Deployment Challenges Research Articles

Challenges for Large-Scale Deployment of WBG in Power Electronics

Challenges for Large-Scale Deployment of WBG in Power Electronics

Edge Computing and Microservices: Extending Scalability Beyond the Cloud

The convergence of edge computing and microservices architecture represents a transformative shift in how organizations process and manage distributed data systems. This article explores the limitations of traditional cloud-centric approaches and presents edge computing as a solution to address these challenges. Through a detailed analysis of real-world implementations across various sectors, including autonomous vehicles, smart cities, and industrial automation, the article examines how edge computing integrated with microservices architecture improves system performance, reduces latency, and optimizes resource utilization. The article investigates deployment challenges and security considerations and provides insights into effective monitoring strategies. Furthermore, the article explores emerging technologies and future research directions that will shape the evolution of edge computing and microservices integration.

Innovative Ghost Channel Spatial Attention Network with Adaptive Activation for Efficient Rice Disease Identification

To address the computational complexity and deployment challenges of traditional convolutional neural networks in rice disease identification, this paper proposes an efficient and lightweight model: Ghost Channel Spatial Attention ShuffleNet with Mish-ReLU Adaptive Activation Function (GCA-MiRaNet). Based on ShuffleNet V2, we effectively reduced the model’s parameter count by streamlining convolutional layers, decreasing stacking depth, and optimizing output channels. Additionally, the model incorporates the Ghost Module as a replacement for traditional 1 × 1 convolutions, further reducing computational overhead. Innovatively, we introduce a Channel Spatial Attention Mechanism (CSAM) that significantly enhances feature extraction and generalization aimed at rice disease detection. Through combining the advantages of Mish and ReLU, we designed the Mish-ReLU Adaptive Activation Function (MAAF), enhancing the model’s generalization capacity and convergence speed. Through transfer learning and ElasticNet regularization, the model’s accuracy has notably improved while effectively avoiding overfitting. Sufficient experimental results indicate that GCA-MiRaNet attains a precision of 94.76% on the rice disease dataset, with a 95.38% reduction in model parameters and a compact size of only 0.4 MB. Compared to traditional models such as ResNet50 and EfficientNet V2, GCA-MiRaNet demonstrates significant advantages in overall performance, especially on embedded devices. This model not only enables efficient and accurate real-time disease monitoring but also provides a viable solution for rice field protection drones and Internet of Things management systems, advancing the process of contemporary agricultural smart management.

A Review of Deep Learning Applications for Sustainable Water Resource Management

Deep learning (DL) techniques and algorithms have the capacity to significantly impact world economies, ecosystems, and communities. DL technologies have been utilized in the development and administration of urban structures. However, there exists a dearth of literature reviewing the present level of these applications and exploring potential directions in which DL can address water challenges. This study aims to review demand projections, leakage detection and localization, drainage defect and blockage, cyber security and wealth surveillance, wastewater recycling and management, water safety prediction, rainfall conversation, and irrigation regulation. The application of DL techniques is currently in its early stages. Most studies have adopted standard networks, simulated information, and experimental or prototype settings to evaluate the efficacy of DL approaches. However, there have been no reported instances of practical adoption. Compared to other reviewed problems, leakage detection is currently being implemented practically in daily operations and handling of water facilities. The major challenges for the practical deployment of DL in water management include algorithmic development, multi-agent platforms, virtual clones, data quality and availability, security, context-aware data analysis, and training efficiency. We validate our study by using several case studies that employ DL for water treatment. Prospective exploration and deployment of DL systems are anticipated to advance water systems toward increased cognition and flexibility. This research aims to encourage further research and development in utilizing DL for feasible water usage and digitalization of the global water sector.

FPGA-Based Sparse Matrix Multiplication Accelerators: From State-of-the-Art to Future Opportunities

Sparse matrix multiplication (SpMM) plays a critical role in high-performance computing applications, such as deep learning, image processing, and physical simulation. Field-Programmable Gate Arrays (FPGAs), with their configurable hardware resources, can be tailored to accelerate SpMMs. There has been considerable research on deploying sparse matrix multipliers across various FPGA platforms. However, the FPGA-based design of sparse matrix multipliers still presents numerous challenges. Therefore, it is necessary to summarize and organize the current work to provide a reference for further research. This article first introduces the computational method of SpMM and categorizes the different challenges of FPGA deployment. Following this, we introduce and analyze a variety of state-of-the-art FPGA-based accelerators tailored for SpMMs. In addition, a comparative analysis of these accelerators is performed, examining metrics including compression rate, throughput, and resource utilization. Finally, we propose potential research directions and challenges for further study of FPGA-based SpMM accelerators.

Shuffle-PG: Lightweight feature extraction model for retrieving images of plant diseases and pests with deep metric learning

Disease and pest diagnosis plays a critical role in managing and controlling the damage caused by plant diseases and pests. This study employs a content-based image retrieval approach to diagnose diseases and pests, suggesting similar candidate images to assist in decision-making. Previous research in disease and pest diagnosis has relied on large models for feature extraction, posing challenges for deployment in resource-constrained environments like mobile devices. To address these challenges, this study proposes a lightweight feature extraction model, Shuffle-PG, which integrates the computationally efficient ShuffleNet v2 model with pointwise group convolution. Additionally, a method for fine-tuning the feature extraction model using deep metric learning based on contrastive loss was developed to enhance discriminative feature extraction. To validate the effectiveness of the proposed method, experiments were conducted using plant disease and pest datasets specifically collected for this study. The results show that the proposed Shuffle-PG model uses approximately 20 times fewer parameters and reduces computational costs by an order of magnitude compared to existing benchmark models, while achieving higher mean average precision scores of 97.7 % and 98.8 % for the disease and pest datasets, respectively.

Harnessing bacterial antimicrobial peptides: a comprehensive review on properties, mechanisms, applications, and challenges in combating antimicrobial resistance.

Antimicrobial resistance (AMR) is a significant global health concern, due to the persistence of pathogens and the emergence of resistance in bacterial infections. Bacterial-derived antimicrobial peptides (BAMPs) have emerged as a promising strategy to combat these challenges. Known for their diversity and multifaceted nature, BAMPs are notable bioactive agents which exhibit potent antimicrobial activities against various pathogens. This review explores the intricate properties and underlying mechanisms of BAMPs, emphasizing their diverse applications in addressing AMR. Additionally, the review investigates the mechanisms, analyses the challenges in utilizing BAMPs effectively, and examines their potential applications and associated deployment challenges providing comprehensive insights into how BAMPs can be harnessed to combat AMR across different domains.

AI-powered Predictive Maintenance for Solar Energy Systems: A Case Study

This article presents the implementation of an AI-powered predictive maintenance system at SolarTech Solutions' 75MW solar installation in Arizona, examining the transformation from traditional reactive maintenance to an advanced predictive maintenance framework utilizing IoT sensors and artificial intelligence. The implementation, spanning 18 months, integrated 12,000 distributed sensors with a sophisticated AI/ML infrastructure, achieving remarkable results across multiple dimensions: 94.3% accuracy in anomaly detection, 98.2% precision in fault localization, and a 47% reduction in unplanned downtime, generating annual savings of $425,000. Key operational improvements include a 64% increase in Mean Time Between Failures (MTBF), reduction in maintenance response time from 72 to 4 hours, and a 3.2% improvement in panel efficiency. The system's environmental impact is equally significant, contributing to a reduction of 1,960 metric tons of CO2 emissions annually and conserving 1.2 million gallons of water per year. The article demonstrates the viability and effectiveness of AI-driven maintenance solutions in utility-scale solar operations, providing a comprehensive framework for similar implementations across the renewable energy sector, while addressing critical challenges in deployment, integration, and optimization of these advanced systems.

AI-Powered Geotechnics: Enhancing Rock Mass Classification for Safer Engineering Practices

AbstractRock mass classification is fundamental for evaluating rock mass quality, essential for stability analysis and geotechnical design. Traditional classification methods are limited by joint observation technology, which typically gathers joint information from one-dimensional or two-dimensional perspectives, failing to comprehensively capture three-dimensional joint occurrences. This often necessitates empirical formulas for joint distribution, resulting in less precise joint parameter calculations. This paper reviews 44 seminal articles on rock engineering classification in construction and subterranean projects, tracing the evolution from foundational methods like Rock Quality Designation, Rock Mass Rating, Q-system, Basic Quality, and Hydropower Classification to contemporary techniques. It highlights the transformative impact of data science, particularly artificial intelligence, on rock engineering. The analysis reveals 73 distinct algorithms used 162 times in literature, with Support Vector Machines Support, Vector Regression, K-means clustering, K-Nearest Neighbors, Artificial Neural Networks and Random Forest being the most successful. This paper examines each method's advantage and limitations, discussing the challenges of algorithm deployment in the scientific community. The findings underscore the integration of machine learning and meta-heuristic optimization methods in rock engineering classification, offering valuable insights for future research and applications.

Continual Learning and Adaptation In Resource-Constrained Environments (CLAIRE)

As climate-related extreme events continue to increase and impact the world, of particular importance are the threats posed to food, agriculture, and water (FAW) systems. Deep learning could benefit FAW systems for classification of threats and for forecasting potential future events given historical patterns. However, many FAW systems are faced with operational environments that are resource-constrained, which could present challenges in deploying deep learning models. Continual learning offers a way to overcome certain deployment challenges by enabling deep learning models that are more robust to data distribution changes, without the need for GPUs or off-line training. We describe a continual learning approach to forecasting extreme air quality events developed for the National Oceanic and Atmospheric Administration to provide operational air quality guidance to the Continental United States. We describe how this deep learning model is resilient to future data distribution changes by performing curriculum learning, and how it can be deployed as a continual learner, offering better predictive performance for resource-constrained environments.

From dishwasher to river: how to adapt a low-cost turbidimeter for water quality monitoring.

This study presents the process of design and development of a low-cost turbidimeter for monitoring water quality, facilitating rigorous spatial-temporal variability analysis within large-scale hydrological systems. We propose a low-cost optical turbidimeter, modifying the existent SEN0189 turbidity sensor, Arduino boards, and additional sensors for temperature compensation. We compared a low-cost system with high-tech sensors, modifying the original low-cost SEN0189 probe for enhanced environmental performance. The three-step methodological framework involved prototype development, compensation for environmental factors, and preparation for future field deployment. Calibration equations with a high coefficient of determination and a temperature correction equation were established. We made adaptations to overcome field deployment challenges, including a 3-D printed sensor case, defining the relationship between measurement uncertainty and energy consumption, and specifying field installation guidelines. In summary, this study presents a comprehensive approach to a low-cost optical turbidity system, demonstrating its potential for accurate and affordable field deployment. We aim to address the critical need for sustainable inland water management tools, making this system a valuable contribution to environmental monitoring practices. We also aim to inspire similar development of open-source monitoring systems within our community.

Efficient Small Object Detection You Only Look Once: A Small Object Detection Algorithm for Aerial Images.

Aerial images have distinct characteristics, such as varying target scales, complex backgrounds, severe occlusion, small targets, and dense distribution. As a result, object detection in aerial images faces challenges like difficulty in extracting small target information and poor integration of spatial and semantic data. Moreover, existing object detection algorithms have a large number of parameters, posing a challenge for deployment on drones with limited hardware resources. We propose an efficient small-object YOLO detection model (ESOD-YOLO) based on YOLOv8n for Unmanned Aerial Vehicle (UAV) object detection. Firstly, we propose that the Reparameterized Multi-scale Inverted Blocks (RepNIBMS) module is implemented to replace the C2f module of the Yolov8n backbone extraction network to enhance the information extraction capability of small objects. Secondly, a cross-level multi-scale feature fusion structure, wave feature pyramid network (WFPN), is designed to enhance the model's capacity to integrate spatial and semantic information. Meanwhile, a small-object detection head is incorporated to augment the model's ability to identify small objects. Finally, a tri-focal loss function is proposed to address the issue of imbalanced samples in aerial images in a straightforward and effective manner. In the VisDrone2019 test set, when the input size is uniformly 640 × 640 pixels, the parameters of ESOD-YOLO are 4.46 M, and the average mean accuracy of detection reaches 29.3%, which is 3.6% higher than the baseline method YOLOv8n. Compared with other detection methods, it also achieves higher detection accuracy with lower parameters.

Optimal capacity allocation and scheduling strategy for CSP+PV hybrid standalone power plants

Solar photovoltaic (PV) power generation, as an important clean technology, has been widely adopted globally, especially in remote island areas where access to the main power grid is unavailable. PV power can serve as the primary energy source for standalone island grids. However, due to its dependence on sunlight, PV power output fluctuates, particularly during nighttime and under poor lighting conditions, necessitating the integration of energy storage technology or alternative power generation methods to ensure continuous power supply. Concentrating solar power (CSP) generation, as an emerging technology, can provide efficient power output when solar radiation is abundant and ensure continuous power supply through thermal energy storage systems during adverse weather conditions or nighttime. Although CSP offers advantages such as dispatchability, its high construction and maintenance costs may pose challenges in commercial deployment. Hybrid solar power plants combining both PV and CSP technologies leverage the strengths of both, ensuring more stable and economically viable power output. This study establishes a model for hybrid solar power plants, considering the impact of PV and CSP component capacities and proportions on performance and costs, i.e., capacity allocation. To maximize the overall benefits of standalone microgrids while ensuring the stability of the power station, a capacity allocation method guided by economic dispatch is proposed. Through iterative analysis, the optimal configuration is determined to minimize the system's equivalent annual costs. Simulation experiments validate the financial feasibility of hybrid solar power plants and the reliability of the proposed configuration method.

A growth posture identification method of immature peaches in natural environments

The precise bagging of immature peaches by the fruit bagging robot requires the identification of the target young fruits as well as the acquisition of their growth angles. The network models employed in single detection algorithms exhibit complexity and pose challenges for deployment on mobile terminals as they heavily rely on the availability of labeled samples. A semi-supervised learning and lightweight strategy based on YOLOv8n-seg was proposed to identify the growth posture of immature peaches in the work. Firstly, the self-training method and data enhancement in semi-supervised learning were used to generate efficient pseudo-label data. A tremendous labeling workload was solved. Secondly, the YOLOv8n-seg backbone network was replaced with an improved MobileNetv3 structure for better real-time detection on MT to reduce network parameters and calculations. The model was easily deployed with improved detection speed. Meanwhile, the original upsampling module was replaced with the CARAFE module to enhance the recognition capability of global features, which considered the impact of immature peaches on the nearby color background. The CIOU loss function was ultimately substituted with the SIOU loss function to further optimize the boundary frame loss and target detection accuracy. The enhanced model could predict the coordinate information of immature peaches and calculate growth angles. The experimental results show that the improved peach seedling growth posture network model has a weight size of only 23.1% of the original model. Additionally, the algorithm achieved a remarkable rate of 87.8% in identifying young peach fruits, with an average error of ±3.3° in estimating the growth angle. It took an average of 31 ms to detect a 3024 × 4032 pixels image in the edge computing device JETSON AGX ORIN CLB development kits. The method could rapidly identify immature peaches and estimate the growth angle, which ensured accurate bagging.

Analyzing SD-WAN Deployment Challenges in the Retail Sector

This article presents a comprehensive case study analysis of a significant SD-WAN deployment failure within a large retail chain, offering critical insights into the challenges and complexities associated with network modernization in the retail sector. Through an in-depth examination of the implementation process, we identify key factors contributing to the failure, including inadequate pre-deployment assessment, poor vendor selection, overlooked security requirements, and difficulties integrating with legacy systems. The article employs a rigorous methodology, combining semi-structured interviews, document analysis, and on-site observations to provide a nuanced understanding of the technical, organizational, and human factors at play. By analyzing the consequences of the failed deployment, including increased operational costs, customer dissatisfaction, and reputational damage, we derive valuable lessons and best practices for future SD-WAN implementations. The article concludes with a discussion of risk mitigation strategies, the importance of tailored solutions, and long-term considerations for SD-WAN success in retail environments. This article contributes to the growing body of knowledge on SD-WAN adoption, offering practical guidance for retailers and IT professionals navigating the complexities of network transformation in an increasingly digital retail landscape.

Navigating the ethical and governance challenges of ai deployment in AML practices within the financial industry

The rapid deployment of Artificial Intelligence (AI) in Anti-Money Laundering (AML) practices within the financial industry presents significant ethical and governance challenges that must be navigated effectively. As financial institutions increasingly adopt AI technologies to enhance their AML efforts, concerns regarding data privacy, algorithmic bias, and transparency emerge. This review explores the ethical implications of AI in AML and offers governance strategies to mitigate risks while ensuring compliance with regulatory frameworks. One of the primary ethical challenges in deploying AI for AML is the potential for algorithmic bias. AI systems trained on historical data may inadvertently perpetuate existing biases, leading to discriminatory practices in transaction monitoring and customer profiling. This raises serious concerns about fairness and equity in the financial sector. Addressing algorithmic bias requires the implementation of rigorous testing and validation processes to ensure AI systems function impartially across diverse populations. Data privacy is another critical issue. The extensive data collection required for effective AML monitoring raises questions about the protection of sensitive customer information. Financial institutions must establish robust data governance frameworks that prioritize privacy and comply with regulations such as the General Data Protection Regulation (GDPR). Ensuring transparency in how data is used and providing clear communication to customers about data practices is essential for building trust. Effective governance frameworks are crucial in navigating these ethical challenges. Financial institutions should adopt a multi-disciplinary approach that includes ethical guidelines, compliance measures, and risk management strategies. Establishing oversight committees can help ensure that AI deployment aligns with ethical standards and regulatory requirements. Furthermore, ongoing training for employees on the ethical use of AI in AML can foster a culture of responsibility and accountability. This review highlights the need for a balanced approach to AI deployment in AML, emphasizing the importance of ethical considerations and governance structures. As the financial industry continues to evolve, addressing these challenges will be essential for maintaining trust, ensuring compliance, and leveraging AI’s potential to enhance AML practices effectively.

RESPONSE OF LIQUID MANURES AS ALTERNATIVE NUTRIENT ON GROWTH AND YIELD OF TOMATO

This research explores the transformative role of advanced technologies, particularly AI, machine learning (ML) and block chain, in the development of smart e-commerce solutions that enhance user experiences and operational efficiencies. It also explores how customized algorithms can improve customer engagement, boost conversion rates and encourage sustainable shopping practices. Additionally, the research investigates AI-driven solutions to encourage sustainable shopping behaviors and optimize shopping post-purchase experiences. Specifically, it aims to assess the impact of AI-based personalization algorithms, examine the trade-offs between data privacy and user experience and analyze AI's influence on consumer trust and decision-making processes. Primary data from surveys and interviews are supplemented with secondary data on customer behavior and sales metrics to evaluate the effectiveness of technologies like AI, machine learning and block chain in driving personalized, ethical and efficient e-commerce. A mixed-method research design, combining qualitative and quantitative approaches, employed to capture insights from key stakeholders (e.g., consumers, developers and e-commerce managers) and assess the impact of smart e-commerce platforms on shopping behavior, trust and loyalty. Implementing the Agile Model, the research iteratively addresses system requirements, testing and deployment challenges to create a flexible, scalable solution. Key findings underscore the role of predictive modeling in inventory management, the ethical potential of sustainable product recommendations and the positive consumer responses to transparency initiatives. The study contributes valuable insights into the development of secure, personalized and ethical e-commerce platforms that align with modern consumer expectations for convenience and responsibility.

Enhancing IoT Security through an Artificial Neural Network Approach

This study aims to fortify Internet of Things (IoT) security through the strategic implementation of Artificial Neural Networks (ANNs). With the rapid expansion of IoT devices, traditional security measures have struggled to cope with the dynamic and complex nature of these environments. ANNs, known for their adaptability, are explored as a promising solution to enhance security. The central objective is to significantly improve the accuracy of IoT security measures by optimizing ANN architectures. Using a curated dataset with key environmental parameters, the study evaluates three ANN models—Backpropagation Neural Network (BPNN), Multilayer Perceptron (MLP), and Long Short-Term Memory (LSTM). The evaluation metrics include accuracy, precision, recall, and F1-score across different train-test splits. Results show that LSTM consistently outperforms BPNN and MLP, demonstrating superior accuracy and the ability to capture temporal dependencies within IoT security data. Implications stress the importance of aligning model selection with specific application goals, considering factors like computational efficiency. In conclusion, this research contributes valuable insights into the practical implementation of ANNs for IoT security, guiding future optimization efforts and addressing real-world deployment challenges to safeguard sensitive data and ensure system resilience in the evolving IoT landscape.

Forecasting short-term subway passenger flow using Wi-Fi data: comparative analysis of advanced time-series methods

Accurately monitoring passenger demand fluctuations is crucial for streamlined operations of subway systems and informed decision-making. This study presents a detailed Time Series Analysis of the Toronto subway system using Wi-Fi data connection from devices as a predictor of passenger volume. Various time series models were tested for short-term forecasting, including Linear Regression, Exponential Smoothing, ARIMA, Random Forest, N-BEATS, and T-GCN. An end-to-end modeling implementation process was carried out, and the performance of each model was evaluated. The primary objective was to assess the effectiveness of short-term prediction models for univariate time series at the system level and discuss deployment challenges. While conventional time series models are fast to implement and interpretable, they require a more in-depth data exploration phase for validation, making scaling at the system level difficult. Additionally, maintenance is more challenging with conventional models, and their exploratory analysis phases need to be repeated when the models degrade over time. Prediction difficulty varied across each subway station, indicating the need for a more thorough calibration or hybrid approach, especially for transfer stations. Despite the different uses and qualities of each model in our scenario, Random Forest and Exponential Smoothing emerged as the best performers and could be a satisfactory option for robust demand forecasting at the system level.

Task-Level Customized Pruning for Image Classification on Edge Devices

Convolutional neural networks (CNNs) are widely utilized in image classification. Nevertheless, CNNs typically require substantial computational resources, posing challenges for deployment on resource-constrained edge devices and limiting the spread of AI-driven applications. While various pruning approaches have been proposed to mitigate this issue, they often overlook a critical fact that edge devices are typically tasked with handling only a subset of classes rather than the entire set. Moreover, the specific combinations of subcategories that each device must discern vary, highlighting the need for fine-grained task-specific adjustments. Unfortunately, these oversights result in pruned models that still contain unnecessary category redundancies, thereby impeding the potential for further model optimization and lightweight design. To bridge this gap, we propose a task-level customized pruning (TLCP) method via utilizing task-level information, i.e., class combination information relevant to edge devices. Specifically, TLCP first introduces channel control gates to assess the importance of each convolutional channel for individual classes. These class-level control gates are then aggregated through linear combinations, resulting in a pruned model customized to the specific tasks of edge devices. Experiments on various customized tasks demonstrate that TLCP can significantly reduce the number of parameters, by up to 33.9% on CIFAR-10 and 14.0% on CIFAR-100, compared to other baseline methods, while maintaining almost the same inference accuracy.