Microscale robotics represents a promising future for minimally invasive medicine. However, one of the biggest challenges of microrobots moving through the human body is represented by the complex 3D structure of biological lumina and tissues, which obstructs the navigation of micron-sized devices. Here, we fabricate ultra-deformable magnetic microrobots, consisting of ferrofluid-loaded lipid vesicles, and we magnetically pull them through chambers that exert upon them a gradually more forceful confinement. We thus analyze their capability to face interstices comparable to or smaller than their characteristic size and their consequent behavior in terms of stability, velocity, and deformation. The results show that the inherent compliance of these vesicle-based magnetic microrobots allows them to infiltrate successfully in interstices slightly smaller than their size. Further enhancement of their compliance and the development of specific control strategies may lead to microrobots able to move through interstices and traverse complex biological environments.This article has been submitted to IEEE Transactions on Medical Robotics and Bionics (manuscript TMRB-07-24-BR-0017) and presented at the IEEE RAS EMBS 10th International Conference on Biomedical Robotics and Biomechatronics (BioRob 2024), 1-4 September 2024 Heidelberg, Germany.

Offshore wind is set to be competitive with fossil fuels within the next decade. The International Energy Agency estimates in its Stated Policy Scenario from 2019 that the Levelized cost of energy (LCoE) from offshore wind will decrease 40 % by 2030, where transmission infrastructure accounts for one-third of the costs. The modular high-voltage direct current (ModHVDC) generator connects converters in a series of modules for a segmented HVDC generator, which limits the number of conversion steps in comparison to a typical offshore wind farm. This paper aims to validate the thermal design proposal for ModHVDC generator. It is especially important for larger generators utilised in wind- or hydro-power applications, as replacing components is complex, time-consuming, and expensive. Thermal management in electric machines is important, as poor heat dissipation leads to a loss of efficiency and can severely reduce the expected lifetime of the machine. In this paper, an analytical approach to determine the heat loss coefficients for the machine is deduced. Magnet, copper, and iron losses are provided to estimate the heat generated in different parts of the machine. Two different core loss calculation methods are used. The analysis shows that the temperature does not exceed critical levels for both methods. The end results prove that the design solution is viable with the use of forced convection for cooling in the air gap with maximum temperatures of 66.8 °C in the permanent magnets and rotor.  

Differential microphone arrays (DMAs) use multiple omnidirectional microphones for synthesising higher-order microphone directivity patterns. In their most basic form, they can be used to obtain fixed-directivity or horizontally steerable beamformers that can satisfy certain constraints. We propose a vector differential microphone array (VDMA) which is frequency-and direction-invariantly steerable in three dimensions. The proposed design comprises pressure and particle velocity sensors positioned on a circular constellation in a plane and allows extracting the third-order spherical harmonic decomposition of the sound field. This decomposition can then be used to obtain spherically directioninvariant steered beams. Synthesis of a maximum directivity factor (MaxDF) directivity pattern is demonstrated. A closed-form expression for the proposed array's white noise gain (WNG) is derived. The robustness of the proposed design to noise is analysed.

This extensive analysis examines the role of artificial intelligence (AI) in contemporary society, specifically emphasizing its influence through intelligent devices and services. As AI technology becomes more extensively incorporated into different facets of everyday existence, it has favourable and unfavorable consequences. AI has significantly transformed convenience and efficiency by introducing smart gadgets such as smartphones and wearable technology into our daily lives. These technologies have become essential in our routines, enabling us to communicate, monitor our health, and access information. AI-powered services have revolutionized sectors such as healthcare, transportation, and customer service, resulting in enhanced results and increased availability. Nevertheless, these progressions are accompanied by substantial ethical and societal concerns. One major concern is the impact of AI on job opportunities, as it can automate tasks and replace occupations. This might result in economic disruptions and create a larger divide between individuals with advanced skills and those without. This paper also discusses moral issues, namely privacy and algorithmic bias. It highlights that the utilization of AI in surveillance and data gathering may harm individual private rights, while algorithms have the potential to perpetuate or intensify preexisting biases. The assessment examines the policy obstacles and recommends the implementation of frameworks that guarantee the responsible utilization of AI, fair distribution of advantages, and safeguards for privacy, employment, and impartiality. The review seeks to provide a more balanced understanding and assist the creation of policies that can limit risks and promote responsible integration of AI into society by assessing both the beneficial implications and possible risks of AI technology.

As artificial intelligence (AI) systems continue their rapid advancement, a framework for contextualising the major transitional phases in the development of machine intellect becomes increasingly vital. This paper proposes a novel chronological classification scheme to characterise the key temporal stages in AI evolution. The Prenoëtic era, spanning all of history prior to the year 2020, is defined as the preliminary phase before substantive artificial intellect manifestations. The Protonoëtic period, which humanity has recently entered, denotes the initial emergence of advanced foundation models exceeding human capacities within specialised domains. The forthcoming Mesonoëtic epoch is anticipated to commence with the advent of Artificial General Intelligence (AGI), potentially facilitated by quantum computing capabilities. Ultimately, the Kainonoëtic age is posited to begin upon the rise of superintelligent systems, likely catalysed by an AGI undergoing recursive self-improvement. This novel period taxonomy provides a structured conceptualisation for the key milestones in the evolution towards advanced artificial intellect.This classification framework offers significant value across various disciplines, researchers and stakeholders in the AI  landscape, by providing a clearly defined chronological nomenclature. AI policy-makers and regulators will find it instrumental in crafting stage-appropriate policy and legal interventions. Scientists and academics will benefit from a standardised terminology in the approach to AI periodisation, facilitating unambiguous communication and collaborative research. Finally, this proposed taxonomy can serve as an accessible tool for public understanding, offering a clearer roadmap of AI development that can inform open discourse and decision-making.

Implementing pilot SF 6-free DC gas insulated switchgear (GIS) connected to the rectifier station poses challenges due to the combination of discharge severity metrics, harmonic control, and gas characteristics. This paper examines the impact of DC harmonic voltages produced by line-commutated converters (LCC) rectification on corona discharge phenomena in GIS. With a focus on the various types of harmonic distortions that present electric field fluctuations, this study evaluates corona discharge behaviour in C 4 F 7 N mixture, CO 2 and technical air under overstressing ripple voltages. Comprehensive descriptions of the methodology encompass: (i) reproducing representative DC terminal voltages of a generic LCC converter in a laboratory setting, and (ii) correlating the harmonics impact of these harmonics with discharge analytics using updated time-resolved analysis. Compared with pure DC voltage, it was found that the LCC harmonic discharge results demonstrate a non-monotonic increase in discharge inception voltage with increasing firing angles under fixed overlap angles, also, a strong correlation between voltage rate of change and discharge pulse sequence patterns was seen for the three gases. The concepts derived from pulse-pulse correlation are explored as indicators of the influence of commutation control logic in discharge generation, emphasising their utility in diagnosing DC-GIS in rectifier stations.

The intersection of robotics and artificial intelligence led to a profound paradigm shift in Robot Learning. Robots have the capacity to replicate human actions and also dynamically adapt, innovate, and excel across a spectrum of tasks. However, the heterogeneity in the deployment of robot platforms and software frameworks poses considerable challenges in terms of systematic testing and comparative analyses. Additionally, the data scarcity of especially force controlled robot manipulation is still restraining the development of advanced foundation models. A reference platform with default software stack can help to increase comparability, reducing development time and collect a large amount of tactile robot manipulation data. To address on this problem, we developed a Parallel and Distributed Robot AI (PD.RAI) framework, comprising a scalable ensemble of Robot Learning Units (RLUs), a global database, and the Robot Cluster Intelligence (RoCI). Each RLU is endowed with robot arms, cameras, and local computational units to autonomously engage in planning, control, and local machine learning of tactile manipulation skills. The RoCI system oversees the learning process and schedules the RLUs tasks. To show the functionality of the system, two black-box optimization algorithms are compared within the robot skill learning domain. An experiment with 24 different optimization tasks is conducted in parallel. The algorithms are incorporated into the same existing default modules acting as a reference environment. This allows for a realistic comparison without sacrificing diversity of possible configurations and testing environments.

In this study, we investigate the optical parameters of plasmonic Titanium Nitride (TiN) thin films using the RefFIT software, employing the Drude-Lorentz model for data fitting. TiN is an emerging material for plasmonic applications due to its high melting point, wear resistance, and stability at various temperatures, positioning it as a viable alternative to traditional noble metals. Our research focuses on fabricating, characterizing and extracting optical and electrical properties of TiN thin films deposited on silicon wafers using ellipsometry and Fourier Transform Infrared (FTIR) spectroscopy.We analyze six different samples, varying in thickness and composition, to determine key parameters such as plasma frequency and damping constants. The simultaneous fitting of reflectance and ellipsometry data in RefFIT provided results consistent with existing literature, establishing the software’s reliability for such analysis. Additionally, we validated the RefFIT results by comparing them with alternative models and experimental data, demonstrating the necessity of incorporating Lorentz terms alongside Drude components for accurate optical characterization.Our findings reveal that TiN films exhibit desirable plasmonic properties, with carrier concentration and relaxation times aligning with known trends in nitrides. The robustness of RefFIT in modeling these properties confirms its utility for future research in plasmonic materials. This study underscores the potential of TiN in advanced plasmonic devices and encourages further exploration using comprehensive modeling techniques.

This paper explores the design and analysis of planar multilayer Metal-Insulator-Metal (MIM) stacks as potential selective emitters for thermophotovoltaic (TPV) systems, specifically targeting medium-grade waste heat applications. These selective emitters are integral to TPV systems, which convert heat energy into electrical power. The study leverages the Transfer Matrix Method (TMM) to evaluate the optical properties of various MIM configurations, focusing on their absorption spectra. Materials such as Tungsten, Aluminum, Gold, Silver and Titanium Nitride were analyzed in combination with Hafnia as the dielectric layer. The results highlight that Silver/Hafnia and Gold/Hafnia stacks exhibit the most favorable narrowband absorption characteristics, matching the bandgap wavelength of Germanium converter cells. Additionally, Aluminum/Silica multilayer structures were optimized for different converter cells, showing potential in medium-grade heat recovery applications. The findings emphasize the importance of material selection and layer thickness in achieving high-efficiency TPV systems. By analyzing the spectral response, the study aims to optimize design parameters to enhance photon-to-current conversion and minimize energy losses in TPV cells.

Ransomware is one of the most critical cybersecurity threats, targeting systems globally and causing substantial financial and data losses. Developing an effective and efficient detection mechanism within operating systems is essential for mitigating these risks, which is where the novel integration of LSTM-based machine learning within the Linux kernel offers a significant advancement. The proposed approach involves the deployment of an LSTM model capable of analyzing system call sequences in realtime, identifying anomalous behavior indicative of ransomware activities with high accuracy. The research outlines the model's design, including data collection, preprocessing, and the intricate integration process into the Linux kernel to facilitate seamless communication and classification. Extensive experimental evaluation demonstrated that the LSTM model achieved superior detection performance when compared to traditional methods, maintaining a high level of accuracy and low false positive rates. The findings indicate that this method not only provides a proactive defense mechanism against emerging ransomware threats but also enhances the overall security framework of Linux systems through its adaptive learning capabilities and efficient resource utilization.

This paper presents a robust adaptive control design for robot manipulators to track desired trajectories amid unknown disturbances and input saturation. The suggested controller integrates backstepping and super-twisting techniques, ensuring system stability and robustness. The backstepping method mitigates unmatched disturbances in a two-step process, while the super-twisting algorithm addresses matched perturbations and overshoot apparitions. A nonlinear observer enhances control efficacy against matched disturbances and input saturation, ensuring fast convergence via a quasi-non-singular terminal sliding surface. This approach enables precise tracking with smooth control signals and avoids large feedback gains. An advanced adaptive reaching law dynamically adjusts the controller's behavior through a potential function, mimicking and enhancing various established reaching control laws. The designed method provides a flexible strategy with rapid convergence, minimal chattering, and adaptability to variation of system dynamics. Stability is confirmed using Lyapunov's direct method, proving uniform boundedness of signals in the closed-loop system. The proposed controller was validated through simulations, experiments, and comparative analysis, demonstrating its superior performance.

Exoskeleton robots hold immense promise in rehabilitation, serving as crucial aids for patient mobility and exercise. However, harnessing their capabilities requires overcoming significant control challenges arising from complex nonlinear dynamics and uncertainties in both models and actuators. This paper introduces a novel adaptive backstepping controller designed specifically for exoskeleton robots navigating uncertain dynamics and actuator parameters. Unlike conventional approaches that rely on basis functions, the proposed controller integrates a Modified Function Approximation Technique (MFAT) to approximate dynamic parameters without the need for such functions. The MFAT effectively manages mismatched perturbations, while the backstepping control compensates for uncertainties associated with state variables, enhancing resilience to disturbances, especially in scenarios where the exoskeleton's dynamics model is unknown. Lyapunov stability analysis ensures uniformly ultimately bounded signals within the closed-loop system. A comparative study conducted on the industrial robot IRB 120, validates the effectiveness of the proposed controller and highlights its superior performance. Real-time implementation on a 7 Degrees of Freedom (DOFs) wearable robot named ETS-MARSE confirms the efficiency of the control algorithm. The results from simulations and experiments underscore the efficacy of the proposed approach. The insights gained from this research pave the way to unlocking the full potential of exoskeleton robots in rehabilitation settings, promising improved patient outcomes and advancing human-technology interaction to new heights.

This paper introduces Conjugate Neural Networks (CoNNs), a novel architecture designed to enhance robustness against adversarial attacks. Adversarial attacks pose significant security challenges to neural networks by manipulating input data to cause incorrect model predictions. The proposed CoNNs architecture leverages a dual-model framework where two complementary neural networks are trained in tandem, each using adversarial samples generated by the other, to cancel out adversarial effects and improve overall resilience. We evaluate CoNNs using various deep learning models, including MLP, ResNet, LSTM, and Agricultural-Informed Neural Networks (AINN), across multiple datasets such as Fashion-MNIST, CIFAR10, the data generated by Lorenz system, and agricultural N 2 O emissions data. The results demonstrate that CoNNs significantly outperform traditional single neural network in resisting both Fast Gradient Sign Method (FGSM) and Basic Iterative Method (BIM) attacks. CoNNs maintain higher accuracy and exhibit greater stability under adversarial conditions, making them a promising approach for improving the robustness of neural network-based systems.

Although modern imaging technologies allow us to study connectivity between two distinct brain regions in-vivo, an in-depth understanding of how anatomical structure supports brain function and how spontaneous functional fluctuations support cognition is still elusive. Meanwhile, tremendous efforts have been made in the realm of machine learning to establish the nonlinear mapping between neuroimaging data and phenotypic traits. However, the absence of neuroscience insight in the current machine learning approaches poses significant challenges in understanding cognitive behavior from transient neural activities. To address this challenge, we put the spotlight on the coupling mechanism of structural connectivity (SC) and functional connectivity (FC) by reformulating this unresolved network neuroscience question into a graph representation learning problem for pathways. Specifically, we introduce the concept of topological detour to characterize how a ubiquitous instance of FC (direct link) is supported by neural pathways (detour) physically wired by SC, which forms a cyclic loop interacted by brain structure and function. The multi-hop detour pathway underlying SC-FC coupling allows us to devise a novel multi-head self-attention within a Transformer. Taken together, we propose a biological-inspired deep model, coined as NeuroDetour , to find putative connectomic feature representations from the unprecedented amount of neuroimaging data, which can be plugged into various downstream applications such as task recognition and disease diagnosis. We have evaluated NeuroDetour on large-scale public datasets including Human Connectome Project (HCP) and UK Biobank (UKB) under different experiment settings of supervised and zero-shot learning, where our NeuroDetour demonstrates state-of-the-art performance.

Sequential manipulation planning has been a critical imperative to achieve a higher level of autonomy in robotics. Classical approaches to address task planning problems are based on symbolic formalisms, such as Planning Domain Definition Language (PDDL) [1], and search for state transition plans to reach task goals. In practice, such task plans are often programmed as Finite State Machines (FSMs), which incorporate expert knowledge specifying control and execution details. Due to its limitation of scalability [2], Behavior trees (BTs), which represent policies in a state-less, hierarchical tree structure, have gained increasing popularity for complex task planning. Its advantages of modularity, reusability and reactivity, make it a more desired formalism for long-horizon manipulation tasks.

This paper presents a comprehensive study on the design, analysis, and performance of Metal-Insulator-Metal-Insulator-Metal (MIMIM) multilayer stacks. The behavior of these stacks was modeled using the Transfer Matrix Method (TMM), allowing for a detailed investigation into their optical properties under varying parameters such as refractive index, layer thickness, polarization, and angle of incidence. The results highlight significant trends, including the effects of changing the incidence angle and layer thickness on reflection dips and absorption peaks, which exhibit wavelength-dependent shifts and intensity variations. These finding are particularly relevant for optimizing selective emitters in thermophotovoltaic (TPV) systems, where specific resonant peaks are desirable for efficient energy conversion. By adjusting the thickness of the metal and insulator layers, the study demonstrates how precise control over multilayer parameters can enhance performance, particularly through the strategic tuning of Fabry-Perot cavities within the stacks. The insights provided pave way for future applications in photonic and optoelectronic devices.

Quantum computers, quantum communication systems, and quantum sensors are expected to revolutionize science and technology by exploiting unique properties of quantum systems like superposition and entanglement. Advances in the underlying science have helped identify an ever-broadening set of compelling use cases for these quantum information technologies. Yet, despite experimental advances, translating design concepts to high-performance hardware remains a significant obstacle preventing these technologies from reaching their revolutionary potentials. In mature engineering fields, these kinds of challenges are typically overcome by leveraging first-principles numerical methods that model the underlying physics (e.g., full-wave methods solving Maxwell's equations) so they can be applied uniformly to any device. For quantum information technologies, such numerical methods are only beginning to be created, but there is a significant opportunity for this. Just as electromagnetic (EM) effects impact many classical technologies, a similarity in the hardware platforms for building quantum information technologies is that the interactions of classical and quantum electromagnetic (QEM) fields with atom-like systems used as qubits in these devices play a central role. In this article, we introduce the basic concepts and emerging trends in the field of numerically modeling these effects on conventional computers, which we broadly term "computational quantum electromagnetics (CQEM)". In an effort to keep this article accessible, we have written it assuming the reader has no detailed background in quantum physics and keep our discussion on the broad opportunities and challenges of the computational strategies in this nascent field.

Sustainable Development Goals (SDGs), adopted by the United Nations in 2015, are presented as a blueprint for our world’s social, economic, and environmental progress. The SDGs have a structured framework spread across 17 goals, 169 targets and 232 unique indicators. The socio-economic, and demographic challenges and opportunities are not uniform for all nations. Hence, the agenda of SDG localisation encourages all member states of the UN to establish and customize suitable national frameworks to localise the 17 interlinked goals following their domestic factors. However, SDG localisation requires conceptualising the key concepts of sustainability and sustainable development. An absence of a suitable representation framework makes it challenging to represent, reason, and compare disparate methods and schemes adopted for SDG localisation. In our work, we address modelling challenges for representing and reasoning around sustainability. We argue that sustainability is a “state of being” making it a state-maintenance problem, rather than a goal-achievement problem. Our central principle around sustainability extends Amartya Sen’s capability approach which states that the well-being of people should be understood in terms of people’s capabilities and their functionings. We add sustainability as a separate dimension, asking how resilient a given set of capabilities are and thus propose SDG indicators to be interpreted as capability indicators. Hence, both the sustainability and capabilities of a state need to be maximised to maintain or improve the indicators. Therefore, sustainable development is defined as achieving capability targets and their progressive sustenance.