مطالب

Through advanced analytical models such as machine learning (ML) and, conceptually, Large Language Models (LLMs), this study explores how Industrial Internet of Things (IIoT) applications can transform educational experiences in the context of smart steel production. To mitigate the shortage of authentic industrial datasets for research, we developed an industry-validated IIoT educational dataset drawn from three months of operational records at a steel plant and enriched with domain-specific annotations—most notably distinct operational phases. Building on this foundation, we propose an IIoT framework for intelligent steel manufacturing that merges ML-driven predictive analytics (employing Lasso regression to optimize energy use) with LLM-based contextualization of data streams within IIoT environments. At its core, this architecture delivers real-time process monitoring alongside adaptive learning modules, effectively simulating the dynamics of a smart factory. By promoting human–machine collaboration and mirroring quality-control workflows, the framework bridges the divide between theoretical instruction and hands-on industrial practice. A key feature is an interactive decision-support dashboard: this interface presents ML model outcomes and elucidates IIoT measurements—such as metallization levels and H 2 /CO ratios—through dynamic visualizations and scenario-based simulations that invite risk-free exploration of energy-optimization strategies. Such tools empower learners to grasp the intricate multivariate dependencies that govern steel manufacturing processes. Our implementation of the Lasso regression model resulted in a 9% reduction in energy consumption and stabilization of metallization levels. Overall, these findings underscore how embedding advanced analytics within IIoT education can cultivate a more engaging, practice-oriented learning environment that aligns closely with real-world industrial operations.

Accurate and early prediction of diabetes is crucial for initiating prompt treatment and minimizing the risk of long-term health issues. This study introduces a comprehensive machine learning model aimed at improving diabetes prediction by leveraging two clinical datasets: the PIMA Indians Diabetes Dataset and the Early-Stage Diabetes Dataset. The pipeline tackles common challenges in medical data, such as missing values, class imbalance, and feature relevance, through a series of advanced preprocessing steps, including class-specific imputation, engineered feature construction, and SMOTETomek resampling. To identify the most informative predictors, a hybrid feature selection strategy is employed, integrating recursive elimination, Random Forest-based importance, and gradient boosting. Model training uses Random Forest and Gradient Boosting classifiers, which are fine-tuned and combined through weighted ensemble averaging to boost predictive performance. The resulting model achieves 93.33% accuracy on the PIMA dataset and 98.44% accuracy on the Early-Stage dataset, outperforming previously reported approaches. To enhance transparency and clinical applicability, both local (LIME) and global (SHAP) explainability methods are applied, highlighting clinically relevant features. Furthermore, probability calibration is performed to ensure that predicted risk scores align with true outcome frequencies, increasing trust in the model’s use for clinical decision support. Overall, the proposed model offers a robust, interpretable, and clinically reliable solution for early-stage diabetes prediction.

Background: The cloud-native architectures have reinvented the original strategies of the companies’ IT infrastructure approach and became popular due to the concepts of modularity, scalability, and resilience. These architectures respond to the shortcomings of the monolithic architectures to meet the new business challenges and workloads, including embracing innovation technologies like Artificial Intelligence and big data processing solutions. Objective: This study was designed with the objective of assessing the performance and business viability of cloud-native systems, based on critical indicators such as availability, resilience to failure, resource use, and compatibility with innovative technologies. The objective was to define the barriers and possibilities for improving cloud native architectures in various enterprises. Methods: A cross-sectional research, consideration, experiment test and case study and performance analysis. Response time, CPU and memory consumption and recovery time were compared across the range of throughput from 1000 to 12000 requests per second. To enhance the interpretational framework, key usage scenarios in the three sectors of healthcare, retail and finance were collected and compared with the results. Results: Cloud-native systems proved to provide high availability rates (> 99.9%), resource scalability, and component resource efficiency. With the use of AI in combination with big data analytics, improvement in performance was realized. But some of the problems that were seen include vendor lock, integration issues, and fluctuating peak load issues. Conclusion: All identified improvements signify the potential of cloud-native architectures for improving enterprise IT functioning. It is thus possible to continue perfecting the identified challenges to enhance their effectiveness, optimal for the current dynamic digital environment.

Background: The growth of the number of connected devices and the extent of Internet of Things (IoT) integration has led to new and emerging needs such as the management of big data, real-time reaction, efficient bandwidth utilization, and security considerations. Due to the intrinsic latency, network load and argue of scalability, standard cloud computing models do not suffice these requirements. In response to this, edge computing the function of analyzing data closer to its source hence leading to performance gains. Objective: This article explores the impact of incorporating edge computing in the optimization of IoT systems specifically in aspects like latency minimization, bandwidth utilization, security, processing capability, flexibility in expansion, and data reliability. Methods: A combined computational model was used to mimic edge and cloud platforms. Performance metrics were evaluated under three primary IoT scenarios: traffic management of smart cities, industrial applications, and health care management applications. Regression models and confidence intervals also provided general support to the findings. Results: The findings showed edge computing to be a more effective substitute for cloud-based systems; proving that latency can be reduced by 82%, and data bandwidth by 65-68%. Perennial threats including interception of data were cut by 50-66% while processing was done at 73% higher efficiency. Other criteria such as scalability and data consistency also pointed out the application of edge computing for resilience in more extensive IoT environment. Conclusion: Essentially, edge computing helps overcome limitations of cloud-based IoT systems, and is therefore imperative to real-time, secure, and scalable IoT. Future work should consider the integration of hybrid edge-cloud models, self-healing schemes, and more robust rigorous security solutions in order to fine-tune its applicability.

Background: Failures of communication usually occur during natural disasters, therefore signaling the importance of flexible networks. Unmanned Aerial Vehicles (UAVs) are anticipated to solve this problem by acting as on-the-move networks in the disaster-stricken regions. However, barriers that include challenges in UAV control coordination, resources allocation as well as security of the data being drawn are still pushing the technology backward. Objective: The article seeks to design and analyze enhanced heuristics for employing UAVs in disaster communications to enhance performance, availability, and security. Methods: Both primary, semi-structured interview and survey and post-disaster reports as well as secondary, computational analysis based on MATLAB and NS - 3 simulations were used as the data collection technique. Five algorithms: Multi-UAV Coordination, Dynamic Resource Allocation with Security, Hybrid Communication Framework, AI-Driven Path Optimization, Privacy-Preserving Data Sharing were implemented and incorporated. Theoretical models built on the basis of multi- objective optimization and the theory of games confirmed work ability to scale. Results: The introduced algorithms increased coverage by 75%, decreased latency by 27 percent, and also introduced 30 percent energy efficiency. Average privacy compliance levels floated above 90%, and an advanced resource allocation model achieved equal distribution. All these enhancements were affirmed in urban, rural and mountainous regions further proving versatility and stability. Conclusion: The article proposes a framework for UAV-enabled disaster communication system that can incorporate advanced algorithm and theoretical models to overcome the coordination challenge while ensuring the efficiency and security of the system. The presented results can be considered to be the reliable base for the UAVs usage in disaster situations.

Background: Telecommunication infrastructure requires regular maintenance and upkeep for its networks’ matrices, but existing approaches have been associated with issues such as time consumption and concern costs, as well as safety hazards. Newer developments in drone technology present progressive opportunity through the improvement of current maintenance processes by means of automation, predictability, and real time computation. Objective: The article seeks to assess whether the use of drone in telecommunication maintenance enhances the operational productivity through increasing the efficiency, reducing cost, safety, environmental and scalability and in different terrains. Methods: The methods followed included the conduct of experimental surveys with drone operations in five different telecommunication settings. These areas of interest were inspection efficiency, the accuracy of condition-based maintenance, signal received signal power, delay reduction through edge computing, and energy consumption. Sophisticated numerical computations, like Kalman filters and various frameworks of edge computing, were used in this context to draw analytical insights on the collected data. Results: The methods that used drones lowered the time needed for inspections by ¾ and cut the expenses by 49.3% and increased safety and quality of the coverage. Predictive maintenance was found to have achieved 89.7% accuracy with the system response time being 246ms at different site. The results of energy consumption model depicted the errors under 2% confirming this approach’s suitability for operational planning. Conclusion: By evaluating the applicability of drones in telecoms maintenance, the paper shows that the notion of drones in this context is promising both now and in the future. These results signal existing and potential applications of drones is to incorporate drone technology into infrastructural management solutions to address emerging needs in the industry.