Fields, a design framework for ubiquitous computing that informs the design of connected products with social grace. Inspired by interactionist theories on social interaction, Fields builds on the idea that the physical space we share with computers can be an interface to mediate interactions. It defines a generalized approach to spatial interactions, and a set of interaction patterns that can be adapted to different ubiquitous computing systems. We investigated its value by implementing it in a set of prototypes and evaluating it in a lab setting., Comment: 21 pages

The non-local dependence of the fermion determinant on the gauge field limits our ability of simulating Quantum Chromodynamics on the lattice. Here we present a factorization of the gauge field dependence of the fermion determinant based on an overlapping four-dimensional domain decomposition of the lattice. The resulting action is block-local in the gauge and in the auxiliary bosonic fields. Possible applications are multi-level integration, master field simulations, and more efficient parallelizations of Monte Carlo algorithms and codes.

We refine and slightly enlarge the recently proposed first-order thermodynamics of scalar-tensor gravity to include gravitational scalar fields with timelike and past-directed gradients. The implications and subtleties arising in this situation are discussed and an exact cosmological solution of scalar-tensor theory in first-order thermodynamics is revisited in light of these results., Comment: 7 pages, no figures

An increasing number of nonspecialist robotic users demand easy-to-use machines. In the context of visual servoing, the removal of explicit image processing is becoming a trend, allowing an easy application of this technique. This work presents a deep learning approach for solving the perception problem within the visual servoing scheme. An artificial neural network is trained using the supervision coming from the knowledge of the controller and the visual features motion model. In this way, it is possible to give a geometrical interpretation to the estimated visual features, which can be used in the analytical law of the visual servoing. The approach keeps perception and control decoupled, conferring flexibility and interpretability on the whole framework. Simulated and real experiments with a robotic manipulator validate our approach., Comment: IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2022

Placing quantum materials into optical cavities provides a unique platform for controlling quantum cooperative properties of matter, via both weak and strong light-matter coupling. Here we report the experimental evidence of reversible cavity control of a metal-to-insulator phase transition in a correlated solid-state material. We embed the charge density wave material 1T-TaS$_{2}$ into cryogenic tunable terahertz cavities and show that a switch between conductive and insulating behaviors, associated with a large change in the sample temperature, is obtained by mechanically tuning the distance between the cavity mirrors and their alignment. The large thermal modification observed is indicative of a Purcell-like scenario in which the spectral profile of the cavity modifies the energy exchange between the material and the external electromagnetic field. Our findings provide opportunities for controlling the thermodynamics and macroscopic transport properties of quantum materials by engineering their electromagnetic environment., Comment: 52 pages, 36 figures (including Supplementary Information)