This article provides a fuzzy control system for a smart home automation system using multiple techniques from different approaches to fulfill the needs of the user in an automation system with ease and customizable way. The fuzzy control system is applied to the automation system to increase its efficiency and the controlling system are characterized in multiple format like use of EMG sensors for controlling the system with the movement of nerves and muscles, wave or gesture control using the infrared or ultrasonic sensors, chatbot system for communicating with the system from anywhere through android, renewable sources and methods are applied for power saving and increasing efficiency. The control system provides freedom for the system to work efficiently according to the environment and data provided to the system. For example, the automatic close of electronic appliances in a room if no one is using it or the automatic start of light and basic environment after entering the room by the user. The chatbot system also provides a way or suggestion for the system to the user in case of emergency and it also keeps track of power usage and monitors the working of the system. The timing can also be set to the system with the help of a chatbot just like an alarm automation.

Technology, including the internet, phones, computers, tablets, platforms, networks, software applications, and databases, has significantly impacted our daily lives. The integration of e-commerce platforms in rural agricultural sectors has opened new opportunities for farmers and producers to reach wider markets. These platforms can facilitate growth, efficiency, and market access for rural farmers. As rural revitalization progresses, e-commerce can help agriculture development through diversification, intelligent agriculture, and e-commerce trading platforms. These platforms have the potential to revolutionize traditional farming by creating direct links between farmers and buyers. This paper aims to study the development strategy of ecommerce in rural areas, focusing on marketing status, problems, and specific development strategies, to promote the development of e-commerce in rural areas under the background of rural revitalization.

Radio Frequency Interference (RFI) in Synthetic Aperture Radar (SAR) is a daunting challenge that affects the SAR sensing reliability and its image quality. To ensure that SAR remains a powerful tool for Earth observation, this paper presents an effective two-dimensional tuneable attenuation space-frequency (azimuth-range) filtration. This framework is based on key components including time-frequency features of level-0 SAR data, radar antenna pattern, and the estimated parameters of the interference signal. Additionally, a signal power localization method is applied to estimate the relative position of the interference source with 95% accuracy to facilitate applying the tuneable filter. Simulated results are obtained using a public open-source spaceborne SAR emulator, SEMUS, for generating emulated clean and contaminated SAR raw data. Furthermore, the filtration framework is successfully tested on real-life interference events on the TerraSAR-X satellite raw data.

In the last few years, our cities become more and more crowded due to the increasing number of cars and old city planes. Small cities as Messina (i.e. an Italian city with 200k inhabitants) have travel time comparable with bigger ones as Adelaide or Cologne (i.e. cities with more than 800k inhabitants situated in Australia and Italy, respectively). A possible solution to mitigate such an issue is to use simulation software and design different city plans to find the most suitable one. However, these tools are often difficult to be used by city traffic managers without good computer skills. In this paper, we used a Multi-Agent Transport Simulation (MATSim) to provide a simple tool that can be used to better organize the direction of roads and public transportation. In particular, starting from the open data provided by the city of Messina, we implemented a software code able to process a MATSim event file through matsim-tools. Moreover, we proposed an index to estimate how much roads are safe for cyclists. From the simulation results, we were able to discover the most saturated links and the travel time distribution by hour of departure time.

Neural Network (NN) technology is revolutionizing the modeling paradigm in the engineering arena, enhancing simultaneously model precision and process speed. Nevertheless, the deficiency of physical interpretability in current NN models causes an unattainable demand for the quantity of training data, especially when dealing with high-dimension physical behaviours. In this work, a Hierarchical Physics-Informed Neural Network (HPINN) framework is proposed to address the 3D magnetic modeling issue of Medium Frequency Transformers (MFTs). By establishing the knowledge transfer channel between the cross section and the equivalent space, the inferior 2D model parameter search space, which is restricted by Dowell’s equation and accessible 2D magnetic data information, can be mapped into the superior 3D search space. Accordingly, with limited 3D training data, the performance of the HPINN framework is still guaranteed around the optimal state. The structural embedding of physics knowledge reduces the model's reliance on 3D training data, thereby relieving the strict computing requirement at the data generation stage. Finally, according to the training results and experiment verifications, with only one-third computational burden used in the classic NN model, the average error of this proposed HPINN framework is reduced to 1%, and the maximum error does not exceed 10%.

User-centric (UC) distributed massive multiple-input multiple-output (D-mMIMO), commonly called cell-free mMIMO, is an important technology to ensure a more uniform coverage as well as higher spectral and energy efficiencies in next generation communication systems. This paper investigates the performance of UC D-mMIMO systems enabled by a swarm of unmanned aerial vehicles (UAVs). Specifically, it presents a comprehensive study on UAVs’ deployment and trajectory optimization as aerial transmission and reception points (TRPs) of D-mMIMO systems, considering systems composed solely of aerial TRPs and those formed combining aerial and terrestrial TRPs. Moreover, user equipment (UE) mobility is modeled using a discrete-time Markov chain, and a novel approach to heuristically optimize the positions of aerial TRPs is proposed, one that considers the continuous movement of UEs in the coverage area. The proposed approach optimizes each UAV’s three-dimensional location under a time discretization framework, with the positioning of the UAVs being adjusted periodically, allowing for iterative trajectory optimization to improve the UEs’ spectral efficiency (SE) performance. Simulation results reveal that the proposed UAV trajectory optimization allows for significant SE improvement, especially for a low UE density scenario. Specifically, comparing the proposed method with a fixed position setup, up to 47.84% increase on average SE is achieved.

The reactive Power Take Off (PTO) force is the key to maximizing mechanical power absorption and electric power generation of Wave Energy Converters (WECs) from ocean waves of variable frequencies and amplitudes, but its study is limited due to its difficulty in physical realization. This paper presents a simple yet effective RLC-tuned WEC that generates a tunable reactive PTO force from tunable resistor R, inductor L and capacitor C in the WEC. A complete closed loop system model of the WEC is derived first, then five quantitative rules are obtained from analyzing the model. These rules are used to tune the RLC network and hence the reactive PTO force for the WEC to resonate with the regular wave input force and generate maximal electric power over a range of wave frequencies, or to significantly enhance electric power generation from irregular wave input force. Mathematical analysis of the WEC and tuning rules reveals the analytical and quantitative descriptions of mechanical power absorption, active and reactive electric power generation and power factor, optimal electric load R, and generator and LC capacity requirements of the WEC. Simulation studies verify the analysis results and show the effectiveness and advantages of the proposed WEC and its robustness to imprecise estimate of wave frequency.

The authenticity of information has been a longstanding issue with it's potential to impact millions of users in the blink of an eye. Recent years has seen a growth in development of fake news detection models. Model Interpretability especially in NLP domain is still challenging, yet it helps in adopting models to various domains. In this study we tried post hoc interpretation techniques like local and global interpretations using LIME and SHAP, Topic modeling and Keyword extraction techniques. We identified these are simple yet powerful techniques to better understand the tagging behavior of fake news detection models used in this paper. With the help of these interpretation and data augmentation techniques we measured model robustness and identified that models built on ML algorithms are not robust to covariance shift in input data. Also, we tried to derive some of the characteristics that better represents style of fake tweets.

Accurately measuring propagation loss is crucial for Citizens Broadband Radio Service (CBRS) network planning, deployment, and successful spectrum sharing among CBRS users. The currently used propagation models, such as the Irregular Terrain Model (ITM) and Extended Hata (eHata) in the CBRS network, disregard significant environmental factors like foliage or clutter data, making the loss prediction unreliable. Additionally, current experimental studies on CBRS propagation are limited, with methodologies often relying on expensive measurement tools. To address these limitations, this paper proposes a CBRS network propagation framework, which includes a signal measurement system made of affordable, commercial off-the shelf electronic devices and advanced empirical data analysis. To validate the framework, an extensive measurement campaign is conducted in a live CBRS network in Buffalo, NY. The empirical path loss results have been compared with existing analytical models, with the alpha-beta (𝛼 − 𝛽) model giving the best path loss prediction. The methodology and framework presented in this paper can be applied to other network environments, helping researchers and engineers estimate network performance during the design phase, thus saving resources.

Intelligent transportation systems (ITS) have been attracting the attention of industry and academia alike for addressing issues raised by the 2030 agenda for sustainable development goals (SDG) approved by the United Nations. However, the diversity and dynamics of present-day transportation scenarios are already very complex, turning the management of ITS into a virtually impossible task for conventional traffic control centers. Recently, the digital twin (DT) paradigm has been presented as a modern architectural concept to tackle complex problems, such as the ones faced by ITS. This survey aims to provide a piece-wise approach to introducing DTs into sustainable ITS by addressing the following cornerstone aspects: i) Why should one consider DTs in ITS applications? ii) What can DTs represent from ITS' new physical environments? And iii) How can one use DTs to address SDG related to efficiency, safety, and ecology in ITS? Our methodological approach for surveying the literature addresses these questions by categorizing contributions and discriminating their ITS elements and agents against the SDG they addressed. Thus, in this survey, we provide an in-depth and contextualized overview of the challenges when approaching ITS through DTs, including scenarios involving autonomous and connected vehicles, ITS' infrastructure, and traffic agents' behavior. Moreover, we propose a functional reference framework for developing DTs of ITS. Finally, we also offer research challenges regarding standardization, connectivity infrastructure, security and privacy aspects, and business management for properly developing DTs for sustainable ITS.

In advanced high-speed integrated systems, the widely distributed and proliferation of temperature sensors to detect hotspots improve the robustness and reliability of the system by preventing overheating. Low area and low energy consumption are essential for integrated temperature sensors in such applications. The band-to-band tunneling (BTBT) current in PDSOI technology increases linearly with temperature. We demonstrate a BTBT-based quasi-linear temperature-dependent oscillator, thermillator. The fabricated thermillator has a low footprint (42.3 μm², 10× less than state-of-the-art temperature sensing cores) and enables low energy consuming temperature to digital converter (0.33 nJ energy/conversion) in GF45RFSOI technology. The proposed thermillator facilitates a low power and low footprint temperature sensor (0 oC to 85 oC) with a simple counter-based digital readout with a best-in-class resolution FoM of 0.16 pJK².

Transformers have achieved great success in natural language processing and computer vision. The core and basic technique of transformers is the self-attention mechanism. The vanilla self-attention mechanism has quadratic complexity, which limits its applications to vision tasks. Most of the existing linear self-attention mechanisms will sacrifice performance to some extent to reduce complexity. In this paper, we propose a novel linear approximation of the vanilla self-attention mechanism named CURSA to achieve both high performance and low complexity at the same time. CURSA is based on the CUR decomposition to decompose the multiplication of large matrices into the multiplication of several small matrices to achieve linear complexity. Experiment results of CURSA in image classification tasks, semantic segmentation tasks, and object detection tasks show that it outperforms state-of-the-art self-attention mechanisms with better data efficiency, faster speed, and higher accuracy.

During the COVID-19 pandemic, there was a notable decrease in student learning rates across public school systems in the United States, undoing years of progress. This research examines nine publicly accessible data outlets within a data science framework to explore their potential to understand and mitigate factors contributing to learning loss. The data was sourced from various entities, including the Census Bureau 2010, USAFACTS, Texas Department of State Health Services (DSHS), the National Center for Education Statistics (CCD), U.S. Bureau of Labor Statistics (LAUS), and three sources from the Texas Education Agency (STAAR, TEA, ADA, ESSER). We present an end-to-end large-scale educational data modeling pipeline that integrates, cleans, and implements automated attribute importance analysis for deriving meaningful insights from educational data. We aim to address several key research inquiries: i) Do students from low-income backgrounds and minority groups exhibit heightened learning loss? ii) How does learning loss vary across different grade levels? iii) What impact does the decision to reopen schools or school districts have on student learning loss? iv) Is there a relationship between the mode of instruction (hybrid, remote, in-person) and learning loss? v) Is the school or district attendance inversely associated with learning loss? vi) Does the local or regional infection rate correlate with increased learning loss? vii) How does the local unemployment rate influence learning losses? Our investigation demonstrated the superior performance of gradient-boosting algorithms, particularly XGBoost and CatBoost, in handling missing values. During this period, the mode of instruction and prior score emerged as the primary resilience factors, alongside low income and grade level, which were found to be the most influential factors in predicting learning loss for both math and reading. We demonstrate a novel data-driven approach to discover insights from an extensive collection of heterogeneous public data sources and offer an actionable understanding to policymakers to identify learning-loss tendencies and prevent them in public schools.

This paper presents a novel Continuously Variable Transmission (CVT) based on a Twisted String Actuator (TSA), capable of automatically adjusting its Transmission Ratio (TR) in response to external loads. Utilizing lightweight hyperelastic slender rods, the CVT achieves a simple, compact, and cost-effective design. By manipulating the distance between two twisted strings through rod deformation, the TR continuously adapts to varying load conditions. Mathematical models of the TSA-based CVT system are derived and experimentally validated, demonstrating a 2.1-fold TR variation from 0.1 kg to 1.5 kg loads. Application in an anthropomorphic robot finger showcases a 6.2-fold TR change between unloaded and loaded states. Our CVT offers unparalleled advantages in weight, cost, simplicity, compliance, and continuous TR adjustability, making it highly suitable for robotic systems.

Students find it difficult to learn the syntax of programming languages when they are new to coding. Some students get discouraged and even quit their coding journey because of having trouble with memorizing all the syntactic rules of a programming language. Flowcharts are an effective way to help students visualize the logic of an algorithm developed to solve a problem. But some students find it challenging to convert their flowcharts into executable code and even when they do, some get frustrated because of the errors they encounter while running the code they have written. So students who are not well versed with the syntax of a programming language can face difficulty debugging their programs even when they understand the logic behind the algorithms. This research aimed at developing a tool that can assist young students with less coding experience to learn programming. An intuitive user interface for the application was developed using a python library called gradio. The functionality of taking flowchart images as user inputs and providing equivalent python code with a brief explanation was incorporated on the UI. Text from the uploaded flowchart image is extracted in 3 steps which include feature extraction using deep learning models like ResNet and VGG, sequence labelling using Long Short-Term Memory networks and decoding the sequence labels using Connectionist Temporal Classification algorithms. The extracted text is then provided to an open source large language model with 70 billion parameters called Llama-2-Chat for the generation of the python code. The generated python code is displayed on the gradio UI along with a brief explanation of the code. The user can copy the code and run it on the IDE of their choice. The application was evaluated with a test dataset of several flowchart images. The generated python codes for the test flowchart images were executed in a code editor and the produced outputs were found to be correct. The generated code got executed without any errors for 75 % of the flowchart images from the test dataset.

In various real-world applications such as machine translation, sentiment analysis, and question answering, a pivotal role is played by NLP models, facilitating efficient communication and decision-making processes in domains ranging from healthcare to finance. However, a significant challenge is posed to the robustness of these natural language processing models by text adversarial attacks. These attacks involve the deliberate manipulation of input text to mislead the predictions of the model while maintaining human interpretability. Despite the remarkable performance achieved by state-of-the-art models like BERT in various natural language processing tasks, they are found to remain vulnerable to adversarial perturbations in the input text. In addressing the vulnerability of text classifiers to adversarial attacks, three distinct attack mechanisms are explored in this paper using the victim model BERT: BERT-on-BERT attack, PWWS attack, and Fraud Bargain’s Attack (FBA). Leveraging the IMDB, AG News, and SST2 datasets, a thorough comparative analysis is conducted to assess the effectiveness of these attacks on the BERT classifier model. It is revealed by the analysis that PWWS emerges as the most potent adversary, consistently outperforming other methods across multiple evaluation scenarios, thereby emphasizing its efficacy in generating adversarial examples for text classification. Through comprehensive experimentation, the performance of these attacks is assessed, and the findings indicate that the PWWS attack outperforms others, demonstrating lower runtime, higher accuracy, and favorable semantic similarity scores. The key insight of this paper lies in the assessment of the relative performances of three prevalent state-of-the-art attack mechanisms.

Traditional assessments of children’s health and behavioral issues primarily rely on subjective evaluation by adult raters, which imposes major costs in time and human resource to the school system. This pilot study investigates the utilization of millimeter-wave radar coupled with machine learning for the objective and semi-automatic detection and classification of children’s activity levels, defined as restlessness, within a real classroom environment. Two objectives are pursued: confirming the feasibility of restlessness detection using millimeter-wave radar and proposing an algorithm for restlessness classification through machine learning. The experiment involves a nine-day observational study, using two radar systems to monitor the activities of 14 children in a primary school. Radar data analysis involves the extraction of distinctive features for restlessness detection and classification. Results indicate the successful detection of restlessness using millimeter-wave radar, demonstrating its potential to capture nuanced body movements in a privacy-protected manner. Machine learning models trained on radar data achieve a classification accuracy of 100%, outperforming other methods in terms of non-invasiveness, lack of body restraint, multi-target applications, and privacy protection. The study’s contributions extend to children, parents, and educational practitioners, emphasizing non-invasiveness, privacy protection, and evidence-based support. Despite limitations such as a short monitoring duration and a small sample size, this pilot study lays the foundation for future research in non-invasive restlessness detection using non-contact monitoring technologies. The integration of millimeter-wave radar and machine learning offers a promising avenue for efficient and ethical trait assessments in real-world educational environments, contributing to the advancement of child psychology and education. This work supports efforts for non-contact monitoring of children’s activity holding promise such as non-invasive, privacy protection, multi-targets, objective evaluation, and computer-aided screening.

A document by Jonas Noeland . Click on the document to view its contents.