Search

Robotic systems are increasingly employed for industrial automation, with contact-rich tasks like polishing requiring dexterity and compliant behaviour. These tasks are difficult to model, making classical control challenging. Deep reinforcement learning (RL) offers a promising solution by enabling the learning of models and control policies directly from data. However, its application to real-world problems is limited by data inefficiency and unsafe exploration. Adaptive hybrid RL methods blend classical control and RL adaptively, combining the strengths of both: structure from control and learning from RL. This has led to improvements in data efficiency and exploration safety. However, their potential for hardware applications remains underexplored, with no evaluations on physical systems to date. Such evaluations are critical to fully assess the practicality and effectiveness of these methods in real-world settings. This work presents an experimental demonstration of the hybrid RL algorithm CHEQ for robotic polishing with variable impedance, a task requiring precise force and velocity tracking. In simulation, we show that variable impedance enhances polishing performance. We compare standalone RL with adaptive hybrid RL, demonstrating that CHEQ achieves effective learning while adhering to safety constraints. On hardware, CHEQ achieves effective polishing behaviour, requiring only eight hours of training and incurring just five failures. These results highlight the potential of adaptive hybrid RL for real-world, contact-rich tasks trained directly on hardware.

Recent advances in cover song identification have shown great success. However, models are usually tested on a fixed set of datasets which are relying on the online cover song database SecondHandSongs. It is unclear how well models perform on cover songs on online video platforms, which might exhibit alterations that are not expected. In this paper, we annotate a subset of songs from YouTube sampled by a multi-modal uncertainty sampling approach and evaluate state-of-the-art models. We find that existing models achieve significantly lower ranking performance on our dataset compared to a community dataset. We additionally measure the performance of different types of versions (e.g., instrumental versions) and find several types that are particularly hard to rank. Lastly, we provide a taxonomy of alterations in cover versions on the web., Comment: accepted for presentation at iConference 2025

We develop a mesoscale transport theory for the charge and spin degrees of freedom of itinerant carriers in a $d$-wave altermagnet. Our effective Lagrangian description is built upon the slave-boson formulation of the microscopic $t-J$ model. We obtain a spin-polarized diffusive contribution to the effective Hamiltonian, with no counterpart in conventional antiferromagnetism and parametrized by the spin splitting, that is responsible for the so-called spin-splitter effect in $d$-wave altermagnets. We also elucidate the spin-transfer response of the itinerant fluid as well as the spin pumping into the altermagnet, which show previously unidentified combinations of the charge current and spatial partial derivatives (namely, {$j_{x}^{e}$,$\partial_{y}$} and {$j_{y}^{e}$,$\partial_{x}$}). The emergent spin-transfer physics in $d$-wave altermagnets opens up new possibilities for the dynamics of spin textures, such as the domain-wall motion driven by transverse charge currents. We also consider the effect of elastic distortions in the aforementioned transport properties., Comment: 18 pages, 2 figures

Detecting music entities such as song titles or artist names is a useful application to help use cases like processing music search queries or analyzing music consumption on the web. Recent approaches incorporate smaller language models (SLMs) like BERT and achieve high results. However, further research indicates a high influence of entity exposure during pre-training on the performance of the models. With the advent of large language models (LLMs), these outperform SLMs in a variety of downstream tasks. However, researchers are still divided if this is applicable to tasks like entity detection in texts due to issues like hallucination. In this paper, we provide a novel dataset of user-generated metadata and conduct a benchmark and a robustness study using recent LLMs with in-context-learning (ICL). Our results indicate that LLMs in the ICL setting yield higher performance than SLMs. We further uncover the large impact of entity exposure on the best performing LLM in our study.

YouTube is a rich source of cover songs. Since the platform itself is organized in terms of videos rather than songs, the retrieval of covers is not trivial. The field of cover song identification addresses this problem and provides approaches that usually rely on audio content. However, including the user-generated video metadata available on YouTube promises improved identification results. In this paper, we propose a multi-modal approach for cover song identification on online video platforms. We combine the entity resolution models with audio-based approaches using a ranking model. Our findings implicate that leveraging user-generated metadata can stabilize cover song identification performance on YouTube., Comment: accepted for presentation at NLP for Music and Audio (NLP4MusA) 2024

Motivated by the ongoing interest in understanding the actual magnetic ground state of the promising axion insulator candidate EuIn2As2, we present here a spin symmetry analysis and ab-initio calculations, aiming to identify specific exchange-dominated physics that could offer insights into the current debate. We investigate two non-collinear coplanar magnetic orders reported in this compound: the helical and broken helical phases. Our symmetry analysis shows that magnetic-exchange alone results in the formation of an out-of-plane odd-wave order in momentum space with a single un-polarized nodal plane in both phases. Additionally, we identified an in-plane g-wave order that emerges exclusively in the broken helical phase, providing a distinguishing feature for this phase. Furthermore, we report a non-relativistic Edelstein effect with a distinct out-of-plane polarized spin density that dominates over spin-orbit coupling effects. Our ab-initio calculations reveal a significant contrast in the magnitude of this effect between both phases, which could serve as a means to identify the magnetic transition and distinguish them from other magnetic ground states proposed for this compound., Comment: 8 pages, 9 figures

Plastic scintillators are widely used for the detection of elementary particles, and 3D reconstruction of particle tracks is achieved by segmenting the detector into 3D granular structures. In this study, we present a novel prototype fabricated by additive manufacturing, consisting of a 5 x 5 x 5 array of 1 cm3 plastic scintillator cubes, each optically isolated. This innovative approach eliminates the need to construct complex monolithic geometries in a single operation and gets rid of the traditional time-consuming manufacturing and assembling processes. The prototype underwent performance characterization during a beam test at CERN's Proton-Synchrotron facility. Light yield, optical crosstalk, and light response uniformity, were evaluated. The prototype demonstrated a consistent light yield of approximately 27 photoelectrons (p.e.) per channel, similar to traditional cast scintillator detectors. Crosstalk between adjacent cubes averaged 4-5%, and light yield uniformity within individual cubes exhibited about 7% variation, indicating stability and reproducibility. These results underscore the potential of the novel additive manufacturing technique, for efficient and reliable production of high-granularity scintillator detectors.

The origin and efficiency of charge-to-spin conversion, known as the Edelstein effect (EE), has been typically linked to spin-orbit coupling mechanisms, which require materials with heavy elements within a non-centrosymmetric environment. Here we demonstrate that the high efficiency of spin-charge conversion can be achieved even without spin-orbit coupling in the recently identified coplanar p-wave magnets. The non-relativistic Edelstein effect (NREE) in these magnets exhibits a distinct phenomenology compared to the relativistic EE, characterized by a strongly anisotropic response and an out-of-plane polarized spin density resulting from the spin symmetries. We illustrate the NREE through minimal tight-binding models, allowing a direct comparison to different systems. Through first-principles calculations, we further identify the p-wave candidate material CeNiAsO as a high-efficiency NREE material, revealing a 25 times larger response than the maximally achieved relativistic EE and other reported NREE in non-collinear magnetic systems with broken time-reversal symmetry. This highlights the potential for efficient spin-charge conversion in p-wave magnetic systems.

Altermagnets constitute a new class of magnetic materials that combine properties previously thought to be exclusive to either antiferromagnets or ferromagnets, and have unique properties of their own. In particular, a combination of symmetries connecting magnetic sublattices gives rise to a band spin splitting exhibiting unconventional d, g, or i-wave character. Their unique electronic properties have already led to new spin-dependent transport effects. Here, we consider their spin and charge dynamics on ultrafast timescales. We use a minimal tight binding model that captures the main features of the altermagnetic candidate material KRu$_4$O$_8$. In the framework of this model, we compute the spin-dependent electronic scattering dynamics after ultrashort-pulse excitation and show through these microscopic calculations how electron-electron and electron-phonon scattering processes redistribute optically excited carriers in a 2D slice of the Brillouin zone. We find that the optically excited spin polarization is long lived (~1ps) compared to the electron-electron momentum scattering lifetime of roughly 10fs. This contrasts remarkably with the much shorter spin lifetimes observed in typical ultrafast electronic spin dynamics in conventional ferromagnets and antiferromagnets, making these pulse-driven spin excitation experiments a key probe of altermagnetism.

We show that magnetic insulators with a collinear and compensated order can exhibit a thermal Hall effect even at zero magnetic field if they have altermagnetic symmetries. We predict a finite thermal Hall conductivity vector $\boldsymbol{\kappa}_\text{H}$ for a rutile-inspired effective spin model with Dzyaloshinskii-Moriya interaction. Within the linear spin-wave theory, we identify two magnon branches that carry identical Berry curvature and give rise to a finite $\boldsymbol{\kappa}_\text{H}$, which can be controlled by the N\'eel vector orientation and by strain. The thermal Hall response is further complemented with a spin Nernst response to contrast spin and heat transport in altermagnetic insulators with those in ferromagnets and antiferromagnets. Our results establish the crystal thermal Hall effect of magnons and we discuss material candidates for experimental realization, such as MnF$_2$, CoF$_2$, and NiF$_2$., Comment: 23 pages

Data is always at the center of the theoretical development and investigation of the applicability of formal concept analysis. It is therefore not surprising that a large number of data sets are repeatedly used in scholarly articles and software tools, acting as de facto standard data sets. However, the distribution of the data sets poses a problem for the sustainable development of the research field. There is a lack of a central location that provides and describes FCA data sets and links them to already known analysis results. This article analyses the current state of the dissemination of FCA data sets, presents the requirements for a central FCA repository, and highlights the challenges for this., Comment: 16 pages

The anomalous Nernst effect generates transverse voltage to the applied thermal gradient in magnetically ordered systems. The effect was previously considered excluded in compensated magnetic materials with collinear ordering. However, in the recently identified class of compensated magnetic materials, dubbed altermagnets, time-reversal symmetry breaking in the electronic band structure makes the presence of the anomalous Nernst effect possible despite the collinear spin arrangement. In this work, we investigate epitaxial Mn5Si3 thin films known to be an altermagnetic candidate. We show that the material manifests a sizable anomalous Nernst coefficient despite the small net magnetization of the films. The measured magnitudes of the anomalous Nernst coefficient reach a scale of microVolts per Kelvin. We support our magneto-thermoelectric measurements by density-functional theory calculations of the material's spin-split electronic structure, which allows for the finite Berry curvature in the reciprocal space. Furthermore, we present our calculations of the intrinsic Berry-curvature Nernst conductivity, which agree with our experimental observations.

We present a phenomenological theory of altermagnets, that captures their unique magnetization dynamics and allows modelling magnetic textures in this new magnetic phase. Focusing on the prototypical d-wave altermagnets, e.g. RuO$_2$, we can explain intuitively the characteristic lifted degeneracy of their magnon spectra, by the emergence of an effective sublattice-dependent anisotropic spin stiffness arising naturally from the phenomenological theory. We show that as a consequence the altermagnetic domain walls, in contrast to antiferromagnets, have a finite gradient of the magnetization, with its strength and gradient direction connected to the altermagnetic anisotropy, even for 180$^\circ$ domain walls. This gradient generates a ponderomotive force in the domain wall in the presence of a strongly inhomogeneous external magnetic field, which may be achieved through magnetic force microscopy techniques. The motion of these altermagentic domain walls is also characterized by an anisotropic Walker breakdown, with much higher speed limits of propagation than ferromagnets but lower than antiferromagnets.

The online implementation of model predictive control for constrained multivariate systems has two main disadvantages: it requires an estimate of the entire model state and an optimisation problem must be solved online. These issues have typically been treated separately. This work proposes an integrated approach for the offline training of an output feedback neural network controller in closed loop. Online this neural network controller computers the plant inputs cheaply using noisy measurements. In addition, the controller can be trained to only make use of certain predefined measurements. Further, a heuristic approach is proposed to perform the automatic selection of important measurements. The proposed method is demonstrated by extensive simulations using a non-linear distillation column model of 50 states.

Plastic-scintillator detectors are devices used for the detection of elementary particles. They provide good particle identification with excellent time resolution, whilst being inexpensive due to the affordability of plastic materials. Particle tracking is achieved by segmenting the scintillator into smaller optically-isolated 3D granular sub-structures which require the integration of multiple types of plastic materials as well as several thousands of tiny holes through a compact volume of several cubic meters. Future particle detectors necessitate larger volumes, possibly with even finer segmentation. However, manufacturing such geometries with current production strategies is challenging, as they involve time-consuming and costly fabrication processes, followed by the assembly of millions of individual parts. The difficulty in scaling up such a workflow can be addressed by additive manufacturing, enabling the construction of complex, monolithic geometries in a single operation. This article presents the fabrication of the first additive manufactured plastic scintillator detector, capable of 3D tracking elementary particles and measuring their stopping power. Its performance is comparable to the state of the art of plastic scintillator detectors. This work paves the way towards a new feasible, time and cost-effective process for the production of future plastic-based scintillator detectors, regardless their size and difficulty in geometry.

For large systems that consider uncertainty the online solution of model predictive control problems can be computationally taxing, and even infeasible. This can be offset by using a shorter horizon, however this can in turn result in poor controller performance. In this work we consider the task of learning convex objective terms to allow the use of a short, potentially single step, control horizon to reduce the online computational cost. We consider two surrogates: (1) a convex interpolating function and (2) an input-convex neural network. We highlight that irrespective of the choice of surrogate the behaviour of the surrogate near the origin and ability of the surrogate to describe the feasible region are key concerns for the closed loop performance of the new MPC problem. We address these concerns by tailoring the design of the surrogate to ensure good performance in both aspects.The paper concludes with a numerical example, showing the clear and significant reduction in computational complexity through the use of these convex surrogates.

Topological semimetal antiferromagnets provide a rich source of exotic topological states which can be controlled by manipulating the orientation of the N\'eel vector, or by modulating the lattice parameters through strain. We investigate via ${ab\ initio}$ density functional theory calculations, the effects of shear strain on the bulk and surface states n two antiferromagnetic EuCd$_2$As$_2$ phases with out-of-plane and in-plane spin configurations. When magnetic moments are along the $\textit{c}$-axis, a $3\%$ longitudinal or diagonal shear strain can tune the Dirac semimetal phase to an axion insulator phase, characterized by the parity-based invariant $\eta_{4I} = 2$. For an in-plane magnetic order, the axion insulator phase remains robust under all shear strains. We further find that for both magnetic orders, the bulk gap increases and a surface gap opens on the (001) surface up to 16 meV. Because of a nonzero $\eta_{4I}$ index and gapped states on the (001) surface, hinge modes are expected to happen on the side surface states between those gapped surface states. This result can provide a valuable insight in the realization of the long-sought axion states., Comment: 5 pages, 4 figures

The p-wave Cooper-pairing instability in superfluid $^{3}$He, characterized by a parity-breaking excitation gap, is regarded as one of the most rich and complex phenomena in physics. The possibility of a counterpart unconventional p-wave ordering of interacting fermions, in which a Fermi surface spontaneously breaks the parity symmetry, has been an open problem for many decades. Here we identify the realization of the counterpart of p-wave superfluidity in magnetism. We demonstrate a strong parity-breaking and anisotropic symmetry lowering of spin-polarized and time-reversal symmetric Fermi surfaces in a representative p-wave magnet CeNiAsO. As a direct experimental signature we predict a large spontaneous anisotropy of the resistivity. Abundant and robust realizations of the unconventional p-wave magnetism can be identified from suitable non-relativistic crystal-lattice and spin symmetries, without requiring strong correlations and extreme external conditions. This opens new prospects in fields ranging from topological phenomena to spintronics., Comment: 6 pages, 3 figures

Altermagnets are compensated magnets with unconvetional $d$, $g$ and $i$-wave spin-channel order in reciprocal space. So far the search for new altermagnetic candidates has been focused on materials in which the magnetic unit cell is identical to the non-magnetic one, i.e. magnetic structures with zero propagation vector. Here, we substantially broaden the family of altermagnetic candidates by predicting supercell altermagnets. Their magnetic unit cell is constructed by enlarging the paramagnetic primitive unit cell, resulting in a non-zero propagation vector for the magnetic structure. This connection of the magnetic configuration to the ordering of sublattices gives an extra degree of freedom to supercell altermagnets, which can allow for the control over the order parameter spatial orientation. We identify realistic candidates MnSe$_2$ with a $d$-wave order, and RbCoBr$_3$, CsCoCr$_3$, and BaMnO$_3$ with $g$-wave order. We demonstrate the reorientation of the order parameter in MnSe$_2$, which has two different magnetic configurations, whose energy difference is only 5 meV, opening the possibility of controlling the orientation of the altermagnetic order parameter by external perturbations., Comment: 10 pages, 4 figures

This paper presents an adaptive horizon multi-stage model-predictive control (MPC) algorithm. It establishes appropriate criteria for recursive feasibility and robust stability using the theory of input-to-state practical stability (ISpS). The proposed algorithm employs parametric nonlinear programming (NLP) sensitivity and terminal ingredients to determine the minimum stabilizing prediction horizon for all the scenarios considered in the subsequent iterations of the multi-stage MPC. This technique notably decreases the computational cost in nonlinear model-predictive control systems with uncertainty, as they involve solving large and complex optimization problems. The efficacy of the controller is illustrated using three numerical examples that illustrate a reduction in computational delay in multi-stage MPC., Comment: Accepted for publication in Elsevier's Journal of Process Control

We use sensitivity analysis to design $\textit{optimality-based}$ discretization (cutting-plane) methods for the global optimization of nonconvex semi-infinite programs (SIPs). We begin by formulating the optimal discretization of SIPs as a max-min problem and propose variants that are more computationally tractable. We then use parametric sensitivity theory to design an efficient method for solving these max-min problems to local optimality and argue this yields valid discretizations without sacrificing global optimality guarantees. Finally, we formulate optimality-based $\textit{generalized}$ discretization of SIPs as max-min problems and design efficient local optimization algorithms to solve them approximately. Numerical experiments on test instances from the literature demonstrate that our new optimality-based discretization methods can significantly reduce the number of iterations for convergence relative to the classical feasibility-based method.

To operate process engineering systems in a safe and reliable manner, predictive models are often used in decision making. In many cases, these are mechanistic first principles models which aim to accurately describe the process. In practice, the parameters of these models need to be tuned to the process conditions at hand. If the conditions change, which is common in practice, the model becomes inaccurate and needs to be re-tuned. In this paper, we propose a hybrid modeling machine learning framework that allows tuning first principles models to process conditions using two different types of Bayesian Neural Networks. Our approach not only estimates the expected values of the first principles model parameters but also quantifies the uncertainty of these estimates. Such an approach of hybrid machine learning modeling is not yet well described in the literature, so we believe this paper will provide an additional angle at which hybrid machine learning modeling of physical systems can be considered. As an example, we choose a multiphase pipe flow process for which we constructed a three-phase steady state model based on the drift-flux approach which can be used for modeling of pipe and well flow behavior in oil and gas production systems with or without the neural network tuning. In the simulation results, we show how uncertainty estimates of the resulting hybrid models can be used to make better operation decisions., Comment: 20 pages, 10 figures, not published in any journal

We propose a new interconnection relation for infinite-dimensional port-Hamiltonian systems that enables the coupling of ports with different spatial dimensions by integrating over the the surplus dimensions. To show the practical relevance, we apply this interconnection to a model system of an actively cooled gas turbine blade. We also show that this interconnection relation behaves well with respect to a discretization in finite element space, ensuring its usability for practical applications., Comment: 6 pages, 1 figure. Submitted to Applied Mathematics Letters

Antennas for wireless body area networks (WBAN) need to be modeled with adapted methods because the coupling with the body tissue does not allow for a clear separation between antenna and channel. Especially for dynamically varying on-body channels due to changing body poses, e.g. with head-worn antennas, modeling is challenging and design goals for optimal antennas are difficult to determine. Therefore, in this paper, the modeling of WBAN channels using spherical wave functions (SWF) is utilized for antenna de-embedding and for deriving optimal antenna characteristics that maximize the transmission coefficient for the respective channel. It is evaluated how typical factors influencing WBAN channels (different body anatomies, body postures, and varying positions of the communication nodes), can be modeled statistically with SWF. An optimized antenna design is developed based on the derived optimization method, specifically adapted to the channel of on-body links with eye-wear applications. The results with the optimized antenna are compared to other standard antenna designs and validated against measurements.

Neural differential equations have recently emerged as a flexible data-driven/hybrid approach to model time-series data. This work experimentally demonstrates that if the data contains oscillations, then standard fitting of a neural differential equation may result in a flattened out trajectory that fails to describe the data. We then introduce the multiple shooting method and present successful demonstrations of this method for the fitting of a neural differential equation to two datasets (synthetic and experimental) that the standard approach fails to fit. Constraints introduced by multiple shooting can be satisfied using a penalty or augmented Lagrangian method., Comment: in IEEE Control Systems Letters