Jul 2026· International Conference on Ubiquitous and Future Networks· pp. 621-623· 0 citations· 15 references
Abstract
The analysis of photovoltaic (PV) degradation data presents significant challenges due to the complexity and volume of heterogeneous information generated by modern monitoring systems. In contemporary deployments, PV installations operate as distributed, networked cyber-physical systems, where multiple sensing devices and monitoring nodes continuously generate multi-modal data streams. This paper proposes a natural language-driven approach based on Large Language Model (LLM) agents to enhance the accessibility of analytics in such networked environments. We design an agent-based architecture that translates natural language intents into executable analytical workflows, enabling intuitive interaction with data generated by IoT-enabled PV monitoring infrastructures. The system integrates LLM reasoning with structured data processing, visualization tools, and domain-specific context. The proposed approach is implemented using Python and LangChain, and evaluated on the PV EL HDR BW Test DB dataset. Experimental results demonstrate an overall success rate of 83.3% across representative analytical tasks, highlighting the practical viability of LLM-based agents for domain-specific data analysis. This work contributes to AI-driven analytics for distributed and networked systems, showing how natural language interfaces can support data-driven decision-making in emerging intelligent energy and IoT infrastructures.
The rapid expansion of the Internet of Things (IoT) has resulted in massive volumes of data being generated by interconnected devices across various domains. Traditional cloud-centric architectures often struggle with issues such as high latency, bandwidth constraints, and data privacy risks when processing this data. Edge computing has emerged as an effective solution by enabling data processing closer to the data source, thereby improving response time and reducing network dependency. In recent years, Generative Artificial Intelligence (GenAI) has gained significant attention for its ability to generate insights, predictions, and adaptive responses from complex and dynamic datasets. This paper examines the integration of Generative AI with IoT and edge computing to enhance intelligent edge systems capable of real-time analytics and autonomous decision-making. It explores architectural frameworks, potential applications in areas such as smart cities, healthcare, industrial automation, and autonomous systems, as well as the advantages of improved efficiency, scalability, and privacy preservation. Additionally, the paper discusses the technical challenges associated with deploying generative models at the edge, including resource constraints, model optimization, security, and data management. Finally, it outlines future research directions aimed at developing scalable, secure, and energy-efficient GenAI-enabled edge computing ecosystems.
Mr. L. S. Shendge, Mrs. S. N. Patel, Ms. D. V. Sharma· International Journal of Lat...· 0 citations
: Smart city systems increasingly depend on data analytics and visualization to support informed and timely decision making in complex urban environments. However, existing middleware solutions predominantly focus on data acquisition and communication, while analytical processing and visualization are typically delegated to external applications, resulting in increased development complexity, reduced reusability, and fragmented system architectures. This study presents analytics and visualization-centric middleware named “Service-Oriented Middleware for Smart City Applications” (SOMSCA), in which these capabilities are embedded directly within the middleware layer. SOMSCA adopts a service-oriented approach, exposing analytics and visualization functionalities as reusable platform services, and incorporates a data-type-oriented visualization strategy along with a template-assisted dash-boarding mechanism to enable dynamic and flexible application development. To validate the proposed approach, a prototype implementation is developed using React, FastAPI, MySQL, and TimescaleDB and evaluated using air-quality data collected from the London Air Quality Network, comprising more than one million observations across multiple pollutants and monitoring locations. The implementation supports real-time, historical, and aggregated analytical services together with dynamic dashboard generation. The results demonstrate the practical feasibility of middleware-integrated analytics and visualization through reusable service creation, flexible dashboard configuration, and interactive environmental monitoring capabilities. These findings highlight the potential of middleware-level intelligence to simplify application development and support data-driven decision making in smart city environment.
Zulfiqar Ali, A. Mahmood, S. Khatoon et al.· Computers, Materials & C...· 0 citations
The Industrial Internet of Things (IIoT) has ushered in a new era of connected devices, real-time data collection, and intelligent analytics. However, the complexity and heterogeneity of IIoT environments demand sophisticated data modeling techniques to harness the full potential of analytics. Hybrid data models—integrating structured, semi-structured, and unstructured data representations—are emerging as a critical solution. This paper presents a comprehensive study on the design and application of hybrid data models tailored for IIoT analytics. It explores data heterogeneity challenges, presents layered data model architectures, and discusses the integration of relational, NoSQL, and time-series databases. A proposed hybrid data model architecture is introduced, with a focus on real-time processing, scalability, and semantic interoperability. The model is validated using an industrial case study involving predictive maintenance in a smart factory. Results indicate a significant improvement in data query performance, storage efficiency, and analytical accuracy compared to traditional single-model approaches. Flowcharts and diagrams illustrate data flow, architectural layers, and analytics pipelines. The paper concludes by emphasizing the role of hybrid data models as enablers for scalable, resilient, and intelligent IIoT ecosystems.
Howard Aiken, J. Eckert, John W. Mauchly· International Journal of Dat...· 0 citations
The rapid growth of heterogeneous data sources—such as IoT devices, social media, enterprise systems, and cloud applications—has led to massive increases in data volume, velocity, and variety. Traditional rule-based and fixed data engineering pipelines are no longer sufficient to handle these complexities. This paper explores AI-driven data engineering frameworks designed to build scalable, adaptive, and intelligent data pipelines. By integrating AI techniques like machine learning, deep learning, and reinforcement learning, these frameworks enable automated data ingestion, intelligent transformation, anomaly detection, and predictive pipeline optimization. Unlike traditional batch-processing systems, modern architectures support hybrid and real-time streaming, improving efficiency and flexibility. The proposed approach introduces a layered architecture where each stage—ingestion, processing, storage, orchestration, and analytics—is enhanced with AI capabilities. Results show significant improvements, including up to 45% reduction in data errors and 60% increase in pipeline efficiency. Overall, AI-based data engineering represents a major advancement, paving the way for more intelligent, self-optimizing systems, with future directions including explainable AI, federated learning, and edge computing.
N. Wirth· International Journal of Dat...· 0 citations
As the Internet of Things (IoT) proliferates, vast volumes of streaming sensor data are being generated in real-time, creating both opportunities and challenges for data-driven decision-making. This paper proposes a cloud-native architecture that integrates Snowflake’s scalable data platform with AWS Lambda's serverless compute capabilities to build real-time predictive analytics pipelines for IoT data. By leveraging AWS services for ingestion (such as Kinesis or MQTT over IoT Core), Lambda for event-driven processing, and Snowflake for scalable storage and analysis, the proposed solution enables rapid deployment of machine learning models to process streaming data. The architecture is designed to be low-latency, cost-effective, and easily extensible for a variety of industrial applications. A prototype implementation and performance evaluation are presented, demonstrating the effectiveness of the architecture in handling high-throughput, low-latency IoT workloads.
Ritu Agarwal· International Journal of Dat...· 0 citations
This paper presents CRAFTER, an automated framework for designing and deploying self-adaptive IoT systems using Causal Reinforcement Learning (CRL). As IoT devices increasingly populate pervasive computing spaces, smart environments are enabled with advanced monitoring and interactive services. The dynamic nature of these environments, such as fluctuating workloads and evolving application demands, poses significant challenges in maintaining consistent Quality of Service (QoS) levels of IoT applications. While existing self-adaptation techniques offer adaptive capabilities, they are often designed to deal with specific application domains, hindering the design of self-adaptive solutions that can be re-used across multiple IoT verticals. In addition, there is a lack of automated pipelines that act on identifying key performance drivers to take effective adaptation decisions. CRAFTER addresses these issues by using Causality as a formal framework for performance analysis of IoT systems. CRAFTER generates causal graphs to uncover dependencies among system components and guide adaptation decisions based on cause–effect relationships. Then, adaptation agents can leverage this knowledge to take more effective adaptation decisions in dynamic situations. Our experimental evaluation demonstrates how CRAFTER enables deriving causal graphs spanning diverse IoT use cases. Furthermore, we showcase how CRAFTER improves self-adaptation performance by 25% compared to state-of-the-art Reinforcement Learning-based approaches.
Houssam Hajj Hassan, A. Kattepur, Denis Conan et al.· SEAMS@ICSE· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
