This study evaluates the effectiveness of a 6-hour Calculus for Engineering preparation course designed for high school seniors (Group A) and incoming first-year engineering students (Group B), utilizing advanced educational technologies. Unlike the traditional 9-15 hour approach at the Engineering Faculty of Rajamangala University of Technology Krungthep (RMUTK), which begins with foundational topics such as numbers, function values, limits, and continuity, our course focuses directly on differentiation using curve slopes and symbolic mathematics. The course employs GeoGebra and SymPy Live web applications to facilitate learning. Paired sample t-tests revealed significant improvements in both groups: Group A (n = 35) showed a mean increase of 4.97 (t(34) =-10.62, p < 0.001), and Group B (n = 27) showed a mean increase of 2.93 (t(26) =-4.05, p < 0.001). An independent samples t-test indicated a significant difference in improvement between the groups (t(60) = 2.47, p = 0.016). The results suggest that the 6-hour calculus preparation course effectively enhanced the calculus competency of both high school seniors and incoming first-year engineering students, with Group A showing greater improvement than Group B. This provides a more efficient alternative to the traditional methods used at RMUTK.

Atmospheric energy harvesting presents a promising solution for sustainable power generation, especially for densely populated cities. This paper explores the integration of triboelectric, electromagnetic, and electrostatic energy harvesting methods, providing a detailed analysis of the underlying physical principles and mathematical models. We derive expressions for the power output of both individual and hybrid systems, considering factors such as environmental conditions, material properties, and system efficiency. Experimental data and computational simulations are discussed to validate the theoretical models. The results suggest that optimized electromagnetic systems offer a significant improvement in energy conversion efficiency, with potential applications in city lighting, fundamental electrical requirements, and electrically powered public transportation. The paper concludes with recommendations for future research directions in atmospheric energy harvesting.

The nuclear spin, which is the fundamental atomic attribute underlying the magnetic resonance imaging, is an exact two-state quantum system. Therefore, it is a suitable candidate to be used as a qubit in quantum computing. The aim of this letter is to review the main equations governing the dynamics of the spin magnetic dipole moment in Magnetic Resonance Imaging. It is also shown that exposing the magnetic dipole moment to a continuous-wave B1-field results in travelling back and forth between its two energy states. It absorbs energy from the background radiation when going from the lower to the higher energy state, and injects the same amount of energy back into its surroundings when relaxing from the higher to the lower energy state. In a complete cycle, the dipole moment behaves therefor purely reactive with zero net energy exchange with its environment.

This investigation introduces an uplink frequencytime (FT) index modulation multiple access (IMMA) (FT-IMMA) scheme designed to augment the spectral efficiency (SE) of conventional orthogonal frequency-division multiple access (OFDMA). The fundamental principle of this approach involves allocating FT resources within OFDMA to users through twodimensional index modulation (IM). A subset of input bits from each user dictates the indices of subcarriers and time slots within the FT resources of OFDMA, while the remaining bits undergo a conventional 𝑀-ary modulation. FT-IMMA obviates the necessity for a prior scheduling mechanism, thereby reducing communication overhead. While this grant-free access approach may potentially introduce data collisions within the system, the collision probability of the proposed model is observed to be markedly lower compared to traditional IMMA systems. However, the two-dimensional IM approach increases the computational complexity of conventional maximum likelihood detection. To address this, the authors propose using orthogonal matching pursuit, a compressive sensing technique that effectively reduces the computational burden. Through analytical investigation and simulation results, FT-IMMA demonstrates superior performance compared to its counterparts, showcasing notable advantages in both bit error rate and SE.

This study introduces a physics-informed diffusion model (PIDM) for super-resolution (SR) reconstruction of optical coherence tomography (OCT) data. Methods: An optimization framework was developed for maximizing the likelihood of observing an OCT image in the dataset, given the super-resolved reconstruction from a physics-informed diffusion model (PIDM) that reverses the degradations in OCT images. The image degradations were modeled as a serialization of three processes accounting for the effects of defocus, speckle noise, and digital sampling in OCT images. An analytical model for lightpropagation model and a statistical model for speckle noise were derived based on the physical properties of the OCT setup. These models were then integrated with a diffusion model to reverse the degradations caused by defocus blur and digital sampling, minimizing susceptibility to noise and defocus-induced artifacts. Results: The proposed method was employed for reconstructing images of a standard resolution target, plant tissue, and in vivo human cornea, using the complex OCT data acquired with a linescan OCT (LS-OCT) system. The results from the PIDM exhibit improved sharpness and contrast compared to the images resulting from a few baseline methods such as standalone superresolution using DM. Conclusion: Complementing DM with the physics of OCT could be a viable solution for obtaining highfidelity SR reconstruction of OCT images. Significance: This work harnesses the power of diffusion models for superresolution in OCT images. Such development could potentially enhance cellular-resolution OCT imaging of ophthalmic tissues, where high-fidelity images are crucial for accurate diagnosis.

The Restricted Hopfield Network (RHN) builds upon the classical Hopfield Neural Network (HNN) by introducing multiple hidden layers and employing the Subspace Rotation Algorithm (SRA) for training. These advancements address the limitations of traditional HNNs, such as limited storage capacity and vulnerability to noise and adversarial perturbations. RHNs with orthogonal weight matrices trained by SRA can form stable attraction basins, enhancing their capacity and adversarial resilience. This paper systematically evaluates RHN against Dense Associative Memory (DAM), Predictive Coding Network (PCN), and Multilayer Perceptron (MLP) in terms of adversarial robustness and retrieval performance using incomplete patterns. Experiments using alphabet and character datasets demonstrate that RHN consistently outperforms the other models across a variety of adversarial scenarios, including Fast Gradient Sign Method (FGSM), Basic Iterative Method (BIM), Projected Gradient Descent (PGD), and Gaussian Noise (GN) attacks. The results show that RHNs trained with SRA not only enhance model capacity but also ensure high retrieval accuracy under adversarial conditions. These findings highlight the robustness and versatility of RHN as an effective solution for adversarially resilient memory systems.

An effective passive source localization technique is proposed as essential for industrial applications. Extensive research has been conducted on multi-station passive localization using two-dimensional angle of arrival (2D AOA), time difference of arrival (TDOA), and their combination. However, certain practical factors limit the observation to only a restricted set of one-dimensional angles of arrival (1D AOA). Additionally, there may be perturbations in the stations' locations. To address these issues, this paper proposes a localization model that hybrid a single 1D AOA with multiple TDOA measurements, taking into account the perturbations in station locations. For clarity, we refer to this combination as hybrid measurements. Two types of localization algorithms are proposed for the hybrid measurement model. The first type approximates the localization problem as a non-convex constrained weighted least squares optimization problem. We introduce two solution algorithms: the two-step weighted least squares algorithm (TWLS) and the semi-infinite relaxation method (SDR), both of which do not require an initial guess. The second type directly transforms the localization problem into a maximum likelihood estimation (MLE) minimum optimization problem. Several iterative algorithms are introduced, including iterative weighted least squares based on Taylor expansion (IRLS-TE), Gaussian Newton (GN), adaptive moment estimation (Adam), and simplification adaptive moment estimation (SAdam). Unlike the first type, these algorithms require an initial guess. Furthermore, we derive the Cramér-Rao Lower Bound (CRLB) for hybrid measurements and analyze the computational complexity and minimum station configuration needed for each solution algorithm. Simulations and the SWellEx96-S5 ocean acoustic experiments demonstrate the effectiveness of the proposed methods. The results indicate that the proposed algorithms can achieve CRLB performance at low noise levels, with the TWLS algorithm showing the lowest complexity, while the Adam algorithm achieves the highest accuracy. These proposed solution algorithms can be selected based on computational complexity, localization accuracy and source of interest location.

This paper explores the cost analysis of vehicle recalls between 2010 and August 2024, focusing on Ford, Tesla, and Toyota. The study highlights Tesla's significantly lower recall volumes and costs compared to traditional automakers, despite rapid growth in the electric vehicle market. This research investigates the importance of integrating software-based solutions and vertical manufacturing processes to reduce recall costs. The approach involved analyzing recall data from the National Highway Traffic Safety Administration (NHTSA). The analysis is primarily based on categories such as air bags, electrical systems, and powertrain issues, considering other major concerns for each automaker and their financial implications. The results show that Tesla's reliance on software updates and fewer mechanical components results in fewer recalls, lower logistics costs, and more efficient issue resolution. In contrast, Ford and Toyota face higher recall volumes due to mechanical complexity. The study concludes that Tesla's innovative approach to recalls sets a benchmark for cost-effective recall management, offering traditional automakers a blueprint for reducing recall costs and improving operational efficiency.

The design of a rural hospital discussed in this article places emphasis on the integration of telehealth services aimed at elevating the standard of healthcare provision in the marginalized areas that lack basic and advanced medical facilities and services. The project seeks to support the population of the island nation of Sudan as they encounter barriers with regard to accessing health services especially due to distance and inadequate facilities. In terms of research methodology, close attention is paid to secondary data, including field studies: assessing health needs in the predefined target population and allocating space in the hospital for such purposes as emergency and outpatient departments and even outpatient diagnostic services. This paper also considers possible issues which can limit the effectiveness of the operations of the hospital in the rural area including staffing limitations, communication issues, and suggests approaches to tackle such issues through training and development of infrastructure. This paper highlights the significance of the project in narrowing the health discrepancies between residents of cities and those in rural regions, providing a sustainable model that can be replicated in more such regions so as to elevate the standards of healthcare and enhance fairness in the distribution of health services.

Autonomous Vehicle decisions rely on multimodal prediction models that account for multiple route options and the inherent uncertainty in human behavior. However, models can suffer from mode collapse, where only the most likely mode is predicted, posing significant safety risks. While existing methods employ various strategies to generate diverse predictions, they often overlook the diversity in interaction modes among agents. Additionally, traditional metrics for evaluating prediction models are dataset-dependent and do not evaluate inter-agent interactions quantitatively. To our knowledge, none of the existing metrics explicitly evaluates mode collapse. In this paper, we propose a novel evaluation framework that assesses mode collapse in joint trajectory predictions, focusing on safetycritical interactions. We introduce metrics for mode collapse, mode correctness, and coverage, emphasizing the sequential dimension of predictions. By testing four multi-agent trajectory prediction models, we demonstrate that mode collapse indeed happens. When looking at the sequential dimension, although prediction accuracy improves closer to interaction events, there are still cases where the models are unable to predict the correct interaction mode even just before the interaction mode becomes inevitable. We hope that our framework can help researchers gain new insights and advance the development of more consistent and accurate prediction models, thus enhancing the safety of autonomous driving systems.

Large language models (LLMs) have revolutionized natural language processing (NLP), demonstrating remarkable capabilities in tasks such as text generation, reasoning, and understanding. However, their substantial computational and memory requirements pose significant challenges for deployment in resourceconstrained environments, including edge devices and real-time applications. Knowledge distillation has emerged as a pivotal technique for addressing these challenges, offering an efficient means of compressing LLMs while preserving their performance. This paper presents a comprehensive examination of knowledge distillation methodologies and their role in enabling scalable and efficient LLMs. We explore foundational techniques, such as soft label transfer, feature alignment, and attention replication, which underpin the distillation process. Advanced methods, including task-specific optimizations, self-distillation, and multimodal distillation, are discussed in detail, showcasing their ability to capture and transfer high-level semantic knowledge. The integration of distillation with other model compression strategies, such as pruning, quantization, and tensor decomposition, is also analyzed, illustrating their synergistic potential to achieve state-of-the-art efficiency gains. While knowledge distillation has achieved remarkable success, several challenges remain unaddressed. These include replicating emergent behaviors, such as reasoning and zero-shot generalization, ensuring robustness across diverse domains, and mitigating the environmental and ethical implications of model training and compression. We highlight promising avenues for future research, including adaptive and dynamic distillation strategies, scalable frameworks for massive teacher models, and automated distillation pipelines leveraging neural architecture search. Additionally, we emphasize the need for standardized evaluation metrics and benchmarks to facilitate consistent and comprehensive assessments of distillation outcomes. Through a detailed synthesis of current advancements and open challenges, this paper underscores the transformative potential of knowledge distillation in democratizing access to cutting-edge LLMs. By advancing the state of the art in this domain, researchers and practitioners can develop efficient, high-performance models that bridge the gap between technical innovation and practical usability, fostering inclusivity and sustainability across a wide range of applications.

As a rigorous and comprehensive foundation for electromagnetic information theory (EIT), we present a theory that elucidates the stochastic structure of radiated electromagnetic (EM) fields and induced currents in general EM information transmission systems, encompassing arbitrary random scatterers and EM mutual coupling. The EM system is modeled as a multiply connected, fully general Riemannian manifold with an arbitrary metric. Our approach leverages the exact Green's function (GF) method to construct a novel electromagnetic random field theory (EM-RFT). The GF, interpreted as a response function on the manifold, enables us to treat the internal physical details of the EM system as a black box, shifting the analytical focus toward external input-output relations-a perspective aligned with signal processing and communication theory. This integration of random fields (RFs), electromagnetics, and GFs provides a robust framework for deriving and characterizing the universal stochastic structures of arbitrary EM information transmission systems. Specifically, we rigorously demonstrate the construction of EM random fields that fully comply with Maxwell's equations using the GFs of systems driven by external information fields. Our theory decouples stochastic input RFs from the random fluctuations inherent to the communication system, such as scatterers, and establishes general stochastic correlation propagators for generic EM communication links. Through the Karhunen-Loève expansion, we represent all EM random fields as sums of random variables, offering a foundation for simulating arbitrary EM RFs, evaluating mutual information between input and output spatial domains at any chosen locations within the system, and proving general theorems on the subject. Overall, this theory offers a rigorous and unified mathematical and conceptual basis for EIT, laying the groundwork for the development of new numerical methods to simulate and evaluate mutual information and capacity in EM communication systems.

Vehicular Ad-Hoc Networks (VANETs) are connected networks of vehicles and infrastructure that allow for communication between parties. VANETs are ever changing now that technology in and around vehicles is rapidly changing as well. Sensors, personal devices, and autonomously driving vehicles all pose new security threats to smart cities. Elliptic Curve Cryptography (ECC) can offer increased security and lowered computational effort in many of these scenarios. Our work will systematically review how ECC is being applied in real world scenarios and how it can be applied in the future to improve vehicle security in VANETs.

This paper presents DA-VINCI, a dynamically configurable and precision-scalable activation function core for versatile iterative neuron implementation to address the increasing demand for diverse activation functions in AI workloads. It leverages the CORDIC methodology to support runtime reconfigurability between Swish, SoftMax, SELU, GELU, Sigmoid, Tanh, and ReLU, achieving a quality of results (QoR) of 98.5%. FPGA evaluations demonstrate reductions of up to 4.5× in LUT usage, 3.2× in FF usage, 1.3× in critical delay, and 5.4× in power consumption compared to state-of-the-art designs. Empirical ASIC analysis further emphasizes the core's efficiency, achieving reductions of up to 16.2× area, 7.8× delay, and 14.3× power at CMOS 45nm, and 1.8× area and 10× delay at CMOS 28nm respectively. These improvements make DA-VINCI a fundamental block for resource-efficient AI accelerators focused on DNNs, RNNs/LSTMs, and Transformers.

This article introduces non-orthogonal frequency and time index modulation multiple access (NFT-IMMA) for sixth-generation (6G) communications. The proposed NFT-IMMA increases the number of communication resources by utilizing power levels, frequency, and time slots as transmission entities, which are selected through a predefined index activation pattern derived from partial user input data. NFT-IMMA increases spectral efficiency (SE) by employing index information and classical constellation symbols as methods of data transmission. Notably, it does not require prior scheduling at the base station, which simplifies the coordination process. Additionally, the use of three-dimensional indices for resource selection helps reduce user collisions. Performance evaluation at the physical layer, focusing on SE and bit error rate, indicates that NFT-IMMA outperforms conventional index modulation (IM) multiple access and IM-based orthogonal frequency division multiple access. At the MAC layer, NFT-IMMA demonstrates superior performance in terms of throughput and packet loss rate when compared to ALOHA and other competing techniques.

This study investigates the mechanical properties of three common infill patterns-grid, gyroid, and cubic-constructed using fused filament fabrication. Test specimens were printed with polylactic acid, and stiffness, strength, and specific energy absorption were evaluated under compression across varying relative densities. The results demonstrate the superior performance of the grid pattern in terms of stiffness and energy absorption, although this advantage is direction-dependent. The gyroid pattern exhibited competitive energy absorption with improved isotropy, while the cubic pattern showed the lowest mechanical properties across all metrics. These findings provide actionable insights for optimizing 3D-printed components for load-bearing applications by balancing material efficiency and mechanical performance.

Perspective-Aware AI (PAi) is a computational innovation in human-AI interaction that allows users to view and interact through each other's perspectives by creating personalized computational models called chronicles. Chronicles capture cognitive and behavioral tendencies from an individual's digital footprint, enabling enhanced decision-making by recognizing and auditing biases. Utilizing federated learning, PAi preserves privacy and ensures data ownership while providing scalability and precision. This approach enhances transparency and clarity in critical decision-making across various domains, including healthcare, education, and business, promoting inclusivity and diverse viewpoints.

Generic elbow prostheses often fail to meet the specific anatomical needs of patients, leading to high misalignment rates in Total Elbow Arthroplasty (TEA). This misalignment negatively impacts surgical efficiency, patient satisfaction, and critical functional outcomes, including range of motion and load-bearing capacity. Although personalized prostheses could address these challenges, current practices rely on ad-hoc intraoperative adjustments that frequently fall short of meeting individual anatomical requirements. A key obstacle to achieving truly personalized designs is the absence of a quantitative method for evaluating alignment. To tackle this issue, we developed a 3D computational model to simulate TEA and introduced the Alignment Improvement Metric (AIM), a novel quantitative tool for assessing prosthesis alignment. Using AIM, we designed three personalized prosthesis approaches optimized for individual, group, and cohort scenarios. Computational validation on 120 clinical participants demonstrated an average alignment improvement of up to 30% with just 3 prostheses, highlighting the potential of this AI-driven solution to enhance TEA outcomes and patient satisfaction.