Spark Plasma Sintering (SPS) has become an increasingly adopted process for hexagonal boron nitride (h-BN) manufacturing. In the SPS process, interparticle diffusion of compressed particles is rapidly achieved due to the concept of Joule heating. Compared to traditional and additive manufacturing (AM) techniques, SPS provides unique control of the structural and microstructural features of Ca components. In this study, the effect of adding calcium oxide to the spark plasma sintering of two commercial hexagonal boron nitride powders, at 1600 ºC, on microstructure, density and dielectric constant was investigated. Scanning electronic microscopy analysis of the sintered h-BN containing 5%wt of CaO samples shows most grains perpendicularly aligned to the pressure direction. Apparent density was 92.2% and 94.4% for the samples originated by h-BN powder with higher and lower initial medium particle sizes, respectively. The corresponding dielectric constant of these samples was 3 and 4.7, indicating a strong relationship of this property to microstructure and density.

This study focuses on predicting temperature using multivariate time series at two locations in Tungurahua province, Ecuador: Mula Corral, situated in the high-altitude grassland ecosystem known as paramo, and Airport Ambato, located in the city center. The prediction integrates key variables such as humidity, precipitation, and wind speed through multivariate neural networks. Six neural network architectures were evaluated: LSTM, Bidirectional LSTM, LSTM with Attention Mechanism, GRU, CNN, and CNN-LSTM. At the Airport Ambato station, the LSTM with Attention Mechanism was the most effective, achieving an RMSE of 0.82 and an R-squared of 0.66. For Mula Corral, the LSTM and GRU networks excelled, with an RMSE of 0.63 and an R-squared of 0.70 for the LSTM, while the GRU achieved an RMSE of 0.89 and an R-squared of 0.60. Conversely, the CNN-LSTM demonstrated the lowest performance, particularly in Mula Corral, with an RMSE of 0.75 and an R-squared of 0.53. These outcomes highlight the superiority of LSTM networks, especially those equipped with an Attention Mechanism, and reveal a preference for Recurrent Neural Networks (RNN) over Convolutional Neural Networks (CNN) for predicting temperatures in sensitive areas such as paramos. The ability to predict temperature with multivariate time series is critical for planning and implementing restoration techniques in the paramos, adapting to climate changes, and maximizing success in the conservation and recovery of these vital ecosystems. The neural network analysis underscores significant climatic differences between the paramo and the city. Mula Corral exhibits lower and more stable temperatures, consistent with the cold, uniform conditions of high-altitude grasslands, whereas the Ambato airport station reflects higher temperatures with greater variability, likely due to urbanization and human activity. These distinctions highlight the importance of accurate neural network models for environmental management, strategic planning, and climate change adaptation.

IntroductionWith no doubt is the prevalence of dementia skyrocketing, with Alzheimer’s, Frontotemporal, and many more which for the patients, the incidence of dementia doubles every 5 years from the ages 65-90 (Corrada et al., 2010). Dementia is also an incredibly difficult condition to manage, with caretakers found to have higher rates of burden and mental health issues, poor physical health, financial problems, and social isolation (Brodaty and Donkin, 2009). Dementia by itself is also characterized by neurological deterioration over time, with a great deal of cognitive functioning being lost in the process. There is no doubt that most caretakers are generally less experienced than traditional hospital staff in managing those with dementia, and so this leaves them open minded to new approaches to help treat their patients (Ma et al., 2024).

Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by difficulties in speech, communication, social interaction, and repetitive behaviors. These challenges hinder language and cognitive development. ASD manifests itself in a wide range of symptoms of varying severity, making it difficult to pinpoint a single cause. Despite numerous efforts, the underlying cause of autism remains unknown. It has been hypothesized that ASD is due to impairment in prediction. The brain constantly processes information and makes predictions about the natural world, which is essential for functioning and habituation. When prediction mechanisms are impaired, the world may feel overwhelming, unpredictable and magical. This has led to the hypothesis about prediction impairment. However, the ability to predict in individuals with ASD has not been extensively studied, particularly across different age groups. In this paper, we developed a machine learning based video game to evaluate the prediction abilities of children with ASD and neurotypical (NT) participants, across two different age groups. We selected subjects from 4 to 7 years and 8 to 12 years, based on the severity of their condition. The study included 12 ASD and 12 NT children aged 4 to 7 years, and 16 ASD and 16 NT children aged 8 to 12 years, all chosen based on specific inclusion and exclusion criteria. Results showed that prediction abilities in the 4-7 age group were significantly lower in ASD children compared to their NT counterparts. However, in the 8-12 age group, ASD children's prediction abilities improved significantly, almost matching NT children in some cases. The study also compared ASD and NT participants within each age group, across various parameters such as success rate, prediction response rate, and accurate response rate. It is observed that the gap between ASD and NT subjects closed significantly for all three parameters with age progression. Also, preliminary usability studies suggest that this tablet-based system has the potential to enhance task performance making it a valuable complementary tool for therapists.

We focus on stochastic geometry analysis of a low Earth orbit (LEO) narrowband terrestrial-satellite uplink with satellite base stations (SBSs) in a uniform constellation equipped with narrow Gaussian beams. The served and interfering omnidirectional user equipments (UEs) are distributed on the Earth's surface according to a homogeneous Poisson point process (HPPP) with Nakagami faded signals. This study presents a detailed but comprehensive mathematical analysis of several key metrics: the signal-to-interference ratio (SIR), the SIR meta distribution (MD), the signal-to-interference-plusnoise ratio (SINR), and the average throughput. Many results are presented in simple analytical and closed forms containing more insight than the expressions proposed in prior works. The results indicate an optimal UE density depending on the altitude, elevation angle, and the width of the antenna gain, maximizing the average throughput. However, this optimal density leads to a significant variance in the user experience regarding link quality (i.e., the users are not treated fairly).

As a generalization of the continuous-time pipeline (CTP), we present a theoretical model that can be used to analyze any conventional continuous-time (CT) analog-to-digital converter (ADC). In addition to providing a unified approach for deriving reconstruction filter coefficients, the general model enables new ADC architectures that do not fit in the framework of neither the cascaded continuous-time sigma-delta modulators (CT-Σ∆Ms), nor the CTP ADCs. As an example exploiting this design flexibility, we present a 7th order Leapfrog ADC. The aggressive 7th order filter enables a low oversampling ratio (OSR), while stability is ensured by local digital feedback with single-bit quantizers. The modular structure makes it well-suited for a programmatic design methodology, and the layout of the ADC is generated using a custom-made Python tool for analog layout compilation. Implemented in a 130nm CMOS technology, the prototype achieves a dynamic range of 82 dB with an OSR of 11, while consuming 2.5 mW from a 1.5 V supply. The resulting Schreier figure-of-merit (FOM) is 168 dB.

The maritime industry is responsible for the transportation of over 80% of worldwide commodities. Currently, the industry heavily relies on traditional maintenance strategies, often leading to excessive and preventable downtime events, failing to ensure operational integrity and safety of the vessels. Machine learning (ML)-based predictive maintenance (PdM) emerges as a promising solution, utilizing vessel related data-driven insights to proactively predict machinery failures and optimize maintenance schedules. Towards this direction, this survey presents state-of-the-art PdM solutions, providing insights into different implementation approaches and their target components, while also mapping available datasets. Furthermore, considering the significant limitations of centralized ML (CML) in data privacy, security and scalability, the paper also explores decentralized architectures based on federated learning (FL). Finally, the paper presents a comparative analysis of the advantages and disadvantages of both CML and FL, with a particular focus on the potential benefits of FL for maritime PdM adoption, while also addressing open issues and future research directions in the field.

Climate change-induced natural disasters are seriously affecting the planet's ecosystem and disrupting the socioeconomic development of humanity. This exigent issue leads to the developing of novel technologies and methodologies to utilize sustainable energy sources such as geothermal energy for energy generation. However, geothermal reservoirs are highly locationspecific and identifying the potential zones is a daunting task for reservoir engineers. Moreover, the geothermal energy density of a place can change over time due to various environmental factors; therefore, it is crucial to monitor and assess its changes periodically. Concurrently, the accelerating advancements in the research of sixth-generation (6G) wireless networks and their possible disruptive technologies, such as integrated sensing and communications (ISAC), have recently received much attention from the broader research communities. In this article, we envision and introduce the general concept of a generalized geothermal sensing and monitoring (GeoSM) model and explore the role of the ISAC system in the localization and periodic monitoring of geothermal energy. Specifically, we illustrate the envisioned system architecture for the GeoSM model, present a general framework of the 6G-enabled ISAC system, and suggest possible enabling technologies to facilitate the distributed sensing and monitoring of the geothermal energy zones. Then, we describe the geothermal heat network as a prospective use case of the GeoSM model. Following that, we present a preliminary case study on the waveform optimization of our proposed ISAC framework of the envisioned GeoSM model for a specific application scenario. Finally, we outline the open research challenges and discuss possible future research directions.

We present an alternative approach to optical tactile sensor design that uses a non-camera-based sensing principle that we term light vector (LiVec) sensing. We propose to present a poster that provides an overarching summary of our work on this design concept to date. Namely, we will review: (i) a proof-of-principle single-element design; (ii) the LiVec Finger design, which consists of an array of sensing elements arranged in a finger pad shape; (iii) work-in-progress toward designs using flexible printed circuit boards to achieve non-planar sensor designs.

We examine the impact of samarium doping on the Curie temperature ( Tc ) of magnetite through density functional theory (DFT) calculations. Upon calculating the total energies of the different spin orientations among the cations in Fe3O 4:Sm, we realized that the Sm atom prefers to substitute an Fe from the octahedral site with a spin opposing that of the atom it replaces. Our results show that Sm doping weakens the ferrimagnetic J coupling between the octahedral and tetrahedral Fe atoms. As a result, the normalized magnetization profile across a broad temperature range shows that the samarium doped compound ( Fe3O 4:Sm ) has a Tc of approximately 319 K, which aligns well with the target range for self-regulated magnetic nanoparticle hyperthermia (MNH) applications. Furthermore, we demonstrated that Sm doping in magnetite with high electron concentrations of 1022 cm −3 and 1023 cm −3 almost nearly preserves the Hall coefficient (RH), implying that Fe3O 4:Sm can be synthesized without significantly altering magnetite's ability for tumor tissue identification based on the Hall effect.

The cross-fertilization of the fast-developing AI technology and spatial indexing has given rise to spatial learned indexes. However, these indexes rely on historical data distributions to build models, which limits their ability to anticipate data that has not yet arrived. To address this, we propose a novel Spatio-Temporal Update Method (STUM) that enhances conventional spatial learned indexes by introducing a Spatial Delta Area (SDA) for updates without altering their hierarchical structure. STUM learns spatio-temporal auto-correlation from historical data and integrates predicted future distributions. We apply STUM to the Spatial Learned Block Range INdex (SLBRIN), resulting in the development of the Spatio-Temporal Updatable learned Block Range INdex (STUBRIN), which adopts Revmap to integrate spatio-temporal sequence predictions with the spatial block range. STUBRIN optimizes the retraining process by learning the temporal continuity from spatial distribution and fusing it into the error threshold control mechanism and historical delta learning mechanism. Our results show that STUBRIN achieves 1.9-2.4×, 1.8-13.3×, 3.4-6.7× better build, query and update performance compared to state-ofthe-art methods. Additionally, STUBRIN offers superior query and update stability. For concurrent learned indexes, we have also designed parallel scheduling for STUBRIN, which improves build, query and update performance by 6.2-6.8×, 0.3-4.2×, 2.6-5.5×, without increasing the index size.

Conventional lithium-ion batteries utilize liquid electrolyte which holds several safety issues including electrolyte leakage, internal short circuits, and explosion amongst other problems, which limits their use. Solid-state (SSEs) and gel-polymer (GPEs) electrolytes in lithium-ion batteries provide excellent safety efficiency and have been on the rise in the lithium battery industry. Poly (Ethylene Oxide) (PEO) and Poly (vinylidene fluoridehexafluoropropylene) (PVDF-HFP) are commonly used compounds in the electrolyte due to their excellent mechanical and electrochemical properties. This paper reviews the use of PEO and PVDF-HFP, along with other additives, in the electrolyte and assesses the battery performance on the metrics of crystallinity, ionic conductivity, discharge-specific capacity, capacity retention, and coloumbic efficiency. With the addition of boron nitride and polymerized dopamine in PVDF-HFP-based GPEs, the electrochemical metrics show promising results in replacing liquid electrolytes in the near future.

The rapid growth of Large Language Models (LLMs) has revolutionized business operations across sectors such as finance, healthcare, customer service, and more. These models are no longer experimental but are becoming core components in streamlining workflows and decision-making. However, the integration of LLMs brings new challenges: ensuring the accuracy of outputs, monitoring model performance, and maintaining strict compliance with industry regulations[1]. This white paper provides an in-depth review of leading LLM tracing and evaluation tools, examining their strengths, limitations, and relevance for real-world use cases. We aim to guide organizations in choosing the right tools to enhance model performance, ensure compliance with data security protocols, and generate actionable insights.

This article investigates the essential cultural, gender, and contextual elements that shape the educational experiences and volunteerism attitudes of Arab and non-Arab students.

This article presents a system of robot identifiers that taxonomically classify and identify robotic products based on their fundamental attributes as autonomous systems. Robotic products have grown increasingly complex, diverse, autonomous, and pervasively ubiquitous in recent years, rapidly outgrowing traditional definitions and categorization standards. New generations of robotic products are no longer limited to traditional mechatronic devices, expanding in various forms such as cyber-physical systems, software platforms, nano and organismic systems, soft and flexible robots, and developmental robotics. Without effective tools to identify and compare their fundamental differences, not only is efficient development stifled due to duplicated and entrenched efforts, but social adoption and safety are also impacted. This article addresses these challenges and provides key contributions in three areas: analysis, synthesis, and application. First, it examines in depth the current industry standards and academic literature, identifying major issues in building an effective taxonomy. Second, it establishes fundamental principles to classify broad classes of machines and develops a coding system for consistent naming, identification, classification, and organization. Third, it provides use cases to demonstrate the utility of the proposed system. The result is a unified taxonomic framework that identifies robots and autonomous systems from the perspectives of anthropogenic, ecological, and phylogenetic factors. This biologically grounded, interdisciplinary approach offers a robust and inclusive framework that encompasses a broad spectrum of technologies, including software platforms and nano robots, thereby future-proofing the taxonomy in anticipation of the imminent emergence of intelligent robots with new capabilities.

This study introduces a novel deep learning framework for analyzing tobacco-related content on social media platforms like YouTube and TikTok. Using a dataset of 5,730 videos, we developed and tested a computer vision-based system that effectively combines visual and textual analysis. The framework achieved 83% accuracy in content classification and strong performance in sentiment analysis (0.79), frame consistency (0.75), and visual-textual integration (0.81). These results demonstrate the framework's potential as a tool for public health surveillance and policy enforcement in monitoring tobacco-related social media content, particularly among youth audiences.

In this letter, we describe a method for transmitting multiple input multiple output (MIMO) orthogonal time frequency space (OTFS) and discrete Fourier transform spread orthogonal frequency division multiplexing (DFT-s-OFDM) in 5G new radio (NR) resource grid and also present their corresponding receivers. These receivers, which use channel estimates obtained from demodulation reference signal (DMRS) pilots of the resource grid, perform iterative MIMO LMMSE equalization using the soft output of the channel decoder. We find that OTFS and DFT-s-OFDM mapped to 5G NR resource grid outperform the existing 5G NR waveform in terms of coded error performance.

Over the past decade, there has been a significant increase in interest in digital twin (DT) technology in a variety of domains. While research on DTs of single assets was initially prevalent, there has been a notable shift towards distributed systems of DTs, which connect to each other to collaborate. Typically, collaboration is enabled by DTs providing services that can be consumed by other DTs. In service-oriented systems, a service is typically identified by type information. However, this is not sufficient in distributed DT systems, where DTs associated with different physical entities may provide the same type of service. Consequently, selecting the appropriate service depends not only on the service type, but also on the associated physical entity. However, requiring DTs to know the mapping of services to their physical environment is not feasible for large dynamic systems. This paper presents a novel proximity-based service discovery method that allows DTs to select services based on service type and their proximity to other objects. That is, service specifications are fully abstracted from the mapping of services to physical objects, relieving DTs from maintaining information about this mapping. Furthermore, service discovery is robust to changes in the physical environment and service population. The proposed service discovery method has been implemented on top of a spatial DBMS. We argue that this implementation is optimal in terms of network utilization and latency, and perform comprehensive evaluations to show the performance of discovery queries as a function of their complexity.