Efficient Handling Research Articles

Integrating Expense Tracking with AI-Driven Investment Recommendations

The Expense Tracker is a comprehensive financial management application developed using the MERN stack—MongoDB, Express.js, React, and Node.js—to simplify expense tracking, budgeting, and investment decision-making. The application features a secure backend powered by Node.js and Express.js, utilizing RESTful APIs for efficient data handling, while MongoDB ensures flexible and scalable data storage. The React-based front end provides a responsive and intuitive interface, enabling users to seamlessly input, categorize, and visualize their expenses. Core functionalities include user authentication for secure and personalized access, secure expense logging to maintain accurate financial records, category management and budget tracking for effective financial organization, and interactive charts offering real-time insights into spending patterns. A key innovation of the application is the PennyDrop feature, which leverages user income and expense data to provide personalized investment suggestions. Using data-driven algorithms, PennyDrop empowers users to make informed financial decisions by aligning their income with tailored investment opportunities. This integration of expense tracking with actionable investment insights creates a holistic platform for financial management. The application is designed to be scalable, maintainable, and user-friendly, making it an ideal solution for individuals and businesses seeking an efficient tool for managing their finances and optimizing their investments. Index Terms— Budget Tracking, Category Management, Data-Driven Investment Insights, Expense Categorization, Expense Input, Expense Logging, Financial Management Application, Income-Based Investment Suggestions, Interactive Charts, MERN, PennyDrop Feature, Personalized Financial Solutions, Real-Time Financial Insights, RESTful APIs, Scalable Data Storage.

Programmable packet-optical network security and monitoring using DPUs with embedded GPUs [Invited

Data processing units (DPUs) with embedded graphics processing units (GPUs) have the potential to revolutionize optical network functionalities at the edge. These advanced units can significantly enhance the performance and capabilities of optical networks by integrating powerful processing capabilities directly at the network edge, where data is generated and consumed. We explore the use cases for DPUs in optical data monitoring with local artificial intelligence (AI) processing and embedded security. This paradigm shift aims to enable more efficient data handling, reduced latency, and improved overall network performance by leveraging local AI processing capabilities embedded within DPUs. In this paper, we show how DPUs can analyze vast amounts of optical data in real-time, implementing advanced data analysis algorithms and security protocols directly on the DPUs to provide robust monitoring and protection for the optical networks. Results indicate that DPUs with embedded GPUs can significantly improve the detection and response times to network anomalies, performance issues, and security threats.

Identification of waste lithium-ion battery cell chemistry for recycling.

Battery technology has attained a key position as an energy storage technology in decarbonization of energy systems. Lithium-ion batteries have become the dominant technology currently used in consumer appliances, electric vehicles (EVs), and industrial applications. However, lithium-ion batteries are not alike and can have different cathode chemistries which makes their recycling more complex. In addition, as larger quantities of batteries are starting to enter their end-of-life (EOL) stage, efficient handling and management of batteries with different cathode chemistry types are required. By identifying the cathode chemistry type prior to mechanical treatment, mixing of different cathode chemistries could be decreased, resulting in an increase in overall recycling efficiency. This study investigated the applicability of a non-destructive battery diagnostic methods, namely incremental capacity analysis (ICA), for identifying EOL lithium-ion battery chemistry. The study conducted ICA both on known reference batteries and EOL batteries from the recycling industry. Next, EOL batteries were crushed and the resulting fine active material was analysed to validate the ICA result. In addition, released gaseous and airborne particles were measured during crushing. The ICA results showed reliable identification of lithium iron phosphate (LFP) from other chemistries. In addition, lithium cobalt oxide (LCO), lithium nickel cobalt aluminum oxide (NCA) and lithium nickel manganese cobalt oxide (NMC) could be identified with various degrees. The identification may suffer if the battery is heavily used, and its state of health is low.

SecEdge: A novel deep learning framework for real-time cybersecurity in mobile IoT environments.

The rapid growth of Internet of Things (IoT) devices presents significant cybersecurity challenges due to their diverse and resource-constrained nature. Existing security solutions often fall short in addressing the dynamic and distributed environments of IoT systems. This study aims to propose a novel deep learning framework, SecEdge, designed to enhance real-time cybersecurity in mobile IoT environments. The SecEdge framework integrates transformer-based models for efficient handling of long-range dependencies and Graph Neural Networks (GNNs) for modeling relational data, coupled with federated learning to ensure data privacy and reduce latency. The adaptive learning mechanism continuously updates model parameters to counter evolving cyber threats. The framework's performance was evaluated in a simulation environment using the NSL-KDD, UNSW-NB15, and CICIDS2017 datasets. Results demonstrated that SecEdge outperformed state-of-the-art methods with a detection rate of 98.8% for DoS attacks on NSL-KDD, 98.5% for MitM attacks on UNSW-NB15, and 98.7% for data injection attacks on CICIDS2017.

A Study of Customer Satisfaction towards BFSI Sector

The BFSI (Banking, Financial Services, and Insurance) sector plays a pivotal role in the global economy by offering essential financial products and services. This study focuses on customer satisfaction within the BFSI sector, exploring key factors that influence customer perceptions and experiences. The research examines aspects such as service quality, accessibility, reliability, digitalization, and employee responsiveness, using both qualitative and quantitative methodologies. Primary data was collected through surveys and interviews with BFSI customers, while secondary data was gathered from industry reports and academic literature. Analytical tools were employed to identify trends and measure satisfaction levels across diverse customer segments. The study applies established frameworks, such as the SERVQUAL model, to evaluate service gaps and suggests actionable strategies for improvement. Findings indicate that digital innovations, personalized services, and efficient grievance redressal mechanisms significantly enhance customer satisfaction. Conversely, issues like delayed services and inadequate communication remain areas of concern. This report concludes with recommendations to strengthen customer relationships, enhance trust, and improve overall service delivery, ensuring long-term competitiveness in the BFSI sector. This study provides valuable insights for stakeholders aiming to foster customer-centric growth in a dynamic financial landscape. Customer satisfaction is a critical measure of success in the BFSI (Banking, Financial Services, and Insurance) sector, which serves as the backbone of the financial ecosystem. This study investigates the factors influencing customer satisfaction in the BFSI sector, focusing on key elements such as service quality, technological advancements, ease of access, and relationship management. The research employed a mixed-method approach, combining surveys, interviews, and secondary data analysis to evaluate customer experiences and expectations. Analytical tools such as the SERVQUAL model were utilized to assess gaps in service delivery. The study also explored the impact of digital transformation and personalized service offerings on customer loyalty and satisfaction. Findings reveal that while technological innovations and efficient grievance handling improve satisfaction, challenges such as service delays, lack of transparency, and limited financial literacy hinder the overall customer experience. Recommendations include enhancing digital platforms, offering tailored financial solutions, and fostering customer trust through effective communication and reliability.

Performance Evaluation of Various Deep Learning Models in Gait Recognition Using the CASIA-B Dataset

Human gait recognition (HGR) has been employed as a biometric technique for security purposes over the last decade. Various factors, including clothing, carrying items, and walking surfaces, can influence the performance of gait recognition. Additionally, identifying individuals from different viewpoints presents a significant challenge in HGR. Numerous conventional and deep learning techniques have been introduced in the literature for HGR, but traditional methods are not well suited to handling large datasets. This research explores the effectiveness of four deep learning models for gait identification in the CASIA B dataset: the convolutional neural network (CNN), multi-layer perceptron (MLP), self-organizing map (SOMs), and transfer learning with EfficientNet. The selected deep learning techniques offer robust feature extraction and the efficient handling of large datasets, making them ideal in enhancing the accuracy of gait recognition. The collection includes gait sequences from 10 individuals, with a total of 92,596 images that have been reduced to 64 × 64 pixels for uniformity. A modified model was developed by integrating sequential convolutional layers for detailed spatial feature extraction, followed by dense layers for classification, optimized through rigorous hyperparameter tuning and regularization techniques, resulting in an accuracy of 97.12% for the test set. This work enhances our understanding of deep learning methods in gait analysis, offering significant insights for choosing optimal models in security and surveillance applications.

CAFE: An Integrated Web App for High-Dimensional Analysis and Visualization in Spectral Flow Cytometry.

Spectral flow cytometry provides greater insights into cellular heterogeneity by simultaneous measurement of up to 50 markers. However, analyzing such high-dimensional (HD) data is complex through traditional manual gating strategy. To address this gap, we developed CAFE as an open-source Python-based web application with a graphical user interface. Built with Streamlit, CAFE incorporates libraries such as Scanpy for single-cell analysis, Pandas and PyArrow for efficient data handling, and Matplotlib, Seaborn, Plotly for creating customizable figures. Its robust toolset includes density-based down-sampling, dimensionality reduction, batch correction, Leiden-based clustering, cluster merging and annotation. Using CAFE, we demonstrated analysis of a human PBMC dataset of 350,000 cells identifying 16 distinct cell clusters. CAFE can generate publication-ready figures in real time via interactive slider controls and dropdown menus, eliminating the need for coding expertise and making HD data analysis accessible to all. CAFE is licensed under MIT and is freely available at https://github.com/mhbsiam/cafe .

Interval-value Pythagorean Fuzzy Prioritized Aggregation Operators for Selecting an Eco-Friendly Transportation Mode Selection

When selecting an environmentally friendly mode of transportation, many factors must be taken into account, such as capacity, delivery time, cost-effectiveness, and environmental impact. This is a difficult decision-making problem since it calls for the simultaneous evaluation and prioritization of several criteria. A thorough and methodical approach is required to make an informed decision that satisfies particular transportation needs while adhering to sustainability goals. The process entails balancing various trade-offs to determine the most environmentally friendly mode of transportation. When solving multi-attribute group decision-making (MAGDM) problems, prioritization is essential. Prioritization in fuzzy systems has been implemented through a variety of techniques and approaches. This paper tackles the MAGDM problem within the Pythagorean fuzzy (PyF) framework, taking into account the different requirements for experts and characteristics. We present new Aczel Alsina aggregation operators (AOs), whose efficient handling of uncertainties makes a major contribution to fuzzy mathematics. We suggest several PyF AOs, such as the PyF-prioritized Aczel Alsina geometric (PyFPAAG) and PyF-prioritized Aczel Alsina averaging (PyFPAAA), which are based on the Aczel Alsina t-norm and t-conorm. We show these AOs satisfy the aggregation criteria by examining their monotonicity, boundedness, and idempotency properties. The prioritization weights are derived from expert knowledge, enabling the suggested operators to capture the prioritization phenomenon among the aggregated arguments. Using a MAGDM technique, the proposed AOs are used to evaluate eco-friendly transportation modes. Their significance is confirmed by contrasting them with other well-known AOs.

An Evaluation of Image Slicing and YOLO Architectures for Object Detection in UAV Images

Object detection in aerial images poses significant challenges due to the high dimensions of the images, requiring efficient handling and resizing to fit object detection models. The image-slicing approach for object detection in aerial images can increase detection accuracy by eliminating pixel loss in high-resolution image data. However, determining the proper dimensions to slice is essential for the integrity of the objects and their learning by the model. This study presents an evaluation of the image-slicing approach for alternative sizes of images to optimize efficiency. For this purpose, a dataset of high-resolution images collected with Unmanned Aerial Vehicles (UAV) has been used. The experiments evaluated using alternative YOLO architectures like YOLOv7, YOLOv8, and YOLOv9 show that the image dimensions significantly change the performance results. According to the experiments, the best mAP@05 accuracy was obtained by slicing 1280×1280 for YOLOv7 producing 88.2. Results show that edge-related objects are better preserved as the overlap and slicing sizes increase, resulting in improved model performance.

Advanced control and optimization strategies for a 2-phase interleaved boost converter

Renowned for their adeptness in smoothing current flow and maintaining balanced operation, 2-phase interleaved boost converters (IBC) demonstrate remarkable efficiency, especially when confronted with demanding loads. This makes them a preferred choice for high-power applications such as renewable energy systems, high-power supplies, and electric vehicle power trains. In contrast, standard boost converters are typically favored in low-power, low-demand scenarios. The control of a 2-phase IBC involves running two boost converters in parallel but with a phase shift to reduce ripple currents, improve efficiency, and increase power handling capabilities. To ensure stability and optimal performance, the control strategies for these converters focus on achieving balanced operation between the phases. Hence, the control of 2-phase IBC presents a significant challenge due to their non-minimum phase behavior. The core focus of this article is the implementation of a composite model predictive control (MPC) technique to regulate a 2-phase interleaved boost converter. It introduces a novel approach, model predictive sliding mode control (MPSMC), which leverages the strengths of both MPC and sliding mode control (SMC). The benefits of this hybrid method, termed MPSMC, are thoroughly developed and simulated using MATLAB/Simulink. The results, as discussed in the respective section, provide an in-depth understanding of its effectiveness in practical applications.

Effective and Efficient Handling of Missing Data in Supervised Machine Learning

Effective and Efficient Handling of Missing Data in Supervised Machine Learning

Quality aspects of paddy grain and seed dried in HSTU mobile grain and seed dryer integrated with a dual heating system

Proper drying and excellence of dried grain and seed are the key concerns of millers, farmers and seed growers. The performance of a new Hajee Mohammad Danesh Science and Technology University (HSTU) mobile grain and seed dryer (HMD) integrated with dual heating system was assessed regarding the quality of dried seed and grain. The HMD was found to be capable of drying 720±10 kg paddy in a batch and drying period varied between 7 to 8 h at drying air temperatures of 65 oC. Paddy seeds of BRRI Dhan 49 and BRRI Dhan 89 were dried at 40 oC and the drying time required 9 to 10 h and 14 to 16 h in Aman and Boro season, respectively. Head rice yield and milling recovery varied from 59.50±0.16% to 61.61±0.54% and 74.95±0.09% to 75.70±0.59% for paddy dried in HMD. In addition, higher germination percentage (GP) of BRRI Dhan 49 varieties was found in Aman season that ranged from 86% to 88% which was dried in HMD in contrary to GP of paddy seeds dried under the sun. Moreover, higher GP was found from BRRI Dhan 89 varieties dried by HMD in Boro season. Also, the vigor index of HMD dried paddy seeds was more than the sun dried samples in both seasons. In conclusion, HMD integrated with a dual heating system is recommended for using at farm level as an alternative drying process for rapid and efficient handling of high-moisture grain and seed in Bangladesh and other subtropical nations in adverse climate.

Enhanced Accuracy in State-of-Charge Estimation for Lithium-Ion Batteries in Electric Vehicles Using Augmented Adaptive Extended Kalman Filter

Accurate assessment of state of charge (SoC) is crucial for the safety and efficiency of lithium-ion batteries used in electric vehicles (EVs). For precise SoC estimation in EV BMSs, an Augmented Adaptive Extended Kalman Filter (AAEKF) has been proposed in this research. To improve the precision of SoC estimation, AAEKF approach combines Kalman filtering methods with adaptive robust control concepts. The effectiveness of the AAEKF algorithm is shown by simulation results and confirmed by comparison with data from LiFePO4 battery research. The proposed AAEKF SoC estimation matches measurement results with errors of less than 3%. Adaptive Extended Kalman Filter (AEKF) is quite sensitive to the accuracy of the battery model, as errors in this region can lead to large estimate errors. Faster convergence to actual SoC, improved real-time responsiveness, accuracy, and resilience to disturbances and parameter fluctuations are all results of AAEKF's efficient handling of nonlinearities in battery systems, adaptation to system changes, errors (Root Mean Square Error (RMSE) and Mean Absolute Error (MAE), and overall performance. The maximum estimate error stays under 5% at various external temperatures.

Multimodal Human–Robot Interaction Using Gestures and Speech: A Case Study for Printed Circuit Board Manufacturing

In recent years, technologies for human–robot interaction (HRI) have undergone substantial advancements, facilitating more intuitive, secure, and efficient collaborations between humans and machines. This paper presents a decentralized HRI platform, specifically designed for printed circuit board manufacturing. The proposal incorporates many input devices, including gesture recognition via Leap Motion and Tap Strap, and speech recognition. The gesture recognition system achieved an average accuracy of 95.42% and 97.58% for each device, respectively. The speech control system, called Cellya, exhibited a markedly reduced Word Error Rate of 22.22% and a Character Error Rate of 11.90%. Furthermore, a scalable user management framework, the decentralized multimodal control server, employs biometric security to facilitate the efficient handling of multiple users, regulating permissions and control privileges. The platform’s flexibility and real-time responsiveness are achieved through advanced sensor integration and signal processing techniques, which facilitate intelligent decision-making and enable accurate manipulation of manufacturing cells. The results demonstrate the system’s potential to improve operational efficiency and adaptability in smart manufacturing environments.

Efficient handling of radiality constraints for large-scaled power distribution networks

Power distribution networks are usually characterized by a radial topology and therefore the related optimization problems require radiality constraints in their formulations. However, practical instances may have very large size, and the number of radiality constraints may grow faster than the size of the instance. Hence, the arising optimization models may become computationally intractable due to their huge dimension. This work proposes a combinatorial optimization approach to overcome this issue. Our approach is based on the combinatorial formulation of the radiality constraints and their delayed and efficient generation in the model, following a separation-optimization scheme. We find an optimal acyclic spanning subgraph subject to both technical side constraints and topological radiality requirements on a graph representing the distribution network. This can be done without considering all the exponentially many radiality constrains from the beginning of the model solution, but introducing only the ones that are needed to make the solution feasible. It turns out that the number of required constraints is much smaller than the number of all possible radiality constraints, also because some of the electric constraints already favor the elimination of infeasible configurations. Computational experiments on reconfiguration benchmarks from the literature show the effectiveness of the proposed approach.

Scalable Transformer Accelerator with Variable Systolic Array for Multiple Models in Voice Assistant Applications

Transformer model is a type of deep learning model that has quickly become fundamental in natural language processing (NLP) and other machine learning tasks. Transformer hardware accelerators are usually designed for specific models, such as Bidirectional Encoder Representations from Transformers (BERT), and vision Transformer models, like the ViT. In this study, we propose a Scalable Transformer Accelerator Unit (STAU) for multiple models, enabling efficient handling of various Transformer models used in voice assistant applications. Variable Systolic Array (VSA) centralized design, along with control and data preprocessing in embedded processors, enables matrix operations of varying sizes. In addition, we propose an efficient variable structure and a row-wise data input method for natural language processing where the word count changes. The proposed scalable Transformer accelerator accelerates text summarization, audio processing, image search, and generative AI used in voice assistance.

Biomimetic Microstructure with Anti-Slip and Anti-Adhesion for Efficient Handling of Brittle Material Surfaces in High-Temperature Environments.

Non-destructive handling such as wafer handling usually requires a high-temperature environment, however, most bionic materials fail in high temperatures due to material decomposition. In this study, inspired by the unique microstructure of locust toe pads with low adhesion and high friction, bionic high-temperature friction pads are designed and fabricated, selecting high-temperature-resistant silicone rubber as the material. The interfacial mechanical properties at high temperatures are analyzed. The samples with terminal bulges possess preferable roughness adaptability, enabling the advantages of low adhesion and high friction in high temperatures. The high-temperature non-destructive handling experiment using a robotic arm verifies the feasibility of bionic high-temperature friction pads in industrial applications and provides a valuable solution for non-destructive handling in high-temperature environments.

PIMSAB: A Processing-In-Memory System with Spatially-Aware Communication and Bit-Serial-Aware Computation

Bit-serial Processing-In-Memory (PIM) is an attractive paradigm for accelerator architectures, for parallel workloads such as Deep Learning (DL), because of its capability to achieve massive data parallelism at a low area overhead and provide orders-of-magnitude data movement savings by moving computational resources closer to the data. While many PIM architectures have been proposed, improvements are needed in communicating intermediate results to consumer kernels, for communication between tiles at scale, for reduction operations, and for efficiently performing bit-serial operations with constants. We present PIMSAB, a scalable architecture that provides a spatially aware communication network for efficient intra-tile and inter-tile data movement and provides efficient computation support for generally inefficient bit-serial compute patterns. Our architecture consists of a massive hierarchical array of compute-enabled SRAMs (CRAMs), which is codesigned with a compiler to achieve high utilization. The key novelties of our architecture are (1) in providing efficient support for spatially aware communication by providing local H-tree network for reductions, by adding explicit hardware for shuffling operands, and by deploying systolic broadcasting, as well as (2) by taking advantage of the divisible nature of bit-serial computations through adaptive precision and efficient handling of constant operations. These innovations are integrated into a tensor expressions-based programming framework (including a compiler for easy programmability) that enables simple programmer control of optimizations for mapping programs into massively parallel binaries for millions of PIM processing elements. When compared against a similarly provisioned modern Tensor Core GPU (NVIDIA A100), across common DL kernels and end-to-end DL networks (Resnet18 and BERT), PIMSAB outperforms the GPU by 4.80×, and reduces energy by 3.76×. We compare PIMSAB with similarly provisioned state-of-the-art SRAM PIM (Duality Cache) and DRAM PIM (SIMDRAM), and observe a speedup of 3.7× and 3.88×, respectively.

A Strategy for Reliable Cargo Loading of Low-Pressure Liquid Carbon Dioxide Carriers

This study addresses the control challenges associated with loading low-pressure liquid carbon dioxide carriers (LCO2Cs), which are crucial components of the carbon capture, utilization, and storage (CCUS) chain. It explores the need for stable pressure and temperature control to prevent dry ice formation and ensure efficient cargo handling. The research employed HYSYS dynamic simulations to assess three different control strategies. The simulations assessed each strategy’s effectiveness in maintaining stable operating conditions and preventing risks, such as dry ice formation and valve blockages. The study concluded by examining the necessity of pressurization for safe and efficient LCO2 loading and by determining which control strategy is most effective and reliable based on the simulation outcomes. Among the three scenarios examined, Case A, which utilized two control valves, exhibited initial instability due to significant flow coefficient differences, resulting in temperature drops below the CO2 triple point and increasing the risk of dry ice formation. Case C, operating without pressurization, experienced severe pressure fluctuations and prolonged exposure to temperatures below the triple point, posing risks of valve blockages. In contrast, Case B, which uses a remote pressure-reducing valve and a control valve, demonstrated the most stable performance, effectively avoiding dry ice formation and pressure fluctuations, making it the most reliable method for safe LCO2 cargo loading.

A Portable Photocollector for the Field Collection of Insects in Biodiversity Assessment.

Arthropod biodiversity research usually requires large sample collections. The efficient handling of these samples has always been a critical bottleneck. Sweep netting along transects is an effective and commonly used approach to sample diverse insects. However, sweep netting requires the time-consuming task of sorting insects from the large amounts of debris and foliage that end up in the sweep net along with the insects. To address this, we introduce a robust, portable, and inexpensive photocollector device with an LED light source to extract insects from sweep net samples in a standardized way. Timed field trials tested the photocollector's efficiency in extracting live insect samples from debris, focusing on Hymenoptera and Diptera. We found that 73% (±13%) of undamaged specimens moved toward the collection bottle within the first hour and 79% (±13%) after four hours. Of the insects failing to move after four hours, most (81%) were damaged and likely unable to move. Accounting only for undamaged specimens, 83% (±11%) moved after 1 h and 90% (±11%) moved after 4 h. We found significant differences in when families of Hymenoptera and Diptera moved. We suggest that the photocollector can be a useful tool in standardized biodiversity assessments.